WO2003025157A1

WO2003025157A1 - A coculture system to identify proteins triggering redifferentiation of tumor cells

Info

Publication number: WO2003025157A1
Application number: PCT/CH2001/000557
Authority: WO
Inventors: Markus O. Imhof; Michel Aguet; Maximilien Murone
Original assignee: The Genetics Company
Priority date: 2001-09-14
Filing date: 2001-09-14
Publication date: 2003-03-27

Abstract

The present invention relates to a novel methd of coculturing cells for the identification of proteins triggering the redifferentiation of dedifferentiated carcinoma cells. The method comprises a simple coculture model using monolayered stroma cells, which are transfected with individual or pooled clones from a cDNA library, and genetically modified reporter tumor cells. Furthermore, the invention relates to the application of said method in a high throughput screen.

Description

A Coculture System to Identify Proteins Triggering Redifferentia- tion of Tumor Cells

Field of the Invention

The present invention relates to a novel method of coculturing cells for the identification of proteins triggering the rediffer- entiation of dedifferentiated carcinoma cells. Furthermore, the invention relates to the coculturing system itself and its use in high throughput screens.

Background of the Invention The term "tumor" is defined as the newly formation of tissue with spontaneous, autonomous and uncontrolled growth. Tumors are classified according to several systems (Pschrembl; Klinisches Worter- buch, 258, Berlin 1998) ; one of these systems is the classification according to their biological behaviour: 1. benign tumors with differentiated cells and slow local growth; 2. malign tumors showing atypical cells with nuclei polymorphism, anaplasia, infiltrating, destructive and fast growth and formation of metastasis; and 3. semi-malign tumors with the histological signs of malign tumors and local infiltration, however without the formation of metastasis. The loss of specific cell and tissue functions, i.e. the dediffer- entiation of already differentiated cells, is a typical character- istic of a malign tumor cell. Cells of a naturally occurring tumor do not dedifferentiate evenly, bot show heterogenous levels of de- differentiation, an observation that led to the assumption that

BESTATIGUNGSKOPIE host cells in the tumor environment influence the differentiation behaviour of tumor cells.

Cytodifferentiation therapies have been tried and used in the treatment of human malignancies for decades. The fundamental mechanism of this approach is to "push" dedifferentiated tumor cells back into a genetic pathway of maturation and/or differentiation and, therefore, to reverse the malignant phenotype of tumor cells. The execution of this therapy, however, is only possi- ble with an understanding of the relevant molecules that control cell differentiation and a realistic approach to manipulate the function of such molecules.

In a study by Guan, R.J. et al. ("Drg-1 as a differentiation- related, putative metastatic suppressor gene in human colon cancer", Cancer Research 60, 749-755 (2000)) an individual candidate tumor suppressor gene has been introduced directly into colon carcinoma cells in order to analyse the cellular and genetic response it provokes. As a result, differentiation was observed in some in- stances. However, the screening of a multiplicity of different compounds or proteins is not possible using such direct approach avoiding a coculture model.

In other attempts, non-proteinaceous chemical compounds were iden- tified, which specifically modulate gene activity in carcinoma cells preventing them to become invasive. Some of these compounds also lead to a certain degree of redifferentiation (Zhou et al. 2000. Rapid induction of histone hyperacetylation and cellular differentiation in human breast tumor cell lines following degra- dation of histone deacetylase-1. J. Biol. Chem. 275:35256-35263).

Models presently used to simulate biological responses to molecules, such as proteins or chemical compounds inducing a rediffer- entiation of tumor cells so far include experimental animals, ex- planted tissue slices and monolayer cell cultures. Recently, 3- dimensional in vi tro models of mammalian tissue have been described simulating the behaviour of cancerous human cells (WO 00/75286) . These models can be used for screening and testing of anticancerous drugs and assessing their toxicity for both the tumor and the surrounding tissue. According to one preferred embodiment described in the WO 00/75286 an in vi tro model of interaction between stro al and tumor cells is described, wherein the stromal cells grow as spheroids on microcarrier beads, where after the tumor cells under study are allowed to adhere on top of the stromal cell layer. The drug to be tested is then added to the spheroids, i.e. into the culture medium and its effectiveness on the respective tumor cells is studied.

Such models of human tissue frequently show a high degree of morphological resemblance to their in vivo counterparts. However, these structures are usually grown following complicated protocols and even though useful for screening for agents with clinical utility, they are poorly adapted to high throughput screening procedures, in particular for screening large libraries of new compounds. Moreover, the presently existing coculture models so far only allow for the screening of compounds added to the "environment", i.e. into the respective medium of the test cell system. That means, that with the present coculture models an "inherent" testing of compounds, such as proteins expressed by the cells of the test system, has not yet been possible. Therefore, the existing models are only able to test compounds aiming at destruction but not at an alteration, i.e. redifferentiation of the tumor cells.

Furthermore, to avoid the shortcomings of unspecific chemical compounds for induction of redifferentiation programs in carcinoma cells proteins have to be identified, which have a specific impact on this process. Targeted application of such newly identified therapeutic proteins will ultimately eliminate the side effects in normal cells and tissues as they are observed with the current ap- plication of unspecific chemicals in cancer therapy.

Thus, there exists a strong need for a test cell system comprising a simple to handle coculture system of monolayered transfected "normal" cells with carcinoma cell lines to score for differentia- tion-promoting proteins being effective on said cancer cells by introducing the genes coding for said proteins into said cells. Furthermore, such a system needs to be suitable for carrying out high throughput screenings.

Summary of the Invention

The here described system employs stroma cells, originating from various cell types, tissues and/or organs (eg. mouse or human fi- broblasts) transfected with a cDNA expression library, along with genetically modified carcinoma reporter cells in order to identify proteins and combinations thereof specifically inducing rediffer- entiation programs.

Accordingly, the most preferred embodiment of the present invention is the provision of a simple coculture system using monolay- ered stroma cells, which are transfected with individual or pooled plasmids from a cDNA expression library, and genetically modified reporter tumor cells.

Another embodiment cf the invention is the identification of pro- teins and combinations thereof that "push" dedifferentiated carcinoma cells towards redif erentiation or, alternatively, induce cy- totoxic or cytostatic effects. In a further embodiment the coculture system of the present invention is applicable in high throughput screenings.

Other advantages, objects and features of the present invention will be readily apparent to those skilled in the art from the following detailed description of the preferred embodiments in conjunction with the accompanying drawings and claims.

Brief Description of the Figures

Figure 1 is a graph showing the progression of a tumor towards metastasis (adapted from Nicolson and Moustafa 1998. Metastasis-associated genes and metastatic tumor progression. In vivo 12:579-588).

Figure 2 illustrates the concept of redifferentiation of a dedif- ferentiated carcinoma and its metastasis.

Figure 3 shows the principle of a coculture system according to the present invention.

Detailed Description of the Invention

The present invention is situated in the field of tumor morphogenesis. Epithelial tumor cells undergo a transition from neopla- sia towards relatively benign adenoma and further to carcinoma in si t u, invasive carcinoma, and finally metastatic cells. This transition is accompanied by gradual cellular dedifferentiation . Adenoma cells are still differentiated; carcinoma cells however, lost most differentiation markers. The objective of this invention is to take advantage of a coculture model to identify proteins, which invert this process and "push" back dedifferentiated carcinoma cells towards differentiated cells or exert cytostatic or cyto- toxic effects. The power of this coculture system and its application in high throughput screens resides in the potential to identify signaling factors, which are expressed in the stroma cells and affect the behaviour of the cocultured cells of varying origin. Obviously, this setup can also be used to identify survival factors for e.g. neuronal cells in coculture, or differentiation inhibitors for co- cultured stem cells, or differentiation inducers for tissue (re) generation with cocultured pluripotent cells.

Stromal cells are seeded into individual wells of a multiwell cell culture dish. The cells are then transfected with a cDNA expression vector: cells in each well receive an individual cDNA or a combination of such cDNAs (pools), which stem from a prearrayed library. Genetically modified (see below) dedifferentiated cells, such as carcinoma cells or cells derived from metastasis are then overlaid and cocultured with the stroma cells for a defined period of time. The cDNAs introduced and expressed in the stroma cells will either or not provoke redifferentiation of the carcinoma cells. Inducers of redifferentiation may belong to a group of proteins, RNAs, fragments and/or derivatives thereof, which are secreted or on the surface of the stroma cells comprising hormones, growth-, otility- and inhibitory factors, extracellular signaling molecules, receptors, secreted enzymes, enzyme inhibitors and structural proteins and the like.

Alternatively, such factors may result from changes in stromal gene expression elicited by the introduced cDNA(s). Such proteins, RNAs, fragments and/cr derivatives thereof may comprise transcription factors, transcription cofactors, intracellular signaling proteins, endogenous enzymes and the like.

However, it goes without saying that these lists of proteins or molecules cannot be exhaustive. Redifferentiation is scored for by assessing the expression of a marker gene, which has been specifically constructed and introduced into the carcinoma cells to monitor differentiation. The whole procedure is performed in a high throughput format.

Variations and substitutions to the system apply to a series of parameters. The type and source of stroma cells is variable: the cells may derive from different cell types such as fibroblasts, myofibroblasts, adipocytes, smooth muscle cells, myoepithelial cells, endothelial cells; or different tissue sources like breast, prostate, bladder, small intestine, colon, pancreas, lung, esophagus, skin, kidney, stomach, testes, ovary, cervix, endometrium, liver tissue or different vertebrates such as human, mouse, rat, rabbit, guinea pig, baby hamster, hamster, cat and dog. The transfected cDNA library can also originate from different cell types, tissues and organisms (vertebrate and invertebrate) , as well as it can stem from different developmental stages of an organism or tissue, or may represent mixtures of the above. Also, the library may encode antisense or inhibitory RNA molecules to specifically eliminate endogenously expressed RNAs in the stromal cells. The introduction of the cDNA into the stromal cells can occur transiently by using standard procedures or by viral transduc- tion, or it can be done in a stable (genomic or episomal) way using procedures known to the skilled in the field.

The period of time for the coculture is variable according to the individual system set-up: scoring for early, intermediate or late effectors is feasible depending on the reporter system used. These effector proteins may be encoded by the transfected cUNA(s), or they may result from a change of the gene expression profile in the stromal cell provoked by the introduction of the cDNA(s) . Tumor cells from epithelial (carcinoma) or non-epithelial origin are genetically modified and used as reporter cells: the expression of a marker gene, quantitative variations of which can easily be measured, is put under control of tumor- or differentiation- specific regulatory DNA sequences, respectively. Alternatively, techniques for quantification of variations of endogenous marker molecules may be applied comprising histological, cytological or morphological staining procedures, quantitative RT-PCR, gene expression profiling with DNA chips or arrays. Thus, during redif- ferentiation, the marker gene expression level is accordingly increased or decreased, depending on the regulatory sequences chosen. These constructs are introduced into the tumor cell either by stable integration into the genome (random or site-specific), or on an episomal vector, methods known to the skilled in the field.

Example 1 — Construction of a carcinoma reporter cell line

In order to have a robust experimental readout system a marker carcinoma cell line is constructed. In general this means that the coding region for an easily measurable reporter enzyme is set under control of a differentiation-specific promoter region and inserted into a carcinoma cell.

1.1. Construction of a differentiation-specific reporter construct Firefly luciferase was chosen as a reporter enzyme. Its coding region is used as it is present on the commercial vector pGL3- Basic from Promega Corporation (Madison, WI, USA).

The minimal promoter from -178 to +92 of the human E-cadherin gene was inserted as a differentiation-specific control region. This region has been shown to be necessary and sufficient to confer epithelial-specific expression to a reporter gene (Hennig et al., J. Biol. Chem. 271:595-602 (1996)), signifying that it is a bona fide differentiation-specific regulatory region. The region was isolated by PCR with primers Xho-Ecad (5'- CCGCTCGAGACTCCAGGCTAGAGG-3' ) and H3-Ecadap (5'- CCCAAGCTTCGGGTGCGGTCGG-3' ) from human genomic DNA. The obtained DNA fragment was digested with restriction enzymes Xhol and 5 Hindlll and inserted into the respective sites of pGL3-Basic. The newly created plasmid was amplified using standard procedures. The integrity and sequence of the inserted E-cadherin region was verified and the plasmid was named pEcad-Luc+ (MOI ) .

10 1.2. Cell culture and transfection

The human breast cancer cell line MDA-MB-231 was chosen as a host carcinoma cell line. This tumorigenic line is from an adenocarcinoma of the mammary gland isolated by pleural effusion. The cell line was obtained from ATCC (Manassas, VA, USA) . MDA-MB-

15 231 cells are dedifferentiated carcinoma cells which do not express E-cadherin. These cells were cultivated using standard techniques in DMEM medium containing glutamax-1 and 10% foetal bovine serum of Australian origin (all media components from LifeTechnologies, Paisley, UK) .

20 To introduce pEcad-Luc+ (MOI) into the genome of MDA-MB-231 cells, the plasmid was digested with the restriction enzyme Pvul which cuts twice in the vector backbone. The crude digest was used directly for transfection of the cells using Lipofectamine 2000 (LifeTechnologies, Gaithersburg, MD, USA) . Specifically, 1 μg di-

25 gested pEcad-Luc+ (MOI) was combined with 0.33 μg of Pvul-digested pSVpuropA, an expression vector for the puromycin resistance gene, and 1.33 μg of calf thymus DNA (LifeTechnologies, Gaithersburg, MD, USA) and brought to a volume of 82.5 μl with Opti-MEM (LifeTechnologies, Gaithersburg, MD, USA) . This mix was then com-

30 bined with an equal volume of Opti-MFM containing 3.3μl of Lipofectamine 2000 and incubated for 20 minutes before dispensing the whole mix into a single cell culture well of a six well plate, containing about 1.5xl0⁵ cells in 2 ml medium (seeded 12-18 hours before) . Two hours later, the medium was replaced with fresh one.

1.3 Isolation of F2 , a cellular reporter clone Cells were resuspended two days after transfection and dispensed into two 10 cm plates and put under selctive pressure applying 2μg/ml puromycine. Medium was replaced every 4 days with fresh one containing puromycine. About 3 weeks after transfection, individual colonies were picked and transfered into 24 well plates and further expanded as a clone. Several of these clones were then characterized, as for instance clone F2.

The puromycine resistant clone F2 (full name: 231-F2, called F2 in the following) was first tested to express luciferase. This was the case with about 100 light units per aproximately 10⁵ cells, which is a low level of expression (range of cells expressing luciferase may vary from 1 to 10⁷ light units) as expected for the E- cadherin promoter in a carcinoma cell line. The presence of the transgene in the genome of F2 was further demonstrated by PCR.

1.4 Validation of F2 as a di ferentiation-specific reporter clone

F2 was then validated as a differentiation-specific reporter cell by applying the differentiation agents quinidine and SAHA. Both compounds stimulated luciferase expression from the integrated reporter construct (about 2 fold for quinidine, >100 fold for SAHA) . In parallel, the accumulation of fat droplets, a sign of differentiation for breast cancer cells, was observed in samples treated with these agents, but not in untreated cells. In conclusion, F2 is a bona fide reporter clone for induction of differentiation, as it contains the easily easureable E-cadherin- luciferase marker construct which is shown to be responsive to chemical differentiation agents . Example 2 - Stromal cell line cultivation

2.1 NIH 3T3 fibroblast cell cultivation

The mouse fibroblast cell line NIH 3T3 was obtained from ATCC . This cell line serves as a stromal feeder cell line, cultivated as a monolayer and transfected with the expression library. Mouse cells are chosen because they can be distinguished from the human carcinoma cells on both the gene and protein levels by virtue of a series of markers. NIH 3T3 cells are cultivated in standard DMEM medium containing glutamax-1 and 10% foetal calf serum (LifeTechnologies, Paisley,

UK) .

2.2 Seeding of NIH 3T3 cells for a HT screen NIH 3T3 cells are seeded into 96 well plates at a density of 10⁴ cells per well in a volume of 100 μl medium. This step is automated using a pipetting robot (Biomek 2000, Beck an Coulter Inc., Fullerton, CA, USA).

Example 3 - cDNA library transfection

3.1 cDNA library properties and preparation

A cDNA library is prepared from selected mRNA and cloned into an expression vector. In the present example, the plasmid library NFL021 is obtained from Invitrogen Inc. (Carlsbad, CA, USA) . This library was made from a mix of mRNA from human testes, lung, and brain tissue. The cDNA was selected to represent full length transcripts. Inserts are normalized to deplete abundant cDNAs and enrich rare cDNAs . The plasmid pCMV-Sport6 is used as a host expression vector. The library is transformed in E. coli bacteria and individual colonies are arrayed into 96 well plates. Each plate contains 91 library samples, 4 negative controls (empty cloning vector), and one expression vector for . luci erase as a downstream transfection control. The arrayed clones are grown in standard media for bacterial culture containing 100 μg/ml ampicilline. The arrayed library is then stored as glycerol stock at -70° C. These masterplates are further used to inoculate 96 well cultures for HT plarnid preparations using standard procedures (Magnesil protocol from Promega Inc., Madison, WI, USA) on a Biomek 2000 (Beckman Coulter Inc., Fullerton, CA, USA).

3.2 Transient transfection of cDNA library into NIH 3T3 cells NIH 3T3 cells are seeded into multiwell plates 12-18 hours prior to transfection. For 96 well plates, 10⁴ cells are seeded in 100 μl medium per well. The transfection mix is prepared as follows: 5 μl plasmid DNA (40 ng/μl) are combined with 5 μl Opti-MEM, before adding 10 μl of a freshly prepared stock solution of Opti-MEM containing 25 μl Lipofectamine 2000 per ml. After an incubation time of 20 minutes the whole mix (20 μl) is added to the cells of a single well. Two hours later medium is aspirated and replaced with 100 μl of fresh medium containg the reporter cell line at a density of 10⁵ cells per ml (see below) .

3.3 Transient transfection of cDNA library for a HT screen

The procedure of transient transfection of NIH 3T3 cells is automated in a 96 well format using a pipetting robot (Biomek 2000, Beckman Coulter Inc., Fullerton, CA, USA). The whole procedure is performed in parallel duplicates in order to eliminate false positives potentially occuring from assay variability .

Example 4 - Coculture of transfected stromal cells with the F2 reporter cells

4.1 Preparation of F2 reporter cells F2 reporter cells are routinely cultivated in puromycine- containing medium (2 μg/ml), however, for the coculture with non- resistant NIH 3T3 cells the antibiotic is removed. This is done by washing the resuspended F2 cells with phosphate buffered saline, pelleting them by centrifugation, and subsequent resuspension in fresh medium at the appropriate cell density.

4.2 Seeding of F2 cells onto transfected NIH 3T3 cells for a HT screen After removal of the medium containing the liposome-DNA complexes from the NIH 3T3 feeder cell layer, the F2 reporter line is seeded in an overlay of fresh medium devoid of puromycine. F2 cells are suspended at a density of 10⁵ cells per ml and 100 μl are used per well in 96 well culture dishes. F2 cells then sediment by gravity and grow on top of the feeder cell layer.

Seeding of the reporter line (F2) for the HT screen is also done with the pipetting robot.

Example 5 - Measurement of differentiation-specific increase of marker gene expression in reporter line F2

At a specific time after transfection - for this example 48 hours cells are lysed and extracts are assayed for luciferase expression using an in vitro procedure.

5.1 Lysis of cells in a HT screen

Medium is removed by aspiration from the cocultivated NIH 3T3 and F2 cells. 50 μl of lysis buffer (2 mM DTT, 2 mM CDTA, 10% glycerol, 0.5% Triton X-100, 25 mM Tris-phosphate, pH 7.8) are pipetted to each well of a 96 well plate. Lysis is performed by incubation for 20 minutes at room temperature.

5 μl of lysate are then transfered into a white 384 well plate for subsequent measurement of luciferase activity. All these steps are performed by using an automated pipetting robot .

5.2 Luciferase assay with cellular lysate for HT screen Luciferase assays are performed in a E'luoroscan Ascent FL luminometer from Labsystems Oy (Helsinki, Finland) by using an injector and a constant glow type assay. Specifically, 25 μl luciferase assay solution (20 mM tricine pH7.8, 0.1 mM EDTA, 1.07 mM (MgC03) ^Mg (OH)2- 5H20, 2.67 mM MgSO-}, 33.3 M DTT, 270 μM CoenzymeA, 470 μM luciferin, 530 μM ATP) are equilibrated to room temperature and injected to the 5 μl lysate in each well of a white 384 well plate. Injections are performed in batches of 24 at 1 sec intervals between the wells. Then light emission from each well is measured for 1 sec. This allows the luciferase, in each well, to reach its peak activity at 24 to 30 seconds after injection (half-life is about 7 minutes), and experimental conditions for well to well data are therefore kept virtually identical . Light emission data are then analyzed by comparison to the negative controls spotted into the arrayed library. Clonal induction (in both of the duplicate wells on the two plates) of luciferase activity above the statistically determined experimental variation is related to a potential induction of differentiation of the F2 reporter line. This is verified by repeating the experiment with the respective library plasmid in a secondary screen using a larger format and further morphologic, cytologic and genetic differentiation markers.

On the other hand, reduction of the basal luciferase signal may indicate cytostatic or cytotoxic effects exerted by the respective library plasmid. Again, these effects are more closely analysed by repeating the experiment and evaluating cell viability and proliferation markers. Equivalents

From the foregoing detailed description of the specific embodiments of the invention, it should be apparent that a unique coculture system has been described using gene library-trasfected feeder cells and dedifferentiated carcinoma reporter cells, which have been genetically modified to provide an easy reporter assay suitable for a high throughput screen. Although particular embodiments have been disclosed herein in detail, this has been done by way of examples with purpose of illustration only, and is not thought to be limiting with respect to the scope of the appended claims which follow. In particular, it is contemplated by the inventor that various substitutions, alterations, and modifications may be made to the invention without departing from the spirit and scope of the invention as defined by the claims. For instance, the choice of a particular differentiation-specific or tumor-specific gene regulatory (promoter) region, the choice of the carcinoma cell type or line, or the way of gene introduction into the feeder cells, as well as the choice of a particular readout sytem are believed to be a matter of routine for a person of ordinary skill in the art with knowledge of the embodiments described herein.

Claims

1. A method of coculturing cells for the identification of proteins, RNAs, fragments and derivatives thereof comprising the steps of: a) seeding stroma cells into individual receptacles; b) transfecting said stromal cells with a cDNA expression vector; c) overlaying said stromal cells with dedifferentiated cells; and d) coculturing the stromal cells with said dedifferentiated cells for a defined period of time.

2. The method according to claim 1, wherein said proteins, fragments and derivatives thereof to be identi- fied are encoded by the cDNA vectors of step b) .

3. The method according to claim 1 or 2, wherein the proteins, fragments and derivatives thereof to be identified are triggering redifferentiation of dedifferentiated cells.

4. The method according to claim 1 or 2, wherein the proteins, RNAs, fragments and derivatives thereof to be identified are survival factors for cells in a coculture, differentiation inhibitors for cocultured cells or differentiation inducers for tissue (re) generation with cocultured cells.

5. The method according to claims 1 to 4, wherein the proteins, RNAs, fragments and derivatives thereof to be identified are either expressed from the introduced cDNA or the result of changed gene expression as elicited by introduction of this cDNA(s).

6. The method according to any of the preceding claims, wherein the stroma cells are mammalian cells originating from a) varying cell types selected from the group consisting of fibroblasts, myofibroblasts, adipo- cytes, smooth muscle cells, myoepithelial cells, endo- thelial cells, b) tissues selected from the group consisting of breast, prostate, bladder, small intestine, colon, pancreas, lung, esophagus, skin, kidney, stomach, testes, ovary, cervix, endometrium, liver tissue and c) vertebrates selected from the group consisting of human, mouse, rat, rabbit, guinea pig, baby hamster, hamster, cat, or dog.

7. The method according to any of the preceding claims, wherein the individual receptacles are wells of a multiwell cell culture dish.

8. The method according to any of the preceding claims, wherein the expression vector contains cDNA derived from invertebrate or vertebrate tissue (s) or cells .

9. The method according to any of the preceding claims, wherein the cDNA is introduced into the stromal cell by way of transient transfection or viral transduc- tion, or in a stable genomic or episomal way.

10. The method according to any of the preceeding claims, v/herein the cDNA is derived from a cDNA library originating from different cell types, tissues and or- ganisms, different developmental stages of an organism or tissue or mixtures thereof.

11. The method according to any of the preceeding claims, wherein the library used codes for antisense or inhibitory RNA.

12. The method according to any of the preceding claims, wherein the dedifferentiated cell is a carcinoma or metastatic cell.

13. The method according to claim 12, wherein the carcinoma or metastasis-cell is genetically modified by introduction of a marker gene.

14. The method according to any of the preceding claims, wherein the period of time for coculturing is variable according to the scoring of early, intermediate or late effector proteins, RNAs, fragments or derivatives thereof.

15. Use of the method according to any of the preceding claims for high throughput screenings of proteins, RNAs, fragments and derivatives thereof, said proteins, RNAs, fragments and derivatives triggering redifferentiation of dedifferientiated cells.

16. A coculturing system comprising: a) stroma cells seeded into individual receptacles, wherein said stromal cells are transfected with a cDNA expression vector and b) dedifferentiated cells, overlaying said stromal cells, wherein said dedifferentiated cells are genetically modified by using marker gene promoters to score for ear]y, intermediate or late effector proteins, RNAs, fragments or derivatives thereof.

17. The coculturing system according to claim 16, wherein said dedifferentiated cells are genetically modified with a marker gene under control of differentiation-specific promoters.

18. The coculturing system according to claim 16, wherein said dedifferentiated cells are genetically modified with a marker gene under control of tumor- specific promoters.

19. The coculturing system according to claim 16 comprising: a) stroma cells seeded into individual recepta- cles, wherein said stromal cells are transfected with a cDNA expression vector and b) dedifferentiated cells, overlaying said stromal cells, wherein said dedifferentiated cells are scored for redifferentiation by analysis of endogenous markers.

20. The coculturing system according to claim 19, wherein changes in endogenous marker expression are scored for by a histological, cytological or morphological staining procedures.

21. The coculturing system according to claim 19, wherein changes in endogenous marker expression are scored for by quantitative RT-PCR.

22. The coculture system according to claim 19, wherein changes in endogenous marker expression are scored for by gene expression profiling with DNA chips or arrays.

23. The coculture system according to claims 19 to 22, wherein the time of coculture is variable according to the scoring of early, intermediate or late endogenous markers induced by the respective type of effector pro- teins or derivatives thereof.

24. Use of the coculture system according to claims 19 to 23 for high throughput screenings of proteins, RNAs, fragments and derivatives thereof, said proteins, RNAs, fragments and derivatives triggering redifferentiation of dedifferientiated cells.

25. Use of the coculture system according to claims 19 to 23 for high throughput screenings of proteins, RNAs, fragments and derivatives thereof, wherein said proteins, RNAs, fragments and derivatives are survival factors for cells in a coculture, differentiation inhibitors for cocultured cells or differentiation inducers for tissue (re) generation with cocultured cells .

26. Use of the coculture system according to claims 15, and 19 to 23 for high throughput screenings of proteins, RNAs, fragments and derivatives thereof, said proteins, RNAs, fragments and derivatives triggering cytotoxic and cytostatic effects on tumor cells.