The present invention relates to the preparation of non-human animals having chimeric livers, whereby some or substantially all of the hepatocytes present are human hepatocytes. It is based, at least in part, on the discovery that rats, tolerized in utero against human hepatocytes, were found to serve as long-term hosts for human hepatocytes introduced post-natally, and the introduced hepatocytes maintained their differentiated phenotype, as evidenced by continued production of human albumin. The present invention further relates to the use of such animals as models of various liver diseases, including viral invention. Such embodiments are based on the discovery that transplanted human hepatocytes in chimeric livers were found to be susceptible to Hepatitis B virus and Hepatitis C virus infection.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to tolerized non-human animals having chimeric livers comprising human hepatocytes, methods for preparing such animals, and the use of such animals either as model systems for assaying toxicology or studying human liver disease or as sources of human hepatocytes for re-introduction into a human host. For purposes of clarity, the description of the invention is presented as the following subsections:

i) producing animals having chimeric livers;

ii) toxicology model systems;

iii) model systems for liver diseases; and

iv) chimeric animals as a source of hepatocytes for liver reconstitution. The subject animals of the invention are referred to herein alternatively as "non-human animals having chimeric livers" or simply "chimeric animals". Both these terms are defined as tolerized non-human animals having livers which comprise human hepatocytes.

A "human hepatocyte" as that term is used herein may be a primary hepatocyte harvested from a human liver or a cultured cell from a differentiated hepatocyte cell line. Examples of differentiated hepatocyte cell lines include cells which express one or more molecular marker associated with the differentiated hepatocyte phenotype, such as, for example but not by way of limitation, the asialoglyoprotein receptor and/or the low density lipoprotein receptor. The definition of differentiated cell lines, as that term is used herein, also includes cell lines which exhibit hepatocyte-specific function, such as, but not limited to, susceptibility to infection by a liver-specific (or selective) pathogen, such as a hepatitis B virus.

In preferred specific non-limiting examples of the invention, the differentiated hepatocyte cell lines Huh7 and HepG2 may be appropriate for certain embodiments. These cell lines are ultimately derived from hepatoblastoma cells, and therefore would not be appropriate for introduction into a human subject for gene therapy or for liver reconstitution purposes. These and other hepatoblastoma-derived differentiated hepatocyte cell lines may be used, however, to produce model systems for human liver diseases in non-human animal hosts. Furthermore, it is not required, according to the invention, that such cell lines be able to cross-tolerize an animal to primary human hepatocytes and vice-versa (in fact, it has been determined that for Huh7 and HepG2 cells, cross-tolerization is either sporadic, incomplete, or absent). The usefulness of differentiated hepatocyte cell lines as an efficient source of hepatocytes for development of model systems for liver diseases is demonstrated in the working examples, infra.

References relating to differentiated hepatocyte cell lines include Aden et al., 1979, Nature 282:615-616; Scwartz et al., 1981, J. Biol. Chem. 256:8878-8881; Wu et al., 1984, Hepatology 4(6):1190-1194; Sells et al., 1987, Proc. Natl. Acad. Sci. U.S.A. 84:1105-1009; Nakabayashi et al., 1984; Jpn J Cancer Res 75:151-151; Liang et al., 1993, J. Clin. Invest. 91:1241-1246; Chang et al., 1987, EMBO J 6:675-680; Sandig et al., 1996, Gene Therapy 3:1002-1009; Dash et al., 1997, Am. J. Pathol. 151(2): 363-373; and Yoo et al., 1995, J Virol 69:32-38.

In addition to the human hepatocytes, the livers of the chimeric animals of the invention may also include hepatocytes and non-hepatocyte elements (e.g., biliary and vascular endothelial cells, Kupffer cells, etc.) endogenous to the animal itself. Human cell types other than hepatocytes may also be present. Preferably, the percentage of human hepatocytes (relative to the total number of hepatocytes present) is at least 10 percent, more preferably at least 20 percent, or at least 50 percent, or at least 80 percent.

In particular, chimeric animals are created by introducing human hepatocytes (and possibly additional cell types) into an animal rendered immunologically tolerant to the introduced human cells. As such, the animals may be referred to as being "hosts" to the human cells, where a human being that is a source of such cells may be referred to as a "donor". The term "tolerant", as used herein, does not refer to a state of general immunosuppression (as might be achieved, for example, by treatment with cyclosporine, or as may exist in an animal with a generalized B cell and/or T cell deficiency) but rather indicates a state of antigen-induced non-responsiveness of lymphocytes achieved by clonal deletion, cell-mediated suppression, or anergy (see, for example, Davies, 1997, "Introductory Immunobiology", Chapman & Hall, London, p. 366) directed specifically toward the introduced human cells.

5.1 Producing Animals having Chimeric Livers

The present invention provides for a method of preparing a non-human animal having a liver comprising human hepatocytes, comprising (i) inducing tolerance in a host animal, where the animal is preferably a fetus or a neonate; and (ii) introducing human hepatocytes into the tolerized animal, preferably postnatally and preferably by intra-splenic injection. In specific embodiments, the host animal is subjected to a selection pressure which favors survival and/or proliferation of human, rather than host animal, hepatocytes. A detailed non-limiting description of these features of the invention is set forth in the following subsections.

5.1.1. Host Animals

Non-human animals which may serve as hosts according to the invention are preferably mammals, and include, but are not limited to, mice, hamsters, rats, rabbits, dogs, goats, sheep, pigs, cattle, etc. In particular non-limiting embodiments of the invention, the host animal is a transgenic animal carrying, as a transgene, a gene which, when expressed in hepatocytes, is directly or indirectly (i.e. via a metabolite) toxic to those cells. Examples of such genes are the urokinase gene which is directly toxic (Sandgren et al., 1991, Cell 66:245), and the Herpes simplex virus ("HSV") thymidine kinase gene ("HSV-TK"); which converts the drug gancyclovir into a toxic form and is therefore indirectly toxic (Smythe et al., 1995, Ann. Surg. 222:78-86). Preferably, the gene is operably linked to a promoter which is selectively active in hepatocytes, such as the albumin promoter, the PEPCK promoter, and the hepatitis B surface antigen promoter. To avoid destroying the animal's liver prior to colonization with human hepatocytes, it is desirable to utilize a promoter that is not particularly active pre-natally. Otherwise, such transgenic animals may die in utero. Other promoters inducible by agents that could be locally administered into the liver may also be suitable, such as the metallothionein promoter (which is inducible by heavy metal ions; Palmiter et al., 1982, Cell 29:701). Such genes are not specifically toxic to human hepatocytes, although there may be some "bystander effect" whereby a limited number of the human hepatocytes are killed.

In one specific, non-limiting embodiment of the invention, transgenic mice carrying an albumin promoter/urokinase transgene may be used as hosts. Urokinase is a plasminogen activator that is useful clinically in dissolving blood clots. When introduced into hepatocytes by an adenoviral vector, it was shown to be toxic to those cells (Lieber et al., 1995, Proc. Natl. Acad. Sci. U.S.A. 92:6210-6214). In addition, Sandgren et al. prepared a transgenic mouse containing the mouse urokinase gene driven by a mouse albumin enhancer/promoter (Sandgren et al., 1991, Cell 66:245-256). Because albumin is not produced by the fetal liver (Krumlauf et al., 1985, Cold Spring Harbor Symp. Quant. Biol. 5-0:371-378), animals survived in utero because urokinase was not produced. However, after birth, with activation of the albumin promoter, the liver was destroyed due to the presence of urokinase. To produce such a transgenic mouse for use as a host, heterozygote transgenic mice, B6SJL background, may be obtained from Jackson Laboratories, Stock No. 002214, which contain the mouse urokinase gene driven by a 3.5 kb mouse albumin promoter sequence with a human growth hormone poly A addition site. Pregnant mice from heterozygotic matings may be used to generate homozygous offspring. The number of copies of the urokinase transgene present in each animal at birth may be determined from DNA extracts of tail snips, where the DNA may be digested with Kpn 1, which cuts once within the urokinase gene, and Southern blotting using a detectably labeled probe specific for the urokinase gene, such as 5′-TGTGCTTATG TAGCCATCCA GCGAGTCCCC-3′ (SEQ ID NO: 1). Because somatic mutations leading to inactivation of the urokinase gene may occur, it may be desirable to use breeding pairs of male and female mice successfully rescued into adulthood by introduction of human hepatocytes to generate litters of homozygous offspring. Further, in previous studies on mice carrying a urokinase transgene, inactivating mutations in the urokinase gene were found to result in proliferation of those cells with that somatic mutation while the homozygous cells failed to grow. The proliferating cells, as expected, had higher ploidy than those less actively proliferating (Sandgren et al., 1991, Cell 66:245-256). Thus, the copy number of human DNA, if measured during proliferation of human hepatocytes may be biased, and not reflect the number of cells due to polyploidy. For this reason, the number of human cells may be better estimated by measuring markers specific for human hepatocytes, such as, but not limited to, the human albumin gene or its protein product.

In another specific non-limiting embodiment of the invention, transgenic mice carrying an albumin promoter/HSV-thymidine kinase gene may be used as hosts. Thymidine kinase of HSV differs from mammalian thymidine kinases in its ability to phosphorylate the drug gancyclovir (Fyfe et al., 1978, J. Biol. Chem. 253:8721-8727). In so doing, it converts the non-toxic agent into atoxic form (De Clerq, 1984, Biochem. Biopharmacol. 33:2159-2169). In a specific non-limiting embodiment, the HSV-TK gene (as present in plasmid pLTR-DTK, as developed by D. Klatzmann, Université Pierre et Marie Curie, Paris, France) may be linked to an albumin promoter prepared by excising a 3.2 kb fragment of the mouse albumin promoter (for example from palb_9-12LDLR, James Wilson, University of Pennsylvania, Philadelphia, Pa.) using Bgl II and Sal 1 restriction enzymes (Wilson et al., 1992, J. Biol. Chem. 267:963-967), and placing the promoter fragment in a polylinker site immediately upstream of the HSV-TK gene. Using this plasmid, founder outbred CD1 mice may be prepared and mated to normal CD1 mice to generate heterozygotes, detected by DNA analysis of tail snips using an HSV-TK specific detectably labeled probe. A breeding pair of heterozygotes may then be used to produce mice homozygous for the albumin promoter/HSV-TK transgene. It should be noted that the natural HSV-TK gene contains elements that activate the gene in the testes, which may result in sterile animals that cannot be used as breeders. Accordingly, a version of the gene which lacks these elements is preferred, such as the gene contained in plasmid pLTR-ΔTK (all such variant genes, as well as the wild-type, are considered HSV-TK genes). Breeding of transgenic mice with this specific construct confirmed the success of the deletion (Salomon et al., 1995, Mol. Cell. Biol. 15:5322-5328). Further, a gancyclovir dose-related (Culver et al., 1992, Science 256:1550-1552) bystander effect of the HSV-TK gene product has been observed whereby nearby cells lacking the transgene are destroyed (Kolberg, 1994, J. NIH Res. 6:62-64). Accordingly, it may be desirable to evaluate different doses of gancyclovir and identify the minimum dose required to produce maximal human hepatocyte proliferation.

In yet another non-limiting embodiment of the invention, a drug which is metabolized to a toxic agent by liver cells may be used to reduce the number of host liver cells. For example, such a drug may be administered subsequent to tolerization but prior to human hepatocyte transplant. Preferably, there is a delay between exposure to the drug and death of host animal hepatocytes, so that the animal can maintain liver function while transplanted hepatocytes proliferate to a point where they are present in sufficient numbers to supply the level of liver function required for viability. According to one embodiment, the drug may be retrorsine, a pyrrolizidine alkaloid, which is metabolized by liver cells to a toxic DNA alkylating intermediate. The dose of such agent should be titrated to establish a dosage which will preserve the viability of the animal. For example, two doses of 30 mg/kg of retrorsine given two weeks apart were lethal to newborn rats, and one dose of 30 mg/kg was not sufficient to eliminate all rat liver cells, but it was found that two doses of 12 mg/kg retrorsine, with the first dose given at birth and the second given two weeks later, were not lethal. Accordingly, the present invention provides for the treatment of tolerized newborn rats with a dose of retrorsine of 10-30 mg at birth and then 10-30 mg two weeks therafter, for a total amount of retrorsine less than 60 mg and preferably les than 40 mg, to prepare newborn rats to receive a human hepatocyte transplant.

5.1.2. Tolerization

Non-human animals which are to be used as hosts for human hepatocytes may be rendered tolerant to those hepatocytes by administration of the relevant antigen(s), preferably in the context of human cells or a lysate prepared from human cells, more preferably using human cells from the same individual who is to serve as the hepatocyte donor, or a genetically related and/or identical individual, or, where a differentiated hepatocyte cell line is used, preferably from the same sub-culture (e.g., a culture used as a source of cells for tolerization is preferably derived from the same laboratory stock, and preferably the same culture separated by ten passages or less) as cells to be used for transplant. Tolerizing antigen(s) may be administered as whole cells, a cell extract or one or more purified component thereof. The source of tolerizing antigen(s) may be hepatocytes, but may alternatively be cells of another type, or a mixture of different types of cells. For example, cells prepared from a specimen of human liver tissue may be used as a source of tolerizing antigen(s); such cells may include not only hepatocytes but also fibroblasts, cells of the biliary system, vascular endothelial cells, Kupffer cells, etc. As another example, human splenocytes or lysates thereof may be used to induce tolerance. Cells to be used in tolerization are preferably cleared of undesirable constituents. For example, if the animal is eventually to be used as a model system for a disease where an immune response to an infectious agent is desirably left intact, the animal should not be tolerized against the infectious agent. Alternatively, if the animal is to be used as a host to support the proliferation of human hepatocytes to be used to reconstitute the liver of a person having liver damage caused by an infectious agent, it is desirable not to tolerize the host animal toward the infectious agent or to introduce the infectious agent into the host animal at any time. The cells or lysate are introduced in a physiologically compatible solution; herein, volumes administered refer to cells or lysate comprised in such a solution.

While the host animal may potentially be of any age when tolerized, tolerization is likely to become more difficult as age of the animal increases. Preferably, the animal is still an infant when tolerized; more preferably, the animal is tolerized during the perinatal period when the animal is a neonate, or tolerized in utero. The terms "neonate" and "newborn" are used interchangeably herein. If the intended host animal is a rat, the preferable upper age limit for tolerization is 18 days post-conception (in utero), and the more preferable age for tolerization is 17 days post-conception (in utero), or within 24 hours after birth. If the intended host animal is a mouse, the preferable upper age limit for tolerization is 18 days post-conception (in utero), and the more preferable age for tolerization is 17 days post-conception (in utero), or within 24 hours after birth. If the intended host animal is a pig, the preferable upper age limit for tolerization is 90 days post-conception, and the more preferable age for tolerization is 80 days post-conception, when the animal is still in utero, or within 24 hours after birth.

Tolerization may be accomplished by any route, including but not limited to intravenous, intraperitoneal, subcutaneous, and intrathymic routes. Preferred methods of tolerization include inoculation of human cells into the thymus or intraperitoneally.

As a specific, non-limiting example, where the intended host animal is a rat, tolerance may be induced by inoculating lysate prepared from 1×10⁴-1×10⁶ and preferably 10⁵ human hepatocytes into the peritoneum of a 15-18 day old, and preferably a 17 day old, rat fetus in utero under transillumination. The lysate may be prepared by sonicating a suspension of the appropriate number of human hepatocytes. The same numbers of whole cells may also be inoculated into the peritoneum during the aforesaid time periods. If the intended host animal is a mouse, the number of human hepatocytes represented in the lysate may be 1×10³-1×10⁵ and preferably 10⁴ and intraperitoneal inoculation may be performed between days 15 and 18 post conception. If the intended host animal is a pig, the number of human hepatocytes represented in the lysate may be between about 10⁵ and 10⁶ or the same number of whole cells and intraperitoneal inoculation may be performed at between about 75 and 90 days post-conception. Alternatively, intraperitoneal inoculation can be performed while the animals are neonates.

As a second non-limiting example, tolerance may be induced by intrathymic injection according to a method as described in Fabrega et al., 1995, Transplantation 59:1362-1364. Either whole cells or a cell lysate may be administered. In particular, where the intended host animal is a rat, about 1×10²-1×10⁵ human hepatocytes (or a lysate thereof) in between about 1 and 10 microliters, preferably about 5 microliters, may be injected into the thymus of a newborn (neonatal) rat, preferably within 1-2 hours of birth. Where the intended host animal is a mouse, about 1×10²-1×10⁴ and preferably 100 human hepatocytes (or a lysate thereof) in between about 1 and 10 microliters and preferably about 5 microliters may be injected into the thymus of a mouse that is up to 3 months old and preferably a neonate, e.g. within 1-2 hours or within 24 hours of birth. Where the intended host animal is a pig, about 10⁵-10⁶ human hepatocytes (or a lysate thereof) in between about 50 and 200 microliters may be injected into the thymus of an infant pig that is preferably up to one week old. As a specific example, a neonatal mouse may be anesthetized by chilling on ice, the thoracic area may be cleaned with alcohol and betadine swipes, the thymus may be visualized through the translucent skin of the newborn, and a 1-2 mm incision may be made with ophthalmic scissors to expose the thymus. The human cells or human cell lysate may then be slowly injected into the thymus, and then the incision may be closed with a sterile nylon suture. The incision area may then be recleaned and the mouse placed on a warming pad and returned to its mother as soon as possible.

The success of tolerization may be assessed by proceeding to introduce human hepatocytes into the animal, and determine whether or not they survive long-term (for example, by monitoring the production of human serum albumin; see infra). Alternatively, the ability of lymphocytes from the animal to react with donor human hepatocytes may be evaluated using standard immunologic techniques, such as methods that determine T cell proliferation in response to donor hepatocytes, the induction of a cytotoxic T cell response, or mixed lymphocyte reaction.

5.1.3. Introduction of Human Liver Cells

Human hepatocytes may then be introduced into host animals rendered tolerant as set forth in the preceding section. The hepatocytes may preferably be introduced via intrasplenic injection, although other routes may also be used, such as direct injection into the liver parenchyma, under the liver capsule, or via the portal vein.

As a specific non-limiting example, where the intended host animal is a rat tolerized as set forth above, between about 10⁶-5×10⁷ human hepatocytes, preferably about 2×10⁶ hepatocytes, may be introduced into a tolerized rat within about 24 hours after birth by anesthetizing the animal, making a 3-4 mm incision in the left paracostal area to visualize the spleen (Marucci et al., 1997, Hepatol. 26:1195-1202), and injecting the donor cells in a volume of approximately about 50-300 microliters, and preferably about 200 microliters, of sterile medium. Where the intended host animal is a tolerized mouse, the number of human hepatocytes introduced by an analogous procedure may be between about 5×10³ and 5×10⁶, preferably about 10⁵ in a volume of about 25-200 microliters, and preferably about 100 microliters, of sterile medium, and the human hepatocytes are administered between about one day and two months, preferably 3-4 days, after tolerization. Where the intended host animal is a tolerized pig, the number of human hepatocytes may be between about 10⁸-10¹⁰, preferably about 10⁹, in a volume of about 10-20 milliliters of sterile medium and the human hepatocytes are administered about one and seven days after birth or about 35 days after tolerization.

Human hepatocytes may be obtained from a commercial source, for example, Clonetics Corporation, 8830 Biggs Ford Road, Walkersville, Md. 21793, which sells normal human hepatocytes as catalog number CC-2591, or Invitro Technologies, Inc., Baltimore, Md.

Alternatively, human hepatocytes may be prepared from a donor as follows. The source of cells may be from a liver biopsy taken percutaneously or via abdominal surgery, or from liver tissue obtained postmortem. The source of cells should be maintained in a manner which protects cell viability. In one specific non-limiting embodiment, human hepatocytes may be prepared using the technique described in Guguen-Guillouzo et al., 1982, "High yield preparation of isolated human adult hepatocytes by enzymatic perfusion of the liver", Cell Biol. Int. Rep. 6:625-628. Briefly, the method of Guguen-Guillouzo et al. involves (i) isolating a liver or a portion thereof from which hepatocytes are to be harvested; (ii) introducing a cannula into the portal vein or a portal branch; (iii) perfusing the liver tissue, via the canula, with a calcium-free buffer followed by an enzymatic solution containing 0.025% collagenase (e.g., Type 4, from Sigma Chemical Company) in 0.075% calcium chloride solution in HEPES buffer at a flow rate of between 30 and 70 milliliters per minute at 37° C.; then (iv) mincing the perfused liver tissue into small (e.g. about 1 cubic millimeter) pieces; (v) continuing the enzymatic digestion in the same buffer as used in step (iii) for about 10-20 minutes with gentle stirring at 37° C. to produce a cell suspension; and (iv) collecting the released hepatocytes by filtering the cell suspension produced in step (v) through a 60-80 micrometer nylon mesh. The collected hepatocytes may then be washed three times in cold HEPES buffer at pH 7.0 using slow centrifugation (e.g., 50×g for five minutes) to remove collagenase and cell debris. Non-parenchymal cells may be removed by metrizamide gradient centrifugation. If the amount of liver tissue is too small to perform the above perfusion procedure, for example, less than 100 g of tissue, then the tissue may be minced and digested with collagenase solution with gentle stirring and processed according to steps (iv) and (v) of this paragraph.

It may be desirable to separate human hepatocytes prepared as set forth above into a subset for introduction into animals and another subset which is undesirable to propagate. For example, if a human subject is to serve as a donor for hepatocytes which are to be propagated in a chimeric animal according to the invention and then reintroduced into the subject, e.g., to reconstitute a liver damaged by infectious disease or malignancy, it would be desirable not to propagate hepatocytes which are infected or which have undergone malignant transformation. In such a situation, it would be desirable to eliminate infected or malignant hepatocytes from the population of hepatocytes which is to be introduced into the host animal. Elimination of unwanted cells can be performed by standard cell sorting techniques, for example fluorescence activated cell sorting using an antibody specific for the infectious agent or for malignant transformation. Alternatively, undesirable cells may be eliminated or attenuated by treatment with antiviral or antimicrobial compounds, radiation, antibody-ligated toxins, culture techniques, etc.

Where a differentiated hepatocyte cell line is to be used for transplantation, such as, but not limited to, Huh7 or HepG2 cells, the cell lines may be obtained from a standard laboratory source (see Liang et al., 1993, J. Clin. Invest. 91:1241-1246). For example, Huh7 may be obtained from individual investigators. HepG2 has the American Type Culture Collection ("ATCC") Accession Number HB-8065; the address of the ATCC is 10801 University Blvd., Manassas, Va. 20110-2209.

5.1.4. Favoring Proliferation of Human Hepatocytes

In particular non-limiting embodiments of the invention, selection pressure may be used to favor the proliferation of human hepatocytes. Such selection pressure is defined herein as including any condition, preexisting in the host animal at the time of introduction of donor cells or imposed thereafter, which results in a greater likelihood that human hepatocytes, rather than host hepatocytes, will proliferate.

For example, the selection pressure may result from the presence of a transgene that decreases the viability of host hepatocytes, either intrinsically (directly) or by administration of an activating agent (indirectly). Alternatively, human donor hepatocytes can be transfected with a protective gene that will enable those cells to survive subsequent exposure to a hepatotoxin. In one specific non-limiting example, the transgene may be the albumin promoter/urokinase construct, whereby as the host animal matures and the albumin promoter becomes active, host hepatocytes may be eliminated by the toxic effects of urokinase. In such cases, the selection pressure is maturation of the animal with consequent transgene activation. In a second specific non-limiting example, the transgene may be the albumin promoter/HSV-TK construct, whereby when gancyclovir is administered to the host animal (e.g., as an intraperitoneal injection of 250 mg/kg gancyclovir in sterile PBS), hepatocytes of the transgenic host may be selectively killed. In such embodiments, the death of host hepatocytes would be expected to favor compensatory proliferation of human hepatocytes. This can occur because of the known property of parenchymal liver cells to proliferate during conditions that stimulate regeneration.

It may be preferable to effect stepwise attenuation of host hepatocytes rather than eliminate a majority in a short period of time, as the sudden loss of liver function could result in death of the animal and/or conditions that would disfavor the establishment of a human hepatocyte population in the host liver. For example, administration of several doses of gancyclovir to a host animal transgenic for the albumin promoter/HSV-TK construct, beginning before and continuing after introduction of donor cells, may result in a gradual elimination of host cells, thereby permitting human hepatocytes to establish a "foothold" before the majority of host hepatocyte function is eliminated.

In another non-limiting embodiment, donor hepatocytes can be transfected with a protective gene. For example, a gene encoding an antisense RNA or ribozyme against the cytochromes 2E1, 1A2, and/or 3A4 (CYP2E1, CYP1A2, CYP3A4, respectively), would prevent activation of the drug acetaminophen. Metabolites of that agent within liver cells results in hepatocyte death. Thus, donor cells containing the transgene would have a survival advantage relative to host cells if massive doses of acetaminophen were administered after cell transplantation. A similar strategy would be to transfect a mutant RNA polymerase II that is resistant to the effects of the hepatotoxin phalloidin. Administration of phalloidin to hosts bearing transfected human hepatocytes producing the mutant polymerase would be protected and have a selective advantage over host cells.

5.1.5. Confirming the Presence of Human Hepatocytes

The presence of human hepatocytes in a host may be evaluated by assaying for specific human markers. The presence of such markers in a blood sample or a liver biopsy collected from the animal (e.g., percutaneously) may be evaluated without affecting the viability of the animal. Alternatively, the success of chimerization may be evaluated retrospectively at necropsy.

As a specific example, the presence or absence of immunologically distinct human albumin may be determined in a blood or tissue sample by Western blot analysis or immunohistochemistry using antibody specific for human, but not host, albumin (see, for example, Wu et al., 1991, J. Biol. Chem. 266:14338-14342; Osborn and Weber, 1982, Meth. Cell Biol. 24:97-132). An example of a publicly available antibody specific for human albumin is Sigma #A6684 monoclonal anti-human albumin HSA II.

5.2. Toxicology Model Systems

In particular non-limiting embodiments of the invention, a chimeric animal prepared as set forth above may be used as a model system for human hepatocyte function in a toxicology study to determine the toxic effect(s) of a test agent on (i) the human hepatocytes present in the animal and/or (ii) the host animal itself. The chimeric animals of the invention provide the opportunity to recapitulate, in a model system, metabolism of the test agent by human hepatocytes, which may result in one or more secondary compounds that may not be produced when the test agent is exposed to non-human hepatocytes.

Because a test agent may have different effects on host hepatocytes and human hepatocytes, it is desirable to determine the relative proportion of human and host hepatocytes in each test animal, for example by quantitation of the amounts of human and non-human albumin in a serum sample. The ability of this measurement to accurately reflect liver cell populations may be established by correlating serum albumin levels with hepatocyte populations as evaluated by immunohistochemistry in liver tissue samples obtained by biopsy or at necropsy. Once the relative proportions of hepatocyte populations for each animal are determined, experimental results relating to the effect of test agent may be compared with the effect of test agent on a control non-chimeric animal which represents a population of 100 percent host hepatocytes. Preferably, the host hepatocytes are less sensitive to test agent than human hepatocytes.

Accordingly, chimeric animals of the invention may be used to evaluate the toxic effect(s) of a test agent on the viability (survival, function) of human hepatocytes in the animal and/or the animal as a whole by subjecting at least one and preferably a plurality of chimeric animals and non-chimeric animals of the same species (as controls) to incremental doses of test agent. At one or a series of time point(s), the animal(s) may be evaluated by standard laboratory tests to determine whether toxic effects have occurred. Such an evaluation may include an assessment of bodily functions, as reflected by weight and/or activity and analysis of blood and/or urine, for example for test agent or its metabolites, markers of liver function and/or hepatocyte viability, kidney function, immune function, etc. As discussed above, such information is considered in view of the percentage of human hepatocytes in each test animal's liver and the relative effects of test agent on human versus host hepatocytes. Further, the percentage of human hepatocytes may change during the course of an experiment, for example, if the test agent is selectively toxic to human hepatocytes so that compensatory proliferation of host hepatocytes occurs. Accordingly, it is desirable to perform measurements of relative quantities of one or more marker specific for human hepatocytes at each time point; for example, the relative amounts of human and host albumin in serum may be measured by Western blot. At one or more time point of the study, an animal(s) may be biopsied and analyzed for human versus host albumin gene or gene product, or human-specific Alu repeat sequence, or sacrificed and a complete necropsy analysis be performed, including immunohistochemical evaluation of hepatocyte populations in the liver.

5.3. Model Systems for Liver Disease

In another non-limiting embodiment of the invention, an animal having a chimeric liver may be used as a model system for human liver disease. Such chimeric animals may be used to create models of liver disease resulting from exposure to a toxin, infectious disease or malignancy. The model systems of the invention may be used to gain a better understanding of these diseases and also to identify agents which may prevent, retard or reverse the disease processes.

Where the chimeric animal is to be used as a model for liver disease caused by a toxin, animals prepared as set forth above may be allowed to mature to a point where the size of the human hepatocyte population is substantial (e.g. has approached a maximum), and then be exposed to a toxic agent. The amount of toxic agent required to produce results most closely mimicking the corresponding human condition may be determined by using a number of chimeric animals exposed to incremental doses of toxic agent. Examples of toxic agents include but are not limited to alcohol, acetaminophen, phenytoin, methyldopa, isoniazid, carbon tetrachloride, yellow phosphorous, and phalloidin.

In embodiments where a chimeric animal is to be used as a model for malignant liver disease, the malignancy may be produced by exposure to a transforming agent or by the introduction of malignant cells. The transforming agent or malignant cells may be introduced with the initial colonizing introduction of human hepatocytes or, preferably, after the human hepatocytes have begun to proliferate in the host animal. In the case of a transforming agent, it may be preferable to administer the agent at a time when human hepatocytes are actively proliferating. Examples of transforming agents include aflatoxin, dimethylnitrosamine, and a choline-deficient diet containing 0.05-0.1% w/w DL-ethionine (Farber and Sarma, 1987, in Concepts and Theories in Carcinogenesis, Maskens et al., eds, Elsevier, Amsterdam, pp. 185-220). Such transforming agents may be administered either systemically to the animal or locally into the liver itself. Malignant cells may preferably be inoculated directly into the liver.

Where the chimeric animal is to be used as a model for infectious liver disease, the infectious agent, or an appropriate portion thereof (e.g. a nucleic acid fragment) may be introduced with the initial introduction of hepatocytes or after the human hepatocytes have begun to proliferate. The infectious agent may be administered as a free entity or incorporated into a human cell such as a human liver cell. Examples of infectious diseases suitable for modeling include but are not limited to hepatitis A, hepatitis B, hepatitis C, hepatitis D, hepatitis E, malaria, Epstein Barr infection, cytomegalovirus infection. and Yellow Fever. For such models, it may be advantageous that the host animal has an immune system that is intact (but for the induced tolerance to the host cells), in that the animal's immune response to the infectious agent and/or infected human hepatocytes may produce a more accurate model of human liver diseases in which the immune system plays a pathogenic role. As such, it may be desirable to ensure that the cells/cell lysate used for tolerization not include infectious agent or related antigens. A working example in which the invention is used to produce a hepatitis B virus model system is set forth below.

Further, where the infectious agent is a virus, the present invention provides for chimeric animals comprising human hepatocytes that contain a nucleic acid of the virus, such as the entire viral genome or a portion thereof, or a nucleic acid encoded by the viral genome or a portion thereof.

5.3.1. HCV Model Produced by Infectious Serum

In a particular non-limiting embodiment, the invention provides for a chimeric animal model for hepatitis C virus infection.

In one set of preferred embodiments, the host animal is tolerized and subsequently transplanted with cells of a differentiated human hepatocyte cell line. In one specific, non-limiting example of such embodiments, the cell line is Huh7. In a more specific non-limiting example, the chimeric animal is a rat tolerized and transplanted with Huh7 cells.

In another non-limiting set of embodiments, the chimeric animal is a mouse transgenic for a gene whose product is selectively toxic to hepatocytes, such as the albumin promoter/urokinase gene or the albumin promoter/HSV-TK gene. Hepatitis C infection of human hepatocytes in such mice may be produced either (i) concurrently with or preferably (ii) after the colonizing introduction of human hepatocytes and after the effects of the toxic transgene have attenuated or eliminated host hepatocytes. Preferably, the chimeric animal has, prior to infection, a liver which comprises substantially (at least about 20 percent, preferably at least 50 percent, more preferably at least 80 percent) human hepatocytes.

The source of infectious agent may be serum from one or more human subject infected with HCV but not demonstrably infected with one or more other agents that infect hepatocytes. Serum samples of genotype Ia may be assayed for viral load by branched DNA (bDNA) assay (Chiron, San Francisco, Calif.). Sera from non-infected subjects and individuals with non-viral hepatitis may be used to pseudo-infect control chimeric animals. Using standard biohazard precautions, serum containing HCV RNA from infectious human serum, at a titer ranging between about 10³-10⁷ particles per milliliter may be injected intravenously into a chimeric animal about 2-4 months and preferably about 6 weeks after colonization with human hepatocytes. Preferably, increasing amounts of HCV RNA in infectious human serum, with the viral titer previously determined (e.g., by National Genetics Institute, Los Angeles, Calif.) may be injected into a panel of such chimeric animals. Where the chimeric animal is a rodent, the site of injection may be the tail vein, and the volume of serum injected may be 0.1-0.5 ml. The serum may preferably be filter sterilized prior to administration. In a preferred embodiment, a chimeric rat is anesthetized, its spleen is exposed, and 100,000 copies of HCV/0.1 ml serum is injected into the spleen; pressure is applied at the injection site and then the incision is closed.

Serum may be collected from the chimeric animal(s) and tested to establish baseline and post-infection levels of liver function markers such as AST (aspartate amino transferase), ALT (alanine aminotransferase) and alkaline phosphatase. For example, baseline and weekly post-infection levels of AST, ALT and alkaline phosphatase in serum may be determined spectrophotometrically using kits from Sigma Chemical Co., St. Louis, Mo., where appropriate standards are used to generate reference curves. Where the animals are rodents, blood samples may be obtained retroorbitally using standard techniques.

The chimeric animal(s) may be tested for seroconversion against HCV by testing for circulating antibody (e.g., anti-C100-3 antibody), for example using the ELISA kit available from Ortho Diagnostics (catalog number 930740: Ortho HCV ver. 3.1 ELISA TEST SYSTEM; Ortho Diagnostics, Raritan, N.J.). Tests for seroconversion may be performed, for example, at weekly intervals for the first month after infection and then monthly.

Viral load may be determined (e.g., weekly) by assay of dilutions of serum for positive strand HCV RNA using thermostable rTth RT-PCR performed under stringent conditions (at 70° C.) to eliminate false priming of the incorrect strand. Branched DNA analysis may also be used, but it is not as sensitive. For positive strand RNA analysis, the cDNA reverse primer may be: 5′-TCGCGACCCA ACACTACTC 3′ (SEQ ID NO: 2) and the forward primer may be 5′-GGGGGCGACA CTCCACCA-3′ (SEQ ID NO: 3). PCR amplification in the absence of reverse transcriptase activity may be accomplished by chelating manganese and magnesium ions as described in (Lanford et al., 1995, J. Virol. 69:8079-8083). The amplified product, which spans nucleotides 15-274 of the 5′-NTR of HCV may be quantitated by Southern blotting using a detectably labeled probe against a region internal to the primers.

Liver tissue obtained by biopsy or from a sacrificed animal may be evaluated for HCV replication and for histopathological changes. Biopsy may be performed by anesthetizing the chimeric animal with intramuscular injections of ketamine (40 mg/kg) and xylazine (5 mg/kg), cleaning the abdominal area with alcohol and betadine wipes, making an incision in the abdominal wall to expose the liver, and collecting a sliver (weighing at least approximately 10 mg) of liver tissue. Afterward, 100 U of sterile thrombin (or another therapeutically effective amount, as needed) may be administered locally at the biopsy site followed by application of gel foam to inhibit bleeding, the abdominal wall may be closed with dissolvable sutures, and the skin may be closed with nylon sutures. Viral replication may be quantitated by measuring the amount of negative strand template HCV RNA in liver RNA (prepared, for example, as set forth in Chomczynski and Sacchi, 1987, Anal. Biochem. 162:156-159), using rTth RT-PCR (Lanford et al., 1995, J. Virol. 69:8079-8083). To assess liver histology, liver tissue may be fixed and sectioned and stained with hematoxylin-eosin or trichrome to evaluate, respectively, inflammation or fibrosis. A standardized scoring method, such as Knodell scoring (Knodell et al., 1981, Hepatology 1:531), may be used. The presence or absence of neoplastic lesions may be evaluated.

To determine the optimum conditions for producing an HCV infected chimeric animal, the time course of serum aminotransferases AST and ALT, alkaline phosphatase levels, and viral RNA loads may be plotted as a function of time and the minimum number of viral equivalents required to sustain an infection determined. Levels of detectable HCV RNA in the serum of an animal may be used as an indicator of the chronicity of infection.

Potential problems associated with the foregoing embodiment are as follows. First, the detection of negative strand HCV template as a measure of HCV replication may be problematic due to the requirement for amplification techniques and the possibility of inadvertent amplification of positive strand. The method of Lanford et al. (supra) using stringent conditions for priming of the RT-PCR and inactivation of the reverse transcriptase by chelation prior to PCR of the cDNA has been shown to reduce false amplification to 1/10⁴-1/10⁵. Second, laboratory animals may harbor an endogenous virus which causes hepatitis (for example, as regards laboratory mice as hosts, the fact that mouse hepatitis virus may be found even in "pathogen free" environments makes it desirable to confirm that host mice are free of the virus, for example using a mouse virus screen as available from Microbiological Associates, Inc., Rockville, Md. (Carlson et al., 1989, J. Clin. Invest. 83:1183-1190)), where animals testing positive are not used as hosts. Third, infection may be improved by increasing the amount of human serum used in the inoculum.

A working example of a chimeric rat model of HCV infection is set forth in Example Section 13, infra.

5.3.2. HCV Model Produced by Infectious Plasmid

In a related embodiment, infection may be introduced by HCV plasmid (Kolykhalov et al., 1997, Science 277:570-574) complexed to a liver-specific protein carrier, such as AsOR-PL or AsORlysine-VSVG, where AsOR-PL is asialoorosomucoid polylysine and AsORlysine-VSVG is asialoorosomucoid covalently linked to L-lysine methyl ester and a synthetic 25 amino acid peptide of the VSVG protein. The DNA-protein complex may be formed by slowly adding protein conjugate in 25 microliter aliquots to DNA in 0.15M NaCl with continuous vortexing at room temperature. After 30 minutes of incubation at room temperature absorption at 260 nm, 340 nm and 400 nm may be measured to detect complex formation. Complexes may be filter sterilized by passage through a 0.22 micron filter. An amount of DNA/protein complex may then be administered. About 10-50 micrograms of the DNA/protein complex in 0.5 milliliters sterile saline may then be injected into the tail vein of a mouse, and 100-500 micrograms of DNA/protein complex in a volume of 1-5 mls may be injected into a rat.

5.3.3. HCV Model Produced by Transplanting Infected Hepatocytes

As an alternative to producing HCV infection by inoculation with infected serum, infection may be produced by transplanting HCV infected hepatocytes into a chimeric animal. Although the infected hepatocytes may be introduced during colonization with human cells, it is preferred that they be introduced into chimeric livers having a substantial population of human hepatocytes. In one non-limiting set of embodiments, the chimeric animal is a mouse transgenic for a gene whose product is selectively toxic to hepatocytes, such as the albumin promoter/urokinase gene or the albumin promoter/HSV-TK gene. In another set of non-limiting embodiments, the chimeric animal is a rat tolerized and transplanted with Huh7 cells.

Infected human hepatocytes may be obtained as described in Lieber et al., 1996, J. Virol. 70:8782-8791. Using appropriate pathogen-containment procedures, human liver specimens may be obtained from HCV-infected liver transplant recipients. An apical piece of liver covered on three sides by capsule may be perfused with buffer without calcium and then with collagenase in perfusion buffer with calcium. Hepatocytes may then be pelleted by low speed centrifugation. Non-parenchymal cells may be separated from parenchymal hepatocytes by metrizamide gradient centrifugation. The viability of isolated hepatocytes may be evaluated by trypan blue exclusion. Hepatocytes may be resuspended in Williams medium at about 10⁷ cells per milliliter.

The infected hepatocytes may then be introduced into the liver of a chimeric animal, for example a chimeric animal whose liver comprises at least about 20 percent, preferably at least 50 percent, more preferably at least 80 percent) human hepatocytes. The infected hepatocytes may be introduced by intrasplenic injection. Where the animal is a mouse, hepatocytes may be introduced by anesthetizing the animal with ketamine (90 mg/kg)/xylazine (10 mg/kg), and then, under aseptic conditions, making a 2-3 millimeter incision in the left paracostal area, exposing the spleen. The spleen may then be exteriorized and infected hepatocytes may be injected slowly into the spleen parenchyma. Gel foam may be used to achieve hemostasis, the spleen may be restored into the body cavity, and the wound may be sutured closed. Monitoring of the resulting infected animals for serconversion, viral load, serum levels of protein markers of liver function, and histopathology may be performed as described in section 5.3.1. Further, these methods may be adapted for use in larger animals.

5.3.4. Use of HCV Models

Chimeric animal models of HCV infection may be used not only to study the biology of HCV, but also to identify agents that may prevent or inhibit HCV infection and/or replication. For example, to determine whether a test agent inhibits infection by HCV, the effect of the agent on preventing infection when administered prior to or contemporaneously with injection of infected serum may be evaluated. Similarly, the effect of a test agent administered during the course of infection may be assessed. Parameters useful in determining the effectiveness of test agent would include whether and when the test animal seroconverts with respect to HCV, the viral load, the ability of serum from the animal to infect other animals, blood levels of proteins/enzymes associated with liver function and/or hepatocyte viability, and liver histology.

5.4. Chimeric Animals as a Source of Hepatocytes for Liver Reconstitution

The present invention further provides for the use of chimeric animals as a source of human hepatocytes for liver reconstitution in a second host subject. Such reconstitution may be used, for example, to (i) produce a "next generation" chimeric non-human animal; (ii) introduce genetically modified hepatocytes for "gene therapy" of the second host subject; or (iii) replace hepatocytes lost as a result of disease, physical or chemical injury, or malignancy in the second host. Human hepatocytes collected from a chimeric animal are said to be "passaged".

For any of these applications, liver tissue from a chimeric animal may be used to produce a cell suspension and then human hepatocytes may be separated from non-human hepatocytes and other cells. The liver tissue may be processed as set forth above to produce a suspension of hepatocytes. As a non-limiting specific example, where the chimeric animal is a mouse or rat, hepatocytes may be prepared by the following method, adapted from Seglen, 1976, "Preparation of rat liver cells", Methods Cell Biol. 13:29. Briefly, a chimeric mouse or rat may be anesthetized with ketamine/xylazine, its abdomen may be shaved and decontaminated, the peritoneal cavity may be opened by incision, the inferior vena cava may be cannulated, the portal vein may be divided and the suprahepatic vena cava may be ligated. Then, the liver may be perfused in situ with calcium free balanced salt solution at 5 ml/min for five minutes at 37° C., followed by perfusion with 0.05% collagenase (e.g., type IV, from Sigma Chemical Co.) in 1% albumin and balanced salt solution for 20 minutes. The liver may then be transferred to a Petri dish, and minced to produce a cell suspension, from which hepatocytes may be collected by passage through a 60-80 micron nylon mesh. The collected cells may then be washed three times in RPMI 1640 or Williams E medium with 10% fetal bovine serum, and then centrifuged at 35×g for five minutes at 4° C. Hepatocytes may be purified through a metrizamide gradient and resuspended in RPMI 1640 or Williams E medium.

Human hepatocytes may be separated from non-human cells using fluorescence activated cell sorting techniques and an antibody which selectively binds to human hepatocytes, for example but not by way of limitation, an antibody that specifically binds to a class I major histocompatibility antigen. Suitable antibodies would include but not be limited to anti-human HLA-A,B,C, PharMingen catalogue #32294X or #32295X, FITC mouse κb, PharMingen catologue #06104D (PharMingen, San Diego, Calif.) See, for example, the procedure described in Markus et al., 1997, Cell Transplantation 6:455-462.

Human hepatocytes may be passaged through cell transplantation of tolerized host animals, using the techniques set forth above. In this manner, cells obtained from an initial human donor may be utilized in a multitude of chimeric animals and over an extended period of time, potentially reducing the variability that may be encountered in chimeric animals produced using hepatocytes obtained from diverse hosts.

Passaged human hepatocytes may also be used for gene therapy applications. In the broadest sense, such hepatocytes are transplanted into a human host to correct a genetic defect. The passaged hepatocytes need not, but are preferably derived originally from the same individual who is to be the recipient of the transplant. However, according to the invention, hepatocytes from a different individual may alternatively be used.

As a specific, non-limiting example, a patient suffering from intermittent acute porphyria, caused by a genetic defect in the expression of uroporphyrinogen I synthase, may benefit from transplantation of human hepatocytes harvested from a chimeric animal of the invention, where the transplanted cells are genetically normal in their expression of that enzyme. The recipient would be "matched" for transplantation antigens with the original donor, or be treated with immunosuppressive therapy. For such applications, chimeric animals prepared from a wide diversity of individual donors could provide the advantage of constituting a "living library" of differentiated hepatocytes having various transplantation antigen profiles, thereby obviating the need for waiting until liver tissue from a genetically suitable donor becomes available.

Preferably, however, the original donor and eventual recipient of passaged hepatocytes are the same person, thereby eliminating the need for immunosuppression. For gene therapy applications, (i) hepatocytes may be harvested from the subject, (ii) the desired genetic construct may be introduced into those hepatocytes, (iii) the resulting genetically engineered human hepatocytes may be used to tolerize a host animal to their presence, (iv) construct-carrying hepatocytes may be introduced into the tolerized animal such that its liver is colonized, and then, once expanded in number, (v) the transgenic hepatocytes may be harvested from the chimeric animal and (vi) reintroduced into the subject. A genetic construct may be introduced into the human hepatocytes by any standard method, including, but not limited to, transfection with naked DNA, microparticles or liposomes, or infection with a viral vector, such as an adenoviral vector, an adeno-associated vector, or a retroviral vector. Hepatocytes used for colonization may be enriched for cells containing the desired construct, for example, by selection by culture conditions, antibody/FACS methods, etc. which eliminate cells lacking the construct.

Alternatively, the hepatocytes may be used to colonize the liver of a tolerized animal prior to or contemporaneous with the introduction of the desired transgene via a gene therapy vector. This approach may be more problematic because the host animal could develop an immune response directed toward either the vector or vector-transformed hepatocytes.

In further embodiments, human hepatocytes passaged through a chimeric animal of the invention may be used to reconstitute liver tissue in a subject as a prelude or an alternative to liver transplant. As a specific non-limiting example, a subject suffering from progressive degeneration of the liver, for example, as a result of alcoholism, may serve as a donor of hepatocytes which are then maintained, through one or several generations, in one or more chimeric animal. As a result of maintenance in such animal(s), the number of hepatocytes is expanded relative to the number originally harvested from the subject (it may be preferable to use larger animals to produce greater numbers of cells). At some later date, when the subject's liver has deteriorated to a medically hazardous condition, hepatocytes passaged in the chimeric animal(s) may be used to reconstitute the subject's liver function. As a second non-limiting example, passaging hepatocytes may be used not only to expand the number of hepatocytes but also to selectively remove hepatocytes that are afflicted with infectious or malignant disease. Specifically, a subject may be suffering from hepatitis, where some but not all of the hepatocytes are infected and infected hepatocytes can be identified by the presence of viral antigens on the cell surface. In such an instance, hepatocytes may be collected from the subject, and non-infected cells may be selected for passaging in one or more chimeric animal, for example by FACS. Meanwhile, aggressive steps could be taken to eliminate infection in the patient. Afterward, the subjects liver tissue may be reconstituted by hepatocytes passaged in a chimeric animal. An analogous method could be used to selectively passage non-malignant cells from a patient suffering from a primary or secondary (e.g. metastatic) liver malignancy. Thus, the chimeric animals of the invention may be used as a means of purging unwanted hepatocytes from a human subject.
 

Claim 1 of 6 Claims

1. A model system for Hepatitis C virus infection In humans, comprising a non-human mammal, wherein the mammal is immunocompetent but has been rendered inmunologically tolerant to human hepatocytes by fetal tolerization and subsequently transplanted with human hepatocytes and infected with Hepatitis C virus, whereby replication of Hepatitis C virus occurs in the model system.

____________________________________________
If you want to learn more about this patent, please go directly to the U.S. Patent and Trademark Office Web site to access the full patent.