WO2013192094A1

WO2013192094A1 - Reconstituted human liver tumor model

Info

Publication number: WO2013192094A1
Application number: PCT/US2013/046144
Authority: WO
Inventors: Samuel J. FARLOW; Yinghui Zhou; Joerg Heyer
Original assignee: Aveo Pharmaceuticals Inc
Current assignee: Aveo Pharmaceuticals Inc
Priority date: 2012-06-18
Filing date: 2013-06-17
Publication date: 2013-12-27
Anticipated expiration: 2014-12-18

Description

RECONSTITUTED HUMAN LIVER TUMOR MODEL

CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of and priority to United States Provisional Patent Application serial number 61/661,030, filed June 18, 2012, the contents of which are hereby incorporated by reference.

FIELD OF THE INVENTION

[0002] The field of the invention is molecular biology and oncology.

BACKGROUND

[0003] Mouse models of human cancer provide a valuable tool for drug treatment and cancer biology studies. (See, Heyer et ah, 2010 Nat. Rev. Cancer 10:470-480, for a review.) For example, conventional human xenograft models offer the advantage of working with human cancer cells in vivo. A disadvantage of xenograft models, however, is that the human cells have been maintained in culture as distinct cell lines (e.g., NCI 60 panel) for many years. This can lead to significant differences between the properties and behavior of the xenograft cells as compared to primary tumor cells. To address the need to work with primary tumor cells, in vivo models that provide spontaneous tumors in mice have been developed. See, e.g., U.S. Patent No. 6,639, 121. The disadvantage of these models, however, is that the tumor cells are mouse tumor cells, and therefore, these models cannot be used effectively in drug studies using anti-human antibodies that do not cross-react with mouse targets.

[0004] In limited instances, human tumors have been produced in mice using human-in- mouse (HIM) transplantation models. For example, a human breast tumor model was generated in mice based on the spontaneous formation of human breast tumors using transduced primary mammary epithelial cells transplanted into a fat pad of an

immunocompromised mouse. See, e.g., U.S. Patent Publication Nos. 2006/0123494 and

2006/0161999. Skin tumor models have also been generated by transducing primary cells from human skin, culturing the skin cells in vitro and then grafting the cells into an immunocompromised mouse (reviewed in Heyer et ah, supra). These models offer the advantage of testing therapies, e.g., anti-human antibodies, on human tumors in a mouse model.

[0005] Although human tumors have been generated using other cell types, e.g., primary mammary epithelial cells and primary skin cells, production of human liver tumors in a mouse has not yet been achieved. To date, there are no rodent models of primary human liver tumors. Unlike most other primary cell types, primary human hepatocytes are known to be fragile cells and can be difficult to manipulate in vitro. For example, Laba et al. describe human parenchymal liver cells (or hepatocytes) as "extremely fragile cellular components" that "require very specific microenvironmental conditions to maintain the cell phenotype in vitro" (2005 Arch Immunol. Ther. Exp. 53 :440-453). Li describes that "most lots of cryopreserved human hepatocytes cannot be cultured as attached, monolayer cultures" and, therefore, notes that their applications are limited to those compatible with suspension culture (2007 Chemico- Biol. Interactions 168: 16-29). Moreover, suppliers of cryopreserved human hepatocytes routinely note that thawed hepatocytes are fragile and suggest that the cells be handled gently to maintain viability.

[0006] Therefore, additional models based on the formation of primary human tumors, such as primary human liver tumors, in a mouse are needed for cancer research.

SUMMARY

[0007] Methods for producing human liver tumors in mice have been discovered. Based on the disclosed methods, the invention features a mouse that contains a reconstituted human liver tumor model.

[0008] The mouse model includes a spontaneous human tumor in a chimeric liver reconstituted to comprise mouse liver cells and integrated human liver cells. The tumor comprises a plurality of human liver cells, e.g., hepatocytes that contain (i) a recombinant human oncogene, e.g., KRAS, and (ii) a genetic modification that reduces or eliminates the function of a tumor suppressor gene, e.g., a recombinant SV40 early region (SV40ER). In exemplary embodiments, the human liver cells do not contain a recombinant hTERT gene.

[0009] Also disclosed herein is a method for making mice that comprise a reconstituted human liver tumor model. The method includes the steps of: (a) providing a non-immortalized human hepatocyte in suspension culture; (b) transducing the hepatocyte, in a non-adherent vessel (i.e., a cell abhesive vessel), by spin infection with (i) a vector comprising a recombinant human oncogene, and (ii) a genetic modification that reduces or eliminates the function of a tumor suppressor gene, e.g., vector comprising a recombinant SV40ER; (c) washing the transduced hepatocyte to remove residual vectors from the suspension culture medium; (d) injecting the transduced and washed hepatocyte, directly or indirectly, into a non-cancerous liver in an immunocompromised mouse; and (e) maintaining the mouse for a suitable latency period, thereby providing: (i) time for the transduced hepatocyte to integrate into the liver, and (ii) time for a tumor to form spontaneously from the transduced hepatocyte and/or a descendent of the transduced hepatocyte. The method may further comprise, after the latency period, the step of isolating primary human liver tumor material from the mouse.

[0010] Other features and advantages of the invention will be apparent from the detailed description, drawings, and the claims.

DETAILED DESCRIPTION OF THE DRAWINGS

[0011] FIG. 1 is a histogram comparing hepatocyte viability and viral infection rate between spin infection versus a viral infection without spinning. The solid black bars are the percentage of viable cells recovered after infection. The hatched bars are the percentage of cell positive for X-gal staining, quantified using the Image J software.

[0012] FIG. 2 is a histogram comparing the human serum albumin level of mice transplanted with hepatocytes transduced by spin infection in a non-adherent plate (solid back bars) versus transduced cells that were allowed to attach to the plates (hatched bars).

[0013] FIG. 3 is a photograph showing a dark field green fluorescent image depicting tumor nodules in the liver that are GFP positive. The boundary of the liver is demarcated by the dashed line.

[0014] FIG. 4 is a pair of photographs showing Fluorescent In Situ Hybridization (FISH) staining of a liver section with a human specific DNA probe (panel A, red channel) and a mouse specific DNA probe (panel B, green channel).

DETAILED DESCRIPTION

[0015] Disclosed herein are methods for making a primary human liver tumor model in a mouse. In this model, tumors (tumor nodules) arise spontaneously from a chimeric liver comprising mouse liver cells and integrated human liver cells. For basic research on human cancer biology and for drug discovery and development, this model offers three major advantages. First, it provides primary liver tumors that arise spontaneously, thereby mimicking naturally occurring human liver cancer, e.g., hepatocellular carcinoma (HCC). Second, the tumors are genetically human, because they arise from human liver cells implanted and integrated into the liver of the host mice. Third, the tumors are generated by defined genetic elements, thereby providing opportunities to study pathway-related tumorigenesis in primary human liver tumors. As described herein, primary human liver tumors (or primary human liver tumor material) may be isolated from the mouse for cancer biology research, drug discovery and/or archiving.

[0016] Published methods for working with human hepatocytes indicate that these cells are fragile (see, e.g., Laba et ah, supra; Li, supra). Consistent with their fragile nature, methods for processing and manipulating human hepatocytes recommend using low-speed

centrifugation (e.g., 40-1 lOxg), wide-bore pipette tips, gentle mixing and the avoidance of severe agitation of the cells. In contrast to the published recommendations and guidelines for handling human hepatocytes, the disclosed methods surprisingly use spin infection, which requires a high speed spin (e.g., about 500-2000xg), to transduce human hepatocytes, prior to transplantation into an immunocompromised mouse. The disclosed methods demonstrate that transduction of human hepatocytes by spin infection leads to higher transduction efficiency compared to the transduction of human hepatocytes in ordinary suspension and on plastic. It was unexpected that human hepatocytes could be transduced with high efficiency using spin infection and that the transduced cells could survive this processing step prior to transplantation into a mouse liver. It was also unexpected that transduction of human hepatocytes using spin infection on non-adherent surfaces would lead to higher rates of human hepatocyte engraftment in mouse livers (compared to transduced human hepatocytes on adherent surfaces), and that the integrated human hepatocytes could produce spontaneous human liver tumors in the recipient mouse.

[0017] Suitable human hepatocytes include hepatocytes freshly isolated from human livers or cryopreserved hepatocytes (thawed prior to use). Cryopreserved hepatocytes are commercially available, e.g., from Invitrogen, Xenotech and Celsis. [0018] To develop human liver tumors, e.g., human HCC tumors, human liver cells, e.g., hepatocytes, are genetically modified prior to transplantation in an immunocompromised mouse. Human hepatocytes are genetically modified to (i) express a recombinant human oncogene, and (ii) to reduce or eliminate the function of a tumor suppressor gene. [0019] Examples of human oncogenes that can be introduced into the human hepatocytes include KRAS, HRAS, NRAS, EGFR, MDM2, RhoC, AKTl, AKT2, MEK, c-MYC, n-MYC, β-catenin (such as -catenin^AN131), PDGF, c-MET, PIK3CA (also known as pi io^aH1047R), CDK4, cyclin Bl, cyclin Dl, estrogen receptor gene, progesterone receptor gene, ERBBl (also known as HER1), ERBB2 (also known as HER2), ERBB3 (also known as HER3), ERBB4 (also known as HER4), TGFa, TGF , β-RAF, RON, ALK, ras-GAP, She, Nek, Src, Yes, Fyn, Wnt, Bci₂ and Bmil. Preferred human oncogenes are KRAS, c-MYC, PIK3CA and β-catenin.

[0020] In some embodiments, the recombinant human oncogene is placed under the control of an inducible promoter. Examples of inducible promoters useful for this purpose include a tetracycline-inducible promoter, a metallothionine promoter, the IPTG/lacI promoter system, and the ecdysone promoter system. In addition, the "lox stop lox" system can be used for irreversibly deleting inhibitory sequences for translation or transcription. An inducible oncogene construct can be used in making genetically modified human liver cells described herein, which are implanted in an immunocompromised host mouse. The implanted human liver cells are maintained in the presence of the inducer, e.g., by administering the inducer in the drinking water of the host mouse, until a tumor forms.

[0021] In addition to comprising a recombinant human oncogene, the human hepatocytes comprise a genetic modification that reduces or eliminates the function of a tumor suppressor gene. Exemplary tumor suppressor genes include p53, Rb, APC, Ink4a/Arf, Axin and pTEN. In some embodiments, the genetic modification that reduces or eliminates the function of a tumor suppressor gene is a dominant negative mutation in a tumor suppressor gene, e.g., p53 175H, APC delta C terminal 1309, Axin with C terminal truncation at residue 596, and pTEN C124S. In other embodiments, the dominant negative mutation is an over-expressed recombinant gene. Alternatively, the genetic mutation may be a recombinant siRNA that targets expression of a tumor suppressor gene. In an exemplary embodiment, the genetic modification is a recombinant SV40 early region (SV40ER). SV40ER encodes both large T and small t antigens and simultaneously disrupts p53 and RB (Cheng et ah, 2009 Semin.

Cancer Biol. 19:218-228; Tilli & Furth, 2003 Breast Can. Res. 5:202-205).

[0022] Human hepatocytes can be transduced using any eukaryotic expression vector that is compatible with spin infection. Exemplary expression vectors include non-replicating viral vectors, e.g., lentiviral vectors, retroviral vectors, adenoviral vectors, and adeno-associated viral (AAV) vectors. Infection may be by a viral vector including lentiviruses, retroviruses and adenoviruses, which are known in the art. Typically, such vectors contain convenient restriction sites for insertion of desired recombinant sequences. The vectors can include a selectable marker, e.g., a drug resistance gene. An exemplary drug resistance gene is the neomycin phosphotransferase (neo) gene (Southern et ah, 1982, J. Mol. Anal. Genet. 1 :327-

341), which confers G418 resistance. Alternatively, genes encoding fluorescence markers, e.g., green fluorescent protein, yellow fluorescent protein or red fluorescent protein, or genes encoding bioluminescent proteins, e.g., luciferase, can be used as selectable markers. Such genes are commercially available. [0023] As described herein, human hepatocytes are transduced by spin infection. Spin infection is a transduction technique that uses a high speed spin (e.g., about 500-2000xg) to introduce, e.g., a virus expressing one or more genes of interest into a cell. The person of ordinary skill would understand how to optimize the g- force for each batch of cells being transduced, e.g., with virus expressing a recombinant human oncogene and a virus expressing SV40ER into a human hepatocyte. In an exemplary embodiment, hepatocytes are transduced on a non-adherent (cell adhesive) surface, e.g., a non-adherent vessel (a cell adhesive vessel).

[0024] Transduced hepatocytes may be washed, e.g., by centrifugation to remove unbound virus particles prior to transplantation in a recipient mouse. Suitable buffers for washing hepatocytes are known in the art and are commercially available. [0025] In some embodiments, infection efficiency is monitored, e.g., by quantitative realtime PCR for integrated lentiviral sequences and/or by percent transduction by GFP

quantitation.

[0026] Mice used in the practice of the disclosed methods are immunocompromised. A compromised immune system is desirable to prevent the mouse from rejecting the implanted human cells. Examples of immunocompromised mice include SCID mice, nude mice, Nod mice, mice whose thymus gland has been surgically removed, and mice whose immune system has been suppressed by drugs or genetic manipulations, e.g., rag-2^7" and rag-2^_/" il-2rg^_/" (also known as rag-2^_/~ yc ^1'). Genetically immunocompromised mice are commercially available, e.g., from Taconic, and selection of immunocompromised mice suitable for purposes of the present invention is within ordinary skill in the art. Additional mice that may be used in the disclosed method include SCID-Alb/UPA mice (as described by Dandri et al., 2001

Hepatology 33 :981-988; Meuleman et al, 2005 Hepatology 41 :847-856 (available from PhoenixBio), fumarylacetoacetate hydrolase (Fah)-null mice (as described by Bissig et al., 2007 PNASUSA 104:20507-20511), and Fah^"7" rag-2^"A il-2rg^"/" (FRG) mice (as described by Azuma et al., 2007 Nature Biotechnol. 25: 903-910; commercially available from Yecuris).

[0027] Transduced human hepatocytes are introduced into a non-cancerous liver of an immunocompromised recipient mouse to generate a chimeric mouse liver comprising both mouse hepatocytes and integrated transduced human hepatocytes. For example, transduced hepatocytes may be injected directly into the liver or introduced indirectly by an intra-splenic injection. Intra-splenic injection may be performed by conventional methods, e.g., as described by Azuma et al., 2007 Nature Biotech. 25:903-910. Alternatively, transduced hepatocytes may be introduced indirectly into the mouse liver by an injection into the portal vein.

[0028] To induce engraftment (i.e., integration) of human hepatocytes into the mouse liver, the host mouse liver is treated to prevent mouse liver cell proliferation and/or injured prior to transplantation of the human hepatocytes. Engraftment or integration generally refers ro the process in which the transplanted human liver cells incorporate into existing portal tracts of mouse hepatocytes to form a chimeric liver (e.g., as described in Rhim et al. 1995 PNAS USA 92:4942-4946). In some embodiments, the host mouse is treated with either retrorsine or monocrotaline (both commercially available from Sigma; see, e.g., Witek et al., 2005 Cell Transplantation 14:41-47) to inhibit endogenous hepatocyte proliferation. Injury to the mouse liver may be produced chemically, physically, or genetically. In certain embodiments, chemical injury methods include treating the mouse with carbon tetrachloride (CC14;

commercially available from Sigma) to stimulate hepatocyte proliferation and regeneration. In other embodiments, the recipient mouse is subjected to a physical injury, e.g., a hepectomy, prior to transplantation to stimulate hepatocyte proliferation and regeneration of liver tissue. In other embodiments, host mice are selected based on genetic modifications that result in liver injury (e.g., Alb/uPA-scid mice, fumarylacetoacetate hydrolase (Fah)-null mice, and Fah^{" "} rag- T<- il-2rg^7" (FRG) mice. In some embodiments, a combination of chemical, physical, and/or genetic injury may be used.

[0029] Following transplantation of the transduced human hepatocytes, the mouse is maintained for a suitable latency period to provide (i) time for the transduced human hepatocyte to integrate into the mouse liver and (ii) time for a tumor to form spontaneously from the transduced hepatocyte and/or a descendent of the transduced hepatocyte. In exemplary embodiments, the transduced human hepatocytes integrate into the mouse liver within approximately 1-2 weeks, e.g., within 2-5 days. Depending on the recombinant human oncogene expressed in the human hepatocyte tumor growth may be observed, e.g., in 2-3 months following transplantation or up to one year following transplantation.

[0030] Engraftment and expansion of human hepatocytes in the mouse liver may be monitored, e.g., by assaying serum levels for human proteins secreted by hepatocytes, e.g., albumin and alpha- 1 antitrypsin. Tumor formation may monitored during the latency period, e.g., by palpation of the abdomen and/or observation of an abdominal bulge or bloating.

[0031] The primary human liver tumors described herein resemble human hepatocellular carcinoma (HCC), which is an invasive adenocarcinoma. Identified histological features of malignancy include marked cytological atypia, cellular pleomorphism, an elevated mitotic rate, the presence of atypical mitoses, and focal tumor necrosis. Features indicative of

hepatocellular differentiation include epithelioid morphology with resemblance to human hepatocellular carcinoma and intratumoral steatotic changes upon exposure to carbon tetrachloride.

[0032] Primary human liver tumor material obtained from the disclosed mice may be isolated for further analysis, e.g., drug discovery to determine anti-tumor effects and/or archiving of tumor material.

EXAMPLES

[0033] The invention is further illustrated by the following examples. The examples are provided for illustrative purposes only, and are not to be construed as limiting the scope or content of the invention in any way. Example 1: Construction of a Liver Tumor Model

Human Hepatocytes

[0034] Cryopreserved primary human hepatocytes were purchased from Invitrogen and stored in liquid nitrogen. To thaw, vials were quickly moved on dry ice to a 37°C water bath for <2 minutes. The thawed hepatocytes were carefully poured into 50 mL of warm

Cryopreserved Hepatocyte Recovery Medium (CHRM, Invitrogen) and the vial was rinsed with 1 mL of CHRM using a wide-bore pipette tip to recover all residual cells. The cells were gently inverted three times and then centrifuged in a swinging bucket rotor at lOOxg for 10 minutes at room temperature. The supernatant was carefully poured off and the cell pellet was gently resuspended in Hepatocyte Plating Medium (Williams Medium E supplemented with Hepatocyte Plating Supplement Pack, Invitrogen). The cell number and viability were determined using acridine orange/propidium iodide dual fluorescence analysis on a Cellometer Vision automated cell counter ( excelom Bioscience). The medium volume was adjusted to lxlO⁶ cells/mL, and polybrene (Sigma) was added to 4 μg/mL. The cells were dispersed onto a 6 well low-binding tissue culture plate (Nunc HydroCell) at a density of approximately 2xl0⁶ cells per well.

Constructs and Virus Production

[0035] Lentivirus vectors were used for transduction of human hepatocytes. The lentivirus backbone used in constructing all of the following lentivirus vectors was pLenti6/V5-D-TOPO, which is commercially available from Invitrogen (Catalog No. K4955-10).

[0036] The vector pLenti-CMV-SV40ER was constructed as follows. A 2.7 kb SV40ER DNA fragment (including LT and st) was obtained by digesting the pSV3-dhfr vector (ATCC #37147) with Sfi I and BamH I. The 2.7 kb fragment was then cloned behind the CMV promoter in vector pLenti6/V 5-D-TOPO. SV40 (promoter)-Blasticidin DNA fragment was then removed from the resulting construct to generate the pLenti-CMV-SV40ER lentivirus construct.

[0037] The vector pLenti-CMV-KRAS+SV40 (promoter)-GFP was constructed as follows. Similar to the pLenti-CMV-SV40ER construct, a 558 bp KRAS cDNA fragment was cloned behind the CMV promoter in vector pLenti6/V5-D-TOPO, and a commercially available GFP (Green-Fluorescent-Protein) cDNA was cloned downstream of the SV40 promoter. The KRAS used in making this construct was the KRAS^G12V. The Genbank accession number for wild- type KRAS cDNA is NM_033360. We used KRAS^G12V (a gift from Lynda Chin, Harvard University Medical School), a mutant form in which amino acid residue 12 is changed from glycine to valine.

[0038] The vector pLenti-CMV-ERBB2+SV40 (promoter)-GFP was constructed as follows. A 3992 bp ERBB2 cDNA fragment was cloned behind the CMV promoter in vector pLenti6/V5-D-TOPO (Invitrogen; Catalog No. K4955-10), and a GFP cDNA was cloned downstream of the SV40 promoter. The ERBB2 used in making this construct was the ERBB2 V659E. The accession number for wild-type ERBB2 cDNA is Ml 1730. Site directed mutagenesis was employed to change amino acid residue 659 from V to E.

[0039] pLenti-LacZ was purchased from Invitrogen (Catalog No. K4955-10).

[0040] Lentiviruses were produced by cotransfection of 293T cells with the lentivirus constructs described above and the optimized packaging plasmid mix (Invitrogen; Catalog No. K4975-00). Transfections were performed using the Lipofectamine™ 2000 Transfection Reagent according to the vendor's instructions (Invitrogen; Catalog No. 11668-019.)

Transducing Human Hepatocytes

[0041] Immediately after placing the cells in suspension, lentivirus was added to the hepatocytes at an MOI of 10-100 and the cells were transferred to 37°C, 5% C0₂ for 30 minutes. Cells were then spin infected with the virus by swinging bucket centrifugation at 1 ,000xg for 90 minutes at room temperature. Following spin infection, cells and virus were incubated for an additional 60 minutes in the cell culture incubator. Hepatocytes were pelleted and washed to remove unbound virus particles by centrifugation at 180xg for 5 minutes at room temperature for a total of five rounds. Cell number and viability were determined using acridine orange/propidium iodide dual fluorescence analysis on a Cellometer Vision automated cell counter. Hepatocytes were resuspended in DMEM medium/10% FBS at a concentration of lxlO⁷ cells/ml and stored on ice for less than 60 minutes, before transplantation.

[0042] The procedure for infecting primary human hepatocytes was established by comparing the viral infection rate and viability of the hepatocytes with and without spin infection. Cells were mixed with pLenti-LacZ virus in equivalent amounts and plated in two standard tissue culture plates. One plate of cells was transduced by spin infection, the other was transduced without spinning. After infection, the two plates were incubated at 37°C, 5% CO₂ for 4 days to allow expression of the LacZ gene before being fixed and stained with X-gal for beta-galactosidase activity, using the LacZ Cell Staining Kit according to the vendor's protocol (Invivogen). The number of infected cells were counted with the Image J software. As shown in FIG. 1, although similar percentages of viable cells were recovered (63% for spin infection vs. 77% for transduction without spinning), the infection rate was two-fold greater with spin infection than transduction without spinning (52% vs. 25%, respectively). Example 2: Transplantation and Engraftment of Transduced Human Hepatocytes into an Immunocompromised Mouse

[0043] All procedures performed on mice were reviewed and approved by the Institutional Animal Care and Use Committee at AVEO Pharmaceuticals. Six-to-eight week old female rag-2^7" il-2rg^_/" were purchased from Taconic. Prior to transplantation, recipient mice were injected with two doses of retrorsine (60 mg/kg intraperitoneally, two to four weeks apart;

Sigma). One day prior to mouse treatment, retrorsine was dissolved in a glass vial overnight at 55°C in 100% EtOH at a concentration of 20 mg/ml. Immediately prior to injection, retrorsine was diluted to 4 mg/ml with 0.9% normal saline.

[0044] Two to four weeks after the final dose of retrorsine, human hepatocytes were transplanted to the recipient liver by intra-splenic injection. Mice were anesthetized with isoflurane and kept on a heating pad for the entire surgical procedure. A mid-abdominal incision was performed and the spleen was retracted. l-2xl0⁶ cells in 100 μΐ were injected over 2 minutes into the lower pole of the spleen using a syringe pump and a 27-gauge butterfly needle. The needle was withdrawn one minute after injection was stopped, and the injection site was compressed for one minute with a PVA surgical spear (Braintree Scientific). The spleen was cauterized and removed, the abdominal wall was closed using Webcryl 4/0 absorbable sutures (Webster Veterinary) and the skin was stapled.

[0045] Two weeks after hepatocyte transplantation, staples were removed and carbon tetrachloride (CC1₄; available from Sigma) treatment was started. Three doses of 1 : 10 diluted CCl₄were injected intraperitoneally (0.5 ml/kg intraperitoneally, one week apart). [0046] To monitor engraftment of the transduced human hepatocytes into the mouse liver, the production of human serum albumin was measured in mouse blood. Blood was collected regularly by saphenous vein bleeding and stored in aliquots at -20°C until analyzed. Human albumin levels in mouse serum were quantitated with the Human Albumin ELISA Quantitation Kit (Bethyl Laboratories) according to the vendor's protocol. Serum samples were diluted 100- fold to fit within the linear range of the standard curve.

[0047] Prior to this study, we were unaware of any previous reports of transplanting or engrafting transduced human hepatocytes into a mouse liver. To establish a baseline for comparison, the engraftment efficiency of three different lots of non-transduced human hepatocytes were tested up to 14 weeks after surgery and the human albumin level were monitored at weeks 1, 3, 7 and 14. Human albumin was detectable one to seven weeks after surgery in mice transplanted with all three lots of hepatocytes. The albumin levels diminished 14 weeks after surgery in all groups. A single lot of hepatocytes (lot 1) was selected for all subsequent experiments. [0048] Next, two different procedures were tested to determine the conditions for achieving high engraftment efficiency with transduced human hepatocytes. In the first procedure, human hepatocytes were spin infected on standard tissue culture plates and allowed to attach during an overnight incubation before being detached and transplanted into recipient mice the next day ("adherent"). In the second procedure, human hepatocytes were spin infected on HydroCell low binding plates (non-adherent plates) and transplanted into recipients the same day to prevent the cells from attaching to the plate ("non-adherent"). In a comparison experiment, five recipient mice were each transplanted with two million hepatocytes infected with the Lenti-LacZ virus under both procedures. The engraftment efficiency was measured by the human albumin assay 4 and 9 weeks after surgery and the data are shown in FIG. 2. The data demonstrated that hepatocytes transduced under the second procedure ("non-adherent") have a 7-to-15 fold higher engraftment efficiency than cells transduced under the first procedure ("adherent"), i.e., 1.53 μg/mL ("non-adherent"; solid bars in FIG. 2) vs. 0.22 μg/mL

("adherent"; hatched bars in FIG.2) on week 4 and 1.13 μg/mL ("non-adherent"; solid bars in FIG. 2) vs. 0.08 μg/mL ("adherent"; hatched bars in FIG.2)on week 9. [0049] These data also demonstrate that hepatocytes transduced under the second procedure (spin infection in a non-adherent plate) have a similar engraftment efficiency compared to non- transduced hepatocytes, i.e., 1.53 μg/mL (spin infection in a non-adherent plate) vs. 1.50 μg/mL (non-transduced hepatocytes).

Example 3: Carcinoma Developed from Human Hepatocytes Transduced with SV40 Early Region and KRAS or HER2 [0050] After establishing conditions for transducing primary human hepatocytes without compromising their ability to engraft in the recipient mouse liver, oncogenic combinations were tested for carcinogenesis. We and others have previously demonstrated that murine hepatocytes with loss of the p53 gene can be transformed by KRAS, MYC and other oncogenes and give rise to propagatable murine hepatocellular carcinomas (Zender et a , 2006 Cell 125: 1253-1267). One of the significant differences between mouse and human cells is the length of their telomeres, and it has been believed in the art that reintroduction of telomerase activity through hTERT is necessary for the development of human carcinomas, including HCC. To test whether human HCC can develop in our mouse model in the absence of recombinant hTERT transgene, primary human hepatocytes were transduced with

SV40ER+KRAS or SV40ER+HER2 and transplanted into recipient mice (n= 20 for

SV40ER+KRAS and n= 10 for SV40ER+HER2). The engraftment efficiency was monitored by human albumin ELISA.

[0051] Mice receiving SV40ER+KRAS transduced hepatocytes started to show bloated abdomen approximately seven weeks after surgery. Upon gross examination, twelve out of eighteen mice (67%) had multiple nodules found in the liver. The average latency for these twelve tumor bearing mice is 8.6 weeks. Four out of nine mice receiving SV40ER+HER2 transduced hepatocytes had similar nodules in the liver and the average latency for these four mice is 13.9 weeks.

[0052] The diameter of these nodules ranged from 2-8 mm and the number of nodules per mouse liver varied from 2 to 10. Under a fluorescent microscope, these nodules emitted green fluorescent light, demonstrating that they were derived from the transduced human hepatocytes (see FIG. 3). FISH analysis with human specific DNA probe confirmed that the tumors derived exclusively from human cells (FIG. 4).

[0053] Portions of the livers from tumor bearing mice (e.g., mice bearing two or more nodules) were harvested, fixed in formalin, and embedded in paraffin. Five micron sections were cut and stained with Hematoxylin and Eosin (H&E) for pathological analysis. The H&E stained slides were scanned using an Aperio slide scanner and the images were reviewed by a pathologist who confirmed that the nodular masses were malignant hepatocellular carcinomas. Features diagnostic of malignancy include marked cytological atypia, cellular pleomorphism, an elevated mitotic rate, the presence of atypical mitoses, and focal tumor necrosis. Features suggestive of hepatocellular differentiation include epithelioid morphology with resemblance to human hepatocellular carcinoma and intratumoral steatotic changes upon exposure to carbon tetrachloride. The histology of SV40ER+KRAS tumors and that of SV40ER+HER2 tumors are not distinguishable, indicating that both combinations result in primary hepatocellular carcinoma.

Example 4: Propagation of SV40ER+KRAS driven HCC tumors

[0054] In this experiment, the genetically engineered tumors were injected into the subcutaneous space of immunocompromised mice (e.g., NCR nude recipient mice) to determine if they propagate in an ectopic environment. Propagation of genetically engineered tumors under these conditions is a stringent test for demonstrating malignancy of the tumors.

[0055] Five SV40ER+KRAS driven tumors (hHK4, hHK6, hHK7, hHKlO and hHK12) were randomly selected and tested for the ability to propagate in the subcutaneous space of NCR nude recipient mice. The viable region of each tumor was chopped to small pieces and digested to single cells in collagenase. Trypan blue staining showed that 30-50% of the resultant cells were viable. Each primary tumor was injected into five recipient mice. For each injection, 300,000-500,000 viable cells were mixed with Matrigel and injected into the right flank of an NCR nude female mouse. After a two week latency period, propagated tumors were visible in the subcutaneous space in NCR nude recipient mice for four of the five tumors (hHK6, hHK7, hHKlO and hHK12). These tumors grew rapidly and reached the IACUC limit in another 2-3 weeks. These results demonstrate the malignancy of the disclosed genetically engineered HCC tumors. Propagated tumors were isolated and cryopreserved.

INCORPORATION BY REFERENCE

[0056] The entire disclosure of each of the patent documents and scientific articles cited herein is incorporated by reference for all purposes. EQUIVALENTS

[0057] The invention can be embodied in other specific forms with departing from the essential characteristics thereof. The foregoing embodiments therefore are to be considered illustrative rather than limiting on the invention described herein. The scope of the invention is indicated by the appended claims rather than by the foregoing description, and all changes that come within the meaning and range of equivalency of the claims are intended to be embraced therein. Further, the term "comprising" is intended to embrace the terms "consisting essentially of and "consisting of."

Claims

What is claimed is: 1. A mouse comprising a human liver tumor model, wherein the model comprises a spontaneous human liver tumor in a chimeric liver comprising mouse liver cells and integrated human liver cells.

2. The mouse of claim 1, wherein the human liver cells comprise: (a) a genetic modification that reduces or eliminates the function of a tumor suppressor gene; and (b) a recombinant human oncogene.

3. The mouse of claim 2, wherein the recombinant human oncogene is selected from the group consisting of KRAS, HRAS, NRAS, EGFR, MDM2, RhoC, AKT1, AKT2, MEK, c- MYC, n-MYC, β-catenin, PDGF, C-MET, PIK3 CA, CDK4, cyclin B 1 , cyclin D 1 , estrogen receptor gene, progesterone receptor gene, ERBBl, HER2, ERBB3, ERBB4, TGFa, TGF , β- RAF, RON, ALK ras-GAP, She, Nek, Src, Yes, Fyn, Wnt, Bcl₂ and Bmil .

4. The mouse of claim 3, wherein the recombinant human oncogene is selected from the group consisting of KRAS, c-MYC, PIK3CA, HER2, and β-catenin.

5. The mouse of claim 4, wherein the recombinant human oncogene is KRAS or HER2.

6. The mouse of claim 2, wherein the genetic modification is a dominant negative mutation in a tumor suppressor gene.

7. The mouse of claim 6, wherein the dominant negative mutation is selected from the group consisting of p53 175H, APC delta C terminal 1309, Axin with C terminal truncation at residue 596, and pTEN C124S.

8. The mouse of claim 6, wherein the dominant negative mutation is an overexpressed recombinant gene.

9. The mouse of claim 2, wherein the genetic modification is a recombinant siRNA that targets expression of a tumor suppressor gene.

10. The mouse of claim 9, wherein the tumor suppressor gene is selected from the group consisting of: p53, Rb, APC, Ink4a/Arf, Axin and pTEN.

1 1. The mouse of claim 2, wherein the genetic modification is a recombinant SV40 early region.

12. The mouse of claim 1 1, wherein the recombinant human oncogene is KRAS or HER2.

13. The mouse of claim 1, wherein the human liver cells do not contain a recombinant hTERT gene.

14. A method of making a human liver tumor model in a mouse, comprising:

(a) providing a non-immortalized human hepatocyte in suspension culture; (b) transducing the hepatocyte, in a non-adherent vessel, by spin infection with (i) a vector comprising a recombinant human oncogene, and (ii) a vector comprising a recombinant SV40ER;

(c) washing the transduced hepatocyte to remove residual vectors from the suspension culture medium;

(d) injecting the transduced and washed hepatocyte into a non-cancerous liver in an immunocompromised mouse; and

(e) maintaining the mouse for a suitable latency period, thereby providing: (i) time for the transduced hepatocyte to integrate into the liver, and (ii) time for a tumor to form spontaneously from the transduced hepatocyte and/or a descendent of the transduced hepatocyte.

15. The method of claim 14, further comprising, after the latency period, isolating primary human liver tumor material from the mouse.

16. The method of claim 14, wherein the recombinant human oncogene is selected from the group consisting of KRAS, HRAS, NRAS, EGFR, MDM2, RhoC, AKT 1 , AKT2, MEK, c- MYC, n-MYC, β-catenin, PDGF, C-MET, PIK3 CA, CDK4, cyclin B 1 , cyclin D 1 , estrogen receptor gene, progesterone receptor gene, ERBBl, HER2, ERBB3, ERBB4, TGFa, TGF , β- RAF, RON, ALK ras-GAP, She, Nek, Src, Yes, Fyn, Wnt, Bcl₂ and Bmil .

17. The method of claim 16, wherein the recombinant human oncogene is selected from the group consisting of KRAS, c-MYC, PIK3CA, HER2, and β-catenin.

18. The method of claim 17, wherein the recombinant human oncogene is KRAS.

19. The method of claim 14, wherein the hepatocyte does not contain a recombinant hTERT gene.

20. The method of claim 14, wherein the transduced and washed hepatocyte is injected indirectly into the liver of the immunocompromised mouse.

21. The method of claim 20, wherein the hepatocyte is injected into the spleen.

22. The method of claim 20, wherein the hepatocyte is injected into the portal vein.

23. Primary human liver tumor material isolated from the method of claim 15.

24. The mouse comprising a human liver tumor generated from the method of claim 14.