WO2006033769A2 - Assays and compositions for identification of cancer associated liver stem cells - Google Patents

Abstract

A transgenic animal model for liver cancer and liver stem cells derived therefrom are provided. Animals are genetically modified to comprise an oncogene that is conditionally expressed in liver cells. After induction of expression, and tumor formation, expression of the oncogene can be inactivated, resulting in regression of the tumor, and generation of a population of liver stem cells with latent tumorigenic properties.

Description

ASSAYS AND COMPOSITIONS FOR IDENTIFICATION OF CANCER ASSOCIATED

LIVER STEM CELLS

INTRODUCTION

[0001] The body depends on the liver to perform a number of vital functions, including regulation, synthesis, and secretion of many substances important in maintaining the body's normal state; storage of important nutrients such as glycogen (glucose), vitamins, and minerals; and purification, transformation, and clearance of waste products, drugs, and toxins. However, its distinctive characteristics and activities render it susceptible to damage from a variety of sources, and such damage can have enormous impact on a person's health.

[0002] Hepatocellular carcinoma arises from malignant hepatocytes. Although much less common than metastatic liver cancer in most areas of the world, hepatocellular carcinoma is the most common internal malignancy and an important cause of death in certain areas of Africa and Southeast Asia. Chronic hepatitis B virus (HBV) infection is largely responsible for the high prevalence of the tumor in endemic areas; the risk is more than one hundredfold higher among HBV carriers, and tumor incidence generally parallels HBV prevalence geographically. In HBV carriers, most of whom are asymptomatic, viral DNA eventually becomes incorporated into the host genome of infected hepatocytes, which leads to malignant transformation, possibly in combination with environmental carcinogens. Chronic hepatitis C virus (HCV) infection has also been recognized as an important factor in the genesis of hepatocellular carcinoma.

[0003] In North America, Europe, and other areas of low prevalence, most patients have underlying cirrhosis unrelated to HBV or HCV infection. Alcoholic, cryptogenic, and especially hemochromatotic cirrhosis are all prone to malignant transformation. The remaining patients have no apparent underlying liver disorder.

[0004] The prognosis for hepatocellular carcinoma is usually grim, and treatment is generally unsatisfactory. Surgical resection provides the best hope but is suitable in only a few cases. Patients with small localized tumors may have prolonged survival after resection, but the diagnosis is usually established late, and death often occurs within a few months. The tumor is not radiosensitive, and chemotherapy is usually unsuccessful, even when hepatic artery infusion or chemoembolization is used. Moderately good long-term survival rates have been reported after liver transplantation, but this may largely reflect selection bias for patients with relatively small localized tumors. Most experts remain wary of transplantation for malignancy.

[0005] Insight into the molecular pathogenesis of liver cancer may provide an opportunity to develop new therapeutic approaches. One of the most commonly activated oncogenes associated with the pathogenesis of liver tumors is the MYC oncogene. Many animal models have confirmed that overexpression of MYC can induce hepatocellular carcinoma, whereas the inhibition of MYC expression in hepatocellular carcinoma cell lines results in a loss in their neoplastic properties. The targeted inactivation of MYC may represent a novel approach for the treatment of liver cancer.

[0006] A fundamental problem in cancer research is identification of the cell type capable of sustaining the growth of the neoplastic clone. There is overwhelming evidence that virtually all cancers are clonal and represent the progeny of a single cell. Recent evidence has suggested that only a limited number of cells (stem cells) are capable of initiating the tumor at high frequency.

[0007] The model of cancer stem cells predicts functional heterogeneity among the cells that comprise the tumor, and that the rare stem cells are different from the vast majority of the cells that make up the tumor. Therefore, tumorigenic pathways may function differently in distinct tumor subpopulations, and studying the bulk of the cells that make up the tumor mass may be misleading as to key properties of the tumor. This model also predicts that, while eradication of the bulk of cancer cells may result in a remission, the disease will relapse if the stem cells cells are not eliminated. Resolution of the stem cell problem requires both purification of tumor cells into subfractions, and functional assays to detect cells with the capacity to initiate tumor growth in vivo. Unfortunately, identification and purification of solid tumor stem cells has been difficult, in part because of the paucity of cell surface markers that enable cell sorting.

[0008] To achieve a further characterization of liver stem cells, and the cells derived therefrom, it is critical to have well defined model systems, that can decipher the complex interplay between "environmental" factors and intrinsic cellular factors that regulate cell renewal, as well as the phenotypic definition of the specific cells capable of giving rise to mature hepatic cells. Identification and characterization of such cells are of great interest. Literature

[0009] Cancer is largely caused by genomic catastrophes that result in the activation of proto-oncogenes and/or the inactivation of tumor suppressor genes. A specific and effective form of cancer therapy may be to target the repair, replacement or inactivation of these mutant genes. It has been demonstrated through transgenic mouse models that even brief inactivation of a single oncogene can be sufficient to induce sustained tumor regression indicating that, at least in some cases, oncogene inactivation may result in the permanent loss of a neoplastic phenotype. See Jain et al. (2002) Science 297, 102-4; and Felsher & Bradon (2003) Drug News Perspect 16, 370-4 (2003). The targeted inactivation of oncogenes may be sufficient to reverse tumorigenesis. [0010] A discussion of hepatic progenitor cells may be found in Susick et al. (2001) Ann.

N.Y. Acad. Sci. 944:398-419; in U.S. Patent no. 5,576,207; U.S. Patent Application no.

20020016000; and in International Patent application WO 03/000848. A model animal for liver regeneration and engraftment is described in Wang et al. (2003) Proc Natl Acad Sci U

S A. 100 Suppl 1 :11881-8; and in Grompe et al. (1993) Genes & Development 7, 2298-

2307. [0011] The biology and implications of cancer stem cells are discussed by Al-Hajj et al.

(2004) Curr Opin Genet Dev. 14(1):43-7; Reya et al. (2001) Nature 414(6859): 105-11 ;

Kopper et al. (2004) Pathol Oncol Res. 10(2):69-73; Singh et al. (2003) Cancer Res.

63(18):5821-8; and Trott (1994) Radiother Oncol. 30(1): 1-5

SUMMARY OF THE INVENTION

[0012] A transgenic animal model for liver cancer, and liver stem cells derived therefrom are provided. Animals are genetically modified to comprise an oncogene that is conditionally expressed in liver cells. Expression of the oncogene is directly or, preferably, indirectly controlled by a liver specific promoter. The ability to turn expression of the oncogene on and off allows manipulation of the liver tumor cells. After induction of expression, and tumor formation, expression of the oncogene can be inactivated, resulting in regression of the tumor, and creation of a population of liver stem cells with latent tumorigenic properties.

[0013] The animal model and cells derived therefrom are useful in screening for compounds and therapies against liver cancer, and provide a model for tumor dormancy. The liver stem cells are also useful to identify factors influencing liver stem cell proliferation, to analyze gene expression patterns or protein expression patterns, to identify new anti¬ cancer drug targets, to predict the sensitivity of tumors from individual patients to existing anti-cancer treatment regimens, to model anti-cancer treatment, to test new therapeutic compounds, to identify and test new diagnostic markers, to treat tumors, to produce genetically modified liver stem cells, and to prepare cDNA libraries and microarrays of polynucleotides and polypeptides from liver stem cells.

BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1. A conditional transgenic model for MYC-induced hepatocellular cancer, a. The Tet system was used to generate mice that conditionally express MYC in their hepatocytes by crossing the LAP-tTA mice with tet-o-MYC mice. In the absence, but not the presence of doxycycline treatment, mice express the MYC transgene in their hepatocytes and succumb to liver cancer, b. Western analysis for MYC transgene expression. Mice transgenic for both LAP-tTA and tet-o-MYC express MYC in the liver, but mice transgenic for LAP-tTA alone do not. c. Kaplan-Meyer survival curve comparing survival of transgenic mice in the presence (square symbol) or absence of MYC transgene expression (diamond symbol). MYC transgene expression was induced in 4 week old mice by removing doxycycline treatment. Each cohort consists of 25 mice.

[0015] Figure 2. Inactivation of the MYC transgene results in sustained tumor regression.

Mice transgenic for both LAP-tTA and tet-o-MYC treated with doxycycline have a. a normal liver grossly and b. histologically. Mice transgenic for both LAP-tTA and tet-o-MYC that have not been treated with doxycycline succumb to c. multi-focal liver tumors that d. histologically are hepatocellular cancers. Mice with liver tumors that are subsequently treated with doxycycline to suppress the MYC transgene, e. undergo complete regression of their tumors and f. histologically the liver appears normal, g. Kaplan-Meyer survival curve of mice with liver tumors that are either not treated (diamond symbol) or treated with doxycycline to suppress MYC transgene expression (square symbol). Each cohort consists of 10 mice.

[0016] Figure 3. MYC inactivation in liver tumors results in rapid tumor regression associated with loss of expression of AFP and increased apoptosis. a. Western blot analysis for expression of MYC and AFP in normal wild type (WT) mouse liver, liver of neonatal mice, liver tumor with MYC overexpression and liver tumor where MYC has been inactivated for 12 hrs, 1 , 4 and 15 days. b. TUNEL assay for liver tumor and for tumor where MYC has been inactivated for 10 days. Upper panels show TUNEL staining and lower panels show DAPI staining of nuclei. Representative data from one of four experiments are shown.

[0017] Figure 4. MYC inactivation is associated with the differentiation of liver tumor cells into normal hepatocytes. a. Liver tumor cells were transplanted subcutaneously into SCID mouse, b. MYC inactivation resulted in tumor regression in the transplanted tumor, c. Histological analysis of site of tumor revealed normal appearing hepatocytes (marked as h) within the epidermis (marked as d). d. Higher magnification of the differentiated hepatocytes. This experiment was performed 3-5 times in 6 different transgenic lines using 3-5 mice in each group.

[0018] Figure 5. MYC inactivation in liver tumors results in the formation of normal hepatic structures, a, d, g, j, m. normal liver, b, e, h, k, n. MYC overexpressing tumor and c, f, i, I, o. regressed tumor. Serial sections were stained with a-c. H&E, lmmunohistochemical analysis was carried out for d-f. - Ki67 g-i. - CEA, j-l. - CK8, and m-o. - AFP. Representative data from one of three experiments.

[0019] Figure 6. Tumor dormancy of liver tumors upon MYC inactivation. a. Kinetics of tumor regression using in vivo bioluminescence imaging (BLI) of Iuciferase labeled liver tumors shows that transplanted tumors undergo rapid regression. Residual transplanted tumor cells with luciferase activity remain at the site of tumor growth until MYC reactivation. Upon MYC reactivation the residual tumor cells resume growth. Tumor regression and re- growth are evident from the plot of light emitted (photons/sec) from the region of interest covering the tumor sites versus time (days). For visualization of tumor growth, a pseudocolor image representing light-intensity is superimposed over a grayscale reference image of the representative animals in each treatment group; b. MYC on, c. MYC on/off 3months (m), d. MYC on/off for 3m/on for 2m. A representative control mouse is represented for the same time points; e. MYC on, f. MYC on/off for 3m, g. MYC on/off for 5m. Data shown are representative of 5 different experiments with 1-10 animals in each group.

[0020] Figure 7. MYC inactivation uncovers stem cell properties in hepatocellular cancer.

Expression of the hepatocellular stem cell marker CK19 in a. normal liver, hepatocellular cancer b. before and c. after MYC inactivation. d. Model for consequence of oncogene inactivation in hepatocellular cancer. MYC induced hepatocellular cancer consist of stem cells and progenitor cells. Upon MYC inactivation progenitor cells die whereas stem cells give rise to differentiated liver lineages of both hepatocytes and bile duct cells. Upon MYC reactivation the stem cells emerge from the dormant state and tumors relapse.

[0021] Figure 8. lmmunohistochemical analysis of MYC transgene expression. Absence of

MYC expression in the liver of a doxycycline treated transgenic mouse liver (MYC off). Robust MYC expression in the liver of a untreated transgenic mouse liver tumor (MYC on). Data is representative for 5 different experiments.

[0022] Figure 9. Sensitivity of in vivo bioluminescence imaging (BLI). a. Bioluminescence imaging was performed after the inoculation of 1x10³ to 1x10⁶ luciferase expressing tumor hepatocytes that had been inoculated subcutaneously into SCID mice. b-f. . BLI could detect luciferase labeled cells with a maximum sensitivity of 1x10³. Experiment was done 4 times.

[0023] Figure 10. Tumor growth measured by BLI. . 0.5x10⁶ luciferase positive tumor cells were injected into SCID mice. Tumor growth was measured by calipers. Luciferase activity was measured by BLI a. Luciferase activity correlated well with tumor size. b. BLI of a newly injected mouse immediately after injection c. BLI of a mouse with a large tumor, d. A control mouse with non-luciferase expressing liver tumor. Experiment was performed 2 times with a cohort of 5 mice.

[0024] Figure 11. BLI is a sensitive measurement of tumor growth to follow the therapeutic consequences of MYC inactivation. a. Tumor regression was measured in the transplanted liver tumor cells by light emission from luciferase labeled viable cells while tumor size was measured using a calipers. Using BLI, the decrease in luciferase activity is detected more rapidly that the decrease in tumor mass. b. Mouse with tumor before MYC inactivation (also indicated in figure a), c. mouse with completely regressed tumor still exhibits luciferase activity from cells by mechanical methods (also indicated in figure a) and d. control mouse with unlabeled tumor. Experiment was performed 2 times with 5 mice in each group. [0025] Figure 12. Normal hepatocytes do not engraft in SCID hosts. A normal liver from a healthy CMV-GFP-LUC mouse was digested with collagenase, and 4*10⁷ hepatocytes were injected subcutaneosly into a SCID host. The luciferase labeled hepatocytes were detectable by BLI 2 hours post-injection (a), but not after 10 days (b). A control mouse injected with PBS is shown for comparison (c). The experiment was carried out 3 times using 3-5 mice in each experiment.

DESCRIPTION OF THE SPECIFIC EMBODIMENTS

[0026] Hepatocellular carcinoma is a common solid malignancy, but one that is usually refractory to clinical treatment. The present invention provides a transgenic animal model, where inactivation of oncogene expression in liver cells is sufficient to induce sustained regression of even highly invasive liver cancers. Oncogene inactivation did not result in the elimination of all tumor cells, but rather some tumor cells retained stem cell properties and differentiated into normal liver cells forming hepatic sinusoids, bile caniculli and duct-like structures. These changes were associated with the loss of the expression of the tumor marker, AFP, and the gain in expression of maturation markers, CK-8 and CEA, and in some cells, the liver stem cell marker CK-19. Using in vivo bioluminescence imaging, it is demonstrated that these differentiated tumor cells remained dormant in the host for as long as the oncogene remained inactivated, but upon reactivation of oncogene expression the cells immediately restored their neoplastic features.

[0027] In the transgenic animals of the invention, animals are genetically modified to comprise an oncogene that is conditionally expressed in liver cells. Expression of the oncogene is directly or, preferably, indirectly controlled by a liver specific promoter. The ability to turn expression of the oncogene on and off allows manipulation of the liver tumor cells. After induction of expression, and tumor formation, expression of the oncogene can be inactivated, resulting in regression of the tumor, and creation of a population of liver stem cells with latent tumorigenic properties.

TRANSGENIC ANIMALS

[0028] The term "transgene" is used herein to describe genetic material that has been or is about to be artificially inserted into the genome of a mammalian cell, particularly a mammalian cell of a living animal. The transgene is used to transform a cell, meaning that a permanent or transient genetic change, preferably a permanent genetic change, is induced in a cell following incorporation of exogenous DNA. A permanent genetic change is generally achieved by introduction of the DNA into the genome of the cell. Vectors for stable integration include plasmids, retroviruses and other animal viruses, YACs, and the like. Of interest are transgenic mammals, e.g. cows, pigs, goats, horses, etc., and particularly rodents, e.g. rats, mice, etc.

[0029] Transgenic animals comprise an exogenous nucleic acid sequence present as an extrachromosomal element or stably integrated in all or a portion of its cells, especially in germ cells. Unless otherwise indicated, it will be assumed that a transgenic animal comprises stable changes to the germline sequence. During the initial construction of the animal, "chimeras" or "chimeric animals" are generated, in which only a subset of cells have the altered genome. Chimeras are primarily used for breeding purposes in order to generate the desired transgenic animal. Animals having a heterozygous alteration are generated by breeding of chimeras. Male and female heterozygotes are typically bred to generate homozygous animals.

[0030] For embryonic stem (ES) cells, an ES cell line may be employed, or embryonic cells may be obtained freshly from a host, e.g. mouse, rat, guinea pig, etc. A large number of species have been demonstrated to be useful in transgenic models, including fish, e.g. zebrafish (Langenau et al. (2003) Science 299: 887-890); primates, e.g. rhesus monkeys (Chan et al. (2001) Science 291 : 309-312); rabbits (Fan & Watanabe (2003) Pharmacol Ther. 99(3):261-82); livestock, e.g. goats, cows, sheep (Schnieke et al. (1997) Science 278: 2130-2133); and laboratory animals such as rodents, including rats, mice, etc. (Dyck et al. (2003) Trends Biotechnol. 21 (9):394-9).

[0031] Such cells are grown on an appropriate fibroblast-feeder layer or grown in the presence of appropriate growth factors, such as leukemia inhibiting factor (LIF). When ES cells have been transformed, they may be used to produce transgenic animals. After transformation, the cells are plated onto a feeder layer in an appropriate medium. Cells containing the construct may be detected by employing a selective medium. After sufficient time for colonies to grow, they are picked and analyzed for the occurrence of homologous recombination or integration of the construct. Those colonies that are positive may then be used for embryo manipulation and blastocyst injection. Blastocysts are obtained from 4 to 6 week old superovulated females. The ES cells are trypsinized, and the modified cells are injected into the blastocoel of the blastocyst. After injection, the blastocysts are returned to each uterine horn of pseudopregnant females. Females are then allowed to go to term and the resulting litters screened for mutant cells having the construct. By providing for a different phenotype of the blastocyst and the ES cells, chimeric progeny can be readily detected. The chimeric animals are screened for the presence of the modified gene and males and females having the modification are mated to produce homozygous progeny. [0032] The transgenic animal of the present invention comprise an exogenous oncogene coding sequence operably linked to a promoter that is directly or indirectly regulated by a soluble factor, e.g. ecdysone, tetracycline, doxycycline, etc., and that is specifically activated in liver cells, e.g. hepatocytes. Examples of such inducible promoters or other gene regulatory elements include, but are not limited to, tetracycline, metallothionine, ecdysone, and other steroid-responsive promoters, rapamycin responsive promoters, and the like (No, et al (1996) Proc. Natl. Acad. Sci. USA, 93:3346-51; Furth et al. (1994) Proc. Natl. Acad. Sci. USA, 91 :9302-6). Additional control elements that can be used include promoters requiring specific transcription factors such as viral, promoters. By "operably linked" is meant that a DNA sequence and a regulatory sequence(s) are connected in such a way as to permit gene expression when the appropriate molecules, e.g. transcriptional activator proteins, are bound to the regulatory sequence(s).

[0033] A preferred promoter is regulated by a specific repressor. For example, the Lac repressor can be used to repress expression of transcriptional units that include the Lac operator. See, e.g., Carraway et al., J. Bacteriol. 175(13):3972-3980. As another example, derivatives of the Tet repressor protein (TetR) can be used to inhibit or drive expression from promoters have tetracycline response elements derived from the Tn10 tetracycline- resistance operon tet operator sequence. See, e.g., Gossen et al., Proc. Natl. Acad. Sci. USA 89(12):5547-51 (1992); Gossen et al., Science 268(5218): 1766-9 (1995); vectors are available commercially that contain the required repressor and response elements (Clontech Labs., Palo Alto, CA USA).

[0034] More preferably, the oncogene is operably linked to a promoter comprising a tetracycline response element (TRE) upstream of a minimal promoter, which is silent in the absence of the Tet-controlled transactivator, tTA. The oncogene is actively transcribed only when tTA binds the TRE element. The gene encoding tTA is operably linked to a liver specific promoter. Many liver specific promoters are known in the art, and include, for example, albumin promoter; factor VII promoter; fatty acid synthase promoter; alpha- fetoprotein promoter (see PCT/US98/04084); liver activator protein promoter; and the like.

[0035] In some embodiments of the invention, the oncogene is myc, or a member of the myc signaling pathway. Such pathway members include myc, erk1/2, akt, etc. The c-Myc protein is a DNA binding protein that is involved in transcriptional control of gene expression and has been shown to be essential for cell proliferation. Coexpression of Ras with Myc allows the generation of cyclin E-dependent kinase activity and the induction of S phase. The c-Myc protein drives the p27^/φ* protein out of Cdk2/CyclinE complexes, which then facilitates the phosphorylation of p27 and thereby marks the protein for ubiquitination and degradation. Sequences of the MYC oncogene have been highly conserved throughout evolution, from Drosophila to vertebrates, and such sequences are known in the art (for example the human sequence has Genbank accession number V00568; Watt et al. (1983) Nature 303 (5919), 725-728; and mouse accession number P01108).

[0036] Optionally, the cells of the transgenic animal comprise a transgene encoding a detectable marker; e.g. green fluorescent protein (GFP); luciferase (LUC); and the like.

[0037] Expression of the oncogene is inactivated by the presence of the repressor, e.g. doxycycline; tetracycline; etc. Preferably expression is inactivated until the animal reaches maturity. The oncogene is then activated for a period of time sufficient for tumorigenesis. Tumors allowed to form, typically for a period of at least about one week, usually at least about 2 weeks, more usually at least about 6 weeks; and may be at least about 8, 10, 12 weeks, or more. MYC induced tumors most resembled hepatoblastomas, a subtype of human hepatocellular carcinomas. In addition, foci of typical hepatocellular carcinoma were also present. A hallmark shared with human liver cancers was the high degree of invasiveness: the transgenic tumors were locally invasive throughout the liver, frequently associated with malignant peritoneal effusions, and in several cases metastasized into the thoracic cavity with invasion into the parenchyma of the lungs.

[0038] Following tumor formation, the oncogene is inactivated by re-administration of the repressor. The tumor may then be allowed to regress for a suitable period of time, at least about one week, usually at least about 2 weeks; and may be for at least 3 weeks or more. Following inactivation of oncogene expression, there is a rapid and sustained tumor regression. Within about two weeks, tumors undergo almost complete regression, accompanied by cell death, for example as determined by TUNEL assay. During tumor regression, many of the areas of liver tumor transform into normal appearing liver, including hepatocytes and biliary cells associated with normal liver structures, e.g. hepatic sinusoids; bile canaliculi; duct-like structures; etc. The differentiated cells are negative for the proliferative cell marker, Ki-67. Animals in this latent period show sustained but reversible tumor dormancy, provide a model for latent stage tumors.

[0039] The transgenic animals of the invention and cells derived therefrom comprise the above-described oncogenic transgene system. The liver cells may actively express the oncogene, and thus have a tumor cell phenotype; or may be inactivated for oncogene expression, either prior to, or following a period of tumor growth. In the latter, the liver stem cells can have latent tumorigenic properties.

LIVER STEM CELLS

[0040] Liver stem cells having latent tumorigenic properties are derived by the methods described above. Such cells may be isolated by expression of cell surface markers, cell cloning in culture, etc. Alternatively, suspensions of the tumor cells can be transferred to a host animal, e.g. a SCID mouse or other immunocompromised recipient, and brought into latent phase in the host animal. In such in vivo transfers, markers that distinguish the recipient from the donor may be used for selection, including, for example, fluorescent transgenic markers.

[0041] Markers for identification of active tumor stage include expression of the embryonic tumor cell marker that characterizes hepatocellular carcinoma, alpha-fetoprotein (AFP). In the latent stage, the liver stem cells may express cytokeratin 19 (CK-19), while differentiated cells express the mature liver markers CEA and Cytokeratin 8 (CK-8), associated with the formation of bile canaliculi and bile-duct-like cells.

[0042] The functional features of a liver cancer stem cell are that they are tumorigenic, they give rise to additional tumorigenic cells ("self-renew"), and they can give rise to non- tumorigenic tumor cells ("differentiation"). In the latent stage, these cells give rise to normal, differentiated hepatic cells. The developmental potential of liver stem cells can be assessed by functional and phenotypic criteria. Functionally, hepatocytes are characterized by their ability to complement FAH deficiency, and by the expression of liver specific proteins, including albumin, alpha-1 -antitrypsin, alpha fetoprotein, etc. Hepatocytes are also functionally characterized by their ability to be infected by hepatitis viruses, e.g. Hepatitis A (HAV); Hepatitis B (HBV), hepatitis C (HCV); Hepatitis D (HDV); Hepatitis E

[0043] Isolated populations of such cells are useful in identifying the genes and proteins expressed by liver cancer stem cells, in order to identify proteins whose function is necessary for tumorigenesis and which represent novel drug targets. The cells can be used to screen potential therapeutic compounds. Markers of the cells are identified, and used to more effectively diagnose the presence of malignant cells. To effectively treat cancer and achieve higher cure rates, anti-cancer therapies are directed against cancer stem cells. Since current therapies are directed against the bulk population, they may be ineffective at eradicating liver cancer stem cells. The identification of cancer stem cells permits the specific targeting of therapeutic agents to this cell population, resulting in more effective cancer treatments. Markers characterized with the cells of the invention may find use in the identification of stem cells from other species, e.g. in human tumors, and the like.

[0044] Cells of the invention are derived from transgenic animals are described above. In the latent stage, these cells are characterized by the absence of AFP expression, and by the presence of CK19 expression. The cells may be further assessed for the expression of other known liver cell markers; or may be used in immunization methods to generate antibodies to novel markers. Cell surface markers can be recognized by reagents that specifically bind to the cell surface markers. For example, proteins, carbohydrates, or lipids on the surfaces of liver cancer stem cells can be immunologically recognized by antibodies specific for the particular protein or carbohydrate. The set of markers present on the cell surfaces of liver cancer stem cells and absent from the cell surfaces of these cells is characteristic for liver cancer stem cells. Therefore, liver cancer stem cells can be selected by positive and negative selection of cell surface markers.

[0045] By selecting for phenotypic characteristics among the cells obtained from a liver cancer, liver cancer stem cells can be isolated from any animal liver cancer, particularly any mammalian liver cancer. It will be appreciated that, taking into consideration factors such as a binding affinities, that antibodies that recognize species-specific varieties of markers are used to enrich for and select liver cancer stem cells.

[0046] Candidate liver stem cells may be isolated from a graft, or are separated from a complex mixture of cells by techniques that enrich for cells having a characteristic of interest. The tumor cells can also be lebeled, e.g. with a fluorescent tag, prior to inactivation of the oncogene, thereby providing a means of identifying the stem cells. Affinity separation can be useful, for example where the liver stem cells comprise a fluorescent marker, or can otherwise be differentiated from a host cell. Techniques for affinity separation may include magnetic separation, using antibody-coated magnetic beads, affinity chromatography, cytotoxic agents joined to a monoclonal antibody or used in conjunction with a monoclonal antibody, e.g. complement and cytotoxins, and "panning" with antibody attached to a solid matrix, e.g. plate, or other convenient technique. Techniques providing accurate separation include fluorescence activated cell sorters, which can have varying degrees of sophistication, such as multiple color channels, low angle and obtuse light scattering detecting channels, impedance channels, etc. The cells may be selected against dead cells by employing dyes associated with dead cells (propidium iodide, 7-AAD). Any technique may be employed which is not unduly detrimental to the viability of the selected cells.

[0047] Of particular interest is the use of antibodies as affinity reagents. Conveniently, these antibodies are conjugated with a label for use in separation. Labels include magnetic beads, which allow for direct separation, biotin, which can be removed with avidin or streptavidin bound to a support, fluorochromes, which can be used with a fluorescence activated cell sorter, or the like, to allow for ease of separation of the particular cell type. Fluorochromes that find use include phycobiliproteins, e.g. phycoerythrin and allophycocyanins, fluorescein and Texas red. Frequently each antibody is labeled with a different fluorochrome, to permit independent sorting for each marker.

[0048] The labeled cells are then separated as to the phenotype described above, e.g. by

FACS, MACS, etc. The separated cells may be collected in any appropriate medium that maintains the viability of the cells, usually having a cushion of serum at the bottom of the collection tube. Various media are commercially available and may be used according to the nature of the cells, including dMEM, HBSS, dPBS, Iscoves medium, Ultra medium, DEM-F12, HCM bullet medium, etc., frequently supplemented with fetal calf serum or serum replacement.

[0049] Compositions highly enriched for liver stem cells are achieved in this manner. The subject population will be at or about 50% or more of the cell composition, and usually at or about 90% or more of the cell composition, and may be as much as about 95% or more of the live cell population. The enriched cell population may be used immediately, or may be frozen at liquid nitrogen temperatures and stored for long periods of time, being thawed and capable of being reused. The cells will usually be stored in 10% DMSO, 50% FCS, 40% medium. Once thawed, the cells may be expanded by use of growth factors and/or stromal cells for proliferation and differentiation.

[0050] The present methods are useful in the development of an in vitro or in vivo model for hepatocyte functions and are also useful in experimentation on gene therapy and for artificial organ construction. The developing hepatocytes serve as a valuable source of novel growth factors and pharmaceuticals and for the production of viruses or vaccines (e.g., hepatitis viruses), as well as for the study of liver parasites or of parasites having a stage of development in the liver, e.g. malarial organisms), for in vitro toxicity and metabolism testing of drugs and industrial compounds, for gene therapy experimentation (since the liver is the largest vascular organ of the body), for the construction of artificial transplantable livers, and for liver mutagenesis and carcinogenesis studies.

[0051] The enriched cell population may be grown in vitro under various culture conditions.

Culture medium may be liquid or semi-solid, e.g. containing agar, collagen, methylcellulose, etc. The cell population may be conveniently suspended in an appropriate nutrient medium, such as Iscove's modified DMEM or Ultra medium, normally supplemented with fetal calf serum (about 5-10%), ITS and antibiotics, e.g. penicillin and streptomycin.

[0052] The subject cells may be grown in a co-culture with feeder layer cells. Stromal cells suitable for use as feeder layers include bone marrow stromal cells, fibroblasts, etc. These cell layers provide non-defined components to the medium and may restrain the differentiation of the pluripotent cells. Culture in the presence of feeder layers is particularly useful for clonal culture, i.e. where a single progenitor cell is expanded to a population.

[0053] The cells may be grown in the absence or presence of the repressor molecule, e.g. tetracycline, doxycycline, etc. Cells grown in the absence of. the repressor will typically have a tumor cell phenotype, while cells grown in the presence of the repressor will have a stem cell phenotype. By expressing the oncogene and then inactivating the oncogene one can enrich for the stem cell in the total population.

[0054] Functional assays may be performed using in vitro cultured cells, particularly clonogenic cultures of cells. For example, cultured cells may be assessed for their ability to express liver specific proteins, including albumin and alpha-1 antitrypsin. Expression may utilize any convenient format, including RT-PCR, ELISA for presence of the protein in culture supernatants, etc. Cultured cells may also be assessed for their ability to express bile duct proteins, e.g. CK19.

[0055] The stem cells may also be assessed for their ability to give rise to residual disease after the incogene is inactivated, e.g. through secondary mutations, escape from the oncogene requirement, and the like.

[0056] The culture may contain growth factors to which the cells are responsive. Growth factors, as defined herein, are molecules capable of promoting survival, growth and/or differentiation of cells, either in culture or in the intact tissue, through specific effects on a transmembrane receptor. Growth factors include polypeptides and non-polypeptide factors. Specific growth factors that may be used in culturing the subject cells include but are not limited to hepatocyte growth factor/scatter factor (HGF), EGF, TGFα, acidic FGF (see Block et al; J. Biol Chem, 1996 132:1133-1149). The specific culture conditions are chosen to achieve a particular purpose, i.e. maintenance of stem cell activity, etc. In addition to, or instead of growth factors, the subject cells may be grown in a co-culture with stromal or feeder layer cells.

[0057] The cultured cells may be used in a variety of ways. For example, the nutrient medium, which is a conditioned medium, may be isolated at various stages and the components analyzed. Separation can be achieved with HPLC, reversed phase-HPLC, gel electrophoresis, isoelectric focusing, dialysis, or other non-degradative techniques, which allow for separation by molecular weight, molecular volume, charge, combinations thereof, or the like. One or more of these techniques may be combined to enrich further for specific fractions that promote progenitor cell activity.

[0058] The liver stem cells may be used in conjunction with a culture system in the isolation and evaluation of factors associated with the differentiation and maturation of hepatocytes. Thus, the cells may be used in assays to determine the activity of media, such as conditioned media, evaluate fluids for growth factor activity, involvement with formation of specific structures, or the like. Cultures may also be used as a means of processing drugs and other compounds, to determine the effect of liver metabolism on an agent of interest. For example, the product of liver metabolism may be isolated and tested for toxicity and efficacy.

[0059] Additional genes may be introduced into the cells prior to culture or transplantation for a variety of purposes, e.g. prevent or reduce susceptibility to infection, replace genes having a loss of function mutation, etc. Alternatively, vectors are introduced that express antisense mRNA or ribozymes, thereby blocking expression of an undesired gene. Various techniques known in the art may be used to transfect the target cells, e.g. electroporation, calcium precipitated DNA, fusion, transfection, lipofection, infection and the like. The particular manner in which the DNA is introduced is not critical to the practice of the invention.

[0060] Many vectors useful for transferring exogenous genes into mammalian cells are available. The vectors may be episomal, e.g. plasmids, virus derived vectors such cytomegalovirus, adenovirus, etc., or may be integrated into the target cell genome, through homologous recombination or random integration, e.g. retrovirus derived vectors such MMLV, HIV-1 , ALV, etc. For examples of progenitor and stem cell genetic alteration, see Svendsen et al. (1999) Trends Neurosci. 22(8):357-64; Krawetz et al. (1999) Gene 234(1): 1-9; Pellegrini et al. Med Biol Enq Comout. 36(6):778-90; and Alison (1998) Curr Qpin Cell Biol. 10(6)710-5.

SCREENING METHODS

[0061] One of the major problems in identifying new cancer therapeutic agents is determining clinically important drug targets and therapies. In a heterogeneous tumor population, it is important to determine which agents act on the stem cells. Purification (enrichment or isolation) of subsets of cancer cells from a liver cancer allows identification of the genes necessary for tumor proliferation and drug resistance. The identification of biological pathways is another important part of the drug discovery process. Biological pathways in cancer stem cells, particularly pathways that originate at a drug target can be identified for use.

[0062] The liver stem cell of the invention is particularly useful in the drug development process because stem cells provide a limited and enriched set of targets for drug development. One of the most important steps in rational drug design is the identification of a target, the molecule with which the drug itself interacts. Frequently, the target will be a receptor on or in a tumorigenic cancer stem cell.

[0063] The subject cells are useful for in vitro assays and screening to detect agents that affect liver stem cells and hepatocytes generated from the liver stem cells. A wide variety of assays may be used for this purpose, including toxicology testing, immunoassays for protein binding; determination of cell growth, differentiation and functional activity; production of hormones; and the like.

[0064] In screening assays for biologically active agents, -viruses, etc., the subject cells, usually a culture comprising the subject cells, is contacted with the agent of interest, and the effect of the agent assessed by monitoring output parameters, such as expression of markers, cell viability, and the like. The cells may be freshly isolated, cultured, genetically altered as described above, or the like. The cells may be environmentally induced variants of clonal cultures: e.g. split into independent cultures and grown under distinct conditions, for example with or without virus; in the presence or absence of other cytokines or combinations thereof. The manner in which cells respond to an agent, particularly a pharmacologic agent, including the timing of responses, is an important reflection of the physiologic state of the cell.

[0065] Parameters are quantifiable components of cells, particularly components that can be accurately measured, desirably in a high throughput system. A parameter can be any cell component or cell product including cell surface determinant, receptor, protein or conformational or posttranslational modification thereof, lipid, carbohydrate, organic or inorganic molecule, nucleic acid, e.g. mRNA, DNA, etc. or a portion derived from such a cell component or combinations thereof. While most parameters will provide a quantitative readout, in some instances a semi-quantitative or qualitative result will be acceptable. Readouts may include a single determined value, or may include mean, median value or the variance, etc. Characteristically a range of parameter readout values will be obtained for each parameter from a multiplicity of the same assays. Variability is expected and a range of values for each of the set of test parameters will be obtained using standard statistical methods with a common statistical method used to provide single values.

[0066] Agents of interest for screening include known and unknown compounds that encompass numerous chemical classes, primarily organic molecules, which may include organometallic molecules, inorganic molecules, genetic sequences, etc. An important aspect of the invention is to evaluate candidate drugs, including toxicity testing, to test the effect of hepatic viruses, e.g. Hepatitis A, B, C, D, E viruses; antiviral agents; and the like.

[0067] In addition to complex biological agents such as viruses, candidate agents include organic molecules comprising functional groups necessary for structural interactions, particularly hydrogen bonding, and typically include at least an amine, carbonyl, hydroxyl or carboxyl group, frequently at least two of the functional chemical groups. The candidate agents often comprise cyclical carbon or heterocyclic structures and/or aromatic or polyaromatic structures substituted with one or more of the above functional groups. Candidate agents are also found among biomolecules, including peptides, polynucleotides, saccharides, fatty acids, steroids, purines, pyrimidines, derivatives, structural analogs or combinations thereof.

[0068] Included are pharmacologically active drugs, genetically active molecules, etc.

Compounds of interest include chemotherapeutic agents, hormones or hormone antagonists, etc. Exemplary of pharmaceutical agents suitable for this invention are those described in, "The Pharmacological Basis of Therapeutics," Goodman and Gilman, McGraw-Hill, New York, New York, (1996), Ninth edition, under the sections: Water, Salts and Ions; Drugs Affecting Renal Function and Electrolyte Metabolism; Drugs Affecting Gastrointestinal Function; Chemotherapy of Microbial Diseases; Chemotherapy of

Neoplastic Diseases; Drugs Acting on Blood-Forming organs; Hormones and Hormone Antagonists; Vitamins, Dermatology; and Toxicology, all incorporated herein by reference. Also included are toxins, and biological and chemical warfare agents, for example see Somani, S.M. (Ed.), "Chemical Warfare Agents," Academic Press, New York, 1992).

[0069] Test compounds include all of the classes of molecules described above, and may further comprise samples of unknown content. Of interest are complex mixtures of naturally occurring compounds derived from natural sources such as plants. While many samples will comprise compounds in solution, solid samples that can be dissolved in a suitable solvent may also be assayed. Samples of interest include environmental samples, e.g. ground water, sea water, mining waste, etc.; biological samples, e.g. lysates prepared from crops, tissue samples, etc.; manufacturing samples, e.g. time course during preparation of pharmaceuticals; as well as libraries of compounds prepared for analysis; and the like. Samples of interest include compounds being assessed for potential therapeutic value, i.e. drug candidates.

[0070] The term samples also includes the fluids described above to which additional components have been added, for example components that affect the ionic strength, pH, total protein concentration, etc. In addition, the samples may be treated to achieve at least partial fractionation or concentration. Biological samples may be stored if care is taken to reduce degradation of the compound, e.g. under nitrogen, frozen, or a combination thereof. The volume of sample used is sufficient to allow for measurable detection; usually from about 0.1 μl to 1 ml of a biological sample is sufficient.

[0071] Compounds, including candidate agents, are obtained from a wide variety of sources including libraries of synthetic or natural compounds. For example, numerous means are available for random and directed synthesis of a wide variety of organic compounds, including biomolecules, including expression of randomized oligonucleotides and oligopeptides. Alternatively, libraries of natural compounds in the form of bacterial, fungal, plant and animal extracts are available or readily produced. Additionally, natural or synthetically produced libraries and compounds are readily modified through conventional chemical, physical and biochemical means, and may be used to produce combinatorial libraries. Known pharmacological agents may be subjected to directed or random chemical modifications, such as acylation, alkylation, esterificationj amidification, etc. to produce structural analogs.-

[0072] Agents are screened for biological activity by adding the agent to at least one and usually a plurality of cell samples, usually in conjunction with cells lacking the agent. The change in parameters in response to the agent is measured, and the result evaluated by comparison to reference cultures, e.g. in the presence and absence of the agent, obtained with other agents, etc. [0073] The agents are conveniently added in solution, or readily soluble form, to the medium of cells in culture. The agents may be added in a flow-through system, as a stream, intermittent or continuous, or alternatively, adding a bolus of the compound, singly or incrementally, to an otherwise static solution. In a flow-through system, two fluids are used, where one is a physiologically neutral solution, and the other is the same solution with the test compound added. The first fluid is passed over the cells, followed by the second. In a single solution method, a bolus of the test compound is added to the volume of medium surrounding the cells. The overall concentrations of the components of the culture medium should not change significantly with the addition of the bolus, or between the two solutions in a flow through method.

[0074] Preferred agent formulations do not include additional components, such as preservatives, that may have a significant effect on the overall formulation. Thus preferred formulations consist essentially of a biologically active compound and a physiologically acceptable carrier, e.g. water, ethanol, DMSO, etc. However, if a compound is liquid without a solvent, the formulation may consist essentially of the compound itself.

[0075] A plurality of assays may be run in parallel with different agent concentrations to obtain a differential response to the various concentrations. As known in the art, determining the effective concentration of an agent typically uses a range of concentrations resulting from 1:10, or other log scale, dilutions. The concentrations may be further refined with a second series of dilutions, if necessary. Typically, one of these concentrations serves as a negative control, i.e. at zero concentration or below the level of detection of the agent or at or below the concentration of agent that does not give a detectable change in the phenotype.

[0076] Techniques for drug screening include high throughput screening of compounds. In this method, large numbers of different small test compounds are synthesized on a solid substrate, such as plastic pins or some other surface. The test compounds are reacted with liver stem cells, or portions thereof, and washed. Bound liver stem cells are then detected by methods well known in the art, using commercially available machinery and methods, for example, the Automated Assay Optimization (AAO) software platforms (Beckman, USA) that interface with liquid handlers to enable direct statistical analysis that optimizes the assays; modular robotic systems, liquid handling systems, readers, and incubators, and the like, which enable a wide range of discovery applications, including HTS, ultra HTS, and high-speed plate preparation.

[0077] For any of these machines and methods, the assays measure a response in the target cells that provides detectable evidence that the test compound is efficacious in its desired effect. The detectable signal is compared to control cells and the detectable signal identified by subtraction analysis. The relative abundance of the differences between the "targeted" and "untargeted" aliquots can be simultaneously compared using a "subtraction" analysis (differential analysis) technique such as differential display, representational difference analysis (RDA), GEM-Gene Expression Microarrays (U.S. Pat. No. 5,545,531), suppressive subtraction hybridization (SSH) and direct sequencing (PCT patent application WO 96/17957). The subtraction analysis can include the methods of differential display, representational differential analysis (RDA), suppressive subtraction hybridization (SSH), serial analysis of gene expression (SAGE), gene expression microarray (GEM), nucleic acid chip technology, or direct sequencing.

[0078] In one set of methods, drugs are screened to determine the binding of test compounds to receptors, in which the binding activates a cell's biological pathway to cause expression of reporter polypeptides. Frequently the reporter polypeptides are coded for on recombinant polypeptides, in which the coding polynucleotide is in operable linkage with a promoter. The detectable signal can be fluorescence, absorbance, or luminescence, depending on the reporter polypeptide, which can be, for example, luciferase (firefly luciferase, Vibrio fisceri luciferase, or Xenorhabdus luminescens luciferase), green fluorescent protein, green fluorescent protein variant, chloramphenicol acetyltransferase, β- glucuronidase, β-galactosidase, neomycin phosphotransferase, guanine xanthine phosphoribosyltransferase, thyridine kinase, β-lactamase, alkaline phosphatase, invertase, amylase (for yeast based assays) human growth hormone (for activity based assays). The fluorescent detectable signal can be fluorescence resonance energy transfer (FRET), bioluminescence resonance energy transfer (BRET), time-resolved fluorescence (TRF) or fluorescence polarization (FP). Where appropriate, the detectable signal is detected by a machine such as a fluorometer, luminometer, fluorescence microplate reader, dual- monochromator microplate spectrofluorometer, spectrophotometer, confocal microscope (laser scanner), or a charge-coupled device (CCD). The detectable signal is determined by comparing the amount of signal produced when the reporter polypeptide is expressed in the tumor stem cell with the signal produced when the reporter polypeptide is not expressed in the tumor stem cell.

EXPERIMENTAL

[0079] The following examples are put forth so as to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the subject invention, and are not intended to limit the scope of what is regarded as the invention. Efforts have been made to ensure accuracy with respect to the numbers used (e.g. amounts, temperature, concentrations, etc.) but some experimental errors and deviations should be allowed for. Unless otherwise indicated, parts are parts by weight, molecular weight is average molecular weight, temperature is in degrees centigrade; and pressure is at or near atmospheric.

[0080] All publications and patent applications cited in this specification are herein incorporated by reference as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference. The citation of any publication is for its disclosure prior to the filing date and should not be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior invention.

[0081] It is to be understood that this invention is not limited to the particular methodology, protocols, cell lines, animal species or genera, and reagents described, as such may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention which will be limited only by the appended claims.

[0082] As used herein the singular forms "a", "and", and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a cell" includes a plurality of such cells and reference to "the protein" includes reference to one or more proteins and equivalents thereof known to those skilled in the art, and so forth. All technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art to which this invention belongs unless clearly indicated otherwise.

Example 1 MYC Inactivation Uncovers Stem Cell Properties and Induces a State of Tumor Dormancy in Hepatocellular Cancer

[0083] To address the possibility that the inactivation of MYC is effective in the treatment of liver cancer, we have developed a novel conditional transgenic model whereby we can regulate MYC expression in murine hepatocytes. We have used this model system to demonstrate that upon MYC inactivation, malignant liver tumors undergo sustained tumor regression. MYC inactivation did not result in the elimination of all the tumor cells, but rather a subset of cells retained their hepatic stem cell potential for self renewal and proliferation. Upon MYC inactivation these cells differentiated into normal hepatocytes, lost expression of cancer markers, gained expression of mature hepatic markers, some cells gained the expression of liver stem cell markers and formed normal liver parenchymal structures such as hepatic sinusoids, bile canaliculi and duct-like structures. Although these differentiated tumor cells appeared to be normal hepatocytes, their malignant potential remained dormant, and reactivation of MYC even months later was sufficient to immediately restore their neoplastic properties. Thus, MYC inactivation in liver tumors can produce a sustained state of tumor dormancy.

Results

[0084] Sustained Regression of Liver Cancers upon MYC /nactivation. To evaluate the role of an oncogene in the initiation and maintenance of hepatocellular cancers, we utilized the tetracycline regulatory system (Tet System; Kistner et al. (1996) Proc Natl Acad Sci U S A 93, 10933-8) to generate transgenic mice that conditionally express the MYC proto- oncogene in their hepatocytes (Fig. 1a). Previously, we have described the generation of transgenic lines that contain the human c-MYC cDNA under the transcriptional regulation of the tetracycline response element (TRE-MYC) (Felsher & Bishop (1999) MoI Cell 4, 199- 207. To conditionally express this MYC transgene in murine hepatocytes, we crossed our TRE-MYC line with a transgenic line, LAP-tTA, in which a transgene composed of the liver activator protein promoter drives expression of the tetracycline transactivating protein (tTA) in hepatocytes (Kistner ef al., supra.) Progeny possessing both transgenes, LAP-tTA and TRE-MYC, expressed the MYC oncogene in the liver, but not the parental lines nor double- transgenic animals treated with doxycycline (Fig. 1 b) . Hence, we have generated transgenic mice in which the MYC oncogene can be conditionally expressed in hepatocytes.

[0085] MYC overexpression in murine hepatocytes has previously been shown to be sufficient to induce tumorigenesis. However, in these model systems MYC was expressed constitutively in the liver throughout the entire development of the mouse. In humans, however, liver cancers generally arise in adults. To address if MYC expression is sufficient for tumorigenesis when overexpressed during adult age, we activated MYC transgene expression in 3-week-old mice by removing doxycycline treatment. Subsequently, all transgenic mice that overexpressed MYC succumbed to liver tumors (Fig. 1c, 2c, 2d), with a mean latency of tumor onset of 12 weeks, whereas transgenic mice continuously treated with doxycycline remained free of disease (Fig. 1c, 2a, 2b). Tumors produced high levels of MYC protein, as measured by immunohistochemistry whereas hepatocytes from mice continuously treated with doxycycline did not express detectable MYC (Figure 8). Thus in this LAP-tTA/tet off MYC conditional transgenic mouse model MYC overexpression in adult mice can reproducibly induce liver cancer.

[0086] Upon histological analysis the MYC induced tumors most resembled hepatoblastomas, a subtype of human hepatocellular carcinomas. In addition, foci of typical hepatocellular carcinoma were also present. A hallmark shared with human liver cancers was the high degree of invasiveness: the transgenic tumors were locally invasive throughout the liver, frequently associated with malignant peritoneal effusions, and in several cases metastasized into the thoracic cavity with invasion into the parenchyma of the lungs. We confirmed that the tumors had a cancer phenotype as they were readily transplantable into SCID mice, as described below. Thus, in our model system, the overexpression of MYC in adult mice is sufficient to cause liver cancers with many of the characteristic features of human liver cancers.

[0087] MYC-induced tumorigenesis is reversible upon cessation of MYC overexpression in hematopoietic tumors, osteogenic sarcoma, breast adenocarcinoma and pancreatic islet cell tumors, but up to 30% of the conditional MYC hematopoietic tumors escape the requirement for sustained MYC expression. Liver cancer is particularly refractory to therapeutic intervention. It was anticipated that oncogene inactivation in a liver tumor would be even less effective in causing tumor regression than in other t/pes of cancer. Surprisingly, over fifty transgenic mice moribund with liver tumors exhibited rapid and sustained tumor regression when treated with doxycycline to inactivate MYC transgene expression (Fig. 2g). Upon gross examination 4 weeks post MYC inactivation, there was no evidence of the tumor persistence and gross and microscopic liver morphology had been restored to normal (Fig. 2e, 2f). These results suggest that targeted inactivation of MYC alone may effectively induce sustained regression of hepatocellular neoplasms.

[0088] Regression of Liver Tumors upon MYC Inactivation. The liver tumors rapidly regressed when the MYC transgene was turned off. Hematopoietic tumors undergo cell cycle arrest, differentiation and apoptosis upon MYC inactivation. In contrast, osteogenic sarcomas responded with cell cycle arrest and differentiation, but without significant apoptosis. Therefore, the consequences of MYC inactivation appear to be highly dependent upon the type of tumor. To explore the long-term consequences of MYC transgene inactivation in liver tumors, we treated mice that were moribund, with doxycycline. Within the subsequent two weeks, tumors underwent almost complete regression, accompanied by cell death as determined by TUNEL assay (Fig. 3b). During tumor regression, many of the areas of liver tumor appeared to transform into normal appearing liver. However, from these experiments we could not distinguish whether the tumor cells were differentiating into normal hepatocytes or whether after tumor regression normal hepatocytes regenerated the liver.

[0089] To unambiguously evaluate the fate of tumor cells upon MYC inactivation, we developed a collagenase method for generating suspensions of tumor cells that could be transplanted subcutaneously into SCID mice (Fig. 4a). When the tumors had reached a diameter of 1 to 2 cms, MYC was inactivated by treating mice with doxycycline. By Western blot analysis, we confirmed that one day after doxycycline treatment (day 1), MYC expression had decreased to 30% of the previous levels and after 4 days was almost undetectable (Fig. 3a). Also, found that the embryonic tumor cell marker that characterizes hepatocellular carcinoma, AFP, was reduced in protein expression to 50% of previous levels on day 1 and abolished on day 4. On visual inspection, tumors began to regress within the first 5 days after MYC inactivation and completely regressed within 30 days, with a scar persisting at the site of initial tumor cell inoculation (Fig. 4b). Thus, transplanted liver tumors also regressed upon MYC inactivation.

[0090] Differentiation of Liver Tumor into Normal Liver after MYC Inactivatfon. Careful examination the histology of the site of initial tumor inoculation after MYC inactivation, revealed normal appearing hepatocytes (Fig. 4c, 4d); these were associated with normal liver structures including hepatic sinusoids, bile canaliculi and duct-like structu res could be detected (see also below, Fig. 6). Thus, normalization of MYC expressiora appears to initiate the differentiation of some of the tumor cells into normal hepatocytes. A trivial explanation for these results could be that we had transplanted some normal liver hepatocytes along with liver tumor cells into the SCID mice. We ruled out this possibility by demonstrating identical results for multiple tumors even when they had been serially transplanted in SCID mice. Identical results were observed from a tumor derived from a distant lung metastasis. Moreover, cell suspensions of normal hepatocytes were not capable of persisting beyond 7 days when transplanted into SCID mice, as described below. Therefore, we conclude that upon normalization of MYC levels hepatocellular carcinomas can differentiate into normal liver.

[0091] Further biochemical and immunohistochemical evidence substantiated that upon

MYC inactivation tumor cells were differentiating into normal hepatocytes and forming normal liver (Fig. 5). The differentiated tumor cells were Ki-67 negative suggesting that they were no longer proliferating. They had also lost expression of the liver cancer tumor marker, AFP, consistent with our Western blot analysis, as described above. Instead they now were positive for the mature liver markers CEA and Cytokeratin 8 (CK-8), associated with the formation of bile canaliculi and bile-duct-like cells. Moreover, rare cells acquired the liver stem cell marker, Cytokeratin 19 (CK-19) (see below Fig. 7). We conclude that liver cancers retain stem cell properties. Upon MYC inactivation, a proportion of these liver tumor cells are able to differentiate into normal hepatocytes and ductal cel ls, and form different liver structures such as sinusoids, bile canaliculi and bile duct-like structures.

[0092] Tumor Dormancy upon MYC Inactivation. Tumor cells derived from MY^'C transgenic mice differentiate and permanently lose their neoplastic properties upon inactivation of MYC. To address this possibility here, we examined the consequences of the reactivation of MYC expression in transgenic mice with liver tumors in which MYC had been inactivated for at least 30 days. Within two weeks of MYC reactivation, we observed gross evidence for tumor regrowth. These tumors were found to have identical histology to the original transplanted tumor. These tumors were still dependent on MYC transgene expression and inactivation of the transgene with doxycycline resulted in tumor regression. Hence, while MYC inactivation can reverse liver tumorigenesis; however, MYC reactivation appears to immediately restore the neoplastic properties of the tumor cell population some or all of which have differentiated.

[0093] To better examine the consequences of MYC inactivation and reactivation in vivo, we generated liver tumors labeled with luciferase and utilized in vivo bioluminescent imaging (BLI), to visualize tumor growth and response to therapy. We generated luciferase- expressing hepatocytes by crossing the LAP-tTA/tet-o-MYC mice with CMV-GFP-LUC mice. Liver tumors derived from triple transgenic mice stably expressed luciferase when transplanted into SCID mice. The number of tumor cells inoculated and the size of tumor correlated well with the light emitted by luciferase activity, allowing us to detect as few as 1000 tumor cells, and quantitatively measure tumor cell number in the recipient animals (Figure 9, 10, 11). Using BLI, we were able to follow tumor regression upon MYC transgene inactivation in vivo in real time. In all cases examined, even 8 months after MYC inactivation, we could continue to detect the presence of luciferase expressing cells even when the tumor was not grossly detectable (Fig. 6). In contrast, when normal hepatocytes that stably expressed luciferase were transplanted into SCID mice no signal above background was detected after 10 days of injection (Figure 12). Moreover, when we restored MYC activation, the tumor cells immediately regained the capacity for proliferation. Resumption of doxycycline treatment resulted again in tumor regression. We conclude that although MYC inactivation results in the loss of the neoplastic properties of liver tumor cells, reactivation of MYC can rapidly restore the neoplastic properties in these differentiated tumor cells. Thus MYC inactivation in liver cancers can induce a state of sustained but reversible tumor dormancy.

[0094] We have demonstrated that highly invasive and malignant liver cancers exhibit rapid and sustained tumor regression upon MYC inactivation. The vast majority of tumor cells undergo cell death, however, some of the tumor cells retain stem cell properties and differentiate into normal liver (Fig. 7). Moreover, the reactivation of MYC was found to restore their neoplastic properties. We conclude that MYC inactivation in liver tumors can result in the differentiation of tumor cells into normal liver, but these apparently normal cells remain in a state of tumor dormancy. These results have general implications for how oncogenes initiate and sustain tumorigenesis, and for the therapy of cancer.

[0095] MYC-induced lymphoma, leukemia, skin papillomas, pancreatic islet cell cancer, breast adenocarcinoma, and osteogenic sarcoma are reversible upon MYC inactivation. However, these hematopoietic and epithelial derived tumors are generally responsive to conventional chemotherapy, radiation therapy and/or hormonal therapies. In marked contrast, the diagnosis of invasive liver cancer portends a dismal prognosis and is not amenable to existing therapeutic modalities. Thus, our results suggest that the targeted inactivation of the MYC oncogene may be an effective strategy for the treatment of some liver cancers. Moreover, in contrast to what we have observed with hematopoietic tumors, only after several serial transplantations did the liver tumors rarely relapse after prolonged inactivation MYC. Hence, MYC inactivation alone appears to be sufficient to induce sustained regression of liver tumors.

[0096] Not all liver cancer cells were eliminated after MYC inactivation. A fraction of the tumor cells possess hepatic stem cell properties (see Thorgeirsson & Grisham (2003) Semin Liver Dis 23, 303-12 for a review). When released from MYC overexpression, these cells were able to resume a physiologic program and differentiate into normal appearing hepatocytes associated with the loss of expression of the immature tumor marker, AFP, and the gain in expression of two normal liver markers, CK-8 and CEA, with rare cells expressing the liver stem cell marker CK19. Moreover, these now differentiated tumor cells formed normal hepatic structures including hepatic sinusoids, bile canaliculi and duct-like structures. We were able to demonstrate that liver tumors differentiated into normal liver even after multiple serial passages. We conclude that normalization of MYC expression in established liver tumors can induce their differentiation into normal liver.

[0097] Several observations confirm that these normal appearing hepatocytes were derived from tumor cells and not from the fortuitous transplantation of rare normal liver cells along with tumor. We found that tumors serially transplanted at least three times or tumors derived from a distant lung metastasis were also capable of differentiating into normal liver tissue upon MYC inactivation. In contrast, normal hepatocytes could not be detected after subcutaneous transplantation into SCID mice. Thus, even highly malignant liver tumors retain an intrinsic capacity to develop into normal hepatocytes and form normal hepatic tissue structures.

[0098] Although inactivation of MYC resulted in sustained tumor regression and differentiation of tumor cells into normal appearing liver tissue, MYC reactivation was capable of immediately restoring the neoplastic properties in at least some of the remaining cell population. We provide the first reported example of oncogene inactivation inducing tumor dormancy. Clinically, it is frequently observed that after therapy, tumors exist in a latent state, and even after many years are still capable of reverting back to a neoplastic state. Experimentally, tumor dormancy has been induced via the suppression of angiogenesis, as well as after treatment with anti-idiotypic antibodies. The mechanism by which tumor cells remain dormant or become reactivated is unknown. Our results demonstrate that some of the liver tumor cells retain stem cell properties and thus exist as cancer stem cells (see Reya et al. (2001) Nature 414, 105-11 ; Al-Hajj et al. (2004) Current Opinion in Genetics and Development 14, 43-47). In general tumor dormancy may reflect changes in epigenetic regulation associated with the differentiation of tumor cells with stem cell features. [0099] MYC inactivation in liver tumor resulted in the death of most of the tumors cells, but some of the tumor cells appeared to have retained stem cell properties and were now able to differentiate into normal liver. Some of these differentiated liver cells upon MYC reactivation apparently are capable of restoring their neoplastic properties. Hence, for liver cancer, MYC induced tumorigenesis is both reversible as well as restorable. Since the liver has the ability to rapidly regenerate itself, liver tumors may similarly retain the capacity to differentiate into normal liver cells to retain their stem cell features, becoming dormant upon MYC inactivation but regaining their neoplastic properties upon MYC reactivation.

[00100] In support of the notion that liver tumors retain their stem cell properties, we showed that upon MYC inactivation the tumor cells differentiated into normal liver parenchymal structures forming hepatic sinusoids and bile ducts. Thus, perhaps the most important implication of our results is that we have demonstrated that liver tumors can retain their stem cell characteristics that can be uncovered through the abrogation of MYC oncogene activation resulting in the differentiation of tumors into mature liver tissue. Our observations are consistent with many recent reports that cancers frequently consist of cellular subpopulations, some of which have retained stem cell properties and most of which are derived from these cells. MYC appears to be an example of an oncogene that sustains malignant transformation by malignantly transforming stem cells, that retain their capacity for cellular differentiation, that is revealed upon termination of oncogene activation.

[00101] A model is provided for how MYC activation induces and sustains tumorigenesis in hepatocytes (Fig. 7d). MYC upregulation results in the malignant expansion of immature hepatocytes with stem cell features, consistent with previous reports suggesting that liver tumors arise from stem cells. Upon MYC inactivation, most of the tumor cells die, but some of the tumor have stem cell properties. These stem cells are now able to undergo differentiation into normal hepatocytes and bile duct cells. Amongst the differentiated tumor cells are retained the stem cells, and upon MYC reactivation these cells are a likely source for the reemergence of the tumor. Alternatively, MYC reactivation results in the dedifferentiation of some of the mature hepatocytes. Our model system provides a strategy to identify and purify these hepatic stem cells.

[00102] These data have important implications for the development of therapies for the treatment of cancer. It is evident that MYC inactivation can result in sustained tumor regression of liver cancers - a type of tumor that can be refractory to clinical treatment. Methods

[00103] Transgenic Mice - Tet-o-MYC transgenic mice were generated previously described in Felsher and Bishop (1999). LAP-tTA mice are described by Kistner et al. (1996) Proc Natl Acad Sci U S A 93, 10933-8. The generation of transgenic CMV-gfp-luc mice on a FVB background is described by Hardy et al. (2001) Exp Hematol 29, 1353-60.

[00104] Tumorigenicity Assays - To suppress MYC transgene expression, mice received doxycycline in their drinking water, changed once per week, at a concentration of 100 μg/ml. For transplantation experiments, tumors were prepared as single cell suspensions by incubating liver tumor pieces in HBSS followed by digestion in 1.5 mg/ml collagenase in 3 mM KCI, 5 mM NaH2PO4, 130 mM NaCI, 10 mM Dextrose Monohydrate. Cells were washed in PBS twice and resuspended in PBS and 10⁷ cells were inoculated intraperitoneal^ into SCID mice.

[00105] Histology - Tissues were fixed in 10% buffered formalin, paraffin embedded and 5 μm paraffin sections were stained with hematoxylin and eosin.

[00106] lmmunohistochemistry - Staining was performed on 4 micron paraffin tissue sections placed on Superfrost Plus (Fisher Scientific) slides and employing the Avidin Biotin staining technique. Sections were deparaffinized and cleared before being treated for endogenous peroxidase activity in 3% solution of hydrogen peroxide (H2O2) and methanol. This treatment was followed by hydration in a graded alcohol series to distilled water. Subsequently, sections were microwave treated for 12 min. in citric acid buffer. Slides were allowed to cool and placed in PBS before adding of antibodies. Sections were incubated with 10% universal blocker (Biogenex) for 20 min. in a humidifier chamber. Slides were blotted and covered with primary antibody in a humidifier chamber and incubated overnight at room temperature. After washing with PBS slides were incubated for 60 min with biotinylated universal link antibody (Biogenex, City. State). Two 5 min. rinses in PBS and a 30 min. incubation with streptavidin-px complex completed the staining. Color development was achieved by treatment with the chromogen DAB (Vector Laboratories, Redwood City, CA) and was carried out for 3-5 min under a microscope. The slides were rinsed in running tap water, counter-stained in Mayer's hematoxylin, dehydrated, cleared and covered. Negative control slides were run without primary antibody. Positive control slides known to react with each antibody were incorporated into each run.

[00107] Western Blots - Western analysis was performed using conventional techniques. MYC protein expression was detected using the sc-788 (SantaCruz Biotechnology Santa Cruz, CA.) antibody. AFP (sc-15375 - SantaCruz Biotechnology, Santa Cruz, CA), α-tubulin (CP06 - Oncogene Inc.)

[00108] In Vivo Bioluminescence Imaging— Transgenic mice were anesthetized with injected Ketamine/Xylazine solution (50 μl/10 g), or a combination of inhaled isoflorane/oxygen delivered by the Xenogen XGI-8 5-port Gas Anesthesia System. An aqueous solution of the substrate D-luciferin (150 mg/kg) was injected into the animal's peritoneal cavity 5 min before imaging. Animals were then placed into a light-tight chamber and imaged with an IVIS-100™ cooled CCD camera (Xenogen, Alameda, CA). First, a grayscale body surface reference image (digital photograph) was taken under weak illumination. After switching off the light source, photons emitted from luciferase expressing cells within the animal and transmitted through the tissues were collected for a period of 30 sec. to 5 min and quantified using the software program "Living Image" (Xenogen) as an overlay on the image analysis program "Igor" (Wavemetrics, Seattle, WA). For anatomical localization, a pseudocolor image representing light intensity (blue, least intense; red, most intense) was generated in "Living Image" and superimposed over the gray scale whole body reference image. Livinglmage was used to collect, archive, and analyze photon fluxes and transform them into pseudocolor images. The camera controls were managed by the Livinglmage software (Xenogen, Alameda, CA) and this software was used to collect, archive, and analyze images.

Claims

WHAT IS CLAIMED IS:

1. A transgenic non-human animal model for liver cancer, comprising an exogenous oncogene coding sequence operably linked to a promoter that is directly or indirectly regulated by a soluble factor, wherein said oncogene is specifically expressed in liver cells; and wherein tumorigenesis caused by expression of said oncogene is reversible.

2. The transgenic animal of Claim 1 , wherein said oncogene is c-myc.

3. The transgenic non-human animal of Claim 2, wherein said promoter is regulated by a specific repressor and by a transactivator, wherein said transactivator is expressed specifically in liver cells.

4. The transgenic non-human animal of Claim 3, wherein said oncogene is operably linked to a minimal promoter, and upstream of said promoter, a tetracycline response element (TRE).

5. The non-human transgenic animal of Claim 4, wherein expression of said oncogene has been activated for a period of time sufficient for tumorigenesis.

6. The non-human transgenic animal of Claim 4, wherein following said tumorigenesis, expression of said oncogene has been inactivated.

7. The non-human transgenic animal of Claim 1 , further comprising a transgene encoding a detectable marker.

8. The non-human transgenic animal of Claim 1 , wherein said animal is a rodent.

9. A liver stem cell derived from the transgenic animal of Claim 1.

10. A liver stem cell derived from the transgenic animal of Claim 6.

11. A method of screening for genetic sequences specifically expressed in mammalian liver stem cells, the method comprising: isolating RNA from an a cell population according to either of Claims 9 or 10; generating a probe from said RNA; screening a population of nucleic acids for hybridization to said probe.

12. The method of Claim 11 , wherein said population of nucleic acids is represented in an array.

13. A method of screening for agents that affect the viability, growth, metabolic function or differentiation of mammalian liver stem cells, the method comprising: contacting a cell population according to according to either of Claims 9 or 10, and determining the effect of said agent on the viability, growth, metabolic function or differentiation of said liver engrafting cells.

14. The method according to Claim 13, wherein said agent is a drug suspected of toxicity on hepatocytes.

15. The method according to Claim 14, wherein said agent is a hepatitis virus.

16. The method according to Claim 14, wherein said agent is an anti-viral agent.

17. The method according to Claim 14, wherein said agent is an anti-cancer agent.

18. A method of screening for agents that affect the viability, growth or metabolic function of latent tumor cells, the method comprising: contacting a cell population according to according to Claim 10 or a transgenic animal according to Claim 6 with a candidate agent, and determining the effect of said agent on the viability, growth or metabolic function of latent tumor cells.