Here, we present a novel humanized mouse liver model generated in Alb-toxin receptor mediated cell knockout (TRECK)/SCID mice following the transplantation of immature and expandable human hepatic stem cells.
A novel animal model involving chimeric mice with humanized livers established via human hepatocyte transplantation has been developed. These mice, in which the liver has been repopulated with functional human hepatocytes, could serve as a useful tool for investigating human hepatic cell biology, drug metabolism, and other preclinical applications. One of the key factors required for successful transplantation of human hepatocytes into mice is the elimination of the endogenous hepatocytes to prevent competition with the human cells and provide a suitable space and microenvironment for promoting human donor cell expansion and differentiation. To date, two major liver injury mouse models utilizing fumarylacetoacetate hydrolase (Fah) and uroplasminogen activator (uPA) mice have been established. However, Fah mice are used mainly with mature hepatocytes and the application of the uPA model is limited by decreased breeding. To overcome these limitations, Alb-toxin receptor mediated cell knockout (TRECK)/SCID mice were used for in vivo differentiation of immature human hepatocytes and humanized liver generation. Human hepatic stem cells (HpSCs) successfully repopulated the livers of Alb-TRECK/SCID mice that had developed lethal fulminant hepatic failure following diphtheria toxin (DT) treatment. This model of a humanized liver in Alb-TRECK/SCID mice will have functional applications in studies involving drug metabolism and drug-drug interactions and will promote other in vivo and in vitro studies.
Mice are commonly used for pharmaceutical testing since biomedical research in humans is restricted1; however, these models are not always useful since they may inaccurately simulate the effects observed in humans. Most drugs in current medical use are metabolized primarily in the liver. However, the same drug can be metabolized into different metabolites in mouse and human livers because of inter-species differences. Thus, it is often difficult to determine during development whether a potential drug poses any risks for clinical applications2,3.
To address this problem, “humanized” mouse livers have been developed by growing human liver cells inside mice4-6; these models exhibit drug responses similar to those observed in the human liver. The primary mouse models currently used for humanized liver generation include uroplasminogen activator (uPA+/+) mice4,7, fumarylacetoacetate hydrolase (Fah−/−) mice6, and the recently reported thymidine kinase (TK-NOG) mice.
However, previous reports have shown that transplanted human immature cells or stem cells are less competitive than adult human hepatocytes in Alb-uPA tg(+/−)Rag2(−/−) mouse livers8-10. Moreover, Fah−/− mice provide a growth advantage only for differentiated hepatocytes and not for immature liver progenitor cells11. The transplantation of human hepatic stem cells (HpSCs) into TK-NOG mice in the lab has been unsuccessful. Hence, no useful mouse model for the efficient engraftment of human immature liver cells currently exists.
Thus, we developed a novel Alb-TRECK/SCID mouse model that could be efficiently repopulated with human immature hepatocytes. This transgenic mouse model expresses human heparin-binding EGF-like growth factor (HB-EGF) receptors under the control of a liver cell-specific albumin promoter. Following the administration of diphtheria toxin (DT), these mice develop fulminant hepatitis due to conditional ablation of hepatocytes, enabling donor cell residency and proliferation12. Although mouse hepatocytes have been successfully transplanted into Alb-TRECK/SCID mice in previous studies13,14, the generation of a humanized liver using Alb-TRECK/SCID mice has yet to be reported.
In this study, humanized livers were generated in Alb-TRECK/SCID mice via transplantation of HpSCs. This humanized liver provides an in vivo environment for universal stem cell differentiation and the ability to predict human drug metabolism patterns and drug-drug interactions.
All animal experimental procedures were performed according to the Animal Protection Guidelines of Yokohama City University.
1. Generation of the Acute Liver Injury Mouse Model
2. Preparation of Human Hepatic Stem Cells
3. Intrasplenic Transplantation of Human Hepatic Stem Cells
4. Detection of Transplanted Human Hepatic Stem Cell-Derived Hepatocytes in the Mouse Liver
Note: For the following procedures, euthanize all animals using an overdose of ketamine and xylazine followed by cervical dislocation.
5. Detection of Human Albumin Secretion and Calculation of the Chimeric Rate
Alb-TRECK/SCID mice hepatocytes express the human DT receptor HB-EGF gene under the control of an albumin promoter and exhibit cytotoxic effects following DT administration12. To evaluate the effects of DT treatment on liver injury, DT doses of 1.5 μg/kg were injected into 8-week-old Alb-TRECK/SCID mice and the pathological changes in the liver 48 hr post DT administration were histologically assessed. Compared with control mice (not treated with DT), the DT-treated mice exhibited a disorganized hepatic architecture showing a correction for congestion with increased serum AST activity. In addition, the hepatocytes of DT-treated mice had multiple, deeply stained acidophilic cytoplasmic inclusions together with dark nuclei (most probably apoptotic hepatocytes); some other hepatocytes appeared preserved or exhibited a vacuolated cytoplasm with dark nuclei with little or no portal vein inflammation (Figure 1). These findings are concordant with those of a previous study that showed that Alb-TRECK/SCID mice represent an ideal lethal fulminant hepatic failure model and can be generated using only a single DT injection12.
Human HpSCs can be used for long-term in vitro culturing and exhibit uniform cell morphology with some ALB and fewer CK19-positive cells. At 6 weeks post transplantation into the mouse livers, the human HpSCs were subjected to basic examination under a microscope and evaluated histologically; human HpSCs were observed to have reconstituted the liver structure by replacing the original mouse hepatocytes (Figure 2). Macroscopically, at 4 days post human HpSC transplantation, small human hepatic clusters that were uniformly distributed around the liver were detected; at 45 days post transplantation, these had proliferated into large clusters (Figure 3). Whole mouse livers reconstituted with human hepatocytes were designated "humanized livers." These results indicate that mice with lethal fulminant hepatic failure that undergo human HpSC transplantation may exhibit liver structures similar to those observed in normal mouse livers.
Prior to testing whether the immature human hepatocytes transplanted into mice had the potential for differentiation and could function in vivo, the presence of human hepatocytes in mouse livers was first confirmed by immunohistochemical analysis at approximately 6 weeks post human HpSC transplantation. Liver sections from mice with 60% human cell repopulation that specifically and positively co-stained with human nuclei and human cytokeratin 8/18 (CK8/18) antibodies were found to contain donor cell-derived human hepatocytes, whereas the original control mice livers were negative for these markers (Figure 4).
To assess the degree of cell differentiation in vivo, the expression of human albumin (ALB) and human cytokeratin 19 (CK19) was assessed immunohistochemically. Human HpSC-derived human hepatocytes were well differentiated in the mouse livers and exhibited upregulated human ALB expression. Human ALB-positive and CK19-negative hepatocytes resembled functional hepatocytes, whereas cells that positively co-stained with human ALB and CK19 exhibited bipotential capability for differentiating into hepatocytes and cholangiocytes (Figure 5). A large-scale scanning method used to analyze all human HpSC-derived humanized liver lobes revealed the presence of multiple round and colony-like clusters surrounding the liver lobes with clear human nuclei and CK8/18 expression (Figure 6), demonstrating the colony-forming capability of human HpSCs in the humanized livers. The repopulation level reached up to 90% one month following transplantation (Figure 7A) and human albumin secretion was detected in the repopulated mice livers (Figure 7B), indicating that human HpSCs could successfully reconstitute the Alb-TRECK/SCID mouse liver. These results demonstrate that human HpSCs have high potential for differentiation into functional hepatocytes in vivo in response to the rescue of defective mouse liver functions.
Figure 1. Liver Damage Occurs in Alb-TRECK/SCID (Toxin Receptor Mediated Cell Knockout) Mice Following Diphtheria Toxin (DT) Injection. Histology (hematoxylin-eosin staining) of mouse livers without and with DT administration after 48 hr. DT dose: 1.5 μg/kg, serum AST activity of normal liver (left panel): 30 IU/L; serum AST activity of DT-treated liver (right panel): 15000 IU/L. Scale bar = 100 µm. Please click here to view a larger version of this figure.
Figure 2. Transplantation of Human Hepatic Stem Cells into Alb-TRECK/SCID Mice with Fulminant Hepatic Failure. Macroscopic view (left panel) and histological analysis (hematoxylin-eosin staining, right panel) of humanized livers with human HpSCs at 6 weeks post transplantation. m, mouse liver region; h, human donor cell-derived human region. Scale bar = 100 µm. Please click here to view a larger version of this figure.
Figure 3. Macroscopic Analysis of Humanized Livers Derived from Human Hepatic Stem Cells Labeled with Green Fluorescent Protein. High-magnification images of the liver structure 4 days (left panel) and 45 days (right panel) following transplantation with human hepatic stem cells (HpSCs). Human hepatic stem cells (HpSCs) were labeled with enhanced green fluorescent protein (EGFP) using a lentivirus vector. Please click here to view a larger version of this figure.
Figure 4. Characterization of Human Hepatic Stem Cell-Derived Human Hepatocytes in Alb-TRECK/SCID Mice. Immunohistochemical analysis distinguishing between human hepatocytes stained with anti-human CK8/18 (green) antigen and anti-human nuclear antigen (aqua blue) in human hepatic stem cell-derived humanized livers at 6 weeks post transplantation. m, mouse liver region; h, human donor cell-derived human liver region. White dashed line: mouse liver region vs. human liver region. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (blue). Scale bars = 100 µm. Please click here to view a larger version of this figure.
Figure 5. Human Hepatic Stem Cells Differentiation in Humanized Livers of Alb-TRECK/SCID Mice. Immunohistochemical analyses of human albumin and human CK19 in HpSC-derived livers at 6 weeks post transplantation. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (blue). Scale bars = 100 µm. Please click here to view a larger version of this figure.
Figure 6. Multiple Clusters of Human Hepatic Stem Cell-Derived Cells are Present in Alb-TRECK/SCID Mice with Humanized Livers. Multiple large clusters derived from human hepatic stem cells were identified in the liver lobes at 6 weeks post transplantation by immunohistochemical analysis for human CK8/18 (green) and human nuclei (aqua blue). Scale bars = 1,000 µm. Please click here to view a larger version of this figure.
Figure 7. Humanized Liver Repopulation Rate and Human Albumin Secretion in Alb-TRECK/SCID Mice. (A) Chimeric rate in Alb-TRECK/SCID mice 1 month following human hepatic stem cell transplantation; data is presented as mean ± SEM (n = 12). Mann-Whitney test: P < 0.0005 for Sham and human HpSCs. (B) Detection of human albumin in mouse sera one month following donor cell transplantation. Data = mean ± SEM (n = 5). Mann-Whitney test: P = 0.0313 for Sham and human HpSCs. ND: undetectable; Sham: mice transplanted with PBS; HpSCTx: mice transplanted with human HpSCs. Please click here to view a larger version of this figure.
Recent studies have shown that the mouse liver can be repopulated with human hepatocytes, including adult hepatocytes and proliferative hepatic stem cells17. These repopulated livers have been used as preclinical experimental models for drug metabolism testing and drug discovery and development18; in addition, they have provided an in vivo environment for cell maturation and differentiation19. The major aim of the present study was to generate a novel liver disease mouse model that allowed human immature hepatocyte proliferation, maturation, and differentiation, for the acquisition of drug metabolism activities.
The course of intrasplenic transplantation has been studied extensively. Cells transplanted into the spleen of rats and mice survive throughout their average life span. The intrasplenic transplantation protocol has proven to be very useful in investigating specific research questions; many laboratories worldwide have attempted to utilize this protocol as a highly reproducible and reliable model for hepatic cell transplantation because of the improved mice survival and liver regeneration. However, implanting cells into the splenic pulp poses a risk, since the majority of implanted cells will translocate to the liver where they can occlude portal branches. Even though this occlusion can be temporary and might not affect a healthy liver environment, in cases of preexisting liver damage, this procedure could further reduce liver function and increase portal hypertension and might also negatively influence cell engraftment. Although we have successfully applied hepatocyte transplantation using intrasplenic transplantation, several issues warrant further study, especially the subject of potential new cell sources. Moreover, when performing intrasplenic hepatocyte transplantation, the cell loading limit must be strictly followed in order to prevent potential disturbance by remnant hepatic microcirculation. The DT receptor has been identified as a membrane-anchored form of the heparin-binding EGF-like growth factor (HB-EGF precursor)20. Whereas HB-EGF precursors derived from toxin-sensitive animals, such as humans and monkeys, bind DT and function as toxin receptors, HB-EGF precursors from mice and rats do not bind DT21. The transgenic Alb-TRECK/SCID mouse model expresses human HB-EGF-like receptors under the control of a liver cell-specific albumin promoter; following the administration of DT, these mice develop fulminant hepatic failure due to conditional ablation of hepatocytes, providing space for donor cell residency and proliferation. Human hepatocytes, even at an immature phase, are known to possess extensive proliferative capacity in vitro, although they lose their drug metabolism functions, limiting their preclinical applications22. Histological and immunohistochemical findings obtained following the transplantation of human HpSCs into Alb-TRECK/SCID mice with lethal fulminant hepatic failure indicated that the human HpSCs could expand and reconstitute the damaged mouse liver structures; the repopulation rate in some mice was nearly 100%.
In our experiments, human HpSC transplantation promoted mouse survival and resulted in increased secretion of human ALB into mouse sera16. These cells also matured and differentiated in vivo and exhibited human drug metabolism activities, indicating that human HpSC-derived humanized mice livers are similar to mature and functional "human organs" and have potential applications in drug development. These results further confirm that the Alb-TRECK/SCID mouse is an ideal model for humanized liver generation. Despite the prospects and advantages of Alb-TRECK/SCID mice as an ideal model for human hepatic stem cell-derived humanized liver generation, they possess one unique disadvantage; when human adult hepatocytes were transplanted into three Alb-TRECK/SCID mice treated with a single dose of DT, no humanized livers were generated, possibly because human adult hepatocytes lack proliferative capacity. Additional doses of DT cannot be administered post transplantation of human adult hepatocyte; since the DT receptor (HB-EGF) in Alb-TRECK/SCID mice hepatocytes is under the control of the albumin promoter, additional DT treatment will destroy the transplanted human adult hepatocytes in the mouse liver.
To the best of our knowledge, this is the first model to generate a humanized liver from immature human cells. Cell replacement rate post-transplant is influenced by the recipient micro-environment (e.g., hepatic niche) and donor cell properties, such as mature vs. immature and proliferative and differential capability. Combining Alb-TRECK/SCID with stem cell transplantation will provide optimized conditions and advantages for the generation of humanized livers. This also constitutes a powerful tool for transplantation of fetal liver cells and potentially as well as iPS or ES-derived hepatic cells since prior to DT treatment, the Alb-TRECK/SCID animals are maintained in a healthy state, liver damage is inducible at any point, and the mice breed normally without renal damage. In contrast, uPA/SCID mice exhibit high neonatal mortality and are difficult to breed. Furthermore, overexpression of urokinase in the animals can lead to severe bleeding and renal damage and thus there is a limited window of opportunity for cell transplantation prior to weaning; Alb-TRECK/SCID can be used for cell transplantation at any age.
The successful generation of a humanized liver mouse requires the following critical steps: 1) Liver injury should be limited to sub-lethal levels (usually by using the lethal dose, 50%); 2) The human primary fetal liver cells should be maintained in a sub-confluent status no greater than 90%; 3) Cell leakage post intrasplenic-transplantation should be prevented; and 4) The levels of human albumin in mouse sera should be monitored at regular intervals. A comprehensive understanding of the critical factors involved in the process of humanized mouse liver generation might contribute to further modifications or new developments utilizing rats or other animals. Possible adaptations include: the induction of kidney injury for examining development and repair with human nephrons, their progenitors, and collecting duct cells and the self-assembly of human nephrons into three dimensional structures in the mouse body; studying the repair of islet damage in the pancreas using pancreatic stem/progenitor cells derived from directly reprogrammed, iPS, or ES cells. Examples of process modifications include: controlling tissue or organ injury with toxins or chemicals such as retrorsine, streptozotocin, or cisplatin to facilitate the competition of donor cells with resident cells; avoiding differential or decreasing proliferation in order to enhance the proliferative capability of the donor cells, thus promoting the replacement of recipient cells, such as ES or iPS- derived cells; selecting a suitable transplant sites and controlling cell infusion to ensure efficient transplantation, for example, using the portal or tail vein for entopic transplantation and the mesentery and renal capsule for ectopic transplantation; monitoring replacement status via the expression of tissue or organ secreted proteins representative of stem/progenitor or immature cells transplant and cancer cell tracing, such as alpha fetoprotein or transferrin.
In summary, Alb-TRECK/SCID mice constitute a suitable model for human hepatic stem cell transplantation, as they provide a beneficial environment that allows for differentiation into mature hepatocytes. This novel model has tremendous applicative potential, not only for ex vivo expansion of human hepatocytes but also for drug screens and testing candidate therapeutic drugs for liver toxicity and metabolism. In addition, these results will aid in the advancement of future studies examining the clinical application of this technique to human hepatic cell transplantation.
The authors have nothing to disclose.
We wish to thank the Mammalian Genetics Project, Tokyo Metropolitan Institute of Medical Science, for providing the mice. We also thank S. Aoyama and Y. Adachi of the ADME (Absorption, Distribution, Metabolism, Excretion) & Toxicology Research Institute, Sekisui Medical Company Ltd., Japan, and K. Kozakai and Y. Yamada for assistance with LC-MS/MS analysis. This work was supported in part by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan to Y-W.Z. (18591421, 20591531, and 23591872); by the Jiangsu innovative and entrepreneurial project for the introduction of high-level talent and the Jiangsu science and technology planning project (BE2015669); and by grants to H.T. for Strategic Promotion of Innovative Research and Development (S-innovation, 62890004) from the Japan Science and Technology Agency.
Human albumin | Sigma | A6684 | Mouse | IgG2a |
Human CK19 | Dako | M088801 | Mouse | IgG1 |
Human nuclei | Millipore | MAB1281 | Mouse | IgG1 |
Human CK8/18 | Progen | GP11 | Guinea pig | Polyclonal |
CDCP1 | Biolegend | 324006 | Mouse | IgG2b |
CD90 | BD | 559869 | Mouse | IgG1 |
CD66 | BD | 551479 | Mouse | IgG2a |
GOT/AST-PIII | Fujifilm | 14A2X10004000009 | ||
DMEM/F-12 | Gibco | 11320-033 | ||
FBS | Biowest | S1520 | ||
0.05% Trypsin-EDTA | Gibco | 25300-054 | ||
Diphtheria Toxin | Sigma | D0564-1MG | ||
Human Albumin ELISA Kit | Bethyl Laboratories | E88-129 | ||
Syringe (1ml) | Terumo | SS-01T | ||
32G 1/2" needle | TSK | PRE-32013 | ||
O.C.T.Compound(118ml) | Sakura Finetek Japan | 4583 | ||
MoFlo high-speed cell sorter | Beckman Coulter | B25982 | ||
DRI-CHEM 7000 | Fujifilm | 14B2X10002000046 |