Docket No. 354. PATENT

RECEPTOR AND KINASE MODULATOR SCREENING METHODS

FIELD OF THE INVENTION

[1] The invention relates to methods to identify modulators of mitogen activated protein kinases of low toxicity. The invention further relates to methods to identify modulators of phosphatidylinositide kinases and cell-surface receptors upstream of the protein and phosphatidylinositide kinases. The modulators identified also include dual modulators of extracellular signal-regulated and phosphatidylinositide kinases. Such modulators are useful to treat diseases related to hyperproliferation, including androgen- associated cancers such as prostate cancer, breast cancer, ovarian cancer, lung cancer, liver cancer, bladder cancer, lymphoma, melanoma, thyroid cancer and colon cancer, or unwanted inflammation, which supports initiation or progression of such cancers.

BACKGROUND

[2] The mitogen-activated protein kinases (MAPK) include the extracellular signal- regulated kinase (ERK) isoforms Erk-1 (also referred to as MAPK-3 or p44 kinase) and Erk-2 (also referred to as MAPK-1 or p42 kinase), c-Jun amino-terminal kinase (JNK) and p38 isoforms. Although the MAPKs respond to a wide array of stimuli and are involved in a wide range of functions, including phosphorylation of phospho-lipids, transcription factors, cytoskeletal protein and other protein kinase termed MAPK-activated protein kinases (MKs or MAPKAPKs), MAPKs contain similar structural binding domains. Those domains include the ATP binding site, a catalytic active site that transfers a phospho- group from bound ATP to a specific serine or threonine of a MAPK substrate, and protein- protein interaction domains, including the CD (common docking) and FD domains, for recognition of substrates and protein binding partners.

[3] The MAPKs are regulated in part through phosphorylation cascades. Activation of MAPK requires the phosphorylation of conserved tyrosine and serine or threonine in the subdomain VIII activation loop by an upstream protein kinase referred to as a MAPKK or MEK. Various isoforms of this upstream kinase exhibits differing levels of selectivity for their MAPK substrates. The MAPKK in turn are regulated by phosphorylation by upstream kinases referred to as MAPKKK, MEKK or MAP3K, which in turn can be activated through interaction with a protein that becomes activated as a consequence of ligand interaction with its cognate membrane bound receptor. Those membrane receptors typically include receptor tyrosine kinases (RTKs). However, it is now appreciated that ligand interactions with G protein-coupled receptors (GPCRs) may also result in MAPK activation though cross-talk with G protein-dependent and/or G protein-independent Docket No. 354. PATENT signaling.

[4] The phosphorylation cascade in MAPK signaling involve multiple protein kinases that amplifies the initial signal entering into the cascade and appears to be a common feature to the diversity of MAPKs signaling effects. That common cascade allows for multiple unique points of regulation and integration of signaling events originating at the cell membrane or within the cytoplasm that flow through each MAPK signaling node to their diverse array of downstream effectors. That signaling cross-talk sometimes becomes aberrant thereby resulting in excessive signaling through one or more of those nodes, which is often responsible for initiating or propagating neoplasms.

[5] Another level of regulation of MAPK signaling is provided by scaffold proteins that pre-assemble some of the components of the protein kinase cascade into sub-cellular compartments in order that the incoming signal into the cascade is properly directed to the appropriate downstream effector proteins or integrated with signaling from other signal transduction pathways. The scaffolding protein may in turn be regulated by proteins that affect its phosphorylation state. Additional regulation is provided by phosphatases, which are also regulated by their own phosphorylation states and interactions with scaffolding proteins. Therefore, a tightly regulated network of proteins is required to properly respond to the signaling input and output through each MAPK component so that signaling coming into this network results in the appropriate outcome.

[6] Erk-1 and Erk-2, which seem to have the same substrate selectivity in vitro and are often lumped together and described as Erk-1/2, are reported to have over 160 substrates [Yoon, S. and Seger, R. (2006)]. By contrast, in cells Erk-1 and Erk-2 show different substrate selectivities because these isoforms are presumed not to be identically localized due to their different preferences for binding partners. Thus, the two Erk isoforms are not influenced identically by the same signaling event. For example, the antiproliferative effect of cytosolic sequestered Erk-1 may be due in part to its improved competition over Erk-2 for the upstream activator MEK as a result of this sequesterization [Vantaggiato, C. et al. (2006)]. That competition between the two Erk isoforms is thought due to differential regulation of Erk-1 , Erk-2 and MEK isoforms by scaffolding proteins.

[7] Furthermore, Erk-1 and Erk-2 differ in their nucleo-cytoplasmic transport properties with Erk-1 shuttling more slowly through the nuclear membrane than Erk-2 [Marchi, M. et al. (2008)]. Additionally, the dimerization state of the activated Erk isoform will also influence the isoform's relative contribution between phosphorylating cytoplasmic effector protein and nuclear transcription factors. For example, pErk-2 may access the nucleus either as a monomer (passive transport) or a homodimer (facilitative transport), whereas Docket No. 354. PATENT its phosphorylation of cytoplasmic proteins seems to require the dimerized state, the formation of which is enhanced by scaffolding proteins [Casar, B. et al. (2009)].

[8] Due to the similarity in MAPKs structures, selective inhibitors of these kinases are lacking. Inhibitors of MAPKs that have been studied typically target the ATP-binding site (i.e., ATP binding site-dependent inhibitors). Compounds that exert their activity through MAPK signaling through the ATP binding site are generally considered to be toxic, particularly to the liver in view of human toxicity that is observed for p38 MAPK inhibitors [Morel, C. et al. (2005); Laufer, S.A. et al. (2006); Kumar, S. et al. (2003)]. Substances that interact at binding domains other than the ATP binding site are sometimes referred to as non-ATP site-dependent inhibitors.

[9] Several non-peptidic MAPK inhibitors have been described that interact with its CD domain [Boston, S.R. et al. (201 1 )], but there are no known low molecular weight, non- peptidic MAPK inhibitors that target the Erk FD domain, nor has any discrimination between the Erk isoforms been shown with any non-ATP binding site-dependent inhibitor. Additionally, to date there are no known studies in humans that use ATP binding site- dependent or ATP binding site-independent inhibitors or other modulators of Erk-1/2 activity where there is direct physical interaction of the modulator with this kinase (cf. inhibition of Erk activity by ATP binding site-dependent inhibitors of an upstream protein kinase, e.g., MEK). Studies with indirect modulators of Erk-1 /2 activity through inhibition of upstream kinases has resulted in skin, cardiac and gastrointestinal toxicity, presumably due to their adverse effects on cellular functions in normal tissue that are dependent upon Erk signaling and undesired cross-reactivities with other protein kinases.

[10] Other kinases having an important role in some hyperproliferation conditions are the phosphatidylinositide kinases. Those kinases include the Class IA phosphatidylinositide-3-kinases (PI3Ks), which are heterodimers composed of a p1 10 catalytic subunit and a p85 regulatory subunit. The Class IA PI3Ks are responsible for production of phosphatidylinositol (3,4,5)-trisphosphate (PI(3,4,5)P3), which stimulates the serine kinase Akt. Hyperproliferation conditions associated with excessive PI3K signaling are often accompanied by mutations in phosphatase and tensin homolog (PTEN), which is a phosphatase that negatively modulates intracellular levels of (PI(3,4,5)P3) to limit AKT activation. For example, it is estimated that 70% of prostate cancers at diagnosis have lost a copy of the PTEN gene [Chen, Z. et al. (2005)]. Other cancers associated with PTEN mutations include glioblastoma and breast cancer [Li, J. et al. (1997)]. Notably, all of those cancers are AR ^+/+ (and thus are androgen-associated cancers) and harbor a membrane androgen receptor (vide infra). Docket No. 354. PATENT

[11] Due to the role of excessive PI3K-AKT signaling in cancer cell proliferation, PTEN is considered a tumor suppressor gene. Therefore, compounds that negatively modulate PI3K-Akt signaling would be useful in treating cancers, particularly those cancer involving PTEN mutations. To date, PI3K inhibitors under clinical investigation for cancer treatment are all ATP-site dependent and like those inhibitors the MAP kinases are thus plagued by undesirable side effects due to the importance of PI3K-Akt signal signaling in normal cells.

[12] Therefore, non-ATP site-dependent MAPK and/or PI3K modulators that suppress (i.e., negatively modulate) excessive pro-proliferative signaling in cancer cells without adversely effecting Ras-Erk and/or PI3K-Akt signaling in normal cells would have improved safety profiles relative to approved standard of care therapies for treating androgen-associated cancers.

SUMMARY OF THE INVENTION

[13] The present invention relates to screening methods to identify Erk MAPK or Erk MAPK and PI3K modulators having anti-proliferative properties. Moreover, the invention is directed to identifying modulators of those kinase activities that are less toxic to humans than inhibitors presently known and to identifying compounds that oppose or negatively modulate adverse biological activities of DHT and its metabolite, 5a-androstane-3a, 17β- diol (3a-diol).

[14] Therefore, the present invention provides for a method to identify a candidate compound, comprising (a) contacting a test compound with a suitable test system; (b) determining phosphorylation states of Erk-1 and Erk-2 resulting from step (a); and (c) selecting a test compound that positively modulated phosphorylation state of Erk-1 activation loop relative to Erk-2 or negatively modulated the phosphorylation state of Erk-2 activation loop relative to Erk-1 of step (b), wherein the test compound selected from step (c) is identified as a candidate compound.

[15] Additionally, the present invention provides screening methods to identify a candidate compound, the method further comprising in step (b) of determining phosphorylation state of a Class 1 PI3K proteins and further comprising in step (c) of selecting a test compound that negatively modulated the tyrosine phosphorylation state of p85a or ρ85β regulatory subunit of the Class 1 PI3K of step (b) (i.e., in addition to positively modulating the phosphorylation state of Erk-1 activation loop relative to Erk-2 or negatively modulating the phosphorylation state of Erk-2 activation loop relative to Erk-1 ).

[16] The present invention also provides screening methods for identifying those candidate compounds having anti-proliferative properties the method comprising steps Docket No. 354. PATENT

(a)-(c) and further comprising the steps of (e) contacting a selected test compound from step (c) with prostate or breast cancer cells of a suitable in vivo test system; (f) determining cancer cell proliferation in the suitable in vivo test system resulting from step (e); and (g) selecting a test compound from step (f) that inhibits cancer cell proliferation statistically significant to cancer cells contacted with vehicle alone.

[17] The present invention also relates to screening methods for identifying those candidate compounds having lower toxicities in comparison to known ATP-site dependent inhibitors of PI3K and/or Erk the method comprising steps (a)-(c) and (e)-(g) and further comprising the step of (h) determining a minimum effective amount of a compound selected from step (g) for treating breast or prostate cancer in a mammal; (i) administering the minimum effective amount to healthy mammals so as to provide treated mammals; (j) determining liver enzymes levels for alanine transaminase, aspartate transaminase, alkaline phosphatase and γ glutamyl transpeptidase of the treated mammals; and (k) selecting a compound from step (i) that did not increase liver enzymes levels for any one of the liver enzymes selected from the group consisting of alanine transaminase, aspartate transaminase, alkaline phosphatase and γ-glutamyl transpeptidase more than about 2-fold compared to normal values in about 70% or more of the treated mammals, wherein the compound of step (k) is identified as a candidate compound wherein the candidate compound is a candidate low toxicity Erk-1 modulator.

BRIEF DESCRIPTION OF THE DRAWINGS

[18] FIG. 1 : E-3a-diol (HE3235) Activity on Prostate Cancer Cells in Suitable In Vitro Test Systems.

[19] FIG. 2: E-3a-diol Inhibition of Cell Cycle and Induction of Apoptosis of LP LNCaP Cancer Cells in a Suitable In Vitro Test System.

[20] FIG. 3: E-3a-diol Activation of Mutant Androgen Receptor in LP LNCaP cells of a Suitable In Vitro Test System.

[21] FIG. 4: Dose Effect of E-3a-diol on Prostate Cancer Tumor Incidence and Tumor Volume in a Suitable In Vivo Test System.

[22] FIG. 5: Time Course Effect of E-3a-diol on Established Prostate Cancer Tumors in a Suitable In Vivo Test System.

[23] FIG. 6: HP LNCaP TransSignal™ SH2-protein Effected by E-3a-diol (T= 5 min).

[24] FIG. 7: HP LNCaP TransSignal™ SH2-proteins Effected by E-3a-diol (T= 15 min).

[25] FIG. 8: Effects of E-3a-diol on Phospho-p85a PI3K Levels Induced by DHT

Stimulation of HP LNCaP cells in a Suitable In Vitro Test System.

[26] FIG. 9: Effects of PI3K Inhibitor (LY295002) and Ras-Erk Signaling Inhibitor Docket No. 354. PATENT

(PD098059) on Prostate Cancer Cells in a Suitable In vitro Test System When Contacted with E-3a-diol.

[27] FIG. 10: Effect of on Erk Phosphorylation of LNCaP Cells in a Suitable In Vitro Test System When Contacted by E-3a-diol (HE3235) and Co-contacted with Kinase Inhibitor.

[28] FIG. 11 : Protein Phosphorylation in HP LNCaP Cells Determined by Kinex™ Antibody Microarray from Contact with E-3a-diol.

[29] FIG. 12: Tumor Volume in Response to Contacting E-3a-diol to Breast Cancer Cells of a Suitable In Vivo Test System.

[30] FIG. 13: Tumor Incidence Resulting From Contacting E-3a-diol to Breast Cancer Cells of a Suitable In Vivo Test System (average number of tumors per animal).

[31] FIG. 14: Tumor Incidence Resulting From Contacting E-3a-diol to Breast Cancer Cells of a Suitable In Vivo Test System (percentage of rats in each group without palpable tumors).

[32] FIG. 15: % Tumor Cells Positive for ERa and PARP after Contacting E-3a-diol to Breast Cancer Cells of a Suitable In Vivo Test System.

[33] FIG. 16: Expression of Proapoptosis and Differentiation Genes in Breast Cancer

Cells of a Suitable In Vivo Test System after Contact with E-3a-diol.

[34] FIG. 17: i-AR Phosphorylation Induced by E-3odiol, DHT or in combination after their Contact to Prostate Cancer Cells of a Suitable In Vivo Test System.

[35] FIG. 18: Mannose-6-phosphate effect on Proliferation of Prostate Cancer Cells of a Suitable In Vitro Test System When Proliferation is Induced by DHT.

[36] FIG. 19: E-3a-diol Anti-Proliferative Effect on Prostate Cancer Cells of a Suitable In Vitro Test System Stimulated by High Dose DHT and Effects of Mannose-6-Phosphate Thereon.

DETAILED DESCRIPTION OF THE INVENTION

[37] 1. Definitions

[38] "Phosphorylation status" or "phosphorylation state" as used herein interchangeably refers to the number or pattern of phosphate groups covalently bound to a phospho- protein, such as a phosphorylated protein kinase, which may be membrane bound or in a protein complex. In some embodiments phosphorylation status refers to the overall extent of phosphorylation of a collection of proteins for a specified protein kinase or to the extent to which specified amino acid residue(s) of a specified protein kinase in collection of such proteins that are capable of being phosphorylated in a suitable test system are actually Docket No. 354. PATENT phosphorylated.

[39] For example, the phosphorylation status of an Erk isoform may refer to the extent of phosphorylation of the activation loop of that isoform or the extent to which the serine or tyrosine residues in the activation loop are phosphorylated before or after contacting a test compound to a suitable test system. An Erk isoform phosphorylated in the activation loop is represented by pErk-1 or pErk-2. Typically, pErk-1 or pErk-2 refers to the isoform that is di-phosphorylated in the activation loop (sometimes identified as ppErk-1 , ppErk-2 or actErk), but also may indicate mono-phosphorylated protein (e.g., pTyr-Erk-1 or pThr-Erk- 1 ) as specified explicitly or implicitly by context.

[40] In another example, the phosphorylation state of intracellular AR (i-AR) may refer to the extent of tyrosine phosphorylation or to the extent or pattern of serine phosphorylation of the nuclear hormone receptor (NHR). Sometimes the phosphorylation state of i-AR refers to specific tyrosine or serine residues as specified explicitly or implicitly by context.

[41] Relative phosphorylation status of a protein or amino acid residues of a protein is stated explicitly by describing the comparator protein or the amino acid residues or by describing the suitable test and control systems, otherwise relative phosphorylation status is implicitly understood by context.

[42] "i-AR" as used herein means any intracellular androgen receptor capable of binding androgen, whether or not that binding events results in its translocation to the nucleus or gene transactivation. Therefore i-AR may refer to functional wild-type AR or a mutant form of the receptor (mt-AR), which may or may not be functional. For suitable test systems, the cells comprising such systems contain either functional wild-type AR or functional mutant AR unless specified otherwise.

[43] "Modulation" of an activity or physical state of a protein as used herein means increasing or decreasing an activity of that protein or a property of the protein's physical state resulting from contacting a test or candidate compound to a suitable test system. The modulation may be relative to another activity or property of a different protein, to the same protein in the basal state or subsequent to external stimulation, including contacting a mitogen to the test system prior to contacting of the test compound, or relative to the change in activity or property from contacting the test system with vehicle or reference compound.

[44] With respect to a defined protein kinase, signal transduction cascade, signal transduction node or signaling pathway, modulation of an activity includes, for example, Docket No. 354. PATENT increasing or decreasing the capacity of a kinase in a suitable test system to phosphorylate one or more of its downstream effector proteins or substrates of that protein kinase or to increase or decrease signaling through that signal transduction node, cascade or pathway upon contacting a test or candidate compound with any suitable test system relative to one or more other kinases or signal transduction nodes, cascades or pathways within the same test system, e.g., including increasing the phosphorylation capacity of Erk-1 while unaffecting or decreasing the phosphorylation capacity of Erk-2.

[45] With respect to Erk-112 activity, modulation of an activity includes increasing or decreasing the potential of one Erk isoform to phosphorylate a shared downstream effector protein or substrate relative to the other Erk isoform in a suitable test system. Erk-1/2 activity modulation also includes increasing or decreasing the relative amounts of the two isoforms that are capable of phosphorylation of their downstream effector proteins, molecules or substrates that existed prior to application or administration of a test compound to a suitable test system.

[46] With respect to Erk-1 /2 property modulation, application of a test compound to a suitable test system may modulate the phosphorylation states of Erk-1 or Erk-2 by, for example, increasing phosphorylation in the activation loop of Erk-1 in comparison to that of Erk-2 or increasing the amount of one phosphorylated isoform in comparison to the other isoform by, for example, increasing the level of phosphorylated Erk-1 (pErk-1 , ppErk-1 or total activation loop phosphorylated Erk-1 protein) in comparison to that of phosphorylated Erk-2 (pErk-2, ppErk-2 or total activation loop phosphorylated Erk-2 protein). Thus, a test compound that increases the amount of pErk-1 and/or ppErk-1 without providing an observable ppErk-2 increase or an observable increase in phosphorylated Erk-2 from basal level(s) when both isoforms are present in the same suitable test system, is a negative modulation of phosphorylation status of Erk-2, since no phosphorylation of this isoform took place in the presence of test compound, but would have taken place in the absence of test compound as evidenced by positive modulation of phosphorylation states of both isoforms. Furthermore, in this instance (i.e., increase in pErk-1 or ppErk-1 level compared to ppErk-2), the phosphorylation status or state of Erk-1 has been positively modulated relative to Erk-2 due to its contact with the test compound (i.e., there has been a change in the relative phosphorylation status between the two isoforms in favor of Erk-1 ).

[47] "Modulation of phosphorylation status" or "modulation of phosphorylation activity" or like terms as used herein means an effected change in activity or phosphorylation state of a specified protein or collection of such proteins in a suitable test system that is capable of being phosphorylated upon contacting a test compound to the test system. Modulation Docket No. 354. PATENT of phosphorylation status or state may mean increasing or decreasing the number of covalently bound phosphate groups in a protein, changing the phosphorylation pattern within a protein, which may or may not be accompanied by an increase or decrease in the number of covalently bound phosphate groups, or increasing the amount of a phosphorylated protein resulting from contacting a test or candidate compound to a suitable test system. Modulation of phosphorylation status or state, may also mean changing the number or pattern of covalently bound phosphate groups in a protein or the amount of a phosphorylated protein in comparison to an effect a test compound has on a reference protein that is present in the same suitable test system resulting from contacting a test compound to the same suitable test system. Modulation of phosphorylation status or activity may be described relative to an isoform of the same protein in the same test system or to the same protein in a control test system to which is contacted the same test or candidate compound or from contact of the same test system with a test compound that is a reference compound (e.g., vehicle or positive or negative control compound). Relative modulation of activity or phosphorylation status or state is stated explicitly by describing the comparator protein or by describing the suitable test and control systems, otherwise relative phosphorylation status is implicitly understood by context.

[48] Examples for modulation of phosphorylation states are exemplified for Erk. An Erk isoform requires activation of tyrosine and threonine residues in the activation loop of the kinase for maximal activity towards its substrates, administration or application of a test compound to a suitable test system will affect one or both of these residue phosphorylations. Therefore, modulation of Erk-1/2 phosphorylation includes decreasing or increasing the phosphorylation state of one Erk isoform relative to the other isoform and includes increasing or decreasing the amount of phosphorylation in the activation loop of one isoform relative to the other isoform.

[49] Thus, an increase or decrease in the percent total phosphorylation (i.e., phosphorylation extent) of a collection of proteins for one Erk isoform protein relative to another protein collection of the other Erk isoform within the same suitable test system represents a relative change in phosphorylation status between the two Erk isoforms. This change in phosphorylation status may be accompanied by or associated with decreased or increased amounts of protein. For example, increased activation loop Tyr/Thr phosphorylation of Erk-1 may occur with decreased Erk-2 protein level or overall decreased Erk-1/2 protein level favoring Erk-1 (i.e., increased formation of pErk-1 relative to pErk-2 may occur along with increase gene transcription for one or both proteins or with overall decreases in gene transcription that favors protein levels of Erk-1 in comparison to Erk-2). Docket No. 354. PATENT

[50] An Erk isoform is typically activated within in vitro cell-based and in vivo systems suitable for determining relative changes in Erk phosphorylation status and usually occurs by sequential phosphorylation of tyrosine and threonine in the activation loop of an Erk isoform protein, wherein the tyrosine is usually first phosphorylated. Therefore, another relative change in phosphorylation status is increasing or decreasing the amount of tyrosine residue phosphorylation in the activation loop for a collection of proteins for one Erk isoform relative to a collection of proteins for the other Erk isoform within the same suitable test system.

[51] Another relative change in phosphorylation status of Erk-1/2 is increasing or decreasing the relative amount of di-phosphorylation in the activation loop of a collection of proteins for one Erk isoform protein relative to the other Erk isoform in the same suitable test system, wherein typically one or both isoforms are initial unphosphorylated (i.e., prior to stimulation with mitogen or contact with a test or candidate compound) or have basal levels (i.e., unstimulated) levels of phosphorylation.

[52] Thus, modulation of phosphorylation status may mean, depending on context, an increase or decrease in the number of phosphate groups covalently bound to a protein, phospho-protein, protein kinase or protein kinase substrate, effector protein other molecule so modulated (i.e., increase or decrease in overall phosphorylation status of the modulated molecule), an increase or decrease in phosphorylation of specified amino acid residues for a collection of specified molecules (e.g., for a specified isoform or a defined set of isoforms), or to an alteration of the phosphorylation pattern of a phospho-protein, which may or may not be accompanied by an increase or decease in the number of covalently bonded phosphate groups.

[53] A test or candidate compound that modulates the phosphorylation state of a protein kinase or transcription factor such that an activity of the specified kinase (e.g., phosphorylation of a downstream effector substrate) or transcription factor (e.g., transactivation activity of a nuclear hormone receptor towards a hormone-inducible gene) is increased is referred to as a positive modulator of that activity, whereas a test compound that modulates the phosphorylation state of a protein kinase or transcription factor such that the kinase or specified transcriptional activity is decreased is referred to as a negative modulator of that activity. A test or candidate compound that increases overall phosphorylation of a protein is referred to as an overall positive phosphorylation state modulator of that protein. A test or candidate compound that decreases phosphorylation of a protein is referred to as an overall negative phosphorylation state modulator of that protein. This positive or negative modulation of the protein's overall Docket No. 354. PATENT phosphorylation state may consequently modulate the activity of that phospho-protein positively or negatively. In those instances where a test compound, when contacted to a suitable test system, diminishes or abrogates phosphorylation of a protein such that the protein appears to remain at its basal phosphorylation state under conditions in which its phosphorylation would occur is also referred to as an overall negative phosphorylation state modulator of that protein.

[54] A test or candidate compound that changes the phosphorylation pattern of a protein, whether or not there has been a change in the number of covalently bonded phosphate groups, (i.e., total number of phosphate groups may be increased, decreased or unchanged) is also referred to as a positive phosphorylation state modulator. This positive modulation based upon a change in phosphorylation pattern may be accompanied by positive or negative phosphorylation state modulation of specified residues and by positive or negative modulation of an activity of the phospho-protein whose phosphorylation pattern has been affected. For example, a phosphoprotein containing pSer/pThr and/or pTyr (e.g. pErk-1 and pErk-2) that undergoes a change in phosphorylation pattern without a change in the total number of phosphate groups such that the number of phosphorylated tyrosines increases concommitant with decreased number of phosphorylated serine/threonines has its phosphorylation state positively modulated due to the change in phosphorylation pattern, but its overall Ser/Thr phosphorylation state has been negatively modulated.

[55] A test or candidate compound that results in removal of one or more phosphate groups from a phosphoprotein without addition of at least one phosphate group elsewhere in the phosphoprotein is not considered a change in phosphorylation pattern and thus is not a positive modulator of that phosphoprotein's phosphorylation state. Rather, it is a negative modulator of overall phosphorylation state. Conversely, test or candidate compound that results in addition of one or more phosphate groups to a protein or phosphoprotein without removal of at least one phosphate group is not considered a change in phosphorylation pattern and thus is not a positive modulator of that phosphoprotein's phosphorylation state. Rather, it is a positive modulator of overall phosphorylation state. Thus, removal of phosphate group(s) from a phosphoprotein to provide a protein absent such groups or addition of phosphate group(s) to a protein absent such group(s) to provide a phosphoprotein is not considered a change in phosphorylation pattern of that phosphoprotein or protein.

[56] Modulation of a phosphorylation pattern is sometimes described functionally such that a modulation by a test or candidate compound of a phospho-protein's phosphorylation pattern that results in negative modulation of pro-survival or pro- Docket No. 354. PATENT proliferative activity or positive modulation of apoptotic or anti-proliferative activity is referred to as a positive modulator of the protein's phosphorylation pattern. For example, modulation of the phosphorylation pattern in a transcription factor that results in negative modulation of nucleo-cytoplasmic transport of that transcription factor, negative modulation of its transactivation of anti-apoptotic or pro-proliferative genes or positive modulation of its transactivation of pro-apoptotic or anti-proliferative genes is referred to as a positive modulator of the transcription factor's phosphorylation pattern.

[57] A test or candidate compound that decreases phosphorylation of a protein is referred to as a negative phosphorylation state modulator. This positive or negative modulation of a protein's phosphorylation status is sometimes determined relative to a comparator protein present in the same suitable test system (i.e., positive or negative relative phosphorylation status). Furthermore, a test compound that increases the amount of a phosphorylated protein due to increased upstream phosphorylation activity acting upon the protein within a kinase cascade (and which may be concommitant with increased protein level) is also referred to as a positive modulator of that protein's phosphorylation state, whereas a test compound that decreases the amount of phosphorylated protein (which may be concommitant with decreased protein level) is referred to as a negative modulator of that protein's phosphorylation status.

[58] Usually, a positive phosphorylation status modulator of a protein kinase will also be a positive catalytic kinase activity modulator of that kinase and a negative phosphorylation status modulator of a protein kinase will also be a negative catalytic kinase activity modulator for that kinase. For other proteins, such as scaffold proteins and transcription factors, modulated phosphorylation states and catalytic kinase activities that are affected by a test or candidate compound may be inversely related.

[59] Sometimes a negative modulator of an activity, phosphorylation state or pattern results in suppression, or lack of an expected qualitative or quantitative increase, of that activity, state or pattern that would otherwise take place in a suitable test system in the absence of test compound.

[60] "Effector protein" or like term means a protein, polypeptide or other molecule that is downstream of a protein kinase, inositol kinase, lipid kinase, lipase or other enzyme (i.e., an upstream enzyme) or is terminal to a kinase cascade or signal transduction pathway whose phosphorylation status or activity is modulated by the upstream enzyme, kinase cascade or signal transduction pathway. This modulation is due to stimulation of a signal transduction pathway that results in modulation of the catalytic activity of the upstream enzyme and results in effect(s) on cellular program(s) not localized to upstream Docket No. 354. PATENT components of the signal transduction pathway. Effector proteins typically refer to those proteins that are direct substrates of an upstream protein kinase or proteins terminal to a kinase cascade or signal transduction pathway within a viable normal or cancerous cell or a cell undergoing apoptosis.

[61] Examples of cellular programs affected by the phosphorylation status of an effector protein include gene transcription, apoptosis, differentiation, cell cycle progression and metabolism. Erk effector proteins include certain apoptotic proteins, transcription factors, MAPKAPKs (MAP kinase activated protein kinases), MSKs (mitogen and stress activated kinases) and other cytosolic and nuclear localized proteins as disclosed herein. Other Erk effector proteins include plasma membrane, endomembrane, mitochondrial and cytoskeletal proteins as disclosed herein.

[62] "Kinase Substrate" or like term means a protein, polypeptide or other molecule that is capable of phosphorylation by a kinase. A kinase substrate includes an effector protein that is directly acted upon by a protein kinase, inositol kinase, lipid kinase or other kinase within a viable normal or cancerous cell or a cell undergoing apoptosis. A kinase substrate also includes a molecule directly acted upon by a protein kinase in a suitable cell-free test system or a suitable test system consisting essentially of cell membrane-disrupted cells. Therefore, a kinase substrate need not be associated with a kinase acting within a viable normal or cancerous cell or a cell undergoing apoptosis.

[63] "Suitable test system" as used herein means an in vitro or in vivo system to which can be contacted with a test compound in order to elicit effect(s) on one or more signal transduction pathways, nodes, complexes, kinase proteins or cascades or effects sequesterization, subcellular localization or nucleo-cytoplasmic transport of one or more protein kinases or signal transduction complexes.

[64] The suitable test systems can constitute cells or tissue in vivo (e.g., in xenograft test systems) or cells in tissue culture (i.e., in vitro cell-based test system) or constitute cell extracts in cell-free tests systems, wherein signaling through target molecules of interest, e.g., Erk-1 , Erk-2, PI3K proteins, GPR-C6a and/or AR, is(are) functional and phosphorylation changes or other biological responses as described herein, as for example for E-3a-diol and 3a-diol, in response to the test (or control) compound can be measured. Other cell-free test systems for measuring phosphorylation changes may be artificial in nature by reconstituting a signal transduction pathway comprising one or more proteins of that pathway in a suitable buffer with a downstream substrate that is capable of phosphorylation by at least one of the signal transduction pathway proteins.

[65] The compound to be contacted with a suitable test system will usually be in a Docket No. 354. PATENT vehicle, composition or formulation that is compatible with the test system, e.g., the test compound can be in a solution or suspension or it can be administered as a solid or liquid formulation to an animal such as a rodent (e.g., mouse or rat), or, for clinical assessment of the test compound, it can be administered to a human patient. Said contact is followed by measurement of target molecule phosphorylation states (e.g., Erk-1 , Erk-2) in cells or tissue samples taken from the animal or patient after administration of the test compound to the animal or patient. For controls, test systems will typically be contacted with a positive, negative, placebo and/or vehicle control or reference compound to confirm proper functioning of the test system in vitro or in vivo. As is apparent from the foregoing, the test compound and any control or reference compound or composition will be contacted with the test system (i) under conditions where the test system is functional, e.g., cells in tissue culture are maintained under standard growth conditions, (ii) for one or more sufficient periods of time, e.g., for about 5 seconds-120 minutes for cells in tissue culture or about 0.1 hour to about 48 hours for cells or tissue in vivo, and (iii) in one or more amounts or concentrations suitable to assess biological responses, e.g., dose- responses to the test compound and/or vehicle, placebo or control compound effects, if any. For assessing non-genomic effects in vitro as described herein that type of biological response is measured typically from within 5 sec to 30 min after contacting the test system with test compound. For genomic effects in vitro as described herein that type of biological response is measured typically from within 60-120 min. after said contact.

[66] Typical test compound final concentrations with the test system will be in a range, e.g., about 0.01 nM to about 20 mM or usually about 1 nM to about 10 mM, which can include one or more of about 0.05 nM, 0.1 nM, 1 nM, 10 nM, 100 nM, 200 nM, 1 mM, 2 mM and 10 mM. Concentrations of other compounds, e.g., components or excipients in the formulation that contains the test compound will typically be tested at the same or nearly the same concentrations they are at when the test compound is contacted with the test system.

[67] Suitable test systems are capable of responding to a test compound to be selected as a candidate compound that modulates an effect as described herein in a qualitatively or quantitatively similar manner when contacted with 17a-ethynyl-5a-androstane-3a, 17β- diol (E-3a-diol) or 5a-androstane-3a, 17p-diol (3a-diol) or another positive or negative control. Typically, suitable test systems comprise cells in vitro maintained or incubated in growth or culture media, in androgen- and growth-depleted media or under serum-starved conditions or are comprised of mammalian cells in vivo (e.g., as xenograft implanted cells in an animal or in situ tumor cells from carcinogen-induced cancer). Docket No. 354. PATENT

[68] Although human mammalian cells having endogenous functional human i-AR protein or transfected to contain functional i-AR gene or genetically engineered to express or overexpress a gene encoding functional i-AR protein are exemplified for some suitable test systems, the invention also contemplates suitable test systems comprising human cells transfected to contain functional rodent i-AR (e.g., murine or rat i-AR) or other mammalian i-AR gene and rodent cells transfected to contain functional rodent, human or other mammalian i-AR gene. The use of a mutated rodent or other mammalian mt-AR protein in mammalian cells transfected to contain functional mt-AR gene having analogous mutations described herein for the human mt-AR gene are also included, although preferred test systems, such as those in the claims or other embodiments described herein, contain functional human wild-type i-AR protein or the T877A mutant AR.

[69] The invention further contemplates suitable test systems comprising mammalian cells other than human cells having endogenous human i-AR protein or transfected to contain functional or i-AR gene or genetically engineered to express or overexpress a functional gene encoding functional i-AR protein. Those human or non-human mammalian cells may contain or be transfected to contain one or more genes encoding other functional nuclear hormone receptors proteins in place of or in addition to i-AR, including estrogen receptor (ER) or ERa and/or ERp. The mammalian cells may also or additionally contain, endogenously or by genetic engineering, one or more surface membrane receptors such as EFGR, ErbB2 or other ErbB receptor or GPR-C6a, an alpha-1 adrenergic receptor or other GPRC coupled to or capable of being coupled to Ga/q. All of the aforementioned human or non-human mammalian cells may also contain an inducible- reporter gene construct that has an upstream promoter for a functional nuclear hormone receptor whose gene is expressed endogenously or is expressed or over-expressed in the cell through genetic engineering.

[70] In some suitable in vivo test systems for evaluating test compounds contemplated as or selected for candidate compounds, a non-human mammal is transplanted with transformed or cancerous mammalian cells to provide a suitable test system (e.g., from cells that comprise the aforementioned suitable in vitro test systems), sometimes referred to as a xenograft tests system, wherein the cells have or are genetically engineered to have a gene encoding for expression or overexpression of functional i-AR protein with or without additional genetic engineering as previously described. Non-human mammals that may be used in those in vivo test systems include mouse, rat, rabbit, dog and non- human primate. Such in vivo test systems typically are xenograft models that have been used for evaluating anti-cancer treatments. Mammals of those xenograft models are Docket No. 354. PATENT usually immuno-compromised to allow for proliferation of the implanted cells, which may be supported by androgen administration, and include CD-1 nude mouse, nu/nu nude mouse, BALB/c nude mouse and RNU nude rat (T-cell deficient), NIH III nude mouse, SCID hairless outbreed mouse, SCID hairless congenic mouse, CB17 SCID mouse, SCID Beige mouse and NOD SCID mouse.

[71] "Control test system" as used herein refers to a suitable test system that is to be sham treated with compound, contacted with vehicle or contacted with a reference compound or composition that, depending on context, may serve as a positive or negative control test compound. Typically, the cells of the control test system are genetically the same as the cells comprising the test system to which test or candidate compound is contacted. Control test systems may also be derived from the suitable test system to which a test or candidate compound is to be contacted by genetic alterations to or by external stimulus of signal transductions pathways of the cells comprising the suitable test system. In this context the same test compound may be applied to both suitable test systems (i.e., the control test system and the original test system). Cells within a control test system used in screening of test compounds are sometimes referred to as control test cells.

[72] "Test compound" as used herein means a compound, or a composition comprising the compound, to be evaluated in a suitable test system for the presence of one or more of the activities for E-3a-diol described herein or for the ability to oppose one or more actions of 3a-diol described herein. Test compounds also include reference compounds whose effect on a suitable test system is known and which is to be compared to an effect (or lack thereof) provided by contact of another test compound to the same test system (i.e., a test or reference compound is contacted with a control test system). The reference compound may be a positive or negative control compound. For example, when assessing Galpha/q-mediated signal transduction a positive control compound may be E- 3a-diol and a negative control compound may be 3a-diol. Positive control compounds may also include androgen conjugates as described herein that are cell-impermeable and are useful for evaluation of test compounds for a subset of activities described herein for E-3a-diol.

[73] One subset of activities are non-genomic signaling that results from interaction of cell-impermeable conjugates with a membrane androgen receptor, and includes longer term non-genomic signal transduction effects subsequent to initial nongenomic signaling. Another subset of activities includes those associated with secondary genomic effects from nongenomic signaling. Typically, initial nongenomic signaling occurs within seconds Docket No. 354. PATENT or peaks at about 5 min to about 30 min of contacting a suitable test system with a test compound. More typically, the initial nongenomic signaling peaks within about 5 to 10 minutes and decays to baseline in about 1 hour and is thus characterized as rapid and transient. Longer term nongenomic signaling occurs subsequent to the peak of the initial nongenomic signaling and may overlap with the decay phase of the initial nongenomic signaling. Typically, the longer term nongenomic signaling is characterized as a plateau and persists after the rapid and transient phase has decayed to baseline. The longer term nongenomic signaling may occur prior to, concurrent with or after induction of secondary genomic effects affected by the initial nongenomic signaling.

[74] Test compounds additionally include test compounds shown to have one or more activities qualitatively or quantitatively similar to E-3a-diol, which are required for consideration or selection as a candidate compound, and which may also serve as a positive control compound for that activity. Test compounds also include candidate compounds to be evaluated for identification as low toxicity Erk inhibitors.

[75] Test compounds typically have a molecular weight of 200-1 ,000 amu or 200-800 amu, and are non-peptidic, but may be peptidic or have higher molecular weight as, for example, when a test compound is used as reference compounds that selectively elicits extracellular effect(s) by binding to a plasma membrane receptor (e.g. a GPCR) in comparison to intracellular effects (e.g., those resulting from binding to protein complexes or scaffold proteins in the cytoplasm or nucleus). Such reference compounds are typically cell-impermeable and include androgen-protein conjugates (e.g., testosterone-bovine serum albumin (T-BSA) or DHT-BSA conjugate), optionally coupled to a fluorophore, as described herein. Other peptidic substrates recognizes the CD or FD protein-protein interaction domain of a MAPK and may be derived from the amino acid sequence of the corresponding protein-protein interaction domain of a substrate or effector protein for that MAPK.

[76] "Candidate compound" as used herein is a test compound that exhibits one or more of the activities of E-3a-diol or opposes one or more activities of 3a-diol in an in vitro or in vivo model(s) predictive or indicative of efficacy for treating a hyperproliferation condition described herein in a mammal or is expected to have that condition, or is shown to have low toxicity in comparison to clinically studied MAPK inhibitors that are directed to the ATP binding site. Typically, candidate compounds positively modulate Erk-1 activation loop phosphorylation or catalytic kinase activity selectively in comparison to Erk- 2, negatively modulate Erk-2 activation loop phosphorylation state or catalytic kinase activity in comparison to Erk-1 or negatively modulate PI3K catalytic kinase activity or Docket No. 354. PATENT negatively modulate the PI3K regulatory subunit tyrosine phosphorylation state. More typically test compounds selected as candidate compounds elicit those Erk-1 and/or Erk-2 and PI3K effects.

[77] "Reference compound" or "control compound" as used herein is test compound that has one or more of the activities as described herein for E-3a-diol or 3a-diol for which comparison is to be made in a suitable test system to another test compound to be screened for that activity (i.e., a positive control). Other reference or control compounds lack one or more of these activities for which comparison is to be made in a suitable test system to another test compound to be screened for that activity (i.e., negative control). Reference or control compounds include E-3a-diol, 3a-diol, 17a-ethynyl-17p-hydroxy-5a- androstan-3-one (E-DHT), 3β, 17p-di-hydroxyandrost-5-ene (βΑΕϋ), 3β, 7β 17β-ιτί- hydroxyandrost-5-ene (βΑΕΤ), 3p,17p-di-hydroxy-5a-androstane (3p-diol), a hormone i- AR agonist, such as testosterone (T) or 17p-hydroxy-5a-androstan-3-one (DHT) for Galpha/i and/or Galpha/q-mediated signaling, or a cell-impermeable GPR-C6a agonist (e.g., T-BSA and DHT-BSA conjugates) for Galpha/i-mediated signal transduction as described herein.

[78] "Liver toxicity" as used herein means an increase of one or more liver enzyme (alanine transaminase, aspartate transaminases, alkaline phosphatase, γ-glutamyl transpeptidase) levels when a test or candidate compound is evaluated in a suitable animal model to an extent that is inconsistent with approvable treatment of an intended hyperproliferation condition in humans. A test or candidate compound having low liver toxicity has about 50% or more of the treated animals, e.g., in about 70% to about 90% of the treated animals with liver enzyme levels that are not increased more than about 3-fold compared to normal values. For humans, a low liver toxicity compound exhibits liver enzyme level increases less than about 2-fold, typically less than about 1 .5-fold, in about 70% or more of treated humans, more typically in about 70% to about 95% of the treated humans. Indications of low liver toxicity of a test or candidate compound for approvable treatment of an intended hyperproliferation condition in humans are provided by suitable animal models showing liver enzyme levels that are not increased more than about 2.5- fold compared to normal values. Initial assessments of liver toxicity are usually conducted with animal models including those with rodents, such as a mouse or rat.

[79] Typically, a test or candidate compound selected as a candidate low toxicity Erk modulator for further evaluation for liver toxicity exhibits more than about 2.5-fold increase in a liver enzyme level in no more than 70% of treated rodents, more typically about 70% to about 95% of the treated rodents do not have more than about 2.5-fold increase in a Docket No. 354. PATENT liver enzyme level when the test or candidate compound is administered to animals comprising the animal model at mg/Kg dose levels of 2X to 10X or more than is expected to elicit a desired therapeutic effect (i.e., has a therapeutic index of at least 2-10). More typically, a test or candidate compound selected for further evaluation of liver toxicity (i.e., a candidate low toxicity Erk modulator) does not have more than about 2.5-fold increase in level for any of the aforementioned liver enzymes in 95% or more of the treated rodents.

[80] "Protein-protein interaction domain" as used herein means a region in the tertiary structure, which may be comprised of continuous or discontinuous amino acid sequences, of a protein wherein the domain is capable of non-covalently binding to another tertiary structural region of the same or different protein. These interaction domains include the Src homology domains SH2 and SH3. Such protein-protein interaction domain are often found in components of protein kinase signal transduction pathways, e.g., in scaffolding proteins, adaptor proteins, membrane-bound receptors, internalized GPCR or RTK receptors and protein kinases and phosphatases as described herein. Protein-protein interaction domains also include CD and FD domains of MAPKs that recognize complementary regions in their binding partners or substrates such as MAPK activator or inhibitor (i.e., regulatory) proteins, adaptor or scaffold proteins, effector proteins, protein phosphatases or another protein kinase.

[81] "Protein complex" as used herein refers to a collection of two or more different proteins wherein at least one protein that is a scaffold or adaptor protein and another protein that affects or modulates the phosphorylation (e.g., a protein, lipid or small molecule kinase or phosphatase) or GTP-GDP binding status of a third protein (i.e. a GTP-binding protein) that is typically activated in the GTP bound state. Those GTP- binding proteins include, e.g., Ga-subunits of heterotrimeric G proteins, such as Ga/q family members, and are described in Takai, Y. et al. (2001 ) and Morris, A.J. and Malbon, C.C. (1999). A protein complex may also refer to a collection of different proteins wherein at least one protein is a protein, lipid or small molecule kinase or a protein that effects or modulates the phosphorylation or GTP-GDP binding status of a GTP-binding protein.

[82] The protein complex typically contains proteins that are components of one or more protein kinase signal transduction pathways or pathways for GPCR or RTK signaling, which often include a scaffolding or adaptor protein such as Ras, SOS, She, β- arrestin, Tfg, protein 14-3-3 or a kinase or nuclear hormone receptor with several protein- protein interaction domains, including Raf, PI3K, AR and ER. The scaffold or adaptor proteins within these complexes are bound non-covalently to one or more effector or Docket No. 354. PATENT modulator proteins, including but not limited to a protein kinase, a protein phosphatase, a G protein monomer or Ga/β heterodimer, a membrane bound or internalized GPCR or RTK or a protein kinase effector protein such as a transcription factor, apoptosis-related protein or cytoskeletal protein.

[83] The protein complexes may reside preassembled in subcellular locations or domains including the cytoplasm, nucleus, nucleolus, mitochondria, lipid rafts, caveolae and clathrin pits or at membranes including the cytosolic plasma membrane or membranes of subcellular structures including late and early endosomes, the endoplasmic reticulum and Golgi apparatus or may reside bound to cytoskeletal structures. In normal cells this pre-assembly increases the efficiency of signal transduction by permitting proper recruitment of other signal transduction components upon activation of a protein complex member, allows appropriate crosstalk with parallel signal transduction pathways for signal integration and properly directs the integrated signal to downstream effector proteins.

[84] In abnormal cells, signal transduction becomes aberrant due excessive stimulation of signal transduction pathways through gene mutations. Examples of "gain of function" mutations include those seen for PI3K, Src, Akt, Ras, Raf, RTKs (e.g., EGFR, PDGFR, Her2/neu), AR, ER, c-myc and c-fos. Examples of "loss of function" mutants include those seen for the tumor suppressor gene PTEN, p53 and Rb. Signal transduction in hyperproliferating cells usually results from several of these mutations leading to excessive or inappropriate cross-talk between signal transduction pathways.

[85] "Signal transduction node" as used herein is a component of a signal transduction pathway capable of having catalytic activity for incoming signal amplification, or is a protein complex containing as a signal transduction component the protein capable of this catalytic activity and one or more other signal transduction components, that signals to multiple downstream effector proteins and/or upstream regulator proteins. The effects of these components on downstream or upstream signaling is dependent on the phosphorylation states of multiple protein kinases that act upon them or on the activities of protein complexes from other signal transduction pathways. Other times the dependency results from sharing of scaffolding protein(s) or other components within a protein complex involved in one signal transduction pathway with a protein complex of another signal transduction pathway.

[86] "Signaling pathway" or "signal transduction pathway" as used herein refers to a sequence of biochemical events or the proteins and relay molecules involved in these events that transfer the consequence of a ligand binding event originating externally or internally to a cell to an effector protein or receptor. The external binding events may Docket No. 354. PATENT result, for example, from binding of a cytokine or mitogen to a receptor tyrosine kinase (RTK) or from an agonist binding to a membrane-bound G protein-coupled receptor (GPCR). Internal binding events that initiate signal transduction may result, for example, from a hormone ligand binding to its nuclear hormone receptor (NHR). The consequence (or signal) from these initial binding events are then transferred to another protein whose catalytic action or its effect on the catalytic action of another downstream protein amplifies the signal, which then may be passed along to yet another protein for further amplification to eventually modulate the activity or phosphorylation state of an effector protein or substrate terminal to the signal transduction cascade. In some hyperproliferating cells the binding event that initiates signal transduction is superfluous, since the receptor has become constitutively active due to mutation and is thus able to initiate signaling on its own (i.e., in absence of agonist ligand binding).

[87] The sequence of biochemical events for signal transduction typically employs a kinase cascade, as defined herein, which is supported by various accessory proteins, including scaffolding and adapter proteins, to guide each biochemical event to the appropriate downstream target. Sometimes the kinase cascade is engaged by activation of a tyrosine kinase receptor or cytoplasmic protein (e.g., EFGR, PDGFR, Her2/neu, Src) the catalytic kinase activity of which may reside within the binding receptor or a protein kinase recruited to the activated receptor. Other times activation of a serine-threonine kinase (e.g., Raf-1 , Ras), which has been recruited to the signal-initiating receptor through intermediary binding by scaffolding, adaptor or other proteins having multiple protein- binding domains (e.g., KSR, Tfg, β-arrestin, SOS, She, i-AR), results. In some instances, the kinase cascade is initiated by G protein activation through a ligand-bound GPCR (e.g., "inverse" agonist activation of a Ga/q coupled GPCR with respect to Galpha/i activation), whereby an activated G protein monomer through scaffolding interacts with a protein kinase, inositol lipid kinase or lipase. In other instances, engagement of the protein kinase cascade is more indirect as for example when a small molecule second messenger is produced by action of an enzyme (e.g., PI3K, ΡΙ_0-β _{1 2} , ΡΙ_0-γ _1/2 ) that has been activated by G proteins (e.g., Ga/q monomer or ΰβ/γ heterodimer) such that subsequent binding of the messenger to a downstream protein kinase results in its activation (PIP3 from PI3K to activate Akt or DAG from PLC to activate certain PKC isoforms).

[88] "Kinase cascade" as used herein refers to a collection of kinases that engage in sequential transfer of ATP γ-phosphate groups catalyzed by one kinase to form another activated protein kinase or to catalyze formation of a kinase or other enzymatic product (i.e., PIP3, DAG, cAMP) that activates a downstream protein kinase within the cascade. A kinase cascade in normal cells typically amplifies a signaling event and directs the Docket No. 354. PATENT outcome to appropriate effector proteins or substrates. A kinase cascade may respond to other signaling events and thus be activated or inhibited by a kinase involved in a different kinase cascade either directly or indirectly through effects on shared scaffolding proteins thereby resulting in signal transduction cross-talk. The signaling events that enter a kinase cascade may originate from a membrane-bound or internalized receptor or from cellular stress response proteins. The signaling event may be triggered by agonist binding to a receptor or spontaneous activity of the receptor. In hyperproliferating cells this spontaneous activity may become constitutively amplified due to mutations of the gene encoding the receptor.

[89] "Cross-talk" as used herein refers to the result of one signal transduction pathway influencing the output from another signal transduction pathway wherein the pathways are normally distinguished by their unique sets of effector substrates (some members of which may be shared).

[90] In cancer or other hyperproliferating cells, typically multiple mutations are required that result in aberrant stimulation of and aberrant cross-talk between signal transduction pathways for the cell to escape its differentiated state and evade apoptosis. Cross-talk between pathways is enabled by shared use of signal transduction nodes and overlapping requirements for components within these and other protein complexes. Through excess or lost activity of those signal transduction components, which may be attributable to dysregulated protein levels or phosphorylation states, protein complexes form improperly. That improper assembly may result in a protein complex forming in an inappropriate subcellular domain or to inhibition or diversion of proper trafficking of its resident components. Cross-talk that is normally present within a cell then becomes dysregulated due to excessive signaling between aberrant pathways or cross-talk becomes established between pathways through inappropriately formed protein complexes.

[91] In cancer or other hyperproliferating cells, aberrant crosstalk typically occurs between the Ras-Erk and PI3K signal transduction pathways. In androgen-associated conditions aberrant cross-talk between those two pathways typically results in dysregulated i-AR nuclear transactivation that supports proliferation. The aberrant cross- talk or i-AR dysregulation may be mediated or enhanced by excessive Src kinase activity on or within protein complexes involving one or more scaffolding and/or adaptor proteins that are shared between Ras-Erk and PI3K signal transduction pathways. That excessive activity may result from RTK activation, which itself is able to inappropriately increase pErk-1/2 levels in an isoform indiscriminate manner. That indiscriminate isoform activation also occurs through protein complexes that includes one or more scaffolding and/or adaptor proteins as defined herein. Docket No. 354. PATENT

[92] Therefore, a test or candidate compound will typically be capable of disrupting mitogenic-promoting protein complexes so as to re-regulate crosstalk between PI3K-Akt and Ras-Erk signal transduction. That outcome will typically re-regulate i-AR nuclear transactivation such that catastrophic cell differentiation (i.e., differentiation that induces apoptosis) occurs. Typically, a test or candidate compounds that re-regulates crosstalk between PI3K-Akt and Ras-Erk signal transduction will do so by redirection of Src activity such that excessive signaling through the Erk and/or PI3K signal transduction node(s) is(are) negatively modulated.

[93] "Translocation", "transport", "Nucleo-cytoplasmic translocation", "nucleo- cytoplasmic trafficking" or like terms as used herein refers to the movement of signal transduction components between the cytoplasm and the nucleus. This translocation or trafficking may result from passive diffusion or facilitative transport. For example, upon activation of a MAPK, its phosphorylated form is transported to the nucleus, which results in phosphorylation of transcription factors responsible for expression of early response genes such as c-fos and myc. Residence time within the nucleus will determine the phosphorylation state of the MAPK when it is translocated back into the cytoplasm, since many of the deactivating phosphatases are found in the nucleus.

[94] A test compound that decreases the rate of pErk-1 or pErk-2 transport from the cytosol to the nucleus or increases the fraction of total pErk in the cytoplasm relative to the nucleus is defined as a negative modulator of nucleo-cytoplasmic translocation. A test compound that increases rate of pErk-1 or pErk-2 transport from the cytosol to the nucleus or increases the fraction of total pErk in the nucleus relative to the cytoplasm is referred to as a positive modulator of nucleo-cytoplasmic translocation.

[95] "Sequester", "sequesterization" or like terms as used herein refers to preferential localization of a signal transduction component or a protein complex containing that component to a membrane, cytoplasm, nucleolus, mitochondria or other subcellular compartment. In normal cells, sequesterization provides proper direction of an incoming signal to a signal transduction pathway (or additional pathways by appropriate signal transduction cross-talk) that results in activation of intended effector protein(s) or substrate(s). In abnormal cells having excessive signaling through a signal transduction pathway, induced sequesterization by a test or candidate compound of a component of that cascade would negatively modulate the aberrant signaling or additionally redirect this activity away from supporting proliferation or survival.

[96] "Hyperproliferation condition" as used herein means a condition or disease state, e.g., a cancer that is characterized by an abnormally high rate or a persistent state of cell Docket No. 354. PATENT division is uncoordinated with that of the surrounding normal tissues, and persists after, e.g., cessation of the stimulus that may have initially evoked the change in cell division. The rate of proliferation by the abnormal cells may be accelerated or similar to that of normal surrounding tissue. This uncontrolled and progressive state of cell proliferation will result in a tumor that is benign, potentially malignant (premalignant) or frankly malignant. Hyperproliferation conditions include those characterized as a hyperplasia, dysplasia, adenoma, sarcoma, blastoma, carcinoma, lymphoma, leukemia or papilloma or other conditions described herein.

[97] "Hormone-associated" cancer, precancer or benign hyperplasia as used herein refers to a hyperproliferation condition that responds negatively (i.e., promotes cell cycle progression) or positively in a therapeutic sense (i.e., retards cell cycle progression or promotes apoptosis), to hormone manipulation or is a condition whose genesis, persistence, invasiveness, refractivity, severity in symptoms or responsiveness to cancer chemotherapy are attributable or related, in part or in whole, to one or more hormone levels. Hormone-associated cancers include prostate cancer, breast cancer, ovarian cancer, cervical cancer, uterine cancer, endometrial carcinoma, adenocarcinoma, malignant melanoma or other conditions as described herein. Some additional hormone- associated cancers are described in Miller, A.B. (1978). The screening methods described herein yield candidate compounds that are low-toxicity drugs for treating hyperproliferation conditions and hormone-associated cancers. Hyperproliferation conditions in which the hyperproliferating cells contain functional i-AR or which respond to androgen(s) are referred to as androgen-associated conditions.

[98] "Scaffold protein" or "adaptor protein" as used herein is a protein capable of bringing together two or more other proteins to facilitate creation of larger signaling complexes and typically contain two or more protein-protein interaction domains for this purpose. By linking specific proteins together, cellular signals can be propagated to elicit an appropriate response from the cell initiated from a membrane-bound receptor or an intracellular receptor.

[99] Designation as an adaptor protein is usually applied to a protein that lacks intrinsic catalytic activity and brings together proteins wherein at least one of these protein is capable of such activity (i.e., catalytic proteins). Adaptor proteins include for example 14- 3-3 protein, CrkL, MyD88, Grb2, Nek and She. A scaffold protein typically refers to a protein that recruits or anchors other proteins to a subcellular location. Scaffolding may therefore result in subcellular localization of the components to direct recruitment of other components to the protein complex formed from that scaffolding or may direct signal transduction to kinase effector proteins or substrates localized to a subcellular Docket No. 354. PATENT compartment. Scaffolding may also contribute to integration of signal(s) emanating from one or more other signal transduction pathways. A scaffold protein may be devoid of intrinsic activity but may include those proteins having transcriptional activity or protein or inositol lipid kinase activity, lipase activity or other catalytic activity. Non-limiting examples of these dual role proteins (i.e., scaffolding and kinase activities) include Raf-1 , Tpl-2, Src, PI3K, ΡΙ_0-βι/2, PLC-Y _{! 2} or are other kinases or lipases associated with components of GPCR, RTK, intracellular AR or ER, Ras-Erk or TNFa-NF-κΒ signal transduction pathways or other signaling pathways disclosed herein.

[100] Scaffold- and adaptor-mediated complexes assemble through protein-protein interactions and engage in signaling resulting from or mediated by a kinase cascade within a signal transduction pathway. Examples of scaffolding and adaptor proteins in MAPK signaling are described in Kolch, W. (2000), Kolch, W. (2005) and Brown, M.D. and Sacks, D.B. (2009). Further examples of scaffolding and adapter proteins for these and other signal transduction pathways are provided in Good, M.C. et al. (2011 ), Ritter, S. L. and Hall, R. A. (2009), Kristiansen, K. (2004) and Hall, R.A. and Lefkowitz, R.J. (2002).

[101 ] Scaffold and adaptor proteins previously found associated with Erk include isoforms of KSR, Sef, 14-3-3, IQGAP1 and β-arrestin. One KSR isoform, KSR-1 , acts as a positive regulator of Ras-Erk signaling by binding Raf-1 and MEK-1/2 to activate Erk-1/2 in response to growth factor stimulation. A second isoform KSR-2 is implicated in the activation of MEK-3 and Erk by the MAP3Ks MEKK-3 and Tpl-2. As discussed herein, Tpl-2, in addition to its protein kinase activity, is also considered to be a scaffold protein and is involved in pro-inflammatory signaling. Undesired inflammation from this signaling may support initiation or progression of hyperproliferation conditions disclosed herein through downstream activation of NF-KB.

[102] Protein scaffolding also supports other protein-protein interactions that result in additional cross-talk between Ras-Erk and other kinase signaling cascades. For example, Raf-1 associates with Tpl-2, which therefore permits cross-talk between the p38 MAPK pathway (via Tpl-2 activation of MEKK-3), the TNFCC-NFKB pathway (via Tpl-2 activation of IKK) and the Ras-Erk pathway (via Raf-1 activation of MEK-1 or MEK-2). These aspects of Raf-1 function indicate participation of Raf-1 in assembling multi-protein signaling complexes. Thus, Raf-1 may also be considered a scaffolding protein in that regard. Tpl- 2 (Cot) is a MAP3K involved in NFKB activation. Intersection of TNFCC-NFKB signaling with Ras-Erk signaling occurs through Tpl-2 and is mediated by the scaffold protein isoform KSR-2, thus providing another avenue of cross-talk between the two signal transduction pathways. Docket No. 354. PATENT

[103] Due to the presence of SH2 and SH3 domains, PI3K may also serve in scaffolding of protein complexes in addition to having kinase activity. For example, the p85 regulatory subunit of PI3K directly interacts with intracellular AR or ERa, which is enhanced by i-AR interaction with androgen. Furthermore, Src positively modulates the interaction of PI3K with AR. Thus, PI3K-AK signaling may also cross-talk with AR and Src signaling. Tfg is also considered a scaffold protein as disclosed herein that lacks intrinsic kinase activity due to the presence of various protein-protein interaction domains that may facilitate cross-talk between the Ras-Erk and PI3K-Akt signal transduction pathways.

[104] The phosphoserine adapter protein 14-3-3 is involved in regulation of cell-cycle check points, proliferation, differentiation and apoptosis and acts primarily by effecting the subcellular localization of its binding partners. The isoform 14-3-3 δ inhibits Ras-Erk signal transduction by acting as a negative regulator of Raf-1 by binding to phoshoserine- 259 of Raf and sequestering this MAP3K away from the cytoplasmic membrane to the cytosol.

[105] The subcellular localization of 14-3-3 δ also affects the pro-apoptotic activity of Bad. Bad promotes apoptosis by binding Bcl-2 and Bcl-x _L at the mitochondrial membrane. However, phosphorylation of Bad generates a binding site to 14-3-3 to cause Bad to disengage from Bcl-2 and Bcl-x _L leading to its sequesterization in the cytosol. Those proteins freed from Bad in turn associate with Bax and Bak to prevent their aggregation on mitochondria, and subsequent cytochrome c release, thereby additionally inhibiting apoptosis. Thus, dysregulated Erk can indirectly inhibit apoptosis though its phosphorylation of p90 ^{RSK 1} , which then phosphorylates Bad to deactivate this protein's pro-apoptotic activity and whose release from Bcl-2 and Bcl-x _L further contributes to inhibition of pro-apoptotic signaling. As described herein, dysregulated Erk may also directly inhibit apoptosis by its phosphorylation of pro-apoptotic proteins, including caspase-9 or Bim.

[106] Another protein that interacts with Erk by protein-protein interactions includes MEK-partner 1 (MP-1 ). MP-1 links Erk-1 with MEK-1 , thus favoring Erk-1 activation over Erk-2. In order to augment Erk-1 signaling, MP-1 requires its interacting protein p14 as for example in EFG-mediated activation of Erk-1 . Therefore, compounds that improve binding interactions between one or more or more of Erk, MP-1 or p14 or can substitute for MP-1 or p14 will typically favor cytosolic signaling through Erk-1 in comparison to nuclear signaling through Erk-2.

[107] Raf kinase inhibitor protein (RKIP) is a member of the phosphatidylethanolamine- binding protein family and behaves as an adaptor protein in Ras-Erk and TNFa-NF-kB Docket No. 354. PATENT signal transduction pathway that disrupts rather than facilitates signaling. In this regard, RKIP has been shown to disrupt the Raf-1 -MEK1/2-ERK1/2 and NF-κΒ signaling pathways, via physical interaction with Raf-1 -MEK1/2 and the ΙκΒ (inhibitor of NF-KB)-IKK (IKB kinase-α/β) complexes, respectively, thereby abrogating the pro-survival and anti- apoptotic properties of these signaling pathways. As a result, RKIP functions as a metastatic tumor suppressor in prostate and breast cancers and sensitizes human prostate and breast cancer cells to drug-induced apoptosis.

[108] IQGAP1 also regulates many signaling pathways and is a scaffold protein that binds directly to B-Raf, MEK-1/2 and Erk-1/2. Analogous to KSR binding to Raf-1 , IQGAP1 may regulate B-Raf signaling. B-Raf is found mutated in various cancers including melanomas, thyroid carcinomas and colorectal cancer.

[109] In EFG-dependent activation, the interaction between IQGAP1 and MEK-1 increases while that with MEK-2 decreases. The changes in these interactions would favor Erk-2 signaling from B-Raf activation since MEK-1 is known to preferentially sequester unactivated Erk-2 in the cytoplasm (in preparation for its activation upon stimulation of the Ras-Erk signal transduction pathway) in comparison to MEK-2. Additionally, IQGAP1 regulates actin and microtubular cytoskeletons, which links Erk signaling to cytoskeletal dynamics. This interaction is supported by the finding that Erk-2 will localize to microtubule-associated protein-2 (MAP-2).

[1 10] Furthermore, stabilization of the protein complex IQGAP1 -B-Raf-MEK-1 -Erk-2 should compete with Erk-2 binding to MAP-2 or slow nuclear entry (i.e., negatively modulate nucleo-cytoplasmic transport) of activated Erk-2 since MEK-1 is known to preferentially retain unactivated Erk-2 in the cytosol pending its activation upon stimulation of the Ras-Erk signal transduction pathway. That effect on scaffolding on MEK-1 -Erk-2 should thus retard aberrant cytoskeletal rearrangements associated with neurodegenerative diseases without adversely effecting basal levels of Ras-mediated Erk activity (i.e., cytoplasmic activity of pErk-1 ). Additionally, a steroid binding site is present in the NH ₂ -terminal region of MAP-2C, which is a region specific to this isoform that recognizes C19 steroids such as DHEA.

[1 1 1 ] Sef or similar expression to FGF is a scaffolding protein first identified as an inhibitor of MEK/Erk-dependent signaling in fibroblast growth. Sef action directs MEK-Erk complexes to the Golgi apparatus to inhibit dissociation of the complex. This in turn results in inhibition of activated Erk translocation to the nucleus and promotes activity at cytoplasmic targets. In aggressive forms of prostate cancer (CaP), Sef is down-regulated. Docket No. 354. PATENT

[112] Other scaffold proteins include Tfg and isoforms of β-arrestin discussed elsewhere. Scaffold proteins are often devoid of protein kinase or transcriptional activity; however, proteins having such activity, but also serving as scaffolding proteins depending on context as disclosed herein, include Raf-1 , Tpl-2 and i-AR. Scaffolding by i-AR at ErbB receptors may depend upon activation of the intracellular hormone receptor by Src phosphorylation, since AR-mediated non-genomic effects from this scaffolding apparently does not require (but is enhanced by) androgen.

[113] Typically, a test or candidate compound having anti-proliferative or pro-apoptotic effects prevents or inhibits the formation of protein complexes involved in aberrant cross- talk between the PI3K-Akt and Ras-Erk signal transduction pathways and/or prevents or inhibits the formation mitogenic Src protein complexes. Such protein complexes include those comprising (1 ) Class I PI3K regulatory subunit p85a or ρ85β and Ras, (2) p85a or ρ85β and intracellular AR or ERa, (3) p85a or ρ85β, i-AR and Src, (4) EFGR, i-AR and i- ERa, and (5) i-AR, i-ERa and Src and 6) protein complexes wherein the aforementioned complexes additionally comprise a scaffold or adaptor protein, wherein i-AR and/or i-Era in the aforementioned protein complexes act as scaffold proteins.

[114] The negative modulation of aberrant crosstalk between PI3K-Akt and Ras-Erk pathways may result from formation or stabilization of alternative protein complexes that reduces signal transduction through one or both PI3K and Erk signal transduction nodes. Those alternative complexes include those comprising p85a or ρ85β and Galpha/q (GTP) or p-Tyr-Galpha/q. A test or candidate compound having anti-proliferative or pro-apoptotic effects may also stabilize formation of proteins complexes that isoform selectively localizes pErk-1 into a cytoplasmic compartment relative to its nuclear translocation or isoform selectively localizes Erk-2 in comparison to Erk-1 to a cellular compartment that prevents or inhibits its phospho-activation, thereby resulting in preferential phospho- activation of Erk-1 on stimulation of Ras-Erk signaling. Such complexes include those comprising Erk-1 and β-arrestin, Erk-1 and MAP-2, or Erk-2 and MEK-1 and may additionally comprise an Erk scaffold protein as described herein.

[115] "Non-genomic effect" as used herein means an effect or activity from a transcription factor not directly related to transactivation of gene transcription or an effect or activity from an agonist of an intracellular transcription factor that is not initiated by binding of agonist to the transcription factor and does not directly result in transactivation of gene transcription. Those effects are typically due to intracellular nuclear hormone receptors (i-NHRs) acting upon cytosolic proteins or from activation of membrane receptors that respond to extracellular hormones or synthetic analogs thereof. These Docket No. 354. PATENT non-genomic effects are typically mediated by kinase cascades that terminate in phosphorylation of cytosolic proteins. An activity from such a kinase cascade is sometimes initiated by i-NHR agonist binding to a GPCR or from scaffolding of an i-NHR to an activated RTK, an intracellular membrane or a cytoskeletal structure. Besides activation of signal transduction pathways, other non-genomic effects include effects on ion transport (e.g., intracellular Ca ²⁺ mobilization) or Erk kinase activation, which typically are hallmarks of GPCR activation. Depending on the signal transduction pathway that is activated or the nature of the stimulated cross-talk between signal transduction pathways, a non-genomic effect may result in apoptosis, survival, cell cycle arrest, proliferation or variation in metabolism. As a secondary consequence of phosphorylation of a cytosolic effector protein by non-genomic action, gene transactivation may eventually result. Those secondary consequences and activities resulting therefrom are indirect and therefore are not considered effects directly related to gene transcription.

[116] "Genomic effect" as used herein refers to transactivation of gene transcription directly resulting from an activated transcription factor binding to its cognate promoter upstream of the gene to be transcribed. Gene transcription that results as a secondary consequence of a primary non-genomic effect is excluded as a direct genomic effect.

[117] "Substantially similar to" as used herein refers to an activity or property of a test or candidate compound equal or nearly equal to that same activity or property determined in identical suitable test systems for a reference compound, usually a positive control test compound.

[118] "Quantitatively similar to" as used herein refers to a numerically determined activity or property of a test or candidate compound or is a numerically determined change in a property or activity of a suitable test system to which a candidate or test compound is contacted that is similar in, or the same as, the magnitude of that same activity or property determined for another test or candidate compound, usually a positive control compound, when contacted to an identical test system or is 50-150%, 60-140%, 80-120% or 90-1 10% of that activity or property numerically determined for a positive control test compound.

[119] "Qualitatively similar to" as used herein refers to a test or candidate compound that exhibits an activity or property or affects a change in a suitable test system to which the compound is contacted in the same direction as that determined or observed for another test or candidate compound, usually in reference to a positive control test (i.e., both result in positive or negative modulation of the determined activity or property). For example, two compounds that negatively modulate the phosphorylation state of protein to observable extents, when applied separately to identical suitable test systems containing Docket No. 354. PATENT the protein to be affected, exhibit qualitatively similar activities. A test or candidate compound that is absent an observable activity or property in relationship to a negative control compound may also be characterized as qualitatively similar to that negative control compound.

[120] "Alkyl" as used here refers to moieties with contiguously linked normal, secondary, tertiary or cyclic carbon atoms, i.e., linear, branched, cyclic or any combination thereof wherein all of the carbon atoms are saturated. An alkyl substituent for a structure is comprised of an alkyl moiety that is single bonded to that structure through a saturated (i.e., sp ³ ) carbon atom of the alkyl moiety. An alkyl moiety that is substituted with one or more moieties as described below for alkenyl, alkynyl, cycloalkyl, aryl and heterocycle (including heteroaryl) provides an alkyl substituent having unsaturated carbons. The number of carbon atoms in an alkyl moiety or substituent is typically 1 to about 10. Expressions such as d- ₆ alkyl or C1 -6 alkyl mean an alkyl moiety or substituent containing 1 , 2, 3, 4, 5 or 6 carbon atoms. When an alkyl substituent of an organic moiety, such as an androst-5-ene, androst-4-ene, androst-1 -ene, androst-1 ,5-diene androst-1 ,4-triene, 1 ,3,5(10)-estratriene, 5a-androstane, 5a-androstan-1 -ene, 5β- androstane or 5p-androstan-1 -ene moiety is specified, an unsaturated carbon of that substituent is covalently attached by a single bond to that organic moiety. Alkyl species include by way of example and not limitation the fully saturated groups methyl, ethyl, 1 - propyl (n-propyl), 2-propyl (/so-propyl, -CH(CH ₃ ) ₂ ), 1 -butyl (n-butyl), 2-methyl-1 -propyl (/so-butyl, -CH ₂ CH(CH ₃ ) ₂ ), 2-butyl (sec-butyl, -CH(CH ₃ )CH ₂ CH ₃ ) and 2-methyl-2-propyl (f- butyl, -C(CH ₃ ) ₃ ). More typically, alkyl moieties or substituents are species or genera selected from the group consisting of Ci _-8 alkyl, d- ₆ alkyl, Ci _-4 alkyl moieties, methyl and ethyl, substituted or unsubstituted (i.e., optionally substituted).

[121] "Alkenyl" as used here means a moiety that contains one non-aromatic sp ² hybridized carbon atom or two contiguous non-aromatic sp ² hybridized carbon atoms and includes a non-aromatic ethenyl (e.g., -CH=CH-) moiety. An alkenyl substituent for a structure is comprised of an ethenyl moiety that is single bonded to that structure through an unsaturated carbon atom of the ethenyl moiety. Alkenyl moieties or substituents may be comprised, in addition to the ethenyl moiety, of contiguously linked normal, secondary, tertiary or cyclic carbon atoms that are fully saturated, i.e., linear, branched, cyclic and/or one or more unsaturated alkyl moieties as described below for alkenyl, alkynyl, and aryl moieties. The number of carbon atoms in an alkenyl moiety typically is 2 to about 10. C ₂ . ₆ alkenyl or C2-6 alkenyl means an alkenyl moiety containing 2, 3, 4, 5 or 6 carbon atoms and is inclusive of the carbons of the ethenyl moiety that defines it as an alkenyl substituent. Docket No. 354. PATENT

[122] When an alkenyl moiety is specified as a substituent of an organic moiety, such as an androst-5-ene, androst-4-ene, androst-1 -ene, androst-1 ,5-diene androst-1 ,4-triene, 1 ,3,5(10)-estratriene, 5a-androstane, 5a-androstan-1 -ene, 5p-androstane or 5β- androstan-1 -ene moiety, an sp ² carbon of that substituent is covalently attached by a single bond to that organic moiety. Species include by way of example and not limitation any of the alkyl moieties or substituents that have an internal double bond such as vinyl (- CH=CH ₂ ), allyl (-CH=CHCH ₃ ), 1 -methylvinyl, butenyl, iso-butenyl, 3-methyl-2-butenyl or 1 - pentenyl. Typically, alkenyl moieties or substituents are species or genera selected from the group consisting of C ₂ . ₈ alkenyl, C ₂ . ₆ alkenyl, C ₂ - ₄ alkenyl, or vinyl, substituted or unsubstituted (i.e., optionally substituted).

[123] "Alkynyl" as used herein refers to linked normal, secondary, tertiary or cyclic carbon atoms where one or more triple bonds are present, usually 1 . Alkynyl moieties or substituents may be comprised, in addition to the ethenyl moiety, of contiguously linked normal, secondary, tertiary or cyclic carbon atoms that are fully saturated, i.e., linear, branched, cyclic and/or one or more unsaturated alkyl moieties as described below for alkenyl and aryl moieties. When an alkynyl moiety is specified as a substituent of an organic moiety, such as an androst-5-ene, androst-4-ene, androst-1 -ene, androst-1 , 5- diene androst-1 ,4-triene, 1 ,3,5(10)-estratriene, 5a-androstane, 5a-androstan-1 -ene, 5β- androstane or 5p-androstan-1 -ene moiety, a non-aromatic sp carbon of that substituent is covalently attached by a single bond to that organic moiety. Thus, an alkynyl substituent is comprised of an ethynyl moiety (i.e., -C≡C-) that is single bonded to that structure through an unsaturated carbon atom of the ethynyl moiety. The number of carbon atoms in an alkynyl moiety or substituent is typically 2 to about 10. C ₂ . ₆ alkynyl or C2-6 alkynyl means an alkynyl moiety containing 2, 3, 4, 5 or 6 carbon atoms and is inclusive of the carbons of the ethynyl moiety that defines it as an alkynyl substituent. When an alkynyl moiety or substituent is specified, species include by way of example and not limitation any of the alkyl moieties or substituents described herein that incorporates a triple bond such as -C≡CH, -C≡CCH ₃ , -C≡CCH ₂ CH ₃ , -C≡CC ₃ H ₇ or -C≡CCH ₂ C ₃ H ₇ . Typically, alkynyl moieties or substituents are selected from the group consisting of C ₂ . ₈ alkynyl, C ₂ . ₆ alkynyl, C ₂ . ₄ alkynyl, more typically ethynyl, 1 -propynyl and 1 -butynyl, substituted or unsubstituted (i.e., optionally substituted).

[124] "Aryl" as used herein refers to an aromatic ring system or a fused ring system with no ring heteroatoms (i.e., the ring(s) that are composed of only carbon atoms) comprising 1 , 2, 3 or 4 to 6 fused rings, typically 1 to 3 fused rings and is characterized by a cyclically conjugated system of 4n+2 electrons (Huckel rule), typically 6, 10 or 14 electrons some of which may additionally participate in exocyclic conjugation (cross-conjugated). When aryl Docket No. 354. PATENT moiety is used as a substituent to a structure the aryl moiety is attached to the structure through an aromatic carbon of the aryl group. When an aryl moiety or substituent is specified, species include by way of example and not limitation phenyl, naphthyl, phenanthryl and quinone. Typically an aryl moiety or substituent is phenyl, substituted or unsubstituted (i.e., optionally substituted). Typically, aryls are unsubstituted phenyl or phenyl substituted with one or more, typically 1 or 2 independently selected halogen, d- ₄ alkyl, C ₂ - ₄ alkenyl, C ₂ - ₄ alkynyl or monovalent oxygen-bound substituents as defined herein such as -OH, ether or ester.

[125] "Heteroaryl" as used here refers to an aryl ring system wherein one or more, typically 1 , 2 or 3, but not all of the aromatic carbon atoms comprising the aryl ring system are replaced independently by a heteroatom, which is a heavy atom other than carbon, , typically, oxygen, nitrogen or sulfur as -0-, -S-, =N- or -N(X)-, wherein X is -H or a protecting group or another organic moiety, including an organic moiety as described herein (e.g., alkyl, alkenyl, alkynyl, and aryl). In heteroaryls the heteroatom(s) of the heteroaryl participates in a conjugated system either through pi-bonding with an adjacent atom in the ring system or through a lone pair of electrons on the heteroatom. The heteroaryl ring system may be optionally substituted on one or more its carbon atoms or heteroatoms, or a combination of both, in a manner which retains the cyclically conjugated system.

[126] By way of example and not limitation, mono cyclic carbon-bonded heteroaryls are bonded at position 2, 3 or 4 of a pyridine, position 3, 4, or 5 of a pyridazine, position 2, 4, 5, or 6 of a pyrimidine, position 2, 3, 5, or 6 of a pyrazine, position 2 or 3 of a furan, thiophene or pyrrole, position 2, 4, or 5 of an oxazole, imidazole or thiazole, position 3, 4, or 5 of an isoxazole, pyrazole, or isothiazole By way of example and not limitation, bicyclic fused carbon-bonded heterocycles are bonded position 2, 3, 4, 5, 6, 7, or 8 of a quinoline, position 1 , 3, 4, 5, 6, 7, or 8 of an isoquinoline, position 2, 3, 4, 5, 6, or 7 of indole, position 2, 3, 4, 5, or 6 of 7-aza-indole, position 2, 4, 5, 6 or 7 of benzimidazole, positions 2, 3, 6, or 8 of a purine, positions 3, 4 or 6 of a pyrazolo-[3,4-d]-pyrimidine, positions 2, 3, 5, 6, or 7 of a pyrazolo[1 ,5-a]pyrimidine or positions 2, 4, 5 of 6 of a pyrrolo- [2,3-d]-pyrimidine. Typically, carbon bonded heteroaryls have one ring oxygen or one or more ring nitrogens, typically 1 or 2, and include by way of example and not limitation 2- furanyl, 2-pyrrolyl, 3-pyrrolyl, 3-pyridyl, 4-pyridyl, 3-pyridazinyl, 4-pyridazinyl, 2-pyrimidinyl,

4- pyrimidinyl, 5-pyrimidinyl, 6-pyrimidinyl, 2-pyrazinyl, 3-pyrazinyl, 2-thiazolyl, 4-thiazolyl,

5- thiazolyl, indol-2-yl, indol-3-yl, 7-aza-indol-3-yl or other nitrogen containing heteroaryls, substituted or unsubstituted (i.e., optionally substituted). Docket No. 354. PATENT

[127] A heteroaryl substituent attached to an organic moiety, such as a steroid moiety that includes by way of example and not limitation an androst-5,16-diene, androst-1 ,5,16- triene, androst-4,16-diene, androst-1 ,4,16-triene 5a-androstan-16-ene, 5p-androstan-16- ene, 5a-androstan-1 ,16-diene and an δβ-androstan-l ,16-diene moiety through a carbon atom of the heteroaryl aromatic ring system is referred to as a carbon-bonded heteroaryl or C-heteroaryl. Similarly a heteroaryl attached to those organic moieties through a nitrogen atom of the heteroaryl aromatic ring system is referred to as a nitrogen-bonded heteroaryl or N-heteroaryl.

[128] Optionally substituted alkyl", "optionally substituted alkenyl", "optionally substituted alkynyl", "optionally substituted heterocycle", "optionally substituted aryl", "optionally substituted heteroaryl" and the like mean an alkyl, cycloalkyl, alkenyl, alkynyl, aryl, heteroaryl, heterocycle or other group or moiety as defined or disclosed herein that has a substituent(s) that optionally replaces a hydrogen atom(s) in the group or moiety. Such substituents are as described above.

[129] Optionally substituted phenyl moieties include unsubstituted phenyl, Ph-N0 ₂ , and Ph-(halogen) _{1 2} or 3, wherein halogen independently selected is -F, -CI, -Br or -I, typically - F, -CI or -Br, more typically -F or -CI, and Ph-(optionally substituted alkyl) _{2 or} 3, wherein optionally substituted alkyl independently selected is typically optionally substituted d- ₆ alkyl, more typically methyl, ethyl or isopropyl.

[130] Optionally substituted heteroaryl moieties include unsubstituted d- ₆ heteroaryl (HetAr), HetAr-N0 ₂ , and HetAr(halogen) _{1 2 0} r 3, wherein halogen independently selected is -F, -CI, -Br or -I, typically -F, -CI or -Br, more typically -F or -CI, -Ph-(optionally substituted alkyl) _! 2 _or 3, wherein optionally substituted alkyl independently selected is typically optionally substituted d- ₆ alkyl, more typically methyl, ethyl or isopropyl.

[131] Optionally substituted phenyl and HetAr moieties further include Ph(0-linked) _{1 2 0} r 3 and HetAR (0-linked) _{1 2 or} 3, wherein the O-linked moiety independently selected is monovalent, typically -OH, ester, ether or divalent, typically -0-(optionally substituted d- ₆ alkyl)-0-. More typically the monovalent and divalent O-linked substituents are independently selected from the group consisting of -OH, -OC(0)CH ₃ , -OC(0)CH ₂ CH ₃ , - OCH ₃ , -OCH ₂ CH ₃ , -OCH ₂ 0-, -OCH ₂ CH ₂ 0-. A substituted phenyl or HetAr having the divalent -0-(optionally substituted d- ₆ alkyl)-0- substituent is considered disubstituted with the substituents open valent oxygen atoms bonded to two adjacent carbons of the aromatic ring and therefore is included within Ph(0-Iinked) ₂ and HetAR(0-linked) ₂ . For Ph(0-Iinked) ₃ and HetAR(0-linked) ₃ all three O-linked substituents independently selected Docket No. 354. PATENT may be monovalent or may contain a divalent O-linked substituent and one monovalent O-linked substituent.

[132] Other optionally substituted phenyl and HetAr moieties have multiple substitutions, typically totaling 2 to 4, more typically 2-3 that are combinations of those previously described and therefore include Ph(0-linked) ₁ . ₂ (halogen), Ph(0-linked)(halogen) ₂ , Ph(0- linked)i _-2 (Ci-8 alkyl), Ph(0-linked)(C ₁ - ₆ alkyl) ₂ , Ph(halogen) _1-2 (Ci-8 alkyl), Ph(halogen)(d- ₆ alkyl) ₂ , Ph(0-linked)i _-2 (Ci-8 alkyl)(halogen), Ph(0-linked)(C ₁ - ₆ alkyl) ₂ (halogen), Ph(0- linked)(d- ₆ alkyl) ₂ (halogen), Ph(0-linked)(d- ₆ alkyl)(halogen) ₂ , HetAriO-linked) _! - ₂ (halogen), HetAr (0-linked)(halogen) ₂ , HetAr (0-linked) ₁ - ₂ (C ₁ - ₆ alkyl), HetAr (0-linked)(d- 6 alkyl) ₂ , HetAr (halogen) ₁ . ₂ (C ₁ - ₆ alkyl), HetAr (halogen)(d- ₆ alkyl) ₂ , HetAr (0-linked) ₁ - ₂ (C ₁ - 6 alkyl)(halogen), HetAr (0-linked)(d- ₆ alkyl) ₂ (halogen), HetAr (0-linked)(d- ₆ alkyl) ₂ (halogen) and Ph(0-linked)(d- ₆ alkyl)(halogen) ₂ , wherein the O-linked, d_ ₆ alkyl and halogen substituents are independently selected and are typically -OH, -OC(0)CH ₃ , - OC(0)CH ₂ CH ₃ ,-OCH ₃ , -OCH ₂ CH ₃ , methyl, ethyl, -F or -CI.

[133] Optionally substituted alkyl includes unsubstituted d_ ₆ alkyl, -CH ₂ Ph, -CF ₃ , - CH ₂ OH, -CH ₂ -halogen, wherein -halogen is -F, -Br, -CI or -I, typically -I or -Br, and optionally substituted alkynyl includes -C≡CCH ₂ OH, -C≡C-halogen, typically C≡C-CI or - C≡C-Si(R ¹³ ) ₃ , with R ¹³ as previously defined for silyl ether. More typically an optionally substituted alkynyl is -C≡C-Si(CH ₃ ) ₃ or -C≡C-Si(t-Bu)(CH ₃ ) ₂ .

[134] "Oxygen-bonded moiety, O-linked moiety" and like terms as used herein refers to an oxygen-containing moiety that is attached to an organic moiety, such as steroid moiety including by way of example and not limitation an androst-5,16-diene, androst-1 ,5,16- triene, androst-4,16-diene, androst-1 ,4,16-triene 5a-androstan-16-ene, 5p-androstan-16- ene, 5a-androstan-1 ,16-diene, δβ-androstan-l ,16-diene, androst-5-en-17-one, androst- 1 ,5-dien-17-one, androst-1 , 4-dien-17-one, 5a-androstan-17-one, 5p-androstan-17-one, 5a-androstan-1 -en-17-one or an 5p-androstan-1 -ene-17-one moiety, directly though an oxygen atom of the oxygen-containing moiety such that the bonded oxygen is divalent and the oxygen-containing moiety is either monovalently or divalently bonded to the organic moiety. Thus, an oxygen-bonded moiety may be a monovalent O-linked moiety and include by way of example and not limitation moieties such as -OH, -OP ^PR , wherein PR is a protecting group as defined herein, an ester, such as acetoxy, i.e., -0-C(0)-CH ₃ , acyloxy, i.e., -0-C(0)-R ¹² , wherein R ¹² is -H (i.e., formyl ester), optionally substituted alkyl, optionally substituted alkenyl, optionally substituted alkynyl, optionally substituted aryl or optionally substituted heteroaryl (including optionally substituted heterocycle). Monovalent oxygen-bonded moieties further include ether and silyl ether moieties such as Docket No. 354. PATENT alkyloxy (Alkyl-O-), aryloxy (Aryl-O-), phenoxy (Ph-O-), benzyloxy (Bn-O-), heteroaryloxy (Het-O-) and silyloxy, i.e., R ¹¹ 0-, wherein R ¹¹ is optionally substituted alkyl, aryl, phenyl, benzyl (i.e., -CH ₂ Ph), heteroaryl or silyl, i.e., (R ¹³ ) ₃ Si-, wherein R ¹³ independently are alkyl, aryl or heteroaryl, optionally substituted. Other monovalent oxygen-linked moieties are carbamates having the structure -0-C(0)N(R ¹⁴ ) ₂ , wherein R ¹⁴ independently are -H, optionally substituted alkyl, optionally substituted alkenyl, optionally substituted alkynyl or another monovalent carbon-bonded moiety, or carbonates having the structure -O- C(0)OR ¹⁵ wherein R ¹⁵ is optionally substituted alkyl or another monovalent C-linked moiety, and -OR ^PR , wherein R ^PR is a protecting group as previously defined, or an O- linked moiety may be divalent, i.e., =0 or -OCH ₂ CH ₂ 0-.

[135] Typically, monovalent O-bonded moieties are -OH, esters that have the structure - 0-C(0)-R ¹² or silyl ethers that have the structure (R ¹³ ) ₃ SiO-. For typical oxygen-bonded esters, R ¹² is d- ₆ alkyl or is -CH ₃ (i.e., acetate), -CH ₂ CH ₃ (i.e., propionate), -Ph (i.e., benzoate), -CH ₂ Ph (phenylacetate) and 4-nitrophenyl (i.e., p-nitrobenzoate) with -CH ₃ especially preferred. For typical oxygen-bonded silyl ethers (i.e., silyloxy moieties), R ¹³ independently are d- ₆ alkyl or aryl including -CH ₃ , -CH ₂ CH _3i f-butyl or -Ph. More typical oxygen-bonded silyl ethers are trimethylsilyloxy and f-butyldimethylsilyl-oxy moieties.

[136] "Carbon-bonded moiety", "C-linked moiety" and like terms as used herein refers to a moiety or substituent that is attached to another organic moiety, such a steroid moiety, including by way of example and not limitation an androst-5,16-diene, androst-1 ,5,16- triene, androst-4,16-diene, androst-1 ,4,16-triene 5a-androstan-16-ene, 5p-androstan-16- ene, 5a-androstan-1 ,16-diene, δβ-androstan-l ,16-diene, androst-5-en-17-one, androst- 1 ,5-dien-17-one, androst-1 , 4-dien-17-one, 5a-androstan-17-one, 5p-androstan-17-one, 5a-androstan-1 -en-17-one or an 5p-androstan-1 -ene-17-one moiety, directly though a carbon atom of the carbon-bonded moiety or substituent.

[137] An C-bonded moiety include groups such as acyl, i.e., -C(0)-R ¹² , wherein R is -H, optionally substituted alkyl, optionally substituted alkenyl, optionally substituted alkynyl, optionally substituted aryl, optionally substituted heteroaryl or optionally substituted C- heterocycle or carboxylate, i.e., -C(0)-OR ¹² , wherein R ¹² is -H or its corresponding salt, - C(0)-0 ^" , or is as previously defined for ester as for non-limiting examples where R ¹² is alkyl, aryl, a C-bonded heteroaryl or a C-bonded heterocycle. C-bonded moieties typically are d- ₄ alkyl, C ₂ . ₄ alkenyl or C ₂ . ₄ alkynyl or more typically are selected from the group consisting of -CH ₃ , -CH ₂ =CH ₂ , and -C≡CH.

[138] "Protecting group" as used here means a moiety that prevents or inhibits the atom or functional group to which it is linked from participating in unwanted reactions. For Docket No. 354. PATENT example, for -OR , R is a protecting group for the oxygen atom found in a hydroxyl, while for =0 (ketone), the protecting group is a ketal or thioketal or the protecting group is an oxime wherein =0 is replaced by =N-OR ¹¹ , wherein R ¹¹ is -as defined for ether or silyl ether. Preferred R ¹¹ for oximes moieties are -H, alkyl or -Si(R ¹³ ) ₃ , with R ¹³ as defined for silyl ether. Ketals include cyclic ketals that contain structures such as -0-C(R ¹⁶ ) ₂ - C(R ¹⁶ ) ₂ 0-, wherein R ¹⁶ independently selected are -H or alkyl. For -C(0)-OR ^PR , R ^PR is a carbonyloxy protecting group, for -SR ^PR , R ^PR is a protecting group for sulfur in thiols, for instance, and for -NHR ^PR or -N(R ^PR ) ₂ -, R ^PR independently selected is a nitrogen atom protecting group for primary or secondary amines. The protecting groups for sulfur or nitrogen are usually used to avoid unwanted reactions with electrophilic compounds. The protecting groups for oxygen are used to avoid unwanted reactions with electrophiles, and are typically esters (e.g. acetate, propionate or benzoate), or avoid interfering with the nucleophilicity of organometallic reagents or other highly basic reagents, and are typically ethers, optionally substituted, including alkyl ethers, (e.g., methyl or tetrahydropyranyl ethers) alkoxymethyl ethers (e.g., methoxymethyl or ethoxy-methyl esters), optionally substituted aryl ethers and silyl ethers (e.g. trimethylsilyl (TMS), triethylsilyl (TES), tert- butyldiphenylsilyl (TBDPS), tert-butyldimethylsilyl (TBS/TBDMS), triisopropylsilyl (TIPS) and [2-(trimethylsilyl)ethoxy]methylsilyl (SEM)).

[139] "Halogen" means fluorine, chlorine, bromine or iodine.

[140] 2. Biological Activity

[141] 2(a) Selective Modulation of MAPK Activity: Erk-1 and Erk-2 have opposing roles in mediating signaling through Ras referred herein as Ras-Erk signaling. Ras is a small G protein with intrinsic GTPase activity that associates with the cell membrane. When in the activated GTP-bound state Ras recruits Raf, which is a MAP3K, by its interaction with the GDP/GTP exchange factor son-of sevenless (SOS). That exchange factor is brought to the cell membrane by growth-factor-receptor-bound protein 2 (Grb-2) adapter protein, which recognizes phosphorylated receptor tyrosine kinase (pRTKs) docking sites. In a feedback inhibitory mechanism, activated Erk phosphorylates SOS to disrupt the cell membrane-based complex that had led to its activation through Ras signaling. Examples of opposing isoform roles include overexpression of ERK1 by not ERK2 gene that was found to inhibit Ras-dependent cell growth, which is an effect independent of its phosphorylation status of the MAPK. Furthermore, Erk-1 deficiency results in the enhanced stimulus-dependent activation of Erk-2 without a compensatory increase in Erk- 2 protein levels, with an opposite effect on proliferation observed with ERK2 knockdown.

[142] Dimerization of Erk-1 and Erk-2, which may effect their translocation to the nucleus Docket No. 354. PATENT when activated, may provide another level of complexity to their regulation and substrate specificity. Only phospho-Erk has been found to dimerize and Erk-1 -Erk-2 heterodimers are considered unstable. Thus, compounds that disrupt Erk-1 homodimerization permit interactions of Erk-1 with scaffolding proteins for retention in the nucleus whereas compounds that stabilize Erk-2 homodimerization are expected, based upon the insights provided by the invention disclosed herein, to inhibit interaction with scaffolding proteins that would retain activated Erk-2 in the cytoplasm. Test compounds or candidate exhibiting such effects are expected, based upon the insights provided by the invention disclosed herein, to decrease survival and proliferation and be useful in treating the hyperproliferation conditions described herein.

[143] Erk-1 and Erk-2 when inactive are retained in the cytoplasm in heterodimeric complexes with MEK-1 or MEK-2. Phosphorylation of the MEK isoforms by upstream kinases triggers phosphorylation of its Erk partner, which then dissociates from the heterodimer complex. However, in MEK-1 -Erk-2 heterodimers, coupling between these proteins is enhanced when a serine residue (Ser-298) in the proline-rich regulatory sequence in MEK-1 is phosphorylated by PAK-1 or Rac (which are protein kinases activated upon integrin-mediated adhesion), thus retaining activated Erk-2 in the cytoplasm. Additional phosphorylation at a threonine residue (Thr-292) by the MAP3K Raf results in dissociation. The activated Erk-2 released from the MEK-1 -Erk-2 heterodimer then enters the nucleus through facilitative transport as a homodimer or by passive diffusion as a monomer. Translocation to the nucleus of activated Erk-2 is therefore dependent on MEK-1 activation and requires Ser-298 and Thr-292 phosphorylation.

[144] Thus, due to the tighter binding between MEK-1 and Erk-2 in adherent cells mediated by Ser-298 phosphorylation in comparison to MEK-2-Erk-2 binding, which remains constant in adherent and suspended cells, MEK-1 preferentially sequesters unactivated Erk-2 in the cytoplasm in comparison to MEK-2 and results in translocation of activated Erk-2 to the nucleus on T292 phosphorylation by Raf. In contrast, Erk-2 activated by MEK-2 after growth factor stimulation is preferentially retained in the cytoplasm and increases survival through p90 phosphorylation. Therefore, Erk-2 intracellular localization sorted by the MEK isoforms determines whether growth factor stimulation mediates survival or adhesion dependent proliferation.

[145] Differences in nucleo-cytoplasmic shuttling between Erk-1 and Erk-2 mediated by the MEK isoforms may also be caused by the unique NH ₂ -terminal domain of Erk-1 since loss of that domain has been found to ablate the inhibitory effect of Erk-1 on Ras- dependent proliferation. The slower shuttling of Erk-1 through the nuclear membrane in Docket No. 354. PATENT comparison to Erk-2 prolongs the presence of Erk-1 in the nucleus, which is the primary location for Erk inactivation by phosphatase dephosphorylation, while favoring nuclear transport of Erk-2. The functional consequence of these effects is segregation of the catalytic activity of Erk-1 to the cytoplasm and transcription-dependent outcomes in the nucleus to be primarily mediated by Erk-2.

[146] Thus, the signaling output from Ras-Erk pathway may be dependent on the relative concentrations or activities of Erk-1 and Erk-2. Test or candidate compounds that increase signaling activity through Erk-1 in comparison to Erk-2 are therefore expected, based upon the insights provided by the invention disclosed herein, to have anti- proliferative properties.

[147] In addition, compounds that disrupt the interaction between Erk-2 and MEK-1 , which is mediated by kinases, whose activity depends upon integrin adhesion, are expected, based upon the insights provided by the invention disclosed herein, to provide an anti-proliferative effect by preventing preferential localization of unactivated Erk-2 in the cytoplasm that would promotes its activation by growth factor stimulation and subsequent translocation of pErk-2 to the nucleus. Since Erk-1 associates with both MEK isoforms and activated Erk-1 typically acts upon cytoplasmic targets, and therefore does not require dissociation as a predicate for activity as does activated Erk-2 for activity in the nucleus, compounds that stabilize the interaction of MEK-1 or MEK-2 with Erk-1 would increase retention of pErk-1 in the cytoplasm. Furthermore, compounds that preferentially stabilize the interaction of MEK-1 with Erk-1 would allow Erk-1 to more effectively compete with Erk-2 for this MAPKK isoform, thus reducing the ability of MEK- 1 to properly sequester inactive Erk-2 in the cytoplasm for activation and subsequent translocation of pErk-2 to the nucleus upon growth factor stimulation. Therefore, test compounds that preferentially stabilize the interaction of MEK-1 with Erk-1 are expected to be useful in the treating hyperproliferation conditions described herein.

[148] Once Erk is dephosphorylated in the nucleus it is rapidly exported out by a mechanism that is mediated in part by re-association with MEK that had entered the nucleus independently. PEA-15, also known as PED, which is a death effector domain (DED)-containing protein, also associates with Erk in the nucleus and due to it nuclear export sequence also mediates relocation of this MAPK to the cytosol. Therefore, some preferred test or candidate compounds modulate the interaction of Erk with PED, for example, by stabilizing the interaction of Erk-2 with PED, thus truncating long-term signaling through Erk-2 due to its more rapid nuclear export. Such test compounds having that characteristic for selection as a candidate compound are thus expected to be useful in the treating hyperproliferation conditions described herein. Docket No. 354. PATENT

[149] MAPKs are also known to associate with MAPK phosphatases (MKPs), also known as dual-substrate phosphatases (DUSPs), since they hydrolyze tyrosine and Ser/Thr phosphate residues, and are mediated by at least two protein-protein interaction sites. The nine known MKP isoforms are unrelated to the Ser/Thr phosphatases but belong to the superfamily of protein tyrosine phosphatases (PTPase). One interaction site on MAPKs recognizing MKPs is the common docking (CD) domain, previously discussed, and another site is a domain referred to as the FXFP or F-domain (FD), so named for the conserved amino acid sequence in this MAPK binding domain. Together the CD and FD domains among other potential protein-protein interaction sites are thought to confer selectivity to MAPK protein associations, including their interactions with MKPs, MNKs and MAPKAPKs, and to confer specificity to substrates having cognate binding sites.

[150] MKP-3 is predominately located in the cytoplasm due to the presence of a putative nuclear export signal, which may, along with CD, FD and other potential protein- protein binding domains, contribute to Erk binding, and are highly specific for deactivating Erk1/2. Although the CD binding is thought more important to binding affinity compared to the other domains it is postulated to be not essential to the Erk- induced activation of MKP-3. However, after phosphorylation of the putative Tyr and Thr activation residues of Erk, productive interaction of activated Erk with MPK-3 is believed to occur. This interaction is expected, based upon the insights provided by the invention disclosed herein, to activate MKP-3 phosphatase activity to efficiently deactivate Erk. Other phosphatases of appropriate specificity to be physiologically relevant regulators of Erk include the tyrosine phosphatase HePTP and the Ser/Thr phosphatase PP2A. Therefore, the strength and duration of the Erk-2 signal from MEK-1 activation is postulated to be balanced by the opposing action of MEK-1 and multiple protein phosphatases that includes single specificity and dual specificity phosphatase and are expected, based upon the insights provided by the invention disclosed herein, to affect the time course and threshold for Erk-2 reactivation. Therefore, test compounds that preferentially stabilizes the interaction of Erk-2 with MKP-3 to preferential retain that isoform in an unphosphorylated or deactivated state MEK-1 are expected to be useful in the treating hyperproliferation conditions described herein.

[151] It has been unexpectedly found that 17a-ethynyl-5a-androstane-3a,17p-diol, referred herein as E-3a-diol, primarily affects Erk MAPK signaling in comparison to p38 or JNK MAPK signaling. Additionally, it has been unexpectedly found that E-3a-diol selectively affects the phosphorylation status of Erk-1 in comparison to the isoform Erk-2. E-3a-diol also unexpectedly exerts different localization effects on the two isoforms that are believed to be mediated by scaffold-adaptor proteins. Thus, preferred candidate Docket No. 354. PATENT compounds positively modulate the phosphorylation state of Erk-1 or negatively modulate the phosphorylation state of Erk-2 in comparison to the other isoform. Other preferred compounds are negative modulators of nucleo-cytoplasmic translocation. Typically, phosphorylation states of the Erk isoforms or nucleo-cytoplasmic translocation of total pErk or for an isoform thereof (i.e., pErk-1 or pErk-2) is evaluated after contacting a suitable cell-based test system, preferably an in vitro cell-based test system, with a test compound as compared to a sham contacted cells in an identical test system. Erk phosphorylation states or pErk nucleo-cytoplasmic transport may also be evaluated relative to that observed when test cells are contacted with a positive control for negative modulation of nucleo-cytoplasmic transport (e.g., E-3a-diol) or when test cells are contacted with a positive control for negative modulation of nucleo-cytoplasmic transport (e.g., 3a-diol) or positive modulation of Erk-2 phosphorylation state.

[152] Based on these discoveries and other insights described herein, the invention provides methods to identify candidate compounds having desirable biological activities, e.g., as anti-cancer drugs or anti-inflammation drugs. Characteristics of preferred candidate compounds include properties such as (i) a capacity of differentially regulate Erk-1 compared to Erk-2, (ii) a capacity to promote or increase apoptosis, e.g., by re- regulating dysregulated AR signaling, (iii) a relatively low toxicity in a mammal, e.g., having a therapeutic index of at least two, more preferably at least about five, most preferably at least about 10 and/or (iv) other biological properties or activities described herein.

[153] Preferred test compounds have a molecular weight of 200-1 ,000 amu or 200-800 amu, and are non-peptidic, but may be peptidic or have higher molecular weight as, for example, when a test compound is used as a reference compound that selectively elicits extracellular effect(s) by binding to a plasma membrane receptor (e.g., a GPCR) in comparison to intracellular effects (e.g., those resulting from binding to protein complexes or scaffold proteins in the cytoplasm or nucleus). Such reference compounds are typically cell-impermeable and include androgen-protein conjugates (e.g., testosterone-bovine serum albumin (T-BSA) or DHT-BSA conjugate), optionally coupled to a fluorophore, as described herein. Other peptidic substrates recognizes the CD or FD protein-protein interaction domain of a MAPK and may be derived from the amino acid sequence of the corresponding protein-protein interaction domain of a substrate or effector protein for that MAPK.

[154] Preferably, a test compound is initially screened for a biological response (e.g., modulation of a phospho-protein's activity or phosphorylation state as described herein) Docket No. 354. PATENT by contacting the test compound with a suitable test system in a concentration range of about 100 μΜ to 0.001 μΜ, preferably in a range including, e.g., about 25 μΜ, 10 μΜ or 1 μΜ concentration (final concentration within the test system). Preferred initial suitable test systems are in vitro test systems that screen test compounds for one or more biological responses qualitatively similar to an activity elicited by E-3a-diol when that reference compound is contacted with a control test system. Other preferred initial suitable test systems are in vitro test systems that screen for test compounds that oppose one or more biological responses elicited by 3a-diol when that reference compound is contacted to those same test systems. A test compound for consideration as a candidate compound that provides a desired biological response within the concentration range of the initial screen provides an EC ₅₀ of 0.1 μΜ or less, more preferably between about 0.1 μΜ to 0.001 μΜ (i.e., 100 nM to 1 nM) or less or about 0.005 μΜ (i.e., 5 nM) or less in a subsequent screen for that biological response (e.g., negative modulation of PI3K p85 regulatory subunit tyrosine phosphorylation state or positive modulation of Erk-1 activation loop phosphorylation state in comparison to that of Erk-2).

[155] Other preferred candidate compounds are test compounds that have the aforementioned Erk-1 and/or Erk-2 effects and additionally have negligible effect on one or more isoforms of p38 or JNK MAPK. Such compounds have a negligible effect on the phosphorylation capacity or rate of pErk-1 or pErk-2 towards an Erk-1/2 substrate or effector protein in a suitable cell-free in vitro test system as described herein. More preferred candidate compounds have negligible binding to the ATP binding site of Erk-1 or Erk-2. Other more preferred candidate compounds are test compounds that exhibit efficacy in treating chemically induced or xenograft implanted tumors and have a therapeutic index of 2 or more when tested in relevant animal model(s). Particularly preferred candidate compounds include test compounds that exhibit efficacy against cancer or hyperproliferating cells and have low liver toxicity or are characterized by a therapeutic index of 10 or more as when contacted with cancer or hyperproliferating cells in a suitable animal model. More preferred are candidate compounds that are efficacious (e.g., demonstrates statistically significant reduction in tumor burden, progression or incidence) when administered at about 0.1 mg/kg to 350 mg/Kg to an animal suffering from a hyperproliferation condition or about 0.01 mg/Kg to 25 mg/Kg in a human. The preferred candidate compounds may be further evaluated for their ability to induce one or more other effects observed E-3a-diol (i.e.,. in addition to effects of E-3a-diol on Erk-1/2 and PI3K phosphorylation states) as described herein.

[156] Other preferred candidate or test compounds will mediate sequesterization of Erk Docket No. 354. PATENT or pErk in the cytoplasm so as to positively modulate Erk-1 phosphorylation or favor cytosolic retention of pErk-1 (i.e., negatively modulate pErk-1 nucleo-cytoplasmic translocation) . More preferred compounds will mediate subcellular localization of Erk-2 so as to negatively modulate this isoform's phosphorylation state. Other more preferred compounds will mediate subcellular localization of cytoplasmic Erk-1 or pErk-1 so to positively modulate this isoform's cytosolic activity (e.g., by promoting Erk-1 phosphorylation or inhibiting pErk-1 nucleo-cytoplasmic translocation) with minimal or no activation of anti-apoptotic proteins. Other more preferred compounds will mediate subcellular localization of cytoplasmic Erk-1 or pErk-1 so to positively modulate this isoform's cytosolic activity (i.e., by promoting Erk-1 phosphorylation or inhibiting p-Erk-1 dephosphorylation) with minimal or no activation of anti-apoptotic proteins. In some instances a preferred candidate or test compound will localize pErk-1 to an endomembrane other than that of mitochondria, with localization to endosomes more preferred.

[157] Other preferred candidate or test compounds modulate spatial or temporal regulation of signaling from the signal transduction nodes Erk-1 and/or Erk-2 that are expected, based upon the insights provided by the invention disclosed herein, to have anti-proliferative or anti-inflammatory effects and may result from candidate or test compounds that (i) promote stabilization of Erk-2 containing protein complexes through sequesterization of one or more of its component(s) in the cytoplasm to prevent Erk-2 signal transmission, (ii) prevent activated Erk-2 from translocating to the nucleus (i.e., negatively modulating pErk-2 nucleo-cytoplasmic translocation), (iii) retaining deactivated Erk-2 in the nucleus (i.e., positively modulating Erk-2 nucleo-cytoplasmic translocation), (iv) promote stabilization or retention of protein complexes containing unphosphorylated Erk-2 in a subcellular compartment less accessible to its activating MAPKK, (v) promote stabilization of Erk-1 containing protein complexes to more efficiently couple kinase cascades involving that isoform to a cytoplasmic compartment favorable to activation of pro-apoptotic proteins, (vi) stabilize interactions between activated Erk-2 and phosphatases to negate long-term signaling of pErk-2, (vii) destabilize protein-protein interactions to interfere with assembly of signaling components that promote signaling through Erk-2, or (viii) some combination of ( i) -(vii) events. Such compounds are expected, based upon the insights provided by the invention disclosed herein, to have anti-proliferative or anti-inflammatory effects and will be useful in treating hyperproliferative and unwanted inflammation conditions that results in or propagates these conditions.

[158] Additional preferred candidate or test compounds mediate cytosolic Docket No. 354. PATENT sequesterization of agonist-activated i-AR (actAR) by negatively modulating nucleo- cytoplasmic translocation of actAR That effect may be mediated by alterations in the phosphorylation pattern in this nuclear hormone receptor by a compound's modulation of Ras-Erk signal transduction, to the cytosol such that pro-proliferative signaling from non- genomic effects as described herein by cytosolic actAR (e.g., from scaffolding of i-AR to RTKs) are not engaged. Downstream outcomes from this negative nucleo-cytoplasmic transport of actAR, includes re-regulation of AR signaling whereby differentiation (and hence apoptosis) is promoted, anti-apoptotic gene transactivation is negatively modulated, pro-apoptotic gene transactivation is positively modulated or some combination of these effects.

[159] Scaffold or adaptor proteins effect the spatial regulation of Erk1/2 activation by directing the activity of Erk-1 and Erk-2 to the appropriate subcellular compartments. Furthermore, the effect of scaffold or adaptor proteins are temporally-dependent with respect to concurrent signaling through mitogenic and pro-inflammatory signaling cascades due to assembly of protein complexes with one or more components in common with complexes that are assembled during Ras-Erk signaling. Thus, the assembly of a multi-component complex will sequester the components within it away from other signaling pathways. Therefore, scaffold proteins will participate in specifically activating some signaling cascade while inhibiting others. Some preferred candidate compounds affect protein complex assembly and localization of these complexes by modulating KSR- mediated scaffolding such that Erk-2 activity or phosphorylation state is negatively modulated in comparison to that of Erk-1 or by modulating the behavior of another or one or more other Erk- interacting scaffolding and adaptor proteins in the manner indicated below.

[160] RKIP interferes with assembly of a competent Ras-Raf-MEK-Erk protein complex by acting as a competitor of MEK binding to Erk. Thus, RKIP negatively modulates Erk activity through negative phosphorylation state modulation of MEK. Therefore, based upon the insights of the present invention, a candidate or test compound that recapitulates or enhances the competitive binding behavior of RKIP may temper long-term mitogenic signaling through RTKs mediated by Erk to avoid triggering an undesired gene transcription program while interfering with pro-inflammatory cascades mediated through Raf-1 and NF-kB. There are no known non-peptidic ligands to RKIP that have been shown to modulate Raf-1 or NF-kB signaling. Such compounds identified by the screening methods described herein should be useful in treating cancers such as breast and prostate cancers and the underlying inflammation that supports their metastatic potential. Docket No. 354. PATENT

[161 ] In addition, a candidate compound that affects IQGAP1 scaffolding of Erk will disrupt cytoskeletal dynamics, which will adversely affect proliferating cells in comparison to quiescent cells. Such compounds will be useful in treating certain cancers, including those dependent on Erb-signaling, and screening methods described herein should be useful in their identification. Therefore, some preferred candidate compounds affect IQGAP1 scaffolding involving Erk to inhibit B-Raf signaling, which should be useful in treating hyperproliferation conditions disclosed herein.

[162] Since 14-3-3 protein is involved in cross-talk between the proinflammatory pathway TNFCC-NFKB and the mitogen-activated Ras-Erk pathway and 14-3-3 δ also affects the pro-apoptotic activity of Bad, then some preferred candidate compounds negatively modulate the adaptor function of 14-3-3 protein. Such compounds should exert antiproliferative effects by re-regulating aberrant or excessive cross-talk between these two signal transduction pathways in cancer cells.

[163] Since scaffolding by Sef inhibits activated Erk translocation to the nucleus and thus promotes activity at cytoplasmic targets that are believed to be involved in differentiation or pro-apoptotic signal transduction pathways, then some preferred candidate compounds positively modulate Sef scaffolding of Erk-1/2. Such compounds should exert antiproliferative effects by restoring or re-regulating apoptotic signaling. Furthermore, compounds that stabilize Sef scaffolding of Erk-1 are expected, based upon the insights provided by the invention disclosed herein, to recover, in part, the desired activity of cytoplasmic Erk-1 (i.e., to support differentiation rather than inhibit apoptosis) upon its activation and prevent its diversion to the nucleus, where it would exert unwanted pro-proliferative activity.

[164] Since chaperone proteins may affect particular protein complexes involving Erk, e.g., by stabilizing Hsp-90-Erk-1 or Hsp-90-Erk-2 complexes, thereby negatively modulate nucleo-cytoplasmic transport of activated Erk (e.g., pErk-1 ) or negatively modulate the phosphorylation state of Erk-2, some preferred test or candidate compounds will positively modulate these interactions in order to direct pErk activity to cytosolic targets required for differentiation. Stabilization of Hsp-90-Erk complexes should also inhibit other signaling cascades that aberrantly or excessively cross-talk with the Ras-Erk pathway (e.g., TNFa- NFKB pathway). Such compounds should then exert concurrent pro-apoptotic and antiinflammatory effects useful for treating the hyperproliferation conditions disclosed herein.

[165] Test or candidate compounds that bind to the NH ₂ -terminal region of MAP-2C may positively modulate binding of Erk-2 to MAP-2C to form a non-productive complex that retards this isoform's phosphorylation and negatively modulates nucleo-cytoplasmic Docket No. 354. PATENT transport of pErk-2 which does form. The resulting cytoplasmic localization of inactive Erk-2 should therefore negatively modulate Erk-mediated deactivation of pro-apoptotic proteins and transactivation of pro-survival genes. Therefore, such compounds are useful in treating one or more of the hyperproliferation conditions described herein.

[166] 2(b). Membrane Androgen Receptor Signaling: Based upon the insights of the present invention it is now appreciated that the natural steroid 5a-androstane-3a-17p-diol (3a-diol) has inherent biological activity resulting from agonist activity on a membrane- bound androgen receptor (m-AR) and exerts anti-apoptotic (i.e., prosurvival) effects in cancer cells mediated at least in part through this transmembrane cell-surface receptor. Previous reports [e.g., Penning, T. (1997)] indicate that 5a-androstane-3a, 173-diol (3a- diol) has significantly lower affinity for i-AR in comparison to DHT (Kd 10 ^"6 M vs. 10 ^"11 M). Thus, 3a-diol was thought to lack intrinsic activity of its own and only serves as a reservoir for the classical androgen [Bauman, D.R. et al. (2006)] despite earlier indications to the contrary [Schultz, F.M. and Wilson, J.D. et al. (1974); Walsh, P.C. and Wilson, J.D. et al. (1976); Ahmad, N. et al. (1978); Mahendroo, M.S. et al. (1999); Shaw, G. et al. (2000); Leihy, M.W. et al. (2001 )]. More recently the ability of 3a-diol to induce cytosolic (i.e., nongenomic) signaling has been recognized [Yang, Y.J. et al. (2008); Dozmorov, M.G. et al. (2007); Nunlist, K.H. et al. (2004); Zimmerman, R.A. et al. (2004)], but that has been attributed strictly to its interaction with sex hormone-binding globulin [Ding, V.D.H. et al. (1998); Nakla, A.M. et al. (1995); Nakla, A.M. et al. (1990)].

[167] It is believed from the insights of the invention that anti-apoptotic signaling from m- AR by 3a-diol is the result of Ga/i activation and its release from this GPRC whereas pro- apoptotic signaling from that same GPCR by E-3a-diol results from activation and release of Galpha/q. Therefore, E-3a-diol is a biased GPCR ligand due to its engagement of an alternative signal transduction pathway.

[168] Other ligands interacting at m-AR are testosterone-BSA (T-BSA) and DHT-BSA conjugates, which are cell-impermeable constructs and therefore are unable to engage intracellular AR (i-AR). Conjugation in those ligands is through a linker to the C-3 position of the steroid, which is a position that is usually considered critical for i-AR activity. The androgen-BSA conjugates may also contain covalently attached fluorophores (to provide T-BSA-fluorophore conjugates). The effect of the cell-impermeable conjugates on LNCaP mirrors some of the activities observed with E-3a-diol. For example, T-BSA cell conjugates contacted with LNCaP cells can induce Erk phosphorylation [Wang, Z. et al. (2008)], which is an effect sensitive to the MEK-1 inhibitor PD098059, inhibit cell proliferation [Hatzoglou, A. et al. (2005)] with an IC ₅₀ of 5 nM and induce PSA secretion Docket No. 354. PATENT

[Kampa, M. et al. (2002)].

[169] However, the effects of androgen-BSA conjugates on downstream signal transduction appear to be highly context-dependent and are sometimes more like that of 3a-diol. For example, 3a-diol increases Akt signal transduction [(Nunlist, E.A. et al. (2004); Zimmerman, R.A. et al. (2004); Yang, Q. et al. (2008); Dozmorov, M.G. et al. (2007)] and T-BSA increases PI3K activity in LNCaP [Kampa, M. et al. (2003); Papakonstanti, E.A. et al. (2003)) , MCF-7 (Kallergi, G. et al. (2007)) and T47D (Pelekanu, V.M. et al. (2010)] cells. However, in DU-145 it has been reported that PI3K is not involved in nongenomic signaling by T-BSA [Papadopoulou, N. et al (2008)]. Additionally, in normal cells (C6 glial and astrocytes) a different androgen-protein conjugate (i.e., DHT-BSA) was shown to decrease pAkt levels [Gatson, J.W. and Singh, M. (2007); Gatson, J.W. et al. (2006)]. With respect to MAPK activity, increases in which are a hallmark of GPCR signal transduction, T-BSA has been shown to increase total pErk in vascular smooth muscle and skeletal muscle cells whereas in such normal cells DHT selectively increases pErk-2, but not pErk-1 [Estrada, M. et al. (2003); Somjen, D. et al. (2004)]. Although changes in phosphorylation states or actual protein levels of the individual isoforms were not discussed, exposures of LNCaP to 3oDIOL or DHT both increase Erk-1 gene expression [Nunlist, K.H. et al. (2004); Zimmerman. R.A. et al. (2004); Dozmorov et al. (2007)]. In contrast, E-3a-diol selectively permits Erk-1 phosphorylation in LNCaP to the exclusion of Erk-2 without increasing gene expression of either isoform. As previously mentioned, in LNCaP cells T-BSA also increases total pErk levels, but no isoform specific effects were described. Finally, T-BSA is reported to have no effect on Erk phosphorylation in hippocampal and i-AR transfected PC12 cells [Nguyen, T.-V. V. (2005)] whereas DHT induces rapid (within 5 min.) preferential Erk-2 phosphorylation over Erk-1 in hippocampal cells with no effect on isoform protein levels and is neuroprotective (i.e., anti-apoptotic). That anti-apoptotic neuroprotection is mediated by ERK-dependent Rsk phosphorylation leading to phosphorylation and inactivation of Bad. In aberrantly activated optic nerve head astrocytes 3a-DIOL and the potent androgen R1881 (173-hydroxy-17a-methyl-estra-4,9,1 1 -trien-3-one) both induce rapid Erk phosphorylation, while only 3a-diol increases Akt phosphorylation [Agapova, O.A. et al. (2006)].

[170] In addition, E-3a-diol and T-BSA have opposite effects on phosphorylation of PI3K with E-3a-diol decreasing tyrosine phosphorylation of the p85 regulatory subunit whereas T-BSA increases that phosphorylation in LNCaP and MCF-7 [Papakonstanti, E.A. et al. (2003); Kampa, M. et al. (2004); Kallergi, G. et al. (2007)]. In DU-145 it is only on longer exposure to T-BSA (2h) is decreased p85 tyrosine phosphorylation observed Docket No. 354. PATENT

[Papadopoulou, N. et al. (2008)].

[171] It thus is unexpected that E-3a-diol, which is structurally unrelated to the aforementioned conjugates by virtue of its free C-3 hydroxy substituent, would promote apoptosis and thus be considered an anti-proliferative m-AR agonist and would do so by engaging a different G-protein (i.e., Galpha/i for T-BSA and Galpha/q for E-3a-diol). Thus, the pro-apoptotic effects of E-3a-diol are mediated by reduced PI3K activity through decreasing tyrosine phosphorylation of it p85 regulatory subunit and by isoform selective increases in p-Erk-1 levels without Erk-2 activation, which are unexpected given the opposite effects observed for cell-impermeable androgen-protein conjugates.

[172] Natural androgens such as T and DHT at physiological concentrations and the synthetic androgen R1881 also interact at m-AR in addition to i-AR. The combined effects from interaction at the two receptors are to promote proliferation. Thus, E-3a-diol will also oppose pro-proliferative effects of T, DHT and R1881 by direct competition of these androgens at m-AR thereby disrupting those combined effects and further to induce nongenomic signaling at m-AR contrary to that induce at m-AR by the aforementioned androgens. Thus, it is believed that contrary nongenomic signaling by E-3a-diol is due to that compound acting as a biased GPCR ligand at m-AR.

[173] 2(b)(i). Ga Protein-dependent Non-qenomic Activity: The m-AR is now identified as the G protein-coupled receptor GPR-C6a [Pi, M. et al. (2010)], which previously had been identified as an amino acid sensor [Wellendorph, P. et al. (2005)]. GPR-C6a, which shares homology with the calcium sensor receptor CaSR, is coupled to the G proteins Ga/i and Ga/q. Unlike CaSR, Ca ²⁺ binding to GPR-C6a is not required for receptor activation although it does positively modulate that activity. GPCR signaling that is dependent on activation of Ga/i, which inhibits adenylate cyclase activity, can enter the Ras-Erk signaling cascade through cross-talk with other signaling pathways that involve, for example, PI3K (phosphoinositide-3-kinase), Akt (a member of the non-specific serine/threonine kinase family, which is a downstream effector protein of PI3K), Src (a non-receptor tyrosine kinase) or transactivation of RTKs (receptor tyrosine kinases).

[174] Signaling to the Ras-Erk pathway may also be effected by Ga/q activation of certain isoforms of phospholipase C (PLC), which generate the second messengers inositol triphosphate (IP3) and diacyl glycerol (DAG), to trigger intracellular mobilization of Ca ²⁺ and to activate phosphokinase C (PKC). PKC then stimulates the Ras-Erk pathway, which typically leads to nuclear accumulation of pErk. A G protein-independent pathway dependent on β-arrestin scaffolding is responsible for slower activation of Ras-Erk signaling through recruitment of Src, which is also dependent on β-arrestin scaffolding. Docket No. 354. PATENT

Those events lead to β-arrestin scaffolding of pErk that preferentially retains that phosphoprotein in the cytoplasm. Thus, two parallel pathways exist for Ga/q coupled GPCRs to activate Erk: (i) a G protein-dependent pathway that produces a transient increase in nuclear pErk and (ii) a beta-arrestin-dependent (i.e. G protein-independent) pathway that leads to a more sustained activation of Erk that is localized to the cytosol and endosomes.

[175] PLC-β _! is also known to be effected through direct interaction with Ga/q in its activated GTP-bound state, Ga/q (GTP). Deactivation of Ga/q (GTP) occurs in a feedback inhibitory manner by the GTPase activity of activated PLC-β _! . In its inactivated state, Ga/q is associated with a sub-population of PLC-β _! as Ga/q (GDP) at the plasma membrane, and the affinity for this association is increased on Ga/q activation. That pre- association of PLC-β _! and Ga/q provides for a rapid cellular response mediated by PLC-β _! upon GPCR-Ga/q activation that does not depend on Ga/q (GTP) release from the receptor.

[176] As well as activating PLC-βι , released Ga/q(GTP) inhibits p1 10a PI3K [Ballou, L.M. et al. (2003)] by binding directly to p100a/p85 and displacing Ras(GTP) [Ballou, L.M. et al. (2006); Jimenez, C. et al. (2002); Carpenter, C.L. et al. (1993)], whereas receptors coupled to the Gi/o family of G proteins increase PI3K activity through stimulation of the isoforms p1 10β- and p1 10γ-ΡΙ3Κ. Phosphorylation of p85 Ser-608 on p85a also is inhibitory and may be effected through feedback inhibition by the intrinsic serine kinase activity of PI3K. As in the case of Ga/q and PLC-β _! , Ga/q and PI3K may also exist in pre-associated complexes, with the Ga/q-PLC-βΙ and Ga/q-PI3K protein complexes probably localized in different subcellular compartments [Golebiewska, U. et al. (2008)].

[177] The p85 regulatory subunit has an SH3 domain, a BcR homology domain (BH) flanked by proline-rich sequences and two SH2 domains separated by a domain that binds to the catalytic domain p1 10a (the inter-SH2 domain). Due to these binding domains, p85 also mediates interaction of PI3K with the deactivating phosphatase SHP- 1. The p100a domain contains a p85-interacting region and a Ras-binding domain. The p85 domain mediates translocation to activated RTKs, which induces an activating conformational change in p1 10a, and phosphorylation of p85 Tyr-688 by Src increases PI3K activity by releasing p1 10a from p85 inhibition [Kodaki, T. et al. (1994)]. Ras (GTP) binding to the p1 10a subunit further increases catalytic activity [Rodriquez-Viciana, P. et al. (1994)]. Thus, tyrosine kinase stimulation releases inhibition of p85, which permits Docket No. 354. PATENT further enhancement of PI3K catalytic kinase activity through interaction of p1 10a subunit with Ras(GTP). Those various interactions between the p85 and p100 subunits with components of other signal transduction pathways provide for cross-talk between PI3K-Akt signaling with that of Ras-Erk, Src and RTK.

[178] Released Ga/q(GTP) therefore will inhibit aberrant crosstalk between PI3K-Akt and Ras-Erk signal transduction pathways by disrupting mitogenic PI3K-Ras protein complexes. Released Galpha/q is known to interact with Src [Liu, W.W., et al. (1996); Umemori, H. et al. (1997)] as does i-AR [Kim, S.B. et al. (2007)] and will thus also out compete formation of Src/PI3K/i-AR mitogenic protein complexes that result from RTK activation [Sun, M. et al. (2003)]. As a result Galpha/q disrupts the p85-Src interaction that otherwise would lead to stimulation of the PI3K/Akt pathway through Src tyrosine phosphorylation of the p85 regulatory subunit. Thus, tyrosine phosphorylation of p85 is negatively modulated by Galpha/q activation.

[179] It is therefore believed based upon the insights from the present invention that re- regulation of apoptotic signaling from E-3a-diol acting upon GPR-C6a results from Ga/q activation and release from this GPRC, scaffolding of activated Ga/q with components of other signal transduction cascades acting in parallel, scaffolding of these other components in G protein-independent signaling subsequent to Ga/q activation or some combination of such effects. Thus, a test compound that acts upon GPR-C6a in qualitatively or quantitatively similar manner to E-3a-diol (i.e., results from activation of Ga/q, which may be considered an inverse agonist activity on Ga/i signaling) is expected to exhibit pro-apoptotic activity thereby re-regulating apoptosis. This pro-apoptotic activity may result from negative modulation of phosphorylation state of a pro-apoptotic protein whose phosphorylation by dysregulated cytosolic Erk would result in negative modulation of its pro-apoptotic this activity. Thus, re-regulation or normalization of Erk activity by E- 3a-diol restores the activity of the pro-apoptotic protein thereby promoting apoptosis. The screening methods disclosed and claimed herein are useful to identify compounds having low toxicity and/or capacity to elicit that pro-apoptotic activity in hyperproliferation conditions such as AR signaling-dependent or androgen-associated cancers.

[180] Based upon the insights from the present invention, a test or candidate compound that has an activity qualitatively or quantitatively similar to E-3a-diol at m-AR will activate Ga/q to negatively modulate the activity of a Class I A PI3K such as ρ1 10α/ρ85α and ρ1 10α/ρ85β. This negative regulation is believed to occur by negative modulation of tyrosine phosphorylation status of the PI3K regulatory subunit p85 (as is observed for E- 3a-diol) and possibly by interfering with the binding of activated Ras to the catalytic Docket No. 354. PATENT subunit p100a. Both of these effects are believed to be mediated by Ga/q, which is activated by "inverse agonist" binding of E-3a-diol to GPR-C6a, probably by its interaction with the COOH-terminal domain of p85. That interaction has the effect of disrupting excessive cross-talk between the PI3K-Akt and Ras-Erk signal transduction pathways that would result from p100a-Ras interaction or phosphorylation of Raf-1 Ser-338, which is catalyzed by p21 -activated kinase (Pak) downstream of PI3K..

[181] Therefore, GPR-C6a signaling through Ga/q counteracts PI3K signaling that would be due to that receptor's alternative engagement of Ga/i. In this respect, the prosurvival effect from GPRC signaling through Ga/i, that would result from Akt activation by PI3K activity through the intermediacy of phosphatidyl inositol-(3,4,5)-triphosphate (PIP3), would be reversed or opposed by GPR-C6a-Ga/q signaling. Furthermore, activation of PLC-β _! by Ga/q activation leads to consumption of PI3K's substrate phosphatidyl inositol (4,5)-bisphosphate (PIP2) to further limit Akt activation.

[182] Potential opposition to this anti-proliferative effect based upon PI3K substrate consumption is the enzymatic action by PLC-β _! to produce diacylglycerol that will activate certain PKC isoforms [Mellor, H. and Parker, P.J. (1998)] to subsequently result in nuclear accumulation of pErk [Caunt, C.J. et al. (2006); Armstrong, S.P. et al. (2009)]. Thus, G protein class switching by inverse agonist binding to GPR-C6a of a test or candidate compound to release Ga/q with qualitatively or quantitatively similar activity in this regard to E-3a-diol will elicit an anti-proliferative effect by preferentially engaging PI3K in comparison to PLC-β _! , or if additionally engaging PLC-β _! doing so such that activation of PKC or nuclear accumulation of pErk associated with that activation does not occur to an extent that obviates the antiproliferative effect from PI3K inhibition. Those dual effects on PI3K and PLC-β _! activities may provide an additional selection criterion for identifying candidate compounds.

[183] Without being bound by theory, it is believed that anti-proliferative effects of E-3a- diol is attributable to biological effects arising from cytoplasmic scaffolding of Erk-1 to retain pErk-1 in a cytosolic subcellular domain. This localization effect may be enhanced by negatively modulating nuclear translocation of pErk-1 that may be formed as a result of PLC-β _! activation (i.e., E-3a-diol redirects pErk-1 activity due to PLC-β _! activation to desired cytosylic effector proteins). Thus, a test or candidate compound having qualitatively or quantitatively similar effect to E-3a-diol on PLC-β _! stimulation of Ras-Erk signaling or PI3K activity should also be expected based upon the insights of the present invention to be anti-proliferative. Docket No. 354. PATENT

[184] Erk-1/2 activity is involved in several cellular processes other than inhibiting apoptosis by phosphorylated deactivation of pro-apoptotic proteins in hyperproliferating cells, and total Erk activity will influence an observed biological outcome [Lefloch, R. et al. (2008)]. For example, in normally proliferating cells, Ras-Erk signal transduction is important for cellular differentiation [Yashuda, T. (201 1 ); Robitaille, H. (2010); Gaest, C. et al. (2009); Jirmanova, L. et al. (2002)].

[185] Furthermore, despite Erk activation being associated with survival and anti- apoptotic signaling in hyperproliferating cells, Ras-Erk signal transduction is important for apoptosis in normally proliferating cells [Ishihara, Y. et al. (201 1 )] and for chemotherapeutic induction of apoptosis in cancer cells [Pal, P. and Kanaujiya, J.K. (201 1 ); Kim, M.J. et al. (2010)]. Additionally, in some malignant cells, particularly those which are PTEN-deficient, Ras-Erk signal transduction is actually inhibited due to Raf phosphorylation by high levels of activated Akt. Thus, in such situations reactivation of Ras-Erk signaling may have an anti-proliferative effect [McCubrey, J. A. et al. (2007)], and based upon the insights of the invention that anti-proliferative effect would most likely occur with a test or candidate compound that selectively phospho-activates Erk-1 vs. Erk- 2.

[186] Raf-1 is inactive when Ser-43, Ser-259 and Ser-621 are phosphorylated, which allows binding of 14-3-3 to induce an inactive configuration. Phosphatases dephosphorylate pSer-259, whereupon 14-3-3 disassociates to allow Raf to be activated by phosphorylation of Ser-338, Tyr-430 and Tyr-431 . Therefore, a test or candidate compound that reactivates Ras-Erk signal transduction in PTEN-deficient cancer cells to have an anti-proliferative effect may do so by negative modulation of the phosphorylation states of one or more of Ser-43, Ser-259 and Ser-621 , positive modulation of the Ser phosphorylate state of inactive Raf-1 (i.e., by altering the serine phosphorylation pattern) or negatively modulate binding of the scaffold protein 14-3-3 to Raf-1 . Such effects on Raf-1 Ser phosphorylation state or 14-3-3 binding are alternative criteria for identifying a candidate compound.

[187] Preferential retention of pErk-1 and suppression of Erk-2 activation induced by E- 3a-diol occurs through subcellular localization such that Erk anti-apoptotic activity in hyperproliferating cells is re-regulated or re-directed to negatively modulate anti-apoptotic signaling or to positively modulate differentiation which thereby induces apoptosis. In PTEN-deficient cancer cells characterized by low basal levels of pErk-1/2, E-3a-diol based upon the insights of the present invention will reactivate Ras-Erk signaling in the manner described such that pro-apoptotic signaling occurs. Without being bound by Docket No. 354. PATENT theory, this engagement of the apoptotic program may result from phosphorylated Elk-1 (a pErk substrate) preferentially retained in the cytoplasm and/or stimulation of p53 signaling in cancer cells retaining functional p53. It is believed those non-genomic effects from reactivation of Ras-Erk signal transduction (i.e., cytosolic p-Elk-1 or p53 activation) stem from negative modulation of pErk-1 nucleo-cytoplasmic transport and is due to scaffolding effects on this MAPK isoform induced by E-3a-diol activity.

[188] Some CaP cell lines such as LNCaP and PC-3 cells have PTEN mutations resulting in high levels of activated Akt and low levels of activated Raf, MEK and Erk. CaP cell lines have varying p53 states, e.g., LNCaP cells (bearing wild-type p53), DU145 cells (bearing de-activated functional p53) and PC3 cells (lacking functional p53). Therefore, some preferred test systems are comprised of hyperproliferating cells having inactive or deleted PTEN, a high level of activated Akt or a preponderance of Raf-1 in its inactive state. Other preferred test systems have wild-type or de-activated functional p53 (i.e., mutated p53 that binds more tightly to MDM2 than wt-p53, thereby repressing p53 pro-apoptotic activity) with inactive or deleted PTEN, a high level of activated Akt or preponderance of Raf-1 in its inactive state. More preferred are test systems comprising cancer cells with inactive or deleted PTEN, wild-type or inactivated functional p53 and no observable pErk-1/2 when incubated under serum-starved conditions or in androgen- and growth factor-depleted serum.

[189] Other preferred test systems comprise cancer cells with inactive or deleted PTEN that harbor gain of function mutant p53 (p53(GOF)) including LNCaP-R273H cells [Carroll, A.G. (1993)) and those described in Tepper, C.G. et al. (2005) and Nesslinger, N.J. et al. (2003). Preferred test or candidate compounds will positively modulate activity of one or more pro-apoptotic proteins or negatively modulate activity of one or more anti- apoptotic proteins in cells of such test systems despite dysregulated p53 activity from the presence of p35(GOF).

[190] A preferred test or candidate compound will reactivate Ras-Erk signal transduction, preferably with negative modulation of nucleo-cytoplasmic translocation of pErk-1 or pElk-1 . In these test systems, the test compound exerts an anti-proliferative effect or sensitizes the cancer cell to apoptosis from co-contact with a cancer chemotherapeutic compound. Those chemotherapeutic compounds include tubulin disrupting agents, DNA damaging agents, topoisomerase I and II inhibiting agents, disrupters of p53-MDM2 binding, inhibitors of RTKs, PI3K, Akt, NF-κΒ, AR and ERa, anti-androgens and ERp agonists. Preferred cancer chemotherapeutic compounds are tubulin disruptors, including taxane-based compounds such as paclitaxel, docetaxel, Docket No. 354. PATENT cabazitaxel and other tubulin disruptors described in Jordan, A. et al. (1998) and Perez, E.A. (2009).

[191] In hyperproliferating cells about 50% of total Erk remains in the cytoplasm primarily associated with other cytoplasmic proteins including microtubules [Reszka, A. A. et al. (1995)]. However, in order for a test compound to exhibit an antiproliferative effect due to preferential cytoplasmic retention of pErk, this sequesterization should be localized to a cytoplasmic subcellular compartment that also avoids or minimizes Erk phosphorylated-activation of anti-apoptotic proteins or deactivation of pro-apoptotic proteins. Therefore, without being bound by theory, it is believed that E-3a-diol-induced localization of pErk-1 in the cytosol results from scaffolding of pErk-1 within a protein complex onto a subcellular cytoplasmic endomembrane. That endomembrane is believed to be those of endosomes and localization of pErk-1 to this structure results from scaffolding, which may result from G protein-independent signaling that is likely initiated by inverse agonist action of E-3a-diol at GPR-C6a (i.e., results from Ga/q release from this GPCR).

[192] In addition to modulating aberrant or excessive (i.e., re-regulating) cross-talk between PI3K-Akt and Ras-Erk signaling, it is believed, based upon the insights provided by the invention disclosed herein, that GPR-C6a-Ga/q signaling may also modulate aberrant or excessive cross-talk between Src and Ras-Erk or Src and PI3K-Akt signal transduction pathways. For example, once activated, Src may tyrosine phosphorylate Ga/q at Tyr-356 (henceforth referred to as pGa/q) to increase the amount of bound GTP [Simon, M.I. et al. (1991 ); Umemori, H. et al. (1997)]. Since formation of pGa/q enhances inactivation of PI3K from GPCR-Ga/q signaling due to the action of Src, G protein-dependent signaling through Ga/q in effect converts pro-proliferative signaling from aberrantly activated Src to anti-proliferative signaling.

[193] In vitro, tyrosine phosphorylation of Ga/q is more active when PLC-β _! is stimulated, and in a feed-forward loop Ga/q subunits may activate PL Thus, in addition to its extracellular activity (i.e., inverse agonist activity at m-AR) additional selection criteria for identifying a candidate compound can may include an intracellular effect on Src activity that redirects PLC-β _! activity from PKC activation (pro-proliferative) to that of stimulating Ga/q phosphorylation (anti-proliferative).

[194] Other Ga monomer subunits, including Ga/i, can also be phosphorylated in vitro at tyrosine residues near their COOH-termini. Thus, one indication of selective engagement of Ga/q instead of Ga/i by a test compound contacted with a suitable in vitro test system Docket No. 354. PATENT is positive modulation of Ga/q Tyr-365 phosphorylation in comparison to the phosphorylation state of the corresponding residue in Ga/i, preferably with concommitant negative modulation of the phosphorylation state of that Ga/i tyrosine. This selective engagement of Ga/q can be another selection criterion for identifying a candidate compound.

[195] GPR-C6a-Ga/q G protein-dependent signaling is also believed to stimulate Ras- Erk signaling in the absence of RTK activation, within the context of re-regulating signal transduction cross-talk. That occurs, through Ga/q activation of PLC-β _! (a G protein- dependent effect) believed to occur at the cytosolic plasma membrane and by G protein- independent effects mediated by β-arrestin scaffolding of a cytosolic Src-containing protein complex or by β-arrestin scaffolding of Erk at early endosomes, particularly phosphorylated Erk-1 .

[196] Without being bound by theory it is believed, given that rapid (within 5 min.) transient signaling through PLC-β _! is associated with nuclear accumulation of pErk and gene transcription (within 15-30 min.), a later onset of Erk activation (between about 5-15 min.) in the cytoplasm results from G protein-independent effects mediated by β-arrestin scaffolding at endosomes. Thus, a test or candidate compound that results in Erk activation and preferential retention of pErk-1 in the cytoplasm by scaffolding proteins at endosomes will recapitulate an indirect intracellular effect associated with E-3a-diol (cf. direct intracellular effect on Erk-2 scaffolding that results in its cytoplasmic sequesterization in an inactivated form). Preferred test or candidate compounds will increase cytosolic Erk-1 within 15 min. without a corresponding increase in nuclear Erk- 1/2. It is further believed that endosomal localization of pErk-1 is responsible for the cytoplasmic subcellular sequesterization described herein that permits cellular functions that support differentiation without deactivation of mitochondrial associated pro-apoptotic proteins by phosphorylation.

[197] Furthermore, the choice of Ga/q effector protein, i.e., between negative modulation of the tyrosine phosphorylation state of PI3K p85 regulatory subunit (inhibition of activity) and positive PLC-β _! phosphorylation modulation (stimulation of activity), which is presumed to take place in different subcellular compartments, will also affect the strength of anti-proliferative activity resulting from engagement of Ga/q signaling. The strength of anti-proliferative activity resulting from choice of Ga/q effector protein or contributions from β-arrestin signaling, which are dependent on the isoform of the scaffolding protein, will be reflected by a resulting degree of retention of pErk-1 in the cytosol (anti-proliferative outcome) in relation to nuclear accumulation of pErk-1/2 in the Docket No. 354. PATENT nucleus (pro-survival or -proliferation outcome).

[198] Thus, based upon the insights provided by the invention disclosed herein, the choice of Ga/q downstream effector protein is also expected to be ligand dependent. Additionally, Ga/q engagement of PLC-β _! may lead to activation of protein kinase C, which is associated with nuclear accumulation of pErk and is thus pro-proliferative, while engagement of Ga/q with PI3K decreases its kinase activity and is thus anti-proliferative. Thus, test compounds that preferentially engage PI3K, or additionally engage PLC-β _! without positive modulation or with minimal increase in PKC activity, are preferred when considering selection of test or candidate compounds for further studies as described herein.

[199] Stimulation of GPR-C6a-Ga/q signaling and its resultant outcomes represent one set of non-genomic effects that are thought desirable for eliciting anti-proliferative outcomes. Another AR non-genomic effect due to inverse agonist engagement of GRPC- 6a is positive modulation of intracellular Ca ²⁺ levels. Those Ca ²⁺ levels may result from enzymatic hydrolysis by a PLC-β _! isoform of phosphatidylinositol 4,5-bisphosphate (PIP3) to inositol 1 ,4,5-trisphosphate (IP3), which then stimulates release of Ca ²⁺ from internal stores. Thus, some preferred test or candidate compounds will positively modulate intracellular Ca ²⁺ levels from mobilization of intracellular stores when contacted with cells having or engineered to have functional GPR-C6a. That would occur without significant accumulation of pErk in the nucleus that would otherwise obviate the antiproliferative effect of negatively modulating the phosphorylation state of p85 PI3K subunit. Non- genomic effects from i-AR recruitment to activated RTKs are discussed elsewhere in respect to other desirable selection criteria.

[200] Although Ca ²⁺ -induced cross-talk between PLC-PKC and Ras-Erk signal transduction pathways will result Erk in activation, mobilized Ca ²⁺ may also stabilize scaffolding interactions to negatively modulate nucleo-cytoplasmic transport thus preferentially localizing pErk to the cytoplasm [Chuderland, D. et al. (2008); Chuderland, D. and Seger, R. (2008)]. Thus, preferential sequesterization or retention of Erk-1 (in active form) and Erk-2 (in inactive form) in subcellular domains, believed to be due to the combined extracellular and intracellular actions of E-3a-diol, may be enhanced by non- genomic effects from GPR-C6a-Ga/q signaling that release Ca ²⁺ from intracellular stores. Thus, negative modulation of a test or candidate compound that negatively modulates the phosphorylation status of PI3K p85 regulatory subunit and positively modulates intracellular Ca ²⁺ levels are in some preferred embodiments characteristics for identifying a candidate compound. Docket No. 354. PATENT

[201 ] Membrane AR Ga/q signaling may also cross-talk with TNFCC-NFKB signaling since TNF-R1 (tumor necrosis factor receptor-1 ) is localized with various GPCRs and many other signaling components in the same subcellular compartments that include lipid rafts and caveolae. Furthermore, it has been reported that the TNF-R1 -TRAF2 (TNF-receptor- associated factor 2) receptor complex associates with Ga/q through β-arrestin-l scaffolding [Kawamata, Y. et al. (2007)], analogous to that observed for Src activation through GPCR signaling, which may employ β-arresin-l and/or p-arrestin-2 depending on the Ga/q coupled GPCR that is activated. Furthermore, scaffolding of Src by a particular β-arrestin isoform may result from Ga/q-dependent or G protein-independent signaling and is also determined by the Ga/q coupled GPCR and the activating ligand at this receptor. Therefore, test or candidate compounds to be identified as a candidate compound simulates G protein-dependent signaling through diversion of Ga/q from β- arrestin-1 to p-arrestin-2 scaffolding are sometimes preferred since such compounds are expected based upon the insights of the present invention to diminish or prevent potential cross-talk between Ga/q and NF-κΒ. That redirection of Ga/q scaffolding provides an anti-inflammatory effect and avoids confounding signaling from Ga/q activation. Since chronic non-productive inflammation is pro-proliferative and promotes metastasis, the antiinflammatory effect from preferential scaffolding should augment an antiproliferative effect resulting from Ga/q inhibition of PI3K instead of this kinase participating in NF-κΒ activation.

[202] Due to the numerous entries into Ras-Erk pathway described herein that may be mediated through GPR-C6a, it has now been discovered unexpectedly that 3a-diol and E- 3a-diol differentially activate the MAPK isoforms Erk-1 and Erk-2, although both compounds are agonists at GPR-C6a, such that 3a-diol prosurvival signaling is switched to E-3a-diol anti-proliferative signaling. That differential activation is unexpectedly accompanied by switching from GPR-C6a Ga/i signaling by 3a-diol to Ga/q signaling by E-3a-diol and results from differing agonist engagement of this GPCR by E-3a-diol. An agonist that redirects signaling away from Ga/i to Ga/q is sometimes referred to an "inverse agonist" (with respect to Ga/i signaling). Those discoveries were unexpected because scaffolding proteins and other signal transduction components are shared between signal transduction pathways, including those between GPR-C6a and Ras-Erk signal transduction pathways. Signaling through these various pathways can result in either prosurvival or anti-proliferative depending on the signaling context, e.g., the source of signal initiation and subcellular localization of signal transduction components and therefore prior to the present invention an anti-proliferative outcome by a test compound Docket No. 354. PATENT was unpredictable.

[203] In addition to the indirect engagement of the Ras-Erk pathway mediated by agonist activity of E-3a-diol at GPR-C6a, it also has been unexpectedly found that E-3a-diol physically interacts selectively with Erk-2 in comparison to the isoform Erk-1. That unexpected effect is more remarkable given that Erk-1 becomes preferentially activated by phosphorylation from contacting E-3a-diol with a suitable test system even though the selective physical interaction is with Erk-2. Those preferential isoform effects (i.e., direct to Erk-2 and indirect to Erk-1 ), without being bound by theory, are believed to be mediated through scaffolding proteins, which contribute to the apoptotic effect of E-3a-diol. Those isoform effects appear at least in part to modulate cross-talk between Ras-Erk and various other signal transduction pathways including signaling initiated at GPR-C6a. Subsequent activation of cytosolic retained Erk-1 in preference to Erk-2 activation negatively modulates Erk-2 activation of nuclear transcription factors, thereby negatively modulating the phosphorylation states of cytosol-residing transcription factors (i.e., inhibits their nucleo-cytoplasmic transport) and pro-apoptotic proteins (i.e., inhibits their deactivating phosphorylation). That would occur in part by Erk-1 outcompeting Erk-2 for the upstream activator MEK. The disruption of Erk-2 activation will therefore re-regulate or normalize apoptotic signaling (i.e., will provide an anti-proliferative effect).

[204] Thus, a test or candidate compound that exhibits one or more of the effects associated with Ga/q activation or "inverse" agonist binding to GPR-C6a when contacted with cells in a suitable test system that have or are genetically engineered to have a GPCR coupled to Ga/q provides selection criteria that can be used in identifying a test compound. G protein-dependent effects associated with Ga/q activation or inverse agonist binding by a test or candidate to GPR-C6a include negative modulation of one or more of (i) PI3K regulator subunit tyrosine phosphorylation state, (ii) ΡΚΟζ activity, (iii) PIP2 phosphorylation, (iv) PIP2 level without conversion to PIP3 or pErk nuclear accumulation, (v) Ras-PI3K interaction and (vi) Akt activity or phosphorylation state. Other effects include positive modulation of one or more of (vii) GTP-bound Ga/q level, (viii) PLC-β _! activity or interaction of Ρίθβτ with Ga/q with minimal or no positive modulation of PKC activity or nuclear accumulation of pErk, (ix) intracellular Ca ²⁺ mobilization with minimal or no positive modulation of PKC activity or nuclear accumulation of pErk and (xi) other non-genomic effects from GPR-C6a signaling that are pro-apoptotic or anti-proliferative.

[205] Particular preferred test or candidate compounds are those exhibiting one or more of those aforementioned GPR-C6a or Ga/q effects and that positively modulates Docket No. 354. PATENT

Erk-1 activity or its phosphorylation state in preference to Erk-2 without significant pErk accumulation in the nucleus, which would tend to counteract effects from pErk-1 retention in the cytosol. Other particularly preferred compounds will affect scaffolding of pErk-1 through G protein-independent signaling subsequent to Ga/q activation and its release from GPR-C6a (i.e., Ga/q-dependent and -independent signaling acting in parallel) such that pErk-1 is preferentially localized to endosomes or to a subcellular compartment that does not deactivate mitochondrial-associated pro-apoptotic proteins.

[206] 2(b)(ii). G Protein-independent Non-qenomic Activity: After ligand stimulation, GPCRs are phosphorylated by GPCR kinases (GRKs) to recruit β-arrestin whose role was thought restricted to agonist-dependent desensitization by internalizing and translocating the receptor to endocytic compartments. In what is known as classical G protein- independent signaling, scaffolding by β-arrestin then promotes recruitment and activation of Src and also recruits Raf, which is then activated by Src. It is now well accepted that certain GPCR ligands have dual efficacy effects by signaling through G protein-dependent and -independent pathways with the latter occurring through the intermediacy of the scaffolding protein β-arrestin. Therefore, in some situations a ligand for the same GPCR exhibits an "inverse agonist" effect by primarily signaling through β-arrestin scaffolding.

[207] Thus, the label of "inverse agonist", which by definition applies to compounds that decrease basal signaling of the receptor to which they bind, is a misnomer in this case since signaling through GPCRs is redirected through a different signaling pathway (e.g., Ga/i to Ga/q) having opposing effects that are agonist dependent rather than basal signaling being diminished overall.

[208] In classical G protein-independent activation of Erk, signaling from Raf scaffolding by β-arrestin occurs from the cytosolic membrane in clathrin-coated pits (Class A β- arrestin signaling). However, depending on the GRK isoform involved in GPCR phosphorylation that brings β-arrestin to this membrane receptor, recruitment and signaling through Erk will occur after internalization of the β-arrestin complex and translocation to early endosomes (Class B β-arrestin signaling).

[209] β-Arrestin scaffolding of Src in G protein-independent signaling is required for receptor internalization to endosomes and transactivation of RTKs such as EGFR. Since β-arrestin will also scaffold with components of the Ras-Erk signal transduction pathway it is believed that endosomal targeting of Erk that is mediated by β-arrestin-Src protein complexes will result in sequestering of pErk to the cytosol whereas β-arrestin scaffolding of Src at the plasma membrane (i.e., from Src recruitment to membrane- Docket No. 354. PATENT bound GPRC-p-arrestin protein complexes) will transactivate RTKs. Therefore, since pErk sequesterization to the cytosol is believed to re-regulate apoptotic activity, whereas RTK activation is considered pro-proliferative, test or candidate compounds that stimulate G protein-independent signaling by scaffolding with a β-arrestin isoform to preferentially target Src to endosomes are preferred.

[210] For activation of Ras-Erk signal transduction that is needed to stimulate pErk-1 formation for its cytosolic retention (i.e., when i-AR signaling-dependent or androgen- associated cancer cells of a suitable in vitro test system are incubated in growth factor- and androgen-depleted media), both Ga/q and the G protein-independent pathway using p-arrestin-2 scaffolding from GPCR-Ga/q signaling will usually provide the necessary Erk- 1 activation. Furthermore, p-arrestin-2 scaffolding is usually associated with cytoplasmic retention of pErk and is believed to be important for restoring apoptotic signaling while scaffolding by β-arrestin-l provides cross-talk to pro-inflammatory or pro-proliferative TNFa-NF-κΒ or MAPK signaling.

[211] Therefore, Src and β-arrestin-l mediate TNFa-induced Erk activation. Since cross-talk between Ras-Erk signaling and PI3K-Akt signaling may occur, β-arrestin-l scaffolding of cytosolic Erk may potentiate growth factor induced phosphorylation of Akt by positively modulating Src-mediated tyrosine phosphorylation state of PI3K regulatory subunit. That positive modulation of PI3K phosphorylation state is mediated in part by Ras acting as a scaffold protein. Thus, G protein-independent signaling downstream from GPR-C6a-Ga/q activation most likely requires p-arrestin-2 scaffolding in preference to that its isoform β-arrestin-l in order for a test or candidate compound to negatively modulate the tyrosine phosphorylation state of the p85 regulatory subunit and to exhibit antiproliferative effects that are enhanced by an anti-inflammatory activity. Thus, a test or candidate compound that induces p-arrestin-2-mediated G protein-independent signaling emanating from GPR-C6a activation in preference to that mediated by β-arrestin-l is one selection criterion for identifying a candidate compound.

[212] For G protein-independent effects mediated by β-arrestin scaffolding, a test or candidate compound that results in positive modulation p-arrestin-2 interaction with Erk- 1 , whereby cytoplasmic localization of pEk-1 is promoted or translocation of pErk to the nucleus is inhibited, is a preferred characteristic for identifying a candidate compound in some embodiments. Also, preferred are compounds that negatively modulate β- arrestin-1 interaction with Src, positively modulate Ga/q interaction with Src or positively modulated β-8 Ό8ΐίη-2 interaction with Raf, MEK or Erk for identification as candidate Docket No. 354. PATENT compounds. Preferred test systems for determining a test compound's activation of Ga/q and β-arrestin pathways include β-arrestin-l , -2 (single) and -1/2 (double) KO and Gq (double) KO mice as described in Kohout, T.A. et al. (2001 ) and Zywietz, A. et al. (2001 ).

[213] 2(c). Other Anti-Inflammatory Activities: As previously discussed some preferred test or candidate compounds with inverse agonist activity at GPR-C6a (i.e., having Ga/q agonist activity) will, through preferential p-arrestin-2 scaffolding, negatively modulate NF-KB signaling. In many cancer cells (including breast cancer, colon cancer, prostate cancer and lymphoid cancers) NF-κΒ becomes constitutively active and resides mostly within the nucleus. In some cancers, this constitutive activation is due to chronic stimulation of the IKK pathway, while in other cases the gene encoding ΙκΒ is defectively mutated. Other cancer cells have mutations or amplifications of genes encoding Rel/NF- KB transcription factors or mutations in genes encoding NF-κΒ signaling regulatory proteins that lead to constitutive activation of NF-κΒ. It is thought that continuous nuclear Rel/NF-κΒ activity protects cancer cells from apoptosis (i.e., is pro-survival) and in some cases is pro-proliferative.

[214] Members of one class of Rel/NF-κΒ proteins (including the NF-κΒ proteins p105 and p100) become activators of gene transcription when dimerized with members of the second class of Rel/NF-κΒ transcription factors after limited proteolysis (p105 to p50, p100 to p52) or arrested translation. The second class of Rel/NF-κΒ proteins, includes cRel, RelB and RelA (p65), referred to collectively as Rel proteins. It has been unexpectedly found that E-3a-diol negatively modulates cRel expression, and based upon the insights of the invention, a test or candidate compound that inhibits NF-KB transcriptional activity by modulating or normalizing Ras-Erk and/or PI3K-Akt signal transduction in the manner described herein for E-3a-diol would be useful as an anti- proliferative compound. That anti-proliferative effect may be observed in single agent therapy using the test compound or may be observed by way of tumor cells sensitized to other cancer chemotherapeutic compounds, including tubulin disrupting agents. That sensitization may be mediated in part by negatively modulating the phosphorylation state of one or more NF-κΒ signaling components that results in negative modulation of activity(ies) of the component(s). That sensitization also may be mediated in part by or negatively modulating gene expression for those components.

[215] Therefore, a test or candidate compound that affects transcription for a Rel protein in qualitatively similar manner to E-3a-diol would negatively modulate NF-KB transactivation activity and thus would have an anti-inflammatory effect. That anti- inflammatory effect should prevent initiation or retard progression of a hyperproliferation Docket No. 354. PATENT condition. Negative modulation of one or more of cRel expression, Rel protein level or NF-KB transcription of a κΒ-inducible promoter genetically engineered into cell of a suitable test system are other selection criteria for identifying a candidate compound. Other selection criteria related to effects of E-3a-diol on NF-κΒ activity for indentifying a candidate compound include negative modulation of (i) Rel/NF-κΒ activity in the nucleus, (ii) NF-kB dimerization, (iii) NF-kB dimer nuclear transport, (iii) phosphorylation state of p100 COOH-terminal region or some combination of the aforementioned effects attributable to decreased cRel expression.

[216] 2(d) Protein Scaffolding: The IKK protein complex includes catalytic IKKa and ΙΚΚβ subunits and the regulatory NF-κΒ essential modulator (NEMO), which is also required for NF-κΒ activation. NEMO and TRAF family member-associated ΝΚ-κΒ activator (TANK) is found associated with TRK fusion gene (Tfg) protein. Thus, Tfg is implicated as a component of the NF-κΒ activation pathway and in RTK signaling leading to Erk activation.

[217] Tfg has no known inherent activity, but contains numerous protein-protein interaction domains. Therefore, Tfg exhibits the hallmarks of a scaffolding protein due to its numerous protein-protein interaction domains (which are also present in other scaffolding proteins, including Ras, β-arrestin and others disclosed herein) and potential binding partners. Tfg is thus capable of bringing together signaling components that are common to several signal transduction pathways. As a further example, Tfg is a direct substrate of c-Src in vitro [Amanchy, R. et al. (2008)] and interacts with TANK and NEMO, which are components of the TNFa-ΝΚ-κΒ signaling pathway. Since Tfg binds to Src, this interaction (mediated by SH3 binding domains) is expected, based upon the insights provided by the invention disclosed herein, to be responsible for additional cross-talk between TNFa-NF-κΒ and Ras-Erk signaling pathways. Those pathways often become aberrant in various hyperproliferation conditions and chronic inflammatory states that support initiation or progression of such conditions. In addition, interaction between Src and Tfg provides another intersection for NF-κΒ and Erk signal transduction cross-talk with G protein-independent GPCR signaling, which is mediated by β-arrestin and PI3K-Akt signaling.

[218] Furthermore, Tfg associates with phosphatase and tensin homolog deleted on chromosome 10 (PTEN), a tumor suppressor that modulates PI3K-Akt signal transduction, which is an activity often lost in caner. Hyperproliferating cells characterized by aberrant RTK signaling harbor mutant PTEN, which results in inappropriate pro-survival signaling from dysregulated PI3K-Akt. As disclosed herein, it is believed that E-3a-diol, through Docket No. 354. PATENT selective activation of Erk-1 in contrast to Erk-2, contributes to restoring normal PI3K activity or reduces aberrant PI3K-Akt signaling in hyperproliferating cells having pro- proliferative RTK or PTEN mutations through its effect on Tfg scaffolding.

[219] Thus, a test or candidate compound that exhibits one or more of the effects associated with E-3a-diol on Tfg scaffolding when contacted with a suitable cell-based or cell-free test system can satisfy a selection criterion for identifying a candidate compound. Those selection criteria include one or more of negative modulation of (i) Tfg with SHP-1 , which will negatively modulate phosphorylation state of PI3K regulatory subunit thereby inhibiting PI3K kinase activity, (ii) Tfg interaction with NEMO or TANK, both of with will negatively modulate NF-kB activity, (iii) Tfg interaction with Src, which will disrupt aberrant Src signaling or redirect Src activity to promote anti-proliferative Ga/q signaling, or (iv) some combination of such effects.

[220] E-3a-diol exhibits favorable biologically active in patients with castrate-resistant prostate cancer (CRPC), in addition to patients with androgen-dependent CaP, without eliciting liver toxicity. It is believed that favorable biological activity is mediated through extracellular interaction of E-3a-diol at GPR-C6a and intracellular interaction(s) of this compound with component(s) of the Ras-Erk signaling pathway, including physical interaction with Erk-2, which is believed to be in association with one or more scaffolding proteins. Those favorable activities are applicable to other hyperproliferation conditions where AR signaling can occur due to the presence of functional i-AR (wild-type or mutant). That assertion is supported by anti-tumor effects in breast cancer models, the tumor cells of which are known to contain i-AR, which were observed or are observable in suitable in vivo test systems as described in the examples.

[221 ] It has also been unexpectedly found that E-3a-diol inhibits PI3K activity, the effect of which is to reduce downstream activation of Akt, and possibly that of NFK-B, all of which would otherwise lead to pro-proliferative signaling. Furthermore, it has been unexpectedly found that induced phosphorylation of the MAPK isoform Erk-1 by E-3a-diol occurs in the absence of external growth factors, which is normally required for Erk activation, with or without androgen stimulation (e.g., in intracellular AR signaling- dependent cancer cells incubated in growth factor and androgen-depleted media, with or without support from externally added androgen) or in the absence of observable constitutive RTK activation. Surprisingly, that effect on Erk-1 phosphorylation occurs without significant observable effect on basal levels of Erk-2 phosphorylation. For example, in i-AR signaling-dependent CaP cells incubated in growth factor- and androgen-depleted media little or no observable pErk-2 is detected and remains so after Docket No. 354. PATENT contacting these cells with E-3a-diol.

[222] Without being bound by theory, it is believed that selective activation of Erk-1 (i.e., by positive modulation of pErk-1 cytosolic level) is mediated in part by non-ATP binding- site dependent binding of E-3a-diol to Erk-2, postulated to occur in the presence of one or more scaffold proteins. That binding through scaffold proteins will sequester Erk-2 in a subcellular compartment such that stimulation of or cross-talk to Ras-Erk signaling will only result in phosphorylation of Erk-1 (and thus negative modulation of pErk-2 nuclear level by sequestering unactivated Erk-2 in a cytosolic compartment or retaining deactivated Erk-2 in the nucleus). It is further believed that recruitment of the scaffolding protein p-arrestin-2 to GPR-C6a occurs to form a G protein-independent signaling complex after stimulation of GPR-C6a-Ga/q from "inverse" agonist activity of E-3a-diol at this receptor. Scaffolding by p-arrestin-2 is believed to retain pErk-1 in the cytosol to a subcellular compartment that supports, after internalization of the receptor, cellular functions required for differentiation without de-activating mitochondrial-localized pro- apoptotic proteins through their phosphorylation. The effect from G protein-independent signaling through p-arrestin-2 is accompanied by G protein-dependent signaling that releases Ga/q (GTP) from GPR-C6a such that tyrosine phosphorylation of PI3K p85 regulatory subunit is negatively modulated. Those effects on GPR-C6a signaling modulation are believed to exert an anti-proliferative effect without significant activation of PKC from PLC-βΙ (i.e., there is insufficient activation of PKC to exert a pro-proliferative effect that negates the effect from inhibiting PI3K activity). Thus, Erk-2 scaffolding is believed to be directly modulated by E-3a-diol acting upon an Erk-2 containing intracellular protein complex, whereas Erk-1 scaffolding is indirectly affected, through the intermediacy of Ga/q and p-arrestin-2, by extracellular action of E-3a-diol on GPR-C6a.

[223] In cancer cells loss of regulation or activity in some components (either due to lack of expression or expression of mutant genes encoding the components) or over-activity in others (either due to excessive activation or over-expression), typically a combination of these effects, is expected, based upon the insights provided by the invention disclosed herein, to cause failures in proper component sequesterization that results in aberrant crosstalk between signal transduction pathways.

[224] Without being bound by theory, it is believed for some embodiments of the invention that E-3a-diol or compounds identified as candidate compounds by methods disclosed herein redirects improper signaling through sequesterization of signal transduction components to restore proper activation of effector substrates. In one aspect of the invention it is believed that Erk-2 is sequestered in a state resistant to activation in a Docket No. 354. PATENT cellular compartment within a protein complex. That protein complex is supported by one or more scaffold proteins that prevent phosphorylation of Erk-2 and transport of any pErk- 2 that may be formed to the nucleus (or alternatively or additionally prevents transport of deactivated Erk-2 from the nucleus for cytosolic reactivation) thereby inhibiting activation of transcription factors that would otherwise support proliferation or survival. It is further believed that sequesterization of Erk-2 allows for activation of Erk-1 so that preferential phosphorylation of cytosolic effector substrates that support differentiation occurs without activating anti-apoptotic signaling. That selectivity for cytosolic effector proteins by pErk-1 is believed to result from pErk-1 (or Erk-1 prior to activation loop phosphorylation) sequesterization to a subcellular cytoplasmic domain.

[225] Therefore, in the aforementioned manner, E-3a-diol and similar test or candidate compounds modulate the phosphorylation states of the Erk isoforms so as to redirect proliferative and survival signaling to pro-apoptotic signaling or differentiation (which induces apoptosis). Additionally, it is believed that aforementioned effects on Erk isoform scaffolding results in alteration in the phosphorylation status of i-AR as described herein such that signaling characteristic of differentiation, which is supported from i-AR activation in normal cells, is restored sufficiently in AR-dependent cancer cells to result in cell death. Thus, pro-proliferative signaling of mt-AR may be redirected by E-3a-diol to support differentiation.

[226] Modulators of PI3K and/or Erk identified by the methods disclosed herein are expected to have anti-proliferative effects while having lower toxicity in comparison to ATP-site dependent inhibitors of these proteins due to the conditional nature of their effects. Thus, basal signaling in normal proliferating cells is not adversely effected in comparison to aberrantly proliferating cells since the result of the intercellular interaction(s) at GPR-C6a and/or on the Erk isoforms is to disrupt formation of inappropriately formed protein complexes that occur primarily in the abnormal cells. Since the non-ATP site dependent modulators identified by the methods disclosed herein effect specific protein interactions that are formed under specified conditions as described herein, their modulating activities are not always "on". That is in contrast to ATP-site inhibitors that offer no distinction between protein complexes in which the kinase participates and hence these inhibitors can exhibit no isoform selectivity.

[227] In support of the various mechanisms described herein that underlie the antiproliferative activity of E-3a-diol and that of test compounds to be selected as candidate compounds, reduced phosphorylation of PI3K regulatory subunits p85a (regulatory subunit 1 polypeptide) and ρ85β (regulatory subunit 2 polypeptide) in LNCaP cells (human Docket No. 354. PATENT lymph node prostate cancer cells), which is a cell line often used for in vitro testing of compounds for activity in prostate cancer, was observed for E-3a-diol when contacted with these cancer cells in a suitable test system. Reduced PI3K phosphorylation by E-3a- diol contributes to the observed anti-proliferative effects of this compound. That activity is now expected, based upon the insights of the invention, to occur when tumor cells expressing GPR-C6a are contacted with E-3a-diol. Since release and phosphorylation of Ga/q (GTP) resulting from activation of a GPCR coupled to this G protein monomer decreases PI3K regulatory subunit phosphorylation, it is believed that E-3a-diol acts as an agonist for GPRC-6a-Ga/q signaling (and thus is an inverse agonist for Ga/i signaling from this same receptor).

[228] In consideration of the foregoing and without being bound by theory, it is postulated that inhibitory effect of E-3a-diol on PI3K phosphorylation within a suitable test system, wherein the cells comprising the system express functional GPR-C6a and functional intracellular AR (wild-type or mutant i-AR), is contributed by inverse agonist engagement of GPR-C6a (i.e., signaling proceeds through Ga/q release). Therefore, it is believed the ability of E-3a-diol to inhibit the pro-proliferative effect of androst-5-ene-3p, 17β -diol (PAED) on low passage (LP) LNCaP is not only opposed by competitive binding of E-3a-diol to GPR-C6a, but is actually reversed due to activation of a counteracting signal transduction cascade due to Ga/q release that inhibits PI3K. Unopposed Ga/i signal transduction is postulated to occur by pAED agonist engagement of intracellular mutant AR (m-AR) and possibly by agonist engagement GPR-C6a such that signaling occurs through activation of Ga/i to increase PI3K activity. E-3a-diol reverses this signaling and results in re-regulated cross-talk between the Ras-Erk and PI3K-Akt pathways such that signaling through i-AR to promote differentiation occurs instead of supporting survival or proliferation when Ras-Erk signal transduction is aberrantly activated. That alternative engagement of i-AR in i-AR signaling-dependent or androgen- associated cancer cells is thus in opposition to pro-proliferative signaling from this nuclear hormone receptor in these cancer cells and contributes to the observed anti-proliferative effects of E-3a-diol.

[229] Without being bound by theory, direct or indirect physical interaction (i.e., direct binding to a protein or indirect binding to a protein through intermediacy of another protein) of the cell-permeable compound E-3a-diol with one or more intracellular components of Ras-Erk signaling is believed to contribute to this compound's antiproliferative effect, whether the Ras-Erk pathway is activated by cross-talk from GPR- C6a-Ga/q signaling or from external mitogen, constitutive RTK stimulation or other Docket No. 354. PATENT aberrant tyrosine kinase activity. Thus, the anti-proliferative effect of E-3a-diol results in part from subcellular sequesterization of Erk-2 and/or, or alternatively from binding Erk-2 in a non-productive protein complex, that prevents its phosphorylation and subsequent translocation of pErk-2 to the nucleus and/or translocation of deactivated Erk-2 in the nucleus to the cytosol. The effect(s) on Erk-2 sequesterization limits cell proliferation or survival by reducing phosphorylation of certain transcription factors in the nucleus, such as Elk1 or CREB. That anti-proliferative effect is augmented by phosphorylation of Erk-1 to result in actions of pErk-1 on cytosolic targets that supports apoptosis and by reduction of excessive PI3K activity often found in tumor cells due to loss of PTEN activity. Suitable test systems can thus comprise androgen-associated tumor cells with lost PTEN activity.

[230] For the aforementioned mechanism, it is also believed that binding of component(s) of one or more signal transduction cascades mediated by scaffolding proteins through protein-protein interaction are modulated by E-3a-diol to affect the Erk isoforms in the aforementioned manner. Thus, inhibition of Erk-2 phosphorylation may occur by E-3a-diol trapping unactivated Erk-2 in non-productive complexes or subcellular sequestering of this isoform to interfere with required protein complex formation required for its phosphorylation. In either scenario, pro-survival or pro-proliferative signaling, through engagement of protein-protein interaction domains within those complexes, is inhibited. In contrast to Erk-2, selective formation of pErk-1 , due to removal of Erk-2 from the Ras-Erk signal transduction pathway, results in anti-proliferative effects from pErk-1 phosphorylation of certain cytosolic targets. Those unexpected differential effects on Erk isoform activity are consistent with the observed binding of E-3a-diol to Erk-2, as shown by Example 1 , with the lack of observable increase in phosphorylation of this isoform.

[231] In the absence of external mitogen activation or constitutive RTK activity, an increase in Erk-1 phosphorylation is believed to be mediated at least in part through the aforementioned agonist engagement GPR-C6a by E-3a-diol that results in Ga/q release, which productively stimulates the Ras-Erk signal transduction pathway. The lack of observable phosphorylation of Erk-2 from this stimulation, concommitant with increased pErk-1 formation, is believed to be mediated by scaffolding protein(s). In the presence of mitogen or constitutive RTK activity excessive Ras-Erk signal transduction is believed to be tunneled by these scaffolding effects to selective pErk-1 formation to exert an antiproliferative effect. Engagement of GPR-C6a Ga/q by E-3a-diol will augment this effect through scaffolding of activated Src with Ga/q (GTP) in preference to PLC-βι . Thus, Ga/q- Src scaffolding not only results in Ga/q (GTP) phosphorylation that leads to negative modulation of PI3K phosphorylation status, but also redirects excess Src activation from Docket No. 354. PATENT pro-proliferative to anti-proliferative effects mediated by Ga/q.

[232] Therefore, in CaP (prostate cancer) cells in vitro and in vivo it is believed that Ga/q activation is responsible at least in part to the observed inhibition of PI3K that occurs as a result of "inverse" agonist action of E-3a-diol at GPR-C6a to activate Ga/q in preference to Ga/i that would otherwise result from agonist action of 3a-diol at this same receptor. Scaffolding of Ga/q from GPR-C6a activation by E-3a-diol with Src is probably enhanced by either intracellular action of E-3a-diol on this scaffolding or is indirectly due to selective recruitment of 3-arrestin-2 (vide infra) to phosphorylated GPRC-6a occurring subsequent to Ga/q release from the receptor. Those effects on Src scaffolding diverts activated Src, which represents a significant signal transduction node often aberrantly stimulated in cancer cells, away from PI3K activation toward its inactivation. Modulation by E-3a-diol of G protein-independent signaling through GPR-C6a may further contribute to the redirection of aberrant Src activity away from PI3K activation through Src subcellular sequesterization away from pro-proliferative effector proteins. That sequesterization of Src may also be mediated by intracellular action of E-3a-diol on protein scaffolding and may involve disruption of protein complexes involving β-arrrestin-l . Scaffolding of Erk-2 that results in its sequesterization to inhibit its activation from stimulation of the Ras-Erk signal transduction pathway may occur through scaffolded protein complexes. Those scaffolded protein complexes contain one or more scaffolding proteins described herein and likely includes the putative scaffolding protein Tfg.

[233] In contrast to the aforementioned cross-talk between G protein-independent signaling with Ras-Erk signaling and between G protein-dependent signaling with PI3K- Akt, scaffolding of Src for Ras-Erk cross-talk through Src phosphorylation of Raf is thought to be negatively modulated by candidate compounds. Modulation of Src scaffolding by E- 3a-diol may thus inhibit aberrant cross-talk between mutant RTK and Ras-Erk signaling by decreasing excessive (i.e., re-regulating) stimulation of the Ras-Erk signal transduction pathway. That re-regulation or normalization may occur by E-3a-diol favoring the scaffolding of Src with p-arrrestin-2 instead of β-arrestin-l , which is believed to occur when aberrant Src signaling is present.

[234] Furthermore, in addition to selective interaction with Erk-2 scaffolding to inhibit pErk-2 formation, it also has been unexpectedly found that E-3a-diol physically interacts (i.e., directly binds) with Tfg to contribute to the observed antiproliferative and antiinflammatory properties of E-3a-diol. Thus, E-3a-diol mediates cross-talk between one or more of GPCR, Ras-Erk, TNFa-NF-κΒ, PI3k-Akt and c-Src signaling pathways, and is believed due to the commonality of scaffold proteins used in these signal transduction Docket No. 354. PATENT cascades, including those involving Tfg. For example, PI3K scaffolding with Tfg, which is postulated to be enhanced by E-3a-diol, may promote PI3K deactivation due to the aforementioned redirection of Src activity to tyrosine phosphorylation of Ga/q. Alternatively, but not to mutual exclusion, E-3a-diol binding to Tfg may dislodge SHP-1 thereby restoring that protein's phosphatase activity to ameliorate excessive signal transduction by PI3K resulting from aberrantly activated tyrosine kinases.

[235] Without being bound by theory, it is believed that E-3a-diol modulates cross-talk between one or more of GPCR, Ras-Erk, TNFa-NF-κΒ, PI3k-Akt and native Src (or its v- Src mutant) signaling pathways initiated through RTK, GPCR G protein-dependent or GPCR G protein-independent signaling by affecting scaffolding of protein complexes that involve one or more of β-arrestin, Ras/Raf, Tfg and other scaffolding proteins (including kinases and phosphatases acting as scaffolding proteins) disclosed herein. It is believed that modulation of cross-talk by E-3a-diol occurs through changes in scaffolding within protein complexes in a manner that reduces pro-survival signaling while increasing pro- apoptotic signaling that may also provide an anti-inflammatory effect. That effect on scaffolding, which is mediated through protein-protein interaction domains, occurs through interaction of E-3a-diol with Erk-2 in a non-ATP binding site-dependent fashion. It is further believed this modulation of cross-talk is initiated by signaling through GPR-C6a- Ga/q and can occur in the absence of other external stimulation, including that arising from RTK signaling either by external mitogen or constitutive activation.

[236] It is therefore believed, without being bound by theory, that E-3a-diol or preferred test or candidate compounds, which exhibits one or more of the intracellular effects of E- 3a-diol as described herein, affects component scaffolding within an Erk signal transduction node. It is further believed this scaffolding effect inhibits or compensates for the abnormal signal transduction cross-talk such that pro-proliferative signaling is attenuated or reversed (i.e., promotes differentiation-induced apoptosis) within cancer cells. It is further believed such test or candidate compounds will influence protein-protein interactions with activated i-AR so as to reverse pro-proliferative signaling to drive back tumor cells to the differentiated state thereby inducing apoptosis.

[237] (2)(e). Modulation of Short-chain Dehydrogenase Activity: Additionally, it has also been discovered that E-3a-diol interacts with 17p-hydroxysteroid dehydrogenase Type 10 (17HSD10), a hydroxysteroid dehydrogenase isoform highly active towards 3a- hydroxysteroids. To date, no selective inhibitors of that enzyme have been described. Since this hydroxysteroid dehydrogenase interconverts 3a-diol to DHT (5a- dihydrotestosterone), inhibition of that process by E-3a-diol will reduce formation of DHT, Docket No. 354. PATENT a powerful cytosolic AR activator (which may also stimulate GPR-C6a Ga/i signaling), to prevent its accumulation in prostatic tissue while counteracting the activity of 3a-diol and DHT at GPR-C6a. That counteraction may occur not only by E-3a-diol acting as a competitive inhibitor for 3a-diol or residual DHT binding to the m-AR, but also by E-3a-diol differentially engaging (i.e., Ga/q vs. Ga/i) the Ras-Erk signaling pathway that is stimulated by GPR-C6a signaling so as to promote apoptosis rather than inhibition of this process by 3a-diol. Stated differently, E-3a-diol opposes Ga/i signaling from GPR-C6a by acting as an "inverse agonist" of that activity.

[238] Therefore, it is additionally believed that contact of E-3a-diol with CaP cells in vivo inhibits the production of DHT from 3a-diol reservoirs by direct interaction with 17HSD10 to diminish intracellular AR (i-AR) signaling that supports survival or proliferation. Activation of this unfavorable i-AR signaling may also be diminished by reducing tyrosine kinase phosphorylation of this nuclear hormone receptor due to the aforementioned action of E-3a-diol on Src signal transduction. That disruption of cross-talk between i-Ar and Src signaling is postulated to occur through re-direction of Src activity either by scaffolding with Ga/q or by scaffolding of Src by p-arrestin-2 (or inhibiting scaffolding of Src with β- arrestin-1 ). Scaffolding by p-arrestin-2 during G protein-independent signaling, or through Tfg scaffolding of PI3K, will also influence Src signaling through these protein complexes such that re-regulation of i-AR tyrosine phosphorylation occurs.

[239] Thus, a test compound contacted with a suitable test system in vitro containing mammalian 17HSD10 that negatively modulates oxidoreductase activity of the enzyme in the presence of NAD+ (oxidative mode) or NADH (reductive mode) and a short chain alcohol or ketone or 3a- or 17p-hydroxy steroid substrate or inhibits turnover of a 17HSD10 substrate or negatively modulates interconversion of 3a-diol and DHT (e.g., test compound competitively inhibits 3a-diol binding to 17HSD10, with or without turnover of the inhibitor, allosterically inhibits substrate oxidation or reduction, inhibits product release or engages in product inhibition) provides an activity that may be used for selecting a candidate compound. Such negative modulators of 17HSD10 activity are not only useful candidate compounds for treating cancer, they are also candidate compounds for treating other conditions or diseases having an unwanted inflammatory component, which also can promote these hyperproliferating conditions.

[240] Furthermore, a test compound contacted with a suitable in vivo test system that decreases the serum or intraprostatic concentration of DHT in comparison to placebo or decreases the serum or intraprostatic concentration of DHT in qualitatively similar manner to a positive control inhibitor of 17HSD10 or to that of E-3a-diol exhibits an activity that Docket No. 354. PATENT can be used in selecting a candidate compound. Additionally, a test compound found to bind to Tfg or to selectively bind Erk-2 in comparison to Erk-1 in intact cells or cell-lysates provides an activity that may be used in identifying a candidate compound.

[241] 2(f). Prostate Cancer Cell Lines: LNCaP clone FGC was originally derived from a lymph node biopsy of a patient with confirmed diagnosis of metastatic CaP. These cells are responsive to DHT (growth modulation and acid phosphatase production) and are androgen receptor, positive; estrogen receptor, positive. Through titration experiments, a concentration of 10 nM βΑΕϋ was found to provide maximal stimulation of LP LNCaP, while no proliferation of LP LNCaP cells was observed without the addition of βΑΕϋ. High-passage (HP) LNCaP cells are castration-resistant, but they still express AR and require the presence of functional i-AR to proliferate. HP and LP LNCaP cell both contain the T877A mt-AR. Other cell lines for evaluating test compound activity in CaP include PC-3 and DU-145

[242] PC-3 cell line was originally derived from advanced androgen-independent, bone- metastasized prostate cancer. DU-145 cell line was originally derived from a brain lesion in a patient with metastatic CaP and history of lymphocytic leukemia. DU-145 and PC-3 cells are reported to be AR-negative and thus are not detectably hormone sensitive. However, other studies have provided evidence that the DU-145 and PC-3 cell lines contain AR mRNA and it has been subsequently found that that DU-145 and PC-3 cell lines produce AR protein, but the relative levels of the AR mRNA and protein were lower than in LNCaP, an AR-positive cell line. Furthermore, treatment of those cell with DHT resulted in measurable increases in the AR protein levels and considerable nuclear accumulation. Although, treatment of DU-145 and PC-3 cells with DHT did not result in stimulation of the activity of an AR-responsive reporter, knockdown of AR expression in PC-3 cells resulted in decreases in p21 ^Cip1 protein levels, and a measurable decrease in the activity of a p21 -luc-reporter [Alimiraha, F. et al. (2006)].

[243] Irrespective of the conflicting reports on AR expression in PC3 and DU-145 cells, it is apparent this nuclear hormone receptor is mutated such that it no longer recognizes its canonical androgen agonists. However, this mt-AR is required for p21 ^Cip1 transcription. Although p21 is typically thought of as an inhibitor of cyclin D-CDK complexes that arrest the cell cycle in G in some cancer cells it may serve a positive role in G1 progression. For example, in androgen-dependent LNCaP, androgen-AR induces expression of p21 that in turn inhibits tumor necrosis factor a-induced JNK and apoptosis. Furthermore, Ras activation, which is often aberrant in cancer cells, upregulates the cyclin D1 gene. Thus, it is thought that cross-talk exists between Ras-Erk and AR signaling in order to stimulate Docket No. 354. PATENT

Rb inactivation and cell cycle progression in cancerous cells.

[244] Therefore, the effect on AR knockdown in PC-3 cells on p21 ^Cip1 indicates that transactivation activity of the mt-AR in this setting has been retained, although mutations, most likely in the LBD, allow this receptor to act independently of androgens. Thus, E-3a- diol may have an effect on the phosphorylation state of this mutant or affect crosstalk between AR, Ras-Erk or TNFa-JNK signaling, but is not sufficient in of itself to promote apoptosis (through inducement of differentiation). That may occur if E-3a-diol is an inadequate agonist (i.e. partial agonist) at this mt-AR for eliciting the required AR signaling that needs to be redirected for cell differentiation. However, the intracellular action of E- 3a-diol, perhaps in combination with its activity at presumptive GPR-C6a, may be sufficient to sensitize PC-3 and DU-145 to apoptosis from co-contacting cancer cells with a tubulin disrupting agent or other cytotoxic cancer chemotherapeutic compound.

[245] The anti-proliferative effect of E-3a-diol can be observed in a suitable in vitro or in vivo test system wherein the cells comprising the test system express or overexpress GPR-C6a and the antiproliferative effect is due in part to re-regulation or normalization of cross-talk between signal transduction through Src and GPCR signaling through Ga/q activation. Cross-talk of Src and GPR-C6a Ga/q signaling is believed to occur through scaffolding of the G protein and the tyrosine kinase or through Src scaffolding by β- arrestin-2 in a G protein-independent manner, and may occur in combination with other additional scaffolding proteins. With either outcome, Src signaling is redirected away from activation of i-AR that occurs through the "outlaw pathway", so named for ligand independent activation of i-AR through its tyrosine phosphorylation by Src.

[246] That androgen-independent route to i-AR activation through the "outlaw pathway", which is mediated by tyrosine AR phosphorylation by Src, causes CaP proliferation or survival that is dependent on AR gene expression to become unresponsive to androgen ablation. Thus, AR activation by this "outlaw pathway" is postulated to become dominant during the transition to CRPC (castrate-resistant prostate cancer). Cross-talk between Ga/q signaling or G protein-independent signaling with Src signaling, both of which are believed mediated by protein complexes with scaffolding of Src, will divert Src activity away from PI3K activation (passive inhibition) while providing Ga/q-mediated PI3K inhibition (active inhibition). Those effects on PI3K signaling will at least in part prevent or delay the transition from androgen-dependent to androgen-independent activation of i-AR. In CRPC i-AR binding by agonist in the cytosol (which results in translocation of the activated nuclear hormone receptor to the nucleus to promote proliferation) is still thought required, albeit at a much lower concentration of ligand that is supportable by residual Docket No. 354. PATENT levels of androgens that are present subsequent to castration.

[247] Prostate cancer cells known to endogenously have m-AR include LNCaP and DU145 [Kampa, M. et al., (2002); Papadopoulou, N. et al. (2008)]. Other cancer cells known to endogenously have m-AR include breast cancer cells T47D and MCF-7 [Kampa, M. et al. (2005); Kallergi, G. et al. (2007)] and colon cancer cells including Caco2 and HCT1 16 [Gu, S. et al. (201 1 ); Gu, S. et al. (2009)]. m-AR is also found in carcinogen- induced colon tumors. Therefore, suitable test systems for screening test compounds for E-3a-diol activity associated with its extracellular interaction at m-AR thus include a suitable cell-based in vitro or in vivo test system comprising LNCaP, MCF-7 or T47D cells, which are AR signaling-dependent, or DU145, and probably PC-3 cells, that are AR ^_/" but are GPR-C6a ^{+ +} (i.e., are androgen-associated). Other suitable tests systems for that purpose are HCT1 16 or transformed Caco2 cells or colon tumor cells that are carcinogen- induced in Balb/c mice [Wang, J.G. (2004)].

[248] For androgen-associated cancer cells that are considered AR ^{_ ~} (e.g., PC-3, DU145, CV-1 ), but nonetheless have or may have the membrane androgen receptor GPR-C6a, the anti-proliferative effect of E-3a-diol will primarily come from its redirection of Ras-Erk signaling to phosphorylate select cytoplasmic targets by inducing preferential activation of Erk-1 . In such situations, particularly when the cancer is indolent or when AR signaling-dependent cancer cells are incubated in growth factor- and androgen-deleted media, there may be insufficient basal Ras-Erk signaling for there to be a frank antiproliferative effect from contacting a suitable test system comprising these AR ^{_ "} cells with E-3a-diol as a single agent. However, Ga/q release from E-3a-diol "inverse agonist" activity in AR ^{_ "} cells endogenously expressing or engineered to express GPR-C6a may provide sufficient Ras-Erk signaling to sensitive those cancer cells to apoptosis. Thus, co- contact of a test system comprising AR ^{_ "} cells with another cancer chemotherapeutic compound such as a tubulin disrupting agent at a sub-therapeutic amount for a single agent therapy is believed to now be effective in inhibiting proliferation of such cells.

[249] Therefore, E-3a-diol contacted with cancer cells in a suitable test system with proliferating cancer cells that are AR ^{_ "} , incubated in growth factor- and androgen-depleted media, and found to have endogenous m-AR may exhibit anti-proliferative effects when the cells are co-contacted with a cancer chemotherapeutic agent. Other suitable test systems are GPR-C6a ^+/+ AR' ^' cancer cells that are engineered to be AR ^/+ and which are proliferating in growth factor- and androgen-depleted media. In those suitable test systems and others comprising GPR-C6a+/+ AR+/+ cancer or transformed normal cells, E-3a-diol is expected to exert an antiproliferative effect that is enhanced or synergistic Docket No. 354. PATENT with a cytotoxic chemotherapeutic compound such as a tubulin disrupting agent. Still other suitable tests are transformed normal cell not endogenously expressing either gene, but are genetically engineered to express GPR-C6a and AR.

[250] Thus, a test or candidate compound with qualitatively or quantitatively similar activity to E-3a-diol with respect to modulation of PI3K activity or its phosphorylation state may result in inhibition of survival or proliferation of the cancer cells through negative modulation of activity(ies) of downstream effector protein(s) in these cells, including that of Akt and/or PKC, when the cells are (1 ) AR positive, with or without GPR-C6a over- expression and dependent on context (i.e., in growth and androgen-depleted media vs. external mitogen or i-AR stimulation) or in (2) AR negative cells with GPR-C6a expression and co-contacted with a cancer chemotherapeutic compound. Other suitable test systems include cancer or transformed cells that have or are engineered to have i-AR with GPR- C6a present, either endogenously or through genetic engineering, incubated in growth factor and androgen-depleted media with or without co-contact of test compound with an externally added androgen or cancer chemotherapeutic compound. A test or candidate compound contacted with such test systems that negatively modulates PKC or Akt activity or negatively modulates phosphorylation states of one or more downstream effector proteins that include adaptor proteins (Bcl-2, BAD, MDM2, PRAS40), other kinases (glycogen synthase kinase-3 [GSK-3], ΙκΒ kinase-β [ΙΚΚ-β], mTOR), GTPases (Rac/cdc42, Rho), caspases (CASP9), metabolic enzymes (endothelial nitric oxide synthase [eNOS], 6-phosphofructo-2-kinase), cell cycle regulators (MDM2, p21 ^Cip1 , p27 ^Kip1 ), transcription factors (FOXO/Forkhead), and cancer susceptibility genes (BRCA1 ) can be used as additional criteria for identifying a candidate compound.

[251] 2(g). Modulation of AR transactivation: In /-AR, there are at least 1 1 serine residues that may be phosphorylated, depending on the cell culture system and the hormonal or growth factor stimulation employed. Although, distributed throughout the length of i-AR, a majority of these serine residues are located in the NH ₂ -terminal domain (NTD) with many of these having an adjacent proline residue COOH-terminal to the phosphorylated serine, which can be recognized by the peptidyl prolyl isomerase Pin-1 . The NTD, which has little intrinsic tertiary structure, contains the hormone-independent coactivator interface AF-1 . Thus, phosphorylation of one or more of the NTD serines may result in a conformational change to a more ordered structure to promote cofactor and DNA binding. In addition the NTD contains several tyrosines that may become phosphorylated by Src, which have importance for nuclear translocation and Cdc42 associated kinase (Ack)-induced activation that promotes CaP progression. Furthermore, there are serine residues that become specifically phosphorylated upon stress kinase Docket No. 354. PATENT activation, EFG stimulation and by the PI3K-Akt signaling pathway. Therefore, different phosphorylation patterns may occur in a context-specific manner to modulate the nucleo- cytoplasmic trafficking of i-AR and recruitment of various co-regulators that effect transcription.

[252] Without being bound by theory, it is believed the inhibitory effect of E-3a-diol on PI3K and Src activity and its selective engagement of the Ras-Erk pathway to activate pErk-1 results in a phosphorylation pattern in the NTD of i-AR that is fundamentally different to that provided by hormone binding. As a result, pro-survival gene transcription by hormone activated i-AR is reduced in hyperproliferation conditions that are dependent on gene transcription activity by this nuclear hormone receptor. It is further postulated that different co-regulator recruitment as a result of this altered NTD phosphorylation may contribute to transactivation of genes for differentiation, which is the role hormone- activated i-AR normally performs in non-hyperproliferating cells.

[253] Furthermore, co-activator stimulation of i-AR may also be dependent on the phosphorylation status of the co-activator. Thus, it is postulated that the anti-proliferative effect of E-3a-diol may also be due in part to modulation of the phosphorylation status of one or more p160 co-activators, including that of Ser-738 in TIF2. Protein levels of this co-activator are increased in various recurrent CaP cell lines, which in some embodiments are included in certain suitable test system as described herein, concomitant with increased phosphorylation status of Ser-738. The increased TIF2 phosphorylation, which increases the interaction between this coactivator and i-AR is presumably due to activation of Ras-Erk signaling resulting from aberrant or excessive cross-talk with RTK signaling, including signaling from EGF stimulation.

[254] Thus, ErbB2 receptors may be important modulators of i-AR activity, since these receptors associate with TGF-R that participates in the "outlaw pathway" through activation of Src, and may mediate Erk activation by EFG signaling. Additionally, EGF- induced transactivation activity of i-AR results in increased levels of TIF2 as well as increased circulating levels of EFG to establish an autocrine loop. Therefore, without being bound by theory, it is believed that Erk-2 sequesterization in activation-resistant form by E-3a-diol disrupts this autocrine loop through the resulting decrease in the phosphorylation status of TIF2 Ser737 thereby contributing to the anti-proliferative effects of this compound. Those effects on i-AR in hyperproliferating cells is coupled with sequesterization of Erk-2 to prevent its activation and subsequent translocation of pErk-2 to the nucleus, or translocation of deactivated Erk-2 from the nucleus for reactivation in the cytosol, or both that would otherwise further support pro-survival signaling. Docket No. 354. PATENT

[255] In artificial constructs derived from HEK293 cells (i.e. a HEK293T cell line) that do not contain GPR-C6a, but in which gene expression produces cytosolic mutant androgen receptor (mt-AR), E-3a-diol exhibits agonist activity for transactivation of an androgen responsive element (ARE) promoter-reporter gene construct. That contrary activity indicates the importance of a functional membrane androgen receptor in order to recapitulate some of the observed anti-proliferative effects for E-3a-diol seen in LNCaP cells. For PC-3 cells, which do not express i-AR, overt anti-proliferative effect of E-3a-diol were not observed indicating the importance of a functional i-AR to mediate signaling for differentiation that leads to apoptosis.

[256] The anti-proliferative effects of E-3a-diol therefore is postulated to result in part from establishment of cross-talk with Ras-Erk signaling resulting from m-AR activation to differentially engage mutant i-AR (mt-AR) signaling from pro-survival to pro-apoptotic gene expression. Those effects are further supported by reduction of Erk-2 stimulation of anti-apoptotic gene expression. Therefore, the action of E-3a-diol at membrane AR is consistent with its role as an mutant-AR agonist that supports an anti-proliferative gene transcription program resulting from its differential engagement of mt-AR.

[257] Contacting LNCaP cells expressing T877A mt-AR with E-3a-diol did not result in the AR-phosphorylation pattern that is seen with DHT. DHT is expected, based upon the insights provided by the invention disclosed herein, to interact at GPR-C6a in similar manner to 3a-diol, i.e., through release of Ga/i. Therefore, it is believed that differences in the resulting cross-talk with Src and Ras-Erk signaling affected by E-3a-diol in comparison to 3a-diol should translate into differences in mt-AR phosphorylation patterns in the cytosol that are reflected in differences in subsequent gene transcription programs in the nucleus.

[258] For low passage (LP) LNCaP, proliferation data disclosed herein in the absence of PAED indicates the importance of i-AR activity separately from that of m-AR. It is believed that E-3a-diol effects in LP LNCaP cells are not mediated to the same extent by binding to i-AR in CRPC cell lines due to reduced or insignificant operation of the outlaw pathway in LP LNCaP. That pathway is thought required for increased sensitivity of i-AR in CRPC cells in order in this setting for there to be the observed AR transactivation for this unnatural agonist. Furthermore, it is important to note that when E-3a-diol is combined with pAED potent inhibitory activity on LP LNCaP tumor cells in vitro is observed and this activity translates to anti-tumor activity in vivo). Thus, the action of pAED in the presence of E-3a-diol indicates the role of pAED as an AR activator that supports pro-apoptotic gene transcription resulting from differential engagement of this nuclear hormone receptor Docket No. 354. PATENT in the presence of pAED alone. Since pAED weakly induces i-AR transactivation that activation may in part be due to direct interaction with i-AR, whose activity is sufficiently enhanced to support proliferation by interaction of pAED at m-AR. That enhancement (due to Ga/i- and G3/Y-mediated signaling) results from PI3K activation and aberrant crosstalk of the PI3K-Akt and Ras-Erk signaling pathways with i-AR. The support by PAED for pro-apoptotic gene transcription induced by E-3a-diol in LP LNCaP instead of proliferation is due to selective activation of the Ras-Erk pathway affected by that compound's alternative engagement of m-AR (due to Ga/q-mediated signaling). As previously discussed, the pre-requisite for cytosolic AR activity in HP LNCaP is provided by E-3a-diol acting at mt-AR sensitized by Src phosphorylation (and is in conjunction with its activity at m-AR).

[259] In order to maintain the differentiated state in normal cells, i-AR signaling is required, which apparently becomes subverted so as to promote proliferation in cancers dependent on aberrant i-AR signaling. Without being bound by theory it is believed that E-3a-diol re-establishes i-AR signaling for differentiation and is thus catastrophic for cancerous cells, and is concommitant with the switch from target activation by pErk-2 to that of pErk-1 localized to a cytosolic subcellular compartment. Initial increases in prostate specific antigen (PSA), whose levels are normally not considered relevant in late stage CaP, occurs during start of human therapy with E-3a-diol. It is therefore believed the initial PSA increases are due to that forced switch towards differentiation.

[260] Without being bound by theory, it is believed the differences in the magnitude of increases in AR-mediated transcription between C4-2B and LuCaP-35V cells (Examples 4-5) resulting from contact of these cells in a suitable test system with E-3a-diol may be attributed to the differences in effects from E-3a-diol acting upon mutated (in C4-2B) and wild-type AR (in LuCaP-35). That attribution to differing effects on i-AR is supported by the greater transactivation (IC ₅₀ = 0.48 ± 0.69) observed with HEK293T cells transiently co-transfected with an ARE-promoter luciferase reporter construct and a cDNA expression vector encoding full-length LNCaP mt-AR than would be expected based upon the binding affinity of E-3a-diol to wt-AR (IC ₅₀ = 1 .5 μΜ) (Example 7). In that artificial in vitro cell- based test system, expression of GPR-C6a is not expected (i.e., no endogenous GPR- C6a protein is present).

[261] Based upon results described herein with a suitable in vivo test system using xenograft LuCaP-35V cells (Example 5), which express wt-AR, it is believed that added PAED is required to elicit AR signaling in these cells. That AR signaling is then redirected to support apoptosis as a result of selective activation of Erk-1 in comparison to Erk-2 Docket No. 354. PATENT resulting from E-3a-diol's effect on the Ras-Erk signal transduction pathway. In contrast, it is believed mt-AR expressing C4-2B cells (Example 4) are able to utilize E-3a-diol as a partial agonist to provide the required AR signaling, which is subsequently redirected towards inducing apoptosis. That belief is based upon the capacity of the T877A mutant to accommodate steroids having a C-linked substituent at position 17 as found in some C21 steroid mt-AR activators.

[262] Therefore, for cancer cells containing wt-AR or a mutant AR, such as T877A mt- AR, and which are known or expected to have the m-AR (either endogenously or through genetic engineering), anti-proliferative effects of E-3a-diol or test compound with qualitatively or quantitatively similar activity may result from redirection of AR signaling. The redirected AR signaling may be elicited by agonist binding of an endogenous or supplemented androgen to wt-AR or in the case of the T877A mt-AR from E-3a-diol partial agonist binding to this mutant AR. Suitable test systems thus include cancer or transformed cells that have or are engineered to have functional mt-AR protein with functional GPR-C6a protein, or wt-AR protein (preferably in the presence of βΑΕϋ) with functional GPR-C6a protein, in order to recapitulate one or more of the cellular activities that have been observed for E-3a-diol. Those proteins may be present either endogenously or through genetic engineering, and the cancer or transformed cell so obtained may by be incubated in growth factor- and androgen-depleted media with or without co-contact with an externally added androgen.

[263] In patients with CaP whose cancer cells are outwardly AR signaling-dependent (i.e., contain functional i-AR) there is postulated to be sufficient endogenous pAED present to provide the required AR signaling to be redirected by E-3a-diol away from pro- survival to differentiation. Although there is some decrease in pAED levels in the serum after androgen deprivation therapy (ADT), these levels remain unchanged in prostate cancer tissues [Mizokami, A. et al. (2004)]. Unfortunately, conventional therapy with hydroxyflutamide and bicalutamide has not proven effective in blocking the stimulatory activity of pAED [Miyamoto, H. et al. (1998)] in CRPC, whose cancer cells oftentimes express wt-AR. Therefore, it is believed that interruption of pAED signaling is an activity of E-3a-diol that is useful when treating prostate cancer, or other cancers that express AR, with this compound, particularly after ADT has failed.

[264] Furthermore, in cancer cells that are presumed "androgen-independent", as when patients progress after maximal androgen blockade (MAB therapy), i.e., where pharmacological or surgical castration is combined with anti-androgen therapy, it is believed that sufficient residual DHT remains to activate i-AR and that aberrant cross-talk Docket No. 354. PATENT with other signal transduction pathways are able to sensitize i-AR to these residual levels. For example, in normal prostate cells, androgens downregulate ErbB signal transduction pathways in order to avoid excessive growth. However, due to inappropriate cross-talk between ErbB receptor and AR signaling in androgen-independent CaP, ErbB receptor signaling is upregulated to promote AR transactivation through stabilization of AR protein and enhanced AR binding to promoters of androgen-regulated genes. Those effects probably result from phosphorylation upon ligand binding of HER2/neu (or ErbB2, a member of the ErbB receptor family whose gene is often overexpressed in breast and ovarian cancers) to activate MAPK and PI3K pathways, which stimulates AR [Traish, A.M. and Morgentaler, A. (2009); Berger, R. et al. (2006)].

[265] In CWR-R1 (a CRPC cell with functional i-AR) it was found that in the presence of 100 pM DHT, EGF and heregulin-induced AR transactivation of androgen-inducible reporter was 10 fold greater than in its absence indicating the importance of AR signaling in CRPC that is supportable by residual serum DHT found after androgen ablation [Gregory, C.W. et al. (2005)].

[266] In PC-3 cells engineered to contain functional wt-AR, 100 pM DHT rapidly increased enzymatic activity of Erk-1/2. This nongenomic effect by AR required PI3K and PKC activity and did not involve activation of Jnk or p38 kinases [Peterziel, H., et al. (1999)]. In DU-145 cells engineered to express AR and overexpress HER2/neu, enhanced phosphorylation of Erk-1/2 and the NTD of AR has been reported, which is abated by HER2/neu inhibitors [Sugita, S. (2004)]. In non-prostatic cells, the LBD or AR is sufficient to stimulate anti-apoptotic effects through Src/Shc/Erk scaffolding with HER2/neu receptor [Kousteni, S. et al. (2001 )]. Thus, rapid signaling from AR activation is now thought to occur in MCF-7 and LNCaP cells from physical interaction of phosphorylated EGFR with a AR/ER/Src complex and requires ER tyrosine-537 phosphorylation elicited by EGF [Migliaccio, A. et al. (2005)].

[267] Thus, ErbB2 or EFGR activation results in AR transactivation by inducing phosphorylation of AR in the NH ₂ -terminal domain by Erk-2 and subsequent AR activation of Erk-1/2 in CaP and other AR signaling-dependent cancer cells to stimulate proliferation [Zhu, X., et al. (1999); Peterzeil, H. et al. (1999); Migliaccio, A. et al. (2000)]. Those are rapid effects (within 2 min.) distinct from AR genotropic effects (i.e., they are non-genomic effects involving i-AR). Phosphorylation of the steroid receptor co-activator-1 (SRC-1 ) by Erk has also been found required for optimal ligand-independent activation of AR through Src phosphorylation [Yeh, S. et al. (1999); Ueda, T. et al. (2002)]. Since AR increases Erb2 expression, which in turn can activate AR through phosphorylation of its NTD at Docket No. 354. PATENT amino acids 51 1 -515, a feed forward loop that supports survival and proliferation of CRPC cells becomes established.

[268] Furthermore, other transcriptional activities besides AR transactivation that promotes survival or proliferation may increase in dominance during progression to androgen-independence. For example, in LNCaP and C4-2 (an androgen-independent cell line derived from LNCaP also containing the T877A mt-AR), phosphorylation of Akt Ser-473 and Thr-308 due to PI3K activity [Burgering, B.M. and Coffer, P.J. (1995)] results in Akt phosphorylates of i-AR at Ser-210 and Ser-790 to induce proliferation [Wen, Y. et al. (2000)], thus establishing cross-talk between PI3K-Akt and AR signaling pathways. Activated Akt is also able to deactivate pro-apoptotic proteins including Bad, capase-3, GSK-3P (through β-catenin) and forkhead transcription factors (through p27 ^Kip1 ) through phosphorylation of these effector proteins. However, cell signaling pathways leading to cell proliferation diverge with C2-4 cells that require phosphorylation of p70 S6 kinase, which is an independent downstream effector of PI3K and promotes cell cycle progression through activation of transcription factors downstream of this protein kinase [Ghosh, P.M. et al. (2005)]. PC-3 cells, which are AR ^1' , also depend on p70 S6 kinase activity for proliferation.

[269] Without being bound by theory, further contributions to anti-proliferative effects of E-3a-diol and the prevention or retardation of the transition from androgen-dependent CaP to CRPC is believed to occur through selective or differential sequesterization of Erk- 2 in comparison to Erk-1 that inhibits formation of pErk-2. That is consistent with the observations described herein where Erk-1 phosphorylation increases relative to Erk-2. Furthermore, retarding pErk-2 formation should inhibit phosphorylation of AR or its co- activators that sensitize i-AR to residual DHT and disrupt the feed forward loop established in CRPC between AR and HER2/neu signaling. Additionally, reducing PI3K activity should disrupt cross-talk between PI3K-Akt and AR signaling mediated by Src. Thus, AR transactivation in intracellular AR signaling-dependent CaP cell lines, or nongenomic AR signaling that supports survival or proliferation, is inhibited or is redirected to support differentiation and apoptosis. Negative modulation of PI3K activity also inhibits activation of other pro-proliferative transcription factors stimulated by Akt.

[270] Therefore, suitable test systems further include those comprising GPR-C6a ^+/+ AR ^+/+ (endogenously or by genetic engineering ) cancer or transformed normal cells that are EFG responsive. Suitable test systems also include cancer cells or transformed cells that have or are engineered to have functional i-AR and GPR-C6a, with or without expression or engineered overexpression of an Erb receptor gene such as HER2/neu. A Docket No. 354. PATENT test or candidate compound in one or more of these suitable test system that is found to negatively modulate the activity of PI3K or Akt or negatively modulate the phosphorylation state of Akt or the NTD of i-AR would exhibit an effect or activity qualitatively similar to E- 3a-diol and thus may be criteria for identifying candidate compounds. Additionally, a test or candidate compound that negatively modulates the phosphorylation state of an effector protein downstream of Akt, such as that of Bad, capase-3 or GSK-3P, disrupts Src or AR scaffolding at ErbB2, negatively modulates the phosphorylation state of an AR co- activator or negatively modulations transactivation activity of a transcription factor activated through PI3K-Akt signal transduction could also exhibit a selection criteria for identifying a candidate compound.

[271] It is further believed that Erk MAPK phosphorylation that occurs in the absence of mitogenic signaling, either from external growth factors or from constitutive activity of mutated RKT, will result from GPCR activation form an agonist that releases Ga/i from the receptor, as seen for example with 3a-diol activation of GPR-C6a. Erk MAPK activation oftentimes seen in this type of GPCR agonist activation is due to Gp/γ heterodimer release after Ga/i activation. In certain GPCRs (i.e., those coupled to Ga/q) signaling through the Ras-Erk pathway results from downstream activation of Raf-1 from intermediary of re- regulated Src, as postulated for E-3a-diol activation of GPR-C6a Ga/q. Release of ΰβ/γ or Ga-subunits leading to activation of the Ras-Erk signaling pathway is normally considered pro-proliferative since no distinction in the absence of E-3a-diol will otherwise be made between Erk-1 and Erk-2 activation. It is therefore believed that selective effects on scaffolding, which is required for the aforementioned cross-talk between Ras-Erk and GPRC-6a signaling, by combined intra- and extracellular action of E-3a-diol contributes to selective increase in cytosolic pErk-1 in comparison to pErk-2 or decreased nuclear localized pErk-1/2 that is observed after contacting intracellular AR-dependent cancer cells with E-3a-diol.

[272] Based upon the experiments disclosed herein, it is likely that Erk MAPK phosphorylation from mitogenic signaling, either from external mitogenic stimulation or from constitutive activity of mutated RKT, will be exacerbated by agonist activation of GPR-C6a-Ga/i signaling. However, it is believed that agonist activation of GPR-C6a coupled to Ga/q, as postulated for "inverse agonist" binding of E-3a-diol to this receptor, will counteract indiscriminate activation of the Ras-Erk signaling pathway occurring through external or constitutive tyrosine kinase activity that would otherwise lead to activation of both Erk isoforms. That counteraction probably results from E-3a-diol (i) re- regulating Src activity, which resulted from the external or constitutive stimulation through Docket No. 354. PATENT modulation of Src scaffolding resulting from engagement of GPR-C6a Ga/q, and/or (ii) suppressing the activation of Erk-2 in favor of Erk-1 through differentially affecting the scaffolding of these Erk MAPK isoforms.

[273] Therefore, a suitable test system will include cancer cells or transformed cells that have or are engineered to have functional i-AR and functional GPR-C6a. A test or candidate compound that positively (i) modulates Ga/q activity, (ii) an activity of a downstream effector protein consistent with this positive modulation of Ga/q activity, or (iii) positively modulates Ga/q phosphorylation state or GTP binding, or negatively modulates (iv) Ga/i activity or (v) activity of a downstream effector protein in a manner consistent with this negative modulation in one or more of these suitable test systems would exhibit an effect or activity qualitatively similar to E-3a-diol and thus may be criteria for identifying a candidate compound. Furthermore, a test or candidate compound that (i) exhibits "inverse agonist" activity at GPR-C6a, (ii) competes with 3a-diol binding at this receptor to inhibit Ga/i release or (iii) opposes an activity or modulation of phosphorylation state effected by 3a-diol in one or more of these suitable test systems would exhibit an effect or activity that may become other criteria for identifying candidate compounds. Additionally, a test compound that re-regulates Src activity on AR or Raf-1 when cells in these suitable test systems are stimulated by external contact with a mitogen may be criteria for identifying a candidate compound.

[274] In certain cancer cells mutated for PTEN "loss of function", the net effect resulting from the aforementioned interruptions or redirections by E-3a-diol of cross-talk between GPR-C6a, Src and Ras-Erk signal transduction pathways, is to effectively compensate for lost PTEN regulatory activity. That compensation occurs through PI3K inhibition by Ga/q signaling and through selective phosphorylation of Erk-1 , which also negatively regulates PI3K activity due to cross-talk between Ras-Erk and PI3K-Akt signal transduction pathways. Cross-talk between these two signal transduction pathways is believed to be mediated in part through the negative feedback mechanism by pErk-1 , which is selectively increased in comparison to pErk-2, to disengage Ras from the cytoplasmic membrane through its phosphorylation of SOS.

[275] Therefore, antiproliferative effects on PTEN mutants, RTK mutants, LNCaP or other hyperproliferating cells that constitute a suitable test system is believed to result from inhibition of PI3K activity either through Ga/q-Src scaffolding or selective formation of pErk-1 , which may be enhanced by suppression of excessive Erk-2 activity (e.g., from mutant RTK signaling), leading to apoptosis of the mutant or hyperproliferating cells. Engagement of GPR-C6a Ga/q in such test system is expected, based upon the insights Docket No. 354. PATENT provided by the invention disclosed herein, to augment PI3K inhibition as a result of effects on Src scaffolding that leads to tyrosine phosphorylation of Ga/q (GTP), which positively modulates that G protein's signaling.

[276] Another contribution to the anti-proliferative effect of E-3a-diol stemming from selective formation of pErk-1 in comparison to pErk-2 is believed due to redirection of i-AR activity from abnormal activation of Src. In normal cells i-AR signaling is required for maintenance of the differentiated state whereas in cancer cells this activity is subverted to support survival or proliferation. Without being bound by theory, it is believed that E-3a- diol re-establishes differentiation activity of i-AR by affecting the phosphorylation state of i- AR. That modulation in phosphorylation state results in redirection of aberrant i-AR activity in the nucleus towards a transcriptional program that supports differentiation or negative modulation of non-genomic i-AR effects in the cytosol that support apoptosis through re-normalization of signal transduction cross-talk. Additionally, the scaffolding of i-Ar to activated RTKs (e.g., EGFR or HER2/neu) that promotes recruitment of Src to the cytoplasmic membrane for its activation may be disrupted thereby inhibiting the feedforward activation of i-AR by the outlaw pathway.

[277] Therefore, a test compound or candidate compound that (i) negatively modulates AR tyrosine phosphorylation status or (ii) positively modulates AR serine phosphorylation status through alteration of the phosphorylation pattern in the NTD in comparison to the pattern obtained from indiscriminate Erk isoform activation, (iii) re-regulates AR nucleo- cytoplasmic translocation in cancer cells to that of the differentiated normal cell from which the cancer cells arose, (iv) negatively modulates AR co-activator binding or recruitment for transactivation of pro-proliferative genes or (v) negatively modulates transactivation of ARE-inducible genes or (vi) exhibits one or more of these effects may be criteria for identifying a candidate compound.

[278] Other preferred test or candidate compounds for consideration as a candidate compound will effect the phosphorylation state, activity or protein levels of one or more Erk effector proteins consistent with negative modulation of pErk nucleo-cytoplasmic translocation, negative modulation of Erk-2 phosphorylation state or positive modulation of cytoplasmic Erk-1 phosphorylation without adverse negative activity modulation of a pro- apoptotic protein or adverse positive activity modulation of an anti-apoptotic protein. Erk effector proteins [Chen, Z. et al. (2001 ); Roux, P.P. and Blenis, J. (2004); Yoon, S. and Seger R. (2006)] that may be effected in this manner include nuclear proteins Elk-1 and Sap-1 a (members of the Ets family of transcription factors), and TIF-IA. Erk phosphorylation of Elk-1 is typically used to assess Erk activity in suitable in vitro test Docket No. 354. PATENT systems and occurs at Ser-383 and Ser-389.

[279] Other Erk substrates are cytoplasmic proteins involved in apoptotic signaling, including FOX03a (Ser-294, Ser-344, Ser-425), p90 ^{Rsk 1} (Ser-364, Thr-574), Bim (Ser-55, Ser-65, Ser-100), capase-9 (Thr-125) and perhaps Bcl-2 (Ser-70). Those proteins and other proteins that are involved in apoptotic signaling are sometimes referred to individually or collectively as apoptotic proteins. Erk phosphorylation of one or more of those substrates in suitable in vitro test systems may also be used to asses Erk activity.

[280] Other soluble cytoplasmic proteins that are Erk substrates include cPLA ₂ (Ser- 505), p70S6 kinase (Thr-421 , Ser-424), phosphodiesterase 4D (Ser-579), MKP-3-DUSP6 (Ser-159, Ser-197), MAP kinase activated protein kinases (MAPKAPKs), which includes p90 , mitogen and stress activated protein kinases (MSKs) and MAP-integrating kinases (MNKs). Additional non-nuclear Erk targets include plasma membrane, endomembrane, mitochondrial and cytoskeletal proteins (e.g., CD120a, calnexin, paxillin, nucleoporins, EFGR, cortactin). Erk phosphorylation of one or more of those substrates in suitable in vitro test systems may also be used to asses Erk activity.

[281] Other preferred test or candidate compounds negatively modulate phosphorylation states, protein levels or activities of one or more transcription factors associated with intermediate early gene expression (e.g., c-Myc, c-Jun, JunB, JunD, c-Fos, FosB, Fra-1 , Fra-2), negatively modulates transactivation by AP-1 (dimers of proto-oncogene products encoded by jun and fos families) or negatively modulates phosphorylation state of an Ets transcription factor (e.g., Elk-1 ). Phosphorylation of c-Fos by Erk occurs at Thr-325 and Thr-331 to extend the half-life of this transcription factor and is dependent on the duration Erk signal. Prolonged Erk signaling for nuclear accumulation of pErk is required for S- phase entry [Balmanno, K. and Cook, S.J. (1999)]. Therefore, other preferred test or candidate compounds will negatively modulate the serine phosphorylation state or half-life of c-Fos.

[282] Additional preferred test or candidate compounds increase retention of pErk in the cytosol (i.e., negatively modulates pErk nucleo-cytoplasmic transport) with minimal or no positive modulation of ser/thr phosphorylation status in one or more of the apoptotic proteins described herein at least one of the indicated residues (in parentheses) such that pro-apoptotic signaling is not negatively modulated or anti-apoptotic signaling is not positively modulated to an extent that cell survival is promoted or anti-proliferative effects from negative modulation of PI3K p85 regulatory subunit tyrosine phosphorylation are not abrogated or adversely diminished.

[283] 2(i) Modulation of gene transcription: Docket No. 354. PATENT

[284] 2(i)(1 ) In vitro gene transcription: E-3a-diol is believed to induce a change in the gene transactivation pattern by i-AR away from proliferation to supporting differentiation- induced apoptosis. An example of that switch in gene expression patterns, is the decreases expression of IGFBP-3 (i.e. negative modulation )in LNCaP cells (which express the T877A mutant AR) and in MDA-MD453 breast cancer cells (which express the wild-type AR) by E-3a-diol. IGFBP-3 has been implicated in cell growth inhibition and initiation of apoptosis in various cancer cell types and androgens.

[285] As reported elsewhere androgens can increase expression of IGFBP-3 protein in LNCaP cells and do so via transactivation of an androgen response element present in the promoter of the IGFBP-3 gene. DHT stimulates production of IGFBP-3 protein whereas E-3a-diol eliminated the DHT response. In contrast, MDA-MB453 cells express the wt-AR and respond to androgens by reducing IGFBP-3 protein expression. This effect was reversed by adding the androgen receptor antagonist Casodex™, or by E-3a-diol, which restored the levels of IGFBP-3 to the same levels achieved by Casodex. The results of that study (see Example 2) indicate E-3a-diol is acting similarly to an AR antagonist with respect to IGFBP-3 expression.

[286] The switch from pro-survival i-AR transactivation program and Erk-2 activation to a pro-apoptotic i-AR transactivation program that is dependent on selective Erk-1 activation and/or inhibition of Erk-2 activation is also consistent with differences observed in the apoptosis profiling, signal transduction pathway finder and drug resistance and metabolism arrays described herein. Various androgen-regulated cell cycle and signal transduction genes modulated by E-3a-diol include CDK2, CDKN1A, KLK2, ODC1, IGFBP-3, TMEPAI, GREB1 and AR. Expression of genes involved in various phases of the cell cycle including G1 phase and G1/S transition effected by E-3a-diol include CCNE1, CDK4 and CDKN1B as well as cell cycle checkpoint and cell cycle arrest (ATM & CHEK2). Among the genes involved in apoptosis and cell senescence, Bcl-2 and Caspase-9 were the most notable. Bcl-2 expression was down-regulated in most samples treated with E-3a-diol. Also down-regulated were ABCG2 and ABCC5, both members of the ABC transporter family involved in drug resistance to cancer chemotherapy.

[287] In some embodiments a test or candidate compound that is identified as a further characterized candidate compound affects transcription of one or more genes of Table 2-5 in qualitatively similar manner. In preferred embodiments the further characterized candidate compound down-regulates Bcl-2 or IGFBP-3 expression; more preferably down-regulates both of these genes.

[288] 2(i)(2) In vivo gene transcription effects: E-3a-diol induced changes in gene Docket No. 354. PATENT expression in prostate tumor xenografts described herein (Example 3) that include numerous changes in AR-regulated genes (i.e., genes under transcriptional control from an upstream ARE) and other gene changes, as provided in Table 6, resulting from contacting a test compound, which is selected as a candidate compound, to a suitable in vivo test system, [see Koreckij, T.D. et al. (2009)].

[289] AR, the proteins levels from which are often found increased in CRPC, is down- regulated by E-3a-diol. AR (i.e., cytosolic androgen receptor) activates the AP-1 (activated protein-1 ) transcription factor, contributing to changes in prostate cancer cell growth. Erk-2 but not Erk-1 was found to complex with AP- 1 [Kumar, N.V. et al. (2001 )]. That observation is consistent with sequesterization of Erk-2 in inactive form due to intracellular action of E-3a-diol.

[290] Another gene down-regulated by is IRX5 (Iroquois homeobox protein 5 inhibits apoptosis, and promotes cell cycle progression in LNCaP prostate cancer cells.

[291] Another down-regulated gene is JUN (v-jun sarcoma virus 17 homolog), which plays a pivotal role in the pathway that connects ligand-activated AR to elevated Ets variant gene 1 (ETV1) expression, leading to enhanced expression of matrix metalloproteases and prostate cancer cell invasion.. JUN is an AR coactivator that stimulates AR transactivation.

[292] Another affected gene that is down-regulated is CD44, which acts as an adhesion molecule for invasion, and a docking receptor for matrix metalloproteinase-9. CD44 is also a receptor for osteopontin, which is involved in prostate tumor metastasis. CD44 also acts as a scaffold protein for the Sialyl Lewis X binding determinant for endothelial cell E- selectin during metastasis. CD44 expression is high on disseminated breast tumor cells in bone marrow, on prostaspheres and prostate cancer stem cells, and potentiates the adherence of metastatic prostate and breast cancer cells to bone marrow endothelial cells, and invasion and growth. In prostate cancer cells, CD44 is associated with the neuroendocrine tumor stem cells. Those cells are more proliferative, clonogenic, tumorigenic and metastatic than the CD44- cells. CD44 is also expressed on colon cancer stem cells, which can reconstitute the original human tumor in vivo, and on pancreatic cancer stem cells. Expression of CD44 variants is correlated with drug resistance during prostate cancer metastasis. Stromal hyaluronan interaction with epithelial CD44 variants promotes prostate cancer invasion by augmenting expression and function of hepatocyte growth factor and androgen receptor [Ghatak, S. et al. (2010)].

[293] Another gene down-regulated by contacting E-3a-diol with LNCaP cells is JAG1. Jagged 1 [Alagille syndrome] and Notchl ligand (JAG1 ) protein levels are increased in Docket No. 354. PATENT prostatospheres, representing tumor-initiating cells. Inhibition of Notch signaling increases susceptibility to apoptosis and targeted knockdown of Notchl inhibits invasion of human prostate cancer cells. Thus, down-regulation of Notchl and Jaggedl inhibits prostate cancer cell growth, migration and invasion, and induces apoptosis via inactivation of Akt, mTOR, and NF-κΒ signaling. Both CD44 and NOTCH are present in breast cancer stem cells, as well as prostate cancer stem cells and Notch signaling is important for the osteomimetic properties of prostate cancer bone metastatic cell lines.

[294] In contrast to the above down-regulated genes, RIS1 is up-regulated by contacting E-3a-diol with LNCaP cells. Expression of RIS1 (Ras-induced senescence 1 ) can be silenced in prostate cancer, and is decreased in non-small cell lung cancer. RIS1 is mutated in melanoma and colorectal cancers.

[295] Another up-regulated gene is TIMP2 (TIMP metallopeptidase inhibitor 2) whose overexpression in prostate cancer is associated with longer disease free survival. Another gene up-regulated by E-3a-diol is RUNX1. Runt-related transcription factor 1 (AML1 oncogene) was specifically bound to and activated the PSA regulatory region in chromatin immunoprecipitation assays. A RUNX1 polymorphism was significantly associated with a higher risk for advanced pathologic stage, a higher risk for lymph node metastasis, and poorer PSA-free survival. RUNX1 induces senescence-like growth arrest in primary murine fibroblasts.

[296] Another up-regulated gene is CASP10. Capase-10 is involved in prostate cancer cell apoptosis induced by various pro-apoptotic agents.

[297] Another gene up-regulated by E-3a-diol is /.OX (lysyl oxidase), whose polypeptide product has been shown to inhibit Ras signaling and the transformed phenotype of prostate and other cancer cells, and to sensitize pancreatic and breast cancer cells to doxorubicin-induced apoptosis. LOX expression is progressively lost in primary human prostate cancer and associated metastatic lesions.

[298] Without being bound by theory, it is believed that E-3a-diol affects expressed of the aforementioned genes by inhibiting pErk-2 activity in the nucleus through sequesterization of unactivated Erk-2 in the cytoplasm or deactivated Erk in the nucleus (or both) and by modulating the phosphorylation status of i-AR or of an AR co-regulator whereby transcriptional activity of agonist-bound AR is redirected to support differentiation.

[299] Based upon the in vitro and in vivo effects of E-3a-diol on gene expression as shown in Tables 2-6 and the roles these genes play in survival and proliferative signaling, Docket No. 354. PATENT particularly with respect to Bcl-2, IGFBP-3, AR, IRX5, JUN, CD44, JAG1, RIS1, TIMP2, RUNX1, CASP10 and LOX, support the contention that E-3a-diol is useful for treating the various hyperproliferation conditions disclosed herein that are dependent on AR signaling or are androgen-associated. Furthermore, results of the oligo arrays and real-time PCR results presented in Example 3 show that levels of AR transcript become decreased by treatment with E-3a-diol in cancer cells comprising a xenograft in vivo test system. Those effects correlated with the effects of E-3a-diol on levels of AR protein.

[300] Therefore, a test or candidate compound that effects transcription of one or more genes in qualitatively similar manner observed for E-3a-diol in some embodiments is a property for identifying a candidate compound. In preferred embodiments, expression or more of AR, IRX5, JUN, CD44 and JAG1 (decreased expression), and RIS1, TIMP2, RUNX1, CASP10 and LOX (increased expression) are affected in the manner indicated.

[301] Association of E-3a-diol to transcription factors other than nuclear hormone receptors was explored (Example 18) with interactions between cRel and Rreb 1 being the most notable that were detected. cRel is a class II member of the Rel/NF-κΒ family of transcription factors. All Rel family members contain an NH ₂ -terminal Rel Homology Domain (RHD) that is involved in dimerization, DNA and ΙκΒ binding, and nuclear localization. Activation of cRel from TNF-a signaling is dependent on PI3K and PLC- ζ phosphorylation of serine residues in the COOH-terminal domain that individually result in distinct, but overlapping, phosphorylation patterns. Furthermore, cRel contains a PKA recognition site in its RHD and an Erk consensus site in its COOH-terminal domain. Bcl _XL is likely to be a direct target for cRel activation. Under normal conditions, cRel is involved in differentiation and apoptosis during T-cell and B-cell maturation including thymic differentiation of Foxp3 ⁺ Treg cells. However, in cancer cells cRel activity becomes aberrant, due to over-expression of cRel and/or loss of function mutations in the gene encoding ΙκΒβ, such that cRel retention by ΙκΒβ in the cytoplasm is impaired. Therefore, as in the case of i-AR, the phosphorylation status of cRel, particularly with respect to the phosphorylation pattern in the C- and NH ₂ -terminal domains, is affected by activities of several protein kinases that include Erk1/2, PI3K, PKA and PLC-ζ.

[302] Therefore, without being bound by theory, E-3a-diol may affect the phosphorylation pattern (i.e., positively modulates phosphorylation status) in cRel in analogous manner to i-AR thereby switching its aberrant activity back towards supporting differentiation and promoting apoptosis or by inhibiting proliferation by binding cRel translocated to the nucleus to components of the transcriptional machinery in non- productive complexes. Therefore, a test or candidate compound that affects the Docket No. 354. PATENT phosphorylation state of cRel in like manner to E-3a-diol is in some embodiments a property for identification of candidate compound.

[303] Rrebl binds and represses expression of the p16(lnk4a) promoter. The p16INK4a (p16) gene is a tumor suppressor involved in regulating cell cycle checkpoints. The p16 protein specifically binds to and inhibits the cyclin-dependent kinases CDK4/6, which regulate cell cycle progression in G1 through phosphorylation of the retinoblastoma protein (pRb) [Goodrich, D.W. et al. (1991 ); Ewen, M.E. et al. (1993); Kato J.-Y. et al. (1993)]. Without being bound by theory, E-3a-diol may affect Rrebl binding activity to inhibit cell cycle progression from d phase. Therefore, a test or candidate compound that interacts with one or more transcription factors in qualitatively similar manner observed for E-3a-diol or which arrests tumor cell cycle in Gi is in some embodiments an additional property for identification of a candidate compound.

[304] 2(k). Modulation of Phosphoprotein Recognition by SH2 Domains: Signal transductions that regulate cell proliferation and differentiation are often transmitted via activation of tyrosine kinases, and increased catalytic activity of tyrosine kinases is often seen in human diseases. Recruitment of adaptor and scaffold proteins to RTKs and nonreceptor protein kinases to regulate catalytic activities of these kinases is also influenced by the phosphorylation states of the kinases and-or that of the adaptor or scaffold proteins. Protein-protein interactions are therefore important in signal transduction and are often mediated by non-catalytic, conserved domains. One of those domains is the SH2 domain.

[305] E-3a-diol was found to modulate phosphoprotein recognition by SH2 domains as shown by Example 13. Because of the binding specificity of the SH2 domain to phosphorylated tyrosine residues, a specific pattern or "fingerprint" of tyrosine phosphorylation can be elucidated. The use of SH2 domains for tyrosine phosphorylation profiling focuses on specific pathways and thus avoids binding preference to dominant populations of tyrosine kinases. The use of SH2 profiling arrays for fingerprinting tyrosine kinase activity within cells experiencing aberrant signal transduction by Western blot analysis are further described in Nallou, P. and Mayer, B.J. (2001 ). Other methods are provided by Machida, K. et al. (2003), which are incorporated by reference herein. Such overall negative modulation of pTyr phosphoprotein recognition may be due to physical disruption of binding to a specific SH2 domain to its cognate phosphoproteins, decreased pTyr phosphorylation of the proteins to be subsequently recognized by a specific SH2 domain, decreased protein levels of the proteins that are to be tyrosine phosphorylated for production of the cognate phosphoproteins or any combination of such effects. Docket No. 354. PATENT

[306] One significant effect due to E-3a-diol activity on SH2 domain phosphoprotein recognition is for the p85 regulatory subunit of PI3K, specifically pTyr-phosphopeptide recognition by the NH ₂ -terminal SH2 domain (D1 ). That domain is predominately responsible for interaction of p85 with mitogenic-induced phosphoproteins that activate PI3K p100 catalytic activity. The observed overall negative modulation in phosphoprotein recognition by the PI3K regulatory subunit's NH ₂ -terminal SH2 domain occurred within 5 minutes and persisted for at least 15 minutes. After 1 h the extent of phosphoprotein recognition by the p85 N-SH2 domain returned to control levels. Those time periods are consistent with non-genomic signaling of E-3a-diol through its inverse agonist activation at GPR-C6a that releases Ga/q (GTP). Released Ga/q (GTP), whose activity is enhanced by tyrosine phosphorylation by Src, binds directly to p100a/p85 to displace Ras(GTP). Thus, decreased p85 N-SH2 domain recognition of its cognate phosphoproteins disrupts sequential activation of PI3K catalytic activity by Src and Ras by redirecting Src kinase activity to Ga/q phosphorylation and by expelling Ras from its protein complex with PI3K. Those disruptions in p85 N-SH2 domain interactions induced by E-3a-diol will therefore negatively modulate PI3K catalytic activity and re-regulate or normalize aberrant or excessive cross-talk between PI3K-Akt and Ras-Erk signal transduction pathways to exert a pro-apoptotic or anti-proliferative effect.

[307] As shown by Figures 6 and 7, at t = 5min and/or t = 15 min time points overall negative modulation of SH2 recognition of pTyr phosphoproteins was also found for the Src-family kinases Hck, Lck and Yes, the adaptor proteins Nck-2, BRDG-1 and CrkL and the non-receptor tyrosine kinases Abl and Btk, whereas overall positive modulation of SH2 recognition of pTyr phosphoproteins by the COOH-terminal SH2 domain (C-SH2 or D2) of PLC-Y1 , the SH2 domain in adaptor protein Shc-2 and the NH2-termianl SH2 domain (N- SH2 or D1 ) of GTPase binding protein RasGAPI was observed.

[308] Thus, redirection of Src signaling resulting from E-3a-diol-induced Ga/q activation will (1 ) negatively modulate PI3K-Akt signal transduction (from positive modulation of Ga/q Tyr-356 phosphorylation or negative modulation of p85 Tyr-688 phosphorylation or a combination of these modulation effects), (2) negatively modulate RasGAP-RhoGAP cytosolic signaling (from negative modulation of RhoGAP Tyr-1087 and/or Tyr 1 105) and/or (3) positively modulate RasGAP translocation to the plasma membrane (from positive modulation of RasGAP Tyr-460 phosphorylation). One or more of those effects from redirection of Src kinase activity is believed due to isoform selective β-arrestin scaffolding of Src (vide infra) that occurs subsequent to GDP-GTP exchange in Ga/q that results from inverse agonist activity of E-3a-diol at GPRC6a. Docket No. 354. PATENT

[309] In consideration of the foregoing, a candidate compound may be characterized as positively modulating p-Tyr phosphoprotein recognition by the N-SH2 domain of RasGAP optionally with no discernable effect of such recognition by the C-SH2 domain. Preferentially this effect on N-SH2 phosphoprotein recognition is observable at about 15- min upon contacting the candidate compound to a suitable test system, preferably with a significantly smaller or no observable effect on this recognition with 5 min. In addition, a candidate compound may be characterized by positive modulation of RasGAP phosphorylation state, preferably with negative modulation of Ser/Thr phosphorylation state or negative modulation of RhoGAP tyrosine phosphorylation state, more preferably by inducing both of these modulations.

[310] 3. Summary of invention embodiments

[311] The following provides a summary of exemplary invention embodiments and biological activities related thereto and is not meant to limit the invention on any way.

[312] It is believed that E-3a-diol antagonizes the pro-proliferative effects of 3a-diol, which is an androgenic compound in its own right, and perhaps the pro-proliferative effects of the powerful androgen DHT (to which 3a-diol is converted by oxidase activity of a 3a-hydroxysteroid dehydrogenase). Without being bound by theory, it is believed this antagonism results from a combination of intracellular and extracellular interactions initiated or induced by E-3a-diol.

[313] The extracellular interaction of E-3a-diol is believed to occur by engaging a GPCR such that Ga/q (GTP) is formed, which then negatively modulates the tyrosine phosphorylation state of PI3K p85 regulatory subunit. In addition to that G protein- dependent signaling, it is further believed that G protein-independent signaling subsequent to Ga/q activation occurs whereby negative nucleo-cytoplasmic of pErk-1 is effected, presumably through scaffolding within a protein complex. That cytoplasmic sequesterization of pErk-1 is believed to occur to a subcellular cytosolic domain that does not allow for phosphorylated- deactivation of pro-apoptotic proteins or activation of anti- apoptotic proteins that would otherwise support anti-apoptotic signaling. The cytoplasmic localization is not considered or required to be absolute, but permits diminished pErk translocation to the nucleus in order that a differentiation program may be engaged as well as permitting non-genomic signaling of cytosolically retained pErk to support this same apoptotic-promoting program. It is believed the aforementioned effects on localization of signal transduction components resulting from GPRC-Ga/q engagement are due to inverse agonist binding of E-3a-diol to GPR-C6a. It is further believed that in PTEN deficient tumor cells with functional p53, cross-talk between GPRC-6a-Ga/q and Docket No. 354. PATENT

Ras-Erk signaling will selectively reactivate Ras-Erk signal transduction whereby apoptosis is promoted, probably as a result from negatively modulation of Mdm2 binding to p53 (i.e., positive modulation of functional p53 activity).

[314] The intracellular interaction of E-3a-diol is believed to occur through stabilization of a protein complex comprising Erk-2 and a scaffold protein or Erk-2 and a dual-substrate phosphatase (which may act as a scaffolding protein). That protein complex stabilization preferentially sequesters Erk-2 to a subcellular compartment to inhibit its phosphorylation due to activation of the Ras-Erk signal transduction pathway (which may result from activation of GPR-C6a-Ga/q signaling through the aforementioned signal transduction cross-talk), thus negatively modulating Erk-2 phosphorylation state, negatively modulating nucleo-cytoplasmic translocation of pErk-2 that may be formed as a result of this crosstalk or preventing or inhibiting translocation of deactivated Erk-2 out the nucleus for reactivation in the cytosol.

[315] Thus, it is believed that the combined intracellular and extracellular actions of E- 3a-diol re-distributes the Erk isoforms through these aforementioned localization effects to increase cytosolic Erk signaling at the expense of transcription factor transactivation of pro-proliferative genes resulting from nuclear activity of pErk while negatively modulating anti-apoptotic activity of pErk in the cytosol. One such consequence from Erk isoform redistribution is to positively modulate the phosphorylation state of actAR or a co-regulator thereof such that aberrant or excessive i-AR signaling in androgen-associated cancer cells is re-regulated in a manner that promotes execution of a differentiation gene transcriptional program leading to apoptosis instead of survival or proliferation.

[316] Furthermore it is additionally believed that E-3a-diol antagonizes the interconversion of 3a-diol and DHT by inhibiting a 3a-hydroxysteroid dehydrogenase believed to be 17HSD10 to negatively modulate prostatic tissue levels of DHT

[317] Consistent with the foregoing theories and biological activities described herein for E-3a-diol and 3a-diol and other test compounds disclosed herein and the disclosed exemplary working examples, the following non-limiting invention embodiments are presented.

[318] The invention provides screening methods in vitro or in vivo that determines one or more of the aforementioned characteristic effects reported herein for E-3a-diol on kinase signal transduction cascade(s). Such test compounds can be anti-proliferative candidate compounds and thus these screening methods would be useful for identifying compounds having anti-proliferative or anti-inflammatory activity(ies) in vivo. Those compounds will Docket No. 354. PATENT typically have low toxicity in vivo, and thus may be particularly useful in comparison to existing therapies for treating a human experiencing a hyperproliferation conditions as disclosed herein, and may additionally be useful in treating unwanted inflammation that initiates or supports these conditions. Additionally, a test or candidate compound found to have a characteristic effect or activity reported herein for E-3a-diol (i.e., desired antiproliferative or pro-apoptotic) or 3a-diol (i.e., undesired pro-proliferative or pro-survival) on kinase signal transduction cascade(s) would be useful as a reference compound or positive or negative control for the effect or activity screened.

[319] The invention further provides methods to identify compounds that are selective or differential modulators of Erk-2 and Erk-1 activity. Test or candidate compounds so identified will usually exhibit low toxicity, e.g., liver, skin or CNS activity, in contrast to ATP binding site-dependent MAPK inhibitors, when administered in therapeutic effective amounts to humans for treating hyperproliferation conditions disclosed herein.

[320] The invention further provides screening methods that identify one or more of the aforementioned characteristic effects on Ras-Erk signaling, as described herein for E-3a- diol resulting from contact of a test compound with a suitable test system that would be useful in selecting candidate compounds or identifying further characterized candidate compounds having such effects. Those screening methods thus would prove useful in identifying compounds having anti-proliferative activity of low liver toxicity in order to treat humans experiencing the hyperproliferation conditions disclosed herein. The test or candidate compounds so identified may also have anti-inflammatory effects that would ameliorate unwanted inflammation that initiates or supports these hyperproliferation conditions in a mammal.

[321] The invention additionally provides screening methods that identifies one or more of the aforementioned characteristic effects on signal transduction pathways or signal nodes including (i) Ras-Erk, (ii) TNFa-NF-κΒ, (iii) PI3K-Akt, (iv) Src or (iv) i-AR, as described herein for E-3a-diol resulting from contact of a test compound with a suitable test system. The characteristic effects so identified may be used to select test or candidate compounds. Screening methods to determine those characteristic effects would be useful in identifying drug candidate compounds (i.e., candidate low toxicity Erk modulators) that are expected, based upon the insight of the present invention, to have anti-proliferative activity and low toxicity (e.g., low liver toxicity). The candidate compounds would thus be useful in treating a mammal having the hyperproliferative condition disclosed herein and/or for treating unwanted inflammation that initiates or supports these hyperproliferation conditions in a mammal. Docket No. 354. PATENT

[322] The invention further provides screening methods that identifies one or more of the aforementioned characteristic effects on GPR-C6a-Ga/q signaling for selecting candidate having such effects and thus would prove useful in identifying compounds having antiproliferative and optionally anti-inflammatory activity(ies) of low liver toxicity in order to treat a mammal experiencing the hyperproliferation conditions as disclosed herein and those compounds additionally treating unwanted inflammation that initiates or supports these conditions.

[323] The invention also provides methods to identify agonists of GPR-C6a that oppose the proliferative effect of 3a-diol by this compound's action at this receptor (i.e., inverse agonists of 3a-diol signaling at GPR-C6a).

[324] The invention additionally provides for methods to identify non-ATP binding site- dependent ligands selective for Erk-1/2 in comparison to p38 and JNK MAPKs.

[325] The invention further provides methods to identify non-ATP binding site-dependent ligands selective for Erk-2 in comparison to Erk-1 .

[326] The invention also provides methods to identify differential modulators of Erk-1 and Erk-2 phosphorylation such that activation of Erk-1 occurs in preference to Erk-2.

[327] The invention also provides methods to identify differential modulators of Erk-1 and Erk-2 phosphorylation such that inhibition of Erk-2 activity occurs in preference to inhibition of Erk-1 activity.

[328] The invention additionally provides methods to identify differential modulators of Erk-1 and Erk-2 phosphorylation such that activation of Erk-1 occurs in preference to activation of Erk-2 activity.

[329] The invention additionally provides methods to identify modulators of protein kinase activit(ies) whereby excessive tyrosine phosphorylation of Raf-1 resulting from external mitogen stimulation or constitutive RTK activity is negatively modulated.

[330] The invention further provides methods to identify modulators of protein scaffolding mediated by scaffold proteins with one or more other components of kinase signal transduction pathway(s) whereby selective or differential modulation of Erk-1 and Erk-2 occurs to reduce survival or decrease proliferation of cancerous cells, preferably by additionally reducing signal transduction through one or more pro-inflammatory signal transduction nodes.

[331] The invention further provides screening methods that identifies modulators of cytosolic AR (wt-AR or mt-AR) phosphorylation to redirect activity of this nuclear hormone Docket No. 354. PATENT receptor away from supporting proliferation towards supporting differentiation and hence apoptosis.

[332] The invention additionally provides screening methods that identifies negative modulators of undesired non-genomic activity of the cytosolic androgen nuclear hormone receptor or negatively modulates RTK signal transduction resulting from external mitogen stimulation of one or more mitogen-responsive RTKs or from constitutive activation of the RTK(s)

[333] The invention further provides screening methods that identifies modulators of Src whereby activity of this tyrosine kinase is directed away from supporting proliferation or directed towards supporting differentiation.

[334] The invention additionally provides methods to identify compounds that interact with the scaffold protein Tfg, or interact with 17pHSD10, or decreases 17pHSD10 conversion of 3a-diol to DHT, or decreases PI3K intracellular activity or downstream effects resulting therefrom, or increases intracellular levels of pErk-1 relative to pErk-2 or a combination of these effects wherein the intracellular activities detected occur in an artificial cell cancer cell line expressing GPR-C6a and wt-AR or mt-AR or artificial transformed cell lines containing gene constructs encoding functional GPR-C6a and wt- AR or mt-AR.

[335] The invention also provides screening methods that identifies modulators of i-AR phosphorylation status such that transcription of pro-survival or anti-apoptotic genes in cancer cells is negatively modulated.

[336] The invention further provides methods to identify compounds that induce differentiation of cancer or normal transformed cells.

[337] The invention further provides methods to identify compounds that induce arrest in proliferating cancer or normal transformed cells.

[338] The invention additionally provides methods to identify compounds that modulate PI3K or Akt phosphorylation status such that prosurvival or anti-apoptotic activity in cancer or transformed normal cells is negatively modulated.

[339] 4. Embodiments

[340] In some embodiments, the phosphorylation status of a protein kinase, e.g., Erk- 1/2, or an isoform of that protein kinase (e.g., Erk-1 or Erk-2) refers to the extent of phosphorylation of a collection of proteins present in the suitable test system for a particular protein kinase, which leads to activation or deactivation or modulation of the Docket No. 354. PATENT activity of this protein kinase or isoform thereof, as specified explicitly or implicitly by context, before or after contacting a test or candidate compound to a suitable test system.

[341] In some embodiments phosphorylation status is stated relative to the number of amino acid residues capable of being phosphorylated or the number of phosphate groups covalently attached to a phospho-protein, when the phospho-protein or protein that is to be phosphorylated is present in a suitable test system.

[342] In other embodiments phosphorylation status of one protein kinase isoform is stated relative to another isoform of the same protein kinase before or after contacting a test or candidate compound to a suitable test system containing the isoforms. In still other embodiments the phosphorylation status of a protein kinase after contact with a test or candidate compound is stated relative to the phosphorylation status of the same protein kinase in the same suitable test system prior to contact to the test or candidate compound or relative to the same protein kinase in a control test system to which is contacted the same test or candidate compound or a reference test compound.

[343] In other embodiments phosphorylation status refers to the pattern of phosphorylation of amino acid residues in a phospho-protein. In those embodiments the phosphorylation status of a protein is sometimes stated relative to one or more specified amino acid or amino-acid type (e.g., Ser/Thr or Tyr) residues that are phosphorylated or capable of phosphorylation in a suitable test system before or after contact with a test or candidate compound.

[344] In some embodiments the phosphorylation pattern in a receptor or a protein kinase prior to contact of a suitable test system containing the receptor or protein kinase with a test or candidate compound is compared with the phosphorylation pattern of the same protein in the same test system after contact with a test or candidate compound or with the same protein in a control test system contacted with the same test or candidate compound or a reference test compound. In those embodiments differences in phosphorylation patterns may not be accompanied by changes in the total number of phosphate groups covalently bonded to the phospho-protein or the overall extent of phosphorylation of a collection of such proteins. Those differing phosphorylation patterns may have in common one or more phosphorylated amino acid residues or have common un-phosphorylated residue(s) that are capable of phosphorylation in the suitable test system.

[345] In some embodiments, modulation of phosphorylation status is based upon comparison of a test or candidate compound's effect on a comparator protein present in the same test system. The compared protein kinases may be the same protein as when Docket No. 354. PATENT comparing changes in basal phosphorylation state or activity of a protein kinase upon contacting a test or candidate compound to a suitable test system. In other embodiments modulation of a protein kinase capacity includes increasing or decreasing the rate of phosphorylation of one or more downstream effectors or substrates of the protein kinase or modulating the phosphorylation status of the effectors or substrates upon contacting a test or candidate compound with a suitable test system relative to the basal state of the system.

[346] For example, the phosphorylation status of one or more isoforms of PI3K of Class I, or their capacity to convert phosphatidyl inositol (4,5)-bisphosphate (PIP2) to phosphatidyl inositol-(3,4,5)-trisphosphate (PIP3), or their capacity to effect activation of the downstream effector protein Akt before and after contacting a test compound to a suitable test system are compared. In those embodiments comparisons with test compound are typically made using 3a-diol or E-3a-diol as control compounds.

[347] In another example modulation of phosphorylation status of Erk includes increasing or decreasing the amount of pErk-1 relative to pErk-2 in a suitable test system after contacting the test compound to the suitable system. The modulation in phosphorylation state of one Erk isoform relative to the other isoform is concommitant with or in absence of an effect on the relative protein levels of the two isoforms.

[348] In some embodiments phosphorylation status of a protein kinase influences its nucleo-cytoplasmic translocation by effecting dimerization of the protein or its interaction with other components of the signal transduction cascade in which it participates either directly or through cross-talk. Without being bound by theory it is believed for some embodiments of the invention that E-3a-diol, or test or candidate compounds identified as having one or more of the activities of E-3a-diol by methods disclosed herein, modulates nucleo-cytoplasmic translocation of Erk-1/2, such that Erk-2 is preferentially retained in a cellular compartment in its un-phosphorylated state. This retention prevents activation of transcription factors in the nucleus that otherwise would support proliferation or survival. It is further believed that Erk-1/2 nucleo-cytoplasmic translocation is modulated by preferentially retaining some fraction of Erk-1 in its phosphorylated state in the cytoplasm (i.e., negatively modulating pErk nucleo-cytoplasmic transport) in order to activate cytosolic targets for promoting differentiation or re-regulating apoptosis.

[349] In other embodiments phosphorylation status of i-AR influences its nucleo- cytoplasmic translocation so that transactivation of genes that support survival or proliferation is inhibited. In these embodiments the induced i-AR phosphorylation additionally influences the transcriptional program of activated i-AR translocated into the Docket No. 354. PATENT nucleus to redirect transactivation away from supporting survival or proliferation to that which promotes differentiation. For E-3a-diol, those effects are believed to result from preferential or increased sequesterization of i-AR within a subcellular compartment of the cytoplasm (i.e., from negative modulation of nucleo-cytoplasmic transport of activated i- AR) that negatively modulates nuclear transactivation or pro-proliferative scaffolding to activated RTKs.

[350] Alternatively, or in addition to cytosolic sequesterization of i-AR, the differentiation promoting effects of E-3a-diol may involve differential recruitment of co-activators as described herein that result in protein complexes that are more favorable for transactivation of pro-apoptotic genes or are sub-optimal for transactivation of anti- apoptotic genes or is a consequence of differential recruitment of co-repressors that result in protein complexes that are less favorable for transactivation of anti-apoptotic genes. Assembly of those i-AR protein complexes through differential recruitment of co-activator or co-repressors re-regulates i-AR nuclear transactivation so that i-Ar signaling is redirected away from proliferation towards differentiation and may be influenced by modulation of the phosphorylation states of these co-activators or repressors. That modulation, based upon the insights of the present invention, should result from selective engagement of the Ras-Erk signal transduction pathway that preferentially forms pErk-1 or which preferentially retains Erk-2 in inactive form.

[351] In other embodiments modulating the phosphorylation status of a nuclear hormone receptor such as i-AR to inhibit proliferation may include decreasing tyrosine phosphorylation, particularly in relationship to decreased aberrant Src activity at i-AR or increased phosphorylation of Ser/Thr residues, particularly in the NH ₂ -terminal domain of i-Ar attributable to increased pErk-1 formation relative to pErk-2 (and which may be concommitant with decreased protein levels of Erk-2).

[352] In some embodiments, contacting a suitable test system comprising LNCaP cells with a test or candidate compound may result in positive modulation of the phosphorylation status of i-AR translocated into the nucleus, which may be accompanied by inhibition of Ser-210 or S-213 phosphorylation in the NH ₂ -terminal domain that is typically associated with transactivation of pro-proliferative genes.

[353] In some embodiments preferred test systems are capable of responding to 3a-diol or E-3a-diol whereupon contact of 3a-diol or E-3a-diol to the test system results in modulation of the aforementioned pathways, nodes, protein complexes, cascades, proteins or cellular trafficking of the complexes or proteins or downstream effects resulting from this modulation, including changes in phosphorylation states or activities of one or Docket No. 354. PATENT more protein kinases, transcription factors or nuclear hormone receptors in qualitatively similar manner to one or more of the effects described herein for 3a-diol or E-3a-diol.

[354] Other preferred test systems are capable to responding to 3a-diol and E-3a-diol in qualitatively similar manner to one or more of the effects described herein such that 3a- diol and E-3a-diol have opposing effects. In more preferred systems demonstrable physical interaction(s) in the manner described herein with an Erk or pErk isoform or with a scaffold protein involved in MAPK signal transduction with E-3a-diol exists or is increased after contacting this compound to a suitable test system.

[355] In other preferred test systems effects of E-3a-diol on phosphorylation status or activity of a phospho-protein that participates in PI3K-Akt or Src signaling are observable.

[356] In some embodiments modulation of a kinase activity includes increasing or decreasing the capacity of a kinase in a suitable test system to phosphorylate one or more of its downstream effectors or substrates of that protein kinase or to increase or decrease signaling through that signal transduction node, cascade or pathway upon contacting a test or candidate compound with a suitable test system relative to the basal state of the system.

[357] Without being bound by theory, it is believed for some embodiments of the invention that E-3a-diol or compounds identified as candidate compounds by methods as disclosed herein positively modulates the phosphorylation status of Erk-1 in its activation loop or negatively modulated the phosphorylation status of Erk-2 in its activation loop in comparison to the other isoform. In other embodiments the modulation of Erk isoform phosphorylation state(s) is accompanied by negative modulation of PI3K p85 phosphorylation state.

[358] In some embodiments in vitro test systems include a cell-based system wherein the cells comprising the test system express genes encoding functional GPR-C6a, wt-AR or T877A mt-AR or other mutant forms of cytosolic androgen nuclear hormone receptor such as those found in tumor tissue cells or in cancer cell lines. Functional GPR-C6a is capable of signaling through Ga/i, Ga/q or Gp/γ (concommitant with activation of Ga/i or Ga/q) release upon binding to the GPCR of a known agonist or reference compound in qualitatively similar manner to the full-length m-AR as described herein. Functional GPR- C6a includes mutants that lack the extracellular Venus fly trap (VFT) domain, which is responsible for sensing certain amino acids. In some embodiments functional GPR-C6a contains the human amino acid sequence. In other embodiments functional GPR-C6a contains the amino acid sequence for a rodent homolog such as that for mouse or rat. Docket No. 354. PATENT

Functional i-AR is capable of stimulation by contact of the cell-based system with a known agonist or reference compound in qualitatively similar manner to full-length wt-AR or mt- AR. In some embodiments agonist stimulation of i-AR in those contexts results from or is enhanced by one or more non-genomic effects initiated at GPRC-6a described herein or results in translation of a gene containing an ARE promoter (i.e., genomic effects).

[359] Candidate compounds are selective Erk modulators or selective modulators of Erk isoform phosphorylation status or activity. In preferred embodiments test compounds that are identified as candidate compounds are potentiators of Erk-2 activity or enhancers of Erk-1 activity. In other embodiments candidate compounds are selective GPR-C6a- Galpha/q agonists, sometime referred to as inverse agonists (with respect to Ga/i activation or release). In preferred embodiments candidate or further characterized candidate compounds are test compounds that exhibit efficacy or low liver toxicity when evaluated in a suitable animal model.

[360] One mutation frequently detected in advanced prostate cancer biopsies is at codon 877 where a switch from threonine to alanine occurs in the ligand-binding domain. This mutation is typically present in LNCaP cells. Another mutation H874Y is found in the recurrent human prostate cancer xenograft cell line CWR-R1 . That mutation retains wild- type sensitivity to androgens, but has increased responsiveness to other steroids. Therefore, some embodiments use a cell line as described herein expressing, or made to express by transfection (stably or transiently), a gene coding one of those particular mutant i-ARs. In preferred embodiments transformed HEK293 (HEK293T) cells that express a gene encoding functional LNCaP mt-AR are used in a suitable test system.

[361] Other gene mutations incorporated into gene constructs used for producing mt-AR expressing cell lines that provide suitable test systems disclosed herein include one or more of the mutations described in Table 1 [see, Taplin, M.E. et al. (1995)] and are numbered according to the gene sequence provided by Lubahn, D.B. et al. (1988).

[362] Table 1 . Some Additional AR Mutations for Derivation of Suitable Test Systems

Docket No. 354. PATENT

[363] Other mutations to be incorporated into cell lines disclosed herein for expressing mt-AR in derivation of suitable test systems are found in Marcelli, M. et al. (1990); Lubahn, D.B. (1989); De Bellis, A. et al. (1992); Newmark, J.R. et al. (1992); Culig, Z. et al. (1993); Suzuki, H. et al. (1993); Gaddipati, P. et al. (2004); Gottlieb, B. et al. (2004), all of which are incorporated by reference herein.

[364] In some embodiments the mt-AR resulting from transfection may be from expression of a gene having single, double, triple, quadruple or more of the point mutations described herein that do not involve the DNA-binding (DBD) domain in order to provide functional i-AR in a cell line disclosed herein. In preferred embodiments mt-AR is present in a cell-line that results from a single point mutation in AR. In other preferred embodiments the mt-AR is constitutively active due to partial or completed deletion of the COOH-terminal region that contains the LBD or is responsive to atypical ligands due to gene shuffling that result in an AR with the LBD of another nuclear hormone receptor.

[365] The NH ₂ -terminal activation function-1 (AF-1 ) is constitutively active in truncated AR that lacks the LBD. It is thus believed that in native i-AR the NTD domain competes with co-activators for AF-2 and this competition is lost in the truncated AR or in certain LBD point mutation that allow for greater co-activator recruitment. Therefore, in some embodiments an mt-AR receptor is introduced into a cell-line with an AR transfection vector containing a point mutation in exon 1 (encodes NTD) of the AR gene and is sometimes combined with a point mutation in a LBD exon that will permit mt-AR activation by wt-AR antagonists.

[366] The Hinge region, located between the NH ₂ -terminal domain (NTD) and the DNA- binding domain (DBD), is important for heat shock protein (Hsp) binding, which also involves the LBD. Hsp dissociation upon agonist binding is required for AR dimerization and eventual co-activator recruitment and binding of the activated receptor complex to ARE. The hinge region is also important for AR transport since a nuclear localization Docket No. 354. PATENT signal spans the region between DBD and the hinge region. Therefore, in some embodiments the mt-AR will contain one mutation in the NTD that forms the AF-2 region or one mutation in the hinge region, each of with is sometimes combined with one mutation in the LBD that permits agonist activation by wt-AR antagonists. In some of those embodiments the mutation in the NTD domain permits or disrupts formation of the AF-2 region on agonist activation. In other embodiments the NTD mutation permits formation of the AF-2 region while disrupting the N/C interaction in preference to a co- activator or co-repressor protein AF-2 interaction or selectively inhibits this protein-protein interaction. In other embodiments the mutation in the hinge region stabilizes or disrupts interaction of the mt-AR with an Hsp protein or disrupts translocated of the activated nuclear receptor to the nucleus (i.e., negatively modulates nucleo-cytoplasmic transport. In additional embodiments an mt-AR is introduced that is constitutively active due to disruption of the N/C interaction that allows for more facile co-activator recruitment.

[367] In other embodiments, the mt-AR will contain a Ser to Ala mutation wherein the wild type Ser residue is located in the NTD which is sometimes combined with one mutation in the LBD that permits agonist activation by wt-AR antagonists. In other preferred embodiments the mutant AR results from expression of a gene lacking one or more of the LBD exons that sometimes may be combined with a point mutation in exon 1 .

[368] The mutant ARs from the aforementioned mutations are useful for developing suitable test systems for screening of test compounds according to the methods described herein that disrupt the survival or proliferative advantage enjoyed by AR signaling- dependent cancers harboring those mutations (and not to be confused with androgen- dependent vs. androgen-independent CaP, both of are postulated to require AR signaling, but which have differing sensitivity to androgen). The advantages conferred include improved binding to co-activators, inhibition of co-repressor binding, improved NF-kB-Rel binding interactions, disrupted internal AR N/C interaction for greater recruitment of co- activators or increased ligand retention due to stabilization of the H12 lid of the ligand- binding pocket.

[369] In other embodiments the cell-based system is comprised of mammalian cells capable of DNA transcription or RNA translation (i.e., are viable cells) and express AR and optionally ERa or GPR-C6a. In more preferred embodiments in vitro test systems comprise viable mammalian cells expressing GPRC6a and wt-AR or mt-AR. In other embodiments the cell-based system is comprised of viable mammalian cells that are derived from a deposited cancer cell line and express AR, which encodes wt or mutant intracellular AR, and are either quiescent or induced to proliferate. In other embodiments Docket No. 354. PATENT the cell-based system is comprised of viable mammalian cells that are derived from a normal cell line and are induced to proliferate, for example, by mitogen or cytokine stimulation, or through transformation.

[370] In other embodiments the suitable cell-based system is comprised of quiescent cancer cells derived from a deposited cell line that expresses mt-AR. In other embodiments the derived cell-based system is comprised of quiescent cancer cells that additionally express GPR-C6a. Mammalian cells are "quiescent" (but not necessarily synchronized), as for example when maintained in charcoal stripped CS or FBS media or in growth factor and androgen-depleted media or in media where steroids are not present at concentrations more than about 100 pM, more than about 10 pM or more than about 1 pM. In some embodiments DHT is present at no more than 10 pM or 100 pM concentration in order to mimic the androgen-depleted state resulting from maximal androgen blockade therapy.

[371] In some embodiments an suitable in vitro cell-based test system will comprise cells expressing AR and GPR-C6a that additionally express one or more genes encoding a mitogen-responsive membrane-bound receptor or a functional RTK, including but not limited to EGFR, IGFR, ErbB2 or Her2/neu, either having native or constitutive activity, or a functional membrane hormone receptor (m-NR) other than GPRC-6a, including but not limited to membrane estrogen receptor (m-ER), membrane progesterone receptor (m-PR) or membrane glucocorticoid receptor (m-GR), or a functional cytosolic nuclear hormone receptor (c-NR) other than a i-AR, including but not limited i-ERa, i-ERp, i-PR or i-GR, or a functional cytokine receptor, including but not limited to receptors for TNFa or IL-6.

[372] In some embodiments a suitable in vitro cell-based test system will comprise cells expressing AR, GPR-C6a and ERfi.

[373] In some embodiments a suitable in vitro cell-based test system will comprise cells additionally expressing one or more genes encoding functional GPR-C6a protein, a functional or constitutively active RTK or one or more of the aforementioned functional hormone membrane or cytosolic receptors wherein the gene encoding the receptor protein is transiently or stably transfected into a normal or cancer cell line. In some embodiments GPRC6a gene is transfected into an androgen- or estrogen-associated cancer cell line in order to express or overexpress GPRC6a. In other embodiments GPRC6a, wt-AR or mt-AR is transfected into a cell line capable of proliferation, optionally containing an ARE promoter-reporter gene. Some preferred embodiments are comprised of cells from a deposited cancer cell line that expresses or are transfected to express AR, GPR-C6a and an ARE promoter-reporter gene. Docket No. 354. PATENT

[374] Therefore, a suitable test system includes a test system whose cells in vitro or in vivo will respond to contact with a naturally occurring androgen or a synthetic agonist at GPR-C6a, wt-AR or mt-AR preferably at concentrations between about 100 μΜ to about 1 pM or less, more preferably between about 10 μΜ to 1 pM or at physiologically relevant concentrations. In some embodiments DHT levels of about 10 pM to about 100 pM are used to mimic circulating concentrations in CaP patients of this hormone after maximal androgen blockade (MAB).

[375] More preferred in vitro cell-based test systems are comprised of hormone- associated cancer cells. Particularly preferred for such test systems are prostate cancer or breast cancer cells lines including but not limited to low passage or high passage LNCaP, PC-3, CV1 , CWR22, DU145, MCF-7, MCF-10A, T47D, MDA-MB-231 , MDA-MB- 438 or MDA-MB-468. Other preferred in vitro cell-based test systems are comprised of transformed normal cells induced to proliferate, by for example, contacting the test system with mitogen or hormone, transformed normal cells from genetic manipulation of normal cells or immortalized cells from a deposited cell and include THP-1 , COS, CHO, HUVEC, HeLa, HEK293, HEK-293T cells. Other preferred in vitro cell-based test systems are comprised or normal, transformed, immortalized or cancerous endothelial or fibroblast cells.

[376] In some embodiments the normal, transformed, immortalized or cancer cells described herein for deriving a suitable test system are transfected with an inducible androgen-responsive promoter-reporter gene or NF-κΒ response element-reporter gene construct and one or more genes encoding for functional i-AR, GPR-C6a, ERp, ErbB1 (EFGR) or ErbB2 (Her2/neu) or a constitutively active RTK or transcription factor.

[377] In some embodiments a suitable in vitro test system comprises low passage (LP) LNCaP cells, optionally transfected with an ARE promoter-reporter gene. In preferred embodiments LNCaP cells, commercially available from American Type Culture Collection (ATCC), are incubated in serum to passage number of 20 to about 35, about 24 to about 32 or about 20 to about 25, preferably about 24 passages, and are used in a suitable test system. Those passaged LNCaP cells are then contacted with a pro-proliferative agent in an effective amount to stimulate proliferation prior to step (a). In some of these embodiments LP LNCaP cells in step (a) are co-contacted with test compound and pAED.

[378] In other embodiments a suitable test system comprises high passage (HP) LNCaP cells, preferably transfected with an ARE promoter-reporter system. In preferred embodiments LNCaP cells from about 80 passages or more, preferably about 84 passages, are used for test compound screening and are contacted with a pro- Docket No. 354. PATENT proliferative agent in effective amount to stimulate proliferation or AR transactivation prior to step (a). In preferred embodiments HP LNCaP cells in step (a) are contacted with test compound while in androgen-depleted media.

[379] In some embodiments a suitable in vitro test system comprises HEK293T cells that are stably transfected with an i-AR ligand-binding domain (LBD) fused to the Gal4 DNA binding domain and which also expresses a reporter gene fused to an Upstream Activator Sequence (Gal4 response elements) both of which may be present in the same gene construct that was used for transfection.

[380] In other embodiments a suitable test system comprises MDA-kb2 cells (ATCC CRL-2713) expressing a gene encoding functional wt-AR and a gene containing an androgen response element upstream to a reporter gene (i.e., an ARE promoter-reporter gene), both of which may be present in the same gene construct that was used for transfection.

[381] In other embodiments a suitable test system comprises cells from a cell line transfected with a gene construct containing an androgen response element upstream to a reporter gene and a gene construct encoding functional mutant AR, both of which may be present in the same gene construct used for transfection. In preferred embodiments MtAR-HEK293 cells, which are transformed HEK293 fibroblasts that have been transiently co-transfected with and ARE promoter-reporter construct and a cDNA expression vector encoding functional LNCaP mutant AR, are used in a suitable cell-based in vitro test system.

[382] In other embodiments a suitable test system comprises MDA-kb2 cells stably transfected with a gene construct containing an MMTV promoter fused upstream of a reporter gene. These cells endogenously express genes for both i-AR (intracellular androgen receptor) and i-GR (intracellular glucocorticoid receptor).

[383] In other embodiments a suitable test system comprises normal cells having Eft/3 expression or AR-dependent cancer cells having ERp expression, either natively or through transfection (stably or transiently). The potential for a test compound to interact with GPR-C6a or i-AR and to effect transactivation of ERp is of importance considering the well-established anti-proliferative, pro-apoptotic role of ERp in prostate tissue. Therefore, in some preferred embodiments a test or candidate compound is contacted with a suitable test system comprising normal cells expressing ERp or AR-dependent cancer cells expressing wt-AR or mt-AR, preferably also expressing GPR-C6a. In other preferred embodiments a test or candidate compound is contacted with a prostate cancer Docket No. 354. PATENT cell line deficient in AR or AR , but expressing GPRC6a or expressing Εβ and GPR C6a.

[384] In some embodiments a suitable test system is derived from AR ^1' cancer cells or transformed normal cells that have been transfected (stably or transiently) with a i-AR expression vector and an androgen inducible reporter construct such as an ARE promoter-reporter, MMT promoter-reporter or a androgen receptor promoter- chloramphenicol acetyl transferase (CAT)-reporter such as GRE ₂ E1 bCAT or -286PBCAT. In preferred embodiments CV1 , COS-1 or PC3 cells are transfected with a gene construct encoding i-AR and a construct containing an androgen receptor promoter-reporter gene.

[385] In other embodiments a suitable test system comprises cancer cells stably transfected with a promoter-reporter gene construct that is sensitive to estrogenic stimulation. In a preferred embodiment T47D-kBluc cells, which endogenously express genes for both forms of the ER (estrogen receptor), namely ERa and ERp, are transfected with a gene construct containing at least three copies of the estrogen response element (ERE) fused upstream of a reporter gene. In other preferred embodiments ERP-HEK293 cells, which are HEK293 fibroblasts transiently co-transfected with an ERE promoter- reporter gene construct and a cDNA expression vector encoding functional ERp, preferably full-length human ERp, are used for deriving a suitable cell-based in vitro test system.

[386] In other embodiments a suitable test system comprises cells from a cancer cell line expressing constitutively active i-AR that lacks the ligand binding domain as when the amino-terminal domain is fused to a DNA binding domain fragment. In a preferred embodiment CWR-R1 cells are transiently infected with a gene construct encoding constitutively active wild-type receptor, which also contains an inducible promoter-reporter gene such as MMTV promoter-luciferase reporter or ARE promoter-luciferase reporter.

[387] In other embodiments a suitable test system comprises cells from a cell line as described herein that is treated with a selective inhibitor of the Erk MAPK pathway. In a preferred embodiment PD098059, which is a selective noncompetitive inhibitor of the Ras- Erk pathway that prevents the activation of MEK-1 by Raf-1 (PD098059 inhibits Raf activation of MEK-2 less effectively), is used prior to contact with a test compound.

[388] In other embodiments a suitable test system comprises cells from a cell line as described herein that is treated with a selective inhibitor of the PI3K-Akt pathway. In a preferred embodiment LY295002 is used, which is a selective inhibitor of PI3K prior to contact with a test compound. Docket No. 354. PATENT

[389] In other embodiments a suitable test system comprises cells from a cell line as described herein that is treated with a compound that directly or indirectly increases intracellular kinase activity or decreases phosphatase activity. Examples of these compounds are 8-bromo-cyclic AMP (cyclic AMP analog that acts through PKA), forksolin (adenylate cyclase activator), okadaic acid (phosphatase 1 and 2A inhibitor), vanadate (phosphotyrosine phosphatase inhibitor), growth factors and certain neurotransmitters.

[390] Preferred suitable in vivo test systems include castrated SCID mouse xenograft models implanted with hormone-associated tumor cells. In some of these systems hormone support is provided by appropriate subcutaneous implant. In one preferred suitable in vivo test system LuCaP-35V tumor cells are implanted. In another preferred test system LNCaP tumor cells are implanted into castrated SCID mouse. In some of those embodiments pAED is implanted subcutaneously to support AR-dependent cancer cell proliferation.

[391] Yet another preferred in vivo suitable test system uses female Lewis rats exposed to the carcinogen N-methyl-N-nitrosourea (MNU) so as to induce hormone- dependent mammary tumors in rats. The MNU-induced tumor model has previously been used to develop tamoxifen therapy in women with breast cancer and is thus appropriate for screening of test compounds potentially useful for the treatment of breast cancer. Substantial evidence suggests this rodent model system mimics human breast cancer since the initiation of cancer occurs primarily at the same site in both humans and rats and the majority of the induced tumors express estrogen and progesterone receptors.

[392] In one variation of the MNU-induced tumor model, rats are co-administered a test compound and a tubulin disrupting agent. Those disrupting compounds interfere with the appropriate function of tubulin in the mitotic spindle by inappropriately stabilizing polymerized tubulin or by inhibiting polymerization. In preferred embodiments the tubulin disrupting agent is a taxane including, for example, paclitaxel, docetaxel and cabazitaxel. That variation is useful to determine if a test compound acts synergistically with a taxane or allows for reduced or less frequent administration of the taxane for achieving similar efficacy of the taxane with reduced toxicity due to enhanced or synergistic activity..

[393] Other suitable in vivo test systems include other xenograft and transgenic models provided in Grossmann, M.E. et al. (2001 ), which are incorporated by reference, particularly Table 1 of page 1688.

[394] In preferred embodiments, suitable test systems are models for hyperproliferation conditions that are driven by cancer cells whose deposited cell line counterparts would be Docket No. 354. PATENT expected, based upon the insights provided by the invention disclosed herein, to favorably respond (i.e., by decreased proliferation or increased apoptosis) to Src, MEK1/2, MAPK1/3, PI3K or Akt inhibitors at physiologically relevant concentrations. In other preferred embodiments suitable test systems are models for hyperproliferation conditions that are driven by cancer cells whose deposited cell line counterparts would be expected, based upon the insights provided by the invention disclosed herein, to respond to i-Ar inhibitors. Those suitable test systems model hyperproliferation conditions that include prostate cancer, breast cancer, liver cancer, e.g., hepatocellular carcinoma, central nervous system cancer, e.g., glioma, astrocytoma or oligodenroglioma, lung cancer, e.g., small cell lung cancer or non-small cell lung cancer, colorectal cancer, myeloma, melanoma, lymphoma, thyroid cancer, pancreatic cancer, bone cancer or a cancer that metastases to bone, e.g., metastatic prostate cancer or breast cancer.

[395] Preferred test compounds for screening for one or more E-3a-diol activities using the suitable test system and methods described herein contain a 5a-androstane, androsten-5-ene or androst-4-ene steroid core structure. More preferred test compounds additionally contain a C3 monovalent O-linked substituent, optionally with disubstitution at C3 wherein the second C3 substituent is a monovalent C-linked group, or contain a C3 =0 substituent, and include compounds with =0 or disubstitution at C17 wherein the second C17 substituent is a monovalent O-linked substituent and the other is a monovalent C- linked substituent as these terms are defined herein. Monovalent C-linked substituents include optionally substituted alkyl, optionally substituted alkenyl, optionally substituted alkynyl, optionally substituted aryl and optionally substituted C-linked heteroaryl as these terms are defined herein.

[396] Preferred test compounds also include the 5a-androstane, androsten-5-ene or androst-4-ene steroids described herein wherein the monovalent O-linked substituent at C17 is replaced with an optionally substituted amine or N-linked heterocycle. Other preferred test compounds also include the 5a-androstane, androsten-5-ene or androst-4- ene steroids described herein wherein the steroid contains one or more double bonds or one or more additional double bonds, preferably containing 1 , 2, 3 or 4 total numbers of double bonds, excluding compounds inherently unstable or having a hypervalent carbon atom. Other preferred test compounds are steroid compounds as described herein wherein one or more of the angular methyl groups (i.e., at C18 and/or C19) is replaced with -H or with independently selected monovalent C-linked substituents, including an optionally substituted alkyl group such as ethyl or hydroxymethyl.

[397] More preferred test compounds have independently selected monovalent O-linked Docket No. 354. PATENT substituents at C3 and C17, optionally with independently selected monovalent C-linked substituents as second R groups at these positions as these terms are defined herein. Preferred monovalent O-linked substituents in these test compounds are -OH or an ester thereof, independently selected. Monovalent C-linked substituents include optionally substituted alkyl, optionally substituted alkenyl, optionally substituted alkynyl, optionally substituted aryl, including optionally substituted phenyl, and optionally substituted C-linked heteroaryl.

[398] Preferred monovalent C-linked substituents in these test compounds are optionally substituted C1 -4 alkyl groups, including methyl, ethyl, propyl, isopropyl, hydroxymethyl, optionally substituted C2-4 alkynyl, including ethynyl, propynyl or chloroethynyl, optionally substituted aryl and optionally substituted heteroaryl as these terms are defined herein. Particularly preferred test compounds are reference or control compounds as described herein, or have independently selected -OH or ester groups at C3 and C17 in the a- and β-configurations, respectively, and additionally contain -H or a monovalent C-linked substituent at C17 in the in the a-configuration, preferably ethynyl or optionally substituted phenyl and optionally substituted monocyclic heteroaryl .

[399] In some embodiments a test compound is evaluated in one or more suitable in vitro or in vivo test systems as described herein relative to E-3a-diol for activity in negatively modulating PI3K phosphorylation-mediated activation of its kinase catalytic activity or PIP3-forming activity and for increasing Erk-1 phosphorylation or activity in comparison to Erk-2.

[400] In other embodiments a test compound is evaluated in one or more suitable in vitro or in vivo test systems as described herein relative to E-3a-diol for activity resulting from the stimulation of Ga/q signaling. Thus, a test compound that activates Ga/q may result in one or more of (i) negatively modulate p1 10a/p85 phosphorylation activity, (ii) negatively modulate tyrosine phosphorylation state of PI3K regulatory subunit p85, (iii) negatively modulate serine phosphorylation state of Raf-1 or (iv) negatively modulate Ras interaction with or activation of p100a. Those downstream effects from Ga/q activation may separately or in some combination thereof provide selection criteria for identifying a candidate compound.

[401] In some embodiments a protein concentration, phosphorylation status or capacity for signal transduction, as determined for example by phosphorylation rate or status of a downstream effector or substrate(s) of that protein, is compared relative to one or more other proteins before and after application or administration of a test compound to a suitable test system such that a relative change in amount, concentration or Docket No. 354. PATENT phosphorylation status of a protein indicates an effect of the test compound on that protein's activity or an effect on signal transduction pathways that involve the compared proteins, either directly or through cross-talk between signal transduction pathways.

[402] In some embodiments a protein kinase concentration, phosphorylation status or capacity for signal transduction is compared relative to other protein kinases with similar amino acid residue specificity for phosphorylation. Thus, the compared protein kinases may all be Ser/Thr protein kinases, as for example when comparing changes in Erk-1/2 phosphorylation activity relative to JNK MAPK activity(ies) or p38 MAPK activity(ies) upon contacting a test compound to a suitable test system.

[403] In some embodiments the compared protein kinases will be homologous in amino acid sequence with at least 50% homology, preferably at least 60% homology, more preferably with at least 80% homology relative to a comparator protein kinase sequence, typically to that protein having the highest abundance in a suitable test system. Sometimes the protein kinases, whose effects from contacting a test or candidate compound to a suitable test system are compared, are of significant homology and substrate specificity in a suitable cell-free test system to be regarded as isozymes, as for example when comparing changes in protein kinase activity of the ERK MAPK isoforms Erk-1 relative to Erk-2 upon or after test or candidate compound contact to the test system.

[404] In some embodiments comparison of activity or a property of a G protein is made to one or more other G -proteins. In some embodiments, comparisons are made between relative release rate or amounts of released monomer Ga/i and Ga/q resulting from a test compound or candidate that engages a GPCR that couples to either of these G proteins. In other embodiments one or more effects from release of a particular Ga monomer after contacting a test compound or candidate with a suitable test system is compared with the basal or mitogen-stimulated state of the system. For example, changes in the amount of tyrosine phosphorylated Ga/q are compared from contacting a test compound or candidate to a suitable test system relative to a suitable reference compound. In those embodiments comparisons with test or candidate compound are preferentially made using 3a-diol or E-3a-diol as control compounds when the GPCR is GPR-C6a.

[405] Due to the homology between Erk-1 and Erk-2 and their ability to phosphorylate the same substrates in certain cell-free cell test systems these MAPK isoforms were considered redundant. However, it is now appreciated that signaling through these extracellular signal-regulated kinases is unique. For example, Erk-2 regulates CD8+ T cell proliferation by promoting their survival through regulation of Bcl-2 family proteins by Docket No. 354. PATENT increasing the activity of pro-survival members (i.e., anti-apoptotic) Bcl-2 and Bcl-x _L and decreasing activity of the pro-apoptotic member Bim at the transcriptional level. In contrast Erk-1 is not necessary for cell proliferation but is required for disease-modifying anti-inflammatory effects as disclosed herein for treating hyperproliferation conditions. Furthermore, increasing activation of Erk-1 relative to Erk-2 results in re-regulation of the apoptotic program and redirection of prosurvival signaling, including those involving other signal transduction nodes such as Src and i-AR, in cancer cells to signal transduction pathways that promote differentiation.

[406] Therefore, in some embodiments comparisons are made between the relative levels of pErk-1 to pErk-2 within a suitable test system prior to and after contacting the test system with test compound. In other embodiments protein levels of Bcl-2, Bcl-x _L or Bim from a suitable test system are determined prior to and after contacting a test compound with the test system. In either embodiment comparisons with test or candidate compound are preferentially made using 3a-diol and/or E-3a-diol as control compounds.

[407] As previously discussed, due to differences in nucleo-cytoplasmic shuttling between Erk-1 and Erk-2 mediated by the MEK isoforms, signaling output from Ras-Erk pathway is dependent on the composition of the Erk-containing protein complexes. Since test or candidate compounds that increase signaling activity through Erk-1 in comparison to Erk-2 are expected, based upon the insights provided by the invention disclosed herein, to have anti-proliferative properties, test compounds that disrupt the interaction between Erk-2 and MEK-1 provide an anti-proliferative effect by preventing preferential localization of unactivated Erk-2 in the cytoplasm that would promote its activation by growth factor stimulation and subsequent translocation of pErk-2 to the nucleus. However, test compounds that stabilizes that interaction, but do so to preferentially retain Erk-2 in inactive form in the cytosol will also provide an anti-proliferative effect through cytosolic sequesterization of that isoform. Furthermore, compounds that stabilize the interaction of MEK-1 with Erk-1 in preference to Erk-2 allow Erk-1 to reduce the ability of MEK-1 to properly present inactive Erk-2 in the cytoplasm for activation and subsequent translocation of pErk-2 to the nucleus upon growth factor stimulation. Therefore, a test compound that stabilizes the interaction between Erk-1 and MEK-1 will also be antiproliferative.

[408] Therefore, in some embodiments comparison is made between nucleo- cytoplasmic trafficking of pErk-1 or pErk-2 or subcellular localization of Erk-1 , pErk-1 , Erk- 2 or pErk-2 within a suitable test system prior to and after contacting the test system with test compound. In other embodiments the phosphorylation status of MEK-1 , Raf-1 or B- Docket No. 354. PATENT

Raf produced within a suitable test system is compared prior to and after contacting the test system with test compound. In preferred embodiments comparisons of test or candidate compound on those phosphorylation states are made using 3a-diol or E-3a-diol as control compounds. In other embodiments comparisons of test or candidate compound on MEK-1 interaction with Erk-1 and/or Erk-2 are made.

[409] Test or candidate compounds that improve protein-protein interaction between Erk-2 and MPK-3 such that phosphatase activity is increased are expected, based upon the insights provided by the invention disclosed herein, to inhibit cytoplasmic pro-survival signaling from Erk-2 and thus would prove useful in treating hyperproliferation conditions. Improved binding with MPK-3 of deactivated Erk-2 is also expected, based upon the insights provided by the invention disclosed herein, to compete with its binding to MEK-1 to slow reactivation of Erk-2 and subsequent nuclear translocation of the reactivated isoform.

[410] Therefore, in some embodiments comparisons of MKP-3 phosphatase activity are compared prior to and after contacting the test system with test or candidate compound. In these embodiments optional comparisons of test or candidate compound are made using 3a-diol or E-3a-diol as control compounds. In some of these embodiments MKP-3 activity is assayed by monitoring hydrolysis of p-nitrophenol phosphate ester.

[411] The results of kinase inhibitor experiments of Example 15 also support the hypothesis that E-3a-diol is binding to the membrane androgen receptor on LNCaP cells, which activates Erk-1 while inhibiting PI3K. In those experiments LY295002 (PI3K inhibitor) further increased inhibited proliferation of HP LNCaP cells by E-3a-diol, whereas PD098059 (inhibits activation of MEK-1 by Raf) decreased the inhibition by E-3a-diol. Therefore, in some embodiments comparisons are compared prior to and after contacting the test system with test or candidate compound in the presence or absence of a PI3K or MEK inhibitor, such as LY295002 or PD098059.

[412] As previously discussed the membrane androgen receptor can be specifically bound by cell-impermeable androgen-protein conjugate, and when contacted with LNCaP cells superficially behave like E-3odiol. For example, T-BSA can induce Erk phosphorylation (but without apparent isoform selectivity), inhibit cell proliferation and induce PSA secretion. However, the effects of androgen-BSA conjugates on downstream signal transduction are sometimes more like that of 3a-diol. For example, 3a-diol increases Akt signal transduction and T-BSA increases PI3K activity in LNCaP, MCF-7 and T47D. Docket No. 354. PATENT

[413] In contrast, E-3a-diol selectively permits Erk-1 phosphorylation in LNCaP to the exclusion of Erk-2 without increasing gene expression of either isoform. As previously mentioned, in LNCaP cells T-BSA also increases total pErk levels. In addition, E-3odiol and T-BSA have opposite effects on phosphorylation of PI3K with E-3a-diol decreasing tyrosine phosphorylation of the p85 regulatory subunit whereas T-BSA increases that phosphorylation in LNCaP and MCF-7. In DU-145 it is only on longer exposure to T-BSA (2h) is decreased p85 tyrosine phosphorylation observed [Papadopoulou et al. (2008)].

[414] As previously discussed, it is believed that E-3a-diol binds to the m-AR GPR-C6a to release Ga/q to inhibit P85a PI3K phosphorylation and to activate Ras-Erk signaling to selectively activate Erk-1 in preference to Erk-2. It's therefore postulated that androgen- BSA conjugates unexpectedly behave in similar manner to E-3a-diol by inducing apoptosis through interaction at GPR-C6a, but does so on time scales more consistent with a secondary genomic effect and does not engage in a non-genomic effect that selectively engages Erk-2 intracellular^ such that Erk-1 is preferentially phosphorylated. In some embodiments those conjugates are useful for positive controls for PI3K activation as measured by increased pAKT levels, p85 tyrosine phosphorylation and/or downstream effects on effector proteins in PI3K-Akt signal transduction and for non-specific Erk isoform activation as measured by increased pErk-1 and p-Erk-2 levels with control for total protein levels.

[415] Therefore, in some embodiments a cell-impermeable androgen conjugate with C3 steroid attachment is used as a positive control for screening test compounds for secondary genomic activity that is associated with the non-genomic activity of E-3a-diol due to its engagement of GPR-C6a. Those conjugates include T, DHT, 3a-diol or βΑΕϋ covalently attached through a linker to a protein such as BSA, optionally with fluorophore. Preparation of T and DHT-conjugates are described in De Goeij, A.F. (1986). Example use of T-BSA conjugates is described in Hatzoglou, A. et al. (2005). The βΑΕΤ and 3a- diol conjugates are prepared by contacting a suitably protected hydroxysteroid (free C3- OH) with a hindered base and contacting the resulting alkoxide with a suitable protected bifunctional PEG linker wherein one functional group is a terminal a-halo ether that undergoes nucleophillic substitution by the alkoxide intermediate and the other functional group is a carboxylic acid ester. After protecting group removal, the steroid-0-CH ₂ - (OCH2) _n -COOH (wherein n is preferably 1 , 2, 3 or 4) intermediate is covalently attached to BSA in analogous manner described for preparation of T and DHT conjugates. Other 5a- androstane-protein conjugates suitable as test compounds include conjugates of 17a- ethynyl-5a-adrostan-3-one-173-ol (or 17a-ethynyl DHT, henceforth E-DHT). In preferred Docket No. 354. PATENT

5a-androstane-protein conjugates, E-DHT is conjugated to BSA in similar manner as conducted for testosterone and DHT (i.e., through an oxime linker utilizing its 3-one substituent).

[416] It is thought the short term effects of Erk activation results from signaling through scaffold protein complexes, which may also involve cytoskeleton components, whereas prolonged Erk signaling results from its nuclear transport and subsequent expression of early-response genes such as c-fos. It is believed that short term signaling is mediated by activated Erk-1 in the cytosol while long-tern signaling effects are mediated by Erk-2 in the nucleus. These long term signaling effects usually result from chronic mitogenic or pro- inflammatory stimuli (external or constitutive).

[417] Therefore, in some embodiments levels of one or more protein products from transcription by Elk-1 (which is an Erk substrate) or dependent on c-fos expression (with Elk-1 as its transcription factor) are compared prior to and after contacting the test system with test compound. In those embodiments comparisons of test or candidate compound are preferentially made using 3a-diol and/or E-3a-diol as control compounds.

[418] As previously discussed, once activated, Erk does not always enter the nucleus, but may be retained in the cytoplasm, which is dependent in part on the activating MEK isoform. Activated Erk may also be retained in the cytoplasm by to one or more scaffold proteins whose subcellular location influences the outcome of the signal flowing through this signal trandsuction node. It is therefore believed the preferential retention of pErk-1 , selectively in a cytosolic subcellular domain with concommitant sequesterization of ErK-2 in an inaccessible domain for phospho-activation occurs by E-3a-diol acting intracellular^.

[419] Thus, compounds that are able to redirect Erk signaling to favor Erk-1 cytoplasmic signaling in comparison to Erk-2 signaling either in the nucleus or the cytoplasm, are expected, based upon the insights provided by the invention disclosed herein, to have anti-proliferative or additional anti-inflammatory properties without unduly effecting basal kinase signaling, thus proving useful in treating hyperproliferation or unwanted inflammation conditions that support initiation or progression of the hyperproliferation conditions as disclosed herein.

[420] Therefore, in some embodiments the aforementioned time course effect of a test or candidate compound contacted to a suitable test system on protein kinase activity including that of Raf, MEK, Erk, PI3K, Src or an isoform thereof or nucleo-cytoplasmic Erk isoform trafficking is determined prior to (i.e., at t=0 min) or after contact of test compound, preferably at one or more time points selected from 5, 10, 15, 30 and 60 min. In preferred embodiments comparisons of test compound are made using 3a-diol and/or E-3a-diol as Docket No. 354. PATENT control compounds. The short term effects of pErk-1 acting in the cytosol are expected to occur within the 5-15 min time frame whereas pErk-2 effects acting in the nucleus are expected to occur within the 30-60 min time frame. Furthermore, non-genomic AR signaling effects, resulting either from the membrane AR or from i-AR scaffolding effects within the cytosol, will occur in the 5-15 min time frame whereas genomic effects from transactivation activity of i-AR acting in the nucleus will occur in the 30-60 min time frame. G protein-dependent signaling is expected to be rapid (within 5 min) and transient in nature while G protein-independent signaling is expected to occur subsequent to G protein-dependent signaling and be more persistent.

[421] In some embodiments transformed HEK-293 are transfected (transiently or stably) with an ARE promoter-reporter construct and T877A mutant AR to provide a suitable in vitro test system. In one such suitable test system (see Example 7) E-3a-diol was found to stimulate transcription of the reporter, presumably due to partial agonist activity at mt- AR, Other suitable in vitro test systems for evaluating test compounds for nuclear receptor binding and transactivation activity employ artificial cell constructs that contain intracellular forms of various steroid receptors, as exemplified by Table 7,. Furthermore, nuclear binding assays as described herein for evaluating test compounds show that E- 3a-diol does not bind strongly to the wild type AR (wt-AR), yet appears to bind the T877A mutant receptor, and not at all to PR, GR or ERp, and has a weak affinity for ERa.

[422] In some embodiments C4-2B cells are transfected (transiently or stably) with an ARE promoter-reporter construct. C4-2B is a CRPC cell line that express the same mutated AR gene as in LNCaP cells. In one such suitable test system E-3a-diol as a positive control compound activated the ARE promoters in these mutated /^-expressing cells, but did not inhibit transactivation of i-AR upon co-contacting the cells with β-AED or DHT. Instead, E-3a-diol appeared to activate the ARE promoter-reporter even though it inhibited proliferation and increased apoptosis. That result supports the assertion that residual i-AR transactivation activity is required for differentiation-induced apoptosis with E-3a-diol acting as a partial agonist.

[423] In some embodiments apoptotic activity of a test or candidate compound is examined on LuCaP-35V tumors implanted in SCID mice in the presence of pAED, which is a suitable in vivo test model for evaluating test compounds for activity against CaP, particularly for CRPC. In this xenograft model, tumor cells contain wt-AR, which is often found in CRPC. Results with E-3a-diol in one such suitable in vivo test system model (see Example 5) further supports its activity against castrate-resistant prostate cancer in the manner described for the C4-2B cell-based test system (whose cells contain T877A Docket No. 354. PATENT mt-AR) since treatment with E-3a-diol significantly inhibited growth of LuCaP-35V in mice supplemented with pAED. Treatment with E-3a-diol also resulted in slight increases in serum PSA levels versus the control AED-LuCaP-35V animals, but these differences did not reach significance. However, these elevations in PSA are in concordance with in vitro studies demonstrating increases in AR-mediated transcription by E-3a-diol treatment.

[424] Without being bound by theory, it is believed that wt-AR cannot be activated directly by E-3a-diol (i.e., not a sufficient agonist for wt-AR), but is activated indirectly through the Ras-Erk pathway through cross-talk with GPR-C6a-Ga/q signaling stimulated by binding of E-3a-diol to the GPCR, thus providing the required AR signaling that supports differentiation. It is further believed that if pAED is an agonist for GPR-C6a Ga/i signaling this activity is effectively competed by E-3a-diol binding to the GPCR to induce inverse agonist activity by activating Ga/q signaling. If pAED is not an agonist at the m- AR, a metabolite thereof may be an agonist for GPR-C6a Ga/i signaling, particularly if that metabolite is DHT. It is additionally believed that GPR-C6a-Ga/q signaling redirects agonist action of pAED (or a metabolite thereof) on wt-AR that otherwise would be pro- proliferative to AR signaling that supports differentiation. Those beliefs are supported by binding studies to wt-AR (IC ₅₀ PAED = 247 nM vs. IC ₅₀ E-3a-diol = 1 .5 μΜ) that would indicate, in the absence of inverse agonist action of E-3a-diol at GPR-C6a to activate Erk- 1 and its sequesterization of Erk-2 in inactive form, direct agonist action at wt-AR to support proliferation would occur.

[425] In some embodiments, a suitable in vitro test system comprises PC-3 or DU-145 cells. In those embodiments a preferred test or candidate compound influences p21 ^Cip1 activity to inhibit cyclin D-CDK activity or decreases transcription of a p21 ^Cip1 -inducible reporter gene transfected into PC-3 or DU-145 cells. In other embodiments a more preferred test or candidate compound induces those cells to undergo apoptosis when contacted in combination with a DNA-alkylating or DNA-intercalating agent or with a tubulin disrupting agent.

[426] The effect of test or candidate compound on LuCaP-35V proliferation in the absence of pAED is examined (see Example 5 as exemplified for E-3a-diol), which represents another suitable in vivo test system. Those experimental conditions mimic the clinical scenario of patients treated with agents aimed at blocking adrenal synthesis of androgens (e.g., ketoconazole). In that setting, E-3a-diol significantly inhibited the tumor doubling times of LuCaP-35V with significant increases in serum PSA levels.

[427] Without being bound by theory, it is believed the increased doubling time observed Docket No. 354. PATENT with E-3a-diol administration as a positive control compound in the absence of pAED from 13 days in the presence of pAED to 18 days (with controls for both studies at 10 days) in its absence is due to E-3a-diol not having to compete with pAED binding (or with that of a metabolite thereof) to GPR-C6a so as to effect Ga/q signaling without contrary signaling from Ga/i activation.

[428] In another suitable in vivo test system, C4-2B cells are injection into tibiae as exemplified by Example 4. In this model for evaluating test compounds in metastatic CaP, xenograft tumors that arise exhibit an osteoblastic response when grown in the bone environment similar to that seen in patients with CaP bone metastases. Results in one such suitable in vivo test system demonstrate that E-3a-diol administered as a positive control compound decreases weight of tumored tibiae in comparison to control tumored tibiae and immunohistochemical analysis of the tumored tibiae showed osteoblastic reaction associated with growth of the C4-2B tumors in the bone and tumor foci between the newly formed woven bone in both E-3a-diol-treated and control tibiae. In contrast to Example 5, which uses LuCaP-35V cells that contain wt-AR, E-3a-diol administered as a positive control compound in this model, where C4-2B cells endogenously contain T877A mt-AR, resulted in decreased serum PSA.

[429] The results with E-3a-diol in the C4-2B and LuCaP-35V indicate its activity for treating late-stage disease as well as earlier stage disease where progression to "androgen-independence" has not occurred. Furthermore, activity is indicated for wild- type and mutant AR-bearing tumors provided that the mutant-AR remains capable of transactivation. Furthermore, transient increases in PSA in a treated subject may actually be associated with an apoptotic effect on tumor cells due to redirection of aberrant AR signaling towards differentiation.

[430] In another suitable in vivo test system test or candidate compounds are evaluated in for activity towards breast cancer (which often contains i-AR) in mammary tumors induced by carcinogen (see Example 19 as exemplified for E-3a-diol). Treatment with high and low dose E-3a-diol administered as a positive control compound was compared with accepted therapeutic compounds, tamoxifen and anastrozole. When treatment started, the number of large tumors (>300 mm ³ ) ranged between five and eight for all active groups except anastrozole (four), and was highest in the E-3a-diol-docetaxel combination group, and lowest in the tamoxifen comparator. E-3a-diol aggressively shrank established tumors and prevented the appearance of new tumors. The rate of tumor volume reduction and degree of tumor suppression after cessation of treatment was similar for both high dose E-3a-diol monotherapy and tamoxifen. When E-3a-diol Docket No. 354. PATENT was combined with docetaxel, not only was the tumor-ablative activity enhanced, but tumor suppression was also maintained for sixty days after cessation of treatment.

[431] Treatment with E-3a-diol alone resulted in a rapid reduction in tumor burden, similar to tamoxifen and anastrozole, and the combination of E-3a-diol with docetaxel was more effective than any of the agents used alone as monotherapy. After cessation of treatment, tumors began to grow in all monotherapy groups, but not in animals treated with the combination of E-3a-diol and docetaxel. Tumors from E-3a-diol treatment animals showed increased immunohistochemical markers of apoptosis, increased expression of pro-apoptotic genes, decreased expression of anti-apoptotic genes and decreased levels of androgen and estrogen nuclear hormone receptors. Anastrozole is a standard of treatment for reducing endogenous estradiol in patients with breast cancer that have failed first-line therapy. Although active in the MNU model, anastrozole was inferior to E-3a-diol in the study.

[432] One of the most disturbing side effects of aromatase inhibitors is the emerging effect on bone, with a significant increase in fractures during the course of the ATAC trial, an earlier increase in bone density loss, and an even earlier increase in bone turnover markers [Eastell, R. et al. (2006)]. In contrast to anastrozole, E-3a-diol may have a positive effect on bone, having been shown to increase bone mineral density relative to vehicle in an intra-tibial prostate cancer xenograft model (Examples 4-5), which may be relevant to breast cancer considering the importance of bone metastases in this disease [Ye, L. et al. (2009)].

[433] Significantly in the cancer therapy setting, where drugs are frequently used in combination, E-3a-diol does not have appreciable hepatic, hematopoietic, or cardiopulmonary toxicity or deleterious effects on bone at what are currently believed to be pharmacologically relevant doses and thus may be used as a negative control compound for these toxicities in suitable in vivo test systems.

[434] This safety profile thus provides a rationale for its combination with a classic cytotoxic agent, including tubulin disrupting agents such as a taxane. Such combinations would be expected to result in complementary anti-neoplastic mechanisms, with a reduced incidence of escape from treatment. Breast cancer adjuvant therapy (treatment to prevent recurrence) employs various combinations of anthracyclines, taxanes, and cyclophosphamide [Jones, S.E. et al. (2000)]. The data presented here supports the contention that addition of E-3a-diol to adjuvant treatment regime may significantly enhance therapeutic benefit without increasing toxicity. Docket No. 354. PATENT

[435] In one suitable test system for evaluating test or candidate compound's effect in combination with a cancer chemotherapeutic compound, the tubulin disrupting agent docetaxel in combination with the test compound is determined using the C4-2B tibia tumor model (Example 16 as exemplified for E-3a-diol when administered as a positive control compound) in castrated SCID beige mice. In that study the benefit of the combination therapy was also indicated by a positive effect on bone mineral density.

[436] It should be understood to one of ordinary skill in the art that Example 16 and 17 may be modified in order to evaluate a test or candidate compound found to have one or more effects on protein phosphorylation states disclosed herein for E-3a-diol in combination with a cytotoxic agent or other cancer chemotherapeutic compound known to be effective in treating a hyperproliferation condition in which AR signaling exists or is capable of occurring.

[437] Therefore, in some embodiments a preferred test or candidate compound will inhibit proliferation or induce apoptosis of tumor cells in a suitable in vivo test system for a hyperproliferation condition dependent on the presence of functional i-AR. In other embodiments a preferred test or candidate compound exhibits synergistic or enhanced tumor inhibition activity when used in conjunction with a cancer chemotherapeutic compound in those test systems. In some of these embodiment the suitable test system in an in vitro, xenograft or carcinogen-induced model for breast or prostate cancer. In other embodiments, a synergistic or enhanced effect is observed when contacting the test compound in these test system in conjunction with an ERa or /-AR inhibitor. In additional embodiments a synergistic effect is observed when using the test or candidate compound in conjunction with a DNA alkylating or intercalating agent. In additional embodiments, a test or candidate compound identified as a candidate compound provides a positive effect on bone mineral density in qualitatively similar manner to E-3a-diol when contacted to a suitable in vivo test system, preferentially in the presence of a cancer chemotherapeutic compound.

[438] In some embodiment a test or candidate compound additionally affects the phosphorylation state of i-AR in qualitative manner observable for E-3a-diol. In other embodiments, a test compound contacted to a suitable in vivo test system effects PSA secretion in qualitatively similar manner described for E-3a-diol for identification as a candidate compound. In other embodiments a test compound or candidate when contacted to a suitable in vitro test system will induce prostate cancer cells to secrete PSA in conjunction with the anti-proliferative or pro-apoptotic Erk sequesterization effects described herein. Docket No. 354. PATENT

[439] In other embodiments a test or candidate compound contacted to a suitable test system comprising LP LNCaP cells will inhibit their proliferation or induce apoptosis when stimulated with an androgen, such as pAED or DHT, for identification as candidate compound. In still other embodiments a test or candidate compound contacted to a suitable test system comprising castrate-resistant CaP cells, such as HP LNCaP, LuCaP or C2-4B cells, will inhibit their proliferation or induce apoptosis in an androgen-depleted environment for its identification as candidate compound.

[440] In other embodiments a test or candidate compound contacted to a suitable test system comprising cancer cells having functional i-AR will exhibit one or more in vitro gene transcription effects described herein. For example, an examination of E-3a-diol- treated HP LNCaP cultures showed a dramatic increase in secreted PSA in the media. However, this rise in PSA was not accompanied by an increase in proliferation; rather the proliferation of HP LNCaP was inhibited. Those findings described herein are similar to the observations in the aforementioned LuCaP-35V tumor treatment study. The data is therefore consistent with a switch in the transactivation program of i-AR culminating in selective ARE promoter engagement that promotes differentiation.

[441] In some embodiments a test or candidate compound will affect gene transcription by switching the gene expression patterns observed for DHT or 3a-diol that supports proliferation or survival to one that supports differentiation in the manner observed for E- 3a-diol. Examples of induced changes in gene expression by contacting E-3a-diol to LNCaP cells are shown in Tables 2-5 (Example 2).

[442] In some embodiments the test or candidate compound will affect one or more genes in qualitatively similar manner to that indicated in Table 6. Particularly preferred are test or candidate compounds that effect expression of one or more of AR, IRX5, JUN, CD44 and JAG1 (decreased expression), and RIS1, TIMP2, RUNX1, CASP10 and LOX (increased expression).

[443] 5. Numbered embodiments

[444] The following numbered embodiments further illustrate the invention and are not intended to limit the invention in any way.

[445] 1. A method to identify a candidate compound, the method comprising (or consisting essentially of or consisting of) (a) contacting a test compound with a suitable test system; (b) determining phosphorylation states of Erk-1 and Erk-2 resulting from step (a); and (c) selecting a test compound that positively modulated phosphorylation state of Erk-1 activation loop relative to Erk-2 or negatively modulated the phosphorylation state of Docket No. 354. PATENT

Erk-2 activation loop relative to Erk-1 of step (b), wherein the test compound selected from step (c) is identified as a candidate compound.

[446] In this embodiment, preferred test systems constitute cells or tissue in vivo or cells in tissue culture or constitute cell extracts wherein signaling through target molecules of interest, e.g., Erk-1 , Erk-2, PI3K proteins, GPR-C6a and/or AR, is(are) functional and phosphorylation changes or other biological responses as described herein, as for example for E-3a-diol and 3a-diol, in response to the test (or control) compound can be measured. Other preferred test systems are cell-free artificial test systems prepared by reconstituting a signal transduction pathway with one or more proteins of that pathway in a suitable buffer with a downstream substrate that is capable of phosphorylation by at least one of the signal transduction pathway proteins

[447] The test compound to be contacted with a suitable test system will usually be in a vehicle, composition or formulation that is compatible with the test system, e.g., the test compound can be in a solution or suspension or it can be administered as a solid or liquid formulation to an animal such as a rodent (e.g., mouse or rat), or, for clinical assessment of the test compound, it can be administered to a human patient. Said contact is followed by measurement of target molecule phosphorylation states (e.g., Erk-1 , Erk-2) in cells or tissue samples taken from the animal or patient after administration of the test compound to the animal or patient. For controls, test systems will typically be contacted with a positive, negative, placebo and/or vehicle control or reference compound to confirm proper functioning of the test system in vitro or in vivo. As is apparent from the foregoing, the test compound and any control or reference compound or composition will be contacted with the test system (i) under conditions where the test system is functional, e.g., cells in tissue culture are maintained under standard growth conditions, (ii) for one or more sufficient periods of time, e.g., for about 5 sec. to about 120 minutes for cells in tissue culture in vitro or about 0.1 hour to about 48 hours for cells or tissue in vivo, and (iii) in one or more amounts or concentrations suitable to assess biological responses, e.g., dose-responses to the test compound and/or vehicle, placebo or control compound effects, if any. For assessing non-genomic effects in vitro as described herein that type of biological response is measured preferably from within 5 sec to 30 min after contacting the test system with test compound. For genomic effects in vitro as described herein that type of biological response is preferably measured from within 60-120 min. after said contact.

[448] Typical test compound final concentrations with the test system will be in a range, e.g., about 0.01 nM to about 20 mM or usually about 1 nM to about 10 mM, which can include one or more of about 0.05 nM, 0.1 nM, 1 nM, 10 nM, 100 nM, 200 nM, 1 mM, 2 Docket No. 354. PATENT mM and 10 mM. Concentrations of other compounds, e.g., components or excipients in the formulation that contains the test compound will typically be tested at the same or nearly the same concentrations they are at when the test compound is contacted with the test system.

[449] In some of these embodiments the phosphorylation state of Erk-1 activation loop is positively modulated with respect to Erk-2 by the test compound increasing the amount of pErk-1 produced in cells of a suitable test system in comparison to the amount of pErk-2 produced. In those embodiments sometimes Erk-1 protein level is positively modulated without discernable increase in pErk-2 protein level. In other embodiments the phosphorylation state of Erk-2 activation loop is negatively modulated by the test compound decreasing the amount of pErk-2 produced when cells of a suitable test system are stimulated to proliferate in comparison to the amount of pErk-2 that would have been produced in absence of the test compound. In other embodiments the phosphorylation status of Erk-1 activation loop is positively modulated such that the ratio of intracellular pErk-1 to pErk-2 is increased due to contact of test compound with cells of a suitable test system. In those embodiments sometimes pErk-1 is produced and pErk-2 remains at basal levels or is not produced.

[450] 2. The method of embodiment 1 wherein the suitable test system is a suitable in vitro cell-based test system. In some of these embodiments the cells within the suitable test system or suitable in vitro cell-based test system are cancer cells dependent on i-AR signaling for survival or proliferation. These embodiments include androgen-associated cancer cells from a deposited cancer cell line known to contain functional i-AR protein. Those embodiments also include hyperproliferating cells not typically dependent on i-AR signaling, but which become increasingly so as a result of an emerging resistance mechanism.

[451] Typically, effects on cells contacted with test compound are compared to those of control cells (i.e., cells that are sham contacted or are contacted with test compound that is a positive or negative control). Those include changes in phosphorylation state of MAPK3 (Erk-1 ) vs. MAPK1 (Erk-2) or phosphorylation state of PI3K in test cells compared to control cells, i.e., modulation of MAPK3 vs. MAPK1 cells compared to control cells. Thus, for example, test compounds selected for step (c) that have been contacted with test cells positively modulated the phosphorylation state of Erk-1 or negatively modulated the phosphorylation state of Erk-2 relative to the other Erk isoform without qualitatively or quantitatively similar effects observed for sham contact with control cells. In another example, the selected test compound will positively modulated the phosphorylation state of Erk-1 or negatively modulated the phosphorylation state of Erk-2 relative to the other Docket No. 354. PATENT

Erk isoform qualitatively or quantitatively similar to that observed for E-3a-diol contact with control cells as positive control.

[452] 3. The method of embodiment 1 or 2 wherein the suitable test system or the suitable in vitro cell-based test system comprises (or consists essentially of or consists of) cells from a prostate, breast, ovarian cancer or other cancer cell line that is dependent on transactivation of one or more genes with an upstream ARE promoter for survival or proliferation, comprises (or consists essentially of or consists of) cancer cells not having endogenous functional GPR-C6a protein, with or without functional i-AR protein, that are genetically engineered or transfected (stably or transiently) so as to contain functional GPR-C6a protein, cancer cells not having endogenous functional GPR-C6a and i-AR proteins that are genetically engineered or transfected (stably or transiently) so as to contain functional GPR-C6a and i-AR proteins or comprises (or consists essentially of or consists of) transformed normal cells not having endogenous functional GPR-C6a protein or endogenous functional GPR-C6a and i-AR proteins that are genetically engineered or transfected (stably or transiently) so as to contain functional GPR-C6a and i-AR proteins.

[453] In some of these embodiments, the i-AR protein is a functional wt-AR or a functional mutant AR protein that contains the human amino acid sequence, while in other embodiments the functional wt-AR or mt-AR protein contains the amino acid sequence for a rodent homolog as for example that found in mouse or rat. In additional embodiments the gene encoding the androgen-inducible promoter gene and the gene encoding functional i-AR protein are within the same gene construct that is transfected into the cells. In some embodiments functional i-AR gene is engineered into the AR ^1' prostate cancer cells such as DU145 and PC-3 or is engineered into transformed normal cells that do not endogenously contain functional i-AR protein. Preferred embodiments use cancer cells containing or genetically engineered to contain full-length human wt-AR protein, while in other preferred embodiments the cancer cells have a T877A mutant AR protein such as the LNCaP mt-AR. In other preferred embodiments the mt-AR protein is mutated in the NH ₂ -terminal domain such that an advantage to survival or proliferation is conferred and may also include the T877A mutation.

[454] 4. The method of embodiment 1 , 2 or 3 wherein the suitable test system or the suitable in vitro cell-based test system comprises (or consists essentially of or consists of) cancer cells or transformed normal cells that are responsive to DHT or pAED endogenously having or transfected (stably or transiently) or genetically engineered to have functional GPR-C6a protein or comprises (or consists essentially of or consists of) AR ^+/+ cancer cells having a functional GPC-C6a gene or transformed GPC-C6a ^v~ AR ^~ ' ^~ Docket No. 354. PATENT normal cells that are genetically engineered or transfected (stably or transiently) so as to contain functional GPR-C6a and i-AR genes.

[455] In some of these embodiments the prostate cancer cells of step (a) are quiescent. In preferred embodiments the prostate cancer cells are LP LNCaP cells or HP LNCaP that have been passaged in androgen- and growth factor-depleted media. In other preferred embodiments these or other prostate cancer cells are supported by androgen supplanted to the media in order to provide a suitable test system.

[456] In other embodiments the prostate cancer cells are PC-3 or DU-145 cells transfected to contain functional i-AR gene. In additional embodiments the cells are contained within suitable in vitro cell-based test systems for modeling CRPC in vitro and preferably include C4-2B or HP-LNCaP cells. In other embodiments the cancer cells are supported under conditions of androgen and growth factor depletion using an i-AR agonist. For cancer cells that have or are transfected or genetically engineered to have wt-AR, media support is typically with DHT. For cancer cells having or transfected or genetically engineered to have T877A mt-AR, support may be affected not only by wt-AR agonists, but with other natural hormones that are agonists for other nuclear hormone receptors (e.g., estradiol) and further include other natural hormones including DHEA or PAED. In some preferred embodiments these T877A mt-AR cancer cells are stimulated with DHT or pAED.

[457] 5. The method of embodiment 1 , 2, 3 or 4 wherein the suitable test system or the suitable in vitro cell-based test system are cancer cells or transformed normal cells containing endogenous functional GPR-C6a protein. In preferred embodiments the cancer or transformed cells of the suitable test system or suitable in vitro test system in step (a) further endogenously contain, or are transfected (stably or transiently) or genetically engineered so as to contain, functional i-AR protein. In more preferred embodiments the cells of the suitable test system or suitable in vitro test system that have endogenous functional GPR-C6a protein and have or are transfected or genetically engineered to have function i-AR protein are further genetically engineered or transfected (stably or transiently) to contain an androgen-inducible reporter gene.

[458] 6. The method of embodiment 1 , 2, 3 or 4 wherein the cancer cells or transformed normal cells of the suitable test system or the suitable in vitro cell-based test system are genetically engineered or transfected (stably or transiently) to contain functional GPR-C6a gene. For some embodiments, this method uses a suitable test systems where the cells of the test system have or are genetically engineered to contain functional GPR-C6a gene in Ar ^~ cancer or transformed cells or the cells of the test system have or are genetically Docket No. 354. PATENT engineered to contain functional i-AR gene in GPR-C6a ^{+ +} cancer or transformed cells. In more preferred embodiments the cells of the suitable test system or suitable in vitro test system having or genetically engineered to have functional GPR-C6a and i-AR genes are further transfected (stably or transiently) or genetically engineered to contain an androgen-inducible reporter gene.

[459] 7. The method of any one of embodiments 1 -6 wherein the cells of the suitable test system or the suitable in vitro cell-based test system are cancer cells wherein the cancer cells are or derived from an androgen signaling-dependent prostate cancer cell line. The androgen-dependent prostate cancer cells typically have minimal baseline AR transactivation activity for a suitable test system when placed into androgen- and growth factor-depleted media and thus in some embodiments are supported by added androgen or another i-AR agonist, (i.e., in these embodiments the suitable test system comprises androgen-dependent prostate cancer in androgen- and growth factor-depleted media supplemented with i-AR agonist). Androgen precursors may also be used if the cancer cells are competent to biotransform the androgen precursor to active androgen.

[460] 8. The method of any one of embodiments 1 -6 wherein the cells of the suitable test system or the suitable in vitro cell-based test system are cancer cells wherein the cancer cells are an androgen-independent or CRPC prostate cancer cell line or are derived therefrom. Those prostate cancer cells typically have sufficient baseline AR transcriptional activity for a suitable test system when placed into androgen-depleted media. In some embodiments this baseline activity is increased by added i-AR agonist.

[461] 9. The method of any one of embodiments 1 -8 wherein the cells of the suitable test system or the suitable in vitro cell-based test system are cancer cells wherein the cancer cells are transfected (stably or transiently) to contain an androgen-inducible reporter gene. In some of these embodiments the upstream promoter in the inducible reporter gene is the ARE promoter for PSA. In other embodiments this promoter is the probasin promoter containing ARE. In other embodiments the reporter gene encodes for luciferase or chloramphenicol acetyl transferase (CAT). In preferred embodiments the cancer cells are transfected with a gene construct that contains the probasin promoter upstream to a luciferase reporter.

[462] 10. The method of any one embodiments 1 -9 wherein the cells of the suitable test system or the suitable in vitro cell-based test system are cancer cells wherein the cancer cells comprise (or consists essentially of or consists of) prostate cancer cells incubated or passaged under hormone- and growth factor-depleted conditions.

[463] In preferred embodiments incubation of the prostate cancer cells is supplemented Docket No. 354. PATENT by androgen or androgen precursor, preferably 3a-diol, pAED, testosterone or DHT, or another i-AR agonist. In more preferred embodiments incubation of the prostate cancer cells is LP LNCaP supplemented by pAED. In some embodiments use of supplemented androgen or androgen precursor requires the cancer cells to contain functional T877A mt- AR protein so that supplementation stimulates i-AR signaling or requires cancer cells competent to biotransform androgen precursor to active androgen.

[464] 11 . The method of embodiment 10 wherein the suitable test system or the suitable in vitro cell-based test system comprises prostate cancer cells passaged or incubated in androgen- and growth factor-depleted media wherein the androgen- and growth factor- depleted media is charcoal-stripped RPMI serum.

[465] 12. The method of embodiment 10 or 1 1 wherein the androgen- and growth factor-depleted media is supplemented by an androgen or other i-AR agonist.

[466] 13. The method of embodiment 10, 1 1 or 12 wherein the prostate cancer cells are LNCaP cells.

[467] 14. The method of embodiment 10, 1 1 , 12 or 13 wherein the suitable test system or the suitable in vitro cell-based test system comprises (or consists essentially of or consists of) LP LNCaP cells co-contacted in step (a) with pAED in an amount effective to activate AR transactivation in the absence of test compound.

[468] 15. The method of embodiment 10, 1 1 , 12 or 13 wherein the suitable test system or the suitable in vitro cell-based test system comprises (or consists essentially of or consists of) HP LNCaP cells.

[469] 16. The method of any one of embodiment 1 -13 or 15 wherein the suitable test system or the suitable in vitro cell-based test system comprises (or consists essentially of or consists of) C4-2B cells.

[470] 17. The method of any one embodiments 1 -16 further comprising (or further consists essentially of or further consists of): (a-2) contacting a test compound with an initial suitable in vitro cell-based test system comprising (i) MDA-kb2 cells transfected to contain a MMTV promoter-reporter gene, (ii) T47D-kBluc cells transfected to contain an estrogen-inducible promoter-reporter gene, (iii) HEK293T cells transfected to contain an androgen-inducible reporter gene and a gene encoding functional i-AR protein, (iv) HEK293T cells containing an estrogen-inducible reporter gene and a gene encoding functional ERp protein, or (v) a combination of (i)-(iv); and (a-1 ) selecting for the conduct or performance of step (a) a test compound from step (a-2) that induces transcription of the reporter(s) in a manner that is qualitatively similar to E-3a-diol as positive control. Docket No. 354. PATENT

[471] In some of these embodiments this step conducted prior to step (a) is used to screen for test compounds that are capable of inducing reporter gene transcription in one or more of the initial suitable in vitro cell-based test system. Such compounds are then selected for determining modulation effects on Erk phosphorylation states according to step (b). In preferred embodiments the test compound induces reporter gene transcription quantitatively similar to E-3a-diol as positive control.

[472] 18. The method of embodiment 17 wherein the functional i-AR protein is wt-AR, T877A mt-AR or an i-AR ligand-binding domain (LBD) fused to the AR Gal4 DNA binding domain.

[473] 19. The method of any one of embodiments 1 -18 further comprising (or further consisting essentially of or further consisting of): (d) determining modulation of redox activity of a functional oxidoreductase protein with the test compound selected from step (c) in the same test system of step (a), a different suitable in vitro cell-based test system, or a suitable cell-free test system; and (e) selecting the test compound that additionally negatively modulates oxidoreductase 3a-hydroxy steroid oxidation or steroid 3-one reduction activity.

[474] In preferred embodiments the oxidoreductase is capable of operating in oxidative mode with concommitant NAD(P)+ conversion to NAD(P)H towards 3a-hydroxy steroids. In other preferred embodiments the oxidoreductase is present in a suitable in vivo cell- based test system and is preferentially active or has physiologically relevant activity towards the 3a-hydroxy group of 3a-hydroxy steroids. In more preferred embodiments the oxidoreductase is active towards the steroid 3a-diol or is capable of interconverting 3a- diol and DHT.

[475] 20. The method of embodiment 19 wherein the oxidoreductase is an aldo-keto reductase, a short-chain dehydrogenase-reductase (SDR), a hydroxyacyl CoA dehydrogenase or a 17p-hydroxysteroid dehydrogenase (17p-HSD) having 3a- hydroxysteroid oxidoreductase activity. In preferred embodiments a suitable cell-free test system for determining modulation of oxidoreductase activity operates the enzyme in oxidative mode and is capable of oxidizing naturally occurring 3a-hydroxy steroids to the corresponding 3-one or reducing a 3-one containing steroid to the corresponding 3a- hydroxysteroid.

[476] 21 . The method of embodiment 20 wherein the test compound negatively modulates oxidative activity of 17p-HSD Type 10 (17HSD10). In preferred embodiments the ability of a test compound to negatively modulate this activity is evaluated in a suitable Docket No. 354. PATENT cell free test system by monitoring the production of NADH from oxidation of an alcohol substrate in the absence and presence of test compound. Example assay conditions and alcohol substrates are described in Shafquat, N. et al. (2003), which are incorporated by reference herein.

[477] 22. The method of embodiment 20 wherein the test compound inhibits 17p-HSD Type 10 interconversion of 3a-diol and DHT. In some of these embodiments the identified candidate compound selected from step (c) inhibits the oxidation of 3a-diol to DHT as determined in the cell-free test system at about 10 μΜ concentration or less, preferably at about 100 nM or less, more preferably at 10 nM or less.

[478] 23. The method of any one of embodiments 1 -22 further comprising (or further consisting essentially of or further consisting of): (f) determining binding interaction, optionally by SILAC, of test compound selected from step (c) or step (e) to the scaffold protein Tfg or to Erk protein(s), optionally in the presence of Tfg, Erk-1 and Erk-2 or another scaffold protein or a MAPK phosphatase, and (g) selecting a test compound that in addition to the characteristics of step (c) or steps (c) and (e) binds to Tfg or selectively binds to Erk-2 protein in comparison to Erk-1 .

[479] In preferred embodiments the identified candidate compound positively modulates protein-protein interaction with Tfg or positively modulates protein-protein interaction with Erk-2 in comparison to Erk-1 . In some of those embodiments the protein-protein interaction positively modulated is between Erk-2 and another signal transduction protein, such as a scaffold protein or a phosphatase, to inhibit Erk-2 phosphorylation, enhance pErk-2 de-phosphorylation or inhibit pErk-2 translocation to the nucleus in comparison to the other Erk isoform. In more preferred embodiments, protein-protein interaction of Erk-2 with Tfg, either directly or indirectly through other signal transduction components within a protein complex, is enhanced by the selected test or candidate compound. In more preferred embodiments a selected test compound identified as a candidate compound in step (g) has the characteristics of steps (c) and (e).

[480] 24. The method of any one embodiments 1 -23 comprising further comprising (or further consisting essentially of or further consisting of): (h) contacting a selected test compound from step (c) or step (e) with cancer cells of a suitable in vivo test system or co-contacting the selected test compound with a positive control cancer chemotherapeutic compound; and (j) selecting the test compound that inhibits cancer cell proliferation statistically significant to cancer cells contacted with vehicle alone, inhibits cell proliferation to similar or greater extent than cancer cells contacted to the positive control alone, or synergistically inhibits cancer cell proliferation when cancer cells are co- Docket No. 354. PATENT contacted with positive control.

[481] In preferred embodiments the in vivo test system is a xenograft animal model resulting from implantation into a rodent such as mice or rats of (1 ) cancer cells from a hormone-associated cancer cell line that is dependent on i-AR signaling for survival or proliferation (2) cancer cells containing functional i-AR or (3) cancer cells that are transfected (stably or transiently) to express or overexpress a gene or genes encoding functional i-AR and/or GPR-C6a or (4) transformed normal cells not containing endogenous i-AR and GPR-C6a proteins that are transfected (stably or transiently) to contain functional i-AR and GPR-C6a proteins. In preferred embodiments the implanted cells are cancer cells from a hormone-associated cancer cell line that is dependent on i- AR signaling for survival or proliferation. In some of those preferred embodiments the cancer cells are from a prostate or breast cancer cell line such as LNCaP, LuCaP, C4-2B, CWR22 or MCF-7. In other preferred embodiments a selected test compound identified as a candidate compound in step (j) in addition to the characteristics of step (c) or steps (c) and (e) or steps (c) and (g) or steps (c), (e) and (g) binds to Tfg or selectively binds to Erk- 2 protein in comparison to Erk-1 . In more preferred embodiments a selected test compound identified as a candidate compound in step (j) has the characteristics of steps (c), (e) or steps (c) and (g). In particularly preferred embodiments a selected test compound identified as a candidate compound in step (j) has the characteristics of steps and (c), (e) and (g).

[482] 25. The method of embodiment 24 wherein the in vivo test system is a xenograft animal model resulting from implantation of cancer cells from a hormone-associated cancer cell line or transformed cell line that are dependent on AR transactivation for survival or proliferation into an immune-compromised rodent. In some embodiments the cells are from a prostate or breast cancer cell line implanted into castrated SCID mice wherein cancer cell proliferation is optionally supported by androgen or estrogen administration.

[483] 26. The method of embodiment 25 wherein the cancer cells are prostate cancer cells implanted into immune-compromised rodent(s), wherein the immune-compromised rodent(s) are castrated SCID or athymic nu/nu mice, and wherein AR transactivation is supported by androgen supplementation.

[484] 27. The method of embodiment 26 wherein the prostate cancer cells are LNCaP cells implanted into castrated SCID mice supplemented with βΑΕϋ implant.

[485] 28. The method of embodiment 26 wherein the prostate cancer cells are LuCaP- 35V cells implanted into castrated SCID mice supplemented with βΑΕϋ implant. Docket No. 354. PATENT

[486] 29. The method of embodiment 26 wherein the prostate cancer cells are CWR22- R1 cells in nu/nu athymic mice supplemented by testosterone. In some of these embodiments testosterone supplementation is by administration of a prodrug such as hydroxyl ester of the testosterone 3p-hydroxy group. These hydroxyl esters include acetate, enanthate, propionate, isopropionate, isobutyrate, butyrate, valerate, caproate, isocaproate, hexanoate, heptanoate, octanoate, nonanoate, decanoate, undecanoate, phenylacetate, benzoate and hydroxyl esters described in Becker, K.L., Ed. (2001 ), the disclosure of which is incorporated by reference herein.

[487] 30. The method of embodiment 26 wherein the prostate cancer cells are C4-2B or LuCaP-35V cells implanted into castrated SCID mice or CWR22-R1 cells implanted into nu/nu athymic mice.

[488] 31 . The method of any one of embodiment 24-30 wherein the in vivo test system is a xenograft animal model resulting from implantation into a rodent of cancer cells from a hormone-associated cancer cell line or transformed cell line dependent on AR transactivation for survival or proliferation, wherein the rodent is immune-compromised and wherein the cancer cells implanted into an immune compromised rodent are co- contacted with test compound and a cancer chemotherapeutic compound.

[489] Those chemotherapeutic compounds include tubulin disrupting agents (i.e., compounds that aberrantly stabilize or destabilize tubulin polymers or interfere with tubulin polymerization). Other cancer chemotherapeutic compounds include anti-metabolites (e.g., methotrexate, fluorouracil, azathioprine and mercaptopurine), tyrosine kinase inhibitors, (e.g., imatinib, gefitinib, erlotinib and others reviewed in Shawver, L.K. et al. (2002) and Hsp90 inhibitors (e.g., glendamycin and others reviewed in Neckers, L. (2002)).

[490] Additional cancer chemotherapeutic compounds contemplated for this embodiment are DNA damaging agents. These agents include alkylating agents (e.g., the nitrogen mustards cyclophosphamide, mechlorethamine, uramustine, melphalan, chlorambucil and Ifosfamide, the nitrosoureas carmustine, lomustine and streptozocin and the alkyl sulfonate busulfan). Other DNA damaging agents are free-radical generating and DNA crosslinking compounds (e.g., platinum-containing compounds, such as cisplatin, carboplatin and oxaliplatin, and calicheamicins, such as calicheamicin-γι and calicheamicin T). Other contemplated cancer chemotherapeutic compounds are topoisomerase I inhibitors (e.g., anthracyclines, which includes daunosamine and tetrahydro-naphthacenedione-based compounds, such as doxorubicin, duanorubicin, valrubicin, idarubicin and epirubicin, which may also act as DNA damaging agents, and Docket No. 354. PATENT camptothecins, such as irinotecan and topotecan) and topoisom erase II inhibitors. In addition, anti-tumor antibiotics such as bleomycin, plicamycin and mitomycin, and epipodophyllotoxins, including etoposide and teniposide, which are topoisomerase II inhibitors, may also be co-contacted with a test compound to a suitable test system comprised of (or consisting essentially of or consisting of) cancer cells. These and other cancer chemotherapeutic compounds that may be used for co-contacting with a test or candidate compound to cancer cells of a suitable test system are provided in Boik, J. (2001 ).

[491] Preferred cancer chemotherapeutic compounds are tubulin disrupting agents, which include taxanes (e.g., paclitaxel, docetaxel and cabazitaxel) and vinca alkaloid compounds (e.g., vinblastine, vincristine, vinorelbine and vindesine).

[492] 32. The method of embodiment 24 wherein the in vivo test system is a rodent with induced or spontaneous mammary tumors. Induced and spontaneous mammary tumor models are reviewed in Medina, D. (2000) and Medina, D. and Thompson, H. (2000). Preferred mammary tumor models are carcinogen-induced tumors derived from exposure of a rodent (e.g., mouse or rat) to 7,12-dimethylbenz[a]anthracene (DMBA) or N-methyl-N- nitrosourea (MNU). Such carcinogen-induced tumor models are reviewed in Medina, D. (2008), which are incorporated by reference herein.

[493] 33. The method of embodiment 32 wherein the rodent is Lewis rat and the mammary tumor is carcinogen-induced by N-methyl-N-nitrosourea.

[494] 34. The method of embodiment 32 or 33 wherein the cancer cells are contacted with test compound or co-contacted with test compound and a cancer chemotherapeutic compound by systemic administration(s) of the compounds to the rodent. In preferred embodiments the co-contacted cancer chemotherapeutic is administered intraperitoneally.

[495] In some of these embodiments the cancer chemotherapeutic compound is a tubulin disrupting agent. These agents include a taxane compound, such as paclitaxel, docetaxel or cabazitaxel or a vindoline compound such as vinblastine or vincristine. In some embodiments rodents implanted with cells containing functional i-AR derived from a prostate cancer cell line are treated with an anti-androgen which antagonizes binding of DHT to the androgen nuclear hormone receptor, which inhibits the transformation of testosterone to DHT or which inhibits adrenal production of androgens. In certain of the embodiments the anti-androgen is a AR antagonist, 5a-reductase inhibitor, a cytochrome P450 lyase inhibitor or a LH-RH agonist or antagonist and includes MDV3100, flutamide, hydroxyflutamide, nitulamide, bicalutamide, cyproterone, ketoconazole, abiraterone, finasteride, dutasteride, diethylstilbestrol, leuprolide, bruserelin, goserelin and abarelix. In Docket No. 354. PATENT other embodiments rodents implanted with cells containing functional i-AR derived from a breast cancer cell are treated with an anti-estrogen. In certain embodiments the anti- estrogen is a ERa receptor antagonist, an anti-aromatase or a LHRH super-agonist including tamoxifen, fulvestrant, letrozole anastrozole and leuprolide.

[496] 35. The method of embodiment 34 wherein the cancer chemotherapeutic compound is a tubulin disrupting agent.

[497] 36. The method of embodiment 35 wherein the tubulin disrupting agent is a taxane compound.

[498] 37. The method of embodiment 36 wherein the taxane compound is docetaxel or paclitaxel.

[499] 38. A method to identify a candidate compound, the method comprising (or consisting of or consisting essentially of): (a) contacting a test compound with a suitable test system (a first test system); (b) determining phosphorylation states of Erk-1 and Erk-2 proteins resulting from step (a); (c) determining the phosphorylation state of a Class 1 PI3K resulting from step (a), or

[500] (a') contacting a test compound with a first suitable test system; (a") contacting the test compound of step (a') with a second suitable test system; (b') determining phosphorylation states of Erk-1 and Erk-2 proteins resulting from step (a'); (c') determining the phosphorylation state of a Class 1 PI3K resulting from step (a"), wherein the first and second suitable test systems are the same or different test systems;

[501] and (d) selecting a test compound that positively modulated the phosphorylation state of Erk-1 activation loop relative to Erk-2 or negatively modulated the phosphorylation state of Erk-2 activation loop relative to Erk-1 of step (b) or (b') and negatively modulated the tyrosine phosphorylation state of p85a or ρ85β regulatory subunit of the Class 1 PI3K of step (c) or (c'),

[502] wherein the selected test compound from step (d) is identified as a candidate compound.

[503] In some embodiments the suitable test systems are suitable in cell-based in vitro test systems. In preferred embodiments the method is conducted with the first and second suitable test systems of steps (a') and (a") wherein the test systems are suitable in cell-based in vitro test systems. In more preferred embodiments the suitable cell-based intro test systems are comprised of mammalian cancer cell or transformed normal cells endogenously having or genetically engineered to be GPR-C6a ^+/+ AR ^+/+ . In those embodiments AR may be wild-type or encode a mutant intracellular AR protein such as Docket No. 354. PATENT

T877A AR. In particularly preferred embodiments the GPR-C6a AR ^+/+ cells harbor a disabling PTEN mutation.

[504] 39. The method of embodiment 38 wherein the suitable test system or systems independently are suitable cell-based in vitro test systems. In preferred embodiments the method is conducted with the first and second suitable test systems of steps (a') and (a") wherein the test systems are suitable cell-based in vitro test systems, more preferably with those test systems being comprised of cells from the same cell line.

[505] 40. The method of embodiment 38 or 39 wherein the suitable cell-based test system or systems comprises (or consists essentially of or consists of) cells from a prostate, breast, ovarian cancer or other cancer cell line dependent on transactivation of one or more genes with an upstream ARE promoter for survival or proliferation or comprises (or consists essentially of or consists of) cancer cells or transformed normal cells genetically engineered or transfected to contain functional i-AR gene.

[506] In preferred embodiments the suitable cell-based in vitro test system of step (a) or the suitable in vitro cell-based test systems of steps (a') and (a") consists essentially of cells from or derived from the same prostate or breast cancer cell line or the same transformed cells.

[507] 41 . The method of embodiment 40 wherein the suitable in vitro test system of step (a) or the suitable in vitro cell-based test system of step (a') or (a") comprises (or consists essentially of or consists of) cancer cells that are responsive to DHT or pAED or are i-AR' ^~ cancer cells or transformed normal cells not having functional endogenous i-AR protein that are genetically engineered or transfected to contain functional i-AR gene.

[508] In those embodiments either or both suitable in vitro cell-based test systems of steps (a') and (a"), i.e. the first and second test system comprises (or consists essentially of or consists of) cancer cells that are responsive to DHT or pAED or are i-AR ^' ' ^' cancer cells or transformed normal cells not having functional endogenous i-AR protein that are genetically engineered or transfected to contain functional i-AR gene. Preferred methods use steps (a') and (a") wherein the first and second in vitro cell-based test systems consists essentially of the same line of i-AR ^+/+ cancer or transformed cells.

[509] 42. The method of embodiment 38, 39, 40 or 41 wherein the cancer cells or transformed normal cells of the suitable in vitro cell-based test system of steps (a) or the cancer cells or transformed normal cells of the suitable cell-based test systems of steps (a') and (a") express functional endogenous GPR-C6a gene or are genetically engineered or transfected (stably or transiently) to express functional GPR-C6a gene. Docket No. 354. PATENT

[510] Preferred methods use steps (a') and (a") wherein the first and second in vitro cell- based test systems consists essentially of the same line of GPR-C6a ^+/+ cancer or transformed cells. In more preferred embodiments the cancer cells or transformed normal cells express functional endogenous GPR-C6a and i-AR genes or are genetically engineered or transfected (stably or transiently) to express functional GPR-C6a and i-AR genes.

[511] 43. The method of embodiment 38, 39, 40 or 41 wherein the cancer cells or transformed normal cells of the suitable in vitro cell-based test system of steps (a) or the cancer cells or transformed normal cells of the suitable cell-based test systems of steps (a') and (a") contain endogenous GPR-C6a protein or are genetically engineered or transfected (stably or transiently) so as to have functional GPR-C6a protein.

[512] 44. The method of embodiment 38, 39, 40, 41 , 42 or 43 wherein the cancer cells are from an androgen signaling-dependent prostate cancer cell line.

[513] 45. The method of embodiment 38, 39, 40, 41 , 42 or 43 wherein the cancer cells are from a CRPC cancer cell line.

[514] 46. The method of any one of embodiments 38-45 wherein the cancer cells are genetically engineered or transfected (stably or transiently) to express a functional androgen-inducible reporter gene.

[515] 47. The method of any one of embodiments 38-46 wherein the suitable test systems or cell-based in vitro test system(s) independently are comprised of (or consists essentially of or consists of) prostate cancer cells that were incubated or passaged under hormone- and growth factor-depleted conditions. In preferred and more embodiments the cells of suitable test system(s) or in vitro cell-based in vitro test system(s) are as described in embodiment 10.

[516] 48. The method of embodiment 47 wherein the prostate cancer cells are passaged or incubated in charcoal-stripped RPMI serum.

[517] 49. The method of embodiment 48 wherein the androgen- and growth factor- depleted media is supplemented by an androgen or other i-AR agonist.

[518] 50. The method of embodiment 49 wherein the prostate cancer cells are LNCaP cells.

[519] 51 . The method of embodiment 50 wherein the prostate cancer cells are LP LNCaP cells co-contacted in step (a) or (a') and (a") with pAED in effective amount to activate AR transactivation in the absence of test compound. Docket No. 354. PATENT

[520] 52. The method of embodiment 49 the prostate cancer cells are HP LNCaP cells.

[521] 53. The method of any one of embodiments 38-48 wherein the suitable test system or the suitable cell-based in vitro test system of step (a) or the first and second suitable test systems or the first and second suitable cell-based in vitro test systems of steps (a') and (a") comprises (or consists essentially of or consists of) C4-2B cells incubated or maintained in hormone- and growth factor-depleted medium.

[522] 54. The method of any one of embodiment 38-53 further comprising (or further consisting essentially of or further consisting of): (a-2) contacting a test compound with an initial suitable in vitro cell-based test system comprising (i) MDA-kb2 cells transfected to contain a MMTV promoter-reporter gene, (ii) T47D-kBluc cells transfected to contain an estrogen-inducible promoter-reporter gene, (iii) HEK293T cells transfected to contain an androgen-inducible reporter gene and a gene encoding functional i-AR protein, (iv) HEK293T cells containing an estrogen-inducible reporter gene and a gene encoding functional ERp protein, or (v) a combination of (i)-(iv); and (a-1 ) selecting for the conduct or performance of step (a) or steps (a') and (a") a test compound from step (a-2) that induces transcription of the reporter(s) in a manner that is qualitatively similar to E-3a-diol as positive control.

[523] In some of those embodiments the steps (a-2) and (a-1 ) is conducted prior to step (a) or steps (a') and (a") and is used to screen for test compounds that are capable of inducing reporter gene transcription in one or more of the initial suitable in vitro cell-based test system. In such embodiments the test compounds eliciting reporter gene transcription are then selected for determining modulation effects on Erk phosphorylation states according to step (b) or (b') or tyrosine p85 phosphorylation states according to step (c) or (c'). In preferred embodiments the test compound induces reporter gene transcription quantitatively similar to E-3a-diol as positive control. In other preferred embodiments the test compounds eliciting reporter gene transcription are then selected for determining modulation effects on Erk and p85 phosphorylation (i.e., step a-1 is conducted prior to steps (a), (b) and (c), or steps (a'), (a"), (b') and (c')

[524] 55. The method of any one of embodiments 40-54 wherein the functional i-AR is wt-AR, T877A mt-AR or an i-AR ligand-binding domain (LBD) fused to AR Gal4 DNA binding domain.

[525] 56. The method of any one of embodiments 38-55 further comprising (or further consisting essentially of or further consisting of): (d) determining modulation of redox activity of a functional oxidoreductase protein with the test compound selected from step (c) or step (c') in the same test system of step (a), (a') or (a"), a different suitable in vitro Docket No. 354. PATENT cell-based test system, or a suitable cell-free in vitro test system; and (e) selecting the test compound that additionally negatively modulates oxidoreductase 3a-hydroxy steroid oxidation or steroid 3-one reduction activity.

[526] In preferred and more preferred embodiments the oxidoreductase is as described in embodiment 19.

[527] 57. The method of embodiment 56 wherein the oxidoreductase is an aldo-keto reductase, a short-chain dehydrogenase-reductase (SDR), a hydroxyacyl CoA dehydrogenase or a 17p-hydroxysteroid dehydrogenase (17p-HSD) having 3a- hydroxysteroid oxidoreductase activity. In preferred embodiments a suitable cell-free test system for determining modulation of oxidoreductase activity is as described in embodiment 20.

[528] 58. The method of embodiment 57 wherein the test compound negatively modulates oxidative activity of 17p-HSD Type 10 (17HSD10). In preferred embodiments the ability of a test compound to negatively modulate this activity is evaluated in a suitable cell free test system by monitoring the production of NADH from oxidation of an alcohol substrate in the absence and presence of test compound. Example assay conditions and alcohol substrates are given by embodiment 21 .

[529] 59. The method of embodiment 58 wherein the test compound inhibits 17p-HSD Type 10 interconversion of 3a-diol and DHT. In some of these embodiments the identified candidate compound selected from step (c) or step (c') inhibits the oxidation of 3a-diol to DHT as determined in the cell-free test system at about 10 μΜ concentration or less, preferably at about 100 nM or less, more preferably at 10 nM or less.

[530] 60. The method of any one of embodiments 38-59 further comprising (or further consisting essentially of or further consisting of): (f) determining binding interaction, optionally by SILAC, of test compound selected from step (c), step (c') or step (e) to the scaffold protein Tfg or to Erk protein(s), optionally in the presence of Tfg, Erk-1 and Erk-2 or another scaffold protein or a MAPK phosphatase, and (g) selecting a test compound that in addition to the characteristics of step (c) or steps (c) and (e) binds to Tfg or selectively binds to Erk-2 protein in comparison to Erk-1 or (g') selecting a test compound that in addition to the characteristics of step (c') or steps (c') and (e) binds to Tfg or selectively binds to Erk-2 protein in comparison to Erk-1 .

[531] In preferred embodiments the identified candidate compound positively modulates protein-protein interaction with Tfg or positively modulates protein-protein interaction with Erk-2 in comparison to Erk-1 . In some of those embodiments the protein-protein Docket No. 354. PATENT interaction positively modulated is between Erk-2 and another signal transduction protein, such as a scaffold protein or a phosphatase, to inhibit Erk-2 phosphorylation, enhance pErk-2 de-phosphorylation or inhibit pErk-2 translocation to the nucleus in comparison to the other Erk isoform. In more preferred embodiments, protein-protein interaction of Erk-2 with Tfg, either directly or indirectly through other signal transduction components within a protein complex, is enhanced by the selected test or candidate compound. In more preferred embodiments a selected test compound identified as a candidate compound in step (g) or step (g') has the characteristics of steps (c) and (e) or steps (c') and (e), respectively.

[532] 61 . The method of any one embodiments 38-60 comprising further comprising (or further consisting essentially of or further consisting of): (h) contacting a selected test compound from step (c), step (c') or step (e) with cancer cells of a suitable in vivo test system or co-contacting the selected test compound with a positive control cancer chemotherapeutic compound; and (j) selecting the test compound that inhibits cancer cell proliferation statistically significant to cancer cells contacted with vehicle alone, inhibits cell proliferation to similar or greater extent than cancer cells contacted to the positive control alone, or synergistically inhibits cancer cell proliferation when cancer cells are co- contacted with positive control.

[533] In preferred embodiments the in vivo test system is a xenograft animal model as described in embodiment 24.

[534] In other preferred embodiments a selected test compound identified as a candidate compound in step (j) in addition to the characteristics of step (c) or steps (c) and (e) or steps (c) and (g) or steps (c), (e) and (g) binds to Tfg or selectively binds to Erk-2 protein in comparison to Erk-1. In more preferred embodiments a selected test compound identified as a candidate compound in step (j) has the characteristics of steps (c) and (e) or step (g).

[535] In more preferred embodiments a selected test compound identified as a candidate compound from step (j) has the characteristics of steps and (c') and (e) or step

(g')- [536] 62. The method of embodiment 61 wherein the in vivo test system is a xenograft animal model resulting from implantation of cancer cells from a hormone-associated cancer cell line or transformed cell line that are dependent on AR transactivation for survival or proliferation into an immune-compromised rodent. In some embodiments the cells are as described in embodiment 25. Docket No. 354. PATENT

[537] 63. The method of embodiment 62 wherein the cancer cells are prostate cancer cells implanted into immune-compromised rodent(s), wherein the immune-compromised rodent(s) are castrated SCID or athymic nu/nu mice, and wherein AR transactivation is supported by androgen supplementation.

[538] 64. The method of embodiment 63 wherein the prostate cancer cells are LNCaP cells implanted into castrated SCID mice supplemented with βΑΕϋ implant.

[539] 65. The method of embodiment 63 wherein the prostate cancer cells are LuCaP- 35V cells implanted into castrated SCID mice supplemented with βΑΕϋ implant.

[540] 66. The method of embodiment 63 wherein the prostate cancer cells are CWR22- R1 cells in nu/nu athymic mice supplemented by testosterone.

[541] In some of these embodiments testosterone supplementation is by administration of a prodrug as described in embodiment 29.

[542] 67. The method of embodiment 63 wherein the prostate cancer cells are C4-2B or LuCaP-35V cells implanted into castrated SCID mice or CWR22-R1 cells implanted into nu/nu athymic mice.

[543] 68. The method of any one of embodiments 61 -67 wherein the in vivo test system is a xenograft animal model resulting from implantation into a rodent of cancer cells from a hormone-associated cancer cell line or transformed cell line dependent on AR transactivation for survival or proliferation, wherein the rodent is immune-compromised and wherein the cancer cells implanted into an immune compromised rodent are co- contacted with test compound and a cancer chemotherapeutic compound. Those chemotherapeutic compounds are described by embodiment 31 and include tubulin disrupting agents, anti-metabolites, tyrosine kinase inhibitors, DNA damaging agents, topoisomerase inhibitors, and anti-tumor antibiotics.

[544] 69. The method of embodiment 61 wherein the in vivo test system is a rodent with induced or spontaneous mammary tumors. Induced and spontaneous mammary tumor models are described in embodiment 32.

[545] 70. The method of embodiment 69 wherein the rodent is Lewis rat and the mammary tumor is carcinogen-induced by N-methyl-N-nitrosourea.

[546] 71 . The method of embodiment 69 or 70 wherein the cancer cells are contacted with test compound or co-contacted with test compound and a cancer chemotherapeutic compound by systemic administration(s) of the compounds to the rodent.

[547] Preferred embodiments and exemplary cancer chemotherapeutic compounds Docket No. 354. PATENT including tubulin disrupting agents, anti-androgens and anti-estrogens administered to rodents implanted with cells containing functional i-AR derived from a prostate cancer or breast cancer cell line are as described in embodiment 34.

[548] 72. The method of embodiment 71 wherein the cancer chemotherapeutic compound is a tubulin disrupting agent.

[549] 73. The method of embodiment 72 wherein the tubulin disrupting agent is a taxane compound.

[550] 74. The method of embodiment 73 wherein the taxane compound is docetaxel or paclitaxel.

[551] 75. A method of screening for a low toxicity Erk modulator comprising (or consisting essentially of or consisting of): (a) contacting a test compound with a suitable test system of a previous claim such as claim 1 , 2, 3, 4, 38, 39, 40 or 41 or a suitable test system comprising (or consisting essentially of or consisting of) (ii) mammalian cells that expresses a gene or genes encoding functional Erk-1 , Erk-2 and G protein-coupled receptor C6a (GPR-C6a) proteins, or mammalian cells that expresses a gene or genes encoding functional Erk-1 , Erk-2, G protein-coupled receptor C6a (GPR-C6a) and functional PI3K proteins, or a suitable in vitro cell-free test system containing an artificial Erk substrate, and one or more components in the mammalian Ras-Erk signal transduction pathway needed for the Ras-Raf-MEK-Erk or Raf-MEK-Erk or MEK-Erk signal transduction; (b) determining modulation of an activity or phosphorylation state from step (a) of one or both Erk isoforms or of one or more upstream activators, regulators or downstream effector proteins or artificial substrates of the Erk isoforms whose activity(ies) or phosphorylation state(s) is(are) indicative of the activity or phosphorylation state of one or both Erk isoforms; (c) determining modulation of Ga/q subunit phosphorylation state, GTP/GDP bound states, GPR-C6a receptor activity or phosphorylation state, or activity of a regulator or downstream effector protein of the GPR-C6a receptor whose activity, phosphorylation state or Galpha/q GTP/GDP bound or phosphorylation states is indicative of an activity of the receptor or of released Ga/q subunit; (d) selecting the test compound that results in or is indicative of positive modulation of kinase activity, or phosphorylation state of the activation loop, of the Erk-1 isoform in comparison to that of Erk-2, or of negative modulation of kinase activity or activation loop phosphorylation state of the Erk-2 isoform in comparison to Erk-1 , and results in or is indicative of positive modulation of Ga/q tyrosine phosphorylation state or positively modulates the GTP-bound state of released Ga/q or modulates an activity of the GPRC-6a receptor, or phosphorylation state or activity of a downstream effector protein, indicative of inverse agonist activation of the Docket No. 354. PATENT receptor; (e) administering the test compound of step (d), now identified as a candidate compound, to an animal in a suitable in vivo test system to obtain a treated animal and determining toxicity to the treated animal; and (f) selecting the candidate compound from step (e) that has a therapeutic index of 2 or greater, wherein the selected candidate compound from step (f) is identified as a low toxicity Erk modulator.

[552] In preferred embodiments the downstream effector protein of the GPR-C6a receptor whose activity or phosphorylation state is indicative of an activity of that receptor or of released Ga/q subunit from that receptor is PI3K or Akt. In other preferred embodiments the downstream effector protein of the GPR-C6a receptor whose activity or phosphorylation state is indicative of the GTP/GDP Galpha/q bound or phosphorylation states is PI3K. In other preferred embodiments positive modulation of Ga/q tyrosine phosphorylation state or positive modulation of released GTP-bound Ga/q from GPR-C6a is indicative of inverse agonist activation of GPR-C6a. In more preferred embodiments negative modulation of the tyrosine phosphorylation state of the p85a or ρ85β regulatory subunit of a PI3K is indicative of inverse agonist activation of GPR-C6a. In other more preferred embodiments the test compound identified as a candidate compound for administration in step (e) positively modulates the phosphorylation state of the activation loop of the Erk-1 isoform with no substantial modulation of the activation loop of the Erk-2 isoform.

[553] 76. The method of embodiment 75 further comprising (or further consisting essentially of or further consisting of) (g) administering the low toxicity Erk modulator of step (f) to a human having a hyperproliferation condition or a cancer in one or more amounts effective to assess the toxicity and/or efficacy of the low toxicity Erk modulator to treat the hyperproliferation condition in the human so as to obtain a treated human and assessing the toxicity and/or efficacy of the low toxicity Erk modulator on the treated human; and (h) selecting the candidate compound from step (g) that has a therapeutic index of 2 or greater, wherein the selected candidate compound from step (h) is identified as a low toxicity Erk modulator.

[554] 77. The method of embodiment 75 wherein the determined modulated activity(ies) or phosphorylation state(s) determined in step (b) or (c) is(are) for functional (i) GPR-C6a receptor protein, (ii) GPR-C6a downstream effector protein(s) (iii) Erk-1 or an Erk-1 substrate or downstream effector protein or an upstream activator or regulator of Erk-1 , (iv) Erk-2 or an Erk-2 downstream substrate or effector protein or an upstream activator or regulator of Erk-2, (v) protein kinase C, a Src kinase, a Src-like kinase, a Ras kinase, a Raf kinase or a MEK kinase, optionally MEK-1 or MEK-2, (vi) a class 1 phosphoinositol-3 Docket No. 354. PATENT kinase regulatory subunit, (vii) protein kinase A, (viii) intracellular free cAMP or Ca ⁺ concentration, or (ix) a combination of (i)-(viii), wherein the determined modulated activity(ies) or phosphorylation state(s) effected by the candidate compound for administration in step (e) is qualitatively, quantitatively or substantially similar to the modulation(s) determined for E-3a-diol when used as positive control.

[555] In preferred embodiments the downstream GPR-C6a effector protein is a Ga monomer, wherein a test compound identified as a candidate compound for administration in step (e) positively modulates Galpha/q. In more preferred embodiments a test compound identified as a candidate compound for administration in step (e) positively modulates Galpha/q activity in preference to Galpha/i. In more preferred embodiments a test compound identified as a candidate compound for administration in step (e) positively modulates Galpha/q activity without positively modulating Galpha/i activity wherein said Ga monomer modulations are qualitatively, quantitatively or substantially similar to the modulations determined for E-3a-diol when used as positive control. In other more preferred embodiments a test compound identified as a candidate compound for administration in step (e) positively modulates the phosphorylation state of an effector downstream of Erk-1 (i.e., an Erk-1 substrate) without substantial positive modulation of the phosphorylation state of an effector protein specific to Erk-2. Examples of specific Erk-2 substrates are described in Carlson, S.M. et al. (201 1 ) and are incorporated by reference herein.

[556] 78. The method of embodiment 75 wherein the modulated activity(ies) or phosphorylation state(s) determined or measured for step, is(are) (i) phosphorylation state or phosphorylation rate (i.e., time course of pErk-1 formation), or cytosolic concentration of pErk-1 , (ii) phosphorylation state or phosphorylation rate of Erk-2, or the nuclear concentration of pErk-2, (iii) phosphorylation state or phosphorylation rate for an Erk-1/2 artificial substrate or downstream effector protein, (iv) activity of or expression of a gene encoding an apoptosis-associated protein, (v) nucleo-cytoplasmic translocation of pErk-1 or pErk-2 (vi) cytosolic activity of pErk-1 , (vii) nuclear activity of pErk-2, (viii) proliferation of the mammalian cells in the suitable test system, (ix) differentiation or gene expression indicative differentiation of mammalian cells in a suitable test system; (x) sensitivity of a mammalian cells to apoptosis or to a cytotoxic compound, (xi) amounts of phosphorylated Erk-1 and Erk-2 or ratio of pErk-1 to un-phosphorylated Erk-2 or rate of phosphorylation of an artificial Erk substrate by pErk-1 , pErk-2 or pErk-1/2, (xii) binding interactions of Erk-1 or Erk-2 in the presence of one or more scaffold proteins or components of the Ras-Erk signal transduction pathway, or (xiii) a combination of (i)-(xii). Docket No. 354. PATENT

[557] In some embodiments the rate of phosphorylation of an artificial Erk substrate by pErk-1 , pErk-2 or pErk-1/2 substrate is determined in a suitable cell-free test system by phosphorylation of myelin basic protein or a protein or peptide that is capable of phosphorylation by Erk-1 , Erk-2 or Erk-1/2 that is conjugated to a peptide or contains an amino acid sequence capable of binding to the Erk CD domain wherein the suitable cell- free test system additionally comprises an Erk-binding scaffold protein. In those embodiments a candidate compound identified for administration in step (e) has no discernable effects when the cell-free test system contains no Erk scaffold proteins, but exhibits negative modulation of pErk-2 protein levels when contacted to cells of a suitable cell-based test system.

[558] In other embodiments differentiation or gene expression in mammalian cells in a suitable test system that is indicative of differentiation include PSA gene expression in mammalian prostate cancer cells, or expression of a suitable reporter gene genetically engineered into prostate cancer cells, in the absence of cell proliferation. In some of those embodiments the mammalian prostate cancer cells are LNCaP cells or mammalian cancer cells expressing genes encoding functional wt-AR or mt-AR protein.

[559] In other embodiments phosphorylation states or phosphorylation rates for an Erk- 1/2 downstream effector protein that are determined include one or more s6 kinases (i.e., p90rsk, RSK-1 or RSK-2), MSK or transcription factors, including Elk-1 , c-Myc, c-Fos, SRF or CREB. In other embodiments activit(ies) for one or more apoptosis-associated proteins, including Bim, Bad or Bcl-2 or effects on gene expression for genes encoding one or more of those proteins are determined.

[560] In preferred embodiments a candidate compound for administration in step (e) increases pErk-1 levels relative to pErk-2 within 5-30 min of contacting the compound to cells of a suitable test system, more preferably without increasing p-Erk-2 10% or more above basal levels in quiescent cells. More preferably the increase in p-Erk-1 is preceded by increased intracellular Ca2+ concentrations with an increase due to intracellular mobilization particularly preferred. In other preferred embodiments a candidate compound for administration in step (e) decreases Bcl-2 gene expression when contacted to cells of a suitable test system wherein the cells are mammalian cancer cells.

[561] 80. The method of embodiment 75 wherein the phosphorylation state of functional Erk-2 in mammalian cells of a suitable test system is negatively modulated (i.e., decreased or unchanged relative to basal levels) wherein the mammalian cells are cancer or transformed normal cells. In preferred embodiments the mammalian cells are endogenously expressing AR ^+/+ GPR-C6a ^+/+ prostate or breast cancer cells, are Docket No. 354. PATENT endogenously expressing GPR-C6a CRPC cells genetically engineered to have functional i-AR., or transformed HEK-293 genetically engineered to have functional i-AR and GPR-C6a genes.

[562] In embodiments 75-80 modulations of kinase, receptors or Ga protein activities or protein phosphorylation states are preferably determined relative to one or more suitable negative and positive control test compounds. For example, positive control compounds for GPR-C6a Ga/q activation include E-3a-diol which is also a negative control compound for GPR-C6a Galpha/i activation. Positive control compounds for GPR-C6a Galpha/i activation include 3a-diol which is also a negative control compound for GPR-C6a Ga/q activation. Other positive control compounds for GPR-C6a-Galpha/i activation include T, DHT, T-BSA.

[563] 81 . The method of embodiment 75 wherein the low toxicity of the low toxicity Erk-1 modulator does not (i) limit administration of the low toxicity Erk-1 modulator to the animal of step (e) or the treated human of step (g) to sub-optimal or non-effective amount, (ii) cause a toxicologically or clinically significant increase in liver enzymes in the treated animal of step (e) or the treated human of step (g), (ii) cause a toxicologically or clinically significant skin rash in the animal of step (e) or the treated human of step (g), and/or (iii) cause a toxicologically or clinically significant degree of a central nervous system (CNS) toxicity such as one, two or more of a seizure, confusion, loss of consciousness, coma, depression, psychosis or delirium, optionally wherein the CNS toxicity is characterized by seizure and/or confusion.

[564] 82. The method of embodiment 75 wherein the mammalian cells endogenously express a gene encoding functional intracellular androgen receptor (i-AR) protein. Preferred mammalian cells include cancer cells, with prostate and breast cancer cells particularly preferred.

[565] 83. The method of embodiment 75 wherein the mammalian cells express a gene encoding functional intracellular estrogen receptor (i-ER) protein or express genes encoding i-AR and i-ER proteins.

[566] 84. The method of embodiment 75 wherein serum albumin or one or more liver enzymes selected from the group consisting of alanine transaminase, aspartate transaminase, alkaline phosphatase and γ glutamyl transpeptidase are not toxicologically or clinically significantly increased, preferably in about 50% or more of treated animals in one or more groups of treated animals or in about 70% to about 90% of treated animals.

[567] 85. The method of embodiment 76 wherein human serum albumin or one or more Docket No. 354. PATENT of human alanine transaminase, aspartate transaminase, alkaline phosphatase and γ glutamyl transpeptidase are not increased more than about 2-fold compared to normal values, preferably in about 70% or more of the treated humans in one or more groups of treated humans or in about 70% to about 95% of the treated humans.

[568] 86. The method of embodiment 84 wherein rodent serum albumin or one or more of rodent alanine transaminase, aspartate transaminase, alkaline phosphatase and γ- glutamyl transpeptidase are not increased more than about 2.5-fold or about 3-fold compared to normal values. In preferred embodiments the rodent is a mouse or rat and about 70% or more of the treated mice or rats or in about 70% to about 95% of the treated mice or rats have none of the recited proteins increased more than about 2.5 fold..

[569] 87. The method of embodiment 75 wherein the treated animals of step (e) have a cancer, precancer or hyperplasia.

[570] In some embodiments the treated animals of step (e) have prostate cancer breast cancer, ovarian cancer, endometriosis, liver cancer, a central nervous system cancer, a lung cancer (e.g., small cell lung cancer, non-small cell lung cancer), colorectal cancer, myeloma, melanoma, lymphoma, thyroid cancer, pancreatic cancer or bone cancer or a cancer that metastasized to bone or liver.

[571] In preferred embodiments the treated aminals of step (e) have prostate cancer, benign prostatic hypertrophy, breast cancer, ovarian cancer, endometriosis, hepatocellular carcinoma, glioblastoma, astrocytoma or oligodendroglioma.

[572] In more preferred embodiments the treated animals of step (e) have prostate or breast cancer, including metastatic prostate cancer or metastatic breast cancer, or cancer that has metastasized to bone or liver.

[573] 88. The method of embodiment 75 wherein functional GPR-C6a gene in the cell is introduced by a stably or transiently transfected expression vector.

[574] 89. The method of embodiment 75 wherein the suitable test system of step (a) is a suitable in vitro system comprised of mammalian cells derived from a tumor or transformed cell line. In preferred embodiments the mammalian tumor cells are LP LNCaP cells or CRPC cells derived therefrom including HP LNCaP cells.

[575] 90. The method of embodiment 75 wherein determining the modulation of Erk-1 and Erk-2 activities by the test compound of step (a) is by

[576] (i) contacting a test compound of step (a) with a suitable cell-free test system comprising p-Erk-1 , pErk-2, or p-Erk-1/2 and a suitable artificial substrate of the Erk Docket No. 354. PATENT isoform(s) in the absence of Erk scaffold proteins;

[577] (ii) contacting a test compound of step (i) with a suitable cell-free test system comprising p-Erk-1 , pErk-2, or p-Erk-1/2 and the suitable Erk isoforms substrate(s) of step (i) and at least one Erk scaffold proteins, wherein the suitable cell-free test systems of steps (i) and (ii) are substantially the same except for the presence or absence of the Erk scaffold protein(s);

[578] (iii) determining amounts or rates of pErk-1 and pErk-2 or pErk-1/2 phosphorylation of the suitable artificial substrate wherein the kinase activity(ies) of the Erk isoform(s) determine modulations of activities or phosphorylation states of step (b) in the presence of test compound;

[579] (iv) determining the amounts or rates of pErk-1 , pErk-2 or pErk-1/2 phosphorylation of the suitable artificial substrate of steps (i) in the absence of the test compound of step (i), wherein step (iv) serves as the vehicle control;

[580] (v) selecting a test compound that negatively modulates pErk-2 or pErk-1/2 kinase activity in preference to pErk-1 in comparison to vehicle controls in the presence of Erk scaffold protein(s) and wherein the selected test compound has no discernable effect on pErk-1 and pErk-2 kinase activities in the absence of the Erk scaffold protein(s).

[581] In some embodiments the amounts of increased pErk-1 relative to pErk-2 kinase activities are additionally determined in the presence of an Erk ATP binding site- dependent inhibitor subsequent to contacting a suitable test system with the test compound of step (a), wherein the selective negative modulation of p-Erk2 kinase activity of the selected candidate compound for conducting step (e) is attenuated or abrogated (i.e., both isoforms become negatively modulated).

[582] 91 . The method of embodiment 75 wherein determining the modulation of Erk-1 and Erk-2 activities by the test compound of step (a) is by

[583] (i) contacting a test compound of step (a) with a suitable in vivo test system comprising mammalian cancer cells;

[584] (iii) determining amounts or rates of formation of pErk-1 and pErk-2 proteins relative to total Erk protein wherein the amounts or rates determine modulations of activities or phosphorylation states of step (b) in the presence of test compound;

[585] (iv) determining the amounts or rates of formation of pErk-1 and pErk-2 proteins relative to total Erk protein of the suitable artificial substrates of steps (i) in the absence of the test compound of step (i), wherein step (iv) serves as vehicle control; Docket No. 354. PATENT

[586] (v) selecting a test compound that negatively modulates pErk-2 or pErk-1/2 kinase proteins levels relative to total Erk protein level in preference to pErk-1 in comparison to vehicle controls in the presence of Erk scaffold protein(s).

[587] In some embodiments increased pErk-1 protein relative to pErk-2 kinase activities is additionally determined in the presence of an inhibitor of Ras-Erk, TNFa-NF-κΒ, PI3K- Akt, GPCR, β-arrestin, Raf-1 , Raf-B, Tpl-2, RTK or Src signal transduction wherein the selective negative modulation of p-Erk2 protein level by the candidate compound selected for conducting step (e) is attenuated or abrogated (i.e., both isoforms become negatively modulated).

[588] 92. The method of embodiment 75 further comprising (or consisting essentially of or consisting of): (j) determining binding interaction of the test compound selected from step (c) to the scaffold protein Tfg in the presence of Erk-1 and Erk-2); and (k) selecting the test compound that binds to Tfg and selectively binds to Erk-2 protein in comparison to Erk-1 for administration in step (e).

[589] In some embodiments the binding interaction to the scaffold protein Tfg of step (j) is conducted in the presence of another scaffold protein or a MAPK phosphatase

[590] 93. The method of embodiment 75 or 92 further comprising (or consisting essentially of or consisting of): (m) determining modulation of activity of a functional oxidoreductase protein; and (n) selecting the test compound that negatively modulates 3a-hydroxy steroid oxidation by the oxidoreductase for administration in step (e).

[591] 94. The method of embodiment 93 wherein the oxidoreductase is 17β- hydroxysteroid dehydrogenase type 10.

[592] 95. The method of embodiment 93 wherein the test compound negatively modulates 3a-hydroxy steroid oxidation of 3a-diol.

[593] 1A. A method to identify a candidate compound, the method comprising (a) contacting a test compound with a suitable test system (a first test system); (b) determining phosphorylation states of Erk-1 and Erk-2 proteins resulting from step (a); (c) determining the phosphorylation state of a Class 1 PI3K using the suitable test system of step (a) or using a different suitable test system (a second test system), after contacting the test compound with the first test system or the second test system; wherein the first and second suitable test systems are in vitro cell-based test systems comprising cells derived from a prostate or breast cancer cell line having endogenous GPR-C6a and intracellular AR proteins or transformed normal cells not having endogenous GPR-C6a and intracellular AR proteins that have been transfected to contain those proteins; and (d) Docket No. 354. PATENT selecting a test compound from step (c) that positively modulated the phosphorylation state of Erk-1 activation loop relative to Erk-2, wherein the phosphorylation state of Erk-2 activation loop is not positively modulated, and negatively modulated the tyrosine phosphorylation state of a p85a or ρ85β regulatory subunit of the Class 1 PI3K, wherein the selected test compound from step (d) is identified as a candidate compound.

[594] 2A. The method of embodiment 1 A wherein the in vitro cell-based test systems of steps (a) and (b) comprises low passage (LP) LNCaP cells and wherein the LP LNCaP cells of step (a) are co-contacted with pAED in an amount effective to induce AR transactivation in the absence of test compound.

[595] 3A. The method of embodiment 1 A further comprising (a-2) contacting a test compound with a suitable cell-based test system prior to step (a) wherein the prior suitable test system comprises (i) MDA-kb2 cells transfected to contain a MMTV promoter-reporter gene, (ii) T47D-kBluc cells transfected to contain an estrogen-inducible promoter-reporter gene, (iii) HEK293T cells transfected to contain an androgen-inducible reporter gene and a gene encoding functional i-AR protein, or (iv) HEK293T cells containing an estrogen-inducible reporter gene and a gene encoding functional ERp protein; and (a-1 ) selecting a test compound from step (a-1 ) that induces transcription of at least one of the reporter for conducting step (a).

[596] 4A. The method of embodiment 1 A further comprising (e) contacting a selected test compound from step (d) with prostate or breast cancer cells of a suitable in vivo test system; (f) determining cancer cell proliferation in the suitable in vivo test system resulting from step (e); and (g) selecting a test compound from step (f) that inhibits cancer cell proliferation statistically significant to cancer cells contacted with vehicle alone.

[597] 5A. The method of embodiment 4A wherein the in vivo test system is a xenograft animal model resulting from implantation of cancer cells from a prostate or breast cancer cell line into an immune-compromised rodent.

[598] 6A. The method of embodiment 5A wherein the prostate cancer cells are LP LNCaP cells implanted into castrated SCID mice supplemented with pAED implant.

[599] 7A. The method of embodiment 5A wherein the prostate cancer cells are LuCaP- 35V cells implanted into castrated SCID mice supplemented with pAED implant.

[600] 8A. The method of embodiment 5A wherein prostate cancer cells are CWR22-R1 cells implanted into castrated nu/nu athymic mice supplemented by testosterone.

[601] 9A. The method of embodiment 5A wherein prostate cancer cells are C4-2B or Docket No. 354. PATENT cells implanted into castrated SCID mice.

[602] 10A. The method of embodiment 4A wherein the in vivo test system is a rodent with induced or spontaneous mammary tumors.

[603] 11 A. The method of embodiment 10A wherein the rodent is Lewis rat and the mammary tumor is carcinogen-induced by N-methyl-N-nitrosourea.

[604] 12A. The method of embodiment 4A further comprising (h) determining a minimum effective amount of a compound selected from step (g) for treating breast or prostate cancer in a mammal; (i) administering the minimum effective amount to healthy mammals so as to provide treated mammals; (j) determining liver enzymes levels for alanine transaminase, aspartate transaminase, alkaline phosphatase and γ glutamyl transpeptidase of the treated mammals; and (k) selecting a compound from step (i) that did not increase liver enzymes levels for any one of the liver enzymes selected from the group consisting of alanine transaminase, aspartate transaminase, alkaline phosphatase and γ glutamyl transpeptidase more than about 2-fold compared to normal values in about 70% or more of the treated mammals, wherein the compound of step (k) identified as a candidate compound is a low toxicity Erk-1 modulator.

[605] 13A. The method of embodiment 12A further comprising (I) determining the minimum effective amount of a compound selected from step (k) to elicit more than about a 2-fold increase levels in any of the liver enzymes selected from the group consisting of alanine transaminase, aspartate transaminase, alkaline phosphatase and γ-glutamyl transpeptidase; and (m) determining the therapeutic index of a compound based upon the effective amounts of steps (h) and (I); and (n) selecting the compound from step (I) having a therapeutic index of at least 5.

[606] All references cited herein are incorporated herein by reference. To the extent not already indicated, it will be understood by those of ordinary skill in the art that any of the various specific embodiments, analysis methods, compounds or compositions described herein may be modified to incorporate other appropriate features, e.g., where one or more protocol steps in a disclosed embodiment is added to or combined with any other compatible protocol step, method or embodiment described herein.

[607] 6. Examples The following examples further illustrate the invention and are not intended to limit the invention in any way.

[608] All animal procedures described herein were performed in compliance with National Institutes of Health guidelines and within established IACUC guidelines. E-3a-diol was administered in these procedures as a suspension in 30% cyclodextrin- Docket No. 354. PATENT

Sulfobutylether (Captisol from CyDex, Lenexa, KS) in water. Statistical analyses of E- 3a-diol in vivo effects were performed using unpaired Student t-tests. (Prism Graphpad, Graphpad Software) unless otherwise indicated. Results with differences yielding P < 0.05 were considered significant.

[609] Example 1 . Test Compound Binding Partner Determination by SI LAC Method

[610] Chemicals: All chemicals and solvents, including, L-arginine- ¹³ C ₆ ¹⁵ N ₄ -HCI, L- lysine- ¹³ C ₆ ¹⁵ N ₂ -HCI, 4-iodobenzylamine hydrochloride, 9-fluorenylmethyloxycarbonyl chloride (FMOC-CI), triethylamine, ethanolamine, palladium diacetate, triphenyl phosphine, cuprous iodide, piperidine, urea, iodoacetamide, ammonium bicarbonate, Tris(2-carboxyethyl)phosphine-HCI (TCEP), sodium chloride, formic acid, THF (anhydrous), DMF (anhydrous), and acetonitrile (ACN, HPLC grade) were purchased from Sigma-Aldrich (USA). The Pierce NHS-activated agarose slurry (26200) and SILAC RPMI media (89984) were purchased from Thermo Scientific (USA). The HEPES buffer was purchased from Mediatech (USA) and the dialyzed fetal bovine serum was purchased from Sigma-Aldrich.

[611] Agarose Coupling to 17a-Ethynyl Steroids: 17a-ethynyl-5a-androstane-3a, 17β- diol (E-3a-diol) was synthesized as described in U.S. Pat. Appl. Pub. No. 2009-0291932- A1 (1 1/26/2009), which is incorporated by reference herein. The preparation of an amino analog of E-3a-diol for linkage to NHS-activated agarose beads was completed using a Sonogashira palladium cross-coupling reaction between the C-17 alkynyl group and a FMOC protected 4-iodobenzylamine, as exemplified in Arterburn, J.B. et al. (2000), which is incorporated by reference herein, an in particular, the procedure for Sonogashira couplings described therein. The FMOC protecting group was removed following the coupling reaction by heating the reaction mixture in piperidine.

[612] The E-3a-diol-benzyl amine was coupled to the NHS-activated bead according to standard protocol described in van Sommeren, A.P.G. et al. (1993), which is incorporated by reference herein. For evaluation of non-specific binding of proteins onto the affinity sorbent, control agarose beads, which were subjected to the same treatments as the E- 3a-diol-agarose beads, were also prepared by omitting the E-3a-diol-amine and quenching the NHS-activated beads with ethanolamine. Other bead-bound steroid test compounds are prepared analogously by introducing an ethynyl group to the test compound by using a suitably protected steroid precursor having a =0 substituent.

Preferred ethynyl group introduction is to a steroid precursor having a C17 =0

substituent. Docket No. 354. PATENT

[613] SILAC Media Preparation and Cell Culture Conditions: The mouse leukemic monocyte macrophage cell line RAW 264.7 (ATCC CRL-2278) cells were grown for at least 6 cell divisions in media supplemented with 5% dialyzed FBS, in a humidified air atmosphere with 5% C0 ₂ . All standard SILAC media preparation and labeling steps were followed as previously described in Ong, S.-E. and Mann, M. (2006), which is incorporated by reference herein, and in particular the methodology for SILAC media preparation and stable isotopic labeling described therein. Briefly, a base media was divided into two portions and either "light" forms of L-arginine and L-lysine or "heavy" L-arginine- ¹³ C ₆ ¹⁵ N ₄ and L-lysine- ¹³ C ₆ ¹⁵ N ₂ were added to generate the 2 SILAC labeling RPMI media. Each growth medium contained the full complement of amino acids and were sterile filtered through a 0.22 μΜ filter (Millipore).

[614] Biochemical Purification with Small Molecule Affinity Matrices: Separate RAW 264.7 cell cultures, SILAC labeled with either light or heavy amino acids, were lysed in ice-chilled Tris buffer (50mM pH 8) that included protease inhibitors (complete tablets, Roche Applied Science, Indianapolis, IN). Lysates are sonicated intermittently (3 x 10 pulses) while chilled on ice. Light and heavy lysate protein concentrations were estimated using the DC Protein Assay (Bio-Rad, Hercules CA) and equalized by dilution with buffer. The lysate protein concentrations varied between 1 .7 and 2.2 mg/mL and affinity enrichments were performed in 1 -1 .4 mL lysate volumes in a 1 .5 mL microcentrifuge tube.

[615] General procedures for biochemical enrichment of SILAC-labeled proteins using affinity baits and quantitative proteomics are described in Ong, S.-E. (2010), and are incorporated by reference herein. In forward bead control (BC) experiments, 2 mg of heavy RAW 264.7 lysate was incubated with 100 μί of 50% (v:v) control-bead slurry while 2 mg of light lysate was incubated with 100 μί of 50% E-3a-diol affinity beads. In the "reverse" experiment, the lysates were swapped for each bead type. In a "forward" soluble competitor (SC) experiment, soluble E-3a-diol (competitive) or 17a-ethynyl- androst-5-ene-3p ,7β, 17p-triol (non-competitive) in DMSO was added to 2 mg of heavy RAW 264.7 lysate (an equal volume of DMSO was then added to 2 mg of light RAW 264.7 lysate as a control). In this case, 100 μί of 50% HE3286 affinity beads was added to both the light and heavy lysates. Affinity enrichment mixtures were incubated overnight (4 °C, 16 h) on an end-over-end rotator, the beads pelleted by bench top centrifugation at (3 min, 1000 x g) and the supernatant aspirated. In BC experiments, beads were combined at the first wash for subsequent washing steps. For SC experiments, each tube in a set was washed with 50mM Tris buffer (pH 8) at least twice to remove excess soluble small molecule competitor and then combined for the later washing steps. After the third wash, beads were pelleted (3 min, 1000 x g) and the final wash was aspirated (leaving -25 μί of Docket No. 354. PATENT buffer). SILAC protein affinity enriched pull-downs were reduced and alkylated on bead by re-suspending in 200μΙ_ 8M urea, then treated with 10 μί of 100 mM TCEP at 37 °C for 30 minutes. 20 μΙ_ of 100 mM iodoacetamide was then added and the solution held in the dark for 30 min. with slow turning. Ammonium bicarbonate solution (600 μΙ_ of 25 mM) was added to create a 2M urea solution and 16 μΙ_ of trypsin (20 μg diluted with 40 μΙ_ resuspension solution) was added. The vials were incubated (37 °C) overnight on a thermomixer and then enzymatic digestion was stopped by the addition of 100% formic acid solution (30 μΙ_). The vials were allowed to sit at room temperature for 5 min and then spun (3 min, 1000 x g) to pellet the beads. The supernatant was removed to a separate vial for LC/MS/MS analysis.

[616] LC-MS/MS Analysis: The LC-MS/MS analysis followed a standard protocol for the MuDPIT technique described in the referenced protocol. [Schieltz, D.M. and Washburn, M.P. Cold Spring Harb. Protoc. (2006), incorporated by reference herein]. Briefly, LC-MS data was obtained on a quaternary Agilent 1 10 series HPLC coupled to an LTQ ion trap mass spectrometer (ThermoElectron) equipped with a nano-LC electrospray ionization source. The LTQ was controlled by Xcalibur™ data system software (ThermoElectron). LC-MS mobile phase buffers were composed in water with 0.1 % formic acid with the following additional modifiers: A (5% ACN), B (80% ACN), C (500mM ammonium acetate, 5% ACN).

[617] Fused silica microcapillary columns (100 μιη i.d. x 365 μιη o.d.) were pulled to generate 5 μιη tips using a Model P-2000 C0 ₂ laser puller (Sutter Instrument). Biphasic columns were packed with 10 cm of 5 μηι Aqua C18 reverse phase resin (RP; Phemomenex) followed by 3 cm of Partisphere strong cation exchange resin (SCX; Whatman). Loading/desalting tips were prepared by packing 4 cm of RP resin into a 250 μιη silica microcapillary fitted with a 2 μιη inline microfilter (Upchurch Scientific). Column packing was performed using a high pressure loading device (600 psi helium). Columns and tips were equilibrated in buffer A shortly before use.

[618] The desalting tip was loaded with sample and connected to a biphasic column and equilibrated with buffer A for 10 minutes before connecting to the MS. Peptides were eluted in steps beginning with a salt wash protocol (increasing %C), followed by an ACN gradient. For each sample MS run, five salt-wash steps (0%, 25%, 50%, 80%, and 100% C) were used. The flow rate was set to approximately 0.25 L/min and the applied distal spray voltage to 2.5-2.7 kV. MS2 data was collected using one full scan (400-1800 MW) followed by 7 data dependent MS2 scans of the most abundant ions with dynamic exclusion enabled (repeat count = 1 ; exclusion list size = 300, exclusion duration = 60). Docket No. 354. PATENT

[619] Database Searching and Protein Identification: Tandem mass spectra were searched using the Sequest algorithm (3.0) against the mouse database (ipi.MOUSEv368.fasta) from the European Bioinformatics Institute. The mass window for peptides searched was given a tolerance of 3 Da. between the measured average mass and the calculated average mass and the b and y ions were included. All samples were searched with a static mod of +57 Da. for cys residues. DTASelect™ (v2.0.25) was used to render Sequest output files. For tryptic rendering, default parameters were used, along with constraints for tryptic ends and exclusion of protein subsets.

[620] Example 2. Gene Transcription Effects in LNCaP Cells from Contact with 17-E- 3cc-diol. An Example Suitable In Vitro Test System using mt-AR CaP cells for Determining Gene Expression Effects.

[621] The PCR arrays, cDNA synthesis reagent (Cat# C-03), and PCR Master Mix (Cat# PA-MI) were purchased from SuperArray Bioscience Corporation. PerfectPure™ RNA (Fisher Cat# 29003 19) was used to extract RNA from the cells. The cDNA was synthesized on iCycler™ (Biorad) and the arrays were processed on real time PCR machine (iCycler IQ, Biorad). The results were analyzed using the software provided online by SuperArray™.

[622] LNCaP clone FCG cells were grown in RPMI 1640 medium (ATCC Cat# 30-2001 ) containing 2 mM L-glutamine modified to contain: 10% FBS, 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/L glucose, 1.5 g/L sodium bicarbonate. The cells were transferred to poly D-lysine coated 96 well plates in RPMI 1640 medium phenol red free (Gibco Cat# 11 835-055) containing 2 mM L-Glutamine modified to contain: 10% Charcoal / Dextran- treated FBS, 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/L glucose, 1 .5 g/L sodium bicarbonate and contacted with 30 nM E-3a-diol and 30 nM DHT or 10 nM βΑΕϋ.

[623] After 24 hours, RNA from the cells was extracted with PerfectPure™ RNA kit (5 Prime, Inc., Gaithersburg, MD) following the manufacturer's protocol including a DNase step. cDNA synthesis was then performed using CO-3 reagent following the manufacturer's protocol. After proceeding with PCR array preparation following the SuperArray™ protocol the PCR plates were run on a real time PCR machine using the settings recommended by the SuperArray protocol for the equipment used.

[624] Cancer pathway profiling arrays (n = 7) were performed with LNCaP cells of varying passages with each array consisting of 84 pathway- or disease-focused genes. Genes involved in cell proliferation, cell death, cell adhesion, invasion, metastasis and signal transduction were probed in the first study using a human cancer pathway finder array (SuperArray Cat# APHS-033A). In further studies, arrays specifically designed to Docket No. 354. PATENT probe the expression of genes involved in apoptosis (SuperArray Cat# APHS-012A), or signal transduction (SuperArray Cat# APHS-014A) or drug resistance and metabolism (SuperArray Cat# PAHS-004A) were used.

[625] The arrays were performed on different passages of LNCaP cells at different times, and some variability was observed in the results. This may be due to passage number, and to unknown differences in culture conditions. CDK4 expression was decreased in 2 of 7 Cancer Pathway arrays (-2.93 and -2.35 fold). Caspase-9 expression was upregulated in 1 of 3 apoptosis arrays (5.31 fold). A few genes were present on different pathway arrays. For example, CDK2 expression was downregulated in 1 of 7 cancer pathway arrays (-2.71 fold) and 1 of 2 signal transduction arrays (-2.06 fold). CDKN1 A expression was downregulated in 1 of 7 cancer profiling arrays (-3.43 fold), 1 of 2 signal transduction arrays (-2.2 fold), and 2 of 4 drug resistance and metabolism arrays (-2.39 and -2.43 fold). Bcl-2 expression was decreased in 6 of 7 cancer pathway arrays (average -3.56 fold), in 2 of 3 apoptosis arrays (-8.56 and -2.6 fold), in 2 of 2 signal transduction arrays (-3.36 and -4.1 1 fold), and in 3 of 4 drug resistance and metabolism arrays (-3.16, -3.43 and -1 .97 fold). These and other genes that were up-regulated (increased expression) or down-regulated (decreased expression) by at least 50% compared to controls in at least 3 of 6 repeats are listed in Table 2.

626] Table 2. E-3a-diol Effect on Cancer Pathway Profiling Array

Docket No. 354. PATENT

[627] In summary, from these gene array experiments E-3a-diol modulates various androgen-regulated cell cycle and signal transduction genes including CDK2, CDKN1A, KLK2, IGFBP-3, TMEPAI, ODC1, GREB1 and AR. E-3a-diol affected the expression of genes involved in various phases of the cell cycle including Gi phase and Gi/S transition including CCNE1, CDK4 and CDKN1B as well as cell cycle checkpoint and cell cycle arrest (ATM & CHEK2). Among the genes involved in apoptosis and cell senescence, Bcl2 and CFLAR (Caspase 9) were the most notable. Bcl-2 expression was down- regulated in most samples treated with E-3a-diol. E-3a-diol also down-regulated ABCG2 and ABCC5, both members of the ABC transporter family involved in drug resistance to chemotherapy.

[628] Table 3. E-3a-diol Affect on Apoptosis Pathway Profiling Array Docket No. 354. PATENT

[629] Apoptosis profiling arrays (n = 2) were performed in LNCaP cells treated with 30 nM E-3a-diol for 24 hours. Genes that were regulated compared to controls are listed in Table 4. IGF1 R was the only gene that was up-regulated in the apoptosis array.

[630] Table 4: E-3a-diol Affect on Signal Transduction Pathway Finder Array

[631] Signal transduction pathway finder arrays (n = 2) were performed in LNCaP cells treated with 30 nM E-3a-diol for 24 hours. Genes that were up-regulated (increased expression, most of these genes respond to androgens), or down-regulated (decreased expression) by at least two-fold compared to controls are listed in Table 5.

[632] Table 5: E-3a-diol Affect on Drug Resistance and Metabolism Array Docket No. 354. PATENT

^* Failed determination-non physiological change

[633] Drug resistance and metabolism arrays were performed in LNCaP cells treated with 30 nM E-3a-diol for 24 hours. Most of these genes were down-regulated. Estrogen receptor beta was the only gene that was up-regulated compared to controls. The data are presented in Table 5.

[634] Example 3. Gene Transcription Effects from Contacting E-3cc-diol with a Xenograft Model using LuCaP Cells. An Example Suitable In Vivo Test System using wt-AR CaP cells for Determining Gene Expression Effects.

[635] Gene expression analysis was performed on the control LuCaP-35V and LuCaP- 35V + E-3a-diol groups described in Example 5. Tumor fragments (50-100 mg) were placed into 1 ml_ of STAT-60™ solution (Tel-Test, Inc, Friendswood, TX) and homogenized using Omni Tips (Omni International, Marietta, GA). The RNA extraction was carried out as recommended by the manufacturer's protocol.

[636] Analyses of the oligo arrays, used four pools of RNA from group 1 (control LuCaP- 35V) and four pools of RNA from group 2 (LuCaP-35V+E-3a-diol). Each pool contained Docket No. 354. PATENT an equal amount of RNA from three different tumors from the specific group. A reference standard RNA for use in two-color oligo arrays was prepared as described in Coleman, I.M. et al. (2006). Total RNA was amplified using the Ambion MessageAmp™ aRNA Kit (Ambion Inc, Austin, TX). Amplified amino-allyl aRNA from each pooled sample was labeled with Cy3 fluorescent dye (reference amino-allyl aRNA was labeled with Cy5) and hybridized to Agilent 44K whole human genome expression oligo microarray slides (Agilent Technologies, Inc., Santa Clara, CA).

[637] Fluorescence array images were collected using the Agilent DNA oligo array scanner G2565BA (Agilent Technologies, Inc.), and Agilent Feature Extraction software was used to grid, extract, and normalize the data. Spots of poor quality or average intensity levels <300 were removed from further analysis. The Statistical Analysis of Microarray (SAM) program (http://www-stat.stanford.edu/~tibs/SAM/) was used to analyze expression differences between treated and untreated specimens.

[638] Unpaired, two-sample t-tests were calculated for all probes passing filters and controlled for multiple testing by estimation of q-values using the false discovery rate (FDR) method. A false discovery rate (FDR) of q-values < 5% was considered significant and set the threshold for differential expression at >1 .6 fold change with treatment. At this level of significance, there were 355 unique genes up-regulated and 30 unique genes down-regulated in E-3a-diol-treated LuCaP-35V vs. control LuCaP-35V. Those results were then reduced to unique genes by eliminating all but the highest scoring probe for each gene. Gene set enrichment analysis of GO (http://www.geneontology.org/) categories was performed using the EASE (http://david.abcc.ncifcrf.gov/) software.

[639] In subsequent experiments, real-time PCR using the same pools of RNA were used as for the oligo array analysis and performed according to a previously described procedure [Coleman, I.M. et al. (2006)] confirmed the oligo array analysis results showing significant down-regulation of AR mRNA in the E-3a-diol -treated tumors and the up- and down-regulation of several AR-regulated genes (although not all of the genes examined resulted in statistically significant changes). In animals supplemented with βΑΕϋ, the effect of E-3a-diol on gene expression was diminished and in some instances, abolished.

[640] Immunohistochemical (IHC) analysis used 5-μιη sections of paraffin-embedded subcutaneous and intra-tibial tumors were used and was performed by standard procedure as described in Kiefer J. A. et al. (2004), using an anti-human AR mouse monoclonal antibody (1 :60 dilution, BioGenex, San Ramon, CA). For analysis, conducted according to the procedure described in Lai J.S. et al. (2004), a quasi-continuous score was created by multiplying each nuclear intensity level (1 for no stain, 2 for faint stain, 3 Docket No. 354. PATENT for intense stain) by the corresponding overall percentage of cells at that intensity for the entire tumor, and then summing the results. All evaluations were performed in a blinded fashion and statistics were performed using Student t-tests. IHC for PSA expression in intra-tibial tumors was done using an anti-human PSA rabbit polyclonal antibody (Dako, 3 μg/mL, Carpinteria, CA).

[641] IHC analysis showed significant decreases in the nuclear localization of AR after E-3a-diol treatment; the AR score of LuCaP-35V+ E-3a-diol was 248 ± 8 vs. 198 ± 28 of control (P=0.0048), while alteration of the AR score in AED-LuCaP-35V+ E-3a-diol vs. AED-LuCaP-35V did not reach significance (228 ± 36 vs. 208 ± 8, P=0.28).

642] Table 6. Changes in Androgen-Regulated Genes by E-3a-diol.

Docket No. 354. PATENT

Docket No. 354. PATENT

[643] The i-AR regulated genes of Table 6 are classified into the following categories. Metabolism: Fatty Acids (FADS1 , FABP5, AZGP1 , AMACR, LONPL, LIPG), Polyamines (LOX, SAT, ODC1 ) Purines (GDA, AMPD3); Mitochondrial and Ribosomal Proteins: RBM24, MRPL33, RPL21 , RPLP1 ; Stress, Apoptosis: IL1 R1 , NFKBIZ, SAA2, SAA3, GPX3, CASP10, NMES1 (C15orf48), SUM02, MICAL1 , RAB32, MT1 G; Adhesion, morphology: Rho signaling (CDC42EP2, PVRL3, POPD3, ARHGEF10, RHOBTB3 GRHL2)CDH26, CLDN8; Development: Wnt, Notch, Hox (PBXN02, PXDN, LGALS3, CTBP1 , LFNG) ARG2, RFPL1 ; Cell Cycle: S100A1 1 , PPM1 E, ZNF33A, TIMP2

[644] Several genes upregulated by E-3a-diol are involved in fatty acid metabolism. FA de novo synthesis of long chain FAs is normally associated with proliferation. However, when in G1 arrest, which E-3odiol induces, continued FA synthesis consumes ATP to no effect. FA beta-oxidation of short chain FAs in order to regenerate the consumed ATP becomes excessive resulting in ROS formation, which promotes apoptosis. That ROS production is aggravated by upregulation of glutathione peroxidase (GPX3) and down-regulation of SLC3A1 (cysteine transport) and adenylate kinase (AK5)

[645] Other genes upregulated by E-3a-diol are involved in polyamine synthesis, which is normally associated with proliferation (in conjunction with DNA synthesis). However, When in G1 arrest for which E-3a-diol induces, there is no DNA replication that accompanies polyamine synthesis. The resulting cellular stress from overproduction of polyamines, which is aggravated by nucleoside(tide) catabolism (GDA and AMPD3 upregulated) and aberrant polyamine transport (SLC3A1 down- regulated), results in apoptosis.

[646] Still other genes upregulated by E-3a-diol are involved in development. Hox- and Wnt-related gene expression is normally associated with immature cells and cancer cell proliferation. However, Wnt is also required for epithelial differentiation Docket No. 354. PATENT and Hox for functional differentiation in mature cells. Thus upregulation of those genes in the context of G1 arrest is believed to promote differentiation that leads to apoptosis. In addition, the HOX gene IRX5 is down-regulated and forced down- regulation in LNCaP is known to induces apoptosis. Finally, Downregulation of v- JUN and JAG1 will relieve inhibition to differentiation. Other genes known or implicated in G1 arrest and apoptosis subsequent to it include PPM1 E, TIMP2, CASP10, which are upregulated by E-3a-diol.

[647] Example 4. Anti-Proliferative Effect of E-3cc-diol in Xenograft Model using C4-2B cells. An Example Suitable In Vivo Test System using mt-AR CaP cells for Determining Anti-Proliferative Effects.

[648] C4-2B cells were directly injected into murine bones (intra-tibial) to determine the effect of E-3a-diol on bone CRPC tumors as described in Koreckij, T.D. et al. (2009). C4- 2B cells were used since their xenograft tumors exhibit an osteoblastic response when grown in the bone environment similar to that seen in patients with CaP bone metastases. These cells are a CRPC subline of LNCaP cells, which also harbors the T877A mt-AR. These cells were derived from a bone metastasis [Thalmann, G.N. et al. (2000)] and were maintained in vitro under standard tissue culture conditions.

[649] For the intra-tibial study using C4-2B cells, thirty male CB-17 SCID mice were castrated, and after two weeks of recovery intra-tibial injections, as described in Corey E. et al. (2002), were performed. Blood samples were drawn weekly for determination of serum PSA levels, which were used to evaluate tumor growth. Four weeks after tumor-cell injection when tumors were established in the bone, animals with PSA >0.6 ng/mL and <5 ng/mL were randomized into two study groups: Control C4-2B group receiving vehicle, n=9 and 2) and Treated C4-2B + E-3a-diol group receiving E-3a-diol (160 mg/kg in 200 μΙ_ i.p.), n=10. Animals were dosed daily, seven days a week for four weeks.

[650] Animals were sacrificed after four weeks of treatment or if otherwise compromised. Tibiae were excised, weighed, decalcified in EDTA and embedded in paraffin for analyses. Effects of tumor growth and E-3a-diol treatment on bone was examined using radiographs (Faxitron Specimen Radiography System, Model MX-20, Faxitron x-ray corporation, Wheeling, IL) and bone mineral density measurements (PIXImus Lunar densitometer, GE Healthcare, Waukesha, Wl), which were obtained prior to sacrifice.

[651] Results demonstrate that E-3a-diol treatment decreases weight of tumored tibiae in comparison to control tumored tibiae (0.056 ± 0.004 g vs. 0.068 ± 0.015 g; P=0.034). Hematoxylin and eosin (H&E) staining of the tumored tibiae showed osteoblastic reaction associated with growth of the C4-2B tumors in the bone and tumor foci between the newly Docket No. 354. PATENT formed woven bone in both E-3a-diol-treated and control tibiae. In contrast to Example 5, which uses LuCaP-35V cells, E-3a-diol treatment in this model resulted in decreases of serum PSA. When normalized to enrollment, E-3a-diol lowered serum PSA levels over time resulting in significantly lower sacrifice serum PSA levels of C4-2B + E-3a-diol vs. control C4-2B animals (6.01 ± 1 .18 ng/mL and 1 1 .6 ± 4.54 ng/mL, respectively; P=0.0076) (Fig. 2). Since the serum PSA levels were lower in the animals with C4-2B tumors treated with E-3a-diol vs. animals with control C4-2B tumors, IHC analysis of PSA expression was performed. The E-3a-diol-treated tumor exhibited stronger PSA immunoreactivity, despite the lower serum levels. Furthermore, no differences in BMD were detected between E- 3a-diol-treated vs. control C4-2B tibiae (0.05 ± 0.004 g/cm ² vs. 0.049 ± 0.004 g/cm ² ; P=0.65) was observed. Similarly no differences in BMD were detected between normal contralateral tibiae of E-3a-diol-treated and control animals.

[652] Example 5. Anti-Proliferative Effect of E-3cc-diol in Xenograft Model using LuCaP- 35V cells. An Example Suitable In Vivo Test System using wt-AR CaP cells for Determining Anti-Proliferative Effects.

[653] LuCaP-35V cells are CRPC subline of the LuCaP 35 CaP xenograft derived from the lymph node metastasis of a patient who had previously undergone an orchiectomy [Corey, E. et al. (2003)]. LuCaP-35V cells harbor wt-AR and are maintained by serial passage in castrated severe combined immunodeficient (SCID) male mice.

[654] In one subcutaneous study, the effect of E-3a-diol on LuCaP-35V proliferation in the presence of βΑΕϋ was examined. Castrated SCID mice were supplemented with βΑΕϋ pellets to closer mimic the human adrenal gland secreting βΑΕϋ because the adrenal gland in mice lacks the enzymes necessary for steroidogenesis [vanWeerden, W.M. et al. (1992)].

[655] For this study 50 male CB-17 SCID mice (Charles River Laboratories, Wilmington, MA) were castrated, and after a 2-week recovery period, half of the mice then received subcutaneous βΑΕϋ pellets (5 mg, 60-day time release; IRA, Sarasota, FL). The other half of animals received placebo pellets. LuCaP-35V was implanted subcutaneously (~ 20 mg tumor bits) 3 days after implantation of the pellets. Animals were randomized into the following study groups when tumor volume exceeded 100 mm ³ : 1 ) Control LuCaP-35V group receiving a placebo pellet + placebo treatment (vehicle), n = 12; 2) LuCaP-35V + E- 3a-diol, group receiving the placebo pellet + E-3a-diol, n = 12; 3) AEDLuCaP-35V, group receiving the AED pellet + vehicle, n = 12; and 4) AED-LuCaP-35V + E-3a-diol, group receiving the AED pellet + E-3a-diol, n = 1 1 . E-3a-diol was administered through Docket No. 354. PATENT intraperitoneal (i.p.) injection once daily, 5 d/wk for 4 weeks at dose of 160 mg/kg.

[656] Tumor volumes were measured twice weekly, and blood samples were drawn weekly for PSA determinations (IMx Total PSA Assay; Abbott Laboratories, Abbott Park, IL). Exponential growth equations were used for calculations of tumor doubling times. Animals were sacrificed after 4 weeks of treatment when tumors exceeded 1000 mm ³ or if otherwise compromised. Sacrifice PSA index was calculated by dividing the serum PSA levels by the tumor volume. At sacrifice, half of each tumor was processed for paraffin embedding and immunohistochemistry (IHC), and the other half was flash frozen for gene expression analysis and determinations of intratumoral androgen levels.

[657] Treatment with E-3a-diol significantly inhibited growth of LuCaP-35V in mice supplemented with pAED; doubling time of pAED-LuCaP-35V + E-3a-diol was 13.15 ± 2.96 days (mean ± SD) in comparison to 9.87 ± 1 .64 days of AED-LuCaP-35V (P = .007). Treatment with E-3a-diol also resulted in slight increases in serum PSA levels versus the control AED-LuCaP-35V animals, but these differences did not reach significance. Similarly, the PSA index at the end of the study was also higher in E-3a-diol-treated animals, and this difference did not reach significance (AED-LuCaP-35V, 0.09 ± 0.04 ng/ml per cubic millimeter; and AED-LuCaP-35V + E-3a-diol, 0.13 ± 0.09 ng/ml per cubic millimeter, P = .17). However, these elevations in PSA are in concordance with in vitro studies demonstrating increases in AR-mediated transcription by E-3a-diol treatment and are consistent with induced differentiation by E-3a-diol.

[658] In another subcutaneous study, the effect of E-3a-diol on LuCaP-35V proliferation in the absence of pAED was examined. These experimental conditions mimic the clinical scenario of patients treated with agents aimed at blocking adrenal synthesis of androgens (e.g., ketoconazole). In this setting, E-3a-diol significantly increased the tumor doubling times of LuCaP-35V (LuCaP-35V + E-3a-diol, 18.2 ± 6.28 days; untreated LuCaP-35V, 10.44 ± 1 .8 days; P < .0001 . E-3a-diol treatment resulted in significant increases in serum PSA levels in the treated animals versus control animals bearing LuCaP-35V tumors in the period of 1 to 3 weeks after treatment initiation (P < .0001 ). PSA levels were not significantly higher in the treated animals at the end of the study. It is believed this is due to the decreases in tumor volume. This explanation is supported by results showing a significantly higher PSA index in the LuCaP-35V + E-3a-diol animals than PSA index in the LuCaP-35V animals, 0.18 ± 0.07 versus 0.07 ± 0.02 ng/mL per cubic millimeter, respectively (P < .0001 ).

[659] Example 6. Anti- Proliferative Effect of E-3cc-diol on C4-2B cells. An Example Docket No. 354. PATENT

Suitable In Vitro Cell -Based Test System using Mt-AR CaP Cells for Reporter and Proliferation Assays.

[660] Experiments were performed in triplicate and repeated twice. Similar proliferation studies with LuCaP-35V are not possible since these cells do not proliferate in vitro.

[661] For the transiently transfected reporter assay, C4-2B cells were grown in RPMI 1640 (Invitrogen, Grand Island, NY) supplemented with 10% fetal bovine serum (Atlanta Biological, Atlanta, GA) under standard tissue culture conditions. Cells were transiently transfected with an androgen response element (ARE) reporter or with a 5.8-kb PSA luciferase plasmid using the Amaxa Nucleofector with solution V on program 27 as per the manufacturer's instructions (Amaxa Biosystems, Inc, Gaithersburg, MD). The hTK renilla- luciferase plasmid was transfected under the same conditions to allow for normalization of transfection efficiencies. Control cells were mock-transfected.

[662] After transfection, C4-2B cells were plated in a six-well plate (200,000 cells per well) in RPMI 1640 medium with 5% charcoal-stripped serum, and E-3a-diol was added at 0, 10, and 50 nM concentrations. Cells were incubated for 48 hours, and luciferase activity was detected with a Dual-Luciferase Reporter Assay™ (Promega, Madison, Wl) using a Tecan GENios Plus™ illuminometer (Phenix Research Products, Hayward, CA).

[663] For the proliferation assay C4-2B cells were plated in a six-well plate (200,000 cells per well) in RPMI 1640 medium with 5% charcoal-stripped serum, and E-3a-diol was added at 0, 10, and 50 nM concentrations. After three days viable cells were counted using the Trypan blue exclusion assay. Results show that E-3a-diol increases AR- mediated transcription using either reporter plasmid in C4-2B cells (expresses mutated AR) AR-mediated transcription was also increased in LuCaP-35V cells (express wild-type AR), but this activation did not reach statistical significance.

[664] In the reporter and proliferation studies, C4-2B cells, which express the same mutated AR gene as in LNCaP cells and are castration-resistant tumor cells, were used and demonstrate that E-3a-diol activates the ARE promoters in these mutated AR- expressing cells. However, in the intra-tibial studies (Example 4). Although there is ARE activation from agonist binding to mt-AR, E-3a-diol caused a significant decrease in the proliferation of C4-2B cells in a dose-dependent fashion (ANOVA; P < 0.0001 ). The AR score of E-3a-diol-treated C4-2B tumors (Example 4) was also lower vs. control C4-2B tumors (236 ± 15 vs. 182 ± 15; P <0.001 ), i.e., E-3a-diol treatment resulted in decreased AR mRNA and lowered levels of nuclear AR in the E-3a-diol treated tumors. Those results suggest that AR mediated transcription may not be occurring simply through direct Docket No. 354. PATENT interaction of E-3a-diol with the mt-AR. AR signaling not attributable to exogenously added androgens (e.g., added pAED) is sometimes referred to as "androgen- independent" AR signaling that is associated with kinase activation [for example, see Carey, A.-M. et al. (2007)].

[665] Example 7. Determining Binding and Transactivation Activity of Nuclear Hormone Receptors by Test Compound as Exemplified for E-3cc-diol.

[666] Evaluation of steroid nuclear receptor (NR) binding affinity was accomplished by a highly sensitive in vitro fluorescence polarization (FP)-based competition binding assay using fluorophore-conjugated high affinity NR-ligands (FP Nuclear Receptor Binding Assay System; Invitrogen). In this assay, a change in fluorescence resulting from binding between a NR and a high affinity proprietary fluorescent ligand (Fluormone™) is measured. Compounds that behave as NR competitive ligands with respect to the Fluormone/NR interaction will cause a concentration-dependent suppression in the extent of fluorescence polarization (FP) according to their relative binding affinity. Thus, the concentration of the test compound that results in half-maximal decrease of the polarization signal corresponds to the IC ₅₀ value, which is a measure of the relative affinity of the test compound for a given NR.

[667] Briefly, recombinant human NR GST-fusion proteins of recombinant origin were incubated in assay buffer in the presence of the appropriate Fluormone (final concentration 200-400 nM) and increasing concentrations of E-3a-diol or the indicated reference ligand. The resulting Fluormone-INR complexes are allowed to reach equilibrium binding for 2 hr at ambient temperature and the extent of FP is then determined in a Tecan Genios Pro™ (fluorimeter mode), using the Magellan 5™ software.

[668] The extent of binding competition was compared with high affinity specific reference ligands, i.e., dihydrotestosterone (DHT) for AR, l7p-estradiol (E2) for ERa or ERp, dexamethasone (DEX) for glucocorticoid and progesterone (PR) for progesterone receptors. The results are expressed in terms of IC ₅₀ competition values (in nM concentration units) as compared to E-3a-diol in each assay, Binding experiments using the AR system (1 ) involve the use of a recombinant GST-fusion protein comprising the ligand binding domain of the human AR (Cat. No. P3018), whereas for ERa and ERp, (4- 7) the human full-length GST-fusion versions of these proteins are used (Cat. No. P2614 & P2615). Evaluation of E-3a-diol on glucocorticoid (GR) and progesterone (PR) nuclear receptor binding was done using fluorescence polarization assays (Invitrogen Cat# P2816 and Cat# P2895). Docket No. 354. PATENT

[669] The GST-fusion proteins used in this study are described in Chang, C.Y. et al. (1999); Green, S. et al. (1986); Greene, G.L. et al. (1986); Mosselman, S. et al. (1996) ; Paech, K. et al. (1997) ; Kuiper, G.G.J. M. et al. (1 997); Kuiper, G.G.J. M. et al. (1996) ; Chang, C.S. et al. (1998).

[670] The nuclear receptor binding profile of E-3a-diol is reported in Table 7. DHT and E ₂ are included as controls for their respective receptors. E-3a-diol does not bind strongly to the wild type AR, or at all to PR, GR or ERp, and has a weak affinity for ERa.

[671 ] Table 7: Nuclear Receptor Binding Profile of E-3a-diol.

[672] *Results are expressed as IC ₅₀ values in nM, representing the statistical mean ± SEM. The numbers in parenthesis represent the number of independent experiments performed (n value). Abbreviations: AR - Androgen receptor; ERa - Estrogen receptor a; ER3 - Estrogen receptor β; GR - Glucocorticoid receptor; PR - Progesterone receptor; DHT - Dihydrotestosterone; E ₂ - 173-estradiol; ND - Not determined

[673] For transactivation experiments of the sex steroid receptors ERa, ERp, AR and GR the following cell lines were utilized. The human MDA-kb2 breast cancer cell line utilized for evaluation of AR and GR transactivation activity was initially developed by the U.S. EPA [Wilson, V. et al. (2002)]. This cell line is commercially available as ATCC Cat. No. CRL-2713 and stably expresses an androgen and glucocorticoid-responsive luciferase reporter for detection of AR or GR agonists and antagonists (both AR and GR receptors are expressed endogenously in this cell line). The T47D-kBluc human breast cancer cell line was also initially developed by the U.S. EPA (Ibid.) and stably expresses an estrogen-responsive luciferase reporter for detection of estrogen receptor agonists and antagonists (both ERa and ERp are expressed endogenously in this cell line). This cell line is commercially available as APCC Cat. No. 286 and was used in transactivation assays for ERa and ERp.

[674] Transient transfections were done utilizing human embryonic kidney 293 fibroblasts (ATCC Cat. No. CRL-1573). In ERp-HEK 293 experiments, cells were Docket No. 354. PATENT transiently co-transfected with a cDNA expression vector encoding the full-length human ΕΡιβ and an estrogen sensitive luciferase reporter gene (these cells exhibit negligible levels of endogenous ERP). Similarly, human mutant AR receptor (T877A) and AR sensitive luciferase reporter gene constructs were utilized for the MtAR-HEK293 experiments.

[675] For the T47D-KBIuc & MDA-kb2 in vitro cell-based assays, cells were contacted with increasing concentrations of E-3a-diol or reference ligands for 24 hours at 37°C in a humidified C0 ₂ incubator. For HEK293T in vitro cell-based assays, cells were contacted with increasing concentrations of E-3a-diol or reference ligands 24 hours post- transfection, and maintained in culture for an additional 24 hours. Total cellular lysates were then prepared using a luciferase cell culture lysis reagent (Promega Cat. No. E1531 ). The resulting level of expression of the reporter luciferase gene in the extracts was measured by the extent of luciferase enzymatic activity as relative light units (RLU) in a Pecan Genios Pro (illuminometer mode) when incubated with an appropriate luciferase substrate (luciferase assay system; Promega Cat. No. E1501 ). Results are expressed as EC50 values (in nM concentration units).

[676] The ability of E-3a-diol to transactivate the same panel of nuclear receptors is reported in Table 8. As shown, DHT and E-3a-diol transactivate the T877A mutant AR with similar potency. E-3a-diol exhibits weak activity for both ER isoforms. E ₂ is also active on the mutant AR, demonstrating the promiscuous nature of the mutant AR.

[677] Table 8: Transactivation of Nuclear Receptors by E-3a-diol.

Ligand MtAR-HEK293 MDA-kb2 ^a T47D-kBluc ^b ERp-HEK293 ^c

DHT 0.5 0.06 ± 0.03 (9) 1 12 408 ± 224 (6)

E ₂ 1 .1 987 ± 293 (5) 0.002 ± 0.002 (10) 0.04 ± 0.04 (17)

AED 144 + 171 (5) 2969 + 790 (7) 2.5 + 0.76 1 .7 +0.26 (7)

E-3a-diol 0.48 + 0.69 (4) 1 1 .4 + 9.3 (3) Inverted V ^d 164 (3)

[678] ^* Results are expressed as EC ₅₀ values in nM, representing the statistical mean ± SEM. The numbers in parenthesis represent the number of independent experiments performed (n value). No parentheses indicate n=1 . Ligand abbreviations are as given in Table 7. MtAR-HEK293 cells are HEK293 fibroblasts transiently co-transfected with an ARE/luciferase promoter/reporter construct and a cDNA expression vector encoding the full-length LNCaP AR. These cells exhibit virtually undetectable levels of endogenous sex steroid receptors. ^a MDA-kb2 cells are stably transfected with a promoter/reporter Docket No. 354. PATENT construct sensitive to sex steroid receptor stimulation (MMTV promoter) fused upstream of a luciferase reporter gene. These cells endogenously express both AR (androgen receptor) and GR (glucocorticoid receptor). The MDA-kb2 cells are engineered to detect activated AR stimulation of androgen response elements (AREs) that are normally present on androgen-responsive genes such as PSA. Using DHT as the cognate ligand for AR, the MBA-kb2 assay induced an 8- to 10-fold increase in luciferase at the optimum DHT concentration. ^b T47D-kBluc cells are stably transfected with a synthetic promoter/reporter construct sensitive to estrogenic stimulation, consisting of 3 copies of the estrogen response element (ERE) fused upstream of a luciferase reporter gene. These cells express endogenously both forms of the ER (estrogen receptor), namely ERa and ERp. ^c ERp-HEK293 cells are HEK293 fibroblasts transiently co-transfected with an ERE/luciferase promoter/reporter construct and a cDNA expression vector encoding the full-length human ERp. These cells exhibit virtually undetectable levels of endogenous sex steroid receptors. ^d Possibly an antagonist at higher concentrations and a partial agonist at low concentrations.

[679] Example 8. Proliferation Assay by Colorimetric Determination. Example for Determining Anti-Proliferative Effects of Test Compounds, Optionally in the Presence of βΑΕΰ, on Cancer Cells in a Suitable Test System.

[680] The MTT proliferation assay is based on the enzymatic reduction of the tetrazolium salt MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide] in living, metabolically active cells. The reaction is carried out in situ, and the reaction product, a purple-colored formazan soluble in dimethylsulfoxide, is measured colorimetrically. The assay is conducted generally according to the methodology in Romijn, J. C. et al. (2006) using the following procedure.

[681] Transformed or cancer cells, such as LNCaP cells (6,000 cells per well), are plated out into wells of a poly-D-lysine coated assay plate each containing 200 μΙ_ RPMI 1640 medium containing 2 mM L-glutamine modified to further contain 10% FBS, 10mM Hepes, 1 mM sodium pyruvate, 4.5 g/L glucose and 1 .5 g/L sodium bicarbonate. The assay plate is incubated overnight at 37°C, 5% C0 ₂ . Test compounds in 10 mM DMSO to be screened are diluted to 100 μΜ then to 10 μΜ in RPMI 1640 medium containing 2 mM L- Glutamine modified to further contain 10% Charcoal/Dextran treated FBS, 10mM Hepes, 1 mM sodium pyruvate, 4.5 g/L glucose and 1 .5 g/L sodium bicarbonate. Media in the wells of the assay plate are replaced with 200 μΙ_ of the 100 μΜ or 10 μΜ test compound solution. After incubation for 1 hour at 37°C, 5% C0 ₂ , 5 μΙ_ of a 410 nM androstenediol (βΑΕϋ) solution, which is prepared from dilution of a 10 mM stock solution in DMSO with Docket No. 354. PATENT the same media that is used to prepared the compound solution, is optionally added to each well for a final concentration of 10 nM βΑΕϋ per well, and otherwise is sham added. The plate is then incubated for 5 days at 37°C, 5% C0 ₂ . To each well is then added 1 10 μΙ of diluted MTT reagent prepared from 10.5 mL of the same media used to prepared the compound solutions at 37 °C followed by addition of 100μΙ of Detergent Reagent after incubation for 2 h at 37 °C, 5% C0 ₂ . After standing at RT in the dark the absorbance in each well at 570 nm using a reference wavelength of 690 nm is determined. Using a pre- established standard curve of cells plated out vs. absorbance reading the number of cells per well is determined from the absorbance average after background subtraction of each triplicate reading.

[682] Example 9. Proliferation Assay by Fluorometric Determination. Example for Determining Anti-Proliferative Effects of Test Compounds, Optionally in the Presence of βΑΕΰ, on Cancer Cells in a Suitable Test System.

[683] Fluorometric determination relies upon the CyQuant™ assay, which is based on dye fluorescence enhancement upon binding to cellular nucleic acids by addition of a buffer containing the CyQuant-GR dye to lysed cells. This assay does not rely on cellular metabolic activity.

[684] LNCaP-FGC, PC3 and DU145 cells are obtained from the American Type Tissue collection (ATTC, Manassas, Virginia). To identify test compounds with anti-proliferative activity for LNCaP, cells are seeded in poly-D-lysine coated 96 well plates at 3.75 x 10 ⁴ / mL in a total volume of 200 μί of RPMI/10% fetal bovine serum (FBS). After incubation overnight at 37 °C, 5% C0 _2, the media is removed and the cells are washed once with 150 μΙ-Jwell of phenol red-free RPMI containing 10% charcoal/dextran-treated FBS (CS-RPMI) and optionally containing 10 nM βΑΕϋ. The appropriate dilution of each test compound is added and the cells are incubated for five days at 37 °C, 5% C0 ₂ . For IC ₅₀ determinations of test compounds with anti-proliferative activity, a dilution sequence of test compound is analyzed and the data are modeled using a nonlinear regression, sigmoidal dose response, variable slope equation. The curve fit is constrained with a bottom of 0 and a top of 100. IC ₅₀ values are generated by GraphPad Prism 4. Cell proliferation is measured on day five using the CyQuant assay (Invitrogen, Carlsbad CA). The effect of test compound on cell growth of LNCaP, PC3 and DU145 cells are assayed by plating 3 x 10 ⁵ cells in 5% CS-RPMI with optional supplementation with 10 nM βΑΕϋ. After 4 days cells are manually counted with a hemocytometer.

[685] Example 10. Cell Cycle and Apoptosis Assay. Examples for Determining Effects on Cell Cycle and Apoptosis of a Test Compound in a Suitable Test System Comprising Docket No. 354. PATENT

Cancer Cells.

[686] DHT can cause a Gi arrest but not cell death of LNCaP cells at concentrations above 10 ^"10 M [Lee, C. et al. (1995)]. Based on the results of ARE studies (see Example 15), an experiment was conducted to determine if E-3a-diol had a similar cytostatic effect on LNCaP and to determine if, unlike DHT, exposure to E-3a-diol caused apoptosis in LNCaP cells.

[687] LP LNCaP cells (5 x 10 ⁵ ) were seeded in phenol-red free RPMI with 5% charcoal stripped serum (CSS) in 6-well plates and allowed to adhere overnight, then cultured in CSS supplemented with 10 nM AED with or without 50 nM E-3a-diol for four days. At the end of the incubation period, floating and adherent cells were harvested for cell cycle analysis. For cell cycle analysis, LP LNCaP cells were re-suspended in a 10 mg/mL solution of 4, 6-diamidino-2-phenylindole (DAPI) and 0.1 % NP-40 in a Tris-buffered saline solution (pH 7.0) and analyzed using an Influx cytometer (Cytopiea, Seattle, WA). Analyses were performed with MultiCycle software (Phoenix Flow Systems, San Diego, CA). Apoptosis was measured using the Annexin V-FITC Apoptosis Detection Kit (Calbiochem, La Jolla, CA) according to the instructions of the manufacturer to detect apoptosis. PI and Annexin positive cells were considered apoptotic. For obtaining cell growth results of Figure 1 [LNCaP (■), PC3 ( A) and DU145 ( T)] cells were incubated with E-3a-diol (HE3235) at 1 nM, 10 nM and 50 nM and assayed in duplicate for inhibition of cell growth relative to vehicle.

[688] As shown by Figure 2, contacting LP LNCaP cells with E-3a-diol resulted in G1 arrest accompanied by an increase in apoptosis. The fold change in the number of apoptotic cells after 4 days of incubation with 50 nM E-3a-diol relative to the vehicle control is shown on the Y-axis. These effects are reminiscent of other agents reported to block the proliferation of prostate cancer cells in vitro and in vivo, most notable the histone deacetylase inhibitors (HDACi). Using the LNCaP tumor model, sodium butyrate caused Gi cell cycle arrest, inhibited LNCaP cell growth and increased PSA gene expression [Gleave, M.E. et al. (1998); Kruh, J. (1982)]. Similar findings with respect to inhibition of cell proliferation and PSA secretion have been reported for vitamin D ₃ [Esquenet, M, et al. (1996)], calcitriol [Bauer, J.A. et al. (2003)], Activin A [Fujii, Y. et al. (2004)], and phenylacetate [Walls, R. et al. (1996)].

[689] These effects on apoptosis, cell cycle progression and PSA secretion observed for E-3a-diol are therefore consistent with this compound inducing catastrophic cancer cell differentiation. Docket No. 354. PATENT

[690] Example 1 1 . Prostate Specific Antigen Secretion by CaP Cells

[691] Prostate specific antigen is an androgen-responsive gene product frequently used as a reporter for AR activation, examples of reporter gene constructs with PSA as the reporter are provided by Langeler, E.G. et al. (1993); Lee, C. et al. (1995); Arnold, J.T. et al. (2007), all of which are incorporated by reference.

[692] An examination of E-3a-diol treated HP LNCaP cultures showed a dramatic increase in secreted PSA in the media. However, this rise in PSA was not accompanied by an increase in proliferation; rather the proliferation of HP LNCaP was inhibited. These findings, presented in Table 9, are similar to the observations in the preclinical LuCaP 35V in vivo tumor treatment study of Example 5 and in the in vitro LP LNCaP cells of Example 12.

[693] Table 9: Proliferation and PSA Secretion by HP LNCaP

[694] ^a Proliferation determined by cell count, input cell number = 5 x 10 ⁵ . ^b PSA determined by ELISA.

[695] Example 12. Androgen-mediated transcription assays. Example for Transfection of an Androgen-Inducible Reporter into Cancer Cells.

[696] To determine if E-3a-diol inhibited the activation of LNCaP mt-AR, LP LNCaP cells were transfected with an ARE-promoter reporter construct and incubated with E-3a-diol with or without DHT and βΑΕϋ present.

[697] An ARE-luciferase reporter construct (probasin promoter containing ARE) and a 5.8-kb PSA promoter-luc plasmid and an Amaxa Nucleofector device was used for transfection of LNCaP cells using solution V and program T-27 according to the manufacturer's directions (Amaxa Biosystems Inc., Gaithersburg, MD). LNCaP cells (1 x 10 ⁶ ) were transfected with ARE reporter plasmid or PSA-promoter reporter plasmid (1 μg). Docket No. 354. PATENT

The hTK renilla-luciferase plasmid (2 ng) was transfected under the same conditions to enable normalization of transfection efficiencies. Cells were seeded in 24-well plates in phenol-red free RPMI with 5% CSS to allow adherence overnight, and then DHT, E2 or PAED were added (1 nM or 10 nM final concentrations) with or without 50 nM E-3a-diol in DMSO. Cells were incubated for -48 hours, lysed and subjected to a dual-luciferase assay according to manufacturer's recommendations (Promega, Madison, Wl) using Tecan Genios plus illuminometer. The luciferase activity was measured and signal normalized to transfection efficiencies based on TK transfection.

[698] As can be seen in Figure 3, E-3a-diol did not inhibit the activation of the LNCaP mt-AR when the cells were stimulated with either DHT or βΑΕϋ. Instead, E-3a-diol activated mt-AR in this setting. Consistent with this finding, an increase in amount of PSA secretion per cell was observed after incubation with E-3a-diol. PSA rose from 8.5 ng/mL/10 ⁵ cells to 175 ng/mL/10 ⁵ cells by day 4 in LP LNCaP cells culture incubated with 50 nM E-3a-diol.

[699] LP LNCaP cells were treated with 1 or 10 nM DHT or βΑΕϋ in charcoal-stripped serum (CSS) and 50 nM E-3a-diol (HE3235). The fold change above background is shown on the Y-axis of Figure 5.

[700] Example 13. Anti-Proliferative Effect of E-3cc-diol in Xenograft Model with LP LNCaP cells. An Example Suitable In Vivo Test System for Determining Effects of Test Compound on Tumor Incidence and Progression.

[701] To test the effect of test compound dose on tumor incidence, as exemplified for E- 3a-diol [Trauger, R. et al. (2009)], a total of 48 castrated male SCID mice were implanted with LNCaP tumor cells as described below, and the mice were randomized into 4 groups of 12 animals each (vehicle control, 4 mg/mouse/day, 1 mg/mouse/day, and 0.4 mg/mouse/day). Administration of E-3a-diol (200 μί i.p.) or vehicle began 24 hours after tumor inoculation. All animals were dosed daily for 28 consecutive days.

[702] To test the effect of E-3a-diol on established LNCaP tumors, 36 castrated SCID mice received βΑΕϋ pellets and were implanted with LNCaP tumor cells and monitored as described below. Once the tumors reached 15 - 25 mm ³ , the mice were paired by tumor volume, and each mouse in a pair was assigned to the vehicle or 4 mg/mouse/day E-3a- diol group. Animals were dosed i.p. with 200 μί E-3a-diol or vehicle once a day for 21 days.

[703] Castrated SCID mice (six week-old) were obtained from Jackson Laboratories, Bar Harbor, Maine, USA. After four days of acclimation, the mice were implanted with Docket No. 354. PATENT

PAED pellets (5 mg, 60-day time-release, IRA, Sarasota, FL). Three days later, all mice were injected subcutaneously in the right flank with 100 μΙ_ of 7.5 x 10 ⁶ LNCaP tumor cells in phenol red-free RPMI mixed 1 :1 with Matrigel (BD, Franklin Lakes, NJ). Tumor volumes were measured weekly and calculated as a2 x b/2 with a being the width and b the length of the tumor in mm (reported as mm ³ ). E-3a-diol was prepared for injection by dilution in Captisol™ (CyDex, Lenexa, Kansas).

[704] The incidence of a measurable tumor was estimated as the relative frequency of mice having a measurable tumor at least once any time in the experiment. The significance of the difference in incidence from vehicle was tested by means of Fisher's exact test, adjusted for multiplicity of comparisons [Westfall, P. H. et al. (1999)]. Dose response in the proportion of mice with tumor was tested for significance via exact Cochran Armitage test. Time to first measurable tumor volume is analyzed via Kaplan Meier product limit estimates, with the exact log-rank test applied to test for the significance of the differences [Cantor, A. (1997)] Tumor volumes and tumor growth rates are analyzed non-parametrically via exact Wilcoxon-Mann-Whitney test. Resulting p- values are corrected for multiplicity of comparisons (Westfall, op. cit).

[705] Reduction of tumor volume was defined as a reduction in volume of at least 20% of the baseline volume, persisting to the end of the study. Fisher's exact test was applied to test for the significance of the difference in the incidence of such reduction relative to vehicle. Time to first measurable tumor volume was analyzed via Kaplan-Meier product limit estimates, with the exact log-rank test applied to test for the significance of the difference. To detect the difference between active and control group, Fisher's exact test and exact 95% CI for the difference were applied. A tumor of non-measurable volume was a tumor that, with the methodology at hand, measures 0 to the end of the study. The growth rate of a tumor was also analyzed via the mixed model.

[706] As seen from Figure 4, treatment with E-3a-diol resulted in a significant reduction in tumor incidence (compared to vehicle) in the two highest dose groups, (40mg/kg, p = 0.006, n = 1 1 ; 160mg/kg, p < 0.001 , n = 12), with decreases in tumor volume apparent in all three dose groups. The mean tumor volume in the animals that developed tumors were also significantly affected, 157 mm ³ (±153) (vehicle) vs. 0 (p<0.001 ), 4 (±7) (p<0.001 ), 34 (±41 ) (p=0.004) mm ³ in the E-3a-diol treated animals (descending dose). There was also a statistically significant delay in the time to a measurable tumor volume in all treated groups (p < 0.01 ) relative to the vehicle control group.

[707] As seen in Figure 5, vehicle treated animals showed a progressive increase in tumor volume over the course of the study. In contrast, the E-3a-diol (160 mg/Kg) Docket No. 354. PATENT treatment significantly blocked the growth of tumors (p < 0.001 ). Significant differences in tumor volumes between the control and treated groups were observed by the first week and were maintained through the course of this study (p < 0.001 ). A significantly greater percent of mice in the treatment group reduced their tumor volumes by 20% or more (p < 0.03) while no tumor reduction was seen in the vehicle group. Furthermore, tumors in two animals receiving E-3a-diol became non-measurable by day 15 of the study.

[708] In conclusion, when treatment was applied to mice with existing tumors, a reduction in tumor volume was observed by the end of the first week of treatment, and tumor growth continued to be depressed throughout the course of the study. In addition, two out of nine animals treated with E-3a-diol in this study had no detectable tumor volume by the end of the observation period. These results show that E-3a-diol had a cytoreductive effect on tumor volume.

[709] Example 14. Phosphotyrosine Profiling of HP LNCaP Cells contacted with E- 3cc-diol. Example for Phosphotyrosine Profiling of Cancer Cells in a Suitable Test System Contacted with Test Compound.

[710] Studies evaluating changes in signal transduction pathways were initiated using a SH2 profiling array (TranSignal™ Phosphotyrosine Profiling Array from Panomics Cat# MA3041 ). The profiling array kit includes all the reagents necessary to perform the assay. The membranes included in this kit are spotted in duplicate with 100 ng of 40 different SH2 domain proteins.

[711] LNCaP cells at passage 83 (p83) were plated at 8 x 10 ⁶ into each of four 10 cm tissue culture dishes in RPMI 1640 medium ATCC Cat# 30-2001 containing 2 mM L- glutamine modified to contain: 10% FBS, 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/L glucose, 1 .5 g/L sodium bicarbonate for 3 days until the cells reached 80% confluency.

[712] Membranes, one for each treatment, from the TranSignal™ Phosphotyrosine Profiling Array kit (Panomics, Inc., Redwood City, CA) were washed once in wash buffer for 30 minutes, blocked until the membranes appeared uniformly wetted (about 1 hour). The membranes were then rinsed once in wash buffer. The next day the confluent HP LNCaP cells were washed in pre-warmed PBS. Three dishes of cells was treated with 50 nM E-3a-diol diluted from 10 μί 10mM DMSO in pre-warmed RPMI 1640 medium phenol red free Gibco Cat# 1 1835-055 containing 2 mM L-Glutamine modified to contain: 10% Charcoal/Dextran-treated FBS, 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/L glucose, 1.5 g/L sodium bicarbonate and one plate was treated with DMSO diluted in this pre- warmed medium. The dishes were incubated at 37 °C. After 5, 15 or 60 minutes the Docket No. 354. PATENT medium was removed, the plates were placed on ice and the cells were washed once with ice-cold PBS. The control plate was incubated for 1 hour. The cells were lysed in 2x ice- cold lysis buffer, scraped off the plates and passed through 20-gauge syringe needle 20 times, keeping the cells on ice as much as possible. The lysed, syringed cells were transferred to microcentrifuge tubes and centrifuged at 16,000 x g for 5 minutes at 4 °C. Each lysate was removed avoiding the pellet of cell debris and diluted in 3 ml_ PBS.

[713] The diluted cell lysates (3 ml_) were added to the membranes (one lysate per membrane), covered tightly and incubated overnight at 4 °C on a shaker. After incubation, the membranes were washed 3 times for 10 minutes each in wash buffer and incubated for 2 hours with biotinylated anti-phosphotyrosine antibody. The membranes were washed 3 times for 10 minutes and incubated with streptavidin-HRP antibody for 1 hour and washed again as before.

[714] Fresh detection reagent was prepared, floated on top the drained membranes and covered with a square of clear plastic taking care to avoid air bubbles. The covered membranes were transferred to the Biorad Chemdoc XRS™ system to capture the chemiluminescence emitted. The data was analyzed with Quantity One version 4.5.0 (Biorad). Each dot of was circled and its intensity recorded using Quantity One software The background was automatically subtracted. The ratio of values of each E-3a-diol dot to its corresponding control dot was calculated and is presented in Figures 6 and 7.

[715] Table 10. Phosphotyrosine Modulation in SH2-proteins Effected by E-3a-diol ¹

_D . Factor Change From DMSO Control ²

T = 5 min I T=15 min

ABL1 -0.27 -0.30

BRDG1 -0.77 -0.13

BTK -0.54 -0.50

CRKL -0.52 -0.43

HCK -0.34 -0.14

LCK -0.15 -0.70

NCK2 -0.55 -0.24

P85A-D1 -0.47 -0.40

P85B-D1 -0.25 -0.27

PLCG1 -D2 +0.15 +1.33

RASGAP1 -D1 +0.18 +0.50

SHC2 +1.44 +2.13

YES -0.21 -0.23

[716] Modulation of pTyr by factor of 0.10 (absolute value) or greater at T= 5 min

2

and/or T=15 min. Average of two determinations. Positive modulation of overall phospho-tyrosine state given by positive factor change and negative modulation given by negative factor change. Docket No. 354. PATENT

[717] The results presented in Figures 6 and 7 and summarized in Table 10 indicate significant effects by E-3a-diol on overall tyrosine phosphorylation states for proteins that are recognized by members of the SH2 domain array.

[718] Example 15. Determination of the Phosphorylation State of PI3K in HP LNCaP Cells Contacted with E-3cc-diol. Example Determination of Phosphorylation State of PI3K in Cancer Cells of a Suitable Test System Contacted with Test Compound.

[719] PI3 kinase (PI3K) is the dominant growth-promoting pathway in LNCaP cells [Castoria, G. et al. (2004)]. Phospho-inositide-3-kinase (PI3K) regulatory subunits polypeptide 1 (p85a) and 2 (ρ85β) bear two SH2 domains each (p85a-D1 and p85a-D2 or ρ85β-ϋ1 and ρ85β-ϋ2). Phosphorylation of the p85 subunits renders the kinase functional and able to regulate diverse biological functions including cell proliferation, cell differentiation and survival.

[720] Analysis of PI3K in lysates of HP LNCaP cells that had been treated with E-3a-diol or DMSO with phospho-specific antibody from Example 13 revealed that the p85 subunits were less phosphorylated (i.e., phosphorylation status negatively modulated) after 15 minutes from contacting E-3a-diol with HP LNCaP cells (see Figure 1 1 ) while levels of total PI3K protein were not affected. Phosphorylation of Akt, a downstream substrate of PI3K, was also reduced after E-3a-diol treatment. To confirm this result, an in-cell ELISA that looked specifically at the levels of phospho-p85a was performed.

[721] HP LNCaP (originally from ATCC Medium) are plated at 20,000 cells/ well in poly- lysine-coated 96-well plates in 10% FBS overnight. The next day, the medium is replaced with RPMI 1640 medium Phenol Red Free Gibco Cat# 1 1835-055 containing 2 mM L- Glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/L glucose, 1 .5 g/L sodium bicarbonate. On the 3 ^rd day, the cells were treated with 10 μί aliquots of 10 mM DHT dissolved in DMSO to induce p85 phosphorylation. At the same time some wells were also treated with 10 μί aliquots of 10 mM E-3a-diol dissolved in DMSO. As shown in Figure 8, E-3a-diol rapidly reduced the levels of phosphorylation on p85a and ρ85β in the presence of DHT. After one hour from contact with E-3a-diol to the cells of this suitable in vitro test system, DHT induction of phosphorylation of the p85a subunits was inhibited down to control levels. Phosphorylation signals represented in Figure 8 were normalized to crystal violet staining in the same wells as a surrogate for cell number.

[722] Example 16. Effect of PI3K Inhibitor or MAP Kinase Inhibitor on Proliferation of LNCaP Cells Contacted with E-3cc-diol. Examples of Co-contacting a Test Compound and a Kinase Inhibitor with Cancer Cells of a Suitable Test System. Docket No. 354. PATENT

[723] Effects of PI3K inhibitor (LY295002) and Ras-Erk signaling inhibitor (PD098059) co-contacted on cancer cells with test compound is exemplified by E-3a-diol-contacted LNCaP cells.

[724] PD098059 is a selective noncompetitive inhibitor of the MAPK pathway. That inhibitor prevents the activation of MEK-1 by Raf or MEK kinase (a MAP3K) with an IC ₅₀ of 2-7 μΜ but does not inhibit Raf-activated MEK-1 . PD098059 inhibits Raf activation of MEK-2 less effectively with an IC ₅₀ of 50 μΜ. LY295002 is a selective inhibitor of PI3K, the dominant growth-promoting pathway of LNCaP cells.

[725] LNCaP cells were grown and treated in 10% FBS and transferred to 10 % CA- FBS for 24 hours before treatment. A dilution series (1/2 log ₁₀ ) of each compound was tested to assess the effect of each inhibitor by itself or in combination with E-3a-diol. Activity was determined by calculating the IC ₅₀ for inhibition of cell proliferation on day 3 of culture relative to cells incubated without any compounds by a commercial BrdU assay.

[726] The effects of E-3a-diol and the PI3K inhibitor (LY295002), MAP Kinase inhibitor (PD098059), and Ras inhibitor (farnesyl thiosalicylic acid - FTSA) with and without E-3a- diol (HE3235) on LNCaP Proliferation as determined by cell proliferation are shown in Table 1 1.

[727] Table 11. Effect on Proliferation of LNCaP Cells Co-Contacted with E-3a-diol and PI3K or Ras-Erk Signal Transduction Inhibitors.

[728] Although the Ras pathway is active in cells treated with E-3a-diol the Ras inhibitor did not inhibit the proliferation of untreated cells. The MAPK inhibitor was weakly effective at inhibiting proliferation of untreated cells. The PI3K inhibitor strongly inhibited LNCaP proliferation of untreated cells. The IC ₅₀ for the Raf-MEK-1 inhibitor PD098059 by itself was about 25 μΜ but in the presence of 30 nM E-3a-diol the inhibitor reversed the inhibiting effect of 30 nM E-3a-diol, with PD098059 returning proliferation to control levels Docket No. 354. PATENT in the presence of 30 nM E-3a-diol. This result indicates that the effect of E-3a-diol on inhibition of LNCaP cell proliferation is dependent on the MAP kinase pathway.

[729] The Raf-MEK-1 inhibitor PD098059 was weakly effective at inhibiting proliferation of LNCaP cells, while the PI3K inhibitor strongly inhibited LNCaP proliferation of cells stimulated only with 10% charcoal stripped serum (CSS). When combined with E-3a-diol, the anti-proliferative activity of LY295002 was increased. However, the presence of PD098059 reversed the inhibiting effect of E-3a-diol, with 30 μΜ PD098059 returning proliferation to control levels (Figure 9). This result supports the contention that E-3a-diol acts, at least in part, by activating the RAS/MEK/ERK pathway in LNCaP cells.

[730] E-3a-diol inhibited the proliferation of LNCaP cells stimulated only with 10% CSS (3,000,000 relative light units [RLU] from anti-BrdU chemiluminescence) to 1 ,000,000 RLU (0 μΜ inhibitor). LY295002 further increased inhibited proliferation, whereas PD098059 decreased the inhibition by E-3oc-diol.

[731] Example 17. Determination of Erk-1/2 Phosphorylation State in HP LNCaP Cells After Contact with E-3cc-diol. Example Determination of Protein Phosphorylation State in Cancer Cells of a Suitable Test System After Contacting with Test Compound.

[732] In general, LNCaP cells are grown and treated in Gibco Cat# 1 1835-055 containing 2 mM L-Glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/L glucose, 1 .5 g/L sodium bicarbonate and transferred to Charcoal stripped media modified to contain 10% Charcoal/Dextran FBS for 24 hours before treatment. The cells are contacted with test compound for 10 - 120 minutes depending on the experiment. Proteins were quantified, separated by electrophoresis on 4 - 12% acrylamide gels, transferred to PVDF membrane and probed with specific phospho-antibodies.

[733] LNCaP cells were grown and treated in 10% FBS and transferred to 10 % CA- FBS for 24 hours before treatment. The cells were treated with test compounds as indicated for 10 - 120 minutes depending on the experiment. The proteins were quantified, separated by electrophoresis on 4 - 12% acrylamide gels, transferred to PVDF membrane and probed with specific phospho-antibodies.

[734] E-3a-diol induced phosphorylation of Erk-1/2 within 10 minutes of activation. The blot of Figure 10 shows E-3a-diol, but not DMSO, induces phosphorylation of ERK1 (p44) but not ERK2 (p42) at 10 minutes, and that the ERK phosphorylation inhibitor PD 098059 blocks this phosphorylation (confirming that the bands seen in the presence of E-3a-diol alone is Erk). ERK2 (p42) is the lower band that is seen at 120 minutes with DMSO at 120 minutes. The time course of pERK1 formation relative to that of pERK2 is consistent Docket No. 354. PATENT with preferred nongenomic phospho-activation of ERK1 over ERK2 mediated through GPCR activation.

[735] Example 18. Determination of Phosphorylation States of Phospho-Proteins in HP LNCaP Cells Contacted with E-3cc-diol. Example Determination of Protein Phosphorylation State After Contacting Cancer Cells in a Suitable Test System with Test Compound.

[736] The phosphorylation states of signal transduction proteins in LNCaP cells were determined 15 minute after contacting with E-3a-diol and compared to DMSO-treated cells by signal transduction protein profiling using an antibody microarray.

[737] HP LNCaP cells (passage 91 and 92) were seeded at 3x 10 ⁶ cells per 10 cm plates in 10% FBS. Two plates were used for each passage and each treatment (8 plates in total). The next day the medium was replaced with 10% charcoal-adsorbed FBS for 24 hr. The cells were treated for 15 minutes with 50 nM E-3a-diol, washed once with ice-cold PBS, trypsinized, re-suspended in an equal volume of FBS, centrifuged at 500 x g for 2 minutes at 4 °C in a swinging bucket bench-top centrifuge. As much medium from the cell pellet as possible was removed without disrupting cells, which were then washed twice with ice-cold PBS. PBS was removed and the cell pellets were quick-frozen in a dry ice- ethanol bath for subsequent analysis.

[738] An adequate amount of ice-cold lysis buffer was added to the cell pellets followed by sonication four times for 10 seconds each time with 10-15 second intervals on ice to rupture the cells and to shear nuclear DNA. The resulting homogenate was centrifuged at 90,000 x g for 30 min at 4°C. The supernatant fraction obtained was transferred to a 1 .5- ml_ microcentrifuge tube. The samples were assayed for protein concentration using the standard protocol of Bradford [Bradford, M.M. (1976)] using bovine serum albumin as the protein standard.

[739] Lysates were then screened with 377 pan-specific and 273 phospho-specific antibodies on the Kinex™ antibody microarray (Kinexus Bioinformatics, Vancouver, CA). As with other antibody microarrays, non-denatured proteins analyzed by the this array provide increased opportunity for false positives and false negatives due to antibody cross-reactivity, protein-protein interaction, and blocked epitopes in protein complexes. From the manufacturer's internal studies with cells from different species, between 30 to 45% of the protein changes detected on the Kinex™ antibody microarray were reproduced by immunoblotting.

[740] From the results of the array, 17 proteins were selected for western blot Docket No. 354. PATENT confirmation. Six of the proteins were either not quantifiable or not detected on the westerns. Phosphorylation of Erk-1 , MEK-1 , MEK-2 and CDK-1/2 are clearly induced by E-3a-diol (Figure 1 1 ). There was no discernable or slight effect on the phosphorylation state of Erk-2. Phosphorylation of PEA-15 was slightly increased by E-3a-diol. Phosphorylated PEA-15 blocks Ras-mediated inhibition of integrin activation and modulates the Erk-MAP kinase cascade. Phosphorylation of PEA-15 at both serine 104 (a Protein Kinase C site) and serine 1 16 (a substrate of CaMKII and Akt) is required for PED/PEA-15 function. PEA-15 inhibits tumor cell invasion by inhibiting pERK translocation to the nucleus.

[741] Example 19. Combined Effects of E-3 -diol and Docetaxel on C4-2B Intratibial Tumors in Castrated SCID Beige Mice. Example Determination of Test Compound Effects in a Suitable In Vivo Test System Co-Contacted with a Chemotherapeutic Compound.

[742] Test compound effect in combination with a cancer chemotherapeutic compound, as exemplified for E-3a-diol with the tubulin disrupting agent docetaxel, is determined using the tibia tumor model in the following manner.

[743] Sixty-five SCID Beige male mice were castrated at 4 weeks of age and allowed to recover for 2 weeks before receiving injections of 2 x 10 ⁵ C4-2B cells (20 μΙ_) in the right tibia (see Example 4). After tumors were established in the bone, the mice were randomized into four treatment groups: Group 1 , vehicle; Group 2, E-3a-diol alone; Group 3, docetaxel alone; Group 4, E-3a-diol plus docetaxel (Table 12). The mice were treated for 4 weeks. Body weights and serum PSA were measured weekly. After four weeks, the animals were sacrificed. The combined weight of the prostate gland and seminal vesicles was recorded for each animal, and the right and left tibia in each animal were examined by X-ray and analyzed for mineral bone density and weight.

[744] One-tenth of a milliliter of 20 mg/mL solution E-3a-diol was administered by intraperitoneal injection (i.p.) once per day, five days per week to groups 2 and 4 (100 mg/kg, 5 mL/kg for a 20 gram mouse). A docetaxel stock solution was prepared by dissolving 20 mg in 0.5 ml_ of polysorbate 80, and diluting with 1 .8 ml_ of 13% ethanol in water. The stock solution was diluted to 0.5 mg/mL with saline immediately prior to use. Four-tenths of a milliliter of this diluted solution was administered by i.p. injection on Day 1 and Day 14 to groups 3 and 4 (10 mg/kg, 20 mL/kg for a 20 gram mouse). Group 1 received 30% β-cyclodextrin sulfobutyl ether (Sb-p-CD) vehicle and saline vehicle to match E-3a-diol and docetaxel schedules. Group 2 received saline vehicle in addition to E-3a-diol. Group 3 received Sb-p-CD vehicle in addition to docetaxel. Docket No. 354. PATENT

[745] Table 12. Study Design

^* based on 20 gram body weight

[746] The mean values for tibia bone mineral density (BMD) are shown in Table 13. The mean tumor bearing tibia BMD (the right tibia) was significantly larger in E-3a-diol + docetaxel treated mice compared to E-3a-diol alone, and there was a statistical trend toward a difference between the E-3a-diol + docetaxel and vehicle groups (Tables 14).

[747] Table 13. Mean Tibia Bone Mineral Density (right tibia is tumor bearing)

[748] Table 14. T Test Results for Tibia BMD

[749] Mean tibia and seminal vesicle (SV) plus prostate gland weights are shown in Table 15. There was a statistical trend to reduce the mean right tibia weight in E-3a-diol + docetaxel treated animals compared to vehicle (Table 16). Note that in groups 1 , 3, and 4, there was no correlation between tibia weight and serum PSA (Pearson correlation p = 0.5563, 0.0051 , 0.5747, and 0.4554 for groups 1 -4 respectively). There were no statistically significant differences in left tibia or seminal vesicle plus prostate weights Docket No. 354. PATENT between any groups. Prostate plus seminal vesicle weights did not significantly correlate to serum PSA (Pearson correlation p ranged from 0.23-0.97 for the 4 groups). The significance of differences between means and Pearson correlation was calculated with MS Excel or GraphPad Prism 4.0.

[750] Table 15. Mean Tibia and Seminal Vesicle Plus Prostate Weights (grams)

[751] Table 16. T Test Results Tibia and Seminal Vesicle Plus Prostate Weights

[752] Benefit of combining E-3a-diol therapy with other treatments was realized in this study. Using serum PSA as a surrogate for tumor burden, E-3a-diol plus docetaxel was the only significantly effective treatment in this study. The benefit of the combination therapy was also indicated by a positive effect on bone mineral density. The combination of treatments produced a significantly greater reduction in serum PSA without an apparent increase in toxicity (using clinical observations and body weights to assess toxicity), noting that the morbidity observed in a prior experiments in the LuCaP 35V and C4-2B tumor models that used a higher dose of E-3a-diol as a monotherapy (160 mg/kg) in Examples 4 and 5 is reported as absent in the current study.

[753] The results of the current experiment are unexpected particularly in view of the potential for drug holiday, since E-3a-diol has a half-life in mice of only about 2 hours, but was only administered once daily, five days per week in this study. The effectiveness of E-3a-diol treatment appeared to decrease in week 4. The reason for this decrease is not apparent. The week 4 blood collection for serum PSA measurement occurred on a Monday instead of a Friday, thus following two days of drug holiday. Docket No. 354. PATENT

[754] Example 20. Determination of Anti-Proliferative Activity of a Test Compound in a Carcinogen-Induced Breast Tumor Model and Effect of Co-Contact with a Chemotherapeutic Compound

[755] The efficacy of a test compound, alone and in combination with a cancer chemotherapeutic compound, as exemplified for E-3a-diol with or without the tubulin disrupting agent docetaxel, was compared to an estrogen receptor antagonist (tamoxifen), an aromatase inhibitor (anastrozole) and docetaxel in the carcinogen- induced MNU model [Ahlem, C. et al. (201 1 )].

[756] The carcinogen N-methyl-N-nitrosourea (MNU) induces hormone-dependent breast tumors in rats. This model has previously been used to develop therapies for the use of tamoxifen in women with breast cancer [Bentrem, D.J and Jordan, V.C. (1999)], and is considered to be appropriate for studies of novel compounds potentially useful for the treatment of breast cancer. Substantial evidence suggests that this rodent model system mimics human breast cancer, since the initiation of cancer occurs primarily at the same site in both humans and rats, the tumor cells express estrogen and progesterone receptors and tumor development is dependent on the reproductive history, diet and hormonal milieu [Gusterson, B.A. and Williams, J.C. (1981 )]. Thus, the model provides an opportunity to examine cause-and-effect relationships of the in situ environment fully impacted by systemic factors [Medina, D. and Thompson, H. (2000)].

[757] Materials and Methods: Docetaxel (Taxotere™, National Drug Code 0075-8001 - 20, Sanofi-Aventis U.S. LLC) was serially diluted with 13% aqueous ethanol and 0.9% saline according to the manufacture's instructions to yield a 0.74 mg/mL solution, which was used immediately after preparation. Anastrozole (AK Scientific, Inc, Mountain View, CA) was dissolved in 30% aqueous cyclodextrin to yield a 10 mg/mL solution. Tamoxifen (free base, Tocris Bioscience, Ellisville, MO) was dissolved in olive oil to yield a 1 .25 mg/mL solution. Solutions of E-3a-diol in 30% Sb-p-CD were prepared at either 20 or 33 mg/mL (w/v).

[758] The significance of differences between means or paired values were calculated using Student's t-test. Tumor volumes censored for death used the last observation carried forward (LOCF) for the purpose of data analysis. The significance of treatment effects on animal survival was determined by Fisher's exact test using SAS™ software (Cary, NC).

[759] Virgin Lewis rats were purchased from Harlan Sprague Dawley (Indianapolis, IN and San Diego, CA). The rats were housed in a temperature-controlled room with a 12- hour light and dark schedule, and fed a standard lab diet with access to food and water Docket No. 354. PATENT ad libitum. At seven weeks of age, all rats were treated with a single intra-peritoneal (i.p.) injection of 50 mg/kg of N-methyl-N-nitrosourea (Sigma, St. Louis, MO) as previously described [McCormick, D.L. et al. (1981 ); Rajkumar, L. et al. (2001 )].

[760] When the rats had developed palpable tumors (approximately 5 mm x 5 mm, 90 days after MNU) they were divided into seven treatment groups of thirteen animals each: 1 ) β-cyclodextrin, 2) 6.6 mg E-3a-diol (high E-3a-diol), 3) 4 mg E-3a-diol (low E-3a-diol), 4) high E-3a-diol + docetaxel, 5) docetaxel, 6) anastrozole, and 7) tamoxifen. E-3a-diol was administered daily by i.p. injection (0.200 mL, 4 or 6.6 mg/day, nominally 16 or 26.4 mg/kg) for four weeks. Two milliliters of diluted docetaxel in saline (1 .5 mg, 6 mg/kg, 8.1 mL/kg) were administered by i.p. injection once weekly for four weeks. Anastrozole was administered daily by i.p. injection (2.5 mg/day; 1 mL/kg) for four weeks. Tamoxifen was administered by subcutaneous injection (SC) once weekly (0.25 mg, 1 mg/kg, 0.8 mL/kg) for four weeks. Treatment doses per body mass (mg/kg) assume a 250 gram body weight at initiation of therapy on Day 101 . Doses were not adjusted for potential weight changes during the course of the experiment. Additional groups of animals were treated for two weeks with either vehicle or 6.6 mg E-3a-diol (N=10/group) for purposes of the evaluation of histopathology, immunohistochemical staining, and gene expression (realtime polymerase chain reaction).

[761 ] The dose of an antineoplastic agent is generally selected to approach the limits of toxic exposure. The doses of the comparator monotherapies used in this experiment were chosen based on doses found in the literature, and did not elicit observable toxicity during the course of this study. The doses of the antineoplastic agent E-3a-diol were selected after evaluating the compound in a dose-ranging pilot study in female Lewis rats. The high dose of E-3a-diol used in this study was approximately 2-fold lower than the multi-dose toxic threshold found in 28-day toxicology studies. The combination of E-3a- diol with docetaxel did not result in an observable toxicity.

[762] Animals were palpated once every week beginning one month after carcinogen exposure until the end of the experiment to monitor mammary cancer development. Tumor dimensions of length (r^ and width (r ₂ ) were measured with a vernier caliper, and tumor volumes were estimated with the formula: V = (4π/3) * x? * r ₂ (mm ³ ) [DeSombre, E. R. and Arbogast, L.Y. (1974)].

[763] The histopathology of paraffin sections was examined from one subset of animals (treated for 2 weeks) to confirm the carcinomatous nature of the palpable cancers. As the study progressed, animals that developed large tumor burdens (generally about 4 grams) were humanely sacrificed. Docket No. 354. PATENT

[764] Results: All animals had least one palpable tumor at initiation of therapy. On the first day of treatment (Day 101 ), the mean group total tumor volume for all groups was 4485 ± 717 mm ³ (range 4080 mm ³ [anastrozole group] to 6004 mm ³ [high E-3a-diol + docetaxel combination]). When treatment began, large tumors (>339 mm ³ ) were present in all groups. Although statistically significant, the variance in tumor volumes between groups at the start of treatment represents less than four days of growth, i.e., the smallest tumor volume on Day 101 [anastrozole group] was larger than the largest volume in any other group four days earlier (Figure 12).

[765] Tumors in vehicle-treated animals grew rapidly, with animals sacrificed for humane reasons in this group beginning on Day 139, with the last animal euthanized on Day 153. No animals treated with the combination of E-3a-diol plus docetaxel were sacrificed because of tumor burden (p = 0.0149 vs. docetaxel alone); one animal was sacrificed in each of the groups treated with high or low E-3a-diol monotherapy or tamoxifen (p = 0.073 vs. docetaxel alone); six were sacrificed in the docetaxel group and seven in the anastrozole group; and all animals were sacrificed in the vehicle group.

[766] Treatment with high E-3a-diol (HE3235) alone had a rapid and potent anti-tumor effect as indicated by a steep decline in the tumor volume after initiation of treatment (Figure 12). The cytoreductive activity of all active treatment groups was similar for the first two weeks of therapy, but the tumor ablative activity of low E-3a-diol, docetaxel, and anastrozole waned during the second half of the treatment period. In addition, tumor volume increased substantially in the docetaxel and anastrozole groups during the observation period after cessation of treatment. None of these three treatments showed statistically significant activity at the end of the treatment period when compared to their respective tumor volumes at baseline (p » 0.1 ). The failure of the low E-3a-diol treatment group to significantly ablate the mean tumor volume was due to substantial growth of one tumor in one animal, whereas with docetaxel it was the contribution from one tumor in each of six animals, and with anastrozole, one tumor in each of three animals. In all three instances, treatment appeared to be more effective in reducing or eliminating small tumors, while larger tumors were generally more resistant.

[767] In contrast, monotherapy with either high E-3a-diol (p = 0.01 1 ) or tamoxifen (p = 0.0042) aggressively ablated tumor volume (Figure 13) through the end of the treatment period, with a modest volume increase during the observation period. The combination of high E-3a-diol and docetaxel was more effective at the end of treatment than either alone (p = 0.01 13 vs. high E-3a-diol and p = 0.0390 vs. docetaxel), and prevented tumor growth through the last day of observation (Day 195). The mean tumor volume in the Docket No. 354. PATENT combination therapy group was not significantly different than the tamoxifen group at the end of treatment (p = 0.3451 ) or at the end of the observation period (p = 0.1383).

[768] The relative effectiveness of each treatment was also scored with the incidence of tumors in each group at the end of treatment and at the end of the study, compared to baseline (Figure 12). The number of tumors increased dramatically (from 16 to 56, p = 0.0001 ) in the vehicle treated group during the dosing period (Day 101 to 128), and as expected, decreased in response to all active treatments (p < 0.05). The numbers of tumors in the monotherapy groups compared to the high E-3a-diol and docetaxel combination were not significantly higher on Day 131 , except for the docetaxel group (p < 0.0089). However, after cessation of dosing, the incidence of tumors increased sharply in the docetaxel (12 to 27, p = 0.0051 ) and anastrozole (9 to 23, p = 0.0051 ) groups, and increased slightly in the tamoxifen group (2 to 10, p = 0.0362), while the E-3a-diol tumor incidence continued to decline through the observation period when treatment was given as either a monotherapy or in combination with docetaxel, although the decline during the observation period was not statistically significant when compared to the end of treatment.

[769] The incidence of disease free animals was comparable between the E-3a-diol and tamoxifen groups at the end of treatment. Eight tamoxifen rats were disease free, compared to six and four in the low and high E-3a-diol groups respectively, and nine in the high E-3a-diol and docetaxel combination group (Figure 14). At the end of the observation period, tumor incidence increased by two in the tamoxifen group (6 of 12 surviving animals were disease free), whereas in contrast, the tumor incidence decreased by four in the low E-3a-diol group (8 of 12), two in the high E-3a-diol group (8 of 12), and 2 in the high E-3a-diol-docetaxel combination group (1 1 of 13, p = 0.0405 vs. tamoxifen; not significant vs. high or low E-3a-diol monotherapy).

[770] For the results presented in Figure 12, seven-week old female Lewis rats were treated with a single i.p. injection of 50 mg/kg MNU. Tumors developed for 90 days, prior to treatment (n=13) for 28 days with: 1 ) cyclodextrin vehicle daily (filled square), 2) 6.6 mg E-3a-diol (HE3235) daily (open inverted triangle), 3) 4 mg E-3a-diol daily (filled inverted triangle), 4) 6.6 mg E-3a-diol daily + 1 .5 mg docetaxel weekly (open diamond), 5) 1 .5 mg docetaxel weekly (open triangle), 6) 2.5 mg anastrozole daily (not shown), and 7) 0.25 mg tamoxifen weekly (open circle). E-3a-diol in combination with docetaxel was more effective than comparator monotherapies at decreasing the mean tumor volume per animal. In Figure 12 split and expanded Y-axis are displayed. Anastrozole results were similar to docetaxel, but not plotted to improve clarity. The mean tumor volume for Docket No. 354. PATENT vehicle on day 101 was 0.41 cm ³ , which was not plotted to improve clarity. Tax = docetaxel, Tarn = tamoxifen, Tx = treatment period.

[771] The results of Figures 12-14 shows that E-3a-diol as a monotherapy or in combination with docetaxel was more effective at decreasing the number of tumors and rendering animals disease free, than comparator therapies. Figure 14: The effect of treatment on the average number of tumors per animal. ^* p start vs. end of treatment = 0.0008 (6.6 mg E-3a-diol), 0.0007 (4 mg E-3a-diol), and <0.0001 (E-3a-diol + Tax); ^§ p end of treatment vs. end of study = 0.0003 (Tax), 0.0016 (Ana), <0.0001 (Tarn). Figure 15: The percentage of rats in each group without palpable tumors. (There were no disease free animals in the vehicle group, not plotted.). ^** p vs. E-3a-diol + Tax = <0.0001 , ^** p vs. E-3a-diol + Tax = 0.0405. Tax = docetaxel, Ana = anastrozole, Tarn = tamoxifen.

[772] Immunohistochemistry: The effects of E-3a-diol (HE3235) on tumor tissue were examined by in satellite groups of animals treated with either vehicle or high E-3a-diol for 2 weeks. As shown in Figure 15, E-3a-diol increased the frequency of PARP stained cells two-fold (2,590 of 3,123 [82.9%] treated cells scored positive versus 1 ,212 of 3,079 [39.4%] vehicle treated cells scored positive, p < 0.0001 ), and decreased the frequency of ERa staining approximately 4-fold (2,349 of 3,051 [77.0%] vehicle treated cells scored positive versus 603 of 3,058 [19.7%] of E-3a-diol treated cells scored positive, p < 0.0001 ). Increases in frequency of cells staining positive for the apoptotic maker, PARP, and decreased frequency of ERa is associated with tumor survival.

[773] The expression of genes associated with cell proliferation, apoptosis and metastatic potential were consistent with the immunohistochemistry results (Figure 17). Proapoptotic genes were upregulated: Casp3 (9-fold), Casp8 (1 1 -fold), and Casp9 (5- fold), and p53 (15-fold), Bad (13-fold), Bax (10-fold), and ERp (4-fold), while genes associated with malignancy, metastasis, and escape from treatment were down- regulated: AR (25-fold), ERa (3-fold), and VEGF (4-fold). The expression of the autocrine growth factor, amphiregulin, and the anti-apoptotic protein, Bcl-2 were also decreased approximately 4-fold.

[774] E-3a-diol greatly decreased tumoral ERa and AR expression in the MNU model, and also decreased AR expression in LuCaP35V prostate cancer xenografts in mice (Example 5). In human breast cancer, decreased ERa and AR expression are associated with improved prognosis and reduced escape from therapy [Smollich, M. et al. (2009)]. Docket No. 354. PATENT

[775] For the results of Figures 15 and 16, seven-week old female Lewis rats were treated with a single i.p. injection of 50 mg/kg MNU. Tumors developed for 90 days, prior to treatment (n=10) for 14 days with cyclodextrin vehicle daily, or 6.6 mg E-3a-diol (HE3235) daily. Paraffin sections of tumors were examined for histopathology, and stained for immunohistochemistry with antibodies to PARP or ER.

[776] Gene expression was measured by RT-PCR. Figure 16 show ratios of gene expression in E-3a-diol (6.6 mg) treated tumor samples relative to vehicle treated, normalized to β-actin. Amphiregulin (Areg), androgen receptor (AR), tumor protein 53 (p53), Bcl2 antagonist of cell death (Bad), apoptosis regulator BAX (bax), B-cell CLL/lymphoma 2 (Bcl-2), caspase 3 (Casp3), caspase 8 (CaspS), caspase 9 (Casp9), Cyclin D1 (Ccndl), estrogen receptor alpha (ERd), estrogen receptor beta (ERfi, progesterone receptor isoform A (PR-A), and vascular endothelial growth factor ( VEGF). E-3a-diol treatment up-regulates pro-apoptotic genes in tumors and down-regulates tumor proliferation and malignancy genes.

[777] Example 20. Determination of Non-AR Transcription Factor Activation

[778] Based on the rapid phosphorylation of Erk-1 induced by E-3a-diol, an experiment was performed to determine if E-3a-diol also binds to transcription factors other than nuclear hormone receptors. Binding of several transcription factors related to NF-KB, namely cRel (a member of the NF-κΒ family), and RREB1 (Ras Response Element Binding 1 ) to proteins of bind to E-3a-diol. E-3a-diol also induced nuclear extract transcription factor protein binding to the Ahr/ARNT dimer. This dimer activates transcription of genes (such as Cyp 1 A1 , Cyp 1 B1 , and NADP[H] oxidoreductase) involved in the metabolism of xenobiotics. Transcription of this class is dependent on prior binding of Ahr (Aryl hydrocarbon receptor) to a xenobiotic ligand.

[779] Example 22. E-3cc-diol Nongenomic Signaling and AR Phosphorylation Status

[780] E-3a-diol exhibits minimal binding to the wild-type i-AR, but an appropriate assay system to determine if E-3a-diol directly binds to the mutated AR expressed in LNCaP cells is lacking. However, it is well established that DHT results in a rapid phosphorylation of AR in LNCaP cells. Thus, rapid i-AR phosphorylation by E-3a-diol in the manner similar to that observed with DHT.

[781 ] After incubating LNCaP cells with 50 nM E-3a-diol for 10 and 30 minutes there was no increase in /-AR phosphorylation that could be attributed to nongenomic signaling, whereas 30 minute incubation with 10 nM DHT increased AR phosphorylation. As shown by Figure 17, the increased /-AR phosphorylation by Docket No. 354. PATENT

DHT was not inhibited when 10 nM DHT and 50 nM E-3odiol are co-administered (i.e., E-3odiol does not inhibit DHT-induced /-AR phosphorylation). Thus, E-3odiol is not interacting with /-AR to antagonize DHT's interaction with that nuclear hormone receptor. After 2 hours, which is a time course not consistent with a direct non- genomic action, incubation with 50 nM E-3odiol (HE3235) results in i-AR phosphorylation. Again 10 nM DHT and the combination of 10 nM DHT and 50 nM E-3odiol showed the same DHT-induced i-AR phosphorylation at the longer time point.

[782] Example 23. Comparison of Anti-prol iterative Effect of High Dose Androgen with that of E-3cc-diol

[783] DHT at pM to nM concentrations is pro-proliferative but induces apoptosis at μΜ concentrations [Sonnenschein et al. (1989); Kim et al (1996); Szelei et al. (1997); Prehn (1999); Bruckheimer et al. (2001 )]. The dependence of proliferation with DHT concentration may be due to further engagement of GPR-C6a at concentrations where saturation of i-AR has already occurred. Suppression of LNCaP proliferation by high dose dihydrotestosterone is reported to occur through production of TGF-β. Mannosed-6-phosphate (M6P) inhibits the activation of latent TGF-β thus reversing that DHT proliferation inhibition. To determine if inhibition of LNCaP proliferation by E-3a-diol was occurring through the same above-mentioned high-dose DHT pathway the activity of E-3a-diol in the presence or absence of M6P was performed.

[784] The cells were plated at 7,000 c/w in lysine-coated 96 well plates and incubated overnight in 10% FBS. Next day, the medium was changed to 10% CA- FBS and treated as shown in the graph with DMSO, 1 μΜ or 100 μΜ M-6-P + indicated dose of DHT or E-3odiol for 4 days. Proliferation was measured with BrdU. Results of Figure 21 show M-6-P reduced the proliferation response of LNCaP when proliferation was induced by low concentration of DHT. However, M-6-P prevented growth inhibition induced by high concentration of DHT.

[785] The results of Figure 18 and 19 indicate that E-3a-diol (HE3235) does not inhibit LNCaP proliferation in a M6P sensitive manner. Example 22 demonstrates that E-3odiol does not induce phosphorylation of AR within 10 minutes of incubation of LNCaP cells, but DHT does. Taken together those data indicate that if DHT and HE3235 are both agonists at GPR-C6a, the rapid nongenomic signaling they induce must have differential longer term effects on i-AR. Those differential effects may reflect differences in their direct interaction with the mutant /-AR in LNCaP.

[786] 7. Citations Docket No. 354. PATENT

[787] Agapova, O.A. et al. (2006) "A neuroactive steroid 5a-androstane-3a,17b-diol regulates androgen receptor levels in astrocytes" J. Neurochem. 98: 355-363

[788] Ahlem, C. et al. (201 1 ) "17a-Ethynyl-5a-androstane-3a, 17β- ϊοΙ treatment of MNU-induced mammary cancer in rats" Int. J. Breast Cancer 6o\A 0.4061/201 1/618757

[789] Ahmad, N. et al. (1978) "The effects of 5a-reduced androgens on maintenance and regeneration of prostate glands and seminal vesicles in castrated and

hypophysectomized rats" Anat. Rec. 192: 543-554.

[790] Alimiraha, F. et al. (2006) "DU-145 and PC-3 human prostate cancer cell lines express androgen receptor: Implications for the androgen receptor functions and regulation" FEBS Lett. 580(9): 2294-2300

[791] Amanchy, R. et al. (2008) "Identification of c-Src tyrosine kinase substrates using mass spectrometry and peptide microarrays" J. Proteome Res. 7: 3900-10

[792] Armstrong, S.P. et al. (2009) "Gonadotropin-releasing hormone and protein kinase C signaling to Erk: spatiotemporal regulation of Erk by docking domains and dual- specificity phosphatases" Mol. Endocr. 23: 510-9

[793] Arnold, J.T. et al. (2007) "Androgen receptor or estrogen receptor-beta blockade alters DHEA-, DHT-, and E(2)-induced proliferation and PSA production in human prostate cancer cells" Prostate 67(1 1 ): 1 152-62

[794] Arterburn, J.B. et al. (2000) "Novel 17a-ethynylestradiol derivatives: Sonogashira couplings using unprotected phenylhydrazines" Tet. Lett. 41 : 839-842

[795] Ballou, L.M. et al. (2003) "Activated G alpha q inhibits p1 10 alpha phosphoinositol 3-kinase and Akt" J. Biol. Chem. 278: 23472-23479

[796] Ballou, L.M. et al. (2006) "Gaq binds to p100a/p85a phosphoinositide 3-kinase and displaces Ras" Biochem. J. 394: 557-562

[797] Bauer, J. A. et al. (2003) "Growth inhibition and differentiation in human prostate carcinoma cells induced by the vitamin D analog 1 alpha, 24-dihydroxyvitamin D ₂ " Prostate 55: 159-67

[798] Bauman, D.R. et al. (2006) "Identification of the major oxidative 3a-hydroxysteroid dehydrogenase in human prostate that converts 5a-androstane-3a,173-diol to 5a- dihydrotestosterone: a potential therapeutic target for androgen dependent disease" Mol. Endocrinol. 20: 444-458

[799] Becker, K.L., Ed. Principles and Practice of Endocrinology, 3rd ed., Lippincott Docket No. 354. PATENT

Williams & Wilkins, Philadelphia, Pa. (2001 ), pp. 1183-1 194

[800] Ben-Ami, I. et al. (201 1 ) "Gq protein-induced apoptosis is mediated by Akt kinase inhibition that leads to protein kinase C-induced c-Jun N-terminal kinase activation" J. Biol. Chem. 286: 31022-31031

[801] Berger, R. et al. (2006) "Androgen-dependent regulation of Her-2/neu in prostate cancer cells", Cancer Res. 66: 5723-28

[802] Boston, S.R. et al. (201 1 ) "Characterization of ERK docking domain inhibitors that induces apoptosis by targeting Rsk-1 and capase-9" BMC Cancer 1 1 :7

[803] Bradford, M.M. (1976) "A rapid and sensitive method for quantification of microgram quantities of protein utilizing the principle of protein-dye binding" Anal.

Biochem. 72: 248-254

[804] Brown, M.D. and Sacks, D.B. (2009) "Protein scaffolds in MAPK signaling" Cell Signal. 21 : 462-469

[805] Burgering, B.M. and Coffer, P.J. (1995) "Protein kinase B/c-Akt in

phoshatidylinositol-3-OH kinase signal transduction", Nature 376: 599-602

[806] Cantor, A. (1997) "Extending SAS survival analysis techniques for medical research" SAS Institute Inc. pp.59-61

[807] Carlson, S.M. et al. "Large-scale discovery of ERK2 substrates identifies ERK- mediated transcriptional regulation by ETV3" Science Signaling 4 (196), rs1 1 . [DOI: 10.1 126/scisignal.2002010]

[808] Carpenter, C.L. et al. (1993) "Phosphoinositide 3-kinase is activated by phosphopeptides that bind to the SH2 domain of the 85-kDa subunit" J. Biol. Chem. 268: 9478-9483

[809] Casar, B. et al. (2009) "ERK dimers and scaffold proteins" Cell Cycle 8: 1007- 1013

[810] Castoria, G. et al. (2004) "Rapid signaling pathway activation by androgens in epithelial and stromal cells" Steroids 69(8-9): 517-22

[811] Caunt, C.J., et al. (2006) "Arrestin-mediated Erk activation by gonadotropin- releasing hormone receptors" J. Biol. Chem. 281 : 2701 -2710

[812] Chang, C.S. et al. (1998) "Molecular cloning of human and rat complementary DNA encoding androgen receptors" Science 240:324-326 Docket No. 354. PATENT

[813] Chang, C.Y. et al. (1999) "Dissection of the LXXLL nuclear receptor-coactivator interaction motif using combinatorial peptide libraries: discovery of peptide antagonists of estrogen receptors a and β" Mol. Cell Biol. 19: 8226-8239

[814] Chen, Z. et al. (2001 ) "MAP kinases" Chem. Rev. 101 : 2449-2475; Roux, P.P. and Blenis, J. (2004) "ERK and p38 MAPK-activated protein kinases: a family of protein kinases with diverse biological functions" Microbiol. Mol. Biol. Rev. 68: 320-344

[815] Chuderland, D. et al. (2008) "Calcium-mediated interactions regulate the subcellular localization of extracellular signal-regulated kinases" J. Biol. Chem. 283: 1 1176-1 1 188;

[816] Chuderland, D. and Seger, R. (2008) "Calcium regulates Erk signaling by modulating its protein-protein interactions" Comm. Integr. Biol. 1 : 4-5

[817] Coleman, I.M. et al. (2006) "Inhibition of androgen-independent prostate cancer by estrogenic compounds is associated with increased expression of immune-related genes" Neoplasia 8: 862-78

[818] Corey E. et al. (2002) "Establishment and characterization of osseous prostate cancer models: intra-tibial injection of human prostate cancer cells" Prostate 52: 20-33

[819] Corey, E. et al. (2003) "LuCaP 35: a new model of prostate cancer progression to androgen independence" Prostate 55: 239-246

[820] Culig, Z. et al. (1993) "Mutant androgen receptor detected in an advanced-stage prostatic carcinoma is activated by adrenal androgens and progesterone" Mol. Endocr. 7: 1541 -50

[821] De Bellis, A. et al. (1992) "Single base mutations in the human androgen receptor gene causing complete androgen insensitivity: rapid detection by a modified denaturing gradient gel electrophoresis technique" Mol. Endocr. 6: 1909-20

[822] Defea, K. (2008) "β-arrestin and heterotrimeric G proteins: collaborators and competitors in signal transduction" Br. J. Pharmacol. 153: S298-309

[823] De Goeij, A.F. (1986) "Steroid-bovine serum albumin conjugates: molecular characterization and their interaction with androgen and estrogen receptors" J. Steroid Biochem. 24: 1017-1031

[824] DeSombre, E.R. and Arbogast, L.Y. (1974) "Effect of the antiestrogen C1628 on the growth of rat mammary tumors" Cancer Res. 34:1971 -6

[825] Ding, V.D.H. et al. (1998) "Sex hormone-binding globulin mediates prostate Docket No. 354. PATENT androgen receptor action via a novel signaling pathway" Endocrinol. 139: 213-218 (1998).

[826] Dowal, L. et al. (2006) "Stable association between Go _q and phospholipase Οβ1 in living cells" J. Biol. Chem. 281 : 23999-24014

[827] Dozmorov, M.G. et al. (2007) "5a-androstane-3a,173-diol selectively activates the canonical PI3K/Akt pathway: a bioinformatics-based evidence for androgen-activated cytoplasmic signaling" Genomic Med. 1 : 139-146

[828] Eastell, R. et al. (2006) "Effect of an aromatase inhibitor on BMD and bone turnover markers: 2-year results of the Anastrozole, Tamoxifen, Alone or in Combination (ATAC) trial (18233230)" J. Bone Miner. Res. 21 :1215-23

[829] Esquenet, M, et al. (1996) "Control of LNCaP proliferation and differentiation: actions and interactions of androgens, 1 alpha,25-dihydroxycholecalciferol, all-trans retinoic acid, 9-cis retinoic acid, and phenylacetate" Prostate 28: 182-94

[830] Estrada, M. et al. (2003) "Testosterone stimulates intracellular calcium release and mitogen-activated protein kinases via a G protein-coupled receptor in skeletal muscle cells" Endocrinol. 144: 3586-3597

[831] Ewen, M.E. et al. (1993) "Functional interactions of the retinoblastoma protein with mammalian D-type cyclins" Cell 73: 487-497

[832] Flynn, D.C. (2001 ) "Adaptor proteins" Oncogene 20: 6270-72

[833] Fujii, Y. et al. (2004) "Regulation of prostate-specific antigen by activin A in prostate cancer LNCaP cells" Am. J. Physiol. Endocrinol. Metab. 286: E927-31

[834] Gaddipati, P. et al. (2004) "Frequent detection of codon 877 mutation in the androgen receptor gene in advanced prostate cancers" Cancer Res. 54: 2861 -4

[835] Gaest, C. et al. (2009) "Tight control of MEK-ERK activation is essential in regulating proliferation, survival and cytokine production of CD34+-derived neutrophil progenitors" Blood 1 14: 3402-341 1

[836] Gatson, J.W. and Singh, M. (2007) "Activation of a membrane-associated androgen receptor promotes cell death in primary cortical astrocytes" Endocrinol. 148: 2458-2464

[837] Gatson, J.W. et al. (2006) "Dihydrotestosterone differentially modulates the mitogen-activated protein kinase and the phosphoinositide 3-kinase/Akt pathways through the nuclear and novel membrane androgen receptor in C6 cells" Endocrinol. 147: 2028-2034 Docket No. 354. PATENT

[838] Gleave, M.E. et al. (1998) "Butyrate analogue, isobutyramide, inhibits tumor growth and time to androgen-independent progression in the human prostate LNCaP tumor model" J. Cell Biochem. 69: 271 -81

[839] Golebiewska, U. et al. (2008) "Go _q binds two effectors separately in cells:

evidence for predetermined signaling pathways" Biophys. J. 95: 2575-82

[840] Good, M.C. et al. (201 1 ) "Scaffold proteins: hubs for controlling the flow of cellular information" Science 332: 680-686

[841] Goodrich, D.W. et al. (1991 ) "The retinoblastoma gene product regulates progression through the G1 phase of the cell cycle" Cell 67: 293-302

[842] Gottlieb, B. et al. (2004) "The androgen receptor gene mutations database (ARDB): 2004 update" Human Mut. 23(6): 527-533

[843] Gregory, C.W. et al. (2005) "Heregulin-induced activation of HER2 and HER3 increases androgen receptor transactivation and CWR-R1 human recurrent prostate cancer cell growth" Clin. Cancer Res. 1 1 : 1704-12

[844] Green, S. et al. (1986) "Human oestrogen receptor cDNA: sequence, expression and homology to v-erb-A" Nature 320: 134-139

[845] Greene, G.L. et al. (1986) "Sequence and expression of the human estrogen receptor complimentary DNA" Science 231 :1 150-1 154

[846] Grossmann, M.E. et al. (2001 ) "Androgen Receptor Signaling in Androgen- Refractory Prostate Cancer" J. Nat'l Cancer Inst. 93(22): 1687-97

[847] Gu, S. et al. (2009) "Functional membrane androgen receptors in colon tumors trigger pro-apoptotic responses in vitro and reduce drastically tumor incidence in vivo" Mol. Cancer 8: 1 14

[848] Gu, S. et al. (201 1 ) "Activation of membrane androgen receptor in colon cancer cells inhibits the prosurvival signals Akt/Bad in vitro and in vivo and blocks migration via vinculin/actin signaling" Mol. Med. 17: 48-58

[849] Gusterson, B.A. and Williams, J.C. (1981 ) "N-nitrosomethylurea-induced rat mammary tumours as models of human breast cancer" J. R. Soc. Med. 74: 56-9

[850] Hall, R.A. and Lefkowitz, R.J. (2002) "Regulation of G protein-coupled signaling by scaffold proteins" Circ. Res. 91 : 672-680

[851] Hatzoglou, A. et al. (2005) "Membrane androgen receptor activation induces apoptotic regression of human prostate cancer cells in vitro and in vivo" J. Clin. Docket No. 354. PATENT

Endocrinol. Metab. 90(2): 893-903

[852] Horoszewicz, J.S. et al. (1983) "LNCaP model of human prostatic carcinoma" Cancer Res. 43: 1809-1818

[853] Ishihara, Y. et al. (201 1 ) "Increased expression of c-fos by ERK activation under sustained oxidative stress elicits BimEL upregulation and hepatocyte apoptosis" FEBS J. doi: 10.1 1 1 1/j.1742-4658.201 1 .08105.x.

[854] Jimenez, C. et al. (2002) "The p85 regulatory subunit controls sequential activation by Tyr kinases and Ras" J. Biol. Chem. 277: 41556-41562

[855] Jirmanova, L. et al. (2002) "Differential contributions of ERK and PI3-kinase to the regulation of cyclin D1 expression and to the control of the d/S transition in mouse embryonic stem cells" Oncogene 21 : 5515-28

[856] Jones, S.E. et al. (2000) "Phase III trial comparing doxorubicin plus

cyclophosphamide with docetaxel plus cyclophosphamide as adjuvant therapy for operable breast cancer" J. Clin. Oncol. 24:5381 -7

[857] Jordan, A. et al. (1998) "Tubulin as a target for anticancer drugs: agents which interact with the mitotic spindle" Med. Res. Rev. 18(4): 259-296

[858] Kaighn, M.E. et al. (1979) "Establishment and characterization of a human prostatic carcinoma cell line (PC-3)" Invest. Urol. 17: 16-23

[859] Kallergi, G. et al. (2007) "Activation of FAK/PI3K/Rac1 signaling controls actin reorganization and inhibits cell mobility in human cancer cells" Cell. Physiol. Biochem. 20: 977-986

[860] Kampa, M. et al. (2002) "The human prostate cancer cell line LNCaP bears functional membrane testosterone receptors that increase PSA secretion and modify actin cytoskeleton" FASEB J. 16: 1429-1431

[861] Kampa, M. et al. (2004) "The opioid agonist ethylketocyclazocine reverts the rapid, non-genomic effects of membrane testosterone receptors in the human prostate LNCaP cell line" Exp. Cell. Res. 294: 434-445

[862] Kampa, M. et al. (2005) "Opposing effects of estradiol- and testosterone- membrane binding sites on T47D breast cancer cell apoptosis" Exp. Cell Res. 307: 41 -51

[863] Kato J.-Y. et al. (1993) "Direct binding of cyclin D to the retinoblastoma gene product (pRb) and pRb phosphorylation by the cyclin D-dependent kinase CDK4" Genes Dev. 7: 331-342 Docket No. 354. PATENT

[864] Kawamata, Y. et al. (2007) "Tumor necrosis factor receptor-1 can function through a Go _q/ ii-p-arrestin-1 signaling complex", J. Biol. Chem. 282: 28549-28556

[865] Kennedy, J.S. et al. (1999) "Signaling scaffolds in immune cells" Cell. Calcium 26: 227-235

[866] Kiefer J. A. et al. (2004) "The effect of osteoprotegerin administration on the intra- tibial growth of the osteoblastic LuCaP 23.1 prostate cancer xenograft" Clin. Exp.

Metastasis 21 : 381 -7

[867] Kim, J. et al. (2009) "Independent p-arrestin2 and Go _q /protein kinase

Οζ pathways for Erk stimulation by angiotensin type 1 A receptors in vascular smooth muscle cells converge on transactivation of the epidermal growth factor receptor" J. Biol. Chem. 284: 1 1953-1 1962

[868] Kim, M.J. et al. (2010) "AMP-activated protein kinase antagonizes pro-apoptotic extracellular signal-regulated kinase activation by inducing dual-specificity protein phosphatases in response to glucose deprivation in HCT1 16 carcinoma" J. Biol. Chem. 285: 14617-27

[869] Kim, S.B. et al. (2007) "Nongenomic activity of ligands in the association of androgen receptor with Src" ACS Chem. Biol. 2: 484-492.

[870] Kodaki, T. et al. (1994) "The activation of phosphatidylinositol 3-kinase by Ras" Curr. Biochem. 4: 798-806)

[871] Kohout, T.A. et al. (2001 ) "β-arrestin-l and -2 differentially regulate heptahelical receptor signaling and trafficking" Proc. Natl Acad. Sci. (USA) 98: 1601 -1606

[872] Kolch, W. (2000) "Meaningful relationships: the regulation of the

Ras/Raf/MEK/ERK pathway by protein interactions" Biochem. J. 351 : 289-305

[873] Kolch, W. (2005) "Coordinating Erk/MAPK signaling through scaffolds and inhibitors" Nat. Rev. Mol. Cell Biol. 6: 827-837

[874] Koreckij, T.D. et al. (2009) ΉΕ3235 inhibits growth of castration-resistant prostate cancer" Neoplasia, 1 1 (11 ): 1216-25

[875] Kousteni, S. et al. (2001 ) "Nongenotropic, sex-nonspecific signaling through estrogen or androgen receptors: dissociation from transcriptional activity", Cell 104: 719- 30

[876] Kristiansen, K. (2004) "Molecular mechanisms of ligand binding, signaling, and regulation within the superfamily of G protein-coupled receptors: molecular modeling and Docket No. 354. PATENT mutagenesis approaches to receptor structure and function" Pharmacol. Ther. 103: 21 -80

[877] Kruh, J. (1982) "Effects of sodium butyrate, a new pharmacological agent, on cells in culture" Mol. Cell Biochem. 42: 65-82

[878] Kuiper, G.G.J. M. et al. (1996) "Cloning of a novel estrogen receptor expressed in rat prostate and ovary " Proc. Nat'l. Acad. Sci. (USA) 93: 5925-30

[879] Kuiper, G.G.J. M. et al. (1997) "Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors a and β" Endocr.138: 863-70

[880] Kumar, N.V. et al. (2001 ) "Ten Erk-related proteins in three distinct classes associate with AP-1 proteins and/or AP-1 DNA" J. Biol. Chem. 276(34): 32362-32372

[881] Kumar, S. et al. (2003) "p38 MAP kinases: key signaling molecules as therapeutic targets for inflammatory diseases" Nat. Rev. Drug Discov., 2: 717-726

[882] Langeler, E.G. et al. (1993) "Effect of culture conditions on androgen sensitivity of the human prostatic cancer cell line LNCaP" Prostate 23(3): 213-23

[883] Laufer, S.A. et al. (2006) "Substituted isoxazoles as potent inhibitors of p38 MAP kinase" Chem. Med. Chem. 1 : 197-207

[884] Lee, C. et al. (1995) "Regulation of proliferation and production of prostate specific antigen in androgen-sensitive prostatic cancer cell, LNCaP, by dihydrotestosterone" Endocr. 136(2): 796-803

[885] Lefloch, R. et al. (2008) "Single and combined silencing of ERK1 and ERK2 reveals their positive contribution to growth signaling dependent on their expression levels" Mol. Cell Biol. 28: 51 1 -527

[886] Leihy, M.W. et al. (2001 ) "Virilization of the urogenital sinus of the tammar wallaby is not unique to 5oandrostane-3a,173-diol" Mol. Cell. Endocrinol. 181 : 1 1 1 -1 15

[887] Li, J. et al. (1997) "PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer" Science 275 (5308): 1943-7.

[888] Lieberherr, M.L. and Grosse, B. (1994) "Androgens increase intracellular calcium concentration and inositol 1 ,4,5-trisphosphate and diacylglycerol formation via a pertussis toxin-sensitive G-protein" J. Biol. Chem. 269: 7217-7223

[889] Liu, W.W., et al. (1996) "Transformation of Rat-1 fibroblasts with v-src oncogene increases tyrosine phosphorylation state and activity of the a subunit of Gq/G1 1 " Proc. Nat'l. Acad. Sci. (USA) 93: 8258-8263 Docket No. 354. PATENT

[890] Lubahn, D.B. et al. (1988) "The Human Androgen Receptor: Complementary Deoxyribonucleic Acid Cloning, Sequence Analysis and Gene Expression in Prostate" Mol. Endocr. 2: 1265-75

[891] Lubahn, D.B. (1989) "Sequence of the intron/exon junctions of the coding region of the human androgen receptor gene and identification of a point mutation in a family with complete androgen insensitivity" Proc. Nat'l. Acad. Sci. (USA) 86: 9534-8

[892] Luttrell, L.M. et al. (2001 ) "Activation and targeting of extracellular signal-regulated kinases by β-arrestin scaffolds" Proc. Nat'l Acad. Sci. (USA) 98: 2449-2454

[893] Luttrell, D.K. and Luttrell, L.M. (2004) "Not so strange bedfellows: G protein- coupled receptors and Src family kinases" Oncogene 23: 7969-7978

[894] Machida, K. et al. (2003) "Profiling the global tyrosine phosphorylation state" Mol. Cell. Proteomics 2: 215-233

[895] Mahendroo, M.S. et al. (1999) "The parturition defect in steroid 5a-reductase type 1 knockout mice is due to impaired cervical ripening" Mol. Endocrinol. 13: 981 -992

[896] Marcelli, M. et al. (1990) "Definition of the human androgen receptor gene structure permits the identification of mutations that cause androgen resistance:

premature termination of the receptor protein at amino acid residue 588 causes complete androgen resistance" Mol. Endocr. 4: 1 105-16

[897] Marchi, M. et al. (2008) "The N-terminal domain of ERK1 accounts for the functional differences with ERK2" PloS One 3(12): e3783

[898] McCormick, D.L. et al. (1981 ) "Lifetime dose-response relationships for mammary tumor induction by a single administration of N-methyl-N-nitrosourea" Cancer Res. 41 : 1690-4

[899] McCubrey, J.A. et al. (2007) "Roles of the RAF/MEK/ERK partway in cell growth, malignant transformations and drug resistance" Biochim. Biophys. Acta 1773: 1263-1284

[900] Medina, D. (2000) "The preneoplastic phenotype in murine mammary

tumorigenesis" J. Mammary Gland Biol. Neoplasia 5(4): 393-407

[901] Medina, D. (2008) "Premalignant and malignant mammary lesions induced by MMTV and chemical carcinogens" J. Mammary Gland Biol. Neoplasia. 13(3): 271 -7

[902] Medina, D. and Thompson, H. (2000) "A comparison of the salient features of mouse, rat and human mammary tumorigenesis" In: Methods in mammary gland biology and breast cancer research, Ip, M. and Asch, B., Eds., New York: Kluwer Docket No. 354. PATENT

Academic/Plenum publishers, pp. 31 -6

[903] Mellor, H. and Parker, P.J. (1998) "The extended protein kinase C superfamily", Biochem. J. 332: 281 -292

[904] Migliaccio, A. et al. (2005) "Steroid receptor regulation of epidermal growth factor signaling through Src in breast and prostate cancer cells: steroid antagonist action", Cancer Res. 65: 10585-93

[905] Miller, A.B. (1978) "An overview of hormone-associated cancers" Cancer Res. 38: 3985-3990

[906] Miranda, C. (2006) "The Tfg protein, involved in oncogenic rearrangements, interacts with TANK and NEMO, two proteins involved in the NF-κΒ pathway" J. Cell. Physiol. 208: 154-160

[907] Miyake, S. et al. (1997) "The Cbl proto-oncogene product: from an enigmatic oncogene to center stage of signal transduction" Crit. Rev. Oncol. 8: 189-218

[908] Miyamoto, H. et al. (1998) "Delta 5-androstenediol is a natural hormone with androgenic activity in human prostate cancer cells" Proc. Nat'l. Acad. Sci. (USA) 95(19): 11083-8

[909] Mizokami, A. et al. (2004) "The adrenal androgen androstenediol is present in prostate cancer tissue after androgen deprivation therapy and activates mutated androgen receptor", Cancer Res. 64(2): 765-71

[910] Morel, C. et al. (2005) "Cross-interactions of two p38 mitogen-activated protein (MAP) kinase inhibitors and two cholecystokinin (CCK) receptor antagonists with the CCK1 receptor and p38 MAP kinase" J. Biol. Chem. 280: 21384-21393

[911] Morris, A.J. and Malbon, C.C. (1999) "Physiological regulation of G protein-linked signaling" Physiol. Rev. 79: 1373-1430

[912] Mosselman, S. et al. (1996) "ER beta: identification and characterization of a novel human estrogen receptor" FEBS Lett. 392: 49-53

[913] Murphy, G.P., Ed. (1980) "Models for prostate cancer" Liss AR, Inc., New York, pp. 1 15-132

[914] Myrthue, A. et al. (2008) "The iroquois homeobox gene 5 is regulated by 1 , 25- dihydroxyvitamin D3 in human prostate cancer and regulates apoptosis and the cell cycle in LNCaP prostate cancer cells" Clin. Cancer Res. 14(1 1 ): 3562-70

[915] Nakla, A.M. et al. (1995) "5a-Andostane-3a-173-diol is a hormone: stimulation of Docket No. 354. PATENT cAMP accumulation in human and dog prostate "J Clin. Endocrinol. Metab. 80: 2259- 2262

[916] Nakla, A.M. et al (1990) "Biologically active steroids activate receptor-bound human sex steroid-binding globulin to cause LNCaP cells to accumulate adenosine 3', 5'- monophosphate" J. Clin. Endocrinol. Metab. 71 : 398-404 (1990)

[917] Nallou, P. and Mayer, B.J. (2001 ) "Profiling the global tyrosine phosphorylation state by Src homology 2 domain binding" Proc. Nat'l Acad. Sci. (USA) 98: 13531 -36

[918] Nesslinger, N.J. et al. (2003) "Androgen-independent growth of LNCaP prostate cancer cells is mediated by gain-of-function mutant p53" Cancer Res. (2003) 63: 2228-33

[919] Newmark, J.R. et al. (1992) "Androgen receptor gene mutations in human prostate cancer" Proc. Nat'l. Acad. Sci. (USA) 89: 6319-23

[920] Nguyen, T.-V. V. (2005) "Androgens activate mitogen-activated protein kinase signaling: role in neuroprotection" J. Neurochem. 94: 1639-1651

[921] Nunlist, E.H. et al. (2004) "Partitioning of 5a-dihydrotestosterone and 5a- androstane-3a,173-diol activated pathways for stimulating human prostate LNCaP proliferation" J. Steroid Biochem. Mol. Biol. 91 : 157-170

[922] Ong, S.-E. and Mann, M. (2006) "A practical recipe for stable isotope labeling by amino acids in cell culture (SILAC)". Nature Protoc 1 : 2650-2660

[923] Ong, S.-E. (2010) "Unbiased identification of protein-bait interactions using biochemical enrichment and quantitative proteomics" Cold Spring Harb. Protoc.

doi:10.1 101

[924] Paech, K. et al. (1997) "Differential ligand activation of estrogen receptors ERalpha and ERbeta at AP1 sites" Science 277: 1508-1510

[925] Pal, P. and Kanaujiya, J.K. (201 1 ) "2-D gel electrophoresis-based proteomic analysis reveals that ormeloxifene induces G0-G1 growth arrest and ERK-mediated apoptosis in chronic myeloid leukemia cells K562" Proteomics. 1 1 : 1517-29

[926] Papadopoulou, N. et al. (2008a) "Rho/ROCK/actin signaling regulates membrane androgen receptor induced apoptosis in prostate cancer cells", Exp. Cell. Res. 314: 3162- 74

[927] Papadopoulou, N. et al (2008b)"Membrane androgen receptor activation triggers down-regulation of PI-3K/Akt/NF-kappaB activity and induces apoptotic responses via Bad, FasL and capase-3 in DU145 prostate cancer cells" BMC Mol. Cancer 7:88, doi: Docket No. 354. PATENT

10.1 186/1476-4598-7-88.

[928] Papakonstanti, EA. et al. (2003) "A rapid, nongenomic, signaling pathway regulates the actin reorganization induced by activation of membrane testosterone receptors" Mol. Endocrinol. 17: 870-881

[929] Pelekanou, V. M. et al. (2010) "Testosterone membrane-initiated action in breast cancer cells: Interaction with the androgen signaling pathway and EPOR" Mol. Oncol. 4: 135-149

[930] Penning, T.M. (1997) ""Molecular endocrinology of hydroxysteroid

dehydrogenase" Endocr. Rev. 18: 281 -306

[931] Peterziel, H., et al. (1999) "Rapid signaling by androgen receptor in prostate cancer cells", Oncogene 18: 6322-29

[932] Pi, M. et al. "GPRC6A mediates the non-genomic effects of steroids" J. Biol. Chem. 285: 39953-39964 (2010).

[933] Pi, M. and Quarles, L.D. (201 1 ) "GPRC6a regulates prostate cancer progression" Prostate 72: 399-409

[934] Rajkumar, L. et al. (2001 ) "Short-term exposure to pregnancy levels of estrogen prevents mammary carcinogenesis" Proc. Nat'l. Acad. Sci. (USA) 98: 1 1755-9

[935] Reszka, A. A. et al. (1995) "Association of mitogen-activated protein kinase with the microtubule cytoskeleton" Proc. Nat'l. Acad. Sci. (USA) 92: 8881 -5

[936] Ritter, S. L. and Hall, R. A. (2009) "Fine-tuning of GPCR activity by receptor- interacting proteins" Nat. Rev. Mol. Cell Biol. 10: 819-830

[937] Robinson, D. et al. (1996) "A tyrosine kinase profile of prostate carcinoma" Proc. Nat'l. Acad. Sci. (USA) 93: 5958-5962

[938] Robitaille, H. (2010) "The small heat-shock protein Hsp27 undergoes ERK- dependent phosphorylation and redistribution to the cytoskeleton in response to dual leucine zipper-bearing kinase expression" J. Invest. Dermatol. 130: 74-85

[939] Rodriquez-Viciana, P. et al. (1994) "Activation of phosphoinositol 3-kinase by interaction with Ras and point mutation" EMBO J. 15: 2442-51

[940] Romijn, J. C. et al. (2006) "Application of the MTT assay to human prostate cancer cell lines in vitro: Establishment of test conditions and assessment of hormone- stimulated growth and drug-induced cytostatic and cytotoxic effects" Prostate 12(1 ): 99- 110 Docket No. 354. PATENT

[941] Roux, P.P. and Blenis, J. (2004) "ERK and p38 MAPK-activated protein kinases: a family of protein kinases with diverse biological functions" Microbiol. Mol. Biol. Rev. 68: 320-344

[942] Schieltz, D.M. and Washburn, M.P. (2006a) "Analysis of complex protein mixtures using multidimensional protein identification technology (MuDPIT)" Cold Spring Harb. Protoc, doi:10.1 101/pdb.prot4555

[943] Schieltz, D.M. and Washburn, M.P. (2006b) "Digestion of insoluble protein fractions for multidimensional protein identification technology (MuDPIT) analysis" Cold Spring Harb. Protoc, doi:10.1 101/pdb.prot4556

[944] Schultz, F.M. and Wilson, J.D. (1974) "Virilization of the Wolffian duct in the rat fetus by various androgens" Endocrinol. 94: 979-986.

[945] Shafquat, N. et al. (2003) "Expanded substrate screenings of human and

Drosophila type 10 17p-hydroxysteroid dehydrogenases (HSDs) reveal multiple specificities in bile acid and steroid hormone metabolism: characterization of

multifunctional 3α/7α/7β/17β/20β/21 -HSD" Biochem. J. 376: 40-60.

[946] Shaw, G. et al. (2000). "Prostate formation in a marsupial is mediated by the testicular androgen 5a-androstane-3a,173-diol" Proc. Nat'l. Acad. Sci. (USA) 97: 12256- 12259

[947] Sheridan, D.L. et al. (2008) "Substrate discrimination among mitogen-activated protein kinases through distinct docking sequence motifs" J. Biol. Chem. 283: 1951 1 -20

[948] Simon, M.I. et al. (1991 ) "Diversity of G proteins in signal transduction" Science 252: 802-8

[949] Smollich, M. et al. (2009) "Differential effects of aromatase inhibitors and antiestrogens on estrogen receptor expression in breast cancer cells" Anticancer Res. 29: 2167-71

[950] Somjen, D. et al. "Role of putative membrane receptors in the effect of androgens on human vascular cell growth" J. Endocrinol. 180: 97-106 (2004)

[951] van Sommeren, A.P.G. et al. (1993) "Comparison of three activated agaroses for use in affinity chromatography: effects on coupling performance and ligand leakage" J. Chromatogr. 639: 23-31

[952] Stone, K.R. (1978) "Isolation of a human prostate carcinoma cell line (DU 145)" Int. J. Cancer 2λ 274-281 Docket No. 354. PATENT

[953] Sugita, S. (2004) "Effect of type I growth factor receptor tyrosine kinase inhibitors on phosphorylation and transactivation activity of the androgen receptor in prostate cancer cells: Ligand-independent activation of the N-terminal domain of the androgen receptor", Oncol. Rep. 11 (6): 1273-9

[954] Sun, M. et al. (2003) "Activation of phosphoinositol 3-kinase/Akt pathway by androgen through interaction of p85a, androgen receptor, and Src" J. Biol. Chem. 278: 42992-43000

[955] Suzuki, H. et al. (1993) "Androgen receptor gene mutations in human prostate cancer" J. Steroid Biochem. Mol. Biol. 46: 759-65

[956] Szelei, J. et al. (1997) "Androgen induced inhibition of proliferation in human breast cancer MCF7 cells transfected with androgen receptor" Endocrinol. 138:1406- 1412

[957] Takai, Y. et al. (2001 ) "Small GTP-binding proteins" Physiol. Rev. 81 : 153-208

[958] Taplin, M.E. et al. (1995) "Mutation of the Androgen-Receptor Gene in Metastatic Androgen-lndependent Prostate Cancer" New Engl. J. Med. 332(21 ): 1393-1398

[959] Tepper, C.G. et al. (2005) "Profiling of gene expression changes caused by p53 gain-of-function mutant alleles in prostate cancer cells" Prostate 65(4): 375-89

[960] Thalmann, G.N. et al. (2000) "LNCaP progression model of human prostate cancer: androgen-independence and osseous metastasis" Prostate 44: 91 -103

[961] Tohgo, A. et al. (2002) "β-Arrestin scaffolding of the Erk cascade enhances cytosolic Erk activity but inhibits ERK-mediated transcription following angiotensin AT1 a receptor stimulation" J. Biol. Chem. 277: 9429-9436

[962] Traish, A.M. and Morgentaler, A. (2009) "Epidermal growth factor receptor expression escapes androgen regulation in prostate cancer: a potential molecular switch for tumour growth" Br. J. Cancer 101 : 1949-1956

[963] Trauger, R. et al. (2009) "Inhibition of androstenediol-dependent LNCaP tumour growth by 17a-ethynyl-5a-androstane-3a, 17β- ίοΙ (HE3235)" Br. J. Cancer 100: 1068- 1072

[964] Tsihlias, J. et al. (2000) "Involvement of p27 ^Kip1 in d arrest by high dose 5a- dihydrotestosterone in LNCaP human prostate cancer cells" Oncogene 19: 670-679

[965] Ueda, T. et al. (2002) "Ligand-independent activation of androgen receptor by IL-6 and the role of the co-activator SRC-1 ", J. Biol. Chem. 277: 38087-94 Docket No. 354. PATENT

[966] Umemori, H. et al. (1997) "Activation of the G protein G _q/ n through tyrosine phosphorylation of the a subunit" Science 276: 1878-1 181

[967] Vantaggiato, C. et al. (2006) "ERK1 and ERK2 mitogen-activated protein kinases affect Ras-dependent cell signaling differently" J. Biol. 5-14

[968] Walls, R. et al. (1996) "The differentiating agent phenylacetate increases prostate- specific antigen production by prostate cancer cells" Prostate 29: 177-82

[969] Walsh, P.C. and Wilson, J.D. (1976) "The induction of prostatic hypertrophy in the dog with androstanediol" J. Clin. Invest. 57: 1093-1097

[970] Wang, J.G. (2004) "A mouse model for colitis-associated colon carcinogenesis induced by 1 ,2-dimethylhydrazine and dextran sulfate sodium" World J. Gastroenterol. 10: 2958-2962

[971] Wang, Z. et al. (2008) "Rapid membrane effect of testosterone in LNCaP cells" Urol. Int. 81 (3): 353-9

[972] Wellendorph, P. et al. (2005) "Deorphanization of GPRC6A: a promiscuous L-a- amino acid receptor with preference for basic amino acids" Mol. Pharmacol. 67:589-597

[973] Wen, Y. et al. (2000) "Her2/neu promotes androgen-independent survival and growth of prostate cancer cells through the Akt pathway" Cancer Res. 60: 6841 -45

[974] Westfall, P. H. et al. (1999) "Multiple comparisons and multiple tests" SAS Institute Inc., Cary, NC, USA, pp. 29-30

[975] Wilson, V. et al. (2002) "A novel cell line, MDA-kb2, that stably expresses an androgen- and glucocorticoid-responsive reporter for the detection of hormone receptor agonists and antagonists" Toxicol. Sci. 66, 69-81

[976] Wilson, V., et al. (2004) "Development and characterization of a cell line that stably expresses an estrogen-responsive luciferase reporter for the detection of estrogen receptor agonists and antagonists" Toxicol. Sci. 81 , 69-77

[977] Yang, Q. et al. (2008) "5a-androstane-3a,173-diol supports human prostate cancer cell survival through androgen receptor-independent signaling pathways:

implication of androgen-independent prostate cancer progression" J. Cell Biochem. 104: 1612-24

[978] Yashuda, T. (201 1 ) "ERKs induce expression of the transcriptional repressor Blimp-1 and subsequent plasma cell differentiation" Sci. Signal. 4(169):ra25

[979] Ye, L. et al. (2009) "Bone morphogenetic proteins in development and progression Docket No. 354. PATENT of breast cancer and therapeutic potential" Int. J. Mol. Med. 24:591 -7

[980] Yeh, S. et al. (1999) "From HER2/neu signal cascade to androgen receptor and its co-activators: a novel pathway by induction of androgen target genes through MAP kinase in prostate cancer cells" Proc. Natl Acad. Sci. (USA) 96: 5458-63

[981] Yoon, S. and Seger, R. (2006) "The extracellular-regulated kinase: multiple substrates regulate diverse cellular functions" Growth Factors 24: 21 -44

[982] Zhu, X., et al. (1999) "Androgen stimulates mitogen-activated protein kinase in human breast cancer cell" Mol. Cell. Endocr. 152: 199-206

[983] Zimmerman, R.A. et al. (2004) "5a-Androstane-3a,173-diol activates pathway that resembles the epidermal growth factor responsive pathways in stimulating LNCaP cell proliferation" Prostate Cancer Prostatic Disease 7: 364-374

[984] Zywietz, A. et al. (2001 ) "Pleiotropic effects of pasteurella multocida toxin are mediated by Gq-dependent and -independent mechanisms" J. Biol. Chem. 276: 3840- 3845