BACKGROUND OF THE INVENTION Field of the Invention The present invention is in the fields of molecular biology, cancer biology and medical therapeutics. The invention is generally directed to the construction of Pak kinase genes, and to polypeptides encoded by such genes. More specifically, the invention relates to isolated Pak kinase nucleic acid molecules and polypeptides, particu- larly those wherein the Pak kinases comprise a kinase domain and wherein the Pak kinases are substantially catalytically inactive. The invention also relates to methods of inhibiting animal cell transform- ation (particularly that mediated through the activity of the Ras oncogene), and to methods of treating or preventing physical disorders, including neurofibromatosis and cancers, in animals using the Pak kinase genes of the invention. The invention also relates to methods of identifying compounds that modulate the activity of Pak kinases and compoounds identified by these methods, and methods of identifying targets acted on by Pak kinases.

Related Art Mutations in the small G protein Ras are found in about 20% of tumors, making it one of the oncogenes most frequently associated with human cancers. Ras plays a key role in regulating cellular proliferation and differentiation. This is accomplished by mediating at least two different pathways. The first is the stimulation

of mitogen-activated protein kinase (MAPK) cascades which convey signals from the plasma membrane to the nucleus to regulate trans- cription (Barbacid, M., Ann. Rev. Biochem. 56:779-827 (1987); Ginty, D.D., et al., Cell 77:713-725 (1994); Lowy, D.R. and Willumsen, B.R., Ann. Rev. Biochem. 62:851-891 (1993); Marshall, C.J., Cell 80:179-185 (1995)). The second pathway involves regulation of the actin cytoskeleton and results in membrane ruffling (Bar-Sagi, D. and Feramisco, J.R., Science 233:1061-1068 (1986); Hall, A., Annu. Rev.

Cell Biol. 10:31-54 (1994)).

Components of the MAP kinase cascade are well characterized. Growth factor receptors recruit the Ras guanine nucleotide exchange factor, SOS, to the membrane which then activates Ras via nucleotide exchange. Once activated, Ras binds and activates the Raf-1 protein kinase which in turn phosphorylates and activates the MEK kinases, MEK1 and MEK2. The MEK kinases phosphorylate and activate the MAP kinases ERK1 and ERK2, which then phosphorylate and activate transcription factors leading to immediate-early gene expression (Egan, S.E. and Weinberg, R.A., Nature 365:781-783 (1993); Hill, C.S. and Triesman, R., Cell 80:199-211 (1995); Marshall, C.J., Cell 80:179-185 (1995)). Oncogenic Ras differs from wild-type Ras by point mutations that reduce its intrinsic GTPase activity. This causes the mutant Ras to be predominantly GTP-bound, hence activating it without growth factor stimulation of nucleotide exchange (Boguski, M.S. and McCormick, F., Nature 366:643-654 (1993)).

Ras regulation of the actin cytoskeleton does not require the interaction with Raf, since mutant Ras proteins that fail to interact with Raf still induce cytoskeletal changes (Joneson, T., et al., Science 271:810- 812 (1996)). Though the cytoskeletal pathways have not been fully defined, they require the coordinated action of Ras-related small G proteins from the Rho family, Cdc42, Rac (Racl and Rac2), and Rho (RhoA, RhoB and RhoC). Cdc42, Rac, and Rho each induce specific actin structures when microinjected into Swiss 3T3 fibroblasts: Cdc42 induces microspikes and filopodia, Rac causes membrane ruffling and Rho induces stress fibers and focal adhesions. Microinjection of Ras

protein induces membrane ruffling, and the ruffling is blocked by dominant negative mutations in Rac, indicating that Ras regulation of the actin cytoskeleton is mediated by Rac. The actin cytoskeletal events can be further ordered into a cascade of Cdc42 activating Rac followed by Rac activating Rho (Nobes, C.D. and Hall, A., Cell 81:53-62 (1995)).

Members of the Rho family also regulate transcription through another MAP kinase cascade similar to the ERK kinase cascade (Coso, O.A., et al., Cell 81:1137-1146 (1995); Minden, A., et al., Cell 81:1147-1157 (1995); Minden, A., et al. Science 266:1719-1723 (1994)).

Rac and Cdc42 bind and activate a protein kinase called Pak, which then activates a cascade that has not been completely defined but is likely to consist of MEKK, SEK and the Jun kinase (JNK) or the related p38 kinase (Bagrodia, S., et al., J. Biol. Chem. 270:27995-27998 (1995); Frost, J.A., et al., Mol. Cell. Biol 16:3707-3713 (1996); Manser, E., et al., Nature 367:40-46 (1994); Polverino, A., et al., J. Biol. Chem. 270:26067-26070 (1995); Yan, M., et al., Nature 372:798-800 (1994)). Jun kinase phosphorylates transcription factors such as c-jun. In many cell lines Ras also activates JNK (Hibi, M., et al., Genes Dev 7:2135-2148 (1993)).

Ras activation of JNK is inhibited by dominant negative mutations of Rac and Cdc42 suggesting that Rac and Cdc42 mediate Ras activation of JNK (Minden, A., et al., Cell 81:1147-1157 (1995)).

Several lines of research suggest that Rac and Rho are essential for Ras transformation. Dominant negative mutations of Rac inhibit Ras transformation and GTPase deficient Rac and Rho can both weakly transform fibroblasts. Furthermore, both activated Rac and Rho can dramatically stimulate transformation by partially activated Raf mutants (Prendergast, G.C., et al., Oncogene 10:2289-2296 (1995); Qui, R.G., et al., Nature 374:457-459 (1995)). Several oncogenes including Ost, Ect2, Dbl and Tiam-1 are guanine nucleotide exchange factors specific for Rho family members (Quilliam, L.A., et al., BioEssays 17:395-404 (1995)). These observations demonstrate that, although the Rho family members are not often found activated in tumors, they can transform cells and in many cases cooperate with the Ras/Raf signaling pathway to transform cells.

Members of the Pak family of protein kinases are regulated by GTP-bound Rac and Cdc42 and are candidates for effectors that mediate both actin and JNK signaling (Lim, L., et al., Eur. J. Biochem.

242:171-185 (1996); Manser, E., et al., Nature 367:40-46 (1994); Sells, M.A.

and Chernoff, J., Cell. Biol. 7:162-167 (1997)). Three Pak kinases have been found in mammals: Pakl, Pak2 and Pak3. All are related to the STE20 gene of the yeast Saccharomyces cerevisiae, which is regulated by a Cdc42 homolog (Bagrodia, S., et al., J. Biol. Chem. 270:22731- 22737 (1995); Leberer, E., et al., EMBO J. 11:4815-4824 (1992); Marcus, S., et al., Proc. Nat. Acad. Sci. (USA) 92:6180-6184 (1995); Ottilie, S., et al., EMBO J 14:5908-5919 (1995); Ramer, S.W. and Davis, R.W., Proc. Nat. Acad. Sci.

(USA) 90:452-456 (1993)). GTP-bound Rac and Cdc42 both stimulate kinase activity through direct binding to a conserved region near the N terminus of Pak called the p21 binding domain (PBD). Regions homologous to the PBD domain are found in other proteins that bind Rac/Cdc42 in vitro, such as Ste20 (Burbelo, P.D., et al., J. Biol. Chem.

270:29071-29074 (1995)).

SUMMARY OF THE INVENTION The present invention relates generally to Pak kinase genes, and to polypeptides encoded by such genes. More particularly, the invention provides isolated nucleic acid molecules comprising a polynucleotide encoding a mutant Pak kinase, which may be a mutant Pakl kinase, a mutant Pak2 kinase or a mutant Pak3 kinase, most particularly wherein the Pak kinase comprises a kinase domain and wherein the Pak kinase is substantially catalytically inactive.

Preferred such nucleic acid molecules encode a Pak kinase that comprises one or more mutations in its kinase domain, which may be deletions, substitutions or insertions of one or more nucleotides.

Particularly preferred nucleic acid molecules are those wherein the mutation occurs in one or more codons encoding one or more amino acid residues within a span of amino acid residues from about amino acid residue 260 to about amino acid residue 520, from about amino acid residue 270 to about amino acid residue 516, from about amino acid

residue 290 to about amino acid residue 400, from about amino acid residue 295 to about amino acid residue 350, from about amino acid residue 297 to about amino acid residue 300, from about amino acid residue 299 to about amino acid residue 300, or in a codon encoding amino acid residue 299, of the mutant Pak kinase. In one such embodiment, the mutation is a substitution of an arginine residue in place of a lysine residue at amino acid residue 299 in the mutant Pak kinase; particularly preferred examples include the mutant Pak kinases designated PaklR299 and PaklL83,L86,R299. The invention also provides nucleic acid molecules comprising a polynucleotide which hybridizes under stringent hybridization conditions to a polynucleotide having a nucleotide sequence identical to the nucleotide sequence of the above-described isolated nucleic acid molecules.

The invention also relates to vectors (particularly expression vectors) and host cells comprising the above-described nucleic acid molecules.

The invention also relates to methods for producing an isolated Pak kinase polypeptide, comprising culturing the above- described host cells under conditions sufficient to allow the expression of said polypeptide, and isolating the Pak kinase polypeptide. The invention also relates to Pak kinase polypeptides produced according to these methods, and to isolated Pak kinase polypeptides which comprise a kinase domain and which are substantially catalytically inactive.

Preferred such nucleic acid molecules encode a Pak kinase that comprises one or more mutations in its kinase domain, which may be deletions, substitutions or insertions of one or more nucleotides.

Particularly preferred Pak kinase polypeptides are those wherein the mutation occurs in one or more amino acid residues within a span of amino acid residues from about amino acid residue 260 to about amino acid residue 520, from about amino acid residue 270 to about amino acid residue 516, from about amino acid residue 290 to about amino acid residue 400, from about amino acid residue 295 to about amino acid residue 350, from about amino acid residue 297 to about amino acid

residue 300, from about amino acid residue 299 to about amino acid residue 300, or in amino acid residue 299, of the mutant Pak kinase.

In one such embodiment, the mutation is a substitution of an arginine residue in place of a lysine residue at amino acid residue 299 in the mutant Pak kinase polypeptide; particularly preferred examples include the mutant Pak kinase polypeptides designated Pak1R299 and paklL83,L86,R299 The invention also relates to methods of producing isolated antibodies that specifically bind to a Pak kinase polypeptide, the methods comprising immunizing an animal with one or more of the above- described isolated Pak kinase polypeptides and isolating, from the immunized animal, an antibody that specifically binds to a Pak kinase polypeptide. The invention also relates to isolated antibodies produced by these methods, which may be polyclonal or monoclonal antibodies, and which may be detectably labeled and/or immobilized on a solid support.

The invention is also directed to methods of inhibiting the activity of a Ras oncogene in a cell comprising a Ras oncogene, or of inhibiting the transformation of a cell, comprising introducing into the cell an effective amount of one or more of the above-described isolated nucleic acid molecules. According to the invention, the isolated nucleic acid molecules used in these methods may be contained in a vector or a virion which may be derived from a retrovirus, an adenovirus or an adeno-associated virus. Preferred cells for use with these methods include animal cells, such as mammalian cells and particularly human cells. Other preferred cells include cancer cells such as Ras-transformed cancer cells.

The invention also provides pharmaceutical compositions comprising one or more of the above-described isolated nucleic acid molecules and a pharmaceutically acceptable carrier or excipient therefor.

The invention further relates to methods of identifying compounds that modulate the activity of Pak kinases, and compounds identified by these methods. The invention further relates to pharma- ceutical compositions comprising one or more such compounds and

a pharmaceutically acceptable carrier or excipient therefor.

The invention also provides methods of identifying targets acted on by Pak kinases.

The invention is also directed to therapeutic methods for treating or preventing a physical disorder, such as a cancer or a neurological disorder, in an animal comprising introducing into the animal an effective amount of one or more of the above-described isolated nucleic acid molecules, or administering to the animal an effective amount of one or more of the above-described pharmaceutical composi- tions. Animals that may be treated by the present therapeutic methods include mammals, particularly humans. Cancers that may be treated or prevented by these methods include without limitation carcinomas, sarcomas (particularly neurofibrosarcomas), melanomas and leukemias, and most particularly Ras-associated cancers. Neurological disorders that may be treated or prevented by these methods include without limitation neurofibromatosis.

Other preferred embodiments of the present invention will be apparent to one of ordinary skill in light of the following drawings and description of the invention, and of the claims.

DESCRIPTION OF THE DRAWINGS Figures lA-1C Characterization of the Pak mutants used in this study.

Myc tagged Pakl and the indicated mutants were transfected into COS cells and then assayed for kinase activity and Rac/Cdc42 binding. (A) Top, kinase activity of different Pak mutants. Activity was measured by immunoprecipitating with the anti-myc antibody 9E10 and incubating precipitates with, where indicated, GST-Cdc42 or GST Racl using myelin basic protein as a substrate. Bottom, map of Pakl showing p21 binding domain (PBD), kinase domain and mutations used in this study.

(B) Cdc42 and Rac binding. Cdc42 and Rac binding to Pakl and PaklR299 was measured by mixing 50 ,ug of extract from the transfected cells as indicated with -50CLg of purified GST Cdc42 or GST Rac bound either to

GTP or GDP- S and glutathione beads. (C) Cdc42/Rac binding to PBD domain mutants. Similar results were obtained in 3 independent experiments. The same extracts were used in parts A, B, and C.

Figure 2A-2D (A) Effect of Pakl and PaklR299 on K-ras transformation in focus assays. Rat-1 fibroblasts were transfected with 10 Rg of the indicated plasmids and stained with crystal violet to visualize foci as described in Materials and Methods. (B) Effect of Pakl and PaklR299 on Raf transformation. Similar results were obtained in more than 10 (A) or 3 (B) independent experiments. (C) Effect of PakR299 on Ras EJ transformation of NIH 3T3 cells. Cells were transfected with, where indicated, 5 ng of Ras and 2.5 clog of PakR299. Total DNA concentration was brought to 13,ug with carrier DNA (D) Western blots of Pak, Raf and Ras expression. Rat-1 cells were co-transfected with K-ras or v-raf and the various Pakl expression plasmids. Extracts were prepared and 50 ,ug of each was run on a 12% gel and tested on a Western blot probed with anti-K-ras (antibody F234 from Santa Cruz Biotechnology, Inc., Santa Cruz, CA), anti-Pak (the anti-myc epitope antibody 9E10) or anti Raf-1 (antibody C12 from Santa Cruz Biotechnology, Inc.). Detection was performed by enhanced chemiluminescence. Similar results were obtained in two independent experiments.

Figures 3A-3G Effect of Pakl and PaklR299 on K-ras transformation in soft agar assays. Cells were transfected as described in FIG. 1 and then plated on soft agar. (A-F) Representative microscopic views of colonies.

(G) Quantification of the soft agar assays. Similar results were seen in more than 5 independent experiments.

Figures 4A-4C Stable expression of Pakl in Rat-l cells. Rat-1 cells were co-transfected with pCDNA and the various Pakl expression plasmids.

G418 resistant colonies were isolated, expanded into cell lines and tested

as described below. (A) Western blot of 10 Rl of extracts from Rat-1 cells probed with antibody 9E10 which recognizes the Myc tag on the Pakl constructs. Pakl is seen as a 65 kDa band. (B) Growth rates of stable cell lines and an H-ras transformed cell line. (C) Growth of cells in 1% serum. Similar results were obtained in more than 2 independent experiments.

Figure 5 Micrographs of Rat-1 cells expressing Pakl and Pakl mutants.

Figures 6A-6C Ras transformation of stable cell lines. The indicated quantities of K-ras were transfected into the stable cell lines and transformation was scored in (A) focus assays and (B) soft agar colony assays. (C) Raf transformation of Rat-1 cells expressing Pakl mutants.

Colonies were counted if they were -50 CLM or larger for Ras transform- ations or -20 piM or larger for Raf transformations. No differences in transfection efficiencies were detected between the stable cell lines and the parent Rat-1 cells. Similar results were obtained in more than 3 independent experiments.

Figures 7A-7D Effect of Pak mutants on JNK1 and p38 activation. Rat-1 cells were cotransfected with either HA-JNK (A, B and C) or HA-p38 (D) and the plasmids encoding the proteins shown. Fold refers to the fold increase in substrate phosphorylation over that occurring in the Pcmv6 lane, as determined through phosphoimager analysis. Below each panel is a panel is a Western blot showing expression of HA-JNK or HA-p38, as indicated.

Figure 8 Effect of Pak mutants on ERK1 activation. Rat-1 cells were cotransfected with HA-ERK and the test DNA. Fold refers to the fold

increase in substrate phosphorylation over that occurring in the pCMV6 lane as determined through phosphoimager analysis. Below each panel is a Western blot showing expression of HA-ERK.

Figure 9 Photographs of soft agar colony formation by ST88-14 neurofibrosarcoma cells transfected with Pak kinases. The Pakl kinase mutants PaklR299 pak1L83 L86 and PaklL83,L86,R299 are seen to reverse the transformation of ST88-14 cells.

Figure 10 Schwann cells were transfected with the indicated plasmids, plated on soft agar and then stained with thiazolyl blue [3-(4,5-dimethylthiazol-2-yl)-2 , 5-diphenyl tetrazolium bromide; Sigma] to visualize colonies. The data is presented as number of colonies per plate. The error bars indicate standard deviations. Data shown were representative of four independent experiments.

Figures 11A and 11B Effect of Pak on activation of JNK and ERK. Schwann cells were cotransfected with either JNK or ERK and the plasmids encoding the proteins shown. Fold indicates fold increase in substrate phosphoryl- ation over that occuring in the pcmv6 lanes, as determined though phosphorimager analysis. At the bottom of each panel is a Western blot showing expresstion of HA-JNK or HA-ERK, as indicated. MBP, myelin basic protein. Data shown were representative gels of three independent experiments.

DETAILED DESCRIPTION OF THE INVENTION Nucleic Acid Molecules Unless otherwise indicated, all nucleotide sequences determined by sequencing a DNA molecule herein were determined using manual DNA sequencing such as dideoxy sequencing, according to methods that are routine to one of ordinary skill in the art (Sanger,

F., and Coulson, A.R., J. Mol. Biol. 94:444-448 (1975); Sanger, F., et al., Proc. Natl. Acad. Sci. USA 74:5463-5467 (1977)), or by automated sequencing such as by using an Applied Biosystems Automated Sequenator according to the manufacturer's instructions. All amino acid sequences of polypeptides encoded by DNA molecules determined herein were predicted by conceptual translation of a DNA sequence determined as above. Therefore, as is known in the art for any DNA sequence determined by these approaches, any nucleotide sequence determined herein may contain some errors. Nucleotide sequences determined by such methods are typically at least about 90% identical, more typically at least about 95% to at least about 99.9% identical to the actual nucleotide sequence of the sequenced DNA molecule. As is also known in the art, a single insertion or deletion in a determined nucleotide sequence compared to the actual sequence will cause a frame shift in translation of the nucleotide sequence such that the predicted amino acid sequence encoded by a determined nucleotide sequence will be completely different from the amino acid sequence actually encoded by the sequenced DNA molecule, beginning at the point of such an insertion or deletion.

Unless otherwise indicated, each "nucleotide sequence" set forth herein is presented as a sequence of deoxyribonucleotides (abbreviated a, G, C and T). However, by "nucleotide sequence" of a nucleic acid molecule or polynucleotide is intended, for a DNA molecule or polynucleotide, a sequence of deoxyribonucleotides, and for an RNA molecule or polynucleotide, the corresponding sequence of ribonucleo- tides (a, G, C and U), where each thymidine deoxyribonucleotide (T) in the specified deoxyribonucleotide sequence is replaced by the ribonucleo- tide uridine (U). For instance, reference to a mutant Pak kinase RNA molecule having a nucleotide sequence set forth using deoxyribonucleo- tide abbreviations is intended to indicate an RNA molecule having a sequence in which each deoxyribonucleotide a, G or C of the nucleotide sequence has been replaced by the corresponding ribonucleotide a, G or C, and each deoxyribonucleotide T has been replaced by a ribonucleotide U.

The nucleotide sequences of wildtype Pak kinases are known in the art (see, e.g., Osada, S., et al., Mol. Cell. Biol. 12:3930-3938 (1992); Bagrodia, S., et al., J. Biol. Chem. 270:22731-22737 (1995); Burbelo, P.D., et al., J. Biol. Chem. 270:29071-29074 (1995); Manser, E., et al., Nature 367:40-46 (1994); Sells, M.A., et al., Curr. Biol. 7:202-210 (1997); Martin, G.S., et al., EMBO J. 14:1970-1978 (1995); Brown, J.L., Curr.

Biol. 6:598-605 (1996), the disclosures of which are incorporated herein by reference in their entireties). Using these sequences and the informa- tion provided herein, a nucleic acid molecule of the present invention encoding a Pak kinase polypeptide, which may be a mutant Pak kinase, a fragment of a mutant Pak kinase or a fragment of a wild type Pak kinase, may be obtained using standard cloning and screening procedures, such as those for cloning cDNAs using mRNA as starting material. As used herein, a "mutant Pak kinase polypeptide" means a polypeptide or fragment thereof that is encoded by a polynucleotide comprising one or more mutations in its nucleotide sequence from the nucleotide sequence of the corresponding wildtype, catalytically active, Pak kinase. For example, the mutant Pak kinase designated PaklR299 herein is a mutant Pakl kinase in which the lysine residue at position 299 of the wildtype Pakl kinase has been replaced by an arginine residue; the corresponding nucleic acid molecule encoding PaklR299 therefore comprises a codon that encodes arginine at position 299 in the polypeptide instead of lysine. Similarly, the mutant Pak kinase designated PaklLS3,L86 herein is a mutant Pakl kinase in which the histidine residues at positions 83 and 86 in the wildtype Pakl kinase have both been replaced with leucine residues; the corresponding nucleic acid molecule encoding PaklL83,L86 therefore comprises codons that encode leucine at positions 83 and 86 in the polypeptide instead of histidine. The Pak kinase nucleic acid molecules and polypeptides of the invention may therefore be analogues or mutants of any member of the Pak kinase family, including but not limited to Pakl kinase, Pak2 kinase and Pak3 kinase. Preferred cloning and screening methods used in the invention include PCR-based cloning methods, such as reverse

transcriptase-PCR (RT-PCR) using primers such as those described in the Examples below.

Nucleic acid molecules of the present invention may be in the form of RNA, such as mRNA, or in the form of DNA, including, for instance, cDNA and genomic DNA obtained by cloning or produced synthetically. The DNA may be double-stranded or single-stranded.

Single-stranded DNA or RNA may be the coding strand, also known as the sense strand, or it may be the non-coding strand, also referred to as the antisense strand.

By "isolated" nucleic acid molecule(s) is intended a nucleic acid molecule, DNA or RNA, which has been removed from its native environment. For example, recombinant DNA molecules contained in a vector are considered isolated for the purposes of the present invention.

Further examples of isolated DNA molecules include recombinant DNA molecules maintained in heterologous host cells, and those DNA molecules purified (partially or substantially) from a solution whether produced by recombinant DNA or synthetic chemistry techniques.

Isolated RNA molecules include in vivo or in vitro RNA transcripts of the DNA molecules of the present invention.

The nucleic acid molecules of the present invention further include genetic constructs comprising one or more Pak kinase DNA sequences operably linked to regulatory DNA sequences (which may be heterologous regulatory sequences), such as promoters or enhancers as described below, wherein upon expression of these DNA sequences in host cells, preferably in bacterial, fungal (including yeast), plant or animal (including insect or mammalian) cells, one or more Pak kinase polypeptides are produced. In such constructs, the regulatory sequences may be operably linked to a polynucleotide encoding a Pak kinase polypeptide or any of its variants, precursors, fragments or derivatives described herein. As used herein, the terms "a portion" or "a fragment" of a nucleic acid molecule or a polypeptide means a segment of a polynucleotide or a polypeptide comprising at least 15, and more preferably at least 20, contiguous nucleotides or amino acids of a

reference polynucleotide or polypeptide, unless otherwise specifically defined below.

Isolated nucleic acid molecules of the present invention include nucleic acid molecules comprising a polynucleotide encoding a Pak kinase polypeptide wherein the Pak kinase polypeptide comprises a kinase domain and wherein the Pak kinase polypeptide is substantially catalytically inactive. As used herein, the term "kinase domain" refers to the portion of the Pak kinase enzyme that is involved in catalyzing phosphorylation of the Pak kinase targets (such as myelin basic protein); in wildtype Pakl, for example, the kinase domain is located at approxi- mately amino acid residues 270 to 516 of the polypeptide (see Figure 1A).

As used herein, the term "substantially catalytically inactive" means that the Pak kinase or mutant Pak kinase phosphorylates a Pak kinase target (such as myelin basic protein) to a level not greater than about 50%, 40%, 30%, 20%, 10%, 5%, 1%, 0.1% or 0.01%, of the level of phosphorylation catalyzed by a wildtype Pak kinase polypeptide such as the wildtypes of Pakl kinase, Pak2 kinase or Pak3 kinase. In practice, whether a Pak kinase or mutant Pak kinase phosphorylates a Pak kinase target to a level not greater than about 50%, 40%, 30%, 20%, 10%, 5%, 1%, 0.1% or 0.01%, of the level of phosphorylation catalyzed by a wildtype Pak kinase polypeptide may be determined by any number of phosphorylation assays which will be familiar to one of ordinary skill, such as those described in detail in the Examples below.

Preferred nucleic acid molecules of the invention encode a substantially catalytically inactive Pak kinase. Such nucleic acid molecules may, for example, encode a mutant Pak kinase comprising one or more mutations in its kinase domain, which may be deletions, substitutions or insertions of one or more nucleotides and particularly substitutions or insertions of one or more nucleotides. Particularly preferred nucleic acid molecules are those wherein the mutation occurs in one or more codons encoding one or more amino acid residues within the kinase domain which spans amino acid residues about 270 to about 516 (see Figure 1A), including those wherein the mutation occurs in one or more codons encoding a span of amino acid residues from about

amino acid residue 260 to about amino acid residue 520, from about amino acid residue 270 to about amino acid residue 516, from about amino acid residue 290 to about amino acid residue 400, from about amino acid residue 295 to about amino acid residue 350, from about amino acid residue 297 to about amino acid residue 300, from about amino acid residue 299 to about amino acid residue 300, or in a codon encoding amino acid residue 299, of the mutant Pak kinase. More particularly preferred are those isolated nucleic acid molecules wherein a first mutation is a substitution mutation for the lysine residue at position 299 (L299) of the wildtype Pak kinase polypeptide. Preferred such substitutions include replacement of L299 with a basic amino acid such as arginine or histidine. In a particularly preferred such embodi- ment, the mutation is a substitution of an arginine residue in place of L299 in the Pak kinase, an example of which is the mutant Pakl kinase designated PaklR299. Other mutations may, of course, be made within the kinase domain or other domains (such as the Rac- and Cdc42- binding domain (PBD) located at approximately amino acid residues 75 to 132 of the polypeptide; see Figure 1A), provided that they result in a mutant Pak kinase that is substantially catalytically inactive as described above. For example, the present invention also provides a mutant Pakl kinase wherein the histidine residues at amino acids 83 and 86 in the PBD of wildtype Pakl kinase have been replaced by leucine residues, and the L299 has been replaced by an arginine as described above, resulting in a Pakl kinase mutant designated PaklL83,L86,R299 which comprises the kinase domain and is substantially catalytically inactive. Methods for producing such mutations in the kinase and/or other domains of the Pak kinase, to generate the mutant Pak kinase nucleic acid molecules and polypeptides of the invention, include site-directed mutagenesis (Sambrook, J., et al., Molecular Cloning: a Laboratory Manual, 2nd Ed., Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press (1989)) and site-elimination mutagenesis as described in detail in the Examples below and in Deng, W.P., and Nickoloff, J.A., Anal. Biochem. 200:81-88 (1992), as well as other methods that will be familiar to one of ordinary skill.

The present invention also encompasses isolated nucleic acid molecules which comprise a sequence substantially different from those described above but which, due to the degeneracy of the genetic code, still encode a substantially catalytically inactive Pak kinase polypeptide that comprises a kinase domain. Since the genetic code is well known in the art, it is routine for one of ordinary skill in the art to produce the degenerate variants described above without undue experimentation.

Nucleic acid molecules of the present invention which encode a Pak kinase polypeptide may include, but are not limited to, those encoding the amino acid sequence of the mature polypeptide by itself; the coding sequence for the mature polypeptide and additional coding sequences, such as those encoding a leader or secretory sequence, such as a pre-, or pro- or prepro- protein sequence; the coding sequence of the mature polypeptide, with or without the aforementioned additional coding sequences, together with additional, non-coding sequences, including for example introns and non-coding 5' and 3' sequences, such as the transcribed, untranslated regions (UTRs) or other 5' flanking sequences that may play a role in transcription (e.g., via providing ribosome- or transcription factor- binding sites), mRNA processing (e.g. splicing and polyadenylation signals) and stability of mRNA; the coding sequence for the mature Pak kinase polypeptide operably linked to a regulatory DNA sequence, particularly a hetero- logous regulatory DNA sequence such as a promoter or enhancer; and the coding sequence for the mature Pak kinase polypeptide linked to one or more coding sequences which code for amino acids that provide additional functionalities. Thus, the sequence encoding the polypeptide may be fused to a marker sequence, such as a sequence encoding a peptide which facilitates purification of the fused polypeptide. In certain embodiments of this aspect of the invention, the marker amino acid sequence may be a hexa-histidine peptide, such as the tag provided in a pQE vector (Qiagen, Inc.), among others, many of which are commercially available. As described for instance in Gentz et al.,

Proc. Natl. Acad. Sci. USA 86:821-824 (1989), hexa-histidine provides for convenient purification of the fusion protein. The "HA" tag is another peptide useful for purification which corresponds to an epitope derived from the influenza hemagglutinin protein, which has been described by Wilson et al., Cell 37: 767 (1984). Yet another useful marker peptide for facilitation of purification of a Pak kinase is glutathione S-transferase (GST) encoded by the pGEX fusion vector (see, e.g., Winnacker, From Genes to Clones, New York: VCH Publishers, pp. 451-481 (1987)). As discussed below, other such fusion proteins include the Pak kinase fused to immunoglobulin Fc at the N- or C-terminus.

The present invention further relates to variants of the nucleic acid molecules of the present invention, which encode portions, analogs or derivatives of the Pak kinase polypeptides of the invention.

Variants may occur naturally, such as a natural allelic variant. By an "allelic variant" is intended one of several alternate forms of a gene occupying a given locus on a chromosome of an organism (see Lewin, B., Ed., Genes II, John Wiley & Sons, New York (1985)). Non-naturally occurring variants may be produced using art-known mutagenesis techniques.

Such variants include those produced by nucleotide substitutions, deletions or additions. The substitutions, deletions or additions may involve one or more nucleotides. The variants may be altered in coding regions, non-coding regions, or both. Alterations in the coding regions may produce conservative or non-conservative amino acid substitutions, deletions or additions. Especially preferred among these are silent substitutions, additions and deletions, which do not alter the structure, properties and activities of the Pak kinase protein or portions thereof. Also especially preferred in this regard are conservative substitutions.

As a practical matter, whether any particular nucleic acid molecule is identical to, for instance, the nucleotide sequence of one or more of the isolated nucleic acid molecules of the invention can be determined conventionally using known computer programs such as FASTA (Heidelberg, Germany), BLAST (Washington, DC) or BESTFIT

(Wisconsin Sequence Analysis Package, Version 8 for Unix, Genetics Computer Group, University Research Park, 575 Science Drive, Madison, WI 53711), which employs a local homology algorithm (Smith and Waterman, Advances in Applied Mathematics 2: 482-489 (1981)) to find the best segment of homology between two sequences. When using FASTA, BLAST, BESTFIT or any other sequence alignment program to determine whether a particular sequence is, for instance, identical to an isolated nucleic acid molecule according to the present invention, the parameters are set such that the percentage of identity is calculated over the full length of the reference nucleotide sequence.

In another aspect, the invention provides an isolated nucleic acid molecule comprising a polynucleotide which hybridizes under stringent hybridization conditions to substantially all of the polynucleotide in the isolated nucleic acid molecules of the invention described above, or to a fragment thereof comprising the regions of mutated nucleotides described above. By "stringent hybridization conditions" is intended overnight incubation at 42"C in a solution comprising: 50% formamide, 5x SSC (1X SSC = 150 mM NaCl, 15mM trisodium citrate), 50 mM sodium phosphate (pH 7.6), 5x Denhardt's solution, 10% dextran sulfate, and 20 g/ml denatured, sheared salmon sperm DNA, followed by washing the filters in 0.1x SSC at about 65"C.

By a polynucleotide which hybridizes to a "fragment" of a polynucleotide is intended a polynucleotide (either DNA or RNA) hybridizing to at least about 10 nucleotides, more preferably at least about 15 or 20 nucleotides, still more preferably at least about 30 nucleotides, and even more preferably about 30-50 nucleotides of the reference polynucleotide. These hybridizing polynucleotides are useful as diagnostic probes according to conventional DNA hybridization techniques or as primers for amplification of a target sequence by the polymerase chain reaction (PCR), as described, for instance, in Sambrook, J., et al., Molecular Cloning: a Laboratory Manual, 2nd Ed., Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press (1989), the entire disclosure of which is ncorporated herein by reference.

Vectors and Host Cells The present invention also relates to genetic constructs comprising the isolated nucleic acid molecules of the invention, or fragments thereof, operably linked to regulatory DNA sequences as described in detail below, vectors which comprise these genetic constructs or the isolated DNA molecules of the present invention, and host cells which comprise these vectors. In addition, the invention relates to the production of Pak kinase polypeptides or fragments thereof by recombinant techniques using these vectors and host cells.

Vectors comprising the genetic constructs or the isolated DNA molecules or fragments of the present invention may be introduced into host cells using well-known techniques such as infection, trans- duction, transfection, electroporation and transformation. The vector may be, for example, a phage, plasmid, viral or retroviral vector, and is preferably an expression vector as described below. Retroviral vectors may be replication-competent or -defective. In the latter case, viral propagation generally will occur only in complementing host cells.

The polynucleotides may be joined to a vector containing a selectable marker for propagation in a host. Generally, a plasmid vector is introduced into mammalian or avian cells in a precipitate, such as a calcium phosphate precipitate, or in a complex with a charged lipid (e.g., LIPOFECTAMINETM; Life Technologies, Inc.; Rockville, Maryland) or in a complex with a virus (such as an adenovirus; see U.S. Patent Nos. 5,547,932 and 5,521,291) or components of a virus (such as viral capsid peptides). If the vector is a virus, it may be packaged in vitro using an appropriate packaging cell line and then transduced into host cells.

Preferred are vectors comprising cis-acting control regions to the polynucleotide of interest. Appropriate trans-acting factors may be supplied by the host, by a complementing vector or by the vector itself upon introduction into the host.

In certain preferred embodiments in this regard, the vectors provide for specific expression, which may be inducible and/or cell type-specific. Particularly preferred among such expression

vectors are those inducible by environmental factors that are easy to manipulate, such as temperature and nutrient additives.

Expression vectors useful in the present invention include chromosomal-, episomal- and virus-derived vectors, e.g., vectors derived from bacterial plasmids, bacteriophages, yeast episomes, yeast chromo- somal elements, viruses such as baculoviruses, papovaviruses, vaccinia viruses, adenoviruses, fowl pox viruses, pseudorabies viruses and retroviruses, and vectors derived from combinations thereof, such as cosmids and phagemids.

In one embodiment, an isolated nucleic acid molecule of the invention or fragment thereof may be operably linked to an appropriate regulatory sequence, preferably a promoter such as the phage lambda PL promoter, promoters from T3, T7 and SP6 phages, the E. coli lac, trp and tac promoters, the SV40 early and late promoters and promoters of retroviral LTRs and derivatives thereof, to name a few.

Other suitable promoters will be known to the skilled artisan. The expression constructs will further contain sites for transcription initiation, termination and, in the transcribed region, a ribosome binding site for translation. The coding portion of the mature transcripts expressed by the constructs will preferably include a translation initiation codon (AUG) at the beginning and a termination codon (UAA, UGA or UAG) appropriately positioned at the end of the polypeptide to be translated.

As indicated above, the expression vectors will preferably include at least one selectable marker. Such markers include dihydro- folate reductase (dhfr) or neomycin (neo) resistance for eukaryotic cell culture and tetracycline (tet) or ampicillin (amp) resistance genes for culturing in E. coli and other bacteria. Representative examples of appropriate hosts include, but are not limited to, bacterial cells, such as Escherichia spp. cells (particularly E. coli), Bacillus spp. cells (particularly B. cereus, B. subtilis and B. megaterium), Streptomyces spp. cells, Salmonella spp. cells (particularly S. typhimurium) and Xanthomonas spp. cells; fungal cells, including yeast cells such as Saccharomyces spp. cells; insect cells such as Drosophila S2, Spodoptera

Sf9 or Sf21 cells and Trichoplusa High- Five cells; other animal cells (particularly mammalian cells and most particularly human cells) such as CHO, COS, VERO, HeLa, Bowes melanoma cells and HepG2 and other liver cell lines; and higher plant cells. Appropriate culture media and conditions for the above-described host cells are known in the art.

Among vectors preferred for use in bacteria include pQE70, pQE60 and pQE-9, available from Qiagen; pBS vectors, Phagescript vectors, Bluescript vectors, pNH8A, pNH16a, pNH18A and pNH46A, available from Stratagene; pcDNA3 available from Invitrogen; and pGEX, pTrxfus, pTrc99a, pET-5, pET-9, pKK223-3, pKK233-3, pDR540 and pRITS available from Pharmacia. Among preferred eukaryotic vectors are pWLNEO, pSV2CAT, pOG44, pXT1, pBK and pSG available from Stratagene; and pSVK3, pBPV, pMSG and pSVL available from Pharmacia. Other suitable vectors will be readily apparent to the skilled artisan.

Among known bacterial promoters suitable for use in the present invention include the E. coli lacI and lacZ promoters, the T3, T7 and SP6 phage promoters, the gpt promoter, the lambda PR and PL promoters and the trp promoter. Suitable eukaryotic promoters include the CMV immediate early promoter, the HSV thymidine kinase promoter, the early and late SV40 promoters, the promoters of retroviral LTRs, such as those of the Rous sarcoma virus (RSV), and metallo- thionein promoters, such as the mouse metallothionein-I promoter.

Introduction of the construct into the host cell can be effected by calcium phosphate transfection, DEAE-dextran mediated transfection, cationic lipid-mediated transfection, electroporation, transduction, infection, nucleic acid-coated microprojectile bombard- ment or other methods. Such methods are described in many standard laboratory manuals, such as Davis et al., Basic Methods In Molecular Biology (1986).

In some embodiments, the isolated polynucleotides of

the present invention may be operably linked to a regulatory genetic sequence, which may be a homologous or a heterologous regulatory genetic sequence (such as an enhancer, promoter or repressor), to form a genetic construct. Genetic constructs according to this aspect of the invention are intended to encompass not only those comprising a polynucleotide encoding a Pak kinase polypeptide operably linked to a regulatory DNA sequence, but also those constructs comprising one or more regulatory sequences operably linked to a polynucleotide fragment which does not encode a Pak kinase, but which contains a sufficient portion of the Pak kinase nucleotide sequence (a "targeting fragment") to target the genetic construct to the native Pak kinase locus upon introduction into a host cell wherein the Pak kinase gene may be active due to oncogenic transformation of the host cell. These constructs may be inserted into a vector as above, and the vectors introduced into a host cell, the genome of which comprises the target gene, by any of the methods described above. The Pak kinase polynucleotide of the invention will then integrate into the host cell genome by homologous recombina- tion. In the case of a construct comprising a homologous or heterolo- gous regulatory sequence linked to a targeting Pak kinase polynucleotide fragment, the regulatory sequence will be targeted to the Pak kinase locus in the Ras-transformed host cell, and will repress or inhibit (if the regulatory sequence comprises, for example, a repressor or otherwise integrates into the native regulatory sequence to inhibit or repress (i.e., "knock out")) the expression of the catalytically active Pak kinase gene in the host cell, thereby decreasing the level of catalytically active Pak kinase gene expression. Alternatively, such gene targeting may be carried out using genetic constructs comprising the above-described Pak kinase targeting fragment in the absence of a regulatory sequence; such an approach may be used, for example, to correct or introduce point mutations in one or more Pak kinase genes (see Steeg, C.M., et al., Proc.

Natl. Acad. Sci. USA 87(12):4680-4684 (1990) for a description of the use of such approaches to correcting or introducing point mutations in other mammalian genes). Such methods of producing genetic constructs, introducing genes of interest into a host cell via homologous recombin-

ation and producing the encoded polypeptides are generally described in U.S. Patent No. 5,578,461; WO 94/12650 (U.S. Application No. 07/985,586); WO 93/09222 (U.S. Application No. 07/911,535); and WO 90/14092 (U.S.

Application No. 07/353,909), the disclosures of which are expressly incorporated herein by reference in their entireties.

Transcription of the DNA encoding the polypeptides of the present invention by higher eukaryotes may be increased by inserting an enhancer sequence into the vector. Enhancers are cis- acting elements of DNA, usually from about 10 to 300 bp, that act to increase transcriptional activity of a promoter in a given host cell-type.

Examples of enhancers include the SV40 enhancer, which is located on the late side of the replication origin at bp 100 to 270, the cytomegalovirus early promoter enhancer, the polyoma enhancer on the late side of the replication origin, and adenovirus enhancers.

For secretion of the translated Pak kinase polypeptide of the invention into the lumen of the endoplasmic reticulum, into the periplasmic space or into the extracellular environment, appropriate secretion signals may be incorporated into the expressed polypeptide.

The signals may be endogenous to the polypeptide or they may be heterologous signals.

The Pak kinase polypeptide may be expressed by the host cells of the invention in a modified form, such as a fusion protein, and may include not only secretion signals, but also additional heterologous functional regions. For instance, a region of additional amino acids, particularly charged amino acids, may be added to the N-terminus of the polypeptide to improve stability and persistence in the host cell, during purification, or during subsequent handling and storage. Also, peptide moieties may be added to the polypeptide to facilitate purification.

Such regions may be removed prior to final preparation of the poly- peptide. The addition of peptide moieties to polypeptides to engender secretion or excretion, to improve stability and to facilitate purification, among other purposes, is a familiar and routine technique in the art.

a preferred fusion protein comprises a heterologous region from an immunoglobulin that is useful to solubilize proteins. For example,

EP 0 464 533 discloses fusion proteins comprising various portions of constant (Fc) region of immunoglobulin molecules together with another human protein or part thereof. In many cases, the Fc portion of a fusion protein is thoroughly advantageous for use in therapy and diagnosis and thus results, for example, in improved pharmacokinetic properties (EP 0 232 262). On the other hand, for some uses it might be desirable to be able to delete the Fc part after the fusion protein has been expressed, detected and purified in the advantageous manner described.

This is the case when the Fc portion proves to be a hindrance to use in therapy, diagnosis or further manufacturing, for example when the fusion protein is to be used as an antigen for immunizations for the preparation of antibodies.

The Pak kinase polypeptide can be recovered and purified from recombinant cell cultures by well-known methods including ammonium sulfate or ethanol precipitation, acid extraction, anion or cation exchange chromatography, lectin chromatography, gel filtration, hydrophobic interaction chromatography, affinity chromatography (such as via binding to immobilized Cdc42 and/or Rac, except in those Pak kinase mutants with mutations in their PBDs which are decreased in their ability to bind to Cdc42 and/or Rac, e.g., PaklLS3,L86 and Pak1L88L86R299) and hydroxylapatite chromatography. Most preferably, high performance liquid chromatography ("HPLC") is employed for purification. Polypeptides of the present invention include naturally purified products, products of chemical synthetic procedures, and products produced by recombinant techniques from a prokaryotic or eukaryotic host, including, for example, bacterial, yeast, insect, mammalian, avian and higher plant cells. Depending upon the host employed in a recombinant production procedure, the polypeptides of the present invention may be glycosylated or may be non-glycosylated.

In addition, mutant Pak kinase polypeptides of the invention may also include an initial modified methionine residue, in some cases as a result of host-mediated processes.

Pak Kinase Polvpeptides and Fragments The invention further provides isolated Pak kinase polypeptides having the amino acid sequence encoded by the above- described isolated nucleic acid molecules. As used herein, the terms "peptide" and "oligopeptide" are considered synonymous (as is commonly recognized) and each term can be used interchangeably as the context requires to indicate a chain of at least two amino acids coupled by (a) peptidyl linkage(s). The term "polypeptide" is used herein to denote chains comprising ten or more amino acid residues. As is commonly recognized in the art, all oligopeptide and polypeptide formulas or sequences herein are written from left to right and in the direction from amino terminus to carboxy terminus.

It will be recognized by those of ordinary skill in the art that some amino acid sequences of the Pak kinase polypeptides of the invention can be varied without significant effect on the structure or function of the polypeptide. If such differences in sequence are contemplated, it should be remembered that there will be critical areas on the protein which determine structure and activity. In general, it is possible to replace residues which form the tertiary structure, provided that residues performing a similar function are used. In other instances, the type of residue may be completely unimportant if the alteration occurs at a non-critical region of the polypeptide.

Thus, the invention further includes variants or mutants of the Pak kinase polypeptides, including allelic variants, which show substantial structural homology to, or the activity of, the above-described Pak kinase polypeptides or which include regions of the Pak kinase polypeptides such as the portions discussed below. Such variants or mutants may include deletions, insertions, inversions, repeats, and type substitutions (for example, substituting one hydrophilic residue for another, but not strongly hydrophilic for strongly hydrophobic as a rule).

Small changes or such "neutral" amino acid substitutions will generally have little effect on activity.

Typical conservative substitutions are the replacements, one for another, among the aliphatic amino acids Ala, Val, Leu and

Ile; interchange of the hydroxylated residues Ser and Thr; exchange of the acidic residues Asp and Glu; substitution between the amidated residues Asn and Gln; exchange of the basic residues Lys and Arg; and replacements among the aromatic residues Phe and Tyr.

Thus, a fragment, derivative or analog of the Pak kinase polypeptides of the invention may be (i) one in which one or more of the amino acid residues are substituted with a conserved or non-conserved amino acid residue (preferably a conserved amino acid residue), and such substituted amino acid residue may be encoded by the genetic code or may be an amino acid (e.g., desmosine, citrulline, ornithine, etc.) that is not encoded by the genetic code; (ii) one in which one or more of the amino acid residues includes a substituent group (e.g., a phosphate, hydroxyl, sulfate or other group) in addition to the normal "R" group of the amino acid; (iii) one in which the mature polypeptide is fused with another compound, such as a compound to increase the half-life of the polypeptide (for example, polyethylene glycol), or (iv) one in which additional amino acids are fused to the mature polypeptide, such as an immunoglobulin Fc region peptide, a leader or secretory sequence, a sequence which is employed for purification of the mature polypeptide (such as GST) or a proprotein sequence. Such fragments, derivatives and analogs are intended to be encompassed by the present invention, and are within the scope of those skilled in the art from the teachings herein and the state of the art at the time of invention.

The polypeptides of the present invention are preferably provided in an isolated form, and preferably are substantially purified.

A recombinantly produced version of a Pak kinase polypeptide can be substantially purified by the one-step method described in Smith and Johnson, Gene 67:31-40 (1988). As used herein, the term "substantially purified" means a preparation of a Pak kinase polypeptide wherein at least 50%, preferably at least 70%, and more preferably at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% of contaminating proteins (i.e., those that are not Pak kinase proteins) have been removed from the preparation.

The Pak kinase polypeptides of the present invention include those comprising a kinase domain wherein the polypeptides are substantially catalytically inactive (as the terms "kinase domain" and "substantially catalytically inactive" have been defined above), which may be mutant Pak kinase polypeptides. Preferred mutant Pak kinase polypeptides of the invention include those comprising one or more mutations in their kinase domain, which may be deletions, substitutions or insertions of one or more amino acids and particularly substitutions or insertions of one or more amino acids. Particularly preferred such mutant Pak kinase polypeptides are those wherein the mutation occurs in one or more amino acid residues within the kinase domain which spans amino acid residues about 270 to about 516 (see Figure 1A), including those wherein the mutation occurs in one or more amino acid residues from about amino acid residue 260 to about amino acid residue 520, from about amino acid residue 270 to about amino acid residue 516, from about amino acid residue 290 to about amino acid residue 400, from about amino acid residue 295 to about amino acid residue 350, from about amino acid residue 297 to about amino acid residue 300, from about amino acid residue 299 to about amino acid residue 300, or in amino acid residue 299, of the polypeptide. More particularly preferred are those mutant Pak kinase polypeptides wherein the mutation is a substitution mutation for the lysine residue at position 299 (L299) of the wildtype Pak kinase polypeptide. Preferred such substitutions include replacement of L299 with a basic amino acid such as arginine or histidine. In a particularly preferred such embodi- ment, the mutation is a substitution of an arginine residue in place of L299 in the Pak kinase, an example of which is the mutant Pakl kinase designated PaklR299. As described above, other mutations may, of course, be made within the kinase domain or other domains (such as the PBD), provided that they result in a Pak kinase that is substantially catalytically inactive. Therefore, the present invention also provides the substantially catalytically inactive mutant Pakl kinase designated PaklL83,L86,R299 which comprises a kinase domain. The present poly-

peptides also include portions or fragments of the above-described polypeptides with at least 30 amino acids and more preferably at least 50 amino acids.

The polypeptides of the present invention can be used as molecular weight markers on SDS-PAGE gels or on molecular sieve gel filtration columns using methods well known to those of skill in the art.

In addition, as described in detail below, the polypeptides of the present invention can be used to raise polyclonal and monoclonal antibodies which are useful in assays for detecting Pak kinase protein expression, as antagonists capable of inhibiting Pak kinase or mutant Pak kinase protein function, or for the isolation of a Pak kinase protein or a variant, mutant or derivative thereof.

As one of skill in the art will appreciate, the Pak kinase polypeptides of the present invention and fragments thereof may be immobilized onto a solid support, by techniques that are well-known and routine in the art. By "solid support" is intended any solid support to which a peptide can be immobilized. Such solid supports include, but are not limited to nitrocellulose, diazocellulose, glass, polystyrene, polyvinylchloride, polypropylene, polyethylene, dextran, Sepharose, agar, starch, nylon, beads and microtitre plates. Linkage of the peptide of the invention to a solid support can be accomplished by attaching one or both ends of the peptide to the support. Attachment may also be made at one or more internal sites in the peptide. Multiple attachments (both internal and at the ends of the peptide) may also be used according to the invention. Attachment can be via an amino acid linkage group such as a primary amino group, a carboxyl group, or a sulfhydryl (SH) group or by chemical linkage groups such as with cyanogen bromide (CNBr) linkage through a spacer. For non-covalent attachments, addition of an affinity tag sequence to the peptide can be used such as GST (Smith, D.B., and Johnson, K.S., Gene 67:31 (1988)), polyhistidines (Hochuli, E., et al., J. Chromatog. 411:77 (1987)), or biotin. Such affinity tags may be used for the reversible attachment of the peptide to the support. Such immobilized polypeptides or fragments may be useful, for example, in

isolating antibodies directed against Pak kinase polypeptides, as described below.

As one of skill in the art will also appreciate, the Pak kinase polypeptides of the present invention and the fragments thereof can be combined with parts of the constant domain of immunoglobulins (Ig), resulting in chimeric or fusion polypeptides. These fusion polypeptides facilitate purification and show an increased half-life in vivo (EP 0 394 827; Traunecker et al., Nature 331:84- 86 (1988)).

Antibodies to Pak Kinases The Pak kinase polypeptides of the invention may be used to produce antibodies directed against Pak kinase polypeptides according to methods well-known in the art. See, for instance, Sutcliffe, J.G., et al., Science 219:660-666 (1983); Wilson et al., Cell 37: 767 (1984); and Bittle, F.J., et al., J. Gen. Virol. 66:2347-2354 (1985). Antibodies specific for Pak kinase polypeptides can be raised against the intact polypeptides of the invention or one or more antigenic polypeptide fragments thereof.

These polypeptides or fragments may be presented together with a carrier protein (e.g., albumin) to an animal system (such as rabbit or mouse) or, if they are long enough (at least about 25 amino acids), without a carrier.

As used herein, the term "antibody" (Ab) may be used interchangeably with the terms "polyclonal antibody" or "monoclonal antibody" (mAb), except in specific contexts as described below. These terms, as used herein, are meant to include intact molecules as well as antibody fragments (such as, for example, Fab and F(ab')2 fragments) which are capable of specifically binding to a Pak kinase polypeptide or a portion thereof. Fab and F(ab')2 fragments lack the Fc fragment of intact antibody, clear more rapidly from the circulation, and may have less non-specific tissue binding than an intact antibody (Wahl et al., J.

Nucl. Med. 24:316-325 (1983)).

The antibodies of the present invention may be polyclonal or monoclonal, and may be prepared by any of a variety of methods. For example, polyclonal antibodies may be made by immunizing an animal

with one or more of the Pak kinase polypeptides or portions thereof of the invention according to standard techniques (see, e.g., Harlow, E., and Lane, D., Antibodies: a Laboratory Manual, Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press (1988); Kaufman, P.B., et al., In: Handbook of Molecular and Cellular Methods in Biology and Medicine, Boca Raton, Florida: CRC Press, pp. 468-469 (1995)). In one preferred method, the antibodies of the present invention are monoclonal anti- bodies (or Pak kinase polypeptide-binding fragments thereof). Such monoclonal antibodies can be prepared using hybridoma technology that is well-known in the art (Köhler et al., Nature 256:495 (1975); Köhler et al., Eur. J. Immunol. 6:511 (1976); Köhler et al., Eur. J. Immunol.

6:292 (1976); Hammerling et al., In: Monoclonal Antibodies and T-Cell Hybridomas, New York: Elsevier, pp. 563-681 (1981); Kaufman, P.B., et al., In: Handbook of Molecular and Cellular Methods in Biology and Medicine, Boca Raton, Florida: CRC Press, pp. 444-467 (1995)).

Alternatively, antibodies capable of binding to Pak kinase polypeptides or fragments thereof may be produced in a two-step procedure through the use of anti-idiotypic antibodies. Such a method makes use of the fact that antibodies are themselves antigens, and that, therefore, it is possible to obtain an antibody which binds to a second antibody. In accordance with this method, Pak kinase polypeptide- specific antibodies are used to immunize an animal, preferably a mouse. The splenocytes of such an animal are then used to produce hybridoma cells, and the hybridoma cells are screened to identify clones which produce an antibody whose ability to bind to the Pak kinase polypeptide-specific antibody can be blocked by the Pak kinase poly- peptide antigen. Such antibodies comprise anti-idiotypic antibodies to the Pak kinase polypeptide-specific antibody and can be used to immunize an animal to induce formation of further Pak kinase polypeptide-specific antibodies.

In another preferred embodiment of the invention, the present antibodies may be prepared as chimeric antibodies. According to the invention, such chimeric antibodies may comprise an antigen- binding domain (i.e., the region of the antibody binding to a Pak kinase)

from a first species and one or more constant regions from a second species. See U.S. Patent No. 4,816,567, which is directed to methods for the preparation of chimeric antibodies, the disclosure of which is incorporated herein by reference in its entirety.

It will be appreciated that Fab, F(ab')2 and other fragments of the antibodies of the present invention may be used according to the methods disclosed herein. Such fragments are typically produced by proteolytic cleavage, using enzymes such as papain (to produce Fab fragments) or pepsin (to produce F(ab')2 fragments). Alternatively, Pak kinase protein-binding fragments can be produced through the application of recombinant DNA technology or through synthetic chemistry.

The Pak kinase protein-specific antibodies of the present invention may be detectably labeled, most preferably with an enzyme, radioisotopic, non-radioactive isotopic, fluorescent, toxin, chemilumin- escent or nuclear magnetic resonance (NMR) contrast agent label.

Suitable examples of each of these types of labels are well-known to one of ordinary skill in the art. Typical techniques for binding a label to an antibody are provided, for example, by Kennedy et al., Clin. Chim. Acta 70:1-31 (1976), and Schurs et al., Clin. Chim. Acta 81:1-40 (1977), all of which methods are incorporated by reference herein.

In an additional preferred embodiment of the invention, the antibodies produced as described above may be covalently or non- covalently immobilized onto a solid support. By "solid support" is intended any solid support to which an antibody can be immobilized, including but not limited to nitrocellulose, diazocellulose, glass, polystyrene, polyvinylchloride, polypropylene, polyethylene, dextran, Sepharose, agar, starch, nylon, beads (including glass, latex, magnetic (including paramagnetic and superparamagnetic) beads) and micro- titre plates. Linkage of the antibodies of the invention to a solid support can be accomplished by attaching one or more ends of the antibody to the support. Attachment may also be made at one or more internal sites in the antibody. Multiple attachments (both internal and at the ends of the antibody) may also be used according to the invention. Attachment can

be via an amino acid linkage group such as a primary amino group, a carboxyl group, or a sulfhydryl (SH) group or by chemical linkage groups such as with cyanogen bromide (CNBr) linkage through a spacer. For non-covalent attachments, addition of an affinity tag sequence to the antibody can be used such as GST (Smith, D.B., and Johnson, K.S. Gene 67:31 (1988)); polyhistidines (Hochuli, E. et al., J.

Chromatog. 411:77 (1987)); or biotin. Alternatively, an indirect coupling agent such as Protein a or Protein G (available commercially, e.g., from Sigma Chemical Co., St. Louis, Missouri) which binds to the Fc region of antibodies may be attached to the solid support and the antibodies of the invention attached thereto by simply incubating the antibodies with the solid support containing the immobilized Protein a or Protein G.

Such affinity tags may be also used for the reversible attachment of the antibodies of the present invention to the support.

Uses The isolated Pak kinase nucleic acid molecules, polypeptides and antibodies of the invention are useful in a variety of methods, for example in industrial, clinical and research settings.

Included among these uses are the use of the present anti-Pak kinase antibodies in the determination of Pak kinase expression or production by isolated cells or tissues, or by cells and tissues in an animal, accord- ing to standard immunological techniques that will be familiar to one of ordinary skill. In addition, the nucleic acid molecules of the invention, particularly those comprising polynucleotides encoding substantially catalytically inactive Pak kinases, may be used in methods for inhibiting or reversing oncogenic transformation (particularly Ras-mediated transformation) of a cell, preferably an animal cell (such as a mam- malian cell including a human cell) or a cancer cell. Analogously, the present nucleic acid molecules, particularly those comprising polynucleotides encoding substantially catalytically inactive Pak kinases, may be used in methods of treating certain disorders in an animal (preferably a mammal such as a human), such as cancers (including but not limited to Ras-dependent cancers, carcinomas,

sarcomas (particularly neurofibrosarcomas), melanomas or leukemias) and neurological disorders (including but not limited to neurofibro- matosis). In particular, the Pak kinase nucleic acid molecules, polypeptides and antibodies of the invention are useful in treating cancers and cancer cells having altered Ras function, including cancers having mutant ras oncogene(s), cancers characterized by elevated tyrosine kinase activity and cancers related to NF-1. Most particularly, the Pak kinase nucleic acid molecules, polypeptides and antibodies of the invention are useful in treating cancers having mutant ras oncogene(s).

In addition, the present nucleic acid molecules, polypeptides and antibodies may be used to identify compounds that modulate the activity of Pak kinases, and to identify novel targets for Pak kinases.

Inhibitor compounds of Pak kinase actiivty have the above described uses.

Inhibition and Reversal of Cellular Transformation The isolated nucleic acid molecules of the invention, particularly those comprising polynucleotides encoding substantially catalytically inactive Pak kinases, may be used in methods for inhibit- ing or reversing the transformation of a cell in vitro or in vivo. In one preferred such aspect, the nucleic acid molecules of the invention may be used to inhibit or reverse oncogenic transformation of a cell, such as that induced by the Ras oncogene and other viral or cellular oncogenes.

Analogously, the nucleic acid molecules of the invention may be used in methods designed to inhibit the activity of an oncogene, such as a Ras oncogene, in a cell comprising the oncogene.

Methods according to this aspect of the invention may comprise one or more steps which are designed to inhibit or reverse transformation in a cell, whether in vitro, in vivo or ex vivo (i.e., in a tissue section removed from the body of an animal that may or may not be replaced into the animal following treatment to inhibit cellular transformation or the activity of the oncogene). In one preferred such embodiment of the invention, a composition comprising one or more of the above-described isolated nucleic acid molecules of the invention may

be introduced into the cell to inhibit the transformation of, or oncogenic activity in, the cell.

In this approach, one or more of the above-described isolated nucleic acid molecules may be incorporated into a vector or virion suitable for introducing the nucleic acid molecule into cell to be treated, to form a transfection vector. Suitable vectors or virions for this purpose include those derived from retroviruses, adenoviruses and adeno-associated viruses. Techniques for the formation of the transfection vector comprising one or more of the Pak kinase-encoding nucleic acid molecules of the invention are well-known in the art, and are generally described in "Working Toward Human Gene Therapy," Chapter 28 in Recombinant DNA, 2nd Ed., Watson, J.D. et al., eds., New York: Scientific American Books, pp. 567-581 (1992), and in the references cited therein. In an alternative approach, the isolated nucleic acid molecules may be introduced into the cell by other methods which will be familiar to one of ordinary skill in the art, including for example electroporation, transduction, transformation, calcium phosphate treatment, hypotonic poration and resealing, and the like. By undertaking the above-described approaches, the level of substantially catalytically inactive Pak kinase protein expression is increased in the cell being treated, thereby inhibiting or reversing the transformation of, or inhibiting oncogenic activity in, the cell.

Compositions for use in this aspect of the invention may optionally further comprise one or more additional compounds, such as a pharmaceutically acceptable carrier or excipient suitable for use with the isolated nucleic acid molecules contained in the compositions. By a "pharmaceutically acceptable carrier or excipient" is meant a non-toxic solid, semisolid or liquid filler, diluent, encapsulating material or formulation auxiliary of any type. The carrier may also contain minor amounts of additives such as substances that enhance isotonicity and chemical stability, which are well-known in the pharmaceutical art.

Thus, the invention also provides pharmaceutical compositions comprising one or more of the isolated nucleic acid molecules of the

invention and a pharmaceutically acceptable carrier or excipient therefor.

Cells that may be treated by the methods of the invention to inhibit or reverse transformation, or to inhibit oncogenic activity, include those obtained from mammals, including but not limited to mice, rats, monkeys, apes, sheep, cows, horses, dogs, cats, guinea pigs, rabbits, and most particularly humans. Although any mammalian cell may be treated by the above-described methods, these methods are particularly well-suited to treat cancer cells. Included among the types of tumor or cancer cells that are advantageously treated by the methods of the invention are carcinoma cells (particularly liver carcinoma cells, ovarian carcinoma cells, breast carcinoma cells, cervical carcinoma cells, lung carcinoma cells, prostatic carcinoma cells, gastric carcinoma cells, bladder carcinoma cells, testicular carcinoma cells, colorectal carcinoma cells, pancreatic carcinoma cells, oral cavity carcinoma cells, squamous cell carcinoma cells, head and neck carcinoma cells and teratocarcinoma cells), sarcoma cells (particularly Kaposi's sarcoma cells, fibrosarcoma cells, neurofibrosarcoma cells and osteosarcoma cells), melanoma cells and leukemia cells.

Therapeutic Uses The isolated nucleic acid molecules of the invention, particularly those comprising polynucleotides encoding substantially catalytically inactive Pak kinases, may also be used therapeutically, for example in methods for treating or preventing a disorder, such as a cancer or a neurological disorder, in an animal suffering therefrom or predisposed thereto. In such approaches, the goal of the therapy is to delay or inhibit the development or progression of the disorder in those animals predisposed to the disorder, and/or to cure or induce remission of the disorder in those animals suffering from the disorder.

In a first such aspect of the invention, the animal suffer- ing from or predisposed to the physical disorder may be treated by introducing into the animal one or more of the isolated nucleic acid molecules of the invention comprising a polynucleotide encoding a

substantially catalytically inactive Pak kinase polypeptide or a fragment thereof. This approach, known generically as "gene therapy," is designed to increase the level of substantially catalytically inactive Pak kinase gene expression in the cells of the animal and thereby to inhibit or delay the development, or to cure of induce remission, of the physical disorder. Analogous gene therapy approaches have proven effective or to have promise in the treatment of mammalian disorders such as cystic fibrosis (Drumm, M.L. et al., Cell 62:1227- 1233 (1990); Gregory, R.J.

et al., Nature 347:358-363 (1990); Rich, D.P. et al., Nature 347:358-363 (1990)), Gaucher disease (Sorge, J. et al., Proc. Natl. Acad. Sci. USA 84:906-909 (1987); Fink, J.K. et al., Proc. Natl. Acad. Sci. USA 87:2334- 2338 (1990)), certain forms of hemophilia (Bontempo, F.A. et al., Blood 69:1721-1724 (1987); Palmer, T.D. et al., Blood 73:438-445 (1989); Axelrod, J.H. et al., Proc. Natl. Acad. Sci. USA 87:5173-5177 (1990); Armentano, D. et al., Proc. Natl. Acad. Sci. USA 87:6141-6145 (1990)) and muscular dystrophy (Partridge, T.A. et al., Nature 337:176-179 (1989); Law, P.K.

et al., Lancet 336:114-115 (1990); Morgan, J.E. et al., J. Cell Biol. 111:2437- 2449 (1990)), as well as in other treatments for certain cancers such as metastatic melanoma (Rosenberg, S.A. et al., Science 233:1318-1321 (1986); Rosenberg, S.A. et al., N. Eng. J. Med. 319:1676-1680 (1988); Rosenberg, S.A. et al., N. Eng. J. Med. 323:570-578 (1990)).

In preferred such methods, one or more isolated nucleic acid molecules of the invention, or one or more of the above-described pharmaceutical compositions comprising one of more of the isolated nucleic acid molecules of the invention, are introduced into or administered to the animal that is suffering from or predisposed to the physical disorder. Such isolated nucleic acid molecules may be incorporated into a vector or virion suitable for introducing the nucleic acid molecules into the cells or tissues of the animal to be treated, to form a transfection vector. Suitable vectors or virions for this purpose include those derived from retroviruses, adenoviruses and adeno- associated viruses. Alternatively, the nucleic acid molecules of the invention may be complexed into a molecular conjugate with a virus

(e.g., an adenovirus or an adeno-associated virus) or with viral components (e.g., viral capsid proteins).

Techniques for the formation of such vectors or virions are as described generally above. In addition, general methods for construction of gene therapy vectors and the introduction thereof into affected animals for therapeutic purposes may be obtained in the above-referenced publications, the disclosures of which are specifically incorporated herein by reference in their entirety. In one such general method, vectors comprising the isolated nucleic acid molecules of the present invention are directly introduced into the cells or tissues of the affected animal, preferably by injection, inhalation, ingestion or intro- duction into a mucous membrane via solution; such an approach is generally referred to as "in vivo" gene therapy. Alternatively, cells, tissues or organs, particularly those containing cancer cells or tumors, may be removed from the affected animal and placed into culture according to methods that are well-known to one of ordinary skill in the art; the vectors comprising the Pak kinase polynucleotides of the invention may then be introduced into these cells or tissues by any of the methods described generally above for introducing isolated polynucleotides into a cell or tissue, and, after a sufficient amount of time to allow incorporation of the Pak kinase polynucleotides of the invention, the cells or tissues may then be re-inserted into the affected animal. Since the introduction of the Pak kinase gene of the invention is performed outside of the body of the affected animal, this approach is generally referred to as "ex vivo" gene therapy.

For both in vivo and ex vivo gene therapy, the isolated nucleic acid molecules of the invention may alternatively be operatively linked to a regulatory DNA sequence, which may be a promoter or an enhancer, or a heterologous regulatory DNA sequence such as a promoter or enhancer derived from a different gene, cell or organism, to form a genetic construct as described above. This genetic construct may then be inserted into a vector, which is then directly introduced into the affected animal in an in vivo gene therapy approach, or into the cells

or tissues of the affected animal in an ex vivo approach. In another preferred embodiment, the genetic construct of the invention may be introduced into the cells or tissues of the animal, either in vivo or ex vivo, in a molecular conjugate with a virus (e.g., an adenovirus or an adeno-associated virus) or viral components (e.g., viral capsid proteins).

These approaches result in increased production of substantially catalytically inactive Pak kinase by the treated animal via (a) random insertion of the gene encoding the substantially catalytically inactive Pak kinase into the host cell genome; or (b) incorporation of the gene encoding the substantially catalytically inactive Pak kinase into the nucleus of the cell where it may exist as an extrachromosomal genetic element. General descriptions of such methods and approaches to gene therapy may be found, for example, in U.S. Patent No. 5,578,461; WO 94/12650; and WO 93/09222.

For treating or preventing a physical disorder by administering one or more of the above-described pharmaceutical compositions to the animal, the compositions should be formulated and dosed in a fashion consistent with good medical practice, taking into account the clinical condition of the individual patient, the site of delivery of the composition, the method of administration, the scheduling of administration, and other factors known to practitioners.

The "therapeutically effective amount" of the pharmaceutical composi- tions or isolated nucleic acid molecules of the invention for purposes herein is thus determined by such considerations. The key factor in selecting an appropriate dose is the result obtained, as measured, for example, by increases in the level of substantially catalytically inactive Pak kinase expression or by determining a reversal or inhibition of cellular transformation or of the physical disorder or symptoms thereof.

Other useful measures of determining therapeutic effectiveness are known to one of ordinary skill in the art. The length of treatment needed to observe changes, and the interval following treatment for responses to occur, may vary depending on the desired effect.

Pharmaceutical compositions for use in such methods comprise one or more of the isolated nucleic acid molecules of the

present invention and may optionally comprise a pharmaceutically acceptable carrier or excipient therefor, as described above. The isolated nucleic acid molecules and the pharmaceutical compositions of the present invention may be administered by any means that achieve their intended purpose of delaying or inhibiting the progression or development, or inducing the remission, of physical disorder an affected animal. For example, administration may be by oral, ocular, otical, rectal, parenteral, subcutaneous, intravenous, intramuscular, intraperitoneal, intravaginal, topical (as by powders, ointments, drops or transdermal patch), buccal, intrathecal or intracranial routes, as an oral or nasal spray or as ocular or intraotic drops. The term "parenteral" as used herein refers to modes of administration which include intravenous, intramuscular, intraperitoneal, intrasternal, subcutaneous and intraarticular injection and infusion. The dosage administered will be dependent upon the age, health, and weight of the recipient, kind of concurrent treatment, if any, frequency of treatment, and the nature of the effect desired. While individual needs may vary from one animal to another, determination of optimal ranges of effective amounts of each component is within the ability of the clinician of ordinary skill.

The present compositions may also be administered by sustained-release systems. Suitable examples of sustained-release compositions include semi-permeable polymer matrices in the form of shaped articles, e.g., films, or microcapsules. Sustained-release matrices include polylactides (U.S. Patent No. 3,773,919; EP 0 058 481), copolymers of L-glutamic acid and gamma-ethyl-L-glutamate (Sidman, U., et al., Biopolymers 22:547-556 (1983)), poly (2- hydroxyethyl methacrylate) (Langer, R., et al., J. Biomed. Mat. Res. 15:167-277 (1981); Langer, R., Chem. Tech. 12:98-105 (1982)), ethylene vinyl acetate (Langer et al., Id.) and poly-D(-)-3-hydroxybutyric acid (EP 0 133 988). Sustained- release compositions may also include liposomally entrapped nucleic acid molecules of the invention, which may be prepared by any of a variety of methods that have been well-described in the art (See U.S.

Patent Nos. 4,485,045 and 4,544,545; Epstein et al., Proc. Natl. Acad. Sci.

(USA) 82:3688-3692 (1985); Hwang et al., Proc. Natl. Acad. Sci. (USA) 77:4030-4034 (1980); EP 0 036 676; EP 0 052 322; EP 0 088 046; EP 0 102 324; EP 0 142 641; EP 0 143 949; DE 3,218,121; and JP 83-118008).

For parenteral administration, in one embodiment, the present isolated nucleic acid molecules may be formulated generally by mixing them at the desired degree of purity, in a unit dosage injectable form (solution, suspension, or emulsion), with a pharmaceutically acceptable carrier, i.e., one that is non-toxic to recipients at the dosages and concentrations employed and is compatible with other ingredients of the formulation. For example, the formulation preferably does not include oxidizing agents and other compounds that are known to be deleterious to nucleic acid molecules.

Generally, the formulations are prepared by contacting the isolated nucleic acid molecules of the invention uniformly and intimately with liquid carriers or finely divided solid carriers or both.

Then, if necessary, the product is shaped into the desired formulation.

Preferably the carrier is a parenteral carrier, more preferably a solution that is isotonic with the blood of the recipient. Examples of such carrier vehicles include water, saline, Ringer's solution, and dextrose solution.

Non-aqueous vehicles such as fixed oils and ethyl oleate are also useful herein, as are liposomes. The therapeutic compositions comprising the isolated nucleic acid molecules of the invention ordinarily may be stored in unit or multi-dose containers, for example, sealed ampules or vials or as an aqueous solution.

A variety of physical disorders may be treated, prevented or cured in animals by these therapeutic methods of the invention.

Such disorders include but are not limited to cancers and neurological disorders. Cancers suitably treated or prevented by these methods include, but are not limited to, Ras-associated cancers, carcinomas, sarcomas (including neurofibrosarcomas), melanomas and leukemias, particularly those described above. Neurological disorders suitably treated or prevented by these methods include, but are not limited to, neurofibromatosis. The methods of the invention are particularly

well-suited for treating or preventing physical disorders in any animal, preferably in mammals and most particularly in humans. Regardless of the approach, use of the therapeutic methods of the present invention will result in the increased production of substantially catalytically inactive Pak kinase by the cells and tissues of the treated animal, such that the development or progression of the physical disorder will be delayed or inhibited, or such that the physical disorder will go into remission or be cured.

Identification of Modulators of Pak Kinase Activitv The isolated nucleic acid molecules, polypeptides and antibodies may also be used in methods that allow the identification of compounds that modulate (i.e., inhibit or activate) the activity of a Pak kinase. By the "activity" of a Pak kinase as used in this aspect of the invention is meant the physiological function of a Pak kinase enzyme, which includes without limitation the binding of various targets (e.g., Cdc42/Rac, ATP, etc.), catalytic activity (e.g., phosphorylation of various physiological and nonphysiological targets, etc.), intracellular signal- ling and other functions which may be direct or indirect.

Methods according to this aspect of the invention may comprise one or more steps, such as (a) contacting one or more Pak kinase polypeptides of the invention with a compound to be tested for its ability to modulate the activity of the Pak kinase; and (b) determining the effect of the compound on the activity of the Pak kinase. Compounds that may be tested according to such methods include, but are not limited to, peptides (including polypeptides and proteins such as enzymes, enzyme complexes, antibodies, peptide hormones, cytokines and the like), steroids, organic and inorganic compounds and the like, which may be natural or synthetic. Compounds identified by these methods may be used to produce pharmaceutical compositions comprising one or more such compounds and a pharmaceutically acceptable carrier or excipient therefor (which carrier or excipient may comprise one or more of those described above). Such pharmaceutical compositions may then be used in methods of treating or preventing disorders in animals (preferably

mammals such as humans), comprising administering to the animal an effective amount of one or more of the present pharmaceutical compositions. Disorders that may be effectively treated or prevented include those described above.

In practice, the determination of whether a compound to be tested modulates the activity of a Pak kinase may be carried out by a variety of art-known methods. Such methods, which may be carried out in vitro or in vivo, may include, for example, determining the ability of a Pak kinase to bind Cdc42/Rac in the presence and absence of the compound to be tested; determining the ability of a Pak kinase to phosphorylate myelin basic protein in the presence and absence of the compound to be tested; etc. Such binding and kinase assays are known in the art and are described in detail in the Examples below. Other preferred assay methods will be familiar to those of ordinary skill.

In analyzing the results of such tests, those test compounds which induce an increase in the measured activity of the Pak kinase may be said to be "Pak kinase activators" (or "positive modulators").

Similarly, those test compounds which induce a decrease in (or inhibit) the measured activity of the Pak kinase may be said to be "Pak kinase inhibitors" (or "negative modulators").

Identification of Pak Kinase Targets The nucleic acid molecules, polypeptides and antibodies of the invention may also be used in methods of identifying novel targets of Pak kinases. As used herein, a "target" of a Pak kinase means a molecule, complex, organelle, cellular structure, etc., that is acted on (i.e., bound by, enzymatically modified (e.g., phosphorylated), etc.) by a Pak kinase, either in vitro or in vivo. Such targets are typically derived from cells and may be intracellular, such as cytoplasmic, nuclear, organellar or membrane constituents or structures, including proteins (cytosolic and membrane-bound), cytoskeletal elements, second messenger molecules, receptors, enzyme complexes and the like. In the practice of the invention, such targets may be contained in cellular extracts or homogenates, in whole cell preparations or in single-cell

suspensions, etc. Pak kinase targets identified by the present methods may be used in methods of diagnosing, treating or preventing a variety of disorders in animals (including mammals such as humans), such as those described above, or to identify compounds and compositions useful in such diagnostic and therapeutic methods.

Methods according to this aspect of the invention may comprise one or more steps, such as (a) contacting a composition to be tested for its content of a Pak kinase target with one or more of the Pak kinase polypeptides of the invention; and (b) determining the activity of the Pak kinase on one or more of the components of the composition, to identify those components that are acted on by the Pak kinase poly- peptide. As noted above, determination of the activity of a Pak kinase on potential Pak kinase targets may be accomplished by measuring, for example, the binding to the target by Pak kinase, or a structural or conformational change in the target induced by Pak kinase. Such determinations may be accomplished by direct or indirect assays. Direct methods of assaying interactions between Pak kinases and putative Pak kinase targets include, for example, determining the ability of the Pak kinases of the invention to bind directly to a cellular component or structure. Such direct methods may be facilitated by detectably labeling the Pak kinase and/or the cellular component or structure, for example with a radiolabel, a fluorescent label, a chemiluminescent label, an enzyme label or the like. Alternatively, such kinase-target interactions may be measured indirectly via a variety of art-known methods, such as using a Yeast Two-Hybrid system (see Golemis, E.A., et al., in: Current Protocols in Molecular Biology, Ausubel, F.M., et al., eds., New York: John Wiley & Sons, Inc., pp. 20.1.1- 20.1.28 (1996)), or using an anti-Pak kinase antibody of the invention, which may be an epitope-tagged anti- body, to sequentially mask potential target-binding sites on the Pak kinase molecules and thereby identify those targets that bind the Pak kinases. Other methods suitable for determining kinase-target interactions will be familiar to one of ordinary skill.

Methods of determining the ability of a Pak kinase to act

on a target may include, for example, assays of the ability of a Pak kinase to bind to or phosphorylate the target. Such assays have been described in detail above and in the Examples below. Other methods suitable for identifying Pak kinase targets which may advantageously use the polynucleotides, polypeptides and antibodies of the present invention will be apparent to the skilled artisan.

It will be readily apparent to one of ordinary skill in the relevant arts that other suitable modifications and adaptations to the methods and applications described herein are obvious and may be made without departing from the scope of the invention or any embodiment thereof. Having now described the present invention in detail, the same will be more clearly understood by reference to the following examples, which are included herewith for purposes of illustration only and are not intended to be limiting of the invention.

EXAMPLES Materials and Methods The following materials and methods were generally used in each of the Examples.

Plasmids.

cDNA expression plasmids utilizing the CMV promoter to express myc tagged Pakl and PaklR299 based on the plasmid pCMV6M (a modified version of pCMV5) have been described elsewhere (Sells, M.A., et al, Current Biol. 7:202-210 (1997)), Pak1L83L6 and Pak1L83.L86.R299 were constructed using a unique-site-elimination mutagenesis protocol to introduce the desired mutations (Deng, W.P. and Nickoloff, J.A., Analyt. Biochem. 200:81-88 (1992)). Human H-ras and K-ras4B expression systems that utilize pZIP-NeoSV(x)1, a retrovirus vector (neomycin resistant) were prepared as previously described (Oldham, S.M., et al., Proc. Nat. Acad. Sci. (USA) 93:6924-6928 (1996)). Expression of the inserted gene is regulated from the Moloney long terminal repeat promoter. v-raf expression plasmids were described elsewhere (Kohl, N.E., et al., Science 260:1934-1937 (1993)). GST-Racl, and GST-Cdc42

bacterial expression vectors based on the plasmid pGEX-2T were prepared as previously described (for example K. Shinjo et al., Proc.

Nat. Acad. Sci. (USA) 878:9853-9857 (1990)).

Cell culture and transformation assavs Rat-1 cells are from the Merck collection of strains and have been described elsewhere (Kohl, N.E., et al., Science 260:1934-1937 (1993)).

NIH 3T3 cells were obtained from the American Type Culture Collection (ATCC CRL-1658), Rockville, MD. Rat-1 cells were grown in high- glucose (4.5 g/liter) Mediatech Dulbecco's modified Eagle medium (DMEM) purchased from Fisher Scientific (Pittsburgh, PA), supple- mented with 10% fetal bovine serum (Sigma, St. Louis, MO), penicillin (100 units/ml), and streptomycin (100 mg/ml), and kept at 37"C and 5% CO95% air. DNA transfections were performed by the calcium phosphate precipitation technique. 20 pig total DNA (10 pig of each test DNA and, when single plasmids were tested, 10 pig of pUCl9 plasmid) were briefly mixed with 0.5 ml 0.25 M CaCl2 and 0.5 ml 2X BES buffered saline (BBS) and incubated 10-20 min at room temperature. The mixture was then added dropwise to a 25-50% confluent, freshly fed 100 mm dish of cells, swirled gently, and incubated 18-24 h at 37"C and 5% CO2. Cells were washed twice with growth media, re-fed, incubated 24-48 h and then split 1:5 into 100 mm dish. Post-transfection cultures were fed twice a week with fresh growth medium. Cell foci were scored 14-18 days post-transfection by fixing in a 10% acetic acid/10% methanol solution and staining the dishes with 0.4% crystal violet in 10% ethanol.

Soft agar assays were performed as previously described (Cox, A.D. and Der, C.J., Methods Enz. 238:277-294 (1994)). 103 post-transfection cells were plated on 60 mm dishes. After 12-15 days, colonies were examined under a Nikon DIAPhot microscope using phase contrast.

To establish stable Rat-1 cell lines expressing Pakl, PaklR299 paklL83,L86,R299 and PaklL83L86 each construct was co-transfected with pCDNA3 into Rat-1 cells. The transfected cells were selected in growth medium containing 400 pig/ml of Geneticin (G418) (GIBCO/BRL, Grand Island, NY). Protein expression levels were determined by

Western blot (immunoblot) analysis of G418-selected cell lysates using the anti-myc rag monoclonal antibody 9E10 (Calbiochem, Cambridge, MA) with the procedure outlined in the enhanced chemiluminescence kit (Amersham, Arlingtown Heights, Il).

Transfection efficiency assays were performed by trans- fecting Rat-1 cells with the Pak and Ras test plasmids to be tested along with pRSV- -gal as a control. 48 h after transfection cultured cells were rinsed with PBS and then fixed in PBS containing 2% formaldehyde and 0.2% glutaraldehyde for 5 min at 4"C. The cells were then washed with PBS and overlaid with a histochemical reaction mixture containing 1 mg/ml 4-Cl-5-Br-3-indolyl- -galactosidase (X-gal), 5 mM potassium ferricyanide, and 2 mM MgCl2 in PBS. After incubating for 18-24 h, cells were then rinsed with PBS and fixed in PBS containing 2% formaldehyde and 0.2% glutaraldehyde for 15 min. Cells were examined under a microscope and scored positive for LacZ if they turned blue. Transfection efficiencies ranged from 5-10% and were not affected by any of the plasmids used in this study.

Rac/ Cdc42 binding and kinase assays Biochemical assays for Pakl were performed on extracts from COS cells transfected with Pak plasmids. Lipofectamine-mediated transient transfections of COS-7 cells were performed according to the manufacturer's protocol (Life Technologies, Inc., Gaithersburg, MD).

18-24 h prior to transfection 2-3 x 105 COS cells were plated on 35 mm dishes. A total of 1.5 g of DNA and 10 pil of Lipofectamine reagent were added to the plates containing 1 ml of DMEM (in the absence of serum).

After 5 h, 1 ml of DMEM containing 20% fetal bovine serum was added.

After 18-24 h, the medium was replaced with fresh DMEM containing 10% fetal bovine serum. 48-72 h after addition of DNA transfected COS cells were washed with cold phosphate-buffered saline and lysed in 40 mM Hepes (pH 7.4),1% Nonidet P-40, 100 mM NaCl, 1 mM EDTA, 25 mM NaF, 1 mM sodium orthovanadate, 10 g/ml Leupeptin, 10 Fg/ml aprotinin and centrifuged at 12,000 xg for 25 min at 4"C (Bagrodia, S.,

et al., J. Biol. Chem. 270:22731-22737 (1995)). Protein concentrations ranged from 5.3 to 7.2 mg/ml.

Pak kinase assays were performed on immunoprecipitates from COS cells as follows. Extracts were incubated with antibody 9E10 and Protein A beads for two hours at 4"C. Precipitates were washed three times with lysis buffer. Immunoprecipitates were washed twice in 2X phosphorylation buffer (10 mM MgCl2, 40 mM Hepes, pH 7.4) and, where indicated, incubated with soluble GTP-bound GST-Cdc42, GST- Racl ( - 5 pig of protein), and 5 pig of myelin basic protein (Sigma) for 5 min on ice. Kinase assays were initiated by the addition of 10 Ci of (y~32p) ATP (3000 Ci/mmol) and 20 RM ATP (final concentration) followed by incubation for 10 min at 22"C (Bagrodia, S., et al., J. Biol. Chem.

270:22731-22737 (1995)). Reactions were stopped by the addition of 2X SDS sample buffer, heated to 95"C, and the products were resolved by SDS- PAGE (12% gel) and visualized by autoradiography.

Variations on this assay method are also possible. For example, Pak could be expressed recombinantly in E. coli or Sf9 cells and then purified. The purified Pak protein would not have to be immuno- precipitated and the assay could be done with Pak free in solution, allowing for isolation of product by acid precipitation/scintillation counting or by a scintillation proximity assay. Additional physiological protein substrates of Pak kinase or peptide fragments of those substrates could also be used. In all instances, a test compound could be added to the reaction mixture to test for inhibition of Pak kinase. Such inhibition could be due to direct inhibition of Pak kinase activity itself or due to inhibition of Pak binding to Cdc42/Rac which is required to stimulate Pak kinase.

To perform Rac and Cdc42 binding assays, 50 pig of purified GST-Cdc42 or GST-Racl were incubated with lysis buffer for 15 min at room temperature to release any nucleotide, washed with lysis buffer, and then incubated with 1 mM GDP- S [guanosine-5'-O-(2- thiodiphosphate)] or GTP in lysis buffer with 10 mM MgCl2 for 30 min at room temperature, and then washed by lysis buffer with 10 mM MgC12, to remove unbound nucleotides. Next, the proteins were incubated with

10 pil of COS cell lysates and glutathione beads supplemented with 10 mM MgCl2 for 1.5 hours at 4"C (Bagrodia, S., et al., J. Biol. Chem.

270:22731-22737 (1995)). Precipitates were washed three times with lysis buffer containing 10 mM MgCl2. Bound proteins were eluted in SDS sample buffer, subjected to 12% SDS-PAGE, Western blotted, and probed for Pakl with anti-myc epitope monoclonal antibody 9E10.

JNK and MAPIJERK kinase assays Transfections of Rat-1 cells were performed similar to those for the transformation assays above. 5 pig of DNA (1 pig of HA-JNKl, HA-ERK1 or HA-p38; 2 pig of each test DNA; when single plasmids were tested, 2 pig of pUCl9 plasmid) were mixed with 0.125 ml 0.25 M CaCl2 and 0.125 ml 2x BES buffered saline and incubated for 10-20 minutes at room temperature. The mixture was then added dropwise to a 50-60% confluent and freshly fed 35 mm dish of cells, mixed gently, and incubated for 18-24 h. The cells were then washed twice with growth media, re-fed, and incubated for 24-48h. Next, transfected Rat-1 cells were washed two times with cold PBS, lysed in Lysis buffer, and centrifuged at 12,000 xg for 30 min at 4"C. Extracts were incubated with HA- antibody (12CA5) and protein A beads for 3-3.5 h at 4"C. Precipitates were washed three times with lysis buffer and washed two times in 2x phosphorylation buffer. The precipitates were then incubated with 5 g of GST-c-jun (for Jun kinase), MBP (for ERX), or GST-ATF2 (for p38 kinase), 10 Ci of(y~32p) ATP (3000 Ci/mmol) and 20 RM ATP (final concentration) at 30"C for 30 min. Mixtures were washed three times with lysis buffer and two times with 2x phosphorylation buffer.

Reactions were stopped by adding 2x SDS sample buffer and heating to 95"C. All experiments were performed at least twice with similar results.

EXAMPLE 1 Construction and Characterization of Pak Mutants

In the present studies, four myc-tagged Pak mutants were constructed and characterized and then used for further experiments.

The first was PaklR299 which converts lysine 299 to arginine and renders the enzyme catalytically inactive (Sells, M.A., et al., Current Biol.

7:202-210 (1997); Zhang, S., et al., J. Biol. Chem. 270:23934-23936 (1995)).

This was confirmed by performing kinase assays on extracts from COS cells transfected with the expression vectors (Fig. 1A). While anti-Pak immunoprecipitates from Pakl-transfected cells efficiently phosphorylated myelin basic protein, no kinase activity was detected in cells transfected with PaklR299. Expression was confirmed by probing Western blots with antibody 9E10, which recognizes the myc tag on the amino terminus (Fig. 1B and 1C). Another mutant tested was PaklL83,L86, which substitutes leucines for highly conserved histidines in the Cdc42/Rac binding domain. This mutant failed to bind either Cdc42 or Rac. Binding assays were performed by precipitating with the appropriate Rac or Cdc42 GST fusion protein and Western blotting the precipitates for Pakl with antibody 9E10 (Fig. 1B and 1C). PaklL83.L86 also failed to bind either Rac or Cdc42 in overlay assays and the yeast two hybrid assay (data not shown). Other mutant proteins shown in Figure 1 are discussed below.

EXAMPLE 2 Kinase-Deficient Pak Inhibits Ras Transformation To test the role of Pak in Ras transformation we performed co-transfection experiments with human K-ras4B (K-ras) using Rat-1 fibroblasts. As expected, when we fixed and stained the cells 14-18 days later with crystal violet, we observed more than 100 foci in the plates transfected with K-ras alone. However, when we included the Pak1R299 expression plasmid in the transfections, we observed about 90% fewer foci in the plates (Fig. 2A). Wild type Pakl or the vector plasmid did not inhibit K-ras transformation. To test if PaklR299 was a non- specific inhibitor of transformation of Rat-1 cells, we performed co-transfection

experiments with v-Raf (Fig. 2B). We found that PaklR299 did not inhibit Raf transformation. Interestingly, the inhibition was specific for Rat-1 cells as no inhibition was observed when NIH 3T3 cells were substituted for Rat-1 cells (Fig. 2C). To test if Pak altered levels of Ras or Raf expression, we prepared extracts from transfected cells and performed Western blots for Ras, Raf and Pak. None of the mutant Paks affected expression of either Ras or Raf (Fig. 2D). Therefore, PaklR299 specifically inhibits Ras transformation of Rat-1 cells without affecting transformation by Raf or affecting Ras expression.

We also measured transformation by assessing growth on soft agar after co-transfecting with K-ras and Pakl plasmids. We observed numerous colonies on soft agar plates with K-ras transfections and, as seen in the focus assays, we observed very few colonies when cells were co-transfected with PaklR299 (Fig. 3). Furthermore, most of the rare colonies seen in the presence of PaklR299 were substantially smaller than those seen with K-ras alone. As with the focus assays, no inhibition was observed when wild type Pakl was substituted for the mutant Pak1R299 or NIH 3T3 cells were tested instead of Rat-l cells (data not shown for NIH 3T3 cells). Typically, when equal concentrations of K-ras and Pakl DNA were transfected into cells, transformation was inhibited by about 90% in focus assays and about 95% in soft agar assays (Fig. 3G). We found Pak1R299 inhibited transformation by both K-ras and H-ras but not by Raf (data not shown for H-ras).

We further addressed the interaction between Ras, Raf and Pak using stable cell lines that expressed Pakl and PaklR299 We co-transfected Rat-l cells with the plasmid pCDNA3 and selected Geneticin (G418) resistant cell lines. We tested expression of the wild- type and mutant Pak by probing a Western blot for a Myc tag fused to the N-terminus of Pak (Fig. 4A) and found that the stable cell lines each expressed a novel 65 kDa protein at comparable levels. Pakl expression caused a small stimulation of growth rate while PaklR299 expression caused a small inhibition of the growth rate, but these differences were within the margin of error of the experiment (Fig. 4B). None of the cell lines expressing the Pakl mutants proliferated in 1% serum (Fig. 4C)

or on soft agar (data not shown). Stable expression of Pakl and PaklR299 affected the morphology of cells. Pakl expressing cells were elongated while PaklR299 cells lost the spindle shape characteristic of fibroblasts (Fig. 5 A-C). Pak1L83,L86 and PaklL83,L86,R299 also caused changes in cellular morphology, but to a smaller extent (Fig. 5 D-E).

Upon establishing the cell lines for stable expression, we tested them in Ras transformation assays using both focus assays and soft agar colony assays. As predicted from the co-transfection experiments, we found that cells expressing Pakl were transformed as efficiently as control cells, while cells expressing Pak1R299 were highly resistant to K-ras transformation (Fig. 6). In a dose response experi- ment, we determined that about 10 to 100 times as much K-ras was required to transform the PaklR299 expressing cells as compared to Rat-1 cells or Rat-1 cells expressing wild-type Pakl. Similar dose response curves were obtained in both focus formation (Fig. 6A) and soft agar assays (Fig. 6B). As expected from the focus assays above, all cell lines were efficiently transformed by Raf (Fig. 6C). Expression of wild type and mutant Pakl did not affect transfection efficiencies (data not shown).

EXAMPLE 3 A Functional Cdc42/Rac Binding Domain is not Required for Ras Inhibition Since Rac is essential for Ras transformation, it was possible that the dominant negative Pakl mutant inhibited Ras transformation by sequestering Rac or Cdc42 into an inactive complex.

To address this mechanism, we tested a Pakl mutant that fails to bind either Cdc42 or Rac and is also defective in kinase activity. The mutant, PaklL83,L86,R299, has substitutions of leucines for conserved histidines at positions 83 and 86, along with the original R299 in the kinase domain (Fig. 1). The new mutations lie within the p21 binding domain (PBD), a region which is necessary and sufficient for both Rac and Cdc42 binding. Western blots of extracts from transfected COS cells confirmed

expression at levels comparable to other mutants, and no Rac or Cdc42 binding was detected when measured directly (Fig. 1C). We found that the PaklL33,L86,R299 mutant was as potent in Ras inhibition as the original PakR299 (Fig. 3). In dose response experiments, cells expressing PaklL83,L86,R299 were as resistant to K-ras transformation as cells expressing Pak1R299 (Fig. 6). We also tested Pak1L83L86, which expressed an hyperactive kinase that was not further stimulated by Rac or Cdc42 (Fig 1A). Mutations within other conserved residues in the PBD of Pak3 also causes hyperactivity (Bagrodia, S., et al., J. Biol. Chem. 270:27995- 27998 (1995)). No effect on K-ras or Raf transformation was observed by paklL83,L86 suggesting that, despite its essential role in Ras transform- ation, constitutively active Pakl does not appear to be an oncogene nor does it cooperate with Ras or Raf in transforming cells (data not shown).

EXAMPLE 4 Ras Inhibition is Uncoupled from JNK but not from MAPK Signaling In order address if the signaling pathways of Rac and Ras were affected by Pak we measured the effect of Pak on JNK and MAP kinase activation in Rat-1 cells. We co-transfected an HA-tagged JNK with Rac and the various Pak constructs described above, immuno- precipitated JNK with the HA antibody and measured phosphorylation of a GST-Jun fusion protein (Fig. 7A). The activated RacL6i stimulated JNK activity almost 40-fold, as did the activated PaklL83 L86 relative to vector control (Fig. 7A, lanes 3,7,8). No stimulation was observed by Pak1, PaklR299 or Pak1L88,L86,R299 (Fig. 7A, lanes 1,2 and 9). To test if the mutant Paks inhibited activation by Rac we cotransfected them with RacL61. We observed no changes when the active Pakl constructs (Pakl and Pak1L83L86) were cotransfected with RacL61. We found that the PaklR299 construct inhibited JNK activation by - 75%, while no inhibition was observed with the Pak1L83,L86R299 construct (lanes 5 and 6). Similar levels of activation were also observed when p38 kinase was tested in place of JNK (Fig. 7D) and when COS-7 cells were used in place of

Rat-1 cells (data not shown). Similarly, we found that PaklR299, but not PaklL83,L86,R299, inhibited Ras activation of JNK (Fig. 7B). No stimulation of JNK was observed by Raf (Fig. 7C). These observations support a Ras to Rac to Pak/p38 activation model and suggest that JNK inhibition is not obligatory for Pak mutants to inhibit Ras transformation.

To measure if Pakl interacted with the MAPK/ERK kinase pathway, we co-transfected Rat-1 cells with HA-tagged ERK1, K-ras and the various Pak constructs described above, immunoprecipitated ERK1 with the HA antibody and measured phosphorylation of myelin basic protein (Fig. 8). K-Ras stimulated ERK1 activity about 35-fold relative to vector control (Fig. 8A lanes 8,9). No stimulation was observed by Pakl, PaklR299 paklL83,L86 or PaklL83,L86,R299, either alone or in the presence of RacLe1 (Fig. 8A, lanes 1-7; Fig. 8B, lanes 1-2). To test if the mutant Paks inhibited activation by Ras we cotransfected them with K-ras.

We observed no changes when the active Pakl constructs (Pakl and PaklL83L86) were cotransfected K-ras (Fig. 8B, lanes 3 and 5). We found that the PaklR299 construct inhibited ERK activation by about 50% and Pak1L83,L36,R299 also inhibited ERK activation to a similar extent (Fig. 8B, lanes 4 and 6). We also observed similar levels of activity when experi- ments were performed in COS-7 cells (data not shown). To test if these mutants also inhibited Raf activation we determined their effects on Raf activation of ERK kinase. None of the Pak constructs inhibited Raf activation of ERK (Fig. 8C). These observations suggest that the dominant negative Pakl mutants may inhibit Ras transformation by interfering with the MAP/ERK kinase cascade.

Example 5 Role of Pak Kinase in Schwann Cell Transformation and Neurofibromatosis Neurofibromatosis type 1 (NF1) is a common autosomal dominant disorder caused by loss of the NF1 gene. The disease is characterized clinically by neurofibromas, café au lait spots, and in some cases neurofibrosarcomas. The Schwann cell is thought to be

the primary cell affected in NF1 patients. The protein encoded by NF1, neurofibromin, is a negative regulator of the Ras oncogene, so loss of neurofibromin causes elevated levels of activated Ras. Cellular trans- formation by Ras requires Rac, and a candidate for an effector of Rac is Pak kinase.

To study the role of Pak kinase in Schwann cell transform- ation, cells from the Rat-1 rat Schwann cell line were transfected with Pakl or with dominant negative mutants of Pakl or Rac and tested in a Ras transformation assay. Dominant negative mutants of Rac and Pakl (PaklR299) specifically inhibited transformation by Ras of Rat-1 cells, with typically 75-90% fewer transformed colonies observed from co-transfected cells (Figure 10). Moreover, Pak1L83L86,R299, a mutant that fails to bind Rac, also inhibited transformation (Figure 10). These results suggest that Pak binds additional proteins required for Ras transformation besides Rac.

Ras is known to activate the MAP/ERK cascade and a related cascade leading to activation of JNK/SAPK. JNK activation occurs through Rac and Pak. To address which signalling pathway was utilized by Pakl to inhibit transformation, the effects of the Pakl dominant negative mutants were tested on Ras activation of JNK and ERK. Surprisingly, transformation inhibition correlated with ERK inhibition but not JNK inhibition (Figure 11A and 11B), suggesting a functional connection between Pak and ERK. The use of Rat Schwann cells as a model system for neurofibrosarcomas was validated by the observation that dominant negative mutants of Pakl caused tumorigenic reversion of ST88-14, a neurofibrosarcoma cell line from an NF1 patient (Figure 9).

Taken together, these results suggest that Rac and Pak may be promising pharmacological targets for designing therapeutic agents for treating patients with neurofibrosarcomas and neurofibromatosis.

General Discussion The present results have demonstrated that Pakl interacts with an essential component of the Ras signaling pathway in addition

to Rac and Cdc42. A catalytically inactive Pakl mutant inhibited Ras transformation of Rat-1 fibroblasts in both focus assays and soft agar assays, two well established assays for Ras transformation. Neither wild type Pakl nor a hyperactive mutant transformed cells, nor did either significantly affect the transformation frequencies of Ras or Raf.

To extend these studies, a variety of cell lines have been surveyed to determine which ones are sensitive to PakR299 inhibition. PakR299 was found to have no effect on Ras transformation of NIH 3T3 cells but to inhibit Ras transformation of the Rat-1 rat Schwann cell line (Peden, K.W.C., et al., Exp. Cell. Res. 185:60-72 (1989); unpublished observa- tions). Thus, dominant negative Pakl inhibits Ras in Rat-1 cells but not in NIH 3T3 cells; nevertheless, the present observations are not unique to Rat-1 cells.

The major Ras signaling pathway in most organisms is the MAP kinase cascade (Marshall, C.J., Cell 80:179-185 (1995)). Since only Pakl mutants that inhibit Ras activation of MAP kinase inhibit transformation, the present results support a role for MAP inhibition as relevant for Pakl inhibition. Pak1R299 inhibits both JNK activation and MAP activation but since paklL83,L86,R299 inhibits transformation without inhibiting JNK or p38 activation, the results of the present studies suggest that JNK inhibition is not necessary for Ras inhibition.

Interestingly, activation of MAP kinases by Pak1L83,L86, which activates JNK to about the same extent as RacL61, has not been detected. In addition, no evidence has been obtained of the capability of PakLS3,LS6 to transform cells either by itself or in cooperation with Ras, Rac or Raf.

Thus, despite very strong inhibition of Ras transformation, constitutively active Pak is not an oncogene.

Several possibilities may explain these observations. Pak may bind essential components without activating them; alternatively, Pak may be required for Ras activation of MAP signaling but not be present in limiting quantities. Evidence for saturating levels of Pak in cells is suggested by the observation that maximum levels of JNK and p38 activation are obtained with RacL61 alone; cotransfection of Racl with Pakl or Pak1L88,L86 does not further stimulate JNK or p38 (Figure 7).

Additionally, activation of MAP via Pak may require translocation as well as enzymatic activation; recently, activation of the PDGF receptor was shown to recruit the adaptor protein Nck to the membrane through an SH2 domain binding site. Nck, in turn bound and activated Pak through one of its SH3 domains translocating the complex to the membrane. Interestingly, both JNK/p38 and MAP were activated by Pak translocation (Galisteo, M.L., et al., J. Biol. Chem. 271:20997-21000 (1996); Lu, W., et al., Current Biol. 7:85-94 (1997)). Thus translocation of Pak may be important for activation of MAP which could explain why the hyperactive kinase PaklL83,L86 does not activate MAP. Roles for Rho, Rac and Cdc42 have also been suggested in MAP signaling because, although none will activate MAP alone, all will synergize with an activated Raf to activate MAP (Frost, J.A., et al., Mol. Cell. Biol 16:3707- 3713 (1996)). The present data are consistent with this receptor/Nck/Rac/ Pak/MAP pathway playing an essential role in Ras transformation.

Expression of Pak1R299 or the N-terminal half of Pak inhibits Rac and Cdc42 activation of JNK in co-transfection experiments (Figure 7). Additionally, the N-terminal half of Pak inhibits ERK kinase activation (Frost, J.A., et al., Mol. Cell. Biol 16:3707-3713 (1996)). The mechanism of inhibition was proposed to be through sequestering of Rac and Cdc42 via the PBD domain. These same mutants also inhibit JNK activation, which prevented determination if JNK signaling was required. The results of the present studies confirm that mutant Paks inhibit both ERK and JNK. Furthermore, they suggest that Rac/Cdc42 sequestering is only required for JNK inhibition; neither Rac/Cdc42 sequestering nor JNK inhibition is required for ERK inhibition.

The Ras-Raf-MEK-ERK signaling pathway has been established in many cell types, but other routes to transformation occur in most cells, and in some cells the alternate pathways may predomin- ate over the Ras-Raf pathway. Cells where Ras signaling is largely Raf independent include a certain line of NIH 3T3 fibroblasts, rat intestinal epithelial cells and Wistar rat thyroid cells (Al-Alawi, N., et al., Mol.

Cell. Biol. 15:1162-1168 (1995); Khosravi-Far, R., et al., Mol. Cell. Biol.

16:3923-3933 (1996); Oldham, S.M., et al., Proc. Nat. Acad. Sci. (USA)

93:6924-6928 (1996)). In each of these cells Ras transduces mitogenic signals independent of Raf activation. Since the Rat-1 cells used in the present studies can be transformed by both Ras and Raf (Figures 2 and 6), they are not Raf- independent. However, since these cells are sensitive to Pak inhibition there may be a major role for Raf-independent transformation. The growing number of experimental systems where Ras transformation is uncoupled from Raf activation suggests that the components of Ras alternate pathways are possible targets for novel antineoplastic drugs.

The possibility that JNK is involved in mediating an alternate Ras transformation signal is supported by the observation that Rac stimulates JNK and dominant negative Rac mutants inhibit Ras transformation. However, several groups recently constructed Rac mutants that failed to interact with PAK and subsequently failed to activate JNK (Joneson, T., et al., Science 274:1374-1376 (1996); Lamarche, N., et al., Cell 67:519-529 (1996); Westwick, J.K., et al., Mol. Cell. Biol.

17:1324-1335 (1997)). These same groups found that their mutant Rac proteins still transformed cells and caused membrane ruffling. Thus Pak and JNK activation are not required for Rac transformation. In studies designed to address the role of JNK in Ras transformation, a dominant negative mutant of SEK was found to inhibit JNK activation and Ras transformation, but not Ras activation of MAP, suggesting a critical role for JNK in Ras transformation (Clark, G.J., et al., J. Biol.

Chem. 272:1677-1681 (1997)). Thus, while JNK activation may not be essential for Rac transformation it appears essential for Ras transformation. Although MAP inhibition, but not JNK inhibition, has been correlated in the present studies with the dominant negative Pak mutants, these results do not necessarily exclude JNK from an essential role in Ras transformation since Pak1R299, the only mutant that inhibited JNK activation, also inhibited ERK activation. This possibility prevented the use of Pak mutants in the present studies to address the role of JNK exclusive of MAP.

The mechanism by which Ras communicates to Pak presumably operates through an effector that is activated by GTP-bound

Ras, but the relevant protein has yet to be identified. Raf is a strong candidate for the relevant effector, since all mutants that inhibited Ras transformation also inhibited MAP kinase activation. If Raf is indeed the site of inhibition, the inhibition is bypassed by the activating mutant, v-Raf. Other potential Ras effectors that Pak dominant negative mutants might interfere with include Ral GDS, Rinl and phosphatidyl- inositol-3-OH kinase, all of which bind Ras-GTP (Han, L. and Colicelli, J., Mol. Cell. Biol. 15:1318- 1323 (1995); Hofer, F., et al., Proc. Natl. Acad.

Sci. (USA) 91:11089-11093 (1994); Marshall, M.S., FASEB J. 9:1311-1318 (1995); Rodriguez-Viciana, P., et al., Nature 370:527-532 (1994)). Another potential target is p190 Rho GAP because it associates with Ras GAP, an association proposed to mediate Ras activation of JNK (Clark, G.J., et al., J. Biol. Chem. 272:1677-1681 (1997); Westwick, J.K., et al., Mol. Cell.

Biol. 17:1324-1335 (1997)). The effects that the dominant negative Pak mutant have on cell shape (Figure 5) suggest the actin cytoskeleton may also be involved in the inhibition of Ras transformation by Pak (Sells, M.A., et al., Current Biol. 7:202-210 (1997)).

The present observation that a kinase deficient/PBD domain mutant Pak still inhibits Ras transformation suggests that Pak interacts with the Ras signaling pathway independent of Cdc42/Rac sequestering via the PBD domain. At least two mechanisms may account for these observations: (1) multiple binding sites for Rac/Cdc42 on Pakl; and (2) novel interactions between Pak and other proteins required for Ras signaling.

Raf is an example of a small G protein effector with multi- ple binding sites. Two sites on Raf bind Ras: the first is apparently the primary binding site, while the second is a cryptic site, only unmasked after Ras binds the first site (Brtva, T.R., et al., Ras. J. Biol. Chem.

270:9809-9812 (1995); Drugan, J.K., et al., J. Biol. Chem. 271:233-237 (1996); Hu, C.D., et al., The J. Biol. Chem. 270:30274- 30277 (1995); Zhang, X.F., et al., Nature 364:308-313 (1993)). Although Rac/Cdc42 binding to PaklL88,L86 and PaklLS3,LS6 ,R299 could not be detected, there is still a possibility that multiple binding sites exist on Pakl for Rac and Cdc42.

A recent study using chimeras between Rac and Rho found that two

sites on Rac were required for Pak binding and membrane ruffling. The first site (amino acids 30 to 40) is equivalent to the major effector region on Ras while the second site (amino acids 143 to 175) does not correspond to a known effector region of Ras (Diekmann, D., et al., EMBO J. 14:5297- 5305 (1995)). Since there are two effector regions on Rac for Pak there may well be multiple binding sites on Pak for Rac and Cdc42. However, since Rac/Cdc42 binding by the PaklL83,L86 mutant was not detected, it is likely that the primary Rac binding site on Pak is the PBD domain. If a second Rac binding site exists, it most likely does not support significant binding by itself.

Another mechanism by which Pak may inhibit Ras is by sequestering other proteins essential for Ras transformation distinct from Rac and Cdc42. Such factors may interact with the kinase domain in the C-terminus causing, in the case of PakR299, a non-productive interaction. A dominant negative Raf mutation that fails to bind Ras has also been constructed which probably acts by sequestering the downstream kinase MEK (Brtva, T.R., et al., Ras. J. Biol. Chem.

270:9809-9812 (1995); Van Aelst, L., et al., Proc. Natl. Acad. Sci. (USA) 90:6213-6217 (1993)). Similarly, PakR299 may sequester downstream kinases such as MEKK or SEK. Other potential sites for protein-protein interactions are found in the N-terminus of Pakl and includes several proline rich regions that bind to SH3 domains, and an acidic region (Galisteo, M.L., et al., J. Biol. Chem. 271:20997-21000 (1996); Sells, M.A.

and Chernoff, J., Cell. Biol. 7:162-167 (1997)).

The identification of physiological targets for Pakl may elucidate the mechanism of Ras inhibition. Although the protein-protein interactions of mammalian Pak kinases are not well understood, the yeast Pak homolog Ste20p interacts with several other components of the mating signaling complex in addition to Cdc42, including Ste5p, and Bemlp (Leeuw, T., et al., Science 270:1210-1212 (1995)). Homologs of Ste5p and Bemlp have yet to be identified in mammals.

Having now fully described the present invention in

some detail by way of illustration and example for purposes of clarity of understanding, it will be obvious to one of ordinary skill in the art that the same can be performed by modifying or changing the invention within a wide and equivalent range of conditions, formulations and other parameters without affecting the scope of the invention or any specific embodiment thereof, and that such modifications or changes are intended to be encompassed within the scope of the appended claims.

All publications, patents and patent applications mentioned in this specification are indicative of the level of skill of those skilled in the art to which this invention pertains, and are herein incorporated by reference to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated by reference.