WO2003089936A1

WO2003089936A1 - Aberrant-specific pathways

Info

Publication number: WO2003089936A1
Application number: PCT/NL2003/000294
Authority: WO
Inventors: Mathieu Hubertus Maria Noteborn; Jennifer Leigh Rohn
Original assignee: Leadd B.V.
Priority date: 2002-04-19
Filing date: 2003-04-17
Publication date: 2003-10-30
Also published as: AU2003228132A1

Abstract

The invention relates to the pathway that activates Apoptin's ability to induce apoptosis in tumor cells. More in specific, the invention relates to a method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way, comprising: - preparing at least one lysate of aberrant cells and at least one lysate of reference cells - separating the components present in said lysates - incubating the components with a molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate - visualising said molecule and comparing the components and identifying said substrate.

Description

Aberrant-specific pathways

The invention relates to the pathway that activates Apoptin's ability to induce apoptosis in tumor cells. More in particular, the invention relates to methods and means to identify components of the pathway that activates Apoptin's ability to induce apoptosis in tumor cells.

Apoptosis is an active and programmed physiological process for eliminating superfluous, altered or malignant cells (Ea nshaw, 1995, Duke et al., 1996). The terms transformed, tumorigenic and aberrant will be used interchangeably herein. Apoptosis is characterised by shrinkage of cells, segmentation of the nucleus, condensation and cleavage of DNA into domain- sized fragments, in most cells followed by internucleosomal degradation. The apoptotic cells fragment into membrane-enclosed apoptotic bodies. Finally, neighbouring cells and/or macrophages will rapidly phagocytose these dying cells (Wyllie et al., 1980, White, 1996). Cells grown under tissue-culture conditions and cells from tissue material can be analysed for being apoptotic with DNA-staining agents, such as e.g. DAPI, which stains normal DNA strongly and regularly, whereas apoptotic DNA is stained weakly and/or irregularly (Noteborn et al., 1994, Telford et al., 1992).

The apoptotic process can be initiated by a variety of regulatory stimuli (Wyllie, 1995, White 1996, Levine, 1997). Changes in the cell survival rate play an important role in human pathogenesis of diseases, e.g. in cancer development and auto-immune diseases, where enhanced proliferation or decreased cell death (Kerr et al., 1994, Paulovich, 1997) is observed. A variety of chemotherapeutic compounds and radiation have been demonstrated to induce apoptosis in tumor cells, in many instances via wild- type 53 protein (Thompson, 1995, Bellamy et al., 1995, Steller, 1995, McDonell et al., 1995).

Many tumors, however, acquire a mutation in p53 during their development, often correlating with poor response to cancer therapy. Certain transforming genes of tumorigenic DNA viruses can inactivate p53 by directly binding to it (Teodoro, 1997). An example of such an agent is the large T antigen of the tumor DNA virus SV40. For several (leukemic) tumors, a high expression level of the proto-oncogene Bcl-2 or Bcr-abl is associated with a strong resistance to various apoptosis-inducing chemotherapeutic agents (Hockenberry 1994, Sachs and Lotem, 1997). For such tumors lacking functional p53 (representing more than half of the tumors), alternative anti-tumor therapies are under development based on induction of apoptosis independent of p53 (Thompson 1995, Paulovich et al., 1997). For this, one has to search for the factors involved in induction of apoptosis that do not need p53 and/or cannot be blocked by anti-apoptotic activities, such as Bcl-2 or Bcr-abl-like ones. These factors might be part of a distinct apoptosis pathway or might be (far) downstream of the apoptosis inhibiting compounds.

Apoptin (also called VP3, the terms will be used interchangeably herein) is a small protein derived from chicken anemia virus (CAV; Noteborn and De Boer, 1996, Noteborn et al., 1991, Noteborn et al., 1994; 1998a), which induces apoptosis in human malignant and transformed cell lines, but not in untransformed human cell cultures. Apoptin fails to induce apoptosis in normal lymphoid, dermal, epidermal, endothelial and smooth-muscle cells, to name a few. However, when normal cells are transformed they become susceptible to apoptosis by Apoptin. Long-term expression of Apoptin in normal human fibroblasts revealed that Apoptin has no toxic or transforming activity in these cells (Danen-van Oorschot, 1997 and Noteborn, 1996).

In normal cells, Apoptin was found predominantly in the cytoplasm, whereas in transformed or malignant cells i.e. characterised by hyperplasia, metaplasia or dysplasia, it was located in the nucleus, suggesting that the localization of Apoptin is related to its activity (Danen-van Oorschot et al. 1997).

Apoptin-induced apoptosis occurs in the absence of functional p53 (Zhuang et al., 1995a), and cannot be blocked by Bcl-2, Bcr-abl (Zhuang et al., 1995), or the Bcl-2-associating protein BAG-1 (Danen-Van Oorschot, 1997a, Noteborn, 1996).

Therefore, Apoptin is a therapeutic compound for the selective destruction of tumor cells, or other hyperplasia, metaplasia Or dysplasia, especially for those tumor cells that have become resistant to (chemo)- therapeutic induction of apoptosis, due to the lack of functional p53 and (overexpression of Bcl-2 and other apoptosis-inhibiting lesions (Noteborn and Pietersen, 1998). It appears that even pre-malignant, minimally transformed cells are sensitive to the death-inducing effect of Apoptin. In addition, Noteborn and Zhang (1998) have shown that Apoptin-induced apoptosis is suitable for the diagnosis of cancer-prone cells and treatment of cancer-prone cells.

The fact that Apoptin does not induce apoptosis in normal human cells implies that there would be little or no toxic effect of Apoptin treatment in vivo. Noteborn and Pietersen (1998), Pietersen et al. (1999), and van der Eb et al. (2002) have provided evidence that adenovirus- expressed Apoptin does not have a toxic effect in υiυo. In addition, in nude mice it was shown that Apoptin has a strong anti-tumor activity.

In co-pending patent application PCT/NL01/00771 and Rohn et al., 2002 it is disclosed that Apoptin is tumor-specifically phosphorylated. Furthermore, in υiυo and in υitro evidence is provided that T108 is the phosphorylation site, whereas T106, T107 and P109 are important for maintaining the conformation and recognition sequences for the kinase consensus site. Further, recent work has proven that the ability to phosphorylate Apoptin on T108 is a property present in primary human tumor tissue samples from cancer patients but absent in various primary tissues derived from healthy individuals. Hence, any drug targets or diagnostics based on this kinase or any of the members of the aberrant- specific pathway in which it acts are relevant in patients (Rohn et al, 2002) Furthermore, co-pending application PCT/NL01/00771 discloses an antibody, 108-P, which recognizes Apoptin phosphorylated on T108. The present application discloses that this antibody is also capable of recognising an endogenous substrate of the kinase capable of phosphorylating Apoptin in an aberrant-specific way. The present invention therefore provides methods and means to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way. More in particular, the present invention provides methods and means to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way without having to know the identity of the tumor-specific kinase capable of phosphorylating Apoptin in an aberrant-specific way. Identified substrates are then, for example, used for the development of a (tumor-specific) drug-target.

In a first embodiment the invention provides a method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant- specific way, comprising:

- preparing at least one lysate of aberrant cells and at least one lysate of reference cells

- separating the components present in said lysates

- incubating the components with a molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate

- visualising said molecule and comparing the components and identifying said substrate.

The fact that Apoptin is tumor-specifically phosphorylated uncovers the existence of a tumor-specific kinase and a tumor-specific pathway. Said tumor- specific kinase will have a cellular endogeneous substrate or substrates that are possibly essential component(s) in a tumor-specific pathway. Application of the method according to the present invention reveals said endogenous substrate (s).

To identify such a substrate at least one lysate of aberrant cells and at least one lysate of reference cells must be prepared. An aberrant cell is typically defined as a cell that is in some way dysregulated when compared to a non-aberrant/normal (the terms will be used interchangeably herein) cell. An aberrant cell can for example be dysregulated in growth, apoptosis, telomeric maintenance or production of cytokines etc. Preferably, said aberrant cells are tumor cells or are cells involved in an auto-immune disease. The experimental part discloses multiple types of tumor cells/cell lines which are used in a method according to the invention: an SV40-transformed VH10 cell line (VHSV), an SV40-transformed keratinocytes line (SVK14), Saos-2 cells and U20S cells. However, it is clear that other aberrant cell/cell lines can also be used. Non-limiting examples are human lung cell carcinoma cells, human kidney cells transformed with adenovirus 5 DNA and with SV40 large T antigen, the Jurkat human acute T cell lymphoma cell line, COS-7 cells, which are SV40-transformed African green monkey kidney fibroblasts, a human colon carcinoma cell line, and an EBV- transformed B cell line. An example of an aberrant cell type involved in an auto-immune disease is rheumatoid arthritis (RA) cells and more specifically, synoviocytes involved in RA. In a preferred embodiment, the mentioned reference cells are non- aberrant cells, like normal VH10 cells, human mesenchymal stem cells, normal keratinocytes etc. Preferably, a comparison is made between the components in a lysate prepared from aberrant cells/cell line and components in a lysate prepared from their non-aberrant cells/cell lines counterparts (in other words aberrant cells and reference cells are essentially of similar background). As disclosed herein within the experimental part one compares, for example, VHIO skin fibroblasts to an SV40-transformed VHIO cell line (VHSV), or primary passage 1 breast-derived keratinocytes to an SV40- transformed keratinocyte line (SVK14). By using counterpart aberrant and reference cells of essentially similar background, non-relevant components merely present in a particular background are easily detected and can be ignored. In this way the identification of an (universal) endogenous substrate is improved. In yet another embodiment, the invention provides a method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant- specific way, comprising:

- separating the components present in said lysates

- visualising said molecule and comparing the components and identifying said substrate, wherein said reference cells are cells in which phosphorylation of said substrate is suppressed. Preferably, a comparison is made between the components in a lysate prepared from aberrant cells/cell line and components in a lysate prepared from aberrant cells/cell line in which phosphorylation of said substrate is suppressed. One way of suppressing the phosphorylation of the endogenous substrate is by providing aberrant cells with an inhibitor of the kinase capable of phosphorylating Apoptin in an aberrant-specific way. Said inhibitor is for example a (small) peptide or a (small) chemical compound. As disclosed herein within the experimental part, such an inhibitor is for example full -length Apoptin or a C-terminal fragment of Apoptin, such as amino acids 80-121 of the sequence as disclosed in Figure 1. For detection purpose, said fragment can further comprise a means for detection, like a biotinylated tag. Yet another example of an inhibitor of the kinase capable of phosphorylating Apoptin in an aberrant-specific way is a CDK-inhibitor, like roscovitin or butyrolactone. Another way of suppressing the phosphorylation of the endogenous substrate is by providing, for example by transfection or micro- injection, aberrant cells with Apoptin or a functional fragment or a functional equivalent thereof. By providing Apoptin to an aberrant cell, a competition for phosphorylation by a kinase capable of phosphorylating Apoptin in an aberrant-specific way, takes place between Apoptin and the endogenous substrate. A comparison between the components present in a cell lysate prepared from aberrant cells/cell line and between the components present in a cell lysate prepared from aberrant cells/cell line in which Apoptin or a functional fragment or a functional equivalent thereof is available identifies an endogenous substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way. A functional fragment or a functional equivalent of Apoptin is herein defined as a fragment or equivalent which is capable of competing for the kinase (which is capable of phosphorylating Apoptin in an aberrant-specific way) with the endogenous substrate, which means a fragment or an equivalent which is capable of being phosphorylated by said kinase. A functional fragment is for example a synthetic peptide, or is prepared by deletions at the N-terminus and/or the C-terminus of Apoptin. A functional equivalent of Apoptin is for example prepared by making (non- essential) mutations within Apoptin. The functionality of said fragments or equivalents is easily determined by an in υitro phosphorylation assay (see co- pending patent application PCT/NL01/00771 or experimental part).

The preparation of a cell lysate can be performed via different methods, which are known by a person skilled in the art. The experimental part provides a not-limiting example for the preparation of a cell lysate.

The separation of the components present in the lysates is, for example, accomplished by separation techniques such as electrophoresis or chromatography. The proteinaceous components present in said lysates can for example be separated with a SDS-PAGE gel (ID or 2D). The separation of the components present in a lysate is also accomplished by combination of different techniques, for example a separation on a cation or anion exchange column followed by a separation on a SDS-PAGE gel.

To be able to eventually visualise the components present in said lysates it is necessary to incubate the components with a molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate. Preferably said molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate is an antibody or a functional fragment or a functional equivalent thereof. Even more preferably said antibody or a functional fragment or a functional equivalent thereof is capable of recognizing phosphorylated Apoptin which is phosphorylated on a threonine residue, which residue in the Apoptin of Figure 1 is located at amino acid 106 and/or 107 and/or 108. Even more preferably said antibody is 108-P which recognizes

Apoptin which is phosphorylated at amino acid 108. A functional fragment or a functional equivalent of such an antibody is a fragment or equivalent which is capable of recognizing phosphorylated Apoptin, possibly in (slightly) different concentration and conditions. Visualisation can be accomplished in different ways, for example the molecule which is capable of recognizing phosphorylated Apoptin and phosphorylated substrate is provided with a means to facilitate direct visualisation, for example direct conjugation of the detecting molecule with a means to facilitate visualization (for example an enzyme like peroxidase, a tag like biotin or a chromophore). However, it is also possible to use a second molecule which in turn is capable of recognizing the molecule which is capable of recognizing phosphorylated Apoptin and phosphorylated substrate. This second molecule is then equipped with a means to facilitate visualisation (for example an enzyme like peroxidase, a tag like biotin or a chromophore). Alternatively, the substrate can also be visualized without a molecule that directly recognizes the substrate. For example, phosphorylation provides a strong negative charge which causes a visual shift on a ID or 2D gel. If the phosphate group is radioactive, the shift can be seen on the autoradiograph. One can know that the shift corresponds to the desired substrate by first inhibiting the reaction with a specific inhibitor of the phosphorylation, for example with a peptide mimicking the endogenous substrate's phosphorylation site; all those signals that "lose" their shifts upon inhibition correspond to the endogenous substrate(s).

By visualising the molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate and by comparing the (visualised) components, a substrate of a kinase that is capable of phosphorylating Apoptin in an aberrant-specific way, is identified.

It is clear that the order of steps involved in this method is flexible. For example, it is possible to first prepare the lysates, followed by separation of the components present in said lysates, then incubation of the separated components with a molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate and finally visualisation and identification of a substrate of a kinase capable of phosphorylating Apoptin in an aberrant- specific way. However, it is also possible, after preparation of the lysates, to incubate the components present in said lysates with a molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate (for example, by immunoprecipitation with the 108-P antibody), followed by separation of the components, visualisation and identification. It is clear to a person skilled in the art that the order of steps within the method according to the invention can be adapted to the circumstances and needs.

Although it is clear that in principle the method according to the invention can be carried out with one lysate of aberrant cells and one lysate of reference cells, it is preferred to examine a larger panel of aberrant and reference cells/cell lines. In this way it is determined whether the differential (proteinaceous) components/proteins seen are truly universal indicators of transformed or tumorigenic or aberrant activity.

As already mentioned, a comparison can be made between one or more aberrant cells/cell lines and one or more reference cells/cell lines. The reference cells can be either normal/non-transformed/non-tumorigenic/non-aberrant cells or transformed/tumorigenic/aberrant cells in which phosphorylation of an endogenous substrate of the tumor-specific kinase is suppressed. Yet another possibility is to replace the (panels of) matched cells described above by a single normal cell type that was conditionally induced to be transformed and, therefore, active for the kinase. Then the comparison is repeated by using induced versus non-induced cells. The advantage of this method is that besides the transforming agent, the cells are otherwise genetically identical. Such conditional cells are known in the art. For example, a very low-passage normal cell line like VHIO could be stably transfected with either a temperature- sensitive SV40 large T gene (whereby shifting to the permissive temperature would result in activation of large T and thereby Apoptin kinase) or an estrogen-receptor fused SV40 large T (whereby adding 4-hydroxytamoxifen would activate large T and thereby Apoptin kinase). Alternatively, normal cells could be flooded with enough SV40 large T fused to the HIV tat peptide allowing transduction and thereby transformation of essentially 100% of cells. Still alternatively, cells could be infected with a retrovirus or other virus expressing SV40 large T using a protocol that essentially achieves 100% infection of all cells. Even more preferably, SV40 small t bearing a nuclear localization signal is used as a transforming agent, for example via a protein transduction experiment or a retrovirus or a other virus expressing SV40 small t bearing a nuclear localization signal. In yet another preferred embodiment, a so-called minimal Apoptin-activating determinant (MAAD) is used as a transforming agent. An example of such a MAAD is TN136. Use of smaller fragments of transforming agent is very advantageous in eliminating background bands in for example 108-P profiling assays. As disclosed herein, normal cells do harbor an Apoptin kinase activity during a discrete stage of the cell-cycle: namely, G2/M border and mitosis and hence, panels of matched cells or cell lines can also be replaced by normal cells selected in certain stages of the cell cycle. For example, one can compare genetically identical normal, non-adherent cells to parallel cells enriched for mitosis, and then analyze which substrates are specific to the latter populations by methods described herein.

The method according to the invention is, for example, further refined by (preferably a combination of) - a pre-clearance of the cell lysates with control 108-X (co-pending patent application PCT/NLOl/00771), a rabbit polyclonal antibody corresponding to the unphosphorylated epitope of phosphorylated Apoptin in the same region (or in other words, 108-X is the non phosphospecific antibody). This extra step removes all proteins that might be responding weakly to the 108-P antibody, thus reducing the level of background; and/or - providing a metabolic label to the cells followed by a method according to the invention. An example of a suitable metabolic label is ³²P orthophosphate. When the other steps of the method according to the invention are followed, a double detection is possible: (a) the (visualised) molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate and (b) the ³²P label. A coincidence of both detection methods provides strong evidence for the identification of a substrate of a kinase that is capable of phosphorylating Apoptin in an aberrant-specific way, and screens out the background signal of non-phosphorylated proteinaceous cross-reacting molecules. - bringing the aberrant cells and reference cells into cell-cycle staged populations to further identify, enrich or refine an endogenous substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way.

Besides a proteinaceous approach it is also possible to use a nucleic acid approach, for example an RNA-based approach based on standard RNA chips or related technologies could be employed to compare the gene expression pattern of all of the aberrant versus reference cells/cell lines described above, most preferably the conditional/induced normal cells or the inhibited tumor cells. In this case, differential display using techniques known in the art could be employed to identify such genes or pathways that are dysregulated in an aberrant-specific manner on the transcriptional level. In another embodiment the invention provides a substrate obtainable by the method according to the invention. After identification of said substrate the corresponding protein sequence is easily determined, for example, by excising from the gels and determining the identity by techniques known in the art, for example mass spectrometric analyses, MALDI-TOF etc. Knowledge of the identity of such a substrate is further exploited to, for example, develop a (tumor-specific) drug target, or a diagnostic method that has the advantage of relying on the detection of an endogenous substrate (which is called EndoApoCheck), or is used to identify the kinase capable of phosphorylating Apoptin in an aberrant-specific way. In yet another embodiment the invention provides a method for obtaining a modulator of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way comprising:

- incubating aberrant cells with a possible modulator of said kinase

- providing said cells with Apoptin or a functional fragment or a functional equivalent thereof

- determining the phosphorylation state of said Apoptin.

A modulator can either be an enhancer or an inhibitor. Preferably said modulator is an inhibitor and even more preferably said modulator is a proteinaceous substance. Examples of an inhibitor of a kinase capable of phosphorylating Apoptin in an aberrant-specific way, are a C-terminal fragment of Apoptin that comprises the amino acids 80-121 of the Apoptin as depicted in Figure 1 or CDK-inhibitors like roscovitin and butyrolactone. Preferably, said inhibitor binds to the kinase and does not let go. However, it is also possible that the inhibitor is preferentially phosphorylated by the kinase and thereby inhibits the phosphorylation of, for example, an endogenous substrate (by competitive inhibition). It is known in the art that artificially mutated phosphorylation consensus sites are often stronger than wild-type consensus sites, because in nature it is not always advantageous to have an overly strong recognition. However, it is known that if presented in excess, even a peptide with the same affinity as the natural substrate still functions as an inhibitor. Such a (Minimal) Phosphorylatable Peptide (MPP) has many uses. For example, the MPP is used as a substrate in ApoCheck™, our ELISA- based detection of the kinase assay (co-pending patent application PCT/NLOl/00771), as an alternative substrate which can be more easily produced. Furthermore, the MPP is used as a competitive inhibitor of the kinase capable of phosphorylating Apoptin in an aberrant-specific way by competing with the endogenous substrate of said kinase. An artificial MPP can also be used as a minimal tumor specific killing domain of Apoptin and may thus be used for therapy. Because (recombinant) Apoptin protein is phosphorylated specifically on residue T108 by for example tumor cell lysates using a routine in vitro kinase assay, it serves as an excellent read-out for inhibition of the tumor-specific kinase (or of other proteins in the tumor-essential pathway). For instance, ApoCheck™ is performed in a 96-well plate or other high-throughput design, facilitating this read-out even further. Because the read-out is so easy and specific, a large number of potential modulators/inhibitors is screened without knowing the identity of the kinase itself. Such inhibitors, or refined second- generation products based upon them, serve as excellent drugs for intervening in a for example tumor-essential pathway. Furthermore, such inhibitors, even crude ones, are useful for the validation of the efficacy of such drugs before going to the time and expense of developing them. Finally, such inhibitors are employed to learn more about the tumor-specific mechanisms of Apoptin- induced apoptosis, studies of which have already led to the discovery of novel drug targets. Candidate inhibitors are obtained from, for example, a (recombinant) peptide library, a small-molecule library or a chemical library. Another way to obtain candidate inhibitors is the preparation of (library of) peptides based on the sequence of Apoptin. Candidate molecules are, for example, tested by using a high throughput in vitro kinase assay (such as ApoCheck) as a read-out. Besides using Apoptin as a readout system., the endogenous substrate can also be used as a readout system. When the endogenous substrate is used as a readout system, the sensitivity of the assay can be further enhanced by providing extra (endogenous) substrate or a functional fragment or a functional equivalent thereof to the cells to be tested. Therefore, the invention also provides a method for obtaining a modulator of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way comprising:

- incubating aberrant cells with a possible modulator of said kinase

- determining the phosphorylation state of an endogenous substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way.

The effect of a possible modulator of a kinase involved in the phosphorylation of Apoptin in an aberrant-specific way can, for example, be determined in vitro, in the kinase assay, or in vivo, by treating cells with peptide (by, for example, microinjection or by conjugating the peptides to a membrane permeable compound such as fLuoromethylketone) and then seeing if, for example expressed Apoptin loses 108-P reactivity in the immunofluorescence assay.

An (effective) identified inhibitor is then used to analyze the effect of kinase inhibition on tumor cells. An inhibited tumor cell can, for example, be identified by induction of apoptosis, ceasing in proliferation, or losing anchorage independence, or some other reversal of the transformed/tumorigenic state. To test the effect of an identified inhibitor on normal cells, parallel assays are performed on preferably various normal cells to confirm that inhibition of the kinase is not prohibitively toxic. Although standard peptides themselves make suboptimal therapeutic agents, recent advances in the art have led to chemistry that mimics peptides, which can result in a stable, tissue-soluble peptide analogue that is much more appropriate for systemic treatment of patients.

In another embodiment the invention provides a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way. In yet another embodiment the invention provides a nucleic acid encoding a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way. An example of such an inhibitor is Apoptin(80-121), optionally comprising a means to stabilize or a means for detection. Now that it is know that a kinase capable of phosphorylating Apoptin in an aberrant- specific way is inhibited by Apoptin(80-121), it is now well within the scope of the invention to obtain a functional equivalent and/or a functional fragment of Apoptin(80-121). A functional equivalent and/or a functional fragment of Apoptin(80-121) is defined as a equivalent and/or a fragment' that has the capability of inhibiting a kinase capable of phosphorylating Apoptin in an aberrant-specific way, possibly in a different amount. A functional fragment of Apoptin(80-121) is for example prepared by further deleting amino acids from the C-terminal and/or N-terminal side. A functional equivalent is for example obtained by replacing non-essential amino acids by other amino acids. In another embodiment the invention provides a vector encoding a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way.

In yet another embodiment the invention provides a gene delivery vehicle comprising a vector according to the invention which enables using a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way for the treatment of a disease where enhanced cell proliferation or decreased cell death is observed via the use of gene-therapy. By equipping a gene delivery vehicle with a nucleic acid molecule encoding a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way, and by targeting said vehicle to a cell or cells that show over-proliferating behaviour and/or have shown decreased death rates, said gene delivery vehicle provides said cell or cells with the necessary means of inhibiting or decreasing aberrant cells, providing therapeutic possibilities. Furthermore, the invention provides a host cell comprising a vector or a gene delivery vehicle encoding a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way.

In another embodiment the invention provides a pharmaceutical composition comprising a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way or a nucleic acid encoding a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way or a vector according to the invention or a gene delivery vehicle according to the invention or a host cell according to the invention.

In yet another embodiment the invention provides the" use of an proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way or a nucleic acid encoding a proteinaceous inhibitor of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way or a vector according to the invention or a gene delivery vehicle according to the invention or a host cell according to the invention in the preparation of a medicament for the treatment of a disease where enhanced cell proliferation or decreased cell death is observed, as is in general the case when said disease comprises cancer or an auto-immune disease.

The invention will be explained in more detail in the following description, which is not limiting the invention.

EXPERIMENTAL PART

Plasmids

Description of pCMV-Apoptin and pCMVneo The plasmid pCMV- Apoptin was described previously by Danen-Van

Oorschot et al. (1997). In short, the plasmid pCMV- Apoptin contains the human c tome galo irus (CMV) promoter and CAV DNA sequences (nt 427- 868) encoding Apoptin exclusively. The synthesized Apoptin protein harbours apoptotic activity and is identical to GenBank Q99152 except position 116 contains a K > R change. The empty vector pCMV-neo was described by Baker et al. (1990) and is used as a negative control.

Description of SV40 early region-encoding plasmids.

The following plasmids have been described previously: pRSV-LT (wild- type Large T antigen); pRSV- Zll35 (LT lacking 17-21, the DNA-J domain); pRSV-TN136 (LT 1-136]) (Noteborn and Danen-Van Oorschot, PCT- NL98/00688) and pRSV-st (wild-type small t) (Noteborn and Zhang, PCT- NL98/00457). In addition, the plasmid pRSV-nls-st (small t antigen with an N- terminal heterologous NLS) was generated by first ligating a BglWKpnϊ/BglU linker with ATG startcodon and NLS to the N-terminus of a Bgllϊ/BamΗI fragment of SV40 small t antigen contained in plasmid pWZT-Bg-st (kindly provided by Prof Kathleen Rundell, Northwestern University, Evanston, IL, USA) to form a Kpnϊ/BamΗI fragment of NLS-st, and then religating the KpnVBamΗI fragment of NLS-st to the -Kp?ιI/_Bα HI-digested pRSV vector, which is identical to those of the above described SV40 large T mutant constructs. Cloning of C-terminal Apoptin mutants and phosphorylation mutants thereof.

The constructs pcDNA-MBP-80-121 (C-terminal Apoptin wild-type) and pcDNA-MBP-Apoptin (full-length wild-type) have been constructed as follows: pcDNA3.1mychis(-)-MVP3 (Leliveld, SR, Leiden University, personal communication) was digested with BamΗI and Kpnl, removing the full Apoptin open reading frame. Subsequently, the truncated Apoptin(80- 121) part was amplified by PCR and cloned into the linearised vector at the BamΗI and Kpnl sites. The Apoptin open reading frame was cloned at the 5' BamΗI and 3' Sail sites of pMalTB, downstream of the E.coli MalE open reading frame, which codes for MBP. PMalTB is derived from pMal-c2 (NEB), the peptide linker between MBP and its fusion partner contains a thrombin consensus site. Expression is controlled by an IPTG-inducible Ptac promoter. The constructs pcDNA-MBP-80-121-5-ala-106 (encoding the phosphorylation loss-of-function mutation, TTTPS [106- 110] AAAAA [also known as 5-Ala-106]) and pcDNA-MBP-80-121-TAA (encoding the phosphorylation loss-of-function mutation TT[107-108]AA [also known as TAA]) were subcloned by standard methods in parallel using similar strategies. Briefly, the parental pcDNA- MBP -Apoptin (full-length wild-type) was digested with BamΗI (which cleaves in-frame precisely at the end of the MBP coding sequence) and Kpnl (which cleaves precisely at the end of the Apoptin coding sequence) to create a cut vector. Insert was prepared by PCR using a 5'primer with a Sα HI-site- containing tail homologous to Apoptin DNA encoding amino acids 80-85 paired with a 3' primer homologous to Apoptin DNA encoding amino acids 117-121 with a tail encoding a stop codon and a Kpnl cleavage site. PCR was performed on two templates: for pcDNA-MBP-80-121-5-ala-106, the template plasmid was pINTsAPO-5-Ala-106 (Rohn et al, 2002); whereas for pcDNA-MBP-80-121- TAA, the template used was an unpublished plasmid identical to, and constructed the same way as, pINTsAPO-5-Ala-106 except that the mutation is TT[107-108]AA. Neither encoded protein can be phosphorylated by Apoptin kinase because their phosphorylation sites are not intact; this fact was confirmed by standard in vivo ortholabeling assays such as those described in co-pending patent application PCT/NLOl/00771 and Rohn et al, 2002.

Cell lines and culturing

The following established cell lines have been described previously: Saos-2 human osteosarcoma cells (Diller et al. 1990), which are functionally deficient for p53 function; U20S human osteosarcoma cells (Diller et al. 1990), which are functionally competent for p53 function; HeLa cells (Kennedy et al., 1995); VHSV (Danen-Van Oorschot et al., 1997) are SV40-large-T transformed VH10 human fibroblasts; SVK14 (Danen-Van Oorschot et al., 1997) are SV40- large-T-transformed keratinocytes; primary passage 1 human breast keratinocytes were a gift from Dr. M. Ponec, Department of Dermatology, Leiden University Medical Center; low-passage primary human fibroblasts (VHIO) were a gift from Dr. L. Mullenders, Leiden University Medical Center, Department of Chemical Mutagenesis and Radiation Genetics; HT29, a human colon carcinoma cell line, was obtained from the European Cell Culture Collection; and human mesenchymal stem cells were purchased from BioWhittaker (USA). Mouse embryo fibroblasts, isolated from normal FVB mice using standard techniques, were a gift from Alexandra Pietersen, at Leadd, Leiden, The Netherlands. Low-passage CD31- human diploid skin fibroblasts were a kind gift from Dominik Mumberg at Schering AG, Berlin, FRG.

All cells were cultured in Dulbecco's Modified Eagle Medium supplemented with 10% fetal bovine serum and penicillin/streptomycin, and cultured at 10% C0₂ in a humidified 37°C incubator, except for the following: primary and transformed keratinocytes were grown in DMEM/Hams F12 (3:1), supplemented with 5% bovine calf serum (Hyclone), 10"⁶ M isoproleronol, 10'⁶ M hydrocortisone, IO-⁷ M insulin, and penicillin/streptomycin and cultured at 37°C at 7.5% CO2.; mesenchymal stem cells were cultured in dedicated medium purchased from the manufacturer for no more than 4 passages, essentially using their protocols; and MEFs were cultured the same as Saos-2 cells except RPMI medium was used instead of DMEM. CD31- cells were cultured in MEM-EARLE (BIOCHROM AG, Berlin) containing 2.2 g/L NaHC03 and 0.518 g/L N-Acetyl-L- glutamine and supplemented with 10% fetal calf serum, 100 U/ml penicillin and 100 μg/ml streptomycin.

Cells separated into various phases of the cell cycle

The following cell samples were a gift from Dr. Nullin Divecha, the Netherlands Cancer Institute. Murine erythroleukemia (MEL) cells, originally derived from the spleen of a mouse infected with Friend's virus, were cultured in suspension as described above for Saos-2 tumor cells. Cells were treated with 50 ng/ml nocodazole to disrupt the spindle-pole formation and therefore to arrest the cells at the G2/M phase of the cell cycle. 9.5 hours post-treatment, cells were "released" from the nocodazole block by washing two times with warm PBS and replating them in normal medium (time zero). The cells were then returned to the incubator and allowed to proceed in a synchronous fashion through the entire cell cycle, which in this cell type takes approximately 11 hr. Samples were drawn out at 1, 2, 4, 5, 6, 8, 10 and 11.5 hr post-release and further processed: cells were washed one time in PBS, pelleted in a centrifuge, aspirated dry, snap-frozen in liquid nitrogen and transported to our laboratory. Control cell pellets included with the shipment were cells incubated in nocodazole for 9.5 hours but not released, and cells growing exponentially (in an unsynchonized fashion) in normal medium. The phase of the cell cycle corresponding to each sample was confirmed by Dr. Divecha using standard FACS DNA profiling with propidium iodide.

Normal CD31- skin fibroblasts (below passage 8) were plated in 9 cm plates, then the next day processing for separation into various stages of the cell cycle began using the following treatments (see Table 1): Table 1: Conditions used to arrest CD31- cells in different stages of the cell cycle:

Following this separation into stages, the cells were harvested as usual for the in vitro kinase assay (see below).

Recombinant proteins Two different soluble, purified, recombinant Apoptin fusion proteins were used, the first being Apoptin with a C-terminal histidine tag (VP3-His), and the second being Apoptin with a N-terminal maltose binding protein (MBP-Apoptin, also known as MBP-VP3). Both fusion proteins were produced in E. coli bacteria and were a gift from Rutger Leliveld, Department of Chemistry, Leiden University. Details on the construction of the plasmids and the purification of the proteins can be found in co-pending application PCT/NL01/00664. Virus production and infection

A construct encoding SV40-large T cDNA in the backbone of the ecotropic retroviral pBabePuro vector was sub-cloned using standard methods. A construct encoding enhanced Green Fluorescence Protein (GFP) in the same retroviral backbone was sub-cloned using standard methods. Both viruses were packaged using Ecopack II (ClonTech) using essentially the methods recommended by the manufacturer. Briefly, 36 hours after transfection at 32°C, virus supernatant was harvested, filtered through a .22 μm Millipore filter and used to infect naϊve MEF cells in the presence of polybrene. Uninfected MEFs were taken along as a negative control. Cells were lysed for the kinase assay (see below) 48 hours post-infection. As a control for viral transduction and to determine transduction efficiency, parallel infections were used to confirm the presence of the large T or GFP proteins by fluorescence microscopy of whole cells and by Western blot of cell lysates (see below).

Transfections and Microinjections

For biochemical analyses, cells were plated the day before on 10 cm dishes such that cultures were 40% confluent at the time of transfection. Seven μg DNA was transfected using a 3:1 (μl:μg Fugene:DNA) ratio of FuGene 6 (Roche) according to the manufacturer's instructions. The complexes were incubated on the cells in the presence of full serum and were left on until the cells were assayed. For apoptosis immunofluorescence assays, 0.5x10^s- lxlO⁵ cells were plated on 2-well Permanox chamber slides (Nunc) and transfected the same' as for 10 cm plates, except that only 1.5 μg DNA per well was used and the rest of the transfection components were scaled down accordingly. When microinjections were used instead of transfections, the following procedures were followed. Cells were cultured on glass-bottomed microinjection dishes. The cells were micro-injected in the cytoplasm with protein at 3 mg/ml using an Eppendorf micro-injector with the injection- pressure condition of 0.5 psi or in the nucleus in the case of DNA (100 ng/μl). The cells were co-injected with Dextran-Rhodamine (MW: 70 kDa; Molecular Probes, Leiden, The Netherlands) to be able to later identify injected cells.

When CD31- cells were transfected, the transfection technology Nucleofector™ was exerted in the presence of the normal human dermal fibroblasts (NHDF) cell-type specific solution (AMAXA BioSystems, Germany). Each nucleofection sample contained 5 μg DNA, l.OxlO⁶ cells and lOOμl NHDF Nucleofector™, and the transfection was carried out under the program U23 of Nucleofector™ device, as recommended by the manufacturer. The transfected cells were then suspended in appropriate volume of medium and seeded for further culture and analysis.

Immunofluorescence assays

The cells were incubated at 37°C after transfection or microinjection until the cells were fixed with formaldehyde -methanol- acetone. Briefly, cells were immunostained using 111.3 (Danen-van Oorschot et al., 1997) or 108-P (details on this antibody can be found in co-pending application PCT/NLOl/00771) as a primary antibody for Apoptin (mutants), or 9E10 anti- myc monoclonal (Evan et al, 1985) for the lacZ-myc control, using the appropriate fluorochrome-conjugated goat-anti-Ig as a secondary antibody, and counter stained with DAPI. Only positive cells were assessed for apoptosis using nuclear morphology as the criterion. At least 100 cells per well were scored and the assays were done multiple times. The appearance of the Apoptin staining and the cell's overall morphology were also noted.

To confirm viral infection efficiency, the monoclonal antibody 419 was used to detect SV40 large T, and GFP was visualized by its own intrinsic fluorescence (without fixation).

When other SV40 early region mutants were used in immunofluorescence, their presence was confirmed by either antibody 419 or KT3, a mouse antibody raised against an eleven amino acid peptide (KPPTPPPEPET) derived from the carboxy terminal sequence of SV40 LT and non-reactive with SV40 st (Research Diagnostic Inc.) .Monoclonal antibody production

Phosphorylation of a protein often creates a new immunological epitope (Blaydes et al., 2000). We had previously described the successful generation of a rabbit polyclonal antibody specific for phosphorylated Apoptin (i.e., Apoptin in a transformed or tumor environment) using as an inoculant the following peptide: SLITTTPSRPRTA (corresponding to Apoptin residues 103-115) with the third threonine phosphorylated (residue 108). Thus, to increase the utility of this tool, we decided to create a corresponding monoclonal with the same peptide design in a mouse.

The peptide was synthesized at Eurogentec (Belgium) and all subsequent antibody syntheses, hybridoma creation and preliminary testing of cell clones were also performed there. In short, the peptide was coupled to Keyhole Limpet Hemocyanin (KLH) and injected in complete Freund's adjuvant into four Balb/c mice with an immunization schedule of one injection and two subsequent boosts. Blood samples were taken before and after immunization. The titers from three of the four mice were high. The sera were tested for specific reactivity to the phosphorylated peptide by ELISA.

The best-reacting mouse was sacrificed and its spleen removed. The spleen cells, including the immunoglobulin-producing lymphocytes, were fused to the mouse myeloma cell line SP2/0-Agl4 with the aid of polyethylene glycol to create a hybridoma, and the properly fused cells selected by standard HGPRT activity. The supernatants of candidate clones were screened for phosphospecific reactivity against phosphorylated versus unphosphorylated antigen using both ELISA and Western blot analysis. Next, the correct and best hybridoma was cloned using a standard limiting dilution strategy, and finally, the secreted antibody of the final clone was confirmed to represent phosphospecific monoclonal antibody by ELISA and Western blot, and was named ml08-P (to distinguish it from the purified rabbit polyclonal antibody 108-P, which from now on will be referred to as Rbl08-P). The antibodies were produced in mass both in vivo (ascites method) as well as in vitro (cell culture) at EuroGentec using standard methods.

Further validation experiments of ml08-P were performed in-house, using Western blot and ELISA. For tumor samples, we used'lysates of Saos-2 cells transfected with CMV-VP3, known by in vivo phosphorylation assays to harbor the phosphorylated form of Apoptin. For non- tumor samples, we used both MBP-Apoptin recombinant protein, which cannot be phosphorylated by E.coli, or lysates of normal VHIO fibroblast cells transfected with CMV-VP3, which are known from the in vivo assay to harbor Apoptin that is not phosphorylated. Samples were subjected to Western blot analysis as described above, and probed in parallel with ml08-P and as a control for total Apoptin, the monoclonal antibody 111.3 which recognizes Apoptin regardless of its phosphorylation state. We also tested the utility of ml08-P in immunofluorescence analysis of Saos-2 tumor cells or VHIO normal fibroblasts grown on glass cover slips and expressing transduced CMV-Apoptin DNA. These assays were performed as described above except that ml08-P was incubated at a concentration of approximately 0.1 mg/ml in conjunction with the antibody 111.3, and each primary antibody was labelled directly with a differently conjugated secondary antibody (FITC or rhodamine) so that both phosphorylated and nonphosphorylated Apoptin could be visualized in the same cell.

Preparation of cell lysates

Cells were cultured to approximately 80% confluency, washed with ice- cold PBS, then lysed in 1 ml RIPA buffer (10 mM Tris pH 7.5, 150 mM NaCl, 0.1 % SDS, 1 % NP-40, 1 % sodium deoxycholate, supplemented with the following protease or phosphatase inhibitors at standard concentrations: trypsin inhibitor, pepstatin, leupeptin, aprotinin, PMSF, β-glycerophosphate, sodium vanadate, and sodium fluoride). Lysates were incubated on ice for 30 minutes, centrifuged for 10 minutes at 13,000 rpm in a refrigerated microfuge, and the supernatants were kept. In some cases, the proteins were further processed by immunoprecipitation with affinity-purified polyclonal rabbit serum (VP3-C) raised against the C terminal portion of Apoptin, or 108-P antibody, in either case coupled to protein A beads using standard methodology. The final pellet was resuspended in 2x denaturing Laemmli sample buffer and stored at -20°C until processing. Alternatively, for quantitative assays, lysates were not supplemented with sample buffer straightaway; instead, their protein concentrations were determined by a detergent-compatible Bradford assay (Biorad DC assay), and aliquots of known amounts of protein were resuspended in sample buffer prior to analysis.

SDS-PAGE and Western blot analysis

Protein samples in sample buffer were boiled for five minutes and then resolved on SDS-PAGE gels at acrylamide concentrations ranging from 7.5 to 15% depending on the experiment. Protein was electroblotted from gel to PVDF membranes (Immobilin, Millipore) using standard techniques. In the case of radiolabelled lysates (see below), autoradiography was first performed on the Immobilin and then the membrane was Western-immunoprobed for precise overlaying of the radioactive signal with the Western signal. All autoradiographs were exposed along with fluorescent marks to facilitate subsequent orientation for band isolation (see below).

For immunostaining, membranes were blocked in a tris-buffered saline solution supplemented in 0.5% Tween-20 (TBS-T) and 5% non-fat dry milk (TBS-TM) for 30 minutes, washed 3 times 20 min in TBS-T to remove all the soluble milk proteins, then incubated for 1 hr at room temperature with purified primary antibody Rb(rabbit)108-P at a dilution of 1:1000 (approximately 1 μg/ml final) or ml08-P at a concentration of approximately 1 mg/ml, or 111.3 as described (Danen-van Oorschot et al., 1997) in a buffer consisting of 3% bovine serum albumin in TBS-T (TBS-T/B). After 3x 5' washes in TBS-T, membranes were further incubated in the appropriate antibody (anti-rabbit Ig) conjugated to horseradish peroxidase (HRP) in TBS-T/B. To confirm the presence of GFP and SV40 large T after infection, the PVDFs were stained with anti-GFP (Living Colors, ClonTech) and the anti-large T monoclonal supernatant, 419. To confirm the presence of other SV40 early region mutants after transfections, 419 antibodies were used. After 3x 20' washes in TBS-T, membranes were subjected to enhanced chemiluminescence using standard techniques, exposed to x-ray film (Kodak), and films were developed using standard automated methods. In some cases to confirm equal loading of samples, the PVDF membranes were stripped using a standard detergent/β-mercaptoethanol strategy and re-probed for an arbitrary endogenous housekeeping protein, e.g. with cytochrome C or actin (available from BD Biosciences and Sigma, respectively). In some cases to confirm that bands seen with antibody/ECL corresponded to phosphorylated proteins, the PVDF was treated with potato acid phosphatase (from Sigma, 100 ug/ml in a buffer of 40 mM PIPES, pH 6, 1 mM DTT and protease inhibitor cocktail from Sigma, incubated for 30 min at 4 degrees C) to remove all potential phosphorylated epitopes and then re-probed with phosphospecific antibody. Any potential phosphoproteins should have disappeared after this analysis. In some cases to confirm that bands seen with antibody/ECL corresponded to phosphorylated proteins, the analysis was repeated with metabolically labeled cell lysates. Briefly, cell cultures were washed two times with phosphate-free DMEM (PFD; Sigma), then incubated for ten minutes in PFD in the tissue culture incubator to deplete intracellular phosphate. Next, the plates were incubated with PFD supplemented with ³²P-orthophosphate (0.5-1.25 mCi/ml) for four hours to label all de novo-phosphorylated proteins in the cell. Cell lysis, SDS-PAGE, autoradiography and Western blot were then performed as described above. A correspondence of antibody-detected phosphospecific bands (seen by Western blot) to those seen to be radioactive (by autoradiography) served as confirmation that the immunostained bands represented bona fide phospho-protein epitopes. To isolate protein corresponding to aberrant-specific bands on the films, the following procedure was performed for one-dimensional gels. (The procedure was also repeated for two-dimensional gels when too many proteins were found to co-exist in one band, see below). Gels were run in parallel in the same electrophoresis and Western blotting chambers. Extreme care was taken not to stretch or change the shapes of each gel during the blot assembly procedure. Subsequent to electroblotting, orientation marks were clearly made on each PVDF. One PVDF was saved at —20 degrees Celsius, while the other was subjected to immunostaining to visualize aberrant-specific bands reacting with the 108-P antibody as described above. Using both the film and the stained PDVF for orientation, the corresponding region on the clean parallel blot that had been saved at -20°C was excised with a razor blade and put in a tube for further identification (see Mass spectroscopy section, below).

In some cases parallel gels were stained with Coomassie Blue using standard methods to confirm that bands of interest were present in amounts sufficient for further analysis; when they were not, the procedures were scaled up and the analysis repeated.

Two-dimensional (2D) gel electrophoresis. In some cases, the procedure described above was performed with 2D gels instead of standard one-dimensional SDS-PAGE gels in order to achieve maximum separation of individual proteins using the techniques described in Anderson, 1991 (Two-Dimensional Electrophoresis: Operation of the ISO- DALT System, Second Edition, Large Scale Biology Press, Rockville, Maryland USA), with further modifications as suggested by Angelica Gorg (Technical University of Munich, Germany) and Michael Dunn (Harefield Hospital, United Kingdom). Briefly, samples were dissolved in an 8M urea buffer, resolved by isoelectric point in the first dimension using the IPGphor Isoelectric Focusing System of Amersham-Pharmacia, and by molecular mass on standard SDS-PAGE electrophoresis (ISO-DALT apparatus) in the second dimension. Just as for the one-dimensional comparison, above, samples to be compared were run on the same platform in the first dimension, and in the same tank for SDS-PAGE and Western blot in the second dimension. All other procedures were identical, except that spots, not bands, were- excised. In some cases parallel gels were stained with Coomassie Blue using standard methods to confirm that bands of interest were present in amounts sufficient for further analysis; when they were not, the procedures were scaled up and the analysis repeated.

Mass spectrometric analysis and isolation of DNA clones.

PVDF fragments corresponding to excised spots or bands were processed further for protein identification by procedures known in the art. Briefly, proteins were digested with trypsin and subjected to MALDI-TOF analysis to identify the sequences of individual tryptic fragments. In some cases, different methods of mass spectroscopy were also used. Standard software and genetic databases were used to reconstruct the full identity of each protein.

When proteins corresponded to known genes in the database, the clones were obtained from the appropriate source. When the proteins corresponded to genes not in the database, or in the database only incompletely, a standard PCR cloning strategy known in the art, using oligos deduced from the peptide sequences, was used to obtain the final clones.

In vitro kinase assay

All cultured cells to be tested were washed twice with ice-cold PBS, scraped on 0.5 ml ice cold PBS with a rubber policeman, transferred to

Eppendorf tubes and centrifuged two minutes at 4000 rpm in a cold microfuge. (Cell pellets can be snap-frozen in liquid nitrogen and stored at -70°C if not used immediately, with no adverse effect.) Supernatants were removed, and 50-100 μl of kinase buffer was added (20 mM Hepes pH 7.4, 20 M MgCls, 150 mM NaCl supplemented with the following protease and phosphatase inhibitors at standard concentrations: trypsin inhibitor, pepstatin, leupeptin, aprotinin, PMSF, β-glycerophosphate, sodium vanadate, and sodium fluoride). Samples were frozen-thawed three times, alternating between ethanol/dry ice and regular ice to lyse the cells, then were centrifuged at 14_/000 rpm in the microfuge. The supernatants were transferred to fresh tubes as the final cellular lysates.

Protein concentration was determined at OD595 using the Biorad Bradford reagent according to standard procedures. Equal amounts of cellular lysate supernatants (usually 10 μg) were incubated for 30' at 30°C with 3 mg of the recombinant Apoptin protein (either MBP-Apoptin or VP3-His), 50 μM ATP, 0.15 μCi of γ-³²P-ATP (only when reaction is desired to be radioactive ly labelled), 2 M DTT and kinase buffer to bring the total volume to 30 μl. Samples were run on SDS-PAGE gels (same procedure as described in "Western blot analysis"), and transferred by Western blot onto PVDF membranes. In many cases, prior to SDS-PAGE, the samples were further improved by including a purification step. In this case, following the reaction, the protein substrate was purified in RIPA buffer either by standard immunoprecipitation with protein-A agarose beads coupled to VP3-C, a polyclonal antibody recognizing Apoptin, or by amylose beads, which are specific for the MBP fusion component of MBP-Apoptin, or by standard nickel precipitation, which is specific for the His fusion component of VP3-His. Precipitates were washed sufficiently prior to the final samples being loaded on SDS-PAGE gels.

Following Western transfer, the membranes were exposed to x-ray film and then visualized using autoradiography. Subsequently, the same PVDFs were immunostained with the antibody 108-P to visualize specific phosphorylation, and with antibody 111.3 to confirm the presence of the Apoptin protein. For some of the cell cycle experiments described, membranes were also probed with an antibody against Cyclin B (Santa Cruz) to confirm certain stages of the cell cycle. Isolation of Apoptin kinase activity by affinity purification

A standard in vitro kinase assay was performed as described above in tumor lysates using either of two peptide substrates: N-terminally biotinylated 80-121 (called Bio80), and N-terminally biotinylated 80-121 mutated at the phosphorylation site: 106-110 replaced with 5 Ala residues (called Bio80-Null). After the reaction was completed, it was not stopped with EDTA as normal. Instead, the reacted lysates were incubated with sepharose beads coated with Strepavadin (which is known in the art to bind very tightly to Biotin) pre- blocked with BSA and then the beads were purified away from the lysates.

Finally, the reacted lysates were re-assayed with fresh MBP-Apoptin substrate in the in vitro kinase assay to determine whether the kinase activity had indeed been depleted, which is an indication that the kinase had bound to the biotinylated peptide and had therefore been "fished" out of the reacted lysate by the Strepavadin beads. The readout for depletion was the standard conclusion to the in vitro kinase assay, namely SDS-PAGE, Western blot and probing of the MBP substrate with the antibodies 108-P and 111.3

Studying the role of cyclin dependent kinases for their ability to affect the phosphorylation of Apoptin

The ability of the inhibitor roscovitine (a low-molecular weight inhibitor of CDK1, CDK2 and CDK5) to inhibit the ability of HeLa tumor lysates to phosphorylate MBP-Apoptin substrate in the standard ³²P-γATP radioactive in vitro Apoptin kinase assay (as described in co-pending application PCT/NLOl/00771 as well as in Rohn et al, 2002) was performed by standard methods. Briefly, uninhibited kinase reactions (mock) were carried out in parallel to reaction tubes containing an increasing amount of roscovitine (0.1, 1.0, 10, 100, 1,000 and 10,000 nM of inhibitor). After the reaction, MBP- Apoptin was purified from the reactions by standard affinity procedures, washed, and then incorporation of radioactivity indicating phosphorylation was quantitated by scintillation counting. The IC50 was determined by plotting radioactive counts versus increasing amounts of inhibitor. As a control, in some reaction instead of using MBP-Apoptin as a substrate, an equimolar equivalent of the known CDK substrate recombinant Histon'eHl (Sigma) was used.

Finally, to investigate the effect of cyclin-dependent kinases on the inherent tumor-specific kinase activity of intact, living tumor cells, Saos-2 cells growing on coverglasses were transfected with pCMV-VP3 and 3 days later, were treated for 90 min with the CDK inhibitor butyrolactone (100 μM), or with an equal volume of DMSO as a vehicle-only mock control, prior to fixation and immunofluorescence with 111.3 and 108-P antibodies (see section on transfection and immunofluorescence). Apoptin-positive cells were inspected for any visual reduction in the proportion or strength of the Thr- 108 phosphorylation signal as a result of treatment.

RESULTS

Further evidence of the fact that the kinase capable of phosphorylating Apoptin on Thr 108 in an aberrant-specific way is specifically necessary for activating the process of aberrant-specific apoptosis

Previously, we have shown that the Apoptin mutant T108E, in which constitutive phosphorylation of Apoptin is mimicked by a single negatively- charged amino acid substitution (glutamic acid) at residue 108, is sufficient to confer a gain-of-function phenotype upon Apoptin, allowing it to translocate to the nucleus and kill normal, non-aberrant cells (Co-pending international patent application PCT/NLOl/00771; Rohn et al., 2002). Therefore, said phosphorylation is sufficient for aberrant-specific apoptosis. To confirm in another way that said phosphorylation is necessary for aberrant-specific apoptosis by Apoptin, a loss-of-function mutant was prepared by abolishing the aberrant-specific phosphorylation site on Apoptin by mutating the phosphorylation region to non-phosphorylatable residues (e.g. alanines), and then testing its ability to kill tumor cells.

The experiment is complicated somewhat by the fact that Apoptin possesses two autonomous death domains, one in the N-terminus and one in the C-terminus (co-pending European patent application EP 02076597.0). The phosphorylation site was predicted to regulate only the C-terminal death domain, in which it resides. Indeed, as the N-terminal half of Apoptin (1-60), which lacks the phosphorylation site altogether, can still kill tumor cells, it is clear that a full-length version mutated only in the phosphorylation site, though reduced in death activity, might nevertheless still contain residual death activity due to its intact N-terminus. Indeed, earlier experiments confirmed this prediction: a number of experiments showed that Ala- replacements in the region of the triple-T locus (T106-T107-T108) in an otherwise wild type Apoptin (hence, full length) caused an impairment in apoptosis, though did not abolish it completely.

Therefore, the best way to show that the kinase capable of phosphorylating Apoptin on ThrlOδ in an aberrant-specific way is specifically necessary for activating the process of aberrant-specific apoptosis from the C- terminal death domain, was to prepare and test a C-terminal phosphorylation mutant in tumor cells. Specifically, this mutant would consist of three parts: 1) an N-terminal maltose binding protein (MBP) stabilization tag (as C-terminal fragments of wild- type Apoptin are rather unstable); 2) the Apoptin amino acids 80-121, which contain the phosphoryation region and which have been shown previously to induce significant death activity in tumor cells (co-pending European patent application EP 02076597.0, and 3) alanine substitution mutations in the region of phosphorylation. We selected two different loss-of- function mutations, TTTPS [106- 110] AAAAA (also known as 5-Ala-106) and TT[107-108]AA (also known as TAA), both of which were proven to be non- phosphorylatable in vivo (Rohn et al, 2002 and data not shown, respectively) by standard ortholabeling methods. Constructs containing these two mutations were called pcDNA-MBP-80-121-5-ala-106 and pcDNA-MBP-80-121-TAA. As controls, we also used pcDNA-MBP-80-121 (C-terminal Apoptin wild-type) as a control as well as pcDNA-MBP-Apoptin (full-length wild-type).

We transfected pcDNA-MBP-80-121-5-ala-106, pcDNA-MBP-80-121- TAA, pcDNA-MBP-80-121 and pcDNA-MBP-Apoptin, along with a plasmid encoding the non-apoptotic cytoskeletal protein Desmin as a negative control into Saos-2 and HeLa tumor cells and scored for apoptosis several days later. These experiments showed that, in both tumor cell types, whereas pcDNA-

MBP-Apoptin could kill robustly, and the truncated pcDNA-MBP-80-121 had a reduced but significant death activity as expected due to the lack of the N- terminal death domain (about 70% of that the full-length), neither independent loss-of-function phosphorylation mutants pcDNA-MBP-80-121-5- ala-106 and pcDNA-MBP-80-121-TAA induced apoptosis above the background levels induced by the Desmin control. Phosphorylation of the full-length and C- terminal (but wild-type) fragment under these conditions was confirmed by double-staining with 108-P, the phophsophospecific antibody.

These results support our finding that the kinase capable of phosphorylating Apoptin (on Thrl08) in an aberrant-specific way is specifically necessary for activating the process of aberrant-specific apoptosis, and that this action occurs via the C-terminal death domain.

Identification of endogenous substrates of the tumor-specific kinase using an antibody specific for phosphorylated Apoptin

To identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way, we performed the following experiment. First, we prepared lysates of aberrant cells versus normal cells. Because there can be cell-type differences in the phosphorylated proteome, we made sure to include matched sets or transformed cells and their normal cell counterparts: VHIO skin fibroblasts versus an SV40-transformed VHIO cell line (VHSV), and primary passage 1 breast-derived keratinocytes versus an SV40-transformed keratinocytes line (SVK14). To ensure that we did not only analyze transformed or SV40-influenced proteomes, we also included several other normal (human mesenchymal stem cells) and tumor cells (Saos-2 and U20S). Equal amounts of each protein samples were analyzed by SDS-PAGE and Western blot analysis with the 108-P antibody. We also included a positive control, namely recombinant Apoptin protein that had been phosphorylated in υitro on T108 by a tumor cell lysate, and a negative control, namely a parallel PVDF probed only with the secondary antibody.

Using these comparisons, we have shown that several strongly reactive proteins were clearly present in all transformed or tumor cells but either lacking or only very faintly visible in all normal cells tested, including prominent aberrant-specific bands migrating at approximately 170 kD, 140 kD, 97kD, 50kD, 30 kD and 18 kD. Use of SV40 large T retrovirus to help identify endogenous substrates of the tumor-specific kinase.

In order to analyze more genetically similar aberrant vs. counterpart cells, it is desirable to develop a method to turn on the tumor-specific kinase in essentially 100% of normal cells, and then compare the 108-P protein profiles on kinase-activated versus control cells. We have achieved this goal by using a retrovirus expressing the SV40 large T antigen. Namely, we have infected normal murine embryo fibroblasts (MEFs) with an ecotropic SV40 large T retrovirus or with a control GFP-encoding retrovirus, comparing these samples to uninfected cells for an additional negative control. Cells were harvested for the in υitro kinase assay using VP3-His and assayed by 108-P Western blot. (Parallel assays confirmed an essentially 100% transduction efficiency). Indeed, while there was virtually no kinase activity in the uninfected MEFs or in the MEFs infected with the GFP control virus, the SV40 large T virus induced massive activation of the tumor-specific kinase in this assay.

Because these experiments showed that SV40 large T protein did indeed switch on the kinase capable of phosphorylating Apoptin in an aberrant- specific way on T108, we next wanted to determine whether any of the natural endogenous substrates of said kinase were similarly switched on in the same experiment. To this end, we took in parallel the same protein lysates that had been used in the in υitro kinase reaction described in the previous paragraph, ran them on an SDS-PAGE gel and probed the Western blot with 108-P. (Note that these lysate supernatants were prepared in a minimal, detergent-free kinase buffer to facilitate the kinase assay, so not all substrate(s) may have been liberated compared to when we lysed cells in stringent RIPA buffer). This experiment revealed that infection by the large T retrovirus, but not by the control GFP retrovirus or by a mock infection, did indeed lead to the detectable phosphorylation of endogenous substrates of sizes consistent with differential bands already seen in our 108-P profiling panel described above (VH10 vs. VHSV; SVK14 vs. keratinocytes, etc.). Specifically, detection of the 50kD substrate was robustly upregulated compared to controls. Milder but significant increases were also seen of the 97kD, 30kD, 170kD and 140kD substrates. These results confirm that retrovirus infection of non- aberrant cells with SV40 large T help to identify and/or confirm natural endogenous substrates of the kinase capable of phosphorylating Apoptin in an aberrant- specific way. Moreover, use of minimal kinase buffer versus RIPA buffer can also help understand the solubility and compartmentalization of said substrates, which could further aid in identification. Finally, performing the profiling in an in vitro kinase assay optimized for the tumor-specific kinase may boost the level of phosphorylation of the substrates over the normal endogenous levels, thus enriching the relevant signals.

Determination of the minimal Apoptin-kinase-activating domain of SV40 large T to further refine the invention.

The large T antigen of SV40 is a complex protein comprised of multiple transformation domains, including the DNA-J region, a region that binds and inhibits Rb, and a region that binds and inhibits p53. One can expect that the entire large T protein could activate and inhibit a whole host of cellular responses and pathways and genes, only some or one of which is important for activating the kinase capable of phosphorylating Apoptin in an aberrant- specific way. Therefore, it is of importance to minimize the background seen in our various 108-P profiling assays such as those described in this invention by narrowing down the region of SV40 large T required to turn on the kinase capable of phosphorylating Apoptin in an aberrant-specific way. Such a refinement, in decreasing the background, will increase the chances of successfully identifying the relevant and useful substrate(s).

We have already shown in our laboratory using various mutated or deleted large T constructs that a region of large T between amino acids 1-82 in the N-terminus of the protein (including the so-called DNA-J domain), when targeted to the nucleus, is responsible for conferring upon otherwise normal cells the ability to direct Apoptin to the nucleus and to thereby kill that cell by apoptosis. The smallest construct we have that contains this minimal Apoptin- activating determinant (MAAD) is represented by SV40 nls-small t, a molecule consisting of the DNA-J domain of SV40 large T followed by a unique C- terminus, all fused to a C-terminal SV40 nuclear localization signal (NLS).

To confirm that SV40 large T gene was able to turn on the tumor- specific kinase, as already shown in the above retrovirus experiments, and to determine whether the MAAD, as represented by the construct nls-small t, was sufficient to perform this function, we microinjected the nuclei of normal VHIO fibroblasts with the following DNAs: CMV-VP3 plus SV40 large T, CMV- VP3 plus nls-small t, and CMV-VP3 plus an empty vector as a negative control. At six hours and 24 hrs post-injection, we fixed the cells and stained them simultaneously for 108-P (phospho-Apoptin) and 111.3 (basal Apoptin regardless of phosphorylation state), whereas parallel microinjections and stainings confirmed the correct expression and localization of the corresponding SV40 proteins using the antibody 419. In addition to antibody staining, we inspected the apoptotic state of the cells using DAPI.

Our results were as followed. First, as expected, in the control cells co- injected with Apoptin plus empty vector, Apoptin was located primarily in the cytoplasm and did not induce significant levels of apoptosis (1% by 24 hours). In these cells, the ratio of phosphorylated Apoptin to basal Apoptin (i.e., the ratio of 108-P-positive cells to 111.3-positive cells) was 14% and 11% at 6 hours and 24 hours, respectively, which is consistent with the percentage of transiently phosphorylated Apoptin in normal cells (see a more detailed description of this phenomenon below in the section about cell-cycle synchronization). In contrast, cells co-injected with SV40 large T and CMV- Apoptin showed drastically increased nuclear localization of Apoptin and significant levels of apoptosis by 24 hours (17%). Strikingly, the ratio of 108-P to 111.3 correspondingly increased to 45% at 6 hr, although it had decreased to 17% by 24 hours, perhaps indicating that kinase activity starts quickly but soon fades in this ectopic expression system. Finally, cells co-injected with nls- small t and CMV- Apoptin also showed drastically increased nuclear localization of Apoptin and significant levels of apoptosis by -24 hours (19%). Strikingly, the ratio of 108-P to 111.3 correspondingly increased to 66% at 6 hr, and was maintained at 60% even at 24 hours. One possible explanation why small t showed more sustained kinase activity than did large T is that large T acts faster and had already peaked before 6 hours. Another possibility is that regions downstream of the DNA-J domain of large T may actually impair or dampen the activation of the kinase.

As a second method of confirming which portion of SV40 large T antigen is responsible for turning on the kinase capable of phosphorylating Apoptin in an aberrant-specific fashion, we also used a biochemical approach. Namely, we co-transfected low-passage CD31- human diploid fibroblasts, which are normal cells, with pCMV-VP3 plasmid plus the following SV40 early region-encoding plasmids: pRSV-LT (wild-type Large T antigen); pRSV-<iZ1135 (LT lacking 17- 21, the DNA-J domain); pRSV-TNl36 (LT 1-136]); pRSV-st (wild-type small t); pRSV-nls-st (small t antigen with an N-terminal heterologous NLS); or pCMV- desmin as a negative control. Three, 6 or 24 hours post-transfection, cell lysates were harvested, quantitated, equal amounts of protein were resolved on SDS-PAGE and Western blotted to PVDF membranes, and these were probed with 108-P, the phosphospecific antibody against phosphorylated Apoptin. (Subsequent re-probing of PVDF membranes with mAb 111.3 confirmed the presence of basal Apoptin). This experiment showed that all SV40 proteins shown previously to activate Apoptin's ability to induce apoptosis in normal cells, presumably by activating the relevant aberrant- specific pathway(s), were able to cause the phosphorylation of Apoptin on Thrl08. Specifically, co-transfection with LT, TN136 and nls-st resulted in a robustly phsophorylated Apoptin protein on the Western blot after 108-P staining, whereas co-transfection with 1135 and st did not. In parallel, as an independent assessment of kinase activity, the above CD 31- fibroblast transfections had contained glass cover slips with cells growing on top of them that were removed at 3, 6 or 24 hours post- transfection, fixed and stained as described above for microinjected VHIO cells. The ratio of 108-P positive to 111.3 positive transfected cells was scored as described for the VHIO cells. This analysis showed again a correlation with the ability of the SV40 early region mutant to activate Apoptin with its ability to phosphorylate Apoptin; specifically, whereas co-transfections with LT, 136 and nls-st showed that approximately 70% of all Ap op tin-containing cells harbored the phosphorylated form, that number was drastically lower (10-20%) for cells co-transfected with 1135 or st which was similar to background phosphorylation in the presence of the negative control plasmid.

These results confirm, in two different non-aberrant cell strains and using several independent methods that the MAAD resides between residues 1-136 and is heavily influenced by the presence of the DNA-J domain and furthermore should be present in the nucleus to switch on the aberrant-specific pathway. Further, these data show that the activation of the kinase capable of phosphorylating Apoptin in an aberrant-specific fashion on Thrl08 is a downstream event following the over-expression of the MAAD, and can therefore be thought to reside in the same aberrant-specific pathway.

In summary, these results suggest that the MAAD might be useful not merely for eliminating background bands in the 108-P profiling assays, but also may actually act as a more efficient, robust or sustained stimulator of the kinase capable of phosphorylating Apoptin and its natural substrates in an aberrant-specific manner. Therefore, the use of inducible cell fines or viral transduction systems or protein transduction systems in our 108-P profiling assays to identify natural endogenous substrates is likely further facilitated by the use of the MAAD in addition to the already quite efficacious SV40 large T. Use of immunofluorescence assays to help identify the endogenous substrate(s) of the tumor-specific kinase

In addition to using the 108-P antibody in biochemical manipulations, the tumor-specific kinase's substrates can also be visualized -using immunofluorescence analysis (IFA). For example, we cultured untransfected Saos-2 tumor cells on chamber slides and performed IFA analysis to see whether any endogenous signal could be visualized. We saw very faint staining in all cells, but we saw a particularly strong signal in cells undergoing mitosis, beginning when the nucleus began to condense during early prophase, continuing all the way through cytokinesis and being retained in newly- separated daughter cells. The staining pattern was diffuse and slightly speckled. During telophase and cytokinesis, the staining was particularly prominent in the midbody between the nascent daughter cells. Although fluorescence artifacts can sometimes be amplified during mitosis, we did not see a similar level of staining when the analysis was repeated in parallel with either fluorochrome-conjugated secondary antibody alone, or during IFA with an irrelevant primary antibody. Therefore, this protein probably represents one endogenous substrate of the tumor-specific kinase. Although the protein cannot be further identified on the biochemical level by this IFA approach, it does facilitate analysis of other strategies. For instance, the biochemical comparisons performed above could be repeated on fractions of cells arrested in mitosis with for example drugs, or enriched to be in mitosis by for example collection at the appropriate time after synchronization. As a further example, mitotic kinases, especially those known to be active in the midbody, can be assessed specifically for a role in the tumor-essential pathway.

Use of cell cycle synchronization to help identify endogenous substrates of the tumor-specific kinase.

Several lines of evidence indicate that the kinase capable of phosphorylating Apoptin in an aberrant-specific manner might be a cell-cycle- related kinase. First, the above result showing 108-P staining in non- transfected cells most prominently during mitosis, especially late mitosis in the midbody, is supportive of this notion. Secondly, when normal cells are given Apoptin by, for example, transfection or microinjection; a small fraction (5-15%) exhibit what appears to be transient, temporally regulated kinase activity, as assessed by 108-P staining of Apoptin in these cells - a result also consistent with a kinase activity that switches on at some transient point during the cell cycle. We hypothesize that whereas normal cells possess kinase activity only transiently, during one discrete, tightly-regulated time-point in the cycle, tumor cells contain constitutive, dysregulated kinase activity during all points in the cell cycle (perhaps peaking slightly during the time-point when the kinase is normally supposed to switch on). This constitutive activity is clearly seen when tumor cells are transfected with Apoptin: more than 80% of the Apoptin is positive for the 108-P at early and middle time points post- transfection. This kinase dysregulation would lead to or contribute to an aberrant environment. Apoptin is not able to kill normal cells likely, at least in part, because Apoptin is phosphorylated only transiently by the non- aberrantly-regulated kinase, and that this activation does not coincide with the window of activity of one of the effectors crucial for Apoptin's execution of apoptosis; in contrast, Apoptin remains constitutively phosphorylated in tumor cells, including at a time coinciding with said effector, so that Apoption is able to kill these dysregulated cells.

If the kinase is cell-cycle-regulated, it is likely that one, several or all of its natural endogenous substrates are also regulated by in a cell cycle- dependent manner. To investigate this possibility, we took MEL cells enriched for various different stages of the cell cycle and examined on the one hand, kinase activity in the 108-P in vitro kinase assay using VP3-His as a substrate, and on the other, the 108-P staining profile of endogenous phospho-proteins in these kinase assay lysates. MEL cells are aberrant, so as expected we found robust kinase activity in all stages of the cell cycle as well as in untreated, exponentially growing unsynchronized cell cultures. However, we did notice a modest increase in VP3τHis phosphorylation at 5 and 6 hours post-nocodazole release, time-points which correspond to the Gl/S transition and S-phase, respectively.

Furthermore, we noticed that unsynchronized cells possessed the lowest kinase activity; this result is consistent, as statistically, roughly 70% of unsynchronized cells will not be in S-phase at any given time.

In addition to the added Apoptin substrate, we also examined other bands that appeared on the same in vitro kinase assay, as these may correspond to natural endogenous substrates phosphorylated by the reaction conditions on the equivalent of T108 by the same kinase that phosphorylates Apoptin in an aberrant-specific manner. First we inspected the region of approximately 50kD because we had noticed in the past that a "background band" could frequently be detected in that region. Indeed, this 50 K band showed very strong cell-cycle-dependent levels of phosphorylation in this assay. Specifically, the activity was very low in nocodazole-arrested mitotic cells, increased robustly at the late mitosis/Gl boundary and stayed at that high level through to the Gl/S transition. Activity began to wane in S-phase down to low levels, and did not increase while traversing back to M-phase. This band of 50kD could very well be the same substrate as the 50kD protein we noticed in our 108-P profiling panel described above (VH10 vs. VHSV; SVK14 vs. keratinocytes, Saos-2 and U20S vs. mesenchymal stem cells.). Moreover, such a late-mitotic phosphorylated substrate could also correspond to the endogenous substrate seen in the immunofluorescence assay described above in untransfected cells, or may represent another mitotic substrate. Finally, we noticed a band of approximately 97kD that also showed cycle-dependent differences, peaking in nocodazole-arrested cells, dropping steeply at the end of mitosis and remaining low until S-phase, when another modest peak occurred. This band of 97kD could very well be the same substrate as the 97kD protein we noticed in our 108-P profiling panel described above.

These results demonstrate that further separation of reference cells and aberrant cells into cell-cycle staged populations is useful to help further identify, enrich or otherwise confirm natural endogenous substrates of the kinase capable of phosphorylating Apoptin in an aberrant-specific way, especially as our desired kinase activity, as shown above, is higher in discrete stages than in unsynchronized cells.

Use of cell cycle arrest and stage-specific harvesting to help identify endogenous substrates of the aberrant-specific kinase in normal cells

In addition to using aberrant cells to identify activated downstream endogenous substrates of the aberrant-specific kinase, we exploited the fact that the kinase in question may also be transiently activated in normal cells (see the beginning of the previous section on "cell cycle synchronization") to seek out natural endogenous substrates in normal cells. To explore this issue, we decided to see whether we could isolate discrete cell cycle stages in which the kinase that phosphorylates Apoptin on ThrlOδ was transiently activated. After collecting the cells in various cell cycle stages, cells were prepared for the standard in vitro kinase assay as described in the methods and the assay was performed to determine whether the cells contained an activity able to phosphorylate the Apoptin substrate on Thrl08. The cell cycle stage was confirmed by probing parallel Western blots with antibodies against telltale cyclins that are enriched or absent in particular stages, especially cyclin B, which comes up in G2 and peaks at the G2/M border and during mitosis. These experiments showed that an activity that can phosphorylate Apoptin on Thrl08 is so dilute when assayed on a population-wide level in a biochemical lysate (as opposed to on the single-cell level via immunofluorescence, as reported earlier in the Results section) that it was virtually undetectable in normal log-phase cells (cyclin B was also virtually undetectable in the population, indicating that the percentage of mitotic cells is too low to see telltale proteins associated with mitosis under these conditions). Also, both kinase activity and the presence of cyclin B was completely undetectable in stationary-phase cells arrested in GO by natural contact inhibition properties. Moreover, Apoptin kinase activity was negligibly, or only faintly detectable in cells arrested in S-phase (confirmed by absence of cyclin B) and in cells well released into late Gl phase (confirmed by gradual disappearance of cyclin B post-release from mitotic arrest). In strong contrast, Apoptin kinase activity was extremely high in cells that had been arrested in mitotis by two independent drugs (vincristine and nocodazole) as well as in dislodged mitotic cells isolated by standard "tapping" procedures that had never been exposed to drugs (in all three cases, very strong cyclin B signals confirmed that the cells were in mitotis, as did inspection of cells on coverglasses stained with DAPI to see the mitotic figures of the chromosomes). These results show that normal cells do harbor Apoptin kinase activity during one discrete stage of the cell- cycle: namely, G2/M border and mitosis proper.

These results open the door for a procedure to compare cells likely to possess activated downstream substrates of the kinase capable of phosphorylating Apoptin in an aberrant-specific fashion with reference cells that do not. One can simply compare genetically identical normal, non- adherent cells to parallel cells enriched for mitosis, and then analyze which substrates are specific to the latter population as already described herein. While drugs could be used for the arrest, it is more preferable to use a way of harvesting in mitosis that is likely to cause the minimum of irrelevant differences between the test cells and the reference cells, such as the "tapping" procedure mentioned above, based on the fact that mitotic cells are poorly attached to the plate, or other ways known in the art, e.g. elutriation centrifugation, which exploits the shape of cells during the cell cycle to harvest them at mitosis. Inhibition of the tumor-specific kinase

To determine whether one form of Apoptin can successfully inhibit the kinase's ability to phosphorylate another form of Apoptin, we performed a standard non-radioactive in vitro kinase assay using Saos-2 tumor cells as the kinase donor and recombinant VP3-His as the substrate, either alone or inhibited with a 15-fold molar excess of MBP-Apoptin protein. Because these two proteins have very different mobilities on a gel, their relative phosphorylation state could be assessed after Western blot and analysis with the 108-P antibody. This experiment showed that the 15-fold excess of MBP- Apoptin completely inhibited phosphorylation of VP3-His, and that furthermore this inhibition was due to competition, as the MBP-Apoptin was shown to be hyperphosphorylated. In the same experiment, a 200x molar ratio of short unphosphorylated peptide (the same 13 amino acid peptide used to inoculate the animals that produced 108-P antisera) was unable to compete for the kinase, as VP3-His was robustly phosphorylated. This result was not surprising, as this same peptide was too short to even be phosphorylated in vitro by the tumor-specific kinase in the absence of Apoptin. We predict that a peptide corresponding to Apoptin residues 80-121 (a 41-mer) will serve as an inhibitor, because it is known from deletion mutant studies that a region from 80-121 is phosphorylated as well as the full-length protein in vivo, whereas a region from 100-121, although containing the phosphorylation site, is poorly phosphorylated. Furthermore, it is known from alanine replacement studies that a full-length Apoptin containing a stretch of five alanines from positions 91-96 is poorly phosphorylated, and therefore may represent a facilitation domain of kinase activity (e.g. a distal docking site or conformational facilitator). Further modifications to the peptide length and sequence using strategies known in the art, in conjunction with the in vitro kinase assay as a readout, will aid in finding the most optimal inhibitor.

As a first step in generating an optimal, inhibitory (phosphorylatable) peptide substrate, we obtained the following two peptides: N-terminally biotinylated 80-121 (called Bio80), and N-terminally biotinylated 80-121 mutated at the phosphorylation site: 106-108 replaced with 3 Ala residues (called Bio80-Null). The biotin tag was employed merely to aid in detection and stability. These two substrates were evaluated in the in vitro kinase assay described above, compared with the 13 a.a peptide and full-length MBP- Apoptin as known controls. In the direct radioactive phosphorylation assay both the Bio80 and the Bio80-null peptides were phosphorylated by lysates in a tumor non-specific manner. Upon Western blotting, both peptides were recognized equally by the phospho-nonspecific antibody VP3C. However, neither peptide was detected above background by the 108P polyclonal antibody which did specifically recognize MBP-VP3 in the same assay. The smaller 13 a.a. peptides gave the same results as before, which is to say that they are readily radiolabelled in a tumor-non-specific and T108 independent manner and that both are readily recognized by 108P-specific antibody even when un-phosphorylated (e.g. straight from the synthesizer). Although the results implied that the T108 residue of the Bio80 peptide might not be labeled, we tried to determine whether there was any interaction of the peptide with the T108 kinase, by using both Bio80 and Bio80-null as inhibitors of phosphorylation of full-length Apoptin. We performed several radioactive IVKAs using a C-terminally histidine-tagged protein (His-VP3) as the readout and Saos-2 lysates which contain highly active Apoptin-kinase. To this we added an 180x molar excess (peptide over His-VP3) of either Bio80 or Bio80- null. If specific phosphorylation of the T108 residue occurs on the peptide, then it is predicted that the Bio80 peptide is, but the Bio80-null peptide is not capable of inhibiting phosphorylation of His-VP3 in this assay. We found that indeed both at the level of radioactivity as well as by using the 108P-specific polyclonal antibody the Bio80 significantly decreased detection of phosphorylated His-VP3, whereas Bio80-null did not have this effect. Other experiments showed that the inhibition was dose-dependent, although at 50x molar excess the inhibition already became insignificant, showing that the peptide has a low-affinity interaction with the kinase compared to the His-VP3 molecule.

Together the data show that although the 108P antibody was unable to detect the Bio80 peptide, the peptide must indeed interact with the T108- kinase in a lysate. Otherwise the observed inhibition would not occur. These experiments clearly showed that whereas Bio80 was an excellent competitive inhibitor of Apoptin kinase, the same peptide mutated at the phosphorylation site could not inhibit the kinase. Peptides such as Bio80 or its derivatives or analogues clearly have utility for determining the role of the kinase in maintaining the integrity of tumor cells in vivo and form the basis for developing a peptidemimetic drug for cancer therapy.

Taken together, these experiments indicate that a long peptide corresponding to full-length Apoptin (121 amino acids) functions as an excellent inhibitor of the tumor specific kinase, as does a peptide corresponding to 80-121, provided that the phosphorylation site is intact, but that the tested 13 amino acid peptide is too short.

We conclude that it is indeed possible to inhibit the tumor-specific kinase capable of phosphorylating Apoptin in an aberrant-specific way without having to know the identity of said kinase.

Identifying the tumor-specific kinase

In addition to finding aberrant-specific pathways without knowing the identity of the kinase in question, one is also able to exploit knowledge of the kinase to help find endogenous substrates of the kinase. For example, knowledge of the kinase could lead to the identification of a cDNA clone of said kinase which is then overexpressed, compared to reference cells, and then these cells are analyzed for the differential appearance of activated substrates downstream as already described. Another example is that solving the structure of a kinase once it is identified is useful in the generation of a very specific small molecular inhibitor thereof which is then used to treat reference cells for differential analyses as already described.

We evaluated whether isolating Apoptin kinase by affinity was feasible by using the biotinylated Apoptin peptide 80-121 (Bio80) described in the previous section. We expected that the peptide is capable of inhibiting the kinase by binding to the active site of the kinase; if this affinity is long or strong enough, said interaction is then exploited to purify the kinase out of a kinase-active lysate. We performed the standard in vitro kinase assays in tumor lysates as described, using either biotinylated Bio80 or Bio80-null (see previous section) as a substrate. After the reaction was completed, we "fished" out the biotinylated peptides with Streptavidin beads and then re-assayed the reacted lysates (by performing a second in vitro kinase assay but this time using MBP-Apoptin as a substrate) to see whether this "fishing" action had depleted the lysates of active kinase. Our results showed that Bio80 was indeed able to deplete an otherwise kinase-active lysate of its ability to phosphorylate Apoptin on Thrl08. This result showed that the kinase bound tightly enough to the biotinylated substrate to be removed from the lysate by means of the beads along with the peptide that it was phosphorylating. As a confirmation of this inference, the Bio80-Null peptide which is unable to competitively inhibit the kinase

(presumably because in lacking the consensus phosphorylation site it does not bind well or long to the kinase, if at all) was indeed unable to deplete activity in the same experiment under the same conditions. These data suggest that the kinase capable of phosphorylating Apoptin in an aberrant-specific fashion on Thrl08 can be purified and subsequently identified (using techniques known in the art, e.g. scaling up of the procedure and mass spectrometric analysis) using affinity approaches similar to, but not limited to, the one we have described.

Another method to identify the kinase capable of phosphorylating Apoptin in an aberrant-specific fashion on Thr 108 as a prelude to seeking out aberrant-specific pathways is to use the knowledge of the kinase's behavior accumulated thus far, hypothesize on the nature of said kinase, and then to confirm this hypothesis using already existing specific inhibitors of said known kinase. Two pieces of data thus far indicated that Apoptin might be phosphorylated by a cyclin-dependent kinase or a kinase very similar to that family of kinases. First, standard protein database searches indicated that the consensus site of phosphorylation of Apoptin on Thr 108 is very similar to the phosphorylation site of known CDKl (also known as cdc2) substrates. Secondly, Apoptin kinase activity is enriched during mitosis during the normal cell cycle, which coincides with when CDKl is active. To analyze the possibility that Apoptin kinase might be in the cyclin-dependent kinase family, we tested the ability of the inhibitor roscovitine (an inhibitor of CDKl, CDK2 and CDK5) to inhibit the ability of tumor lysates to phosphorylate MBP-Apoptin substrate in the standard radioactive in vitro kinase assay (as described in co-pending application PCT/NLOl/00771 as well as in Rohn et al, 2002). These results showed that the inhibitor could indeed prevent the in vitro phosphorylation of Apoptin substrate in a dose-dependent manner (IC50 in the nanomolar range). Repeating the experiment with the known CDK substrate Histone HI yielded similar kinetics of inhibition with an IC50 in the same nanomolar range. To assess the ability of a standard inhibitor of cyclin-dependent kinases to reduce the phosphorylation of Apoptin in Thr 108 in living Saos-2 cells, we transfected Saos-2 cells growing on coverglasses with pCMV-VP3 and 3 days later, treated them with the CDK inhibitor butyrolactone, or with DMSO as a vehicle-only mock control, prior to fixation and immunofluorescence with 111.3 and 108-P antibodies. This experiment revealed that Apoptin-positive cells demonstrated a visual reduction in both the proportion and strength of the Thr- 108 phosphorylation signal as a result of treatment, suggesting that the inhibitor had reduced somewhat the activity of the kinase capable of phosphorylating Apoptin in an aberrant-specific manner on Thrl08. Taken together, these experiments suggest that Apoptin might be phosphorylated by a cyclin dependent kinase. Hence, if a tumor- specific kinase exists and if such a kinase is necessary for tumor survival then CDK inhibitors can be used in a tumor therapy.

FIGURE DESCRIPTION

Figure 1. The amino acid sequence of Apoptin.

REFERENCES

Baker, S. J., Markowitz, S., Fearon, E.R., Willson, J.K.V., and Vogelstein, B. (1990). Suppression of human colorectal carcinoma cell growth by wild-type p53. Science 249, 912-915.

Bellamy, C.O.C., Malcomson, R.D.G., Harrison, D.J., and Wyllie, H. (1995). Cell death and disease: The biology and regulation of apoptosis. Seminars in Cancer Biology 6, 3-12.

Blaydes, J. P., Vojtesek, B., Bloomberg, G. B., and Hupp, T. R. (2000). The development and use of phospho-specific antibodies to study protein phosphorylation. Methods Mol Biol 99, 177-89.

Danen-Van Oorschot, A.A.A.M., Fischer, D.F., Grimbergen, J.M., Klein, B., Zhuang, S.-M., Falkenburg, J.H.F., Backendorf, C, Quax, P.H.A., Van der Eb, J.A., and Noteborn, M.H.M. (1997). Apoptin induces apoptosis in human transformed and malignant cells but not in normal cells. Proceedings National Academy Sciences, USA: 94, 5843-5847.

Danen-Van Oorschot, A.A.A.M, Den Hollander, A., Takayama, S., Reed, J., Van der Eb, A.J. and Noteborn, M.H.M. (1997a). BAG-1 inhibits p53-induced but not Apoptin-induced apoptosis. Apoptosis 2, 395-402.

Diller, L., Kassel, J., Nelson, C, Gryka, M.A., Litwak, G., Gebhardt, M.,

Bressac, B., Ozturk, M., Baker, S.J., Vogelstein, B., and Friend, S.H. (1990). p53 functions as a cell cycle control protein in osteosacomas. Molecular Cellular Biology 10, 5772-5781. Duke, R.C., Ocjius, D.M., Young, J, D-E. (1996). Cell suicide in health and disease. Scientific American December 1996, 48-55.

Earnshaw, W.C., 1995. Nuclear changes in apoptosis. Currenrt Opinion in Cell Biology 7, 337-343.

Evan GI, Lewis GK, Ramsay G, Bishop JM, (1985) Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product. Mol Cell Biol Dec;5(12):3610-3616.

Hockenberry, D.M. (1994). Bcl-2 in cancer, development and apoptosis. Journal of Cell Science, Supplement 18, 51-55.

Kennedy, A.S., DeMars, L.R., Flannagan, L.M., and Varia, M.A. (1995). Primary squamous cell carcinoma of the endometrium: a first report of adjuvant chemoradiation. Gynecology Oncology 59, 117-123.

Kerr, J.F.R., Winterford, CM., and Harmon, B.V. (1994). Apoptosis: Its significance in cancer and cancer therapy. Cancer 73, 2013-2026.

Levine, A.J. (1997). p53, the cellular gatekeeper for growth and division. Cell 88, 323-331.

Maniatis, T., Fritsch, E.F., and Sambrook, J. (1982). Molecular Cloning: A Laboratory Manual. CSHL Press, New York, USA.

McDonell T.J., Meyn, R.E., Robertson, L.E. (1995). Implications of apoptotic cell death regulation in cancer therapy. Seminars in Cancer Biology 6, 53-60. Noteborn, M.H.M. (1996). PCT application WO 96/41191. Apoptin induces apoptosis in human transformed and malignant cells but not in normal cells as essential characteristic for the development of an anti-tumor therapy.

Noteborn, M.H.M., and De Boer, G.F. (1996). Patent USA/no. 5,941,073 Cloning of chicken anemia DNA.

Noteborn, MHM, and Danen-Van Oorschot, A.A.A.M., (1998). International application PCT-NL98/00688 (WO 99/28461). Methods and means for inducing apoptosis by interfering with Bip-like proteins.

Noteborn, M.H.M., De Boer, G.F., Van Roozelaar, D., Karreman, C, Kranenburg, O., Vos, J., Jeurissen, S., Zantema, A., Hoeben, R., Koch, G., Van Ormondt, H., and Van der Eb, A.J. (1991). Characterization of cloned chicken anemia virus DNA that contains all elements for the infectious replication cycle. Journal of Virology 65, 3131-3139.

Noteborn, M.H.M., and Pietersen, A. (1998). A gene delivery vehicle expressing the apoptosis-inducing proteins VP2 and/or Apoptin. PCT Application no. PCT/NL98/00213

Noteborn, M.H.M., Todd, D., Verschueren, C.A.J., De Gauw, H.W.F.M., Curran, W.L., Veldkamp, S., Douglas, A.J., McNulty, M.S., Van der Eb, A.J., and Koch, G. (1994). A single chicken anemia virus protein induces apoptosis. Journal of Virology 68, 346-351.

Noteborn, M.H.M., and Zhang, Y. (1998). Methods and means for determining the transforming capability of agents, for determining the predisposition of cells to become transformed and prophylactic treatment of cancer using Apoptin-like activity. PCT Application no. PCT/NL98/00457 Noteborn, M.H.M., Danen-van Oorschot, A.A.A.M., Van der Eb, A.J. (1998a). Chicken anemia virus: Induction of apoptosis by a single protein of a single- stranded DNA virus. Seminars in Virology 8, 497-504.

Noteborn M.H.M., Zhang, Y.-H. and Van der Eb, A.J. (1998b). Viral protein Apoptin induces apoptosis in tumor cells and after UN-treatment in untransformed cells from cancer-prone individuals: A review. Mutation Research 400, 447-456.

Noteborn, M.H.M. (1999). Apoptin-induced apoptosis: a review. Apoptosis 4, 317-319.

Noteborn et al. (2002), PCT/NLO 1/00664 (WO 02/20809), A delivery method for the tumor specific apoptosis inducing activity of apoptin.

Noteborn et al. (2002; not published yet), PCT/NLOl/00771 (WO 02/32954), Phosphorylation modifications of Apoptin.

Noteborn et al. (2002; not published yet), EP02076597.0, Apoptin fragments.

Paulovich, A.G., Toczyski, D., Hartwell, H. (1997). When checkpoints fail. Cell 88, 315-321.

Pietersen, A.M., Van der Eb, M.M., Rademaker, H.J., Van den WoUenberg, D.J.M., Rabelink, M.J.W.E., Kuppen, P.J.K., Van Dierendonck, J.H., Van Ormondt, H., Masman, D., Van de Velde, C.J.H., Van der Eb, Hoeben, R.C., and Noteborn, M.H.M. (1999). Specific tumor-cell killing with adenovirus vectors containing the Apoptin gene. Gene Therapy 6, 882-892. Rohn, J. L,.Zhang, Y.-H., Aalbers, Otto, den Hertog, J., van de Velde, C, Kuppen, P., Donner, P., Mumberg, D. and M.H.M Noteborn (2002). A tumor- specific kinase activity regulates the viral death protein Apoptin. J Biol Chem 277, 50820-50827. Sachs, L. and Lotem, J. (1993). Control of programmed cell death in normal and leukemia cells: New implications for therapy. Blood 82, 15-21.

Steller, H. (1995). Mechanisms and genes of cellular suicide. Science 267, 1445-1449.

Telford, W.G., King, L.E., Fraker, P.J. (1992). Comparative evaluation of several DNA binding dyes in the detection of apoptosis-associated chromatin degradation by flow cytometry. Cytometry 13, 137-143.

Teodoro, J.G. and Branton, P.E. (1997). Regulation of apoptosis by viral gene products. Journal of Virology 71, 1739-1746.

Thompson, C.B. (1995). Apoptosis in the pathogenesis and treatment of disease. Science 267, 1456-1462.

Van der Eb, M.M., Pietersen, A.M., Speetjens, F.M., Kuppen, P.J.K., Van der Velde, C.J.H., Noteborn, M.H.M., and Hoeben, R.C. (2002). Gene therapy with Apoptin induces regression of xenografted human hepatomas. Cancer Gene Therapy 9, 53-61.

White, E. (1996). Life, death, and the pursuit of apoptosis. Genes and development 10, 1-15.

Wyllie, A.H. (1995). The genetic regulation of apoptosis. Current Opinion in Genetics and Development 5, 97-104. Wyllie, A.H., Kerr, J.F.R., Currie, A.R. (1980). Cell death: The significance of apoptosis. International Review of Cytology 68, 251-306.

Zhuang, S.-M., Landegent, J.E., Verschueren, C.A.J., Falkenburg, J.H.F., Van Ormondt, H., Van der Eb, A.J., Noteborn, M.H.M. (1995). Apoptin, a protein encoded by chicken anemia virus, induces cell death in various human hematologic malignant cells in vitro. Leukemia 9 SI, 118-120.

Zhuang, S.-M., Shvarts, A., Van Ormondt, H., Jochemsen, A.-G., Van der Eb, A. J., Noteborn, M.H.M. (1995a). Apoptin, a protein derived from chicken anemia virus, induces a p53-independent apoptosis in human osteosarcoma cells. Cancer Research 55, 486-489.

Claims

1. A method to identify a substrate of a kinase capable of- phosphorylating Apoptin in an aberrant-specific way, comprising:

5 - separating the components present in said lysates

10 2. A method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way according to claim 1, wherein said aberrant cells are tumor cells.

3. A method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way according to claim 1, wherein said

15 aberrant cells are cells involved in an auto-immune disease.

4. A method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way according to anyone of claims 1 to 3, wherein said reference cells are non-aberrant cells.

5. A method to identify a substrate of a kinase capable of phosphorylating 20 Apoptin in an aberrant-specific way according to anyone of claims 1 to 3, wherein said reference cells are cells in which phosphorylation of said substrate is suppressed.

6. A method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way according to claim 5, wherein said cells in

25. which phosphorylation of said substrate is suppressed are obtained by providing aberrant cells with an inhibitor of said kinase.

7. A method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way according to claim 5, wherein said cells in which phosphorylation of said substrate is suppressed are obtained by providing aberrant cells with Apoptin or a functional fragment or a functional equivalent thereof.

8. A method to identify a substrate of a kinase capable of phosphorylating Apoptin in an aberrant-specific way according to anyone of claim 1 to 7, wherein said molecule capable of recognizing phosphorylated Apoptin and phosphorylated substrate is an antibody or a functional fragment or a functional equivalent thereof.

9. A substrate obtainable by the method according to anyone of claims 1 to 8.

10. A method for obtaining a modulator of a kinase involved in phosphorylation of Apoptin in an aberrant-specific way comprising: - incubating aberrant cells with a possible modulator of said kinase

- determining the phosphorylation state of said Apoptin.

11. A method according to claim 10 wherein said modulator is an inhibitor.

12. A method according to claim 10 or 11 wherein said modulator is a proteinaceous substance.

13. An inhibitor obtainable according to anyone of claims 11 or 12.

14. A nucleic acid encoding an inhibitor according to claim 13.

15. A vector comprising a nucleic acid according to claim 14.

16. A gene delivery vehicle comprising an inhibitor according to claim 13 or a nucleic acid according to claim 14 or a vector according to claim 15.

17. A host cell comprising an inhibitor according to claim 13 or a nucleic acid according to claim 14 or a vector according to claim 15 or a gene delivery vehicle according to claim 16.

18. A pharmaceutical composition comprising an inhibitor according to claim 13 or a nucleic acid according to claim 14 or a vector according to claim 15 or a gene delivery vehicle according to claim 16 or a host cell according to claim 17.

19. Use of an inhibitor according to claim 13 or a nucleic acid according to claim 14 or a vector according to claim 15 or a gene delivery vehicle according to claim 16 or a host cell according to claim 17 in the preparation of a medicament for the treatment of a disease where enhanced cell proliferation or decreased cell death is observed.

20. Use according to claim 19 wherein said disease comprises cancer or auto-immune disease.

21. A method for selecting a binding molecule comprising providing a collection of binding molecules capable of binding aberrant cell specific apoptin and selecting at least one binding molecule from said collection that is capable of specifically binding at least one cellular protein that is phosphorylated in an aberrant cell and not phosphorylated in at least one reference cell.

22. A method for identifying a protein comprising providing a collection of proteins derived from aberrant cells and a similar collection derived from reference cells, contacting said collections with a binding molecule selected according to a method according to claim 21, and identifying a cellular protein specifically bound by said binding molecule in the collection of proteins from aberrant cells.