THREE HYBRID SCREENING ASSAY USING REVERSE TRANSFECTED MAMMALIAN CELL ARRAYS

TECHNICAL FIELD

The present invention relates to mammalian three-hybrid screening methods, arrays and kits for studying the interaction between small molecules and their target proteins in mammalian cells. The present invention is particularly useful for screening a small molecule against a protein library to identify the protein targets of the small molecule. The invention may also be used for screening a library of small molecules to identify which of these binds to a known protein target.

BACKGROUND

A major area of research during the identification and development of new pharmaceuticals is the study of the interaction between the small molecule drug candidates and their protein targets. A technology, known as yeast three-hybrid, was first reported in 1996, see Liu and Licitra, Proc Natl Acad Sci U S A. 1996 Nov 12; 93(23): 12817-21. It was developed with the aim of facilitating the identification of the protein targets of small molecule drug candidates.

The yeast three-hybrid system, described in detail in US 5,928,868, is an assay system that comprises a DNA binding domain and an activation domain, each expressed as a subunit of a different fusion (hybrid) protein. When these fusion proteins are brought into proximity, via the binding of the small molecule component of a heteroligand (heterocompound) to a target protein, they form a three-hybrid complex, which is active as a transcription factor. When transcription factor activity is reconstituted by the formation of the three-hybrid complex, transcription of a reporter gene is activated and a positive assay result produced.

The yeast three-hybrid assay is carried out in cultured yeast cells that are stably transformed with three different vectors, these are as follows: - 1) a first expression vector, comprising DNA that encodes a first hybrid protein, the first hybrid protein comprising a ligand binding domain, linked to either a DNA binding domain

("DBD"), or an activation domain ("AD"); this vector is sometimes known in the art as the "bait".

2) a second expression vector, comprising DNA that encodes a second hybrid protein, the second hybrid protein comprising a target polypeptide linked either to an activation domain, ("AD"), or a DNA binding domain, ("DBD"); this vector is sometimes known in the art as the "prey".

3) a vector comprising a reporter gene, the transcription of which can be activated when the DBD and AD of the first and second hybrid proteins are brought into proximity of the reporter gene by the binding of a heterocompound.

Yeast three-hybrid is, potentially, a very useful tool in pharmaceutical research. However, there are a number of limitations inherent in the technology that has restricted its applicability, both in drug (small compound) library screening and in the identification of target proteins. Indeed, there are only a limited number of reports describing the successful application of yeast three-hybrid, for a review of these see Kley N, Chemistry & Biology, VoI 11, 599-608, May 2004.

The major drawback of the yeast three-hybrid assay is that it is uses yeast rather than human or mammalian cells to evaluate small molecules and their interaction with target proteins. Post translational modifications of mammalian proteins are missing in the yeast environment and these may be required for compound-target protein interaction to occur, or for the active form of a target protein to be made e.g. through dimerisation. This means that a yeast-based system is expected to miss target protein-small molecule "hits" that might be expected to occur in a mammalian cell environment, thus the yeast three-hybrid system gives misleading results. It would be very advantageous to use a mammalian host cell, ideally human, to test small molecules which are ultimately intended for use as human therapeutics, so that these are tested in a cellular environment that most closely mimics the native protein environment.

In addition, yeast cells, whilst being permeable to small molecules, have a very efficient active transport mechanism for removing small molecules. This means that very high

concentrations of the small molecule (in the form of the heterocompound) must be used. The resulting intracellular concentration of the small molecule can be uncertain, and may be at levels that are far higher than would be used therapeutically. This can lead to an unacceptably high level of false positives and results that are of only limited relevance to a 5 mammalian system. The pharmacokinetics of mammalian cells obviously makes them the most appropriate models for the study of small molecules intended for therapeutic use.

Despite the fact that a mammalian form of the three-hybrid assay would be such useful tool and yeast three-hybrid technology has been available for a number of years there have been o no reports of the assay being performed in mammalian cells. There are, however, a number of references that refer to mammalian three-hybrid systems or assay systems that purport to be mammalian three-hybrid assays, but these either fail to provide any experimental validation or don't disclose how to perform the assay, or they aren't true three-hybrid systems that enable the study of small molecule-target protein interaction. 5

US 5,928,868 discloses the yeast three-hybrid system, it also states that a mammalian version of the yeast three-hybrid assay would be useful and even has a specific claim to such an assay, see claim 4. However, the patent contains no specific examples of a mammalian three-hybrid assay and contains no technical disclosure describing how such an o assay might be carried out.

The recent review by Kley, Chemistry and Biology, VoI 11, 599-608, May 2004 again highlights the need for a mammalian three-hybrid system. It also describes an assay system that it states could be developed into a mammalian three-hybrid system. However, the ₅ assay format and methodology referred to differs significantly from the present invention because it relies on receptor signalling to activate transcription of a reporter gene. No information is provided that indicates how such a three-hybrid assay might be carried out.

In summary, despite the well documented need for a mammalian three-hybrid system and o much effort invested in the development of a mammalian cell assay for the study of small molecule-protein interaction there have been no reports of such a system in the prior art.

Sabatini et al first developed a technique known as "reverse transfection" as a way of introducing nucleic acids into defined areas of a lawn of mammalian cells, to form an array of transfected cells. Reverse transfection has been applied to a number of different assay formats that use mammalian host cells, for a recent review of this field see Palmer E, Pharmacogenomics (2005) 6(5), 527-534. Known applications of reverse transfection include the study of gene function, the identification of new GPCRs and mammalian two- hybrid, as described by Feibtz et al, Poster 167, HGM2004 Poster Abstracts.

SUMMARY OF THE INVENTION

The inventors have now found that using the reverse transfection technique has enabled them to develop a novel three-hybrid assay that is performed in mammalian cells. This makes it possible to very rapidly screen large libraries of different target proteins to detect pharmacologically relevant small molecule: protein interactions. Indeed, the new assay makes feasible the rapid screening of small molecules against the entire proteome. This means that the present invention, in addition to being used to identify the protein targets of a small molecule and thus determine the mechanism of action of the small molecule, also has great potential for toxicity testing. For example, a small molecule, with a known mechanism of action (primary target), may be screened against a library comprising the entire proteome to determine whether it binds to any other targets, including those associated with toxicity, so called "off-target" toxicity testing.

The present invention utilises the components of the previously described yeast three- hybrid assays but the assay is performed in mammalian cells that need only be transiently transfected with the necessary vectors.

The mammalian cells are grown on a substrate and form a living array that is used to study small molecule-target protein interaction. At least one of the vectors that encode the assay components is present on this substrate prior to seeding it with the cells so that the vector enters the cells by what is known as "reverse transfection".

When searching for a target protein of a small molecule in a library, e.g. of vector containing either a cDNA clone from a clone collection fused to a DBD or AD, or cDNA clone or fragment of clone from a library, is spotted onto the substrate at defined locations prior to seeding the substrate with mammalian cells. This is advantageous because it enables the cells to be transfected with a different library member depending upon their location on the substrate. This produces an array that has a group of cells transfected with and expressing a different DNA sequence (from a library or clone collection) at a defined location on the array. The three-hybrid assay is then carried out to deteπnine whether any of the different protein targets, expressed at defined locations on the array bind to a small molecule (in the form of a heterocompound).

STATEMENT OF INVENTION

Thus according to a first aspect of the present invention there is provided a mammalian three-hybrid assay method for identifying a target polypeptide to which a small molecule binds, comprising: (a) introducing into a mammalian cell;

(i) a first expression vector, comprising a DNA encoding a first hybrid protein the first hybrid protein comprising a ligand binding domain linked to a first transcriptional module;

(ii) a second expression vector, comprising a DNA encoding a second hybrid protein, the second hybrid protein comprising the target polypeptide linked to a second transcriptional module; (iii) a third vector comprising a reporter gene,

(iv) a heterocompound comprising an anchor ligand covalently linked to the small molecule, such that binding of the anchor ligand to the ligand binding domain, and binding of the small molecule to the target polypeptide permits expression of the reporter gene; (b) identifying one or more cells that express the reporter gene; and (c) identifying the target polypeptide expressed in the cells identified in step (b); and

wherein the second expression vector is present at defined locations on the substrate prior to seeding the substrate with mammalian cells, such that the vector enters the cells by reverse transfection.

Preferably the DNA encoding the target polypeptide comprises DNA from a library of DNA, preferably the DNA is selected from genomic DNA, cDNA and synthetic DNA. The library of DNA can be a clone collection, PCR product library, protein targets of known identity, a fully characterised library, sequence- validated open reading frame (ORF) library, a tissue a cell specific library such as a liver library, or a library of mutants or fragments of a primary target protein.

The assay method of the present invention is particularly useful for performing competition assays, i.e. to determine whether a small molecule is able to compete with the small molecule of the heterocompound for binding to the protein target. Thus preferably mammalian three hybrid assay method of the invention further comprised the addition of a second small molecule into the cells to determine whether it competes with the heterocompound for binding to the target protein. This second small molecule may be any small molecule. For example, it may be an identical but unmodified form (i.e. no chemical linker) of the small molecule used in the heterocompound, or it may be a compound of similar structure to the small molecule of the heterocompound, or a compound of unrelated structure.

Although the present invention is particularly useful for identifying the protein target of a known small molecule it may also be used to screen a library of small molecules against a known protein target. This requires the synthesis of a different heterocompound for each small molecule to be assayed. The cells on the array will all be transfected with the same vectors but a different heterocompound will be added to separate locations on the array, for example, by embedding the heterocompound in a biodegradable polymer on the substrate, in manner analogous to that described recently by Bailey SN et al in PNAS, November 16th 2004, vol. 101, no. 46, pl6144-16149. Alternatively, each heterocompound is added to a separate array.

Thus is a second aspect the invention provides a mammalian three-hybrid assay method for identifying a small molecule which binds to a known polypeptide, comprising:

(a) introducing into a mammalian cell; (i) a first expression vector, comprising a DNA encoding a first hybrid protein the first hybrid protein comprising a ligand binding domain linked to a first transcriptional module;

(ii) a second expression vector, comprising a DNA encoding a second hybrid protein, the second hybrid protein comprising the known polypeptide linked to a second transcriptional module;

(iii) a third vector including a reporter gene,

(iv) a heterocompound comprising an anchor ligand covalently linked to the small molecule, such that binding of the anchor ligand to the ligand binding domain, and binding of the small molecule to the target polypeptide permits expression of the reporter gene;

(b) identifying one or more cells that express the reporter gene; and

(c) identifying the heterocompound present in the cells identified in step (b); and wherein the second expression vector is present at defined locations on the substrate prior to seeding the substrate with mammalian cells, such that the vector enters the cells by reverse transfection.

When using the present invention to screen a library of protein targets to identify the target or targets of a small molecule a library of second expression vectors is constructed each member of which expresses a hybrid protein comprising a transcriptional module (DBD or AD) linked to a different protein target. Preferably the method or array of the invention uses a second expression vector that comprises a library of vectors constructed from, for example, a cDNA library (clone collection), a library of PCR products. Preferably the library comprises protein targets of known identity, more preferably a fully characterised library, such as the type available from Invitrogen, where the sequence of each clone is known. This enables the instant identification of each target polypeptide that produces a positive result in the assay, without the need for a time consuming sequencing step. Still

more preferably a library comprising a pre-made, sequence-validated open reading frame (ORF) that is already cloned into an entry vector (also known as Gateway enabled) such as the Ultimate™ ORF Clone Collection, available from Invitrogen, can be used; this makes construction of the second expression vector relatively straightforward.

The first expression vector and the third vector (reporter gene) may also be present on the substrate prior to seeding with cells. More preferably both the first and the second expression vector are present prior to seeding with cells, the third vector (reporter gene) being added after seeding the cells onto the substrate and preferably when the cells are o established and growing over the substrate.

When utilising the present invention to screen a library of protein targets, the first expression vector and reporter gene remain constant. Preferably, the mammalian cells will be transiently transfected, as transgene expression persists for sufficient time to perform the s assays. However, the invention is not limited in this respect. If desired, stably transfected mammalian cell lines can be generated that express the "constant" components of the assay, e.g. the first hybrid protein, the reporter gene, or both.

The present invention also concerns arrays that may be useful in the practice of the o invention.

Thus in a third aspect, the invention provides an array for a mammalian three-hybrid assay comprising, at defined locations on a substrate, a first expression vector, comprising a DNA encoding a first hybrid protein the first hybrid protein comprising, a ligand binding 5 domain linked to a first transcriptional module. Preferably a second expression vector, comprising a DNA encoding a second hybrid protein, the second hybrid protein comprising the target polypeptide linked to a second transcriptional module is also present at defined locations on the substrate. The first hybrid protein is capable of interaction with the second hybrid protein in the presence of a heterocompound, said heterocompound comprising an o anchor ligand covalently linked to a small molecule.

The first and second expression vectors used in the practice of the invention encode fusion or hybrid proteins, each hybrid protein including a transcription module (either a DNA Binding Domain or an Activation Domain) and a ligand binding domain or a target protein for binding either anchor A or the small molecule of the heterocompound. Once the three- hybrid complex is formed, and the transcription modules are brought into close proximity, the transcriptional activation of a reporter gene occurs. Preferably the second expression vector comprises DNA encoding a second hybrid protein comprising a target protein linked to an activation domain (AD).

It is well established that many transcription factors possess two modular domains that are separable in function, see Crabtree et al (WO 95/02684). Any of the modular transcription factors known in the art may be useful for the three-hybrid system of the present invention. Indeed, any of those reported as useful in yeast three-hybrid may be used in the mammalian assay, see Liu et al US 5,928,868, incorporated herein by reference. Particularly suitable pairs of DNA binding (DBD) and activation domains (AD) for use in the practice of the invention include the Lex A DNA binding protein (DBD) that binds tightly to Lex A operator and activates transcription of a reporter gene such as Lac Z and a the complimentary B42 activation domain; another suitable pairing is the GAL4 DBD and the GAL4 AD.

A preferred pairing for use in the practice of the invention is GAL4 DBD with a VP 16 AD. Preferably the second transcriptional module is the GAL4 DNA binding domain and the first transcriptional module is the VP 16 activation domain.

The first expression vector for use in the methods, arrays and kits of the invention comprises a DNA encoding a first hybrid protein the first hybrid protein comprising, a ligand binding domain linked to a first transcriptional module and DNA. The first transcriptional module may be a DNA binding domain (DBD) or an activation domain (AD) but preferably it is a DBD, preferably the GAL4 DBD.

The ligand-binding domain of the first hybrid protein acts to capture the anchor ligand of the heterocompound. Any protein that can bind specifically to the compound used as the anchor ligand may be used in the practice of the invention. Prior art three-hybrid assays have generally relied on the specific interaction of a ligand-receptor pair, such as methotrexate and DHFR, to provide the necessary components, methotrexate being used as the anchor ligand of the heterocompound and DHFR as the ligand binding domain of the first hybrid protein. Thus preferably, the assays of the present invention comprise methotrexate as the anchor ligand of the heterocompound, and DHFR as the ligand-binding domain of the first hybrid protein, preferably the DHFR is E.coli DHFR. However, other suitable anchors may be used providing they bind to a known target (to be used as ligand binding domain), and providing it is possible chemically link the anchor ligand to the small molecule to form the heterocompound. For example, dexamethasone may be used as an anchor ligand and paired with the glucocorticoid receptor used as the ligand-binding domain. Other pairings suitable pairs of anchor ligand and ligand binding domain include those disclosed in Liu et al US 5,928,868, incorporated herein by reference.

The second expression vector used in the practice of the invention comprises a DNA encoding a second hybrid protein, the second hybrid protein comprising the target polypeptide linked to a second transcriptional module. As discussed above the transcriptional module used must pair with the transcriptional module of the first hybrid protein. When the present invention is used to screen a small molecule against a library of target proteins a library of second expression vectors is constructed and this is spotted out on to defined locations a substrate. However, when screening a compound library to identify which compounds bind to a know protein target (as described in the second aspect of the invention) the second expression vector remains constant.

The reporter gene used in the mammalian three-hybrid assay of the invention encodes a reporter protein, the transcription of which is switched on in the presence of transcriptional modules brought into proximity with the reporter gene by the binding of the heterocompound to a target protein. Numerous different reporter genes are known in the art and may be used in the practice of the present invention. Preferably the reporter gene used

permits visual screening. Examples of reporter gene products that may be detected visually include Green Fluorescent Protein (GFP), and the New Fluorescent Proteins (NFPs) which are a set of fluorescent proteins from corals. These NFPs are also now referred to as coral fluorescent proteins and as reef coral fluorescent proteins (RCFPs).

The switching on or off of the reporter gene depends on the binding affinity of the small molecule ligand to the target protein so as to activate the reporter gene or to competitively inhibit the activation of the reporter gene. The affinity of a ligand or small molecule for a target molecule may vary substantially in the three-hybrid screen.

The mammalian three-hybrid system of the invention uses a heterocompound. When the heterocompound binds to both the first and second hybrid protein a three-hybrid complex is formed that stimulates transcription of at least one reporter gene. The detection of a positive result may follow from direct binding of a heterocompound or by competitive binding of the heterocompound acting as an agonist or antagonist of a known ligand for a target protein. As discussed above, the heterocompound comprises an anchor ligand, capable of specific interaction with the ligand-binding domain of the first hybrid protein, linked to small molecule.

The heterocompound is generally added to the array of cells after they have been transfected with the assay vectors. However, the heterocompound may be present on the substrate prior to seeding with cells. Preferably it is added after the cells are established on an array. However, if a number of different heterocompounds are to be used, for example, when using the method of the second aspect of the invention, it is advantageous to spot these out on to defined areas of the substrate prior to seeding with cells, e.g. in the manner described by Bailey SN et al in PNAS, November 16th 2004, vol. 101, no. 46, pl6144- 16149. Thus a heterocompound array, i.e. an array of heterocompounds that comprises different small molecules at defined locations on a substrate, can be constructed.

The anchor ligand remains a constant component of the heterocompound. Methotrexate has been successfully used as an anchor ligand in the yeast three-hybrid system but

methotrexate is toxic to mammalian cells. Surprisingly, the present inventors have determined that methotrexate can be successfully used as an anchor ligand in a mammalian system. Thus, preferably, the heterocompound used in the present invention comprises methotrexate as the anchor ligand. The steroid molecule, dexamethasone, which binds the glucocorticoid receptor with high affinity, may also be used as the anchor ligand or any other anchor ligands used in the yeast three-hybrid system.

The small molecule component of the heterocompound can be a known small molecule or a molecule of unknown identity obtained from a combinatorial library, or other small molecule archive. Thus the small molecule of the heterocompound maybe any samll molecule that can be linked to the anchor ligand to form a heterocompund that Examples of combinatorial libraries include but are not limited to peptide libraries, nucleic acid libraries, polysaccharide libraries, and small organic molecules. When searching for the protein targets of a known small molecule the heterocompound is a constant component of the assay, e.g. known small molecule linked to methotrexate. However, in certain circumstances, the protein target may be known but a suitable small molecule capable of binding to the protein target needs to be identified. Thus, it may be desirable to generate a library or series of different heterocompounds from a mixture of small molecules, or a defined library of heterocompounds.

The anchor ligand and small molecule are covalently linked, via a chemical linker, to form the heterocompound. This linkage may be formed by any of the methods known in the art, particular those exemplified and referred to in US 5,928,868, incorporated herein by reference. For example: Jerry March, Advanced Organic Chemistry (1985) Pub. John Wiley & Sons Inc; and HH, House, Modern Synthetic Reactions (1972) pub. Benjamin Cummings) Example 1 and Figure 7 describes an embodiment of a linkage reaction between dexamethasone and FK506. Descriptions of linkage chemistries are further provided by Crabtree et al. WO 94/18317, 95/02684, Schreiber et al WO 96/13613, Holt et al.

The present inventors have found that a straight n-methlyene chain linker works effectively in the mammalian 3-hybrid assay, where n = is an integer from 2-30, preferably n =10 (known as a 10-carbon linker or ClO linker). A ClO linker has been used previously in yeast. Thus preferably the heterocompound for use in the present invention comprises a ClO linker, more preferably the heterocompound comprises a small molecule linked to methotrexate via a ClO linker. A suitable route to synthesis of a DEX-IOC-MTX heterocompound is described in Henthorn DC, et al, Biochemical Pharmacology 63 (2002) 1619-1628.

The inventors have found that using mammalian cells instead of yeast has enabled them to use much lower concentrations of heterocompound, the amount can be reduced by about 10-100 times. This means that small molecule-protein interactions can be studied at concentrations that are much closer to the concentrations available at a therapeutic dose.

Many different mammalian cell types may be used in the practice of the invention. Cells suitable for use in a three-hybrid assay include primary cultures, cultures of immortalized cells or genetically manipulated strains of cells.

One of the main criteria for selection of a particular cell type may be the nature of post translational modification of target proteins expressed where the binding of such modified target proteins to a small molecule may more accurately mimic the natural state. Cells that are associated with a particular disease state, or that originate from a particular tissue type may be chosen. Another criteria is the selection of a suitable cellular background to mimic the activity of small molecule in its target tissue or cell type. If studying toxicity it may be appropriate to select a cell type associated with that toxicity, e.g. liver. Cell lines recognised in the art as easy to transfect are particularly preferred. Different mammalian cell types may also be selected according to their permeability.

Cells may also be selected on the basis of their adherence to the chosen substrate, their rate of growth, and the ease with which they can be maintained in culture. Preferably the cells are human cells. Preferred types of cells include HEK239T.

The substrate can be any solid surface that supports the attachment of mammalian cells. Any of the substrates described in the prior art as useful in cell arrays, particularly reverse transfection arrays may be used, including the substrates described in Pharmacogenomics (2005) 6(5), 527-534 and WO 01/20015 Al (Sabatini), incorporated herein by reference. Preferably the substrate is glass slide. Preferably the substrate comprises a cationic polymer layer such as aminopropylsilane (GAPS). Slides with cationic polymers on the surface may be particularly useful, as these have been reported to permit transfection without the need to use transfection reagents.

Typically a transfection reagent is used in the practice of the invention. However, the invention is not limited in this respect, transfection may be carried out without the aid of transfection reagents, for example, by applying the nucleic acid to an appropriate substrate. If using a transfection reagent it may be added to the substrate, either on its own or mixed with the vector DNA, preferably it is mixed with one or more of the vectors and spotted onto the substrate. There are a number of different methodologies known in the art e.g. Pharmacogenomics (2005) 6(5), 527-534 and WO 01/20015 Al. Generally the DNA is allowed to dry onto the substrate before the addition of the transfection reagent; the substrate is then ready to be seeded with cells.

The methods and arrays of the present invention have a number of useful applications in the identification and development of new small molecule therapeutics. A discussed earlier, toxicity screening is an important application, for example using the method of the first aspect to identify so called "off-target" toxicity by identifying any protein targets that a molecule binds to in addition to its main therapeutic target. A small molecule could be screened against the entire proteome to identify all its protein target or it could be screened against a more defined library. For example, a library of proteins associated with liver toxicity could be used to produce a toxicity profile for a small molecule. Further testing could then be performed using a series of closely related compounds to determine whether any of these has a better toxicity profile, e.g. a compound that binds to less toxicity associated proteins or binds to a lesser degree, for example, by using a competition assay.

Competition assays may also be performed to determine the relative potencies of similar compounds. Another application of the invention is its use to screen a library of closely related target proteins, for example, mutants, fragments, or different polymorphic forms of a primary target protein, in order to study how these changes affect the binding of a known 5 small molecule.

o DESCRIPTION OF FIGURES

Figure 1 shows a diagrammatic representation of the methodology used for screening a protein target library, the specific methodology of Example 1 is shown.

1. A reporter plasmid, pG5 Zs Green, a plasmid expressing a GAL4-DHFR protein chimera s (first expression vector) and a prey plasmid expressing a protein chimera of a variable cDNA and VP 16 (second expression vector) are spotted onto a glass slide.

2. Transfection reagent is added to the DNA on the slide and then mammalian cells are seeded (here shown as HEK293T cells). A test heterocompound is added to the cell media.

3. The GAL4-DHFR chimera protein (first hybrid protein) binds to the upstream activation o sequence (UAS) present on the reporter plasmid. The DHFR part of the chimera then tightly binds the methotrexate (MTX) part (anchor ligand) of the heterocompound. This offers the variable end (small molecule) of the heterocompound (X) as bait to the variable expressed part (target protein) of the VP 16 chimera. If the variable expressed part of the VP 16 chimera binds to the heterocompound bait then theVP16 transcriptional activation ₅ domain is brought into proximity with the start site of the ZsGreen reporter gene resulting in upregulation of reporter gene transcription.

Figure 2 shows cell viability following exposure to methotrexate, a heterocompound comprising methotrexate-CIO linker-dexarnethasone and dexamethasone.

A ™ 0 HEK293T cells were seeded at 5xlO ⁴ in 100ml of DMEM (plus Glutamax ) (Invitrogen) with 10% foetal calf serum per well of a black walled 96 well plate. These cells were

treated with MTX-ClO-DEX, DEX or MTX over an eight-point log titration of compound from 0-1000OnM. Cells were incubated for 24 hours at 37 ⁰ C and cell viability was assessed

TM using the luminescence based Cell Titer-Glo assay (Promega). Luminescence is directly proportional to cell viability so relative luciferase units (RLU x 10 ) are used to express cell viability. MTX alone was found to significantly reduced cell viability but when MTX was part of a heterocompound, cell viability was maintained.

Figure 3 shows a plasmid map of the plasmid that has an VP 16 activation domain it can be used to generate the second expression vector, e.g. activation domain fused to target

TM protein. The pACT Gateway plasmid is a Gateway system (Invitrogen) modified version of the pACT plasmid (Promega). This plasmid can used to generate an N terminal VP 16 fusion expression vector. This is achieved via a gateway system LR reaction with an entry vector that has an in frame cDNA clone set between attL sites, the LR reaction exchanges the DNA cassette between the two attR sites in the pACT Gateway plasmid for the cDNA clone cassette. CAMr = Chloramphenical resistance gene; AMPr = Ampicillin resistance; ccdB = suicide gene.

Figure 4 shows a plasmid map of the plasmid that has a GAL4 DBD, it can be used to generate the second expression vector, e.g. DNA binding domain linked to ligand binding domain. The pBIND Gateway plasmid is a gateway system (Invitrogen) modified version of the pBIND plasmid (Promega). This plasmid can used to generate an N terminal GAL4

TM fusion expression vector. This is achieved via a Gateway system LR reaction with an entry vector that has an in frame cDNA clone set between attL sites, the LR reaction exchanges the DNA cassette between the two attR sites in the pBIND Gateway plasmid for the cDNA clone cassette. CAMr = Chloramphenical resistance gene; AMPr = Ampicillin resistance; ccdB = suicide gene. Note Renilla gene that is used for co-transfection control.

Figure 5 shows a plasmid map of the reporter plasmid. The pG5ZsGreen reporter plasmid is modified version of the pG5 Luc plasmid (Promega) where the luciferase gene has been

swapped for the ZsGreen gene. This plasmid has multiple GAL4 binding sites and the reporter gene responds to VP 16 transcriptional activation. AMP — Ampicillin resistance.

Figure 6. DNA (40ng/ml of each plasmid) was spotted onto slides using a "hand spotter" (Greiner) forming a 500μm diameter spot. DNA spot plan: samples 1. & 2.= pG5 ZsGreen reporter, pBIND DHFR(E.coli), pACT GRHBD(mut); sample 3. = pCMV ZsGreen (control plasmid that constitutively expresses ZsGreen).

TM

Slides were prepared for transfection using Effectene transfection reagent (Qiagen) as per the Sabatini protocol. The slides were placed in a tissue culture dish and 10 ml of

5.OxIO ⁵ /ml HEK293T cells were added.

(A). Cells treated with 100OnM DEX-ClO-MTX heterocompound after 72hours.

(B). Untreated cells after 72hours.

Transfection signal was noted within 18 hours, and reached a peak after 72hours. GRHBD = glucocorticoid hormone receptor binding domain.

MATERIALS & METHODS

Cell Viability Assay

TM 0 HEK293T cells were seeded at 5x10 in 100ml of DMEM (plus Glutamax ) (Invitrogen) with 10% foetal calf serum per well of a black walled 96 well plate. These cells were treated with MTX-ClO-DEX, DEX or MTX over an eight-point log titration of compound from 0-1000OnM. Cells were incubated for 24 hours at 37 ⁰ C and cell viability was assessed

TM using the luminescence based Cell Titer-Glo assay (Promega) as per the manufacturers 5 instructions. In this assay luminescence is directly proportional to cell viability so relative luciferase units (RLU) were used to express cell viability.

Mammalian three-hybrid using Reverse Transfection

DNA (40ng/μl of each plasmid) was spotted onto slides using a "hand spotter" (Greiner) o forming a 500mm diameter spot. DNA spots either contained: 1. The pCMV ZsGreen

plasmid. This plasmid constitutively expresses ZsGreen and was used as a control plasmid; or 2. The pG5 ZsGreen plasmid used as the fluorescent reporter plasmid, the pBIND DHFR (E.coli) plasmid used as the expression plasmid for VP16-DHFR chimeric protein and the p ACT GRHBD (mut) plasmid used as the expression plasmid for the GAL4-GR chimeric protein. Spots were allowed to dry at room temperature for one hour prior to transfection.

TM

Slides were prepared for transfection using Effectene transfection reagent (Qiagen) as

TM per the Sabatini protocol. Briefly for each slide, 16ml of Effectene enhancer was added to 150ml of EC buffer, mixed and incubated at room temperature for 5 minutes; 25ml of

TM TM

Effectene transfection reagent was added to the previous step; a HybriWell (Grace

TM Biolabs) was attached to the glass slide over the array and the Effectene mix was added

TM to this and incubated with the array for 20 minutes; the HybriWell was then removed.

The slides were placed in a tissue culture dish and 10 ml of 5.OxIO ⁵ /ml HEK293T cells were added. Cells treated with either 100OnM DEX-ClO-MTX or control vehicle for 72hours. Transfection signal was noted within 18 hours, and reached a peak after 72hours. Green fluorescence was observed and recorded at 496/506nm (excitation/emission) on an Axiovert inverted microscope (Zeiss).

EXAMPLES Plasmid DNA was diluted in 0.2% gelatin prior to spotting as per the Sabatini protocol, WO 01/20015 Al. We spotted DNA from different plasmids (40ng/μl of each plasmid) slides using a "hand spotter" (Greiner) forming a 500mm diameter spot. The DNA spot plan was as follows: Spot 1. & 2.= pG5 ZsGreen reporter, pBIND DHFR(E.coli), pACT GRHBD(mut); Spot 3. = pCMV ZsGreen (control plasmid that constitutively expresses ZsGreen).

The pCMV ZsGreen plasmid was used as a control plasmid that constitutively expresses ZsGreen. The pG5 ZsGreen plasmid was used as the fluorescent reporter plasmid, the pBIND DHFR(E.coli) plasmid was used as the expression plasmid for VP16-DHFR chimeric protein and the pACT GRHBD(mut) plasmid was used as the expression plasmid

for the GAL4-GR chimeric protein. Plasmid DNA spots were allowed to dry at room temperature for one hour prior to transfection.

TM

Slides were prepared for transfection using Effectene transfection reagent (Qiagen) as per the Sabatini protocol, WO 01/20015 Al. The slides were placed in a tissue culture dish 5 andl 0 ml of 5.Ox 10 ⁵ /ml HEK293T cells were added. Cells treated with either 100OnM DEX-ClO-MTX or control vehicle for 72hours.

Transfection signal was noted within 18 hours, and reached a peak after 72hours. After 72hours green fluorescence was observed and recorded at 496/506nm (excitation/emission) on an Axiovert inverted microscope (Zeiss). 0

RESULTS

Cell Viability Assay

HEK293T cells that were treated with MTX-ClO-DEX or DEX showed no decrease in s viability with any of the tested doses compared to no treatment. This is in contrast to the HEK293T cells that were treated with MTX. The MTX treated cells showed significant loss of viability with doses of InM and above of MTX.

Mammalian three-hybrid usinfi Reverse Transfection o Groups of cells on the HEK293T layer, grown over the array of plasmid DNA on the glass slide, were transfected with the array DNA by reverse transfection. This formed clusters of transfected cells within the 500mm diameter of the original DNA spot. These clusters of cells formed arrays that mirrored the original position of the plasmid DNA spots.

Clusters of cells transfected with pCMVZsGreen exhibited very intense fluorescence 5 indicating high levels of ZsGreen expression (Figure 6). There was no variation in ZsGreen signal from pCMVZsGreen transfected cells between MTX-ClO-DEX treatment and vehicle control treatment.

Clusters of cells transfected with pG5 ZsGreen, pBIND DHFR(E.coli), pACT o GRHBD(mut) showed no signal from the ZsGreen reporter when treated with the vehicle control. However when the same clusters of cells were treated with MTX-ClO-DEX then

there was activation of the reporter gene and generation of ZsGreen signal (Figure 6.)- These observations show that VP 16-GR can only be recruited to the reporter gene in the presence of MTX-ClO-DEX.