Therapeutic efficacy of human hepatocyte transplantation in a SCID/uPA mouse model with inducible liver disease

Abstract

Background: Severe Combined Immune Deficient (SCID)/Urokinase-type Plasminogen Activator (uPA) mice undergo liver failure and are useful hosts for the propagation of transplanted human hepatocytes (HH) which must compete with recipient-derived hepatocytes for replacement of the diseased liver parenchyma. While partial replacement by HH has proven useful for studies with Hepatitis C virus, complete replacement of SCID/uPA mouse liver by HH has never been achieved and limits the broader application of these mice for other areas of biomedical research. The herpes simplex virus type-1 thymidine kinase (HSVtk)/ganciclovir (GCV) system is a powerful tool for cell-specific ablation in transgenic animals. The aim of this study was to selectively eliminate murine-derived parenchymal liver cells from humanized SCID/uPA mouse liver in order to achieve mice with completely humanized liver parenchyma. Thus, we reproduced the HSVtk (vTK)/GCV system of hepatic failure in SCID/uPA mice.

Methodology/principal findings: In vitro experiments demonstrated efficient killing of vTK expressing hepatoma cells after GCV treatment. For in vivo experiments, expression of vTK was targeted to the livers of FVB/N and SCID/uPA mice. Hepatic sensitivity to GCV was first established in FVB/N mice since these mice do not undergo liver failure inherent to SCID/uPA mice. Hepatic vTK expression was found to be an integral component of GCV-induced pathologic and biochemical alterations and caused death due to liver dysfunction in vTK transgenic FVB/N and non-transplanted SCID/uPA mice. In SCID/uPA mice with humanized liver, vTK/GCV caused death despite extensive replacement of the mouse liver parenchyma with HH (ranging from 32-87%). Surprisingly, vTK/GCV-dependent apoptosis and mitochondrial aberrations were also localized to bystander vTK-negative HH.

Conclusions/significance: Extensive replacement of mouse liver parenchyma by HH does not provide a secure therapeutic advantage against vTK/GCV-induced cytotoxicity targeted to residual mouse hepatocytes. Functional support by engrafted HH may be secured by strategies aimed at limiting this bystander effect.

Figure 1. Histology of chimeric SCID/uPA liver at 12 weeks old.

Engrafted human hepatocytes (HH, brown nuclei) in paraffin embedded liver sections were distinguished from mouse hepatocytes (MH, blue nuclei) and other murine constituents (M, blue nuclei) by hybridization with fluoresceinated Alu probe (i, left). Serial sections were analyzed by PAS staining (right) without (ii) or with (iii) prior diastase treatment. Ceroid macrophages (iii, arrowheads). (Original magnification ×50).

Figure 2. Functional expression of vTK in Huh7 cells.

A. Huh7 cells were transfected with pCI-vTK and G418-resistant Huh cellular clones (cl) (1–6) were evaluated for vTK expression (∼43 KDa, arrowheads) by immunoblot analysis. Lmw, low molecular weight markers; 293TvTK, 293T cells transiently transfected with pCI-vTK. B. In vitro cytotoxicity of GCV in vTK expressing cells. HuhvTKcl1 and HuhvTKcl6 cells were incubated with various concentrations of GCV for 72 hours, followed by cell survival quantitation by MTT assay. Data represent the Mean +/− S.D from quadruplicate cell cultures for each GCV dose. Mock, Huh cells that underwent stable selection after transfection with empty pCIneo vector.

Figure 3. Tissue distribution of vTK in transgenic mice.

A. Expression of vTK protein (47 kDa) in various tissues from experimental FVB/N mice (listed in Table 1) were examined by immunoblot analysis (as shown for vTK+ mice 3.3, 3.21 and 3.11 and vTK− mice 3.18, 3.12, and 3.12, left panel). A smaller vTK immunoreactive protein (∼39 kDa) was identified in the testis of vTK+ mice. (right panel). B. Expression of vTK mRNA was evaluated in several tissues obtained from vTK+ mice (as shown for 3.3, left). Expression of delta-vTK mRNA was evaluated in the liver and testis of vTK+ (as shown fo 3.3, right) and vTK− (as shown fo 3.2, right) mice.

Figure 4. GCV induced histopathological changes.

A. vTK+FVB/N mice that did not receive GCV showed normal liver histology. B–C. vTK+FVB/N mice that received 50 mg/kg GCV showed increased cytoplasmic and nuclear enlargement with increased acidophilic bodies (B, arrowhead), apoptotic bodies (B, arrow), and areas of confluent necrosis (C, arrows) (A–C:Hematoxylin and eosin, original magnification ×200). D. TUNEL immunostaining showing increased apoptotic nuclei (arrows) in livers post GCV (Immunoperoxidase, original magnification ×200).

Figure 5. Impact of vTK/GCV on non transplanted SCID/uPA mouse liver.

Gross Appearance (upper panels) and H&E staining (lower panels) of livers from non-transplanted and age-matched vTK+SCID/uPA mice (3 months old) treated with 0, 25, 50 and 100 mg/kg GCV (i.p. every 48 h for 10 days). RN, regenerative nodules; PL, pale liver.

Figure 6. Impact of vTK/GCV on serum aminotranserases levels in chimeric SCID/uPA mice.

ALT (left) and AST (right) concentrations in blood sampled from experimental mice (Table 2) 1 h prior to (baseline, white bars) and at the end (endpoint, black bars) of the GCV treatment period (after 14 days of GCV administration with the exception of V241 whose endpoint sampling was after 10 days of GCV administration at which time V241 exhibited health crisis and had to be euthanized). Data represent Mean +/− S.E.M. P<.05 was considered to be significantly different endpoint levels relative to baseline (**).

Figure 7. Impact of vTK/GCV on hAAT production and human hepatocyte engraftement in chimeric SCID/uPA mice.

A. Baseline blood samples were obtained from experimental mice (Table 2) 1 h prior to the commencement of GCV dosing (t₀) for hAAT analysis. Subsequent samples were obtained from all chimeric mice for hAAT analysis once weekly for 2 weeks with the exception of v241 which had to be euthanized after 10 days of GCV dosing. B. A correlation plot of final hAAT concentration vs. replacement index (RI) was constructed for those experimental mice (Table 2) that were severely impacted by GCV (left) and those that remained healthy during the course of dosing period (right). RI was determined as the ratio of the area occupied by Alu-positive HH relative to the entire area examined in the in situ hybridization sections expressed as percent. Data represent the mean RI +/− SEM of at least 3 separate lobes).

Figure 8. Impact of vTK/GCV on the gross appearance of chimeric SCID/uPA liver.

Chimeric livers from vTK− SCID/uPA mice (Table 2) exhibited multiple small red-colored foci (D, E, and F for mice V245, V244 and V238, respectively) that were clearly evident in liver from chimeric vTK+ SCID/uPA mice that did not receive GCV (A for mouse V250), but were diminished in chimeric livers from vTK+ SCID/uPA mice that received 25 and 100 mg/kg GCV doses (B and C, respectively). Occasionally, chimeric livers from chimeric mice also exhibited large 3-dimensional regenerative nodules (RN).

Figure 9. Histopathological changes induced by vTK/GCV in chimeric SCID/uPA mice.

Engrafted human hepatocytes (HH, bordered by dashed line) in paraffin embedded liver sections from experimental mice (Table 2) were identified by hybridization with fluoresceinated Alu probe (top). Serial sections were analyzed by H& E staining (bottom). MH; mouse hepatoctyes. Original magnification ×100.

Figure 10. Induction of apoptosis by vTK/GCV in chimeric SCID/uPA mice.

Engrafted human hepatocytes (bordered by dashed line) in paraffin embedded liver sections from experimental mice (Table 2) were identified by hybridization with fluoresceinated Alu probe (left). Indirect TUNEL analysis of DNA fragmentation was performed on serial sections (right). Original magnification ×100.

Figure 11. Ultrastructural changes after GCV treatment.

A–B. Chimeric vTK+ SCID/uPA livers after GCV treatment (100 mg/kg) showing swollen electron lucent mitochondria (A) with aberrant cristae and matrices (B). C–D. vTK+ SCID/uPA livers with no GCV showing normal cellular (C) and mitochondrial (D) ultrastructure. E–F. vTK− SCID/uPA livers showing normal cellular (E) and mitochondrial (F) ultrastructure (Uranyl acetate-lead citrate, original magnification, A,C,E: ×1500; B,D,F: ×3600). These micrographs represent ultrastructural features that are present in 10 biopsies with a minimum of 5 random fields examined per mouse.