KR20010040420A - Gene expression methods for screening compounds - Google Patents

Abstract

A method for screening test compounds in vitro for predicted in vivo activity has been disclosed.

Description

Gene expression methods for compound screening {GENE EXPRESSION METHODS FOR SCREENING COMPOUNDS}

<110> SHERING AKTIENGESELLSCHAFT

〈120〉 GENE EXPRESSION METHODS FOR SCREENING COMPOUNDS

<130> PL00214FR

〈150〉 PCT / US 09 / 013,496

<151> 1998-01-26

〈150〉 PCT / US 09 / 236,026

<151> 1999-01-22

〈160〉 2

〈170〉 KOPATIN 1.5

〈210〉 1

<211> 23

<212> DNA

〈213〉 Artificial Sequence

〈220〉

<223> DESCRIPTION OF ARTIFICIAL SEQUENCE: PCR PRIMER

<400> 1

gtcagtggrc agatgctrta ttt 23

〈210〉 2

<211> 27

<212> DNA

〈213〉 Artificial Sequence

〈220〉

<223> DESCRIPTION OF ARTIFICIAL SEQUENCE: PCR PRIMER

<400> 2

aacttcagac atcatttcyg gaaattc 27

One aspect of the invention is a method of grouping test compounds, the method of

(a) exposing a cell culture or a culture comprising at least two gene-cell combinations to a test compound to generate an exposed cell culture or culture,

(b) separating RNA from the exposed cell culture,

(c) screening RNA from (b) for each mRNA of the gene in the gene-cell combination of (a) to yield a gene expression fingerprint (GEF) for the test compound,

(d) repeat (a) to (c) to classify each test compound,

(e) comparing the GEFs for each test compound (d) and grouping the test compounds into at least two taxa groups based on differences in their GEFs.

Representative test compounds in each taxa are further tested for representative activity or activity of interest in vivo.

At least two gene-cell combinations include, for example, at least two different genes, at least two different cell types, or a combination thereof. In some embodiments, the gene or genes in the gene-cell combination include endogenous genes under the control of its own promoter and heterologous genes under the control of heterologous promoters, include internal negative regulatory genes, and are negative in response to the test compound The effect of the regulatory gene on the mRNA level is that the toxic effects of the test compound indicate a non-specific effect.

Screening of RNA involves RCR amplification using specific oligonucleotide primers of each gene. In some embodiments, the RNA is further reverse transcribed into cDNA. In some embodiments, the screening comprises hybridization of nucleic acid sequences specific for each gene to RNA or cDNA of the exposed cell culture. In another embodiment, mRNA levels of at least one gene in at least two gene-cell combinations are quantified.

In some embodiments of the invention a combination of two or more test compounds has been added to the cell culture to yield a GEF of that combination.

Another aspect of the invention is a method of identifying one or more genes for use in a gene-cell combination for grouping test compounds, the method comprising:

(a) exposing a host cell or at least one host cell culture in vivo to a first reference compound,

(b) isolating RNA from a host cell or host cell culture in vivo of (a),

(c) comparing the RNA of (b) with the host cell or the RNA of a host cell culture in vivo that has not been exposed to the first reference compound and wherein at least one gene having an affected mRNA level in response to the first reference compound The genes were identified as usable in gene-cell combinations for grouping test compounds. The RNA of (c) was compared with RNA from a host cell or regulatory host cell culture in vivo, and the host cell or regulatory host cell culture in vivo is exposed to a second reference compound, whereby Genes with mRNA levels affected in response to the reference compound were found to have a specific response to the first reference compound.

Another aspect of the invention is a method of grouping test compounds, the method of

(a) exposing a cell culture or a culture comprising at least two gene-cell combinations to a test compound to generate an exposed cell culture or culture, wherein at least one gene in at least two gene-cell combinations comprises a first and a second agent; To yield differentially expressed and exposed cell cultures at the 2 baseline state,

(b) separating RNA from the exposed cell culture or culture,

(c) screening RNA from (b) for mRNA levels of each gene in the gene-cell combination of (a) to yield a gene expression fingerprint (GEF) of the test compound,

(d) repeat (a) to (c) to classify each test compound,

(e) comparing the GEF of each test compound tested in (d),

Compounds are grouped into at least two taxa based on differences in their GEFs. In some embodiments at least one of the first and second reference conditions is a condition of a disease such as cancer.

In another aspect, the present invention provides a method for calculating a reference gene expression fingerprint (GEF) of at least a first test compound for use in grouping test compounds, the method comprising:

(a) identifying at least two gene-cell combinations, each of the at least two gene-cell combinations comprising a unique combination of a particular gene and a cell of a particular cell type, wherein the first gene-cell combination is

(i) exposing a host cell or host cell culture in vivo of a first cell type to a first reference compound,

(ii) isolating RNA from the exposed in vivo host cell or host cell culture of (i),

(iii) comparing the RNA prepared from a host cell or host cell culture in vivo of a first cell type not exposed to the first reference compound with the RNA of (ii) and responsive to the first reference compound The change in the mRNA level of a gene in identifies a cell of a first cell type with a first gene-cell combination for use in grouping genes and test compounds, and the second gene-cell combination is

(Iii) exposing a host cell or host cell culture in vivo of a first cell type or a second cell type to a first reference compound,

(Iii) separating RNA from the exposed in vivo host cell or host cell culture,

(Iii) compares RNA prepared from in vivo host cells or host cell cultures of cell types as in (v) not exposed to the first reference compound and (i) RNA and changes the mRNA in response to the first reference compound The gene with the level is identified as a gene that can be used in a second gene-cell combination for use in grouping test compounds, wherein the second gene-cell combination is different from the first gene-cell combination, and Contains cells of the same cell type as in (iii),

(b) Screening RNAs of (ii) and (iii) for mRNA of each gene in each of at least two gene-cell combinations to yield a reference GEF for the first reference compound for use in grouping test compounds. It involves doing.

In another aspect, the invention provides a method of grouping test compounds, which method comprises:

(a) calculating a reference GEF for the reference compound according to the methods described above and below;

(b) The method of calculating and grouping GEF for each test compound by

(i) exposing the cell culture or culture comprising at least two gene-cell combinations identified in claim 1 to a test compound to yield an exposed cell culture or culture,

(ii) separating RNA from the exposed cell culture or culture,

(iii) screen RNA of (ii) for mRNA of each gene of at least two gene-cell combinations of (i) to yield a GEF for the test compound,

(Iii) repeat (i) to (iii) to group each test compound for calculating the GEF of each test compound,

(c) comparing the GEF of each test compound produced in (b) with the reference GEF in (a), and the test compounds being grouped into at least two taxonomy based on the difference or similarity between their GEFs and the reference GEFs. Test compounds characterized by:

Differences in gene expression, such as normal cells, activation, disease, and tumor cells, help to understand the cellular processes that cause disease. For example, Zhang et al (Science 276: 1268-1272 (1997)) identified more than 500 transcripts expressed with significant differences in normal and tumor cells to disclose gene expression patterns in gastrointestinal tumors. Bernard et al (Nulc. Acid Res. 24: 1435-1442 (1996)) describe a method for analyzing the expression levels of 47 genes in epithelial and resting and activated T cells.

Microarrays of synthetic oligonucleotides or cDNAs are useful for investigating differences in gene expression. For example, Schena et al (Science 270: 467-470 (1995)) disclosed quantitative monitoring of gene expression patterns in response to transgenes using complementary DNA microarrays. Shena et al (Proc. Natl. Acad. Sci USA 93 (20): 10614-10619 (1996)) describe human cDNA of unknown sequence to quantitatively monitor differences in gene expression under given experimental conditions. A microassay was used. De Resi et al (Nat. Genet. 14 (4): 457-460 (1996)) used cDNA microarray analysis to analyze gene expression patterns in human cancer. Heller et al (Proc. Natl. Acad Sci. U.S.A. 94 (6): 2150-2155 (1997)) disclosed the use of cDNA microarray technology to monitor gene expression in inflammation.

Other methods for screening include detecting and isolating differentially expressed mRNAs using first oligonucleotide primers for reverse transcription of mRNAs and first and second oligonucleotide primers for amplifying synthesized cDNAs ( US 5,580,726). Rosenberg et al (PCT publication WO 95/21944) disclosed the use of expressed sequence tags (EST's) to detect differentially expressed genes in healthy versus diseased specimens. Lee et al (Cell Biology 92: 8303-9307 (1995)) described the use of a comparative EST assay to confirm differential expression of about 600 in RNAs of PC12 cells treated with nerve growth factors and untreated cells. .

Another screening method is described by Nilsson et al (PCT Publication WO 93/07290), which analyzes cellular responses to receptor-responsive substrates based on expression levels of proteins produced by genes regulated by hormone-receptor interactions. Examples include those that disclose methods of in vitro investigating the effects of antagonism or action of receptor-binding substrates on selected cell types, including intercellular hormone receptors. WO / 96/41013 discloses methods for identifying receptor antagonism or action using variants of intercellular receptors such as estrogen (ER), androgen (AR), progesterone (PR), and glucocorticoid (GR) receptors.

Knowledge of environmental agents that alter gene expression has led to the use of specific genes as biological markers of exposure to chemicals and other environmental factors (Links et al., Annu. Rev. Public Health 16: 83-103). (1995)). Such biological labels have been used to screen chemical and biological samples that can alter gene expression (Sewall et al., Clin. Chem. 41: 1829-1834 (1995)).

Thus, there is a need for a method for screening and characterizing differential gene expression in vitro and for screening compounds for their effect on gene expression in vitro. The present invention addresses more than these needs.

Figure 1 includes Figures 1A and 1B. 1A is a plot of GEF results of reference compound (Ref) and test compounds x, y, z in two assays. FIG. 1B shows the GEF results of the reference compound (Ref) and the seven test compounds in three assays. Each square represents the result of one analysis. In certain assays the activity of the compound is indicated by a filled rectangle. Inactive compounds are indicated by empty squares.

Figure 2 includes Figures 2A and 2B. 2A shows the GEF results of the reference compound (Ref) and six test compounds in five assays. FIG. 2B is a single association plot showing the inconsistency ratio between the reference compound and the six test compounds as a result of the GEF activity shown in FIG. 2A.

Figure 3 includes Figures 3A-3C. 3A shows the same GEFs of human breast cells from the stage of progression of normal and malignant tumors.

The same gene expression changes are shown in cell lines classified as either mildly or highly invasive. Values corresponding to mean fold-change relative to the MCF10A baseline state were observed for each gene from the data in Tables 7A-7B. The data shown as "normal" GEF has a change in gene expression observed in 76N MEC cell lines relative to MCF10A. The genes of the "tumor-related" expression change are represented by the left striped bar (the bar has striped angling from left to right) and the weakly invasive cancer-related genes are filled bars, the highly invasive cancer-related genes. Is indicated by the right stripe (the bar has stripe angling from left to right upwards). Dotted bars indicate that the direction or degree of expression of the gene is associated with weak or highly invasive cancer. The figure description to the right of the three graphs is a list of the depicted genes. Each number above represents a specific gene.

3B shows GEFs of two breast cell lines with unknown invasive activity. Changes in gene expression of mammary fibroadenoma cell line 006FA2B and mammary epithelial cell line HBL100 relative to MCF10A were identified using Atlas I cDNA conforming arrays. Data is shown for the 28 genes shown in FIG. 3A. The marking of a particular bar (left striped, right striped, dotted or filled) has the same meaning as in FIG. 3A. 3C shows GEFs for tumor biopsy specimens. Gene expression was monitored by analysis of tumor RNA using AtlasI conforming arrays. Changes in gene expression relative to normal breast tissue samples for the 28 genes listed in the drawing description of FIG. 3A are shown. The marking of a particular bar (left striped, right striped, dotted or filled) has the same meaning as in FIG. 3A.

4 shows changes in gene expression following treatment of MDA231 with various compounds. MDA231 cells were exposed to Taxol, Butyrate, Mevastatin, or Reference Carrier for 72 hours and analyzed for the gene expression effects described in M & M. The data is about the effect on mRNA levels induced by drug treatment relative to baseline conditions for genes with more than twofold changes in expression of at least one treatment condition.

I. Summary of the Invention

The present invention aims at a screening method of measuring changes induced by a compound in the expression of any gene in a cell, grouping the compound and classifying it into a compound having similar activity. The gene or cell used in the assay was identified by function, location, or other physiological variables related to the disease or symptom for the intended or discovered therapeutic agent.

Typically, a reference "gene expression fingerprint" (GEF) is calculated for a reference compound or "state." GEF is calculated for each test compound of interest as a result of the screening process of the present invention. Test compounds are grouped and classified based on comparison with reference GEF.

The basic screening procedure used herein to yield a reference GEF or to screen test compounds is a "gene-cell combination." As used herein, a gene-cell combination is used to refer to a particular gene in a particular host cell type. Other gene-cell combinations may arise from various combinations of specific genes, such as the same gene in two or more host cell types, two or more different genes in the same host cell type, and the like. In addition, a single host cell may comprise one or more such genes to yield two or more gene-cell combinations.

The host cell type used herein relates to cells of a particular source or to tissues of organs, differentiation status, adaptation to specific growth conditions, modification of clones, cell lines, transformation, transduction, viral infections, parasitic infections, bacterial infections, trans Including but not limited to transgenic hosts, species origins, and the like.

Thus, for example, in an embodiment a reference GEF exposes a cell culture or a culture comprising at least two gene-cell combinations, and observes a change in the mRNA level of the gene in the gene-cell combination that responds to the reference compound. Is calculated for. In a preferred embodiment of the invention, a single gene-cell combination is insufficient to yield a GEF. More typically, several gene-cell combinations (also defined herein as "analytes") were examined by comparison of reference conditions to respond to or yield a "reference GEF".

In another embodiment of the present invention, the relative mRNA levels of at least one gene were compared at at least two host cell sources, each host cell source comprising a different reference state to yield a reference GEF for the reference state. As discussed herein, genes are selected based on differentially expressed in the first and second baseline conditions. Typically, at least one of the reference states is a disease state.

In screening test compounds by the methods of the present invention, test compounds or peptides, peptide analogs (p53, not limited to estrogen, raloxifene, tamoxifen, or IFNβ analogs), polypeptides, proteins Agents or natural products (eg, microbial broth, plant or animal cell extracts), such as ribozymes, nucleic acids, oligonucleotides, or libraries of other organic or inorganic compounds, allow GEFs to culture cells or at least two gene-cell combinations in each compound. It is calculated through a screening process for each test compound by exposing a culture containing the same and observing any change in the mRNA level of the gene in the gene-cell combination responsive to the test compound. The results were used to compare the similarities and differences between the screened test compounds. Based on these similarities or differences, test compounds are grouped in the subsequent analysis. Such subsequent analysis is screening according to an in vivo test or other analysis.

In certain embodiments of the invention, the methods of the invention are useful for identifying compounds or agents, such as analogues of protein function (eg, p53-induced changes in gene expression) or unaffected GEF (eg, tumors). Cells versus "normal", plaques of atherosclerosis versus "normal" blood vessels, inflammatory tissues versus "normal" tissues). In such cases, the “reference GEF” is preferably derived from the differential gene expression patterns observed between different cellular states (eg, p53 positive versus negative; metastasis versus non-malignant tumors), and is itself treated with a reference compound. Is not essential.

II. Reference compound and state

As used herein, reference compounds include proteins, polypeptides, peptides, nucleic acids, peptide analogs, ribozymes, nucleic acids, oligonucleotides, or other organic or inorganic compounds, or natural products of microorganisms, plants, and animals. The reference compound is preferably selected to have in vivo activity but inhibits cell growth, stimulates receptors of interest, catalysts of interest, synthesis of compounds of interest, inhibition of replication of the virus of interest, stimulates cell growth, Inhibit cell infiltration, chemotactic response, anti-metastatic activity, anti-atherogenic atherosclerosis activity, anti-inflammatory activity, anti-cell death effect, prevention of atherosclerosis progression, reduced bone loss, reduced inflammation in rheumatoid arthritis, It is not limited to improved cognitive function or prevention of burning. However, GEFs calculated for reference compounds are not directly required for the determination of such activity. To some extent, GEF is only required to show the effect on the mRNA level of a reference compound in a given gene-cell combination. In addition, the genes analyzed for mRNA levels need not be directly or indirectly related to the desired in vivo activity. In the screening methods of the invention, test compounds were grouped through screening relative to the reference compound. The organization of such taxa can be screened for desirable in vivo activity, lack of side effects, or other enhanced features.

It is generally understood that one of the common techniques in the art is the selection of reference compounds based on the issues raised. In this way, generally, a class of compounds that includes reference agents, chemical compounds, proteins, peptides, oligonucleotides, etc., having known or predictable physiological effects on the pathological condition or desirable pharmaceutical properties, for carrying out the methods of the present invention. It is chosen as the basis for its decision.

In certain instances, the reference compound is tamoxifen, raloxifene, interferonα (IFNα), interferonβ (IFNβ), interferonγ (IFNγ), or anti-Ha-ras-ribozyme (Kijima et al, Pharmacol. Ther. 68: 247-267 (1995)), ligands of nuclear receptors, transcription factors such as steroid hormones, retinoids, receptors such as endothelin, endothelin, gastrin releasing peptides, neuregulin, PDGF, cytokines, chemokines, and insulin Ligands for transmembrane receptors, extracellular matrix components such as vitronectin, laminin and collagen, cell adhesion molecules such as N-CAM or I-CAM, enzyme inhibitors or active agents of interest such as L-NAME for oxidative nitric acid synthesis, cisplatin or taxol Including but not limited to the same chemotherapeutic agent.

The reference compound may also be the product of a gene expressed in the host cell. Such genes may be endogenous or heterologous under conditions such as endogenous or heterologous promoters. Example genes include, but are not limited to, transgenes, viral genes, antisense nucleic acids, ribozymes, and the like.

In some cases, the reference state will be used in place of or in combination with the reference compound for the determination of the reference GEF. Differences in mRNA levels between two or more cells or tissues exhibiting relevant physiological / pathological conditions form the basis of baseline GEF. Examples of baseline conditions include blood vessels of atherosclerosis of normal versus various lesions, normal versus malignant carcinoma, sarcoma, melanoma, or the stage of development of lymphoma, normal versus multiple sclerosis, Alzheimer's or Parkinson's. Including but not limited to stages of neurodegeneration associated with other forms of disease.

III. Gene-cell combination

A gene

The present invention utilizes changes in mRNA levels of one or more genes in at least two gene-cell combinations to yield a GEF for each test compound screened, and the mRNA levels of that gene respond to the reference compound. Test compounds bind directly or indirectly, for example, to promoters or other regulatory elements, bind to receptors to induce certain intracellular signals, alter mRNA stability, bind intracellular enzymes such as kinases or phosphatase, and Affects mRNA levels such as binding, changing the redox environment, or affecting intracellular and incoming ion flow. The gene is preferably an endogenous gene, which is preferably under the control of its own promoter. In some embodiments, the cell is infected with a virus, where the gene to respond is a viral gene. In some embodiments, marker genes, such as heterologous genes under the control of a heterologous promoter, are introduced into the cell by internal regulation to detect gene expression or the physiological state of the cell.

One or more sets of reactive genes for screening can be determined in several ways. For example, mRNA of a cell culture exposed to a reference compound can be compared to mRNA of a cell culture that is not regulated or exposed. In some embodiments, the organism or animal is exposed to the reference compound in vivo and such organisms, tissue samples, transplants, primary cultures, etc. are used as the source of mRNA. Changes in the level of specific mRNAs that occur in response to reference compounds can be achieved by subtractive hybridization using normal or abnormal libraries (eg, Gurskaya et al, Anal. Biochem. 240: 90-97 (1996), Bonaldo et al, Genome Res. 6: 791-806 (1996), the use of multiple arrays made of EST's or cDNAs (eg Bernard et al, Nucl. Acids Res. 24: 1435-1442 (1996); Schena et. al, Science 270: 467 (1995)), DD-PCR (Liang et al, Science 257: 967-971 (1992)), SAGE (Velculescu et al., Science 270: 484 (1995)), and the like. Without limitation, it can be confirmed by various means.

Although the response genes in the present invention are not required to have the desired in vivo effects of the reference compound, it is preferred to use the response genes of known nature and function. For example, known genes that respond to a reference compound constitute all or part of a set of response genes. Such genes are identified from cloning of cDNA from cell cultures exposed to literature, reference compounds or other sources. In this way, epidermal growth factor-regulatory genes such as junB, rhoB, EGF receptor, integrin β1 and viculin may be used to screen related compounds for selective EGF receptor action or antagonism. It can include all or part of a set. Genes encoding proteins such as p21, MDR1, hsp70, IGFBP-3, and bax all appear to be regulated by p53 through different mechanisms. These genes make up all or part of the gene set to screen related compounds for p53 analogs.

Preferably, the response genes selected for use in the screening assay should maintain at least a 2 to 5 fold change in the level of that mRNA in response to the reference compound. This change may be an increase or a decrease. Measurement of five or more fold responses provides for the detection of a "weak" activity test compound, which, for example, provides only a "partial" response (eg, a two-fold change in mRNA level compared to a five-fold "complete" response). .

In certain embodiments of the invention, the same set of response genes or portions thereof, or other sets, can be investigated in more than one cell type as part of the screening (ie, to yield another “gene-cell combination”). ).

Preferably 2 to 15 or more gene-cell combinations (or “analytes”) are used for the screening compound. The number of assays used to group and characterize compounds or reference conditions based on GEF can be reduced by further using reference compounds with known in vivo effects. GEF's may be interpreted as being the same as or not the reference compound or state. For example, when the added reference compound has an undesirable in vivo effect, an assay that cannot distinguish between the added reference compound and the first reference compound can be removed from the screening used to yield GEFs. Any of the gene-cell combinations can be internal regulation. For example, a "house-keeping" gene, such as GAPDH, actin, or cyclophilin, generally does not respond to the control compound and can therefore be provided with negative internal regulation. Positive internal regulation can include, for example, recombinant molecules under the control of a promoter that is expected or known to respond to a reference compound.

Further internal regulation may include genes that predict possible "toxic" effects of the reference or test compound. For example, such response genes include, but are not limited to, TNF or lymphotoxin, heat shock proteins such as hsp70, DNA damage inducing genes such as gadd153 or gadd45 and the like. An increase in the mRNA levels of one or more of these genes generally predicts the toxic effects of the control or test compound. Thus, for example, in embodiments, the screening of test compounds for reduced toxic effects is accomplished by examining reduced or unchanged levels of these internal regulatory genes.

B. cells

Cell lines and genes are generally selected from consensus as "informational" gene-cell combinations for the screening of test compounds. Practical considerations include the tissue of origin of the cell line, the level of differentiation of the cell line, the number of expression of the target gene, the efficiency of compounds such as cDNA, peptides, ribozymes, etc. that can be obtained in the cell line. In some embodiments, clinical samples such as tissue transplantation or primary cell culture, tissue transplants or blood samples from experimental animals, clinical samples such as tumor tissue, atherosclerosis are preferred. Thus, for example, although not required for the present application, it is useful for the screening of compounds aimed at developing new prostate tumor therapies using prostate cell lines.

Ⅳ. Screening method

In general, the test compound (preferably in the form of a library) is screened for a set of reactive genes and cells to identify compounds with the same or similar gene expression patterns. In one embodiment, for example, about 10 ⁵ to 10 ⁷ test compound libraries (e.g., peptides, oligonucleotides, ribozymes, peptide mimetics, polypeptides, proteins, nucleic acids, oligonucleotides or other maintenance or inorganic compounds, etc.), the screen of the do. For example, small molecular libraries are screened by exposing the cell culture to a general final level of 1-10 μM. The range of levels (eg, low, medium, high) of each test compound is preferred for the detection of weakly active compounds and for distinguishing compounds that exhibit different activity levels at a given level. For convenience, the cell culture treatment can be done in a 96 well microtiter dish. Exposure generally lasts 24 to 48 hours, but can be 30 minutes or even a week, especially in infected cells. Cells are generally treated in a humid environment containing 5-10% CO ₂ at 37 ° C. but changes to this condition can be indicated by a particular screen. RNA can be obtained from conventionally known techniques, preferably from poly dT capture plates (Mitsuhashi et al, Nature 357: 519-520 (1992)) or silica gel-based membrane adsorption purification (eg, Qiagen's RNeasy Total RNA) from an exposed culture. It can be obtained by the method of obtaining high materials such as, but not limited to, Extraction Kit. mRNA can further be reverse transcribed into cDNA. mRNA or cDNA can be used as a probe or target in a matching reaction and can be fixed or in solution. Known methods such as quantitative PCR are not limited to basic Northern or slot blot matching, nuclease protection, or the number of other RNAs that can be analyzed automatically as well as simultaneously, from mRNAs from 1 to 20 reactive gene sets. Can be quantified by Other preferred methods include membrane filters (eg, Bernard et al, Nucl., Using radioisotopes or fluorescently-labeled RNA or cDNA isolated from intracellular RNA isolated with matching probes for analysis including purified cDNAs). Acids.Res. 24: 1435-1442 (1996)) or glass plates (Schena et al, Science 270: 467-470 (1995)). A modification of this general method uses covalently attached chemically synthesized oligonucleotides to solid substrates instead of cDNA as targets of matching RNA or DNA (Lockhardt et al., Nature Biotech. 14: 1675-1680 (1996) ). Another method is to directly measure RNA or cDNA in accordance with differentially labeled gene-specific oligonucleotides (e.g., mass label, europium, which can be quantified by time-of-flight (TOF) mass spectrometry). ), Fluorescent enhancers such as terbium, samarium and dysprosium (Xu et al., Anal. Chem. Acta. 256: 9-16 (1992)).

GEF for each compound includes the results of the screening process. The compound may be removed from subsequent tests because of the likelihood of toxic effects on the cells, induced nonspecific responses, and the like. GEFs can be modified by testing with added reactive gene-cell combinations, using cells with the same set of reactive genes and cells, or different levels of metabolic compounds, removing reactive gene-cell combinations that do not provide information from the GEF. have.

Ⅵ. Test compound by group

Test compounds screened as described above are classified according to their GEFs. For example, test compounds that induced changes in mRNA levels of all components of a set of reactive gene-cell combinations are grouped separately from test compounds that caused changes in only one example, two examples, and the like. The greater the number of assays used for screening, the more can be grouped.

In this way, for example, the reference compound is defined as “activity” in all GEF assays, and the activity may be increased or decreased relative to the control at the mRNA level of a particular gene by following compound treatment. Compound x or y is found or identified as having activity in at least one GEF assay. Compounds x and y can be separated from the reference compound based on inactivity in at least one assay.

Compounds are classified from one another if they are active in the same assay. In a simple example using two assays (FIG. 1A), four possible classes of compounds may be defined. The number of possible classifications is X ⁿ , X is the number of active states measured (eg, +,-) and n means the number of assays. In this example, the possibility of X ⁿ = 2 ² or 4 is indicated by the reference and the compounds x, y, z. Each compound is distinguished from others by different GEFs. The taxonomy is further subdivided to account for quantitative differences in response to other compounds as the criteria for classification.

By increasing the number of GEF assays investigated, more taxonomies of compounds can be defined. Compounds active in the same assay can be grouped together. For example, in FIG. 1B, seven compounds (x, y, z, a, b, c, d) in the possibility of X ⁿ = 2 ³ or 8 are representative of different categories. In the case of three or more assays (eg, FIGS. 1B, 2A, and 2B), clustering algorithms can be used to determine the similarity of the compounds to the reference compound and each other, respectively. Initially, compound classification systems can be determined by their association distance, which measures the rate of inconsistency with the reference state. The association distance between the compound and the reference state is close when the compound exhibits a high rate of activity consistent with the reference state. The compound shown in FIG. 2A by a simple clustering algorithm based on similarity with the reference state is characterized by the association plot of FIG. 2B. In this analysis, compound z is closest to the reference state (ie the association distance 0.4), and compounds a and x are the same distance. The compound is classified as z by changing the criteria for an associative distance classification of 0.6. In this way the justification of the classification can be adjusted by varying this association distance. Using a smaller association distance as a criterion for classification results in a more generalized classification than is achieved using a larger association distance. Algorithms added depending on the data set can be used to cluster compounds based on their similarities (James, M., Classification Algorithms (1st ed.) New York, NY, John Wiley & Sons (1985)).

Compounds that have activity in only one assay (or less than 20% of eight or more of those assays) are not classified or further investigated if they are not active in the assay that forms the basis for the majority of the active compounds identified. For example, compounds y, b in FIG. 2A are potentially relevant for the following investigation because they are active in assays identifying compounds x, a, and z. Compound c is not examined further.

The decision to increase the justification of the classification can be influenced by the patterns of observed gene expression as well as data from other assays. For example, if the investigation of compounds x, z, a in FIG. 2A is active in cell-based assays where only x and z are important, compounds such as b and y that are active in assays common to x and z are further investigated, alone and in combination. do.

Iii. In-depth investigation of the test compound

After grouping test compounds on the basis of GEF, representative compounds are further characterized in conventional known cell-based assays for meaningful properties. Such assays include, for example, cell growth, anti-viral activity, gel-phoretic migration displacement analysis from extracts of treated cells to DNA-protein complexes, cell invasion through extracellular matrix or reconstituted substrate membranes, anchorage-independent growth, Analysis of chemotaxis, cell death, differentiation, cell adhesion to various substrates, intercellular interactions, secretion, proteolytic activity, osteoclast bone resorption, and the like.

It is beneficial as an example for extending cell-based assays to useful animal models. Some examples of known animal models include uterotropic effects (eg, uterine hypertrophy; Allen-Doisey), fever (eg, rabbit fever), osteoporosis (eg, ovarian ablation or transgenic / knock-out animals). Bone density of the cortex and soju of rats), atherosclerosis (e.g. lipid adhesion in blood vessels of fat-fed rabbits or transgenic / knock-out animals), restenosis (e.g. neovascularization following carotid artery injury) Thickening of the intima), cancer (e.g., tumor induction in rats or mice, tumor xenograft growth in nude mice, dementia or transgenic / knock-out animals), metastasis (e.g. injecting tumor cells into the tail arteries to the lungs) Colony formation), rheumatoid arthritis (eg, swelling of co-induced junctions), multiple sclerosis (eg, EAE models of marmosets or mice, transgenic / knock-out animals), Alzheimer's disease (eg, Tran Including animal models of animals out) transgenic / knockout.

In some embodiments, the GEF's of two or more test compounds are complementary to each other, that is, when GEF's are added, they can infer the GEF of the reference compound or the preferred aspect of the GEF of the reference compound. In such embodiments, two or more test compounds may be used together in combination to determine which combination has the desired biological activity in cell-based or in vivo assays.

The following examples are included with the intention of the above, but do not limit the invention.

Example

I. Selective estrogen compound discovery

A. Background

Epidemiological and experimental data support the protective role of estrogen in reducing the frequency and severity of coronary artery disease, Alzheimer's disease and osteoporosis. Estrogen therapy, however, leads to unexpected effects such as endometrial hyperplasia in women and reduced testosterone levels in men. Therefore, the objectives of the studies described herein are to provide a selective in vivo protective effect on bone (eg, reducing bone loss), the function of neurons (eg, anti-Alzheimer's disease), and the vascular system (eg, anti-atherosclerosis). To determine what the in vitro profile for a compound with is. Such selective compounds are preferably devoid of undesirable adverse effects (eg, uterine directional effects in women, testosterone-lowering in men and reduced sex organ weights).

The research strategy we pursue relies on three basic assumptions: this "estrogenic" biological effect is at least partially regulated by the ligand-induced transcription factor estrogen receptor (ER) (Mangelsdorf et al., Cell 83: 835-839 (1995)), regulation of gene expression by estrogens is limited by the number of different mechanical processes that can be further modified in tissue-specific methods, and compounds with selective in vivo effects are distinguished. Will induce a gene expression pattern.

Useful methods for identifying ER ligands with potential as selective agents in vivo are standard ER ligand binding and cell-based estrogen (E) -dependent proliferation assays or ER-mediated interaction assays (eg, Tzukeman et. al., Mol. Endo. 8: 21-30 (1994)), which uses other E-response promoters to characterize the compound. Screening for other ligands in their ability to alter the ER structure makes it possible to use cleavage analysis of proteolysis (Beekman et al., Mol. Endo. 7: 1266-1274 (1993)). Careful use of this assay may enable the separation of E agonists from partial agonists and antagonists. However, since compounds with distinctly distinct in vivo effects are not distinguished by the assay, this method does not provide sufficient information about the compounds that allow the prediction of selectivity in vivo.

Methods for classifying compounds based on differential gene expression control have been developed here to identify such selective compounds. All 49 compounds were examined by this method and classified according to GEFs. The prediction of in vitro “fingerprints” for the in vivo activity of any of the finally sorted compounds is investigated.

B. Specific Methods

1. Genes and Cells

Known E-responsive genes are literature studies (52 kD of cathepsin D, growth hormone, prolactin, progesterone receptor, pS2, TGFα, IGFBP-1, CBG, amphiregulin, TRHR (thyroid secretion hormone receptor)) and similar The cDNA (or fragment thereof) was cloned and the probe fragment was prepared for northern or slot blot conformance studies by known techniques.

Mammalian cell lines comprising endogenous ER have been described in (GH3 pituitary adenoma, BG-1 ovarian carcinoma, MCF7 breast carcinoma, ZR75-1 breast carcinoma, MDA361 breast carcinoma, Ishikawa human endometrial carcinoma (Nishida et al., Acta Obstet. Jpn. 37: 1103-1111 (1985) and / or by analysis of ER expression (eg, protein by Western blot analysis; RNA by RT-PCR).

In addition, infected cells stably expressing ER were also examined (MDA231-ER-breast carcinoma (Zajchowski et al., Cancer. Res. 53: 5004-5011 (1993)), 185B5-ER-human breast epithelial cell line ( Zajchowski et al., Mol. Endocrin. 5: 1613-1623 (1991)), HepG2-ER-human hepatocellular carcinoma, and Fe33-rat liver cancer (Kaling et al., Mol. Cell. Endo. 69: 167-178 ( 1990).

The first step was to determine which genes and cell lines actually showed measurable responses to E processing. Lastly, the ER-positive cells were cultured without estrogen and natural hormone, 17β-estradiol (E2), or 17α-ethynyl-estradiol (EE; non-metabolic estrogen), respectively, for 3 hours (short term), The cells are grown by treatment for 24 hours (medium) and 72 hours (long term), and RNA is prepared from the cells at each time point. Analysis of the mRNA levels of the genes of interest gives kinetic investigations in response to EE treatment and indications of appropriate conditions for measuring the reactivity of each gene.

2. grouping of active and nonspecific compounds according to GEF; Screening of "informational" analysis

At this stage, anything identified as an E-responsive gene-cell combination can be used in a screen of multiple compounds. However, for the validation experiments of this concept, we decided to simplify the GEF screen by asking if a portion of the gene-cell combination is sufficient to identify other compounds of known pharmacy. To this end, we chose seven additional compounds to test their gene-cell combinations that responded at least three times to mRNA levels in E2 or EE treatment. Seven other compounds are selected based on known properties in in vitro and in vivo assays. Important compounds are tamoxifen (4-OH-tamoxifen (HT) derivatives are used in the first study) and raloxifene (Ral) as implemented at the time of this study, although they will be clearly distinguishable in in vivo reactions. Rado (e.g., even though they have similar anti-estrogenic effects on mammalian glands, tamoxifen has more pronounced uterine tropism than raloxifene (Sate et al., FASEB J. 10: 905-912 (1996)), No distinct method has been reported in vitro, and since we want to find similar compounds as well as others, this compound becomes the reference compound to be added in the assay. Estaladiol (ie 2-OH-17β-estradiol (2HE)), other reported partial work, for this initial study to determine which of the assays can be distinguished using Dragon-antagonist (i.e. RU39411 (RU): Gottardis et al., Cancer Res. 49: 4090-4093 (1989); 119010 (119): Nishino et al., J. Endocrinol. 130: 409-414 ( Centchroman (Cen): Hall, BBRC 216: 662-668 (1995)), and pure antagonists (i.e. ICI164384: Wakeling et al., J. Endocrinol. 112: R7-R10 (1987)); Select structurally related compounds.

The level of all compounds is examined at 1.0 μM since compounds at 1.0 μM concentration induce optimal response in most of the assays. In some cases, it is examined at the 10 μM level. The efficacy of a compound to change the state of a constant level of mRNA corresponding to each gene is quantified by Northern, slot blot, or RT-PCR analysis as described herein (Table 1). The mean fold-up of mRNA levels induced by E2 or EE for each gene / cell analysis is shown in the third column. In certain assays, the compounds that elicited the reaction are marked with (+) and those with no effect are marked with (-). Analysis of 27 different gene-cell combinations with these nine compounds (to yield the GEF of each compound) provided much of the information (the series of compounds in Table 1 shows the same pattern of activity). However, five distinct activity patterns across this set of compounds were indicated by the Roman numerals I to V on the right side of Table 1, showing in all these gene / cell combinations. Pattern I was observed in most of the cell types tested, but other patterns (particularly pattern II) were biased towards the cell type. Such data highlight the importance of using different cells as well as other genes in performing this analysis. Also clear from Table 1 is that compounds may have different activity in activating the same gene (eg, 52 kD) depending on the cell (eg, comparing ZR75-1 to BG-1).

Summary of maximal response of each gene / cell combination (ie, analyte) to compound treatment. +, Active compound; -Inert compounds; nd, not confirmed. The listed cell lines were treated with the indicated compounds, the total RNA was isolated and the expression changes of the listed genes as described in MM were analyzed. The maximum mean gene expression response of each cell line following E2 or EE treatment is provided in the third column (ie fold effect + E). Each assay can be grouped into the taxonomy shown on the right (ie, I-V) depending on the response to the compound.

In addition, the same compound may have different activity for different genes in the same cell (eg, compare PR to pS2 or TGF-α in MDA-ER cells).

In this way, for this set of compounds, five “informational” analyzes were identified during the 27 analyzes analyzed, such as those that could differentiate and combine the compounds into five taxa. The five analysis types (patterns) do not appear identical. A prominent assay type is the response to estradiol derivatives (ie EE and 2HE), whereas the least frequently identified modalities (corresponding to the assays representing Ral and 119) were observed only four-fold. In this way, among the estrogen response assays used herein, some of them randomly selected consist of at least 15 and preferably about 20 for use in GEF screens. The statistical probability of identifying raloxifene as the active compound in such a screen is 96% if 20 assays are used, 91% if 15 assays are used, and 80% if only 10 assays are used (Snedecor et al. , Statistical Methods, 8th ed.Iowa State University Press, Ames, Iowa, Chapter 7, (1989).

To simplify the GEF screens, additional studies were conducted to determine which remaining assays were most suitable for the screening strategy (eg, high productivity and degree of change for the control). IGFBP-1 / Fe33 gene-cell combination (pattern V) was not used in previous studies (because it is difficult to interpret data in carcinoma-induced cells of interest of drug-metabolic activity). Representative assays selected for substudy are shown in Table 2. Data was identified by specific GEFs based on the activity each compound was derived in each of the four assays (marked with + and − in the column below each compound). In this method, compounds having the same GEFs are grouped together to distinguish them from having other GEFs. For example, E2, EE and 2HE are represented by one group (# 1 in Table 2) and HT and RU are represented by another group (# 2). The most important is the observed difference between E2, Ral and HT, indicating that this assay is successful in distinguishing compounds of distinct in vivo drugs.

3. Classification of Additional Compounds Using Selected Gene-Cell Analysis

This method of classification is used to isolate the 30 additional compounds, most of which relate to the nine compounds that were first tested structurally. Compound E1 (estrone). E3 (estriol), DHE (17α-dihydroequilene), DHEN (17α-dihydroequilenin), ZK182491 and ZK155843 are derivatives of 17α-estradiol (17α-E2) or 17β-estradiol. Compounds ZK166780, ZK166781, ZK167466, ZK167957 and ZK180686 are 11β-substituted 17β-estradiol derivatives related to RU39411. Compounds HT, ZK186275, ZK183819, ZK182956, and ZK183955 are tamoxifen derivatives. Compounds ZK185157 and ICI182780 relate to the pure steroid antagonist, ICI164384. Compounds ZK182254, ZK186217, and raloxifene are benzothiophenes. Compounds ZK183659, ZK22496 and ZK185704 are structurally related (ie, include cyclophenyl moieties). Compound ZK167502 is a naphthalene derivative and coumestrol is phytoestrogen (Price et al., Food Addit. Contam. 2: 73-106 (1985)). Most of these were previously classified as agents, partial agonists, or antagonists of ER through ER binding and transcriptional activity analysis. In this experiment, compounds were shown using three levels of activity (e.g., inactive, partial activity <50% of E2 reaction, complete activity> 50% of E2 reaction). It is clear from Table 3 that the compounds are classified into 10 groups by this analysis (Table 3). This classification of compounds is not based primarily on the chemical structure indicated as a result with compounds related to RU39411 (ie, ZK166780, ZK166781, ZK167466, ZK167957 and ZK180686). These six compounds are divided into three different taxa according to their GEFs.

Data represent the mean maximum response (at 10 μM level) of at least three individual experiments in duplicate measurements. Active ++,> 50% E2: +, <50%:-, inactive.

4. Determination of predictive capacity of GEF classification for in vivo effects

Included in the tested compounds is the standard (ie E2, tamoxifen, raloxifene, ICI 164384) in the reported distinguishable in vivo profile. Except for ICI, E2, tamoxifen and raloxifene are effective in reducing the "estrogenic" effect on the bone and cardiovascular system (ie, reducing atherosclerosis) in experimental and / or clinical studies Forming tamoxifen: Williams et al., Arterioscler.Thromb. Vasc. Biol. 17: 403-408 (1997); Raloxifene: Bjarnason et al., Circulation 96: 1964-1969 (1997) and / or protection against hysterectomy-induced bone loss (tamoxifen Love et al., N. Engl. J. Med. 326: 852-856 (1992); Raloxifene: Black et al., J. Clin. Invest. 93: 63-69 (1994) .E2 and Tapoxyphene is readily distinguishable from raloxifene because it has a great effect of inducing uterine directional effects (Sato et al., FASEB J. 10: 905-912 (1996) and Table 4), which is a raloxifene This means that we have tissue-selective activity in vivo. In our analysis of gene expression patterns, we see that four compounds have different GEFs and are separated into groups. Made (table 2 and 3). This data is found not to be selective in vivo efficiency of the compound is distinguished by different gene expression profiles (GEFs) in vitro.

Of particular significance is the group of compounds comprising ZK167466 (Group 9, Table 3). Like the raloxifene group (group 8), this compound shows activity in only one GEF assay. To determine which co-classifications of these compounds have similar in vivo drug properties, they were tested in vivo for uterine directional activity as well as the ability to reduce the loss of bone mass caused by reduced levels of circulating estrogen. (Ie experimentally induced by hysterectomy). Table 4 compares the activity of this group of compounds on E2, Tam, Ral and ICI. Four of the Group 9 compounds differ from others in the same assays. They do not show any effect on the thickening of the endometrium (ie, uterine directional effects) or exhibit mildly irritating effects (indicated by-or-/ + in Table 4). Three of them are significantly efficient in "bone protection" assays that predict efficacy on osteoporosis (Table 4). These data predict that the GEF profile is a new selective compound taxon (i.e., one that has a bone-protective effect and few uterine-orientated responses), which means that in vitro screening methods (isolated from other "partial agents") Can not confirm.

Data from in vitro assays (Table 3) are compared with the activity of the same compounds in in vivo studies. The mean maximum response relative to the reference group is shown for each compound. Uterine response is a change in epithelial cell length of the endometrium by compound treatment. Bone protection is measured from% loss in bone mineral density (quantified by pQCT) of baseline and compound-treated animals. ++,> 70% E2; +, 35-70% E 2; -/ +, 10-35% E2; -, <10% E2; ND, not measured.

C. Materials and Methods

1. Cell Culture and Compound Treatment

MDA-231 ER infected E-28 cells are often 1 milliMolar (mM) HEPES, 2 mM glutamine, 0.1 mM MEM non-essential amino acid, 1.0 mM pyruvate, 50 μg / ml gentamicine (all from GIBCO) , Phenol red-free α-modified minimal essential medium (MEM Gibco BRL; Gaithersburg) supplemented with 1.0 μg / ml insulin (Sigma; St. Louis, MO), and 5% DCC-treated FBS (Intergen) , MD). Cells are plated with approximately 40% identity (1.5 × 10 ⁶ / plate) in a 150 mm culture dish. After attaching the cells overnight, the medium will contain 0.2% ethanol or test compound and further incubate for 48 hours.

GH3 murine pituitary cells were treated with 12.5% horse serum, 2.5% FBS, 25 mM Hepes, 2 mM L-glutamine and 50 μg / ml gentamicin sulfate M-F10 (1: 1) at 37 ° C., 5% CO ₂ . Often cultured in media. Under this condition, the cells are partially adhered and adhered to both sides, and non-adherent cells are maintained during the passage of the cells. For measurement of mRNA expressing cells are seeded in culture medium containing phenol red and DCC--treated serum (10 ^6/100 mm dish). After 3 days, the medium will contain 0.2% ethanol or test compound and the cells are further incubated for 2 days.

BG-1 human ovarian carcinoma cells (Geisinger et al., Cancer 63: 280-288 (1989)) were treated with DMEM-F12 (1: 1) containing 10% FBS, 2 mM L-glutamine and 50 μg / ml gentamicinsulfate. 1) Often cultured in medium. In order to measure the mRNA expression level of 2 × 10 ^6/150 mm before plated in the same medium at a density of plate phenol red containing 5% DCC- FBS-treated - is incubated for 24 hours in the pre-culture medium. The next day, the medium will contain 0.2% ethanol or test compound and incubate for an additional 72 hours.

ZR75-1, MCF7 and MDA361 human breast carcinoma cell lines are often 1 mM HEPES, 2 mM glutamine, 0.1 mM MEM non-essential amino acids, 1.0 mM pyruvate, 50 μg / ml gentamicin, 1.0 μg / ml insulin and 10% Cultured in α-modified MEM supplemented with FBS. Cells were plated in phenol red and insulin-free medium containing 5% FBS-DCC for analysis (ZR75-1: 1.5 × 10 ⁶ / p100; MCF7: 2 × 10 ⁶ / p150; MDA361: 5 × 10 ⁶ / p100). After attaching the cells overnight, the medium will contain 0.2% ethanol or test compound and incubate for an additional 24 hours (ZR75-1), 48 hours (MDA361), or 72 hours (MCF7).

HepG2 human liver carcinoma cells stably infected with ER (clones ER1 and ER2) contained 1 mM HEPES, 2 mM glutamine, 0.1 mM MEM non-essential amino acids, 1.0 mM pyruvate, 50 μg / ml gentamicin and 10% FBS. Cultured in supplemented EMEM (GIBCO). Carcinoma cells of Ishikawa human endometrium were cultured in EMEM supplemented with 2 mM glutamine, 50 μg / ml gentamycin and 10% FBS. Fe33 (ER-infected FTO-2B rat liver cells) is maintained in DMEM-Ham's F12 (1: 1) without phenol red containing 10% DCC-FBS on 0.1% gelatin coated Petri dishes. All cells were plated in insulin-free medium containing phenol red and 5% FBS-DCC for analysis (HepG-ER: 4 × 10 ⁶ / p100; Ishikawa: 2 × 10 ⁶ / p150; Fe33: 2.5 × 10 ⁵ / p150). After attaching the cells overnight, the medium will contain 0.2% ethanol or test compound and incubate for an additional 72 hours.

ER-infected human breast epithelial cells (B5-ER) are maintained and analyzed for changes in gene expression according to the previously described method (Zajchowski et al., Mol. Endocrinol. 5: 1613-1623 (1991)). Compound or carrier treatment is for 72 hours.

17β-estradiol, 17α-ethynylestradiol, estrone, estriol, progesterone, dexatamentason and phenol red were purchased from Sigma Biochemicals (St. Louis, Mo.). All other compounds were synthesized at Schering AG (Berlin). Stock solutions of all chemicals (10 mM) were stored in DMSO and diluted in ethanol for analysis.

2. RNA Isolation and Slot Blot Analysis

At the end of the compound treatment time, cell monolayers are collected with Ultraspec (Biotecx Laboratories, Houston, TX) or RNeasy (Qiagen Inc., Santa Clara, Calif.) RNA separation reagents and proceed according to the manufacturer's suggested method. Total RNA (MDA-231 ER: 10 μg; GH3: 1.0 μg) was imprinted on Zetaprobe-GT nylon membrane using a 48-well slot blot device attached to a vacuum branch. Total RNA (20 μg) from treated and untreated samples of all other cell lines was determined by Northern blot analysis. Matching the membrane labeled probe with ³² P-dCTP has been described and performed previously. Quantification of specific matches at each point was performed using Fuji phosporimager by eliminating the non-specific background detected in the negative control for each mRNA, and the ratio of signal intensity in the compound-treated samples relative to the control was determined for each specific assay. Values for fold-changes used in the investigation of compound activity. Changes in mRNA levels above 2 fold are marked positive.

3. Progesterone Receptor Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)

All RNA samples are diluted to 20 ng / μl in DEPC-treated water. RT-PCR is performed using 100 ng total RNA. The reaction mixture is 5 units of rTth DNA polymerase (Perkin Elmer; Foster City, CA), 1 × EZ buffer (Perkin Elmer; Foster City, CA), 2.5 mM Mn (OAc) ₂ , 300 μM dNTP's in a final volume of 50 μl. (mix from Pharmacia; Alameda, Calif.) and 10 pmol of each biotin treated primer. PCR primers were synthesized by Synthetic Genetics (San Diego, CA) with PR # 1 (5 'GTC AGT GGR CAG ATG CTR TAT TT), PR # 2 (5'11C TTC AGA CAT CAT TTC YGG AAA TTC). Amplification is carried out on a Perkin Elmer 9600 with 33 cycles of 30 minutes of RT at 60 ° C. followed by two PCR reactions (15 seconds at 95 ° C., 45 seconds at 60 ° C.) and a final 7 minutes extension at 60 ° C. The following PCR is quantitatively analyzed by removing the 1/20 reaction volume and using an oligonucleotide probe specific for streptavidin-coated 96-well microplates and PCR targets, the probe being linked to HRP or AP, The addition of a color system (HRP) or chemiluminescent (AP) allows for an initial replication of 300-500 of the specific RNA template in 20-100 ng of total RNA samples. Standard curves for analysis (calculated by nonlinear regression analysis using four ao variables with s-shaped lines) were used, and changes in mRNA levels were labeled positive if> 3 fold.

4. Analysis of Histomorphometry

To measure cervical activity, immature female Sprague-Dawley rats, 35-50 g, 19-21 days old, were administered subcutaneous injections daily for 3 days with compound or vehicle alone. The compound is dissolved in a carrier consisting of 10% ethanol in arachis oil or benzylbenzoate / castor oil (1: 4). On the fourth day, the animals are weighed and suffocated with carbon dioxide and euthanized. The uterus is excised and placed in neutral 3.7% formaldehyde buffer for at least 24 hours. The uterus is then pinched into blue pins and cut across 4 [mu] m and stained with hematoxyln and eosin, and the fragments are obtained from Branham et al. (Branham et al., Biol. Reprod. 53: 863-872). Luminescent epithelial cell height as described by &lt; RTI ID = 0.0 &gt; The difference in epithelial cell height of the estrogen (0.3 μg 17β-estradiol / animal) and the carrier-treated group is calculated and expressed in 100%.

The activity of the compound of interest in the ratio of 17β-estradiol was calculated according to the following formula.

5. Determination of bone mineral density

Three months old female rats (Sprague Dawley) were treated immediately after hysterectomy to determine the efficacy of preventing bone loss. Compounds are treated with benzylbenzoate / caster oil (1: 4) or arachis oil / ethanol (95: 5) once daily. The control group (sharm / ovx treated with carrier) and the treatment group consisted of six animals each. After the animals were sacrificed 4 weeks later, the bones of the left and right tibia were treated for inorganic density measurements. Bone inorganic density (BMD) is measured in the second spongy of the adjacent tibia by means of peripheral quantitive computed tomograohy (pQCT). The results are expressed in percentage of protection from bone loss. Bone protection is expressed in comparison to the effect of estrogen (0.3 μg 17β-estradiol / kg) according to the following formula.

II. Screening for Interferon-β (IFNβ) Similars

A. Background

INFβ has efficacy in the treatment of multiple sclerosis (MS) (The INFβ Multiple Sclerosis Study Group Neurology 43: 655-661 (1993)). The precise mechanism by which INFβ induces its therapeutic efficacy is unknown. However, INFβ interacts with receptors directly as ligands to induce phosphorylation of numerous induced signaling proteins (STATs (Ihle, Nature 377: 591-594 (1995)), eventually leading to specific changes in gene expression directly. (Darnell et al., Science 264: 1415-1421 (1994)), there are many reports of signal transduction processes regulated by IFNβ INFα that can bind to the same receptor protein with similar members of the same cytokine class. Is not yet used in the treatment of MS because of unacceptable adverse effects Other interferons, IFNγ, share some of the effects on gene expression of INFβ but exacerbate signs of MS (Panitch et al., J. Neuroimmunol. 46 (155-164 (1993)). The differences in the biological effects of these three ligands therefore lead to the development of screens for finding selective INFβ analogs that are more efficient and better than INFβ itself. It can be used.

Animal models are available to test drug efficacy in improving the severity of the disease (ie, experimental autoimmune encephalitis (EAE) or T cell metastasis EAE models).

B. Cell Screening and Gene Identification

The cells used in this study are known or suspected INFβ-responsive tissues (eg B cells (eg Daudi), T cells (eg Jurkat), glioblastomas (eg T98G), carcinomas and astrocytes (eg CH235) RNA is discriminated by matching probes prepared from relevant cell lines treated with INFβ and prepared from RNA on microarrays containing more than 100 pre-selected cDNAs, such as Atlas cDNA arrays (Clontech). The cell lines representing the highest number of differentially expressed sequences are selected for studies to identify INFβ-responsive genes Technically, this may be useful for any useful differential gene expression screening strategy (e.g., , DD-PCR, subtractive conformance libraries, etc.) Following identification of differentially-expressed genes, It is desirable to determine which conditions can extend the range of mRNA changes compared to the control, and conditions that allow the analysis of the largest number of genes are used.

C. Analysis Features

For each cell line, the genes representing significant regulation (preferably at least five-fold increases or decreases in baseline levels) are different but have a common effect on IFNβ (eg IFNα, IFNγ, IL-8, IL-12). It is used to screen a set of known compounds. This investigation can be performed using RNA isolated from each compound-treated cell to analyze cDNAs for this related gene and prepare a matching probe. Reactive genes are examined for response to each compound. In embodiments of one or more gene sets comprising gene / cell combinations, there are groups that respond only to IIFNβ, simultaneously to INFα and β, and to IL-8, IFNγ and IFNβ, and the like.

D. Assay Screening

The "best" gene-cell combination (the ratio of noise to signal for the highest response and detection; measurable gene expression in the cell line in which the other "information" gene was measured) from each group of genes was selected for the compound screen. do. Internal regulatory genes have been provided to cell lines to be used as markers of cytotoxicity (eg gadd45, hsp 70)

E. Screening

Test compound libraries were screened for those test compounds that are specific regulators of IFN-responsive genes using methods of sensing active and inactive. An "active" hit is one that induces a change in gene expression prominently over the background of a particular assay. Test compounds are grouped according to their GEF and retested to determine EC ₅₀ for representative compounds.

In the generation of GEFs that predict for efficacy in vivo, it is not clear how close to GEFs of IFNβ “heats” required to have IFN-like activity in vivo. To determine this, test compounds that are active in many assays (ie, gene / cell combinations) were tested in cell-based assays for IFN responses (eg, anti-viral effects) prior to in vivo testing. This screen is used as a method of sorting through GEFs to determine if a "hit" with activity in very few IFN-response assays has IFN-like activity. If none of the hits active in the various GEF assays show activity in the biological assay, the compounds are preferably screened in combination with each other to be processed together to determine their GEF. The combination of compounds yielding new GEFs close to that of IFNβ was further tested for in vitro activity in biological assays.

Representative compounds were selected for in vivo investigation based on their activity in in vitro biological assays, potential in GEF assays and other useful information. If any "heat" meets the criteria for in vivo testing, they are examined for efficacy in the EAE model. If not, additional compound sources may be screened or weak “hits” may be utilized against their GEF to find more potent compounds before being examined in animal models.

F. Screening and "Selective" GEF Decisions

The GEF profile determined in the previous step can be used directly as a means of utilizing "induction" or as the most relevant compound represented. In this step of the assay, the maximum response to the EC50s and the inducing compound for each assay is considered.

"Inducible" compounds are often tested for adverse, undesirable effects in a suitable biological model (e.g., induction of testable fever in rabbit pyrogen assays). If there are "induction" compounds with different GEFs, the GEF corresponding to the "induction" with little activity in the assay is used for optimization. However, if the "derived" compound does not meet the screening requirements for the desired drug, further analysis of the screening panel is required, retesting for bioactivity "heat" is required, and in this new screen, the previously proposed GEF Compounds in the taxa are distinguished from each other by this new analysis (ie because of the different GEFs now found). In such cases, additional in vivo investigations are essential to validate new GEF predictions for in vivo efficacy and selection.

III. Identification of p53 Analogues for Cancer Treatment

A. Background

Mutations and deletions of p53 tumor suppressor genes are frequent in many cancers (Hollstein et al., Science 253: 49 (1991); Weinberg, Science 254: 1138 (1991)). Research over the last decade has indicated that the protein plays a dominant role in maintaining a normal balance between cell proliferation and death. More importantly, experimental evidence from in vitro and in vivo studies demonstrated the possibility of p53 protein replacement as a treatment for cancer (Wills et al., Hum. Gen. Ther. 5: 1079-1088 (1994)).

In addition to its transcriptional regulatory activity, p53 affects DNA replication and recovery as well as apoptosis signaling pathways. The profile of the change in the resulting gene expression from the expression of wild type (WT) p53 in cancer cells is used in the application shown here as a tool of investigation for compounds that mimic the activity of p53. The presence of temperature sensitive (ts) p53 proteins and multiple p53 variants, as well as expression systems that allow for the detection-regulation of protein expression (eg, lac or tet-inducible systems), makes this system suitable for drug-screening efforts. To expand.

B. Cells and Genes

Stably modified cancer cell lines (eg, infection or transduction techniques) that allow the regulation of p53 WT or variant expression are used to identify p53-dependent genes. This study is preferably performed on p53 unlabeled cells, although the criteria are not absolute. Any of the methods described in the previous examples can be used to identify related p53 response genes. RNA for this assay may be either (1) the expression of p53 protein is on or off (eg inducible expression system) or (2) the active versus inactive type of p53 protein (eg, temperature sensitive p53 protein or WT versus variant protein). ) Is isolated from the cells cultured under the conditions present.

In this example, (1) the effector compound is a 53 kD protein (ie p53) and is not a small molecule (ie estradiol) or polypeptide ligand (ie IFN-β), and (2) the study is more selective It is for alternative effector molecules that induce an effect such as p53 in vivo, although they are not or efficient molecules. In this regard, successful p53 analogs are combinations of compounds, each of which performs a “part” of essential p53 function. In the previous example, the cell line that exhibits the greatest response in response to the reference compound is selected for identifying the response gene. In this case, the smallest set of gene / cell information readouts that predict the tumor suppressor function of p53 is the preferred result of the assay screening step. Therefore, the first gene identification approach will examine several different tumor cell lines of tumor development inhibited by p53 induction / activity. The p53 response assay shared by all these cells is selected for further investigation.

C. Analytical Features and Screening

Although not required, in addition, a method for selecting a suitable assay is to investigate the expression of related genes by induction of WT p53 compared to its variants. Genes controlled by truncated or mutated p53 that retain their tumor suppressor function are useful for screening p53 analogues because they are markers of desired p53 function, and genes regulated by variants of p53 that are inactive in tumor suppression are screened from the screen. Removed or used as a "non-selective" analysis. The choice of assay used as the "cytotoxic" read described above differs from this application in this screen, and assays that do not respond to p53 read "cytotoxic" because some of the targets of p53 may be genes such as gadd45. Can be retained with water.

Investigations of gene expression patterns induced by compounds are similar to other studies. "Hits" are grouped according to their GEF and retested to determine EC ₅₀ for each activity assay.

D. Preliminary Cell-Based Assays

"Heat" refers to DNA-laddering induced by exposure to radiation in the presence of a compound (eg, measured by ³ H-thymidine uptake), non-proliferation (eg, soft agar assays), and cell death (eg, in the presence of a compound). Can be examined in an in vitro assay. This preliminary investigation further defines GEF predicting activity (measured by in vitro surrogate assay) in tumor suppression. In vitro systems can also be used to investigate the efficacy of a combination of "hits" that can be synergistic to yield a GEF that predicts tumor suppression function.

E. In Vivo Research

Representative compounds are selected for in vivo investigations based on activity in in vitro biological assays, potential in GEF assays and other useful information. The efficacy of the compounds in inhibiting the growth of human tumors in nude mice implanted with human tumors was investigated by measuring tumor-suppressive activity. The positive baseline state of this study is the same tumor cell made to express the inducible p53 protein, which allows for the regulation of tumor growth in vivo.

F. GEF-limited and derived compound optimization

GEF profiles related to in vitro and in vitro efficacy can be used directly as a means of utilizing "induced" compounds. Combination therapy is a screening step for combinations of compounds that are active in in vitro biological assays because they are difficult to investigate in in vivo assays due to possible pharmacokinetic differences in mixed components. At this stage of the assay, the maximum response to the EC50s and the inducible compound for each assay is considered.

Depending on the screening needs for the preferred drug, it is useful to match further analysis with the screening panel at this stage. In such cases, additional in vivo investigations are necessary to confirm the prediction of new GEFs for in vivo efficacy and selection.

Ⅳ. Identification of agonists that prevent cell invasion for the treatment of cancer

Therapeutic agents that prevent progression from primary cancer to metastasis are important components of anti-cancer drugs. Another aspect of the process by which cancer cells enter blood vessels and re-divide at a distant location past them is a potential target for anti-metastatic drugs. However, this is an attempt to identify anti-metastatic agents because of the lack of in vitro and in vivo models that predict the metastatic-forming ability of human cancer cells.

An important aspect of progression is the process by which cells penetrate through the endothelium and enter the surrounding matrix. Cell invasion through reconstituted vesicles (eg, Matrigel) can be used as an in vitro surrogate instead of in vivo development. However, the assay is not suitable for the screening of large compound libraries. The GEF methodology can be used to develop screens for agents that prevent or reduce cell invasion and / or metastasis.

Rather than using reference compounds to identify gene expression differences, the genes for this screen are identified by comparing the reference status. Predictive baseline conditions include, but are not limited to, invasive versus non-invasive cell lines, normal versus invasive carcinoma tissue or two histopathologically-produced malignant tissues (eg, prostate carcinoma of Gleason Grades III and IV).

A. Cells and Genes

Cell and tissue samples (eg, high invasiveness or metastasis from normal) that represent various stages of cancer progression are used as the source of RNA. Cell lines reported for studies of breast cancer progression have been reported extracellular invasive properties (eg, normal human breast epithelial cells, infinitely proliferating MCF10A or 184B5, weakly invasive MCF7, ZR75-1, MDA468, moderately invasive MDA435 and high) Tissue samples are the basis for invasive MDA231 or BT549 (ATCC, Rockville, MD) Tissue samples are human xenografts of immunodeficient animals, tumors obtained by pathologists, for example by Laser Capture Microdissection and normal and invasive substances or Include in vivo tissue incised to include similarly characterized cells (Emmert-Buck et al., Science 274: 998-1001 (1996)) Comparisons made in biological and tissue samples derived from the same tissue Although there is a scientific reason for this, this suggests that processes common to metastasis of different cancer types may lead to screens using cells and living tissue from other tissues. It can be targeted.

Other approaches can be used to determine similarities in gene expression differences among these RNA samples. RNA isolated from its normal and majority invasive cells (or living tissue) can be compared using the methods described above to identify differences between treated and untreated cells (eg, DD-PCR, subtracks). CDNA library, high density cDNA array). Samples from tumor cell lines or specimens representing different stages of normal versus cancer progression are used to generate this gene expression comparison and are, in fact, preferred because of the larger pool of differentially expressed humans that can be produced. More importantly for tumor cells, because of the individual diversity in the tumor, the difference will be expressed in different gene expression profiles.

Genes differentially expressed between normal and highly invasive cells are selected by the following investigation.

B. Analytical Features and Screening

Genes identified and identified differentially in the first step are examined for inclusion in the GEF based on their expression in the cells considered for use in the screening process. For example, if initial gene identification is performed using RNA isolated from a tissue sample and not from a cell culture material, any gene expressed in vivo may not be similarly expressed or regulated in the culture environment. Preferably cell lines expressing the highest number and high levels of mRNA for differentially expressed genes are selected for GEF analysis.

In the course of investigating the expression of related genes in normal versus invasive cultured cells, it is also desirable to test their relative expression in tumor cells that are not invasive or mildly invasive. In these cells, comparing gene expression enhancement, it can be seen that a portion of the gene is often regulated in only the invasive cells or the majority of the invasive cell lines tested. This part will in particular provide information on inclusion in the GEF.

In some embodiments, regulation of expression of any of the relevant genes of an agonist reported to modulate cancer cell invasion (eg, TGFβ, metastasis inhibition nm23, anti-Ha-ras ribozyme) is measured. Genes with expression affected by the agonist are included in the GEF.

"Best" analysis (eg, gene / cell combination with noise ratio to signal for highest fold response and detection) is selected for the compound screen. Suitable genes used as indicators of cytotoxicity (eg gadd45, hsp70) or internal regulation (eg GAPDH) are also matched to GEF.

C. Compound Screening

Investigations of gene expression patterns induced by the compounds are similar to those of the other investigations described above. "Hits" are grouped according to their GEF and retested to determine EC ₅₀ for activity in each assay.

D. Preliminary Cell-Based Analysis

"Hits" are initially examined in in vitro assays for invasion (eg, modified Boyden chamber (Albini et al., Cancer Res. 47: 3239-3245 (1987)). This preliminary investigation is active in tumor cell invasion. Define a GEF that predicts (as measured by an in vitro surrogate assay) The in vitro system also exhibits no activity when tested alone, but a combination of “hits” with other GEFs that are mixed together to show activity. It can be used to investigate the efficacy of.

E. In Vivo Research

Representative compounds are preferably selected for in vivo investigations based on their potential in GEF assays. The efficacy of a compound that inhibits tumor invasion can be investigated by a number of methods, including metastatic growth of tumors implanted in nude mice or invasion of tumor cells implanted in the renal capsule.

F. Utilization of GEF Limited and Inducible Compounds

GEF profiles related to in vitro and in vivo efficacy can be used directly as a means of utilizing "induced" compounds. This would be an essential step for the combination of compounds that are active in in vitro biological assays because the combination treatment is difficult to investigate in vivo assays with possible pharmacokinetic differences in the components of the mixture. At this stage of the assay, EC50s and maximal response to the inducing compounds of each assay are considered.

Depending on the selective requirements for the desired drug, it is useful at this stage to meet additional assays in the screening panel. In such cases, additional in vivo investigations are necessary to confirm the prediction of new GEFs for in vivo efficacy and selection.

Ⅴ. Identification of agonists that prevent breast tumor progression

A. Background

Hormone-dependent, progression of breast cancer (BC) from carcinoma differentiated into more advanced stage lesions results in loss of estrogen receptor (ER) function, decreased estrogen-cadherin (E-cadherin) expression, and increased ratio Distinguished by mentin expression. This progression is similar to mesodermal metastasis (EMT) of epithelial cells during embryonic time (Hay, Acta Anat. 154: 8-20 (1995)). Advanced stage breast cancer cells are similar in structural and functional characteristics to mesodermal cells. Altered expression of mesofine fibers affects this morphology (eg, reduced expression of some keratin and induction of non-mentin synthesis compared to less advanced cancer cells). The changes added are not only increased proteolytic activity (eg, stromelysin, MMPS) but also signaling proteins (eg, E-cadherin, ZO-1), adhesion components (eg, integrin) and extracellular matrix of the cell junction. Reduced expression / function of proteins (eg, thrombospodine). At a later stage of interest, advanced breast cancer (ABC) may be applied to cultured BC cells that exhibit reduced intracellular signaling and adhesion of hormone dependence, increased motility and increased invasiveness through reconstituted basement membranes (ie, Matrigel). In vitro (Thompson et al., J. Cell Physiol. 150: 534-544 (1992)).

Because motility and invasive capacity are the first salient features of ABC cells, we designed the experiment to identify gene expression fingerprints (GEFs) that could be replaced by the phenotypic analysis commonly used to measure this activity. . Additional GEFs may be used for cancer cell progression, such as proliferation (eg, activity of proliferation), cell death (eg, cell death response), angiogenesis (eg, activity of angiogenesis), differentiation, inflammation, and intercellular or cell-stromal responses. Can be replaced by other assays commonly used to measure. The method is to identify genes with controlled expression during the course of change, tumorigenesis, or tumor / metastasis suppression in the majority of ABCs. Genes in a set of differentially expressed genes with expressions altered by known anti-invasive or anti-transitional agents will preferably be included in the GEF used for drug screening. GEFs predict drug efficacy in the diagnosis and treatment of ABC. Changes in GEF of the screening cell line identify compounds as potential inductions for further optimization.

B. Developing Diagnostic GEFs for Weak and Strongly Invasive Breast Cancer

To induce GEF, which can be used in compound screens for agents that interfere with or prevent progression of breast cancer's invasive and / or metastatic activity, we identify gene expression changes commonly found in BC cell lines compared to normal cells. it started. For this study, we analyzed 14 cell lines derived from clinical specimens cultured from primary or metastatic samples of patients with invasive vascular carcinoma, which is the most common form of breast cancer (Table 5, Group I-III). ). The majority of these cell lines have been widely characterized for their in vivo tumorigenic and metastatic capacity as well as their proliferative characteristics and invasive capacity. Expression of informational marker genes ER, E-cadherin and non-mentin separates BC cells into three groups [Table 5: Group I shows ER-positive (ER +), E-cadherin positive (E-cad +), Non-mentin-negative (Vim-); Group II is negative for all markers; Group III is negative for ER and E-cadherin expression but positive for non-mentin expression. When classified on the basis of their invasiveness in Boyden chamber analysis, these BC cell lines are separated into only two groups: those that are weakly invasive (Inv-w) (including groups I and II) and high invasive (Inv-h). (Group III). All non-mentin-expressing BC cell lines are highly invasive and have star-shaped characteristics when cultured in Matrigel. In vivo, cells in this group are the only BC cell lines capable of forming metastases in the lung and lymph nodes (ie MDA231, Hs578T, MDA435) or brain (ie MDA435) (Price et al., Cancer Res. 50: 717-721 (1990))

For each cell line, the sample origin and pathological assessments indicated in the initial description describing the cell line groves were listed. BP, benzopyrene; PE, pleural effusion; Ca, carcinoma.

Tumor development is recorded as + when xenograft is identified as apparent tumor development in nude or SCID mice.

Met, metastatic cell line. * Estrogen pellets required for tumor formation. HBL-100 cell lines are labeled "various" because they show tumor formation after extended passage in culture.

The activity of the morphology of cells cultured in Matrigel and in the Boyden chamber invasion assays were determined by Sommers, CL et al Breast Cancer Res. & Treat. 31: 325-335 (1994). ** Data from this study.

Expression of mRNAs and proteins of the marker genes estrogen receptor (ER), E-cadherin (E-cad) and bimentin (Vim) is by reference.

nd, not measured.

Gene expression profiles for all BC cell lines that represent different clinical and morphological stages in BC progression are determined using Clontech (ie, Human Atlas I) cDNA arrays. This assay can be expanded to include additional genes (eg, other assays, cDNA libraries) and cell sources. As a basis for this study, we analyzed gene expression patterns in MCF10A, a naturally occurring endogenous proliferative "normal" breast epithelial cell (MEC) line derived from patients with fibrocystic disease (Soule et al., Cancer Res. 50 6075-6086 (1990). Of additional “normal” cell cultures (ie, 76N MEC cell lines (Band and Sager, Proc. Natl. Acad. Sci. USA 86: 1249-1253 (1989))) and 184B5 benzopyrene-infinite proliferation derived from abridged mastectomy samples Gene expression profiles of MEC (Stampfer and Bartley, Proc. Natl. Acad. Sci. USA 82: 2394-2398 (1985)) are also obtained: RNA from each cell line is isolated and radiolabeled for conformation to Atlas I arrays. It is used to prepare cDNA probes, which represent six functional gene classes, including oncogenes and tumor suppressor genes, genes involved in cell cycle regulation, intercellular interactions, genes involved in cell death and signal transduction pathways. CDNA fragments corresponding to 588 different genes: Over half of the genes present in the Atlas I array were expressed in human breast epithelial cells, resulting in approximately 300 out of 588 genes in this assay. It is not. Meets the signal from each cDNA that is quantitative analysis, are compared with the signals obtained for the same gene in an array conforming to the probe prepared from the reference MCF10A RNA.

An important element of the development of GEF for compound screening is the identification of gene expression changes that can be used to distinguish tumor-derived and "normal" cells, as well as highly invasive and weakly invasive tumors. This is particularly important in developing methods for screening for anti-cancer agents because cancer is the result of gene instability and accumulated somatic mutations that cause complex changes in gene expression. Therefore, we investigated the expression of genes commonly expressed in tumors versus "normal" cells or portions of tumor cells (eg, in four highly invasive BC cell lines). Table 6 shows the expression of frequently changed genes in tumor cells for the reference “normal” reference state. The value corresponds to the number of cell lines that changed to at least twice the mRNA levels observed for the indicated genes. Of the 28 genes, 11 were differentially expressed in the majority of tumor cell lines compared to the baseline “normal” baseline state (Table 6). The plectin gene is differentially expressed in all 14 BC cell lines, while the B-myb, transferrin R and ICH-2 protease genes have changed in 8 of 14 cell lines (Table 6).

Table 7A shows fold-differences in the mRNA levels observed for these genes in each of the cell lines compared to expression in reference MCF10A. The expression of most of the genes (ie 8/11) is reduced in BC cells compared to "normal" cells. The other three genes (ie B-myb, MacMarcks and transferrin R) show increased expression in BC cell lines. Other “normal” cells (ie 76N and 184B5) show minimal changes in the expression of these genes (FIG. 3A and data are omitted). Aspects of expression change for these genes (ie, increase or decrease for “normal” cells) represent “tumor-related” changes found in cultured breast tumor cell lines.

We also identified genes with mainly altered expression in BC cell lines that were classified as weak or highly invasive. Table 6 shows the number of cell lines in the weak or high invasive group showing differential expression of the indicated genes. Two of these genes (ie, GST P and Integrin A-3) are differentially expressed compared to "normal" in all 10 cell lines with low invasiveness, and the c-jun gene in all four highly invasive cell lines. Differentially expressed. The actual change in the expression level measured for each of these genes is shown in the table (Table 7B). Most of the genes associated with weak or high invasive cell lines relative to the "tumor-related" genes described above were overexpressed in these cells compared to "normal" cells. Highly invasive cell lines for the c-jun gene express higher mRNA levels than the reference "normal". For the GST P gene, all 14 cell lines express lower mRNAs than those of the baseline, but high invasive BC cell lines have higher levels of GST P mRNA than weak invasive cell lines as indicated by smaller negative value changes. These data demonstrate that some genes are differentially expressed (or inhibited) in weakly invasive cell lines, while other genes are differentially expressed (or inhibited) in more aggressive and highly invasive tumor cell lines.

The number of cell lines with changes in the expression of the indicated genes relative to MCF10A is provided. Only two or more pear-changes were seen.

The direction or extent of expression change differs in weak versus high invasive cells.

* Changes in the expression of BC cell lines relative to MCF10A

Consistent GEFs for weak and high invasive cancers are shown in FIG. 3A. Normal MEC cell lines (ie 76N) GEF are also shown for comparison. Three sub-profiles: tumor-related GEFs containing 11 genes (FIG. 3A, left striped bars (left to right striped bars)), GEF of weakly invasive carcinoma comprising 8 genes (FIG. 3A, filled bars) and GEF diagnosis for highly invasive ABC containing 6 genes (FIG. 3A, striped bars on the right (bars with stripes from left to right). These genes are "normal" Compared to weak and high invasive cell lines compared to (Fig. 3A, dotted bars), and thus the diagnosis of invasive conditions, these data indicate that the expression pattern of 28 genes in the cell line is not specified We suggest that it can be used as a means of predicting tumorigenesis and invasive potential, we suggest GEF of two cell lines that have not been tested for invasiveness. One of these is a cell line derived from our laboratory from breast fibroadenoma that has been cultured and cultured with HPV E6 / E7 tumor genes, and the other is one that encodes the T antigen protein obtained from human steamed epithelial cells. HBL100 cell line that will contain the integrated SV40 gene sequence (Vanhamme and Szpire, Carcinogenesis 9: 653-655 (1988)). The expression profile of these two cell lines is shown in Figure 3B. From this aspect we see the HBL-100 cell line. Is a tumor-induced mesoderm, a highly invasive cell line, on the contrary, 006FA-B cells differ significantly from "normal" endemic HMECs such as MCF10A and 184B5, but the differential gene expression of any of the tumor cell types profiled in this study Growth characteristics in the Matrigel of these two cell lines are not related to the phenotype predicted by their GEF. In agreement with the GEFs for these cell lines, 006FA-2B had a fused form in Matrigel, while HBL-100 had a star-shaped appearance of mesodermal cells with high invasiveness. Multiplies.

GEFs identified in cell culture models of breast cancer are valuable for investigating clinical specimens in response to staged or drug treatment. Gene expression patterns are measured on three cancerous living tissues obtained from patients with moderately differentiated invasive vascular carcinoma and compared to gene expression profiles of normal breast tissue. In the profile shown in FIG. 3C, characteristic cancer-related GEFs are shown in three types of tumors and are prominent in tumors T8911044 and T8911045. In addition, all of these tumors exhibit GEF associated with weakly invasive tumors. These data suggest that GEFs similar to those described herein are useful for the diagnosis of cancer patients and for the treatment of cancer patients. They also suggest that cultured cells reproduce some of the changes in gene expression observed in the tumor environment in vivo.

C. Development of Process-Related GEFs

GEFs identified up to this point are diagnostics of the phenotypic state of high and weakly invasive cells. These gene expression differences are important in diagnostic applications. It is important whether the gene expression differences are possible or sufficient to indicate the activity of the anti-invasive or metastatic drug. The sebset selection of these 28 genes, which is most useful for predicting reserve product efficacy, is made by determining which of these genes are involved in the process of malignant progression. For this purpose, we measured gene expression changes that occur during cell transformation as well as tumor and / or metastasis inhibition. Models for this process include treatment with tumor gene-transformed normal HMECs, tumor suppressor gene-infected tumor cells, and anti-tumor drugs or other agonists. These studies analyze gene expression patterns by treatment of cells in vivo or in vitro under conditions including culture on subsequent Matrigel, low adherent tissue culture plates or other cell types, and the like, but are not limited to these conditions. . Of particular importance is the change in gene expression that occurs during the conversion of weak or non-invasive BC cells to those with high invasiveness treated by infection with growth factors (eg EGF, scatter factor) or tumor genes (eg v-ras). Do. Additional model systems that repeat EMT (eg, treatment with anti-E-cadherin antibodies) can also be used to define genes that report the invasive properties of BC cells. Information regarding changes in gene expression associated with a reduction in invasive efficacy in response to treatment of drug or invasive-inhibitory gene products is also desirable for inducing GEF for compound screening.

Normal, limited lifespan HMECs can be proliferated indefinitely by expression of the SV40 T antigen, HPV E6 tumor gene, or selected p53 variant protein (Band, Intl. J. Oncol. 12: 499-507 (1998)). Using the Atlas I array, we measured the gene expression changes generated in HMECs that were infinitely propagated by mutant p53-expressing retrovirus infections (Gao et al., Cancer Res. 56: 3129-3133 (1996)). The expression levels of the 13 genes are infinitely multiplied with three other p53 variant proteins that act as overwhelming-negative inhibitors of p53 function, six of which are included in the "tumor-related" GEF (Table 8). This data limits that inactivation of p53 is an important determinant of the reduced gene expression observed for those genes. These data also mean that these genes are information providers at critical stages of the oncogenic process (cell proliferation). They are also important for the development of tumors where p53 inactivation is indicated by the majority of these BC cell lines.

Variation of p53 is associated with most of breast carcinomas. It is of interest that tumor biopsies show reduced expression of four of these six genes compared to normal tissue reference conditions (Table 8). These studies demonstrate a means of identifying GEFs that indicate the process of tumor formation. In addition, genes comprising GEF identified as a diagnosis of ABC are included in the gene-cell combinations used in drug screens.

Identification of genes predicting anti-invasive drug activity is confirmed by measuring changes in gene expression as a result of treatment of anti-invasive or anti-metastatic drugs with highly invasive cells. By comparing the effects of anti-invasive compounds with other known mechanisms of activity, a set of genes with expression changes with anti-invasive activity can be derived. Also important are drugs that are ineffective in preventing invasion but that are modulators of signaling pathways that do not inhibit invasion, as well as drugs with other anti-tumor properties (eg, early-cell death, anti-angiogenesis, anti-proliferation). Measurement of gene expression changes caused by In this study, we tested taxol, mevastatin, sodium butyrate, retinoic acid (RA), and caffeic acid (CA). It is reported that the efficacy of Taxol is dependent on the inhibition of microtubule formation, while mevastatin inhibits the HMG CoA reducing agent and indirectly protein lipidation to stop the cell cycle of the G1 phase. Sodium butyrate is a differentiating agent that causes histone acetylation and transcriptional activity. RA has anti-proliferative and differentiating effects in some BC cell lines (ie ER +), but not others (ie ER-). Taxol and mevastatin can both prevent the development of the characteristic star-shaped mesodermal cell morphology of MDA231 cells, while sodium butyrate is ineffective (data omitted). Taxol has also been shown to prevent the invasion of MDA231 in Boyden chamber analysis (Sasaki and Passanti, Biotechniques 24: 1038-1043 (1998)) and mevastatin inhibits breast tumor metastasis in vivo (Alonso et al., Breast Cancer Res. Treat. 50: 83-93 (1998). Highly invasive MDA231 BC cells are treated with these compounds under conditions (ie levels and times) that have been reported to exhibit maximal effects with little toxicity. Taxol, mevastatin and butyrate treatment resulted in a doubling change in the Atlas I array gene expressed approximately 10% (ie Taxol: 27/300; mevastatin: 33/300; butyrate: 39/300), whereas RA Or had little effect on CA treatment. Gene expression profiles for each of these compounds can be easily distinguished from each other (FIG. 4). Importantly, 12 of the 28 genes identified as potential markers of oncogenesis or invasion are regulated by one or more of these drugs. In addition, the direction of gene expression changes induced by these drugs for 11 of these 12 genes indicates a more "normal" or lower invasive GEF (Table 8). For example, the expression of seven genes that are inhibited or expanded in highly invasive MDA231 cancer cells as compared to "normal" is reversed. Changes in the expression of four of these genes (ie RABP II, Integrin A-3, DB1 and GST P) are lower than those of lower invasive GEFs (eg, RABP II expression is similar to changes in expression of weakly invasive cell lines by drug treatment). "Appear at a higher level than the cells" Such data are not essential for identifying genes that these genes can be used to inform drug activity of affecting malignant progression or to predict anti-invasive effects by themselves. Some of the genes regulated by mevastatin and taxol, but not butyrate (ie, GC-Box BP, RABP II, DB1), are anti-invasive when these agonists are based on matrigel morphology, while butyrate is not. It is believed to have an anti-invasive effect. Investigation of additional drug treatments with anti-invasive effects, as well as those with anti-proliferative or pre-cell quenching effects, allows for fine tuning of GEF to best predict drug efficacy, screening for invasiveness and potential toxicity.

The trend of expression change for each of the indicated genes was tabulated under the heading Diagnosis of differences in BC cell lines and tumor biopsies relative to MCF10A and normal breast tissue (data from Tables 7A and 7B and FIG. 3C). . Under the title Process, genes were directed to oncogenic columns. Trends in gene expression change in highly invasive MDA231 cell lines in response to treatment with taxol, mevastatin (mev) or sodium butyrate are provided in anti-cancer drug columns.

D. GEF Limited for Anti-Invasive Drug Screening

The studies described have induced GEFs that correspond to 28 gene expressions useful for distinguishing weak and highly invasive BC cell lines and tumor biological tissues. There are some of the gene expression changes associated with all BC cell lines and tumors (ie tumor-associated GEFs) within GEF. In combination with tumor-associated GEFs, two other distinct sub-GEFs define weak versus high invasive cancer. Experiments using tumor progression model systems (ie, p53 inactivation) and anti-oriented drug treatments have genes identified within 28 genes regulated during the inhibition of oncogenic processes or invasion.

Precise GEF predicting anti-invasive drug efficacy is a change in the expression of a portion of the 28 gene GEF representatives of highly invasive cancer cells. That portion is determined by a screening process similar to that used to induce diagnostic GEFs. Genes often affected by drugs or other agents that control the invasive phenotype are compared to diagnostic GEFs to induce common gene expression changes, which produce a GEF prediction of drug efficacy. Gene-cell combinations used to screen for anti-invasive compounds include at least two genes from each of the highly invasive MDA231 cell line and the sub-GEFs described above (ie, tumor-associated, weakly invasive and highly invasive). . Gene and cell line selection also takes into account data from drug treatments as well as those of weakly invasive cells. GEF screens can be performed in one or more cell lines in mixed or parallel cultures.

E. Materials and Methods

1. Cell Culture and Compound Treatment

76N human MEC cell lines and 184B5 benzopyrene-increasing human MEC cell lines are cultured in DFCI-1 medium (Band and Sager, Proc. Natl. Acad. Sci. USA 86: 1249-1253 (1989)). The 006FA-2B cell line is generated from positive fibroadenoma tissue samples by infecting plasmid vectors and selective SVneo plasmids encoding the HPV16 E6 and E7 tumor genes together in cultured organs using standard calcium phosphorus-mediated procedures. 006FA-2B is G418 (100 μg / ml, Gibco), MCF10A, HBL-100, T47D, ZR75-1, MCF7, BT483, MDA361, BT474, BT20, MDA468, SKBR3, MDA453, from ATCC (Rockville, MD) One of a number of stable epithelial cell clones with extended lifespan selected using BT549, Hs578T, MDA231 and MDA435S cells, and is initially cultured in ATCC-recommended media. To determine the continuous gene expression profile of the breast tumor cell line, the cells were treated with α-MEM medium [1 mM HEPES, 2 mM glutamine, 0.1 mM MEM non-essential amino acid, 1.0 mM pyruvate, 50 μg / ml gentamicin, 1.0 Cultured at 80-90% consensus in μg / ml insulin (both Gibco, Gaitherburg, MD) and α-modified MEM supplemented with 10% FBS (Intergen). To investigate the effect of the compound selected for the gene expression in the cell line MDA231 understand, the cell is such that α-MEM mounting plate being in a culture medium (10 ^6/100 mm dish), overnight. Cells were treated with 3 mM sodium butyrate (Specialty Media, Inc. Lavallette, NJ), 5.0 μM Taxol (Molecular Probes, Inc, Eugene, OR), ^10-8 M caffeic acid, 1.0 M retinoic acid or 10 μM mevastatin (all from Sigma). ), And the cell monolayer is collected after 72 hours for RNA isolation.

2. Gene Expression Analysis

Total RNA from cell lines and compound-treated cells is isolated by guanidinium-isothiocyanate-CsCl gradient process (Chirgwin et al., Biochemistry 18: 5294-5299 (1979)). Total RNA from normal and tumor tissue specimens was obtained from BioChain Institute, Inc (San Leandro, Calif.).

Preparation of radioisotope labeled cDNA from total RNA (5 μg) is performed as described in the Clontech Atlas I cDNA Array Conformance Kit Method. The only exception is the step for the removal of the unattached nucleotide triphosphate, which uses a G50 spin column and the length of the pre-attach is increased by at least 6 hours. Probe levels commonly used in matching reactions range from 0.7 to 1.0 × 10 ⁶ counts per minute / ml (cpm / ml).

3. Phase Analysis of Clontech Atlas I cDNA Expression Arrays

Probe intensity at each target (cDNA) point on the Atlas I array was quantified using an "array vision" software package (Imaging Research, Inc., St. Catherine, Ontario, Canada). The grid confinement method was used as an automatic algorithm in this analysis to apply the grid to cover the target. In the array, each target is a Storm Phsphorimaging System by Molecular Dynamics. Scanned using Inc (Sunnyvale, Calif.), The data table was made with modified mean PSL × area values (PSL values per pixel adjust the area in mm of the target) corrected for background and reference normalization. The average background is measured from the selected blank portion of the array, and the reference value of normalization is calculated using the average of all signals of the target on the array. The ratio and Z-score difference between the two samples are calculated and the differentially expressed genes are identified with a set of common threshold ratios and differences. For this analysis, the threshold ratio was doubled and the z score value was 0.3.

All the criteria cited herein are explicitly met by the criteria in their entirety for the whole intent.

Although any detailed description has been described by way of illustration and example for purposes of understanding the invention, it will be apparent that this is a change and modification that may be made within the scope of the appended claims.

Claims (59)

In the method of grouping the test compound, the method (a) exposing a cell culture or culture comprising at least two gene-cell combinations to a test compound to generate an exposed cell culture or culture, wherein each of the at least two gene-cell combinations comprises a specific gene and a specific cell type Contains unusual combinations of cells, (b) separating RNA from the exposed cell culture, (c) screen RNA from (b) for mRNA of each particular gene of at least two gene-cell combinations to yield a gene expression fingerprint (GEF) for the test compound, (d) repeat (a) to (c) to classify each test compound, (e) comparing the GEFs for each test compound tested in (a) to (d) and grouping the test compounds into at least two taxa groups based on differences or similarities in their GEFs To classify test compounds. The method of claim 1, wherein the at least two gene-cell combinations comprise at least two different genes. The method of claim 1, wherein the at least two gene-cell combinations comprise at least two cell types. The method of claim 1, wherein the screening comprises PCR amplification using specific oligonucleotide primers for each gene. The method of claim 1, wherein the RNA is further reverse transcribed into cDNA. 6. The method of claim 1 or 5, wherein said screening comprises matching the specific nucleic acid sequence of each gene to RNA or cDNA. The method of claim 1, wherein at least one gene in at least two gene-cell combinations comprises endogenous genes under the control of its own promoter. The method of claim 1, wherein at least one gene in the at least two gene-cell combinations comprises a heterologous gene under the control of a heterologous promoter. The method of claim 1, wherein at least one gene in the at least two gene-cell combinations further comprises an internal negative regulatory gene, and the effect on the mRNA level of the negative regulatory gene in response to the test compound is toxic of the test compound. A method for grouping test compounds, characterized by directing influence. The method of claim 1, wherein at least one gene in the at least two gene-cell combinations further comprises an internal negative regulatory gene, wherein the effect on the mRNA level of the negative regulatory gene in response to the test compound is non- A method for grouping test compounds, characterized by indicative of specific effects. The method of claim 1, wherein said screening further comprises quantifying the effect of mRNA levels of at least one gene in at least two gene-cell combinations. The test compound of claim 1, further comprising adding a combination of two or more test compounds to the cell culture in (a), wherein a GEF for the combination of the two or more test compounds is calculated. How to group it. The test compound according to claim 1, wherein the test compound is an estrogen, p53, IFNβ, TNFα, endothelin, tamoxifen, raloxifene, IFNα, IFNγ, or an analogue of anti-Ha-ras ribozyme. How to. The test compound of claim 1, wherein the test compound is a peptide, peptide analog, polypeptide, protein, ribozyme, nucleic acid, oligonucleotide, organic or inorganic compound, or extract of an animal, plant, or microorganism. How to group. The method of claim 1, wherein the method further comprises testing a test compound representative of each taxon for activity of interest in vitro. The method of claim 15, wherein the representative test compound is p53, estrogen, raloxifene, tamoxifen, or an analog of IFNβ. 16. The method of claim 15, wherein said activity of interest is drug suppression. The method of claim 15, wherein the activity of interest is reduced bone loss. The method of claim 15, wherein the activity of interest is anti-metastatic activity, prevents progression of atherosclerosis, reduced inflammation in rheumatoid arthritis, enhanced cognitive function, or prevents runaway. In the method for classifying test compounds, the method is as follows. (a) exposing a cell culture or a culture comprising at least two gene-cell combinations to a test compound to generate an exposed cell culture or culture, each of the at least two gene-cell combinations of a particular gene and a particular cell type A specific combination, wherein at least one gene in said at least two gene-cell combinations is differentially expressed in a first and a second reference state, (b) separating RNA from the exposed cell culture of (a), (c) screen RNA from (b) for mRNA of each specific gene in each of at least two gene-cell combinations to yield a gene expression fingerprint (GEF) for the test compound, (d) repeat (a) to (c) to classify each test compound, (e) comparing GEF for each test compound tested in (a)-(d), And classifying the test compound into at least two taxa based on differences in their GEFs. 21. The method of claim 20, wherein at least one of the first and second baseline conditions is a disease state. 22. The method of claim 21, wherein said disease state is cancer. The method of claim 20, wherein said screening comprises PCR amplification using oligonucleotide primers specific for each gene in at least two gene-cell combinations. The method of claim 20, wherein said RNA is further reverse transcribed into cDNA. 25. The method of claim 20 or 24, wherein said screening comprises matching the specific nucleic acid tamchip of each gene to RNA or cDNA in at least two gene-cell combinations. 21. The method of claim 20, wherein at least one gene in the at least two gene-cell combinations further comprises an internal negative regulatory gene, and the effect on the mRNA level of the negative regulatory gene in response to the test compound affects the toxicity of the test compound. Instructing a test compound to be grouped. The method of claim 20, wherein at least one gene in the at least two gene-cell combinations further comprises an internal negative regulatory gene, and wherein the effect on the mRNA level of the negative regulatory gene in response to the test compound is non-specific. A method for grouping test compounds, characterized by directing their effects. 21. The method of claim 20, wherein said screening further comprises quantifying the level of mRNA of each gene in at least two gene-cell combinations. 21. The method of claim 20, wherein the method further comprises testing representative test compounds in each taxa for desired activity in vivo. The test compound of claim 20, wherein the method further comprises adding a combination of two or more test compounds to the cell culture in (a), wherein a GEF is calculated for the combination of the two or more test compounds. How to group it. A method of calculating a reference gene expression fingerprint (GEF) for at least one reference compound for use in grouping test compounds, the method comprising: (a) identifying at least two gene-cell combinations, each of the at least two gene-cell combinations comprising a particular combination of a particular gene and a cell of a particular cell type, wherein the first gene-cell combination is , (i) exposing a host cell or host cell culture in vivo of a first cell type to a first reference compound, (ii) isolating RNA from the exposed in vivo host cell or host cell culture of (i), (iii) comparing the RNA of (ii) with the RNA isolated from a host cell or host cell culture in vivo of a first cell type not exposed to the first reference compound and reacting with the first reference compound Changes in gene mRNA levels in cells identify cells of a first cell type as a first gene-cell combination for use in grouping genes and test compounds, and a second gene-cell combination, (Iii) exposing a host cell or host cell culture in vivo of a first cell type or a second cell type to a first reference compound, (Iii) separating RNA from the exposed in vivo host cell or host cell culture, (Iii) compare RNA prepared from in vivo host cells or host cell cultures of cell types as in (v) not exposed to the first reference compound and (i) RNA and change mRNA in response to the first reference compound A gene having a level is identified as a gene for use in a second gene-cell combination for grouping test compounds, the second gene-cell combination being different from the first gene-cell combination, and the identified gene ( Cell lines of the same cell type as in iii), (b) screening RNAs of (ii) and (iii) for mRNA of each gene in each of at least two gene-cell combinations to yield a reference GEF for the first reference compound for use in grouping test compounds A reference GEF calculation method for a reference compound for classifying a test compound, characterized in that it comprises a. 32. The method of claim 31, wherein said at least two gene-cell combinations comprise at least two different genes. 32. The method according to claim 31, wherein said at least two gene-cell combinations comprise at least two different cell types. 32. The method according to claim 31, comprising PCR amplification using specific oligonucleotide primers for each gene of at least two gene-cell combinations. 32. The method according to claim 31, wherein the RNA is further reverse transcribed into cDNA. 36. The method of claim 31 or 35, wherein said screening comprises matching the specific nucleic acid sequence for each gene of at least two gene-cell combinations to RNA or cDNA. Method for calculating reference GEF for a compound. 32. The method according to claim 31, wherein at least one gene in at least two gene-cell combinations comprises endogenous genes under the control of its own promoter. 32. The method according to claim 31, wherein at least one gene in at least two gene-cell combinations comprises a heterologous gene under the control of a heterologous promoter. The method of claim 31, wherein at least one gene in the at least two gene-cell combinations further comprises an internal negative regulatory gene, wherein the effect of mRNA levels of the negative regulatory gene in response to the test compound indicates a toxic effect of the test compound. A reference GEF calculation method for a reference compound for classifying test compounds, characterized in that. The method of claim 31, wherein at least one gene in the at least two gene-cell combinations further comprises an internal negative regulatory gene, and the effect of mRNA levels of the negative regulatory gene in response to the test compound is non-specific effect of the test compound. A method for calculating a reference GEF for a reference compound for classifying a test compound, characterized in that it indicates. 32. The method of claim 31, wherein said screening comprises quantifying the effect of mRNA levels of at least one gene in at least two gene-cell combinations. . The test compound of claim 31, wherein the first reference compound is estrogen, p53, IFNβ, TNFα, endothelin, tamoxifen, raloxifene, IFNα, IFNγ, or anti-Ha-ras ribozyme Method for calculating a reference GEF for a reference compound. The test compound of claim 31, wherein the first reference compound is a peptide, peptide analog, polypeptide, protein, ribozyme, nucleic acid, oligonucleotide, organic or inorganic compound, or extract of an animal, plant, or microorganism. A method of calculating a reference GEF for a reference compound for grouping the compounds. 32. The method of claim 31, wherein (a) to (b) are repeated in the second reference compound, whereby the gene having the changed mRNA level in response to only the first reference compound is not reacted to the second reference compound. A method of calculating a reference GEF for a reference compound for classifying a test compound, characterized in that it has been confirmed to have a specific response to the reference compound. The method of claim 44, wherein the second reference compound is different from the first reference compound and is an analog of estrogen, p53, IFNβ, TNFα, endothelin, tamoxifen, raloxifene, IFNα, IFNγ, or anti-Ha-ras ribozyme. A reference GEF calculation method for a reference compound for classifying a test compound, characterized in that. 45. The method of claim 44, wherein the second reference compound is a product of a gene expressed in a host cell. 32. The method of claim 31, wherein the first reference compound is a product of a gene expressed in a host cell. 32. The method according to claim 31, wherein said gene is a p53 gene. In the method of grouping the test compound, the method (a) calculating the GEF for the reference compound according to the method of claim 31; (b) A GEF is calculated to group each test compound, and the method of grouping is as follows. (i) exposing the cell culture or culture comprising at least two gene-cell combinations identified in claim 31 to a test compound, yielding the exposed cell culture or culture, (ii) separating RNA from the exposed cell culture or culture, (iii) screen RNA of (ii) for mRNA of each gene of at least two gene-cell combinations of (i) to yield GEF for the test compound, (Iii) repeat (i) to (iii) to calculate the GEF of each test compound for grouping; (c) comparing the GEF of each test compound produced in (b) with the reference GEF of (a), wherein the test compounds are grouped into at least two taxa groups based on the difference or similarity between their GEFs and the reference GEFs A method of grouping a test compound. The test compound of claim 49, wherein the method further comprises adding a combination of two or more test compounds to the cell culture in (a), wherein the GEF is calculated for the combination of the two or more test compounds. How to group it. The test compound of claim 49, wherein the test compound is an estrogen, p53, IFNβ, TNFα, endothelin, tamoxifen, raloxifene, IFNα, IFNγ, or an analogue of anti-Ha-ras ribozyme. Way. 50. The test compound according to claim 49, wherein the test compound is a peptide, peptide analog, polypeptide, protein, ribozyme, nucleic acid, oligonucleotide, organic or inorganic compound, or an extract of an animal, plant, or microorganism. How to. 50. The method of claim 49, wherein the method further comprises testing representative test compounds of each taxa for activity of interest in vivo. 55. The method of claim 53, wherein said representative test compound is p53, estrogen, raloxifene, tamoxifen, or an analog of IFNβ. The method of claim 53, wherein the activity of interest is tumor suppression. The method of claim 53, wherein the activity of interest is reduced bone loss. 55. The method of claim 53, wherein the activity of interest is anti-metastatic activity, prevents progression of atherosclerosis, reduced inflammation in rheumatoid arthritis, enhanced cognitive function, or prevents runaway. 21. The method of claim 20, wherein at least one of the first and second reference states comprises a change in cell phenotype. 59. The test compound according to claim 58, wherein the change in cell phenotype comprises a change in cell invasiveness, cell death response, angiogenic activity, proliferative activity, inflammation, intercellular interaction or cell-substrate interaction. How to group.