Interruption of binding of MDM2 and p53 protein and therapeutic application thereof

Abstract

A method of identifying a compound which interferes with the binding of MDM2 to human p53 has been determined. This method comprises forming a mixture between MDM2 and a fragment of human p53 consisting of 6 to 28 amino acids comprising TFSDLW (SEQ ID NO:2), adding a test compound to the mixture and determining the quantity of protein bound to the other before and after adding the compound. A compound which decreases the amount of binding of the two proteins to each other is a compound which interferes with the binding of MDM2 to human p53.

Description

FIELD OF THE INVENTION

The invention relates to the area of cancer detection and therapeutics. More particularly it relates to the prevention or disruption of the inactivation of the p53 tumour suppressor which occurs as a result of the binding of a protein more particularly any protein that binds to p53 through the amino acid motif within the region of p53 represented by amino acids 16-30 QETFSDLWKLLPENN (SEQ ID NO:1) of the human p53 protein. More particularly an example of such a protein is the oncogene protein MDM2 (human MDM2).

BACKGROUND OF THE INVENTION

Inactivation of the p53 tumour suppressor is a frequent event in human neoplasia. The inactivation can occur by mutation of the p53 gene or through binding to viral or cellular oncogene proteins, such as the SV40 large T antigen and MDM2. While the mechanism through which wild-type p53 suppresses tumour cell growth is as yet poorly defined it is clear that one key feature of the growth suppression is the property of p53 to act as a transcription factor (Farmer et al., 1992; Funk et al., 1992; Kern et al., 1992). Currently, considerable effort is being made to identify growth control genes that are regulated by p53 binding to sequence elements near or within these genes. A number of such genes have been identified. In cases such as the muscle creatine kinase gene (Weintraub et al., 1991; Zambetti et al., 1992) and a GLN retrovital element (Zauberman et al., 1993) the role these genes might play in the suppression of growth control is unclear. Yet there are other examples, namely mdm2 (Barak et al., 1993; Wu et al., 1993), GADD 45 (Kastan et al., 1992) and WAF1 or CIP1 (El-Deiry et al., 1993; Harper et al., 1993), where their involvement in the regulation of cell growth is better understood.

In the present text "mdm2" refers to the oncogene and "MDM2" refers to the protein obtained as a result of expression of that gene.

Mdm2, a known oncogene, was originally found on mouse double minute chromosomes (Cahilly-Snyder et al., 1987). Its protein product was subsequently found to form a complex with p53, which was first observed in a rat fibroblast cell line (Clone 6) previously transfected with a temperature sensitive mouse p53 gene (Michalovitz et al., 1990). The rat cell line grew well at 37.degree. C. but exhibited a G1 arrest when shifted down to 32.degree. C., which was entirely consistent with an observed temperature dependent switch in p53 conformation and activity. However, the p53-MDM2 complex was only observed in abundance at 32.degree. C., at which temperature p53 was predominantly in a functional or "wild-type" form (Barak and Oren, 1992; Momand et al., 1992). By shifting the rat cell line down to 32.degree. C. and blocking de novo protein synthesis it was shown that only "wild-type" p53 induced expression of the mdm2 gene, thereby accounting for the differential abundance of the complex in terms of p53 transcriptional activity (Barak et al., 1993). The explanation was further developed by the identification of a DNA binding site for wild-type p53 within the first intron of the mdm2 gene (Wu et al., 1993). Reporter constructs employing this p53 DNA binding site revealed that they were inactivated when wild-type p53 was co-expressed with MDM2.

This inhibition of the transcriptional activity of p53 may be caused by MDM2 blocking the activation domain of p53 and/ or the DNA binding site. Consequently, it was proposed that mdm2 expression is autoregulated, via the inhibitory effect of MDM2 protein on the transcriptional activity of wild-type p53. This p53-mdm2 autoregulatory feedback loop provided a novel insight as to how cell growth might be regulated by p53. Up to a third of human sarcomas are considered to overcome p53-regulated growth control by amplification of the mdm2 gene (Oliner et al., 1992). Hence the interaction between p53 and MDM2 represents a key potential therapeutic target.

The cDNA sequence encoding the human MDM2 protein (which is also referred to as "HDM2" in the art) is known from WO/93/20238. This application also discloses that human MDM2 protein binds with human p53 and it has been suggested that molecules which inhibit the binding of MDM2 to p53 would be therapeutic by alleviating the sequestration of p53. However it is also suggested that the p53 and MDM2 binding site is extensive, including amino acid residues 13-41 of p53 as well an additional nine to thirteen residues at either the amino or carboxyl terminal side of the peptide are also involved. This would indicate that a large polypeptide or other large molecule would be required in order to significantly interfere with the binding.

The applicants have therefore sought to immunochemically characterize the p53-MDM2 complex, and also determine in fine detail the MDM2 binding site on p53.

Surprisingly, it has been found that only a relatively small number of amino acids within the p53 protein are involved in binding to MDM2.

SUMMARY OF THE INVENTION

The precise identification of this binding site is vital to allow the rational design of molecules which will disrupt or prevent binding between p53 and MDM2 or proteins containing analogous p53 binding sites. In addition it allows for the design of screening procedures which will enable compounds which can disrupt or prevent the binding interaction to be accurately and rapidly identified.

The applicants have found that the site on the p53 protein which is responsible for binding to MDM2 is a small sequence of only six amino acids. This sequence is represented by the sequence TFSDLW (SEQ ID NO:2) in human (amino acids 18-23 in the sequence) and TFSGLW (SEQ ID NO:3) (amino acids 18-23) in mouse. By disrupting or preventing p53 from binding in this specific region, the deleterious effects of binding to MDM2 or proteins having an analogous p53 binding site can be avoided. Proteins having a p53 binding site which is analogous to that of MDM2 will generally comprise oncogene proteins which bind to p53 through the amino acid motif within the region of p53 represented by amino acids 16-30 (QETFSDLWKLLPENN) (SEQ ID NO:1) of the human p53 protein.

Hence the present invention provides a method for interfering with the binding between p53 and MDM2 or an oncogene protein having an analogous p53 binding site, which method comprises administering a effective amount of a compound, selected from the group consisting of a peptide having up to twenty eight amino acids which is able to disrupt or prevent the binding between p53 and MDM2, or a functional peptide analogue thereof.

It may be expected that small peptides, for example of from 4 to 10 amino acids, suitably from six to 10 amino acids, or peptide analogues thereof would be particularly suitable in such a process. Peptides which would be of particular interest are those which comprise a fragment of p53 protein which includes at least some of amino acids 18-23 within the sequence of human p53, as identified in WO93/20238 or a peptide analogue thereof. Suitably these peptides are those which are circular, linear or derivatised to achieve better penetration of membranes.

Novel peptides or peptide analogues of this type form a further aspect of the invention.

Hence preferred peptides include the sequence TFSDLW (SEQ ID NO:2) or a portion thereof.

Other compounds which may interfere with the binding include organic compounds which are modelled to achieve the same three dimensional structure as the said region of the p53 peptide. Hence in an alternative embodiment the invention provides an organic compound which is modelled to resemble the three dimensional structure of the sequence TFSDLW (SEQ ID NO:2) as it appears in human p53 and which binds to human MDM2.

A suitable oncogene protein is MDM2 but the disruption of binding of p53 to other oncogene proteins containing a p53 binding site analogous to that of MDM2 are included within the scope of the present invention. For example, the adenovirus EIB 58 kD protein, the Tata box binding protein TBP and the transcription factor of the E2F family.

As used herein the expression `peptide analogue` refers to peptide variants or mimetics which are chemical compounds such organic compounds having the same functional activity as the peptide in question, in particular which interfere with the binding between p53 and MDM2. Examples of such analogues will include chemical compounds which are modelled to resemble the three dimensional structure of the sequence TFSDLW (SEQ ID NO:2) as it appears in human p53 and which binds to human MDM2.

In order to identify compounds which are useful in the above described methods, compounds may be screened for interference of the MDM2/p53 interaction where one of the binding partners is bound to a solid support and the other partner is labelled. Hence the invention further provides a method of identifying compounds which interfere with the binding of human MDM2 to human p53, said method comprising

binding a predetermined quantity of a first protein which is detectably labelled to a second protein, adding a compound to be tested; and determining the quantity of the first protein which is displaced from or prevented from binding to the second protein; wherein one of the first protein or the second protein is MDM2 or an oncogene protein having a p53 binding site analogous to that of MDM2, and the other is a fragment of human p53 of from six to twenty eight amino acids including the amino acid residues 18-23 in the sequence of human p53 as set out in WO/93/20238, or a peptide analogue thereof.

Suitably in the method, the oncogene protein is human MDM2 and the other protein comprises a fragment of human p53 of from 12 to 28 amino acids including the sequence TFSDLW (SEQ ID NO:2).

This method can be readily adapted to provide a high throughput screen, for example by carrying out the process in a 96-well format. Automated screening techniques can be applied in these circumstances as would be understood in the art. Compounds from various sources can be screened in large numbers. One potential source of compounds are the available synthetic combinatorial peptide libraries.

Compounds identified by this screening method form a further aspect of the invention.

Methods of treatment of conditions such as cancer and other malignancies are envisaged by the administration of the compounds of the invention.

Hence the invention also provides a method for inhibiting the growth of tumour cells which contain a human MDM2 gene amplification which method comprises administering a effective amount of a compound which interferes with the binding between p53 and an MDM2, said compound being selected from the group consisting of a peptide having up to twenty eight amino acids which is able to disrupt or prevent the binding between p53 and MDM2, or a functional peptide analogue thereof.

Preferably in the above-described method of treatment, the compound is a peptide of from six to twenty eight amino acids which coincide with a region of human p53 and includes the sequence TFSDLW (SEQ ID NO:2).

Alternatively, the compound used in the method is peptide analogue such as an organic compound which binds to the same site on MDM2 as the sequence TFSDLW (SEQ ID NO:2).

For use in these applications, compounds of the invention are suitably applied in the form of compositions with pharmaceutically acceptable carriers. These may be solid or liquid for carriers and the compositions suitable for oral or parenteral application as would be understood in the art. Dosages of the compounds will depend upon the patient, the particular condition and the nature of the specific compound chosen. For example, when the compound is a peptide fragment dosages of from 0.1 to 10 mg/Kg may be effective.

It has been suggested (Picksley and Lane 1993) that mdm2 expression is autoregulated in a feedback loop, via the inhibitory effect of MDM2 protein on the transcriptional activity of wild-type p53. Any interference with the binding between p53 and MDM2 in accordance with the present invention, will affect the p53-MDM2 autoregulatory loop. Given p53's role as guardian of the genome, compounds which have such an effect could enhance the activity of other therapeutic agents.

Hence in a further aspect the invention comprises a pharmaceutical composition comprising synergistic amount of a compound of the invention in combination with another anticancer therapeutic agent.

DNA encoding an MDM2-binding, p53 derived peptide, or multiple copies thereof may also be administered to tumour cells as a mode of administering the peptide. Hence the invention provides a method for inhibiting the growth of tumour cells which contain a human MDM2 gene amplification, the method comprising applying to said tumour cells a DNA molecule which expresses a polypeptide comprising a portion of p53 or a variant thereof, said portion comprising amino acids 18-23 of p53, said polypeptide being capable of binding to human MDM2.

The DNA will typically be in an expression construct, such as a retrovirus, DNA virus, or plasmid-vector, which has the DNA elements necessary for expression properly positioned to achieve expression of MDM2-binding peptide. The DNA can be administered inter alia encapsulated in liposomes, or in any other form known to the art to achieve efficient uptake by cells.

By identifying the binding site so specifically, the applicants have opened up the possibility of making small therapeutic compounds which will target this site specifically. This is advantageous since small molecules are more likely to be able to penetrate into a cell and hence be therapeutically active. Furthermore the diagnostic process can be effected more accurately and using simpler molecules as a result of this discovery.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1A-1B show Western blots of immunoprecipitates of MDM2, p53 and the MDM2-p53 complex obtained from Clone 6 cells.

FIGS. 2A-2B are graphs showing the results of two-site immunoassays to determine the levels of MDM2, p53 and MDM2-p53 complex in Clone 6 cells grown at 32.degree. C. for 24 hours (A) or continuously at 37.degree. C. (B).

FIGS. 3A-3C are graphs showing the results of binding MDM2 to a peptide library as determined by ELISA assay using monoclonal antibody 4B2. The library of human and mouse p53 had been challenged with insect cell extract alone (SF9) and insect cell extract expressing mouse MDM2 (SF9 Mus MDM2. The results of peptides numbers 3-50 of the N-terminal to mid region of human p53 are shown in 3A, and the remainder of the human p53 amino acid sequence and the N-terminal sequence of mouse p53 in 3B. 3C shows the results from a control experiment using certain peptides from B in order to verify the specificity of the detecting antibody 4B2.

FIG. 4 identifies the peptide sequences (SEQ ID NOS:1 and 4-6) to which MDM2 bound (SEQ ID NOS:1,4-6), and defines the consensus binding site (SEQ ID NOS:7-8) on human and mouse p53 (SEQ ID NOS:7 and 8).

FIGS. 5A-5C show the key residues (SEQ ID NOS:9-21) required for binding MDM2 (SEQ ID NOS:1 and 9-21), antibody DO-1 (SEQ ID NOS:9-20) and antibody Bp53-19 (SEQ ID NOS:9-20) respectively.

DETAILED DESCRIPTION OF THE INVENTION

The first indication of an interaction between MDM2 protein and p53 protein emerged from work on a rat cell line, Clone 6, which expressed a temperature sensitive mutant form of mouse p53 (Barak and Oren, 1992; Michalovitz et al., 1990; Momand et al., 1992). MDM2 was readily observed to form a complex at 32.degree. C. with p53 but was just detectable when cells were grown at 37.degree. C.

The formation of a p53-MDM2 complex in Clone 6 cells at 32.degree. C. and 37.degree. C. was re-examined in a quantitative manner. The results confirm previous immunoprecipitation observations that the level of MDM2 at the lower temperature is significantly elevated, approximately 10-30 fold greater than that at 37.degree. C., at which temperature MDM2 is only just detectable. Consequently, the p53-MDM2 complex is readily observed at 32.degree. C. and not at 37.degree. C. The levels of p53 also vary at the two different temperatures. However, the p53 levels are elevated approximately five fold at 37.degree. C. as compared with that at 32.degree. C.,--the opposite behaviour to that of MDM2. Accordingly, the difference in the levels of p53 and MDM2 are likely to have alternative explanations. In the case of MDM2 other groups have established that the increase of MDM2 at 32.degree. C. is due to increased transcription of MDM2 due to a conformational change in p53 to a presumed transcriptional active form (Barak et al., 1993; Wu et al., 1993). The same explanation does not apply for p53 even though wild type p53 is required for p53 expression (Deffie et al., 1993), and is probably explained by the increased half life of the mutant conformation of p53 at 37.degree. C. (Gannon and Lane, 1991). Data described herein after using both direct observation of the p53-MDM2 complex by ELISA and immunoprecipitation combined with the indirect inference of the loss of the Bp53-19 epitope suggested that nearly all p53 molecules are complexed to excess MDM2 protein in C6 cells at 32.degree. C. This is not consistent with the powerful p53 dependant transcriptional response seen in these cells at this temperature and suggests that either that complexing to MDM2 is unable to completely inactivate p53 in vivo or that small amounts of "free" p53 may be very active. The complex between p53 and MDM2 may be regulated in cells to release functional p53 at the individual cell level perhaps as a cell cycle dependant response.

The present invention is based upon the identification of the minimal MDM2 binding site to be TFSD/GLW (SEQ ID NOS:2-3). This site is in a location broadly reported by other groups to be the MDM2 binding domain of p53, specifically aa1-41 and 13-57 (Oliner et al., 1993), aa1-52 (Chen et al., 1993) and aa1-159 (Brown et al., 1993.) Notably, a construct generated by Oliner and co-workers encompassing aa13-41 of p53 was not sufficient for MDM2 binding in a three hybrid protein system, and differs from our observations. The disparity might be explained by the close proximity of the fusion protein sequence adjacent to the TFSDGLW (SEQ ID NO:22) sequence at aa18-23 as our data does show that flanking sequences do contribute in a minor way to MDM2 binding. The TFSD/GLW (SEQ ID NOS:2-3) sequence is very closely adjacent to the transactivation domain aa20-42 (Unger et al., 1992), and as shown by others the binding of MDM2 to this site interferes with the transcriptional activity of p53 (Oliner et al., 1993). While substitution analysis of the MDM2 binding site on p53 identified the TFSD/GLW (SEQ ID NOS:2-3) sequence to be the key region required for MDM2 to bind p53, other residues flanking this site also contribute in a minor way to MDM2 binding, but clearly the TFSD/GLW (SEQ ID NOS:2-3) sequence is a minimal target for agents that might disrupt complex formation without effecting the transactivation activity (for which as yet the key residues are undetermined). The first two residues TF are part of the conserved box I, and the latter four SD/GLW (SEQ ID NOS:23-24) are outside but are also part of a region of p53 that is conserved from Xenopus to man.

The corresponding binding site on MDM2 for p53 has variously been reported to be between aa1-121, 19-102 (Chen et al., 1993) together with aa102-294 or 249-491, and also 1-221 (Brown et al., 1993). Notably, a monoclonal antibody against the N-terminal region of human MDM2, 3G5 (maps at aa59-89) is able to immunoprecipitate MDM2 but not co-immunoprecipitate p53 (Chen et al., 1993), an analogous observation to our findings with antibody Bp53-19.

The binding of MDM2 to p53 peptides has obvious parallels to a similar study that used small peptides to identify the binding sites of Adenovirus E1A and human papilloma virus E7 for a range of proteins including retino-blastoma protein, p107, cyclin A and p130 (Dyson et al., 1992a,b). The MDM2 binding site on p53, appears to be a single domain rather than two domains as in the case of E1A and E7. The MDM2 binding site on p53 overlaps precisely with a highly immunogenic epitope on the protein; many independently isolated monoclonal antibodies to p53 recognise the site, and antibodies to it are present in the sera of cancer patients (Schlichtholtz et al., 1993) This suggests that it has an exposed and defined structure. It is possible that the amino acid sequence of the complementarity determining regions of these antibodies will show homology to the p53 binding site of MDM2. It also suggests that anti-p53 antibodies used to examine p53 levels where high levels of MDM2 are present must be chosen with care. Binding of MDM2 to this site may be regulated by phosphorylation since there is a DNA-dependent kinase site at serine 20 (Lees Miller et al., 1990) and other phosphorylation sites at serine 6, 9 and 15 (Samad et al., 1986; Meek and Eckhardt, 1988).

The following examples are provided to exemplify various aspects of the invention and are not intended to limit the scope of the invention.

In these examples, the following materials and methods were used.

MATERIALS AND METHODS

Cell Culture

Clone 6 cells (Michalovitz et al., 1990) were grown in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% FCS at either 32 or 37.degree. C. The Spododoptera frugiperda cell line, SF9, was grown at 27.degree. C. in ExCell 400 medium (J. R. H. Biosciences, Sera-Lab, UK) supplemented with 5% FCS and glutamine.

Expression of MDM2 in Insect Cells

The mouse mdm2 gene was obtained from a mouse prostate cell line (Lu et al., 1992) by polymerase chain reaction and then cloned into a Spododoptera frugiperda expression vector pVL1393 using standard DNA and baculovirus expression techniques. Art expression clone was identified by the production of a 90-95 kDa protein that was recognized by anti-MDM2 antibodies.

Antibodies

p53 protein was detected using the polyclonal sera CM1 (Midgley et al., 1992), or monoclonal antibodies PAb421 (Harlow et al., 1981) and Bp53-19 (Bartek et al., 1993). MDM2 was detected using rabbit anti-MDM2 polyclonal sera (Barak et al., 1993) or monoclonal antibody 4B2 (Chen et al., 1993) and SMP14 (a previously unreported monoclonal antibody raised by us against a peptide, CSRPSTSSRRRAISE (SEQ ID NO:25), containing part of the human MDM2 sequence from aa154 to 167 (Oliner et al., 1992), the first cysteine is not part of the MDM2 sequence but was added to provide an extra coupling option). An antibody, PAb419, raised against SV40 large T antigen (Harlow et al., 1981) was used as an irrelevant control for immunoprecipitations.

Immunoprecipitation

Cells were lysed in ice-cold NET buffer (50 mM Tris-HCI, pH8.0, 150 mM NaCl, 5 mM EDTA, 1% NP40) containing 1 mM phenylmethylsulphonyl fluoride, for 30 min at 4.degree. C. Debris was removed from the cell extract by centrifugation at 14,000 rpm in a refridgerated Eppendorf centrifuge. The immunoprecipitation procedure was essentially as previously described (Gannon et al., 1990) using 1 .mu.g of purified mouse monoclonal antibody, and Protein G Sepharose beads (Pharmacia) for both pre-absorption of the cell extracts and subsequent isolation of the antibody-protein complex.

Screening of p53 Peptide Library

Peptide libraries of the entire human p53 protein and a partial N-terminal region of the mouse p53 protein was obtained from Chiron Mimotopes P/L (Victoria, Australia). The libraries were in the form of 15 mer peptides linked to biotin via an additional peptide spacer region of serine-glycine-serine-glycine, and each peptide shared a 5 amino acid overlap with the previous peptide in the primary sequence. ELISA plates were coated with 100 .mu.l of 5 .mu.g/ml streptavidin (Vector labs) per well and incubated overnight at 37.degree. C. and then blocked with phosphate buffered saline (PBS) containing 2% bovine serum albumin (BSA) for 1 hour at room temperature. The stock biotinylated peptides were diluted to 5 .mu.g/ml in PBS containing 0.1% BSA and 50 .mu.l of each were plated into designed wells and then incubated at room temperature for 1 hour. The plates were washed four times with PBS containing 0.1% Tween 20 before addition of the cell extract (50 .mu.l of 1-4 mg/ml per well) or purified protein. The plates were incubated at 4.degree. C. for 2-3 hours, before washing four times with PBS containing 0.1% Tween 20 to remove unbound protein. In the case of cell extracts bound protein was detected with the appropriate primary antibody at 1-3 .mu.g/ml, and followed by an anti-mouse horse radish peroxidase conjugate and 3'3'4'4'-tetramethyl benzidine (TMB) substrate as in the standard ELISA assay (Harlow and Lane, 1988).

The levels of p53, MDM2, and complexes thereof were determined by a two site immunoassay using stated antibodies. Mouse monoclonal antibodies were used as the solid phase by incubating Falcon microtitre dish wells with 50 .mu.l of a 30 .mu.g/ml solution of purified antibody overnight at 4.degree. C. The plates were blocked with 2% bovine serum albumin in PBS for 2 h at room temperature, and washed with PBS. Cell extracts were prepared as described for immunoprecipitations and then serially two-fold diluted before adding 50 .mu.l per well and incubating at 4.degree. C. for two hours. The plates were then washed with 0.1% NP-40 in PBS, before addition of 50 .mu.l of detecting polyclonal antisera at 1/1000 dilution. The plates were washed again with 0.1% NP-40 in PBS and 50 .mu.l of 1/1000 dilution peroxidase conjugated swine anti-rabbit Ig serum (DAKO) was added for 2 h, then visualised by the TMB reaction.

EXAMPLES

Example 1

Immunoprecipitation of MDM2, p53 and the MDM2-p53 complex

The observation that the rat cell line, Clone 6 expressed a temperature sensitive mutant form of mouse p53 was reexamined using a panel of p53 monoclonal antibodies.

Western blots were obtained of immunoprecipitates of MDM2, p53 and the MDM2-p53 complex from Clone 6 cells grown at 32.degree. C. for 24 hrs (FIG. 1A) or continuously at 37.degree. C. (FIG. 1B). The immunoprecipitates were obtained using 1 .mu.g of purified antibody which were as follows: in lanes 1 and 4,--PAb421; in lanes 2 and 5,--Bp53-19; and in lanes 3 and 6,--4B2. MDM2 was detected in lanes 1, 2 and 3 using SMP14 antibody supernatant and rabbit anti-mouse horse radish peroxidase conjugate; and p53 detected in lanes 4, 5 and 6 using a 1 in 200 dilution of DM-1 and swine anti-rabbit horse radish peroxidase conjugate. An irrelevant antibody, PAb419, did not immunopecipitate either MDM2 or p53 from cell extracts prepared at either 32.degree. C. or 37.degree. C. (data not shown). The molecular weight of the markers are given in kDa.

It was surprisingly found that one of the antibodies, Bp53-19, failed to immunoprecipitate p53 from Clone C6 cells grown at 32.degree. C. for 24 hours, but efficiently precipitated p53 from cells grown continuously at 37.degree. C. (compare FIG. 1A track 5 with FIG. 1B track 5), whereas PAb421 precipitated p53 at both temperatures (FIGS. 1A track 4 and 1B track 4). investigations were then carried out to determine whether Bp53-19 would co-immunoprecipitate MDM2 with p53. From the immunoprecipitation western data in FIGS. 1A and 1B it is clear that Bp53-19 does not co-immunoprecipitate MDM2 from cell extracts grown at 32 or 37.degree. C. (track 2 in FIGS. 1A and B). Other p53 antibodies such as PAb421 do however co-immunoprecipitate MDM2 with p53 at 32.degree. C. but not at 37.degree. C. (track 1 FIGS. 1A and B). Conversely, antibodies against MDM2 such as 4B2, FIG. 1, and SMP14 (data not shown) co-immunoprecipitate p53 at 32.degree. C. but not at 37.degree. C. (track 6 FIGS. 1A and B). The two bands recognized by 4B2 (and SMP14) at just below 80 kDa are truncated forms of rat MDM2, as full length migrates on an SDS-PAGE gel with an apparent relative molecular mass of 90 kDa, multiple forms of MDM2 are often observed (Chen et al., 1993).

Example 2

Two-site immunoassay to determine levels of MDM2, p53 and MDM2-p53 complex

Two-site immunoassays were carried out to determine the levels of MDM2, p53 and MDM2-p53 complex in Clone 6 cells grown at 32.degree. C. for 24 hrs (FIG. 2A) or continuously at 37.degree. C. (FIG. 2B). In FIG. 2A the coating antibodies were one of the following purified antibodies as stated in the figure legends: 4B2, 421 and Bp53-19, probed with rabbit anti-p53 serum CM1 or rabbit anti-MDM2 serum, and then detected using swine anti-rabbit horse radish peroxidase conjugate and TMB as substrate. At 37.degree. C. the MDM2-p53 complex was undetectable by any combination of antibodies.

The two-site immunoassays of the levels of MDM2, p53 and MDM2-p53 complex at 32.degree. C. and 37.degree. C. are consistent with the immunoprecipitation results of Example 1. A striking feature apparent from the data in FIG. 2A is that the levels of p53 and p53-MDM2 complex are very similar suggesting that most, but not all, p53 is in complex with MDM2 at 32.degree. C. The inability of Bp53-19 to detect a p53-MDM2 complex at 32.degree. C. is again notable since other combinations of antibodies are able to do so.

From comparison of the two-site immunoassays at 32.degree. C. and 37.degree. C. it is clear why MDM2 is not immunoprecipated at 37.degree. C., as the levels of MDM2 protein are very much lower and are only just detectable. No MDM2-p53 complex could be detected by the two-site immunoassay of cell extracts prepared at 37.degree. C., see FIG. 2B, where the data for the 4B2 (as the capture antibody) and CM1 (as the detecting antibody) combination of antibodies is shown (similarly antibodies PAb421 or Bp53-19 and rabbit anti-MDM2 polyclonal did not detect the complex). The diminished level of MDM2 at 37.degree. C., less than 10% of that at 32.degree. C., is in contrast to the situation with p53 which is elevated approximately 5 fold relative to the levels at 32.degree. C.

The explanation for the ability of PAb421 and 4B2 only being able to coprecipitate p53 and MDM2 together at 32.degree. C., but not at 37.degree. C. is consistent with difference in levels of MDM2 at the two temperatures, and also with the published observations that mdm2 expression is dependent on the "wild-type" form of p53 predominantly present at 32.degree. C.

The failure of Bp53-19 to co-immunoprecipitate MDM2 or detect the p53-MDM2 complex at 32.degree. C. is unexpected for two reasons. Firstly, the two-site assay suggests there is MDM2 protein in excess, which is able to form complexes with p53 as detected by the capturing antibodies PAb421 and 4B2. Secondly, the two-site immunoassay at 37.degree. C. suggests that Bp53-19 is almost as efficient as PAb421 at recognizing p53 in the cell extracts. The simplest interpretation for this observation is that Bp53-19 recognizes the same region on p53 that MDM2 binds to.

Example 3

Identification of MDM2-p53 binding site

It has previously been shown that Bp53-19 and MDM2 interact with the amino acid terminal end of p53 (Stephen et al manuscript in preparation; Oliner et al., 1993). A complete peptide library of the human p53 protein, and a partial peptide library of the mouse p53 protein were available to identify the region to which MDM2 binds. The human p53 sequence starts at peptide number 3 and ends at peptide 79, and each peptide consists of 15 amino acids, with the last five amino acids being present in the next peptide along. The mouse p53 sequence is partial and consists of the N-terminal sequence from amino acid 1-92, again each overlapping the next and previous peptide by five amino acids.

These libraries consisted of 15 amino acid long sections of the p53 primary amino acid sequence, that consecutively overlapped by 5 amino acids, and were each attached to biotin via a 4 amino acid long spacer. By immobilizing the biotinylated peptides on streptavidin coated ELISA plates the MDM2 binding site on p53 could be quickly identified if it was encompassed within a stretch of fifteen amino acids or less. Extract containing MDM2 was added to an ELISA plate with the peptide library bound to it, and the bound MDM2 protein was later detected using monoclonal antibody 4B2 and the standard ELISA assay. Several sources of recombinant MDM2 protein were used to challenge the p53 library, these included crude extracts and partially purified preparations of human and mouse MDM2 expressed in E.coli and also mouse MDM2 expressed in insect cells;--all forms identified the same peptides in the p53 library. The results using the mouse MDM2 expressed in insect cells are shown in FIGS. 3A and B. The peptide library was challenged with insect cell extract alone, SF9, and insect cell extract expressing mouse MDM2, SF9 Mus MDM2. Binding of MDM2 to the peptides was determined by an ELISA assay using monoclonal antibody 4B2, and then detecting bound antibody with rabbit anti-mouse Ig conjugated horse radish peroxidase and TMB substrate. In FIG. 3C is shown the results from a control experiment using peptides 59, 71, 83 and 95, as used in FIG. 3B but conducted in the presence or absence of extract to verify the specificity of the detecting antibody, 4B2. The results are presented alongside the ELISA readings for extract of insect cells alone not expressing mouse MDM2. The specificity is remarkable,--suggesting a strong interaction between MDM2 and p53 derived peptides. From the controls shown in FIG. 3C it can be seen that the binding is only observed in the presence of extract expressing MDM2, and is not due to the antibody recognizing the peptide alone, moreover identical results were obtained using SMP14 as the primary detecting antibody (data not shown).

The four peptides that bind MDM2 are shown in FIG. 4. Peptides 5 and 6 identify a site at the N-terminal end of human p53, whereas peptides 83 and 84 identify the corresponding region in the N-terminal end of mouse p53. Collectively, these four peptides define the consensus MDM2 binding site on p53 to be --QETFSD/GLWKL--(SEQ ID NOS:7-8), the aspartate to glycine being the only amino acid difference between the human and mouse sequence. The peptides involved in binding MDM2 are also those recognized by the p53 antibodies DO-1 and Bp53-19 (Stephen et al, manuscript in preparation).

To define key residues on p53 that are involved in the interaction with MDM2 a form of the consensus binding site sequence--QETFSDLWKL--(SEQ ID NO:7) was modified by substituting alanine at each position in the sequence and determining what effect this had on the binding of MDM2 from the insect cell extract expressing MDM2. This experiment was conducted in concert with examining the effect on binding of the antibodies DO-1 and Bp53-19. The results are presented in FIG. 5. The amino acid sequences are as stated. In A the sequence QETFSDLWKLLPENN (SEQ ID NO:1) represents the sequence of peptide 6 from FIG. 3 and SPDDIEQWFTEDPGP (SEQ ID NO:21) is an irrelevant peptide control. Formally the first serine residue on the stated peptide is part of the spacer coupling the consensus peptide to biotin, since serin also precedes the consensus p53 sequence this residue was also substituted with alanine. With regard to MDM2 binding all alanine substitutions in the consensus binding site reduce the level of binding as measured by ELISA, however, the key residues would appear to be TFSDLW (SEQ ID NO:2) as substitutions in these positions reduce the amount of MDM2 binding to less than 15% of that seen with the unchanged consensus sequence. interestingly, a higher level of binding of MDM2 is observed to the smaller consensus peptide rather than to peptide 6 (QETFSDLWKLLPENN) (SEQ ID NO:1) of the p53 peptide library reaffirming the definition of the binding site. In the case of monoclonal antibody DO-1 binding to the consensus sequence the key residues are ETFSDLK (SEQ ID NO:26), with D and K being the most crucial. The importance of the aspartate residue to the DO-1 epitope is consistent with the report that DO-1 only recognizes human p53 and not mouse p53,--the only difference being an aspartate to glycine change. While this difference has a critical affect on DO-1 binding it does not grossly affect the interaction of MDM2 .DELTA. with the protein or peptides. However substitution of alanine for aspartate at this position blocks binding of all three protein ligands. The ability of MDM2 to distinguish alanine from either glycine or aspartate at this position may imply that the polar environment of this region of the binding site is critical for the interaction. It has also been established from phage display libraries that the epitope of DO-1 is FSDLWKL (SEQ ID NO:27) (Stephen et al, manuscript in preparation), which is in agreement with our observations on key residues. For the antibody Bp53-19 the alanine substitution series identifies the key residues to be F--DLW-- with the latter three residues being the most crucial, and is similar to the requirements for MDM2 binding to the consensus binding site. Not surprisingly, it was found that the pre-binding of antibody Bp53-19 onto he SQETFSDLWKL (SEQ ID NO:9) biotinylated peptide blocked binding of MDM2 to the peptide when added later (data not shown).

REFERENCES

Barak, Y., et al. (1993). Embo J., 12, 461-468. Barak, Y. et al. (1992). Embo J., 11, 2115-2121. Bartek, J., et al. (1993). J. Pathol., 169, 27-34. Brown, D. R., et al. (1993). Mol. Cell. Biol., 13, 6849-57. Cahilly-Snyder., L., et al. (1987) Somatic Cell Mol. Genet. 13, 235-244. Chen, J., et al. (1993). Mol Cell Biol, 13, 4107-14. Deffie, A., et al. G. (1993). Mol. Cell. Biol., 13, 3415-3423. Dyson, N., et al. (1992a). J. Virol., 66, 4606-4611. Dyson, N., et al. (1992b). J. Virol., 66, 6893-6902. El-Beiry, W. S., et al. (1993). Cell, 75, 817-825. Farmer, G., et al. (1992). Nature, 358, 83-86. Funk, W. D. et al. (1992). Mol. Cell. Biol., 12, 2866-2871. Gannon, J. V., et al. (1990). Embo J., 9, 1595-1602. Gannon, J. V. et al. (1991). Nature, 349, 802-806. Harlow, E., et al. (1981). J. Virol., 39, 861-869. Harlow, E., et al (1988). Antibodies: a laboratory manual. New York. Cold Spring Harbor Laboratory Press. Harper, J. W., et al. (1993). Cell., 75, 805-816. Houghten, R. A., et al. (1991). Nature, 354, 84-86. Kastan, M. B., et al. (1992). Cell, 71, 587-597. Kern, S. E., et al. (1992). Science, 256, 827-830. Less-Miller, S. P. et al. (1990). Mol. Cell. Biol., 10, 6472-6481. Lu, X., et al (1992). Cell, 70, 153-161. Meek, D. W. et al. (1988). Mol. Cell. Biol., 8, 461-465. Michalovitz, D., et al. (1990). Cell, 62, 671-680. Midgley, C. A., et al. (1992). J. Cell. Sci., 101, 183-189. Momand, J., et al. (1992). Cell, 69, 1237-1245. Oliner, J. D., et al. (1992). Nature, 358, 80-83. Oliner, J. D., et al. (1993). Nature, 362, 857-860. Picksley and Lane. (1993). Bioessays. 15, 10, 689-690. Samad, A., et al. (1986). Proc. Natl. Acad. Sci. U.S.A., 83, 897-901. Schlichtholtz, B., et al. (1993). Cancer Res., 52, 6380-6384. Unger, T., et al. (1992). Embo J., 11, 1383-1390. Weintraub, H., et al. (1991). Proc. Natl. Acad. Sci. U.S.A., 88, 4570-4571. Wu, X., et al. (1993). Genes Dev., 7, 1126-1132. Zambetti, G. P., et al. (1992). Genes Dev., 6, 1143-1152. Zauberman, A., et al. (1993). Embo J., 12, 2799-2808. __________________________________________________________________________SEQUENCE LISTING(1) GENERAL INFORMATION:(iii) NUMBER OF SEQUENCES: 27(2) INFORMATION FOR SEQ ID NO:1:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 15 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:1:GlnGluThrPheSerAspLeuTrpLysLeuLeuProGluAsnAsn151015(2) INFORMATION FOR SEQ ID NO:2:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 6 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2:ThrPheSerAspLeuTrp15(2) INFORMATION FOR SEQ ID NO:3:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 6 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3:ThrPheSerGlyLeuTrp15(2) INFORMATION FOR SEQ ID NO:4:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 15 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4:GluProProLeuSerGlnGluThrPheSerAspLeuTrpLysLeu151015(2) INFORMATION FOR SEQ ID NO:5:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 15 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5:ProLeuSerGlnGluThrPheSerGlyLeuTrpLysLeuLeuPro151015(2) INFORMATION FOR SEQ ID NO:6:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 15 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:6:ThrPheSerGlyLeuTrpLysLeuLeuProProGluAspIleLeu151015(2) INFORMATION FOR SEQ ID NO:7:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 10 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:7:GlnGluThrPheSerAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:8:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 10 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:8:GlnGluThrPheSerGlyLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:9:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:9:SerGlnGluThrPheSerAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:10:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10:AlaGlnGluThrPheSerAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:11:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:11:SerAlaGluThrPheSerAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:12:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:12:SerGlnAlaThrPheSerAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:13:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:13:SerGlnGluAlaPheSerAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:14:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:14:SerGlnGluThrAlaSerAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:15:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:15:SerGlnGluThrPheAlaAspLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:16:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:16:SerGlnGluThrPheSerAlaLeuTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:17:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:17:SerGlnGluThrPheSerAspAlaTrpLysLeu1510(2) INFORMATION FOR SEQ ID NO:18:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:18:SerGlnGluThrPheSerAspLeuAlaLysLeu1510(2) INFORMATION FOR SEQ ID NO:19:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:19:SerGlnGluThrPheSerAspLeuTrpAlaLeu1510(2) INFORMATION FOR SEQ ID NO:20:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 11 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:20:SerGlnGluThrPheSerAspLeuTrpLysAla1510(2) INFORMATION FOR SEQ ID NO:21:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 15 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:21:SerProAspAspIleGluGlnTrpPheThrGluAspProGlyPro151015(2) INFORMATION FOR SEQ ID NO:22:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 7 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:22:ThrPheSerAspGlyLeuTrp15(2) INFORMATION FOR SEQ ID NO:23:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 4 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:23:SerAspLeuTrp(2) INFORMATION FOR SEQ ID NO:24:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 4 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:24:SerGlyLeuTrp1(2) INFORMATION FOR SEQ ID NO:25:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 15 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:25:CysSerArgProSerThrSerSerArgArgArgAlaIleSerGlu151015(2) INFORMATION FOR SEQ ID NO:26:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 7 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:26:GluThrPheSerAspLeuLys15(2) INFORMATION FOR SEQ ID NO:27:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 7 amino acids(B) TYPE: amino acid(C) STRANDEDNESS:(D) TOPOLOGY: linear(xi) SEQUENCE DESCRIPTION: SEQ ID NO:27:PheSerAspLeuTrpLysLeu15__________________________________________________________________________

Claims

1. A method of identifying a compound which interferes with the binding of MDM2 to human p53, said method comprising the steps of:

forming a mixture by combining a labeled first protein with a second protein, wherein one of the first protein or the second protein is MDM2 and the other is a fragment of human p53 consisting of 6 to 28 amino acids comprising the amino acid sequence TFSDLW (SEQ ID NO:2) or a peptide analogue thereof;

adding a test compound to the mixture; and

determining the quantity of the first protein which is bound to the second protein before and after said adding step, wherein a decrease in the quantity of the first protein which is bound to the second protein after the adding step indicates that the test compound is a compound which interferes with the binding of MDM2 to human p53.

2. The method of claim 1 wherein the fragment of human p53 consists of 12 to 28 amino acids comprising the amino acid sequence TFSDLW (SEQ ID NO:2).

3. The method of claim 1 wherein said method is performed as a high throughput screen.

4. The method of claim 1 wherein the test compound is a peptide from a synthetic combinatorial peptide library.

5. The method of claim 1 wherein said MDM2 is human MDM2.