Y.-T. Zeng, Shanghai Institute of Medical Genetics, Shanghai Children's Hospital, Shanghai Jiao Tong University, Shanghai 200040, People's Republic of China Fax: +86 21 6247 5476 Tel: +86 21 6247 2308 E-mail: email@example.com
The herpes simplex virus thymidine kinase/ganciclovir (HSV-tk/GCV) system that selectively depletes cells expressing HSV-tk upon treatment with GCV has provided a valuable tool for developing a new animal model expressing the desired tissue damage. In this paper, an HSV-tk vector with an albumin promoter/enhancer was constructed. Based on the favourable killing effect on Hep-G2 cells by the recombinant construct, the HSV-tk transgenic mouse strains were developed. One strain of the TK transgenic mouse (TK5) was studied intensively. Integration of the target gene was confirmed primarily by PCR. Fluorescence in situ hybridization following G-banding analysis demonstrated that the insertion site was located at 2F1-G3. The hepatocyte-specific transcription and expression of HSV-tkwas verified by reverse transcription (RT)–PCR as well as by immunohistochemical staining. When two second-generation mice (TK5-F1 and TK5-F2) were injected with GCV, the pathogenic alterations in the liver were readily identified, including the appearance of vaculation in the hepatocytes with inflammatory infiltration in the liver, and diffuse proliferation of hepatocytes. In addition, the blood test demonstrates a significant increase of serum alanine aminotransferase, aspartate aminotransferase and total bilirubin. In conclusion, the transgenic mouse model with hepatocyte-specific expressed HSV-tk developed hepatitis with administration of GCV, had morphological and clinical chemical characteristics indicative of hepatocellular disease and should be useful for the the study of inducible liver-specific diseases.
The morbidity of severe liver disease is usually very high, seriously threatening the patient's health. Availability of animal models expressing related hepatic disorders should provide a means of studying the pathogenic mechanism of such diseases. Among these are the genetically engineered animal models. Unfortunately, currently available transgenic models are unsatisfactory for experimental use, because those transgenic mice expressing toxic protein often die too early due to the overexpression of toxic protein in such a vital organ. Others, for example the alb-uPA transgenic mouse and FAH– knockout mouse developed in the 1990s can be maintained only with constant medical treatment [1,2].
In 1989, Heyman et al.  developed a new transgenic mouse in which ablation of a specific cell type is TK-dependent. In such a transgenic mouse, the inserted herpes simplex virus thymidine kinase (HSV-tk) gene products can phosphorylate certain nucleoside analogues such as ganciclovir (GCV) that are not metabolized by conventional cellular enzymes. Phosphorylated nucleoside analogues such as GCV triphosphate are potent toxic metabolites for cells. Nevertheless, neither GCV nor the HSV-tk alone is harmful to cells. Hence, this conditional cell-depleting effect is achieved by expressing HSV-tk with a cell-specific promoter. It has been used for depletion of lymphoid cells, growth hormone-secreting cells, interleukin-2 and interleukin-4-expressing cells, dendritic cells or fibroblasts under the control of a cell-specific promoter [4–9]. Such a system is used in the transgenic rats of Kawasaki et al. , in which the rats develop experimental hepatitis on administration of GCV. The genome of the mouse is much better characterized than that of the rat and the cost of producing and maintaining transgenic mice is less than for rats. The high conservation and strong liver-specific regulatory machinery of the mouse serum albumin cluster makes it appropriate for use as a promoter for hepatic-specific expression [11,12].
In this study, HSV-tk transgenic mice were produced, in which the inserted gene is regulated by an albumin enhancer/promoter; liver injury is readily induced in this model. Among five founder transgenic mice generated, only one (TK5♀) transmitted the transgene to progeny through the germ line by mating with male –/– wild-type KM mice. Therefore, the F1 and F2 generations of TK5 were used for the inducible hepatic injury. In addition, the founder TK3♂ was also used for preliminary analysis of the relationship between expression level and histopathological changes.
Liver damage in transfected Hep-G2 cells after treatment with GCV
The pCMV-TK vectors were transfected into Hep-G2 cells, which were then induced with GCV. The transfected cells started to detach on third day after a single exposure to 40 µmol·L−1 GCV. Hep-G2 cells transfected with pLLTK started to detach on day 5, and maximal expression was achieved on day 7 after GCV treatment. Cell apoptosis was recognized in both of the two groups mentioned above; sick or damaged cells were seen to swell and burst. By contrast, the control cells transfected with vector pcDNA 3.1/zeo(+) grew and proliferated normally. The Hep-G2 transfected pLLTK showed completely different morphology as compared to the control cells (Fig. 1A). Such increased cell death could also be assessed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay. The survival rate of Hep-G2 cells was reduced significantly after transfection with HSV-tk (Fig. 1B).
Detection of inserted transgene and monitoring of the pedigree of mouse family TK5 by PCR
A total of 182 eggs were microinjected and subsequently reimplanted into eight pseudopregnant foster mothers, of which five became pregnant and gave birth to 36 mice. Among them, six mice showed the insertion of the HSV-tk gene as detected by PCR. The integration rate was 16.7% (6/36). One line of transgenic mice is female; the transgene is transmitted to the offspring at a rate of about 50% according to Mendel's laws (Fig. 2 A).
Chromosomal localization of transgene integration as demonstrated by fluorescence in situ hybridization (FISH)
More than 50 metaphases were analysed for each transgenic mouse. All of the metaphase cells showed one positive hybrid signal. According to the standard idiograms of mouse chromosomes, the integration site is located at 2F1-G3 in TK5-F1-455 (Fig. 2B). The integration site of TK5-F2-327 was similar to that of TK5-F1-455 (data not shown).
Reverse transcription (RT)–PCR of tissue-specific expression of HSV-tk in transgenic mice
RT-PCR showed that the 390 bp specific band of HSV-tk was detectable only in the transfected cells, liver and testicle. It was not detectable in the cells used as negative control, or in blood, kidney, pancreas, intestine, brain, skin or heart, even though an internal control band of 190 bp β-actin was present in all of the samples (Fig. 3).
HSV-tk protein expression in the liver of transgenic mice
Immunohistochemical staining was performed using a polyclonal rabbit-(anti-HSV-tk) Ig. The yellowish-brown staining sites were located mainly in the nucleus of hepatocytes integrated by the HSV-tk gene, and the HSV-tk-positive cells were distributed scattered or clustered in liver lobules, located mainly around the central vein, and occasionally in the periportal areas (Fig. 4A), while there was no staining in the liver of the wild-type mice (Fig. 4B). Simultaneously, the positive signal can be observed in both the nucleus and the cytoplasm after GCV treatment (Fig. 4C). The staining cells account for 20–30% of the total hepatocytes of TK5-F1-455 (Fig. 4A), whereas in mouse TK3, there were approximately 60–70% HSV-tk staining hepatocytes, with visible slight yellowish-brown signals in the focal necrosis, but several regenerative foci (regenerating parenchyma hepatocytes) displayed reduced or no staining (Fig. 4D). The percentage of HSV-tk-positive hepatocytes in the F2 mice (TK5-F2-327) of TK5 was similar to those of TK5-F1-455 (data not shown).
Hematoxylin and eosin staining for histological analysis
Microscopic analysis of the livers of GCV-treated HSV-tk mice (F1 and F2) showed that the diseased livers display a number of abnormalities, including the appearance of apoptosis bodies, hepatocyte vaculation, lymphocyte infiltration, hepatocyte megalocytosis, and diffused proliferation of hepatocytes (Fig. 5 A). In transgenic mouse TK3, mutifocal coagulation necrosis was evident in the liver (Fig. 5B). Histological analysis of the kidney showed no apparent abnormity in the GCV-treated transgenic mice and wild-type mice (data not shown).
Biochemical analysis of the blood
Twenty-one days after the injection of GCV, the values of alanine aminotransferase (ALT), aspartate aminotransferase (AST) and total bilirubin were significantly increased in the TK5-F1 transgenic mice (P < 0.05), whereas there were no significant increase in the wild-type control mice. The value of creatinine was not altered significantly in either group (Fig. 6). The changes of the four serum values in the F2 generation of the TK5 family and mouse TK3 showed similar results (data not shown).
We have generated a number of transgenic mice for liver damage, in which the HSV-tk gene was regulated by an albumin promoter/enhancer. The excellence of this mouse model is that liver damage and its extent in HSV-tk mice can be induced and controlled by GCV treatment. When the mice are injected with GCV, the pathologic changes and biochemical abnormalities, including vaculation of the hepatocytes, inflammatory infiltration, diffuse proliferation of hepatocytes as well as a significant increase of serum ALT, AST and total bilirubin, can be easily recognized. However, renal function is not affected by GCV treatment. This indicates that GCV at the dosage used in this study is associated with toxin-mediated hepatocyte damage in our HSV-tk mice.
We used FISH and RT-PCR to investigate transgene integration and HSV-tk expression in various tissues of the transgenic mice. FISH indicated that the HSV-tk gene was stably integrated in the genomes of the mouse family (TK5), and the expression of HSV-tk was readily detectd in the liver and the testis of TK5 family, but was not detectable in other tissues, such as blood, kidney, pancreas, intestines, brain, skin and heart. It indicated that the recombinant construct driven by the albumin/enhancer that we used in this study was highly tissue specific. Immunohistochemical analysis confirmed that the HSV-tk protein was expressed specifically in the liver. In TK3 mice, approximately 60–70% of the total hepatocytes showed HSV-tk expression, and 20–30% of the liver cells in the TK5 family gave positive results. The discrepancies of levels of HSV-tk expression between these two mouse strains may be due to differences in HSV-tk gene integration sites, that may be caused by random integration of the transgene.
Recent studies showed that HSV-tk converts the nontoxic prodrug GCV into GCV-triphosphate, which can cause chain termination and single-strand breakage upon incorporation into DNA. Although blocking of DNA synthesis of GCV is especially toxic for dividing cells, it can also cause damage of nondividing cells, such as hepatocytes, and liver toxicity of HSV-tk [13,14]. This provides the basis for selected hepatocyte killing using a hepatocyte-specific promoter in vivo. Hepatocyte replication was not a prerequisite for this effect, indicating that interference with DNA synthesis during S phase of the cell cycle is not the only mechanism of toxicity of phosphorylated GCV. Furthermore, although the exact mechanism by which suicide genes kill the HSV-tk-expressing cells is not yet clear, apoptosis has been considered to be a major contributor to GCV killing [15–18]. Song et al.  have shown that GCV induced HSV-tk expressing cells into apoptosis, thus inhibiting the growth of ovarian cancer cells. Shibata et al.  injected the HSV-tk vector into rats with bladder cancer and observed apoptosis of bladder cancer cells. Kawasaki et al.  created an AL-HSV-tk transgenic rat that expressed HSV-tk in hepatocytes, in which apoptosis was demonstrated after treatment with GCV. The administration of GCV elicited leukocyte infiltration and induced chronic hepatitis [21,22]. Although in the hepatitis model the precise role of Kupffer cells is unclear, it is possible that they are involved in inflammation . Activated Kupffer cells release cytokines and chemokines that activate and transport T cells [23–25]. It seems that hepatitis in the rat is primed by hepatocyte apoptosis . To examine the immunological mechanisms involved in cell killing using the HSV-tk/GCV system, HSV-tk-transduced human hepatocellular carcinoma (HCC) cells were implanted subcutaneously into immunocompetent syngeneic mice. After GCV treatment, marked infiltration by lymphocytes including CD4+ and CD8+ T cells, apoptosis of cells was induced, and significant reduction or even complete regression of tumours was achieved. Conversely, no significant inhibitory effects on tumour formation were observed in athymic nude mice. The results indicate that T cell-mediated immune responses may be a critical factor for achieving successful cell killing using the HSV-tk/GCV system . Administration of GCV to mice and rats injected with adenovirus encoding HSV-tk caused extensive signs of liver degradation with negligible survival rate . Microscopic analysis of the GCV-treated HSV-tk rat model of Kawasaki et al.  revealed moderate hepatocyte vacuolation and an increased number of inflammatory cells. In this study, we found that there was more severe focal necrosis of the liver tissue in TK3 than in TK5 mice. Moreover, the liver regenerating focus was more evident in TK3, in which clones of transgene expression-deficient cells were formed as detected by immunohistochemistry. The reasons for the different pathological changes between these two mouse strains are not clear. We suggest that these differences may be associated with the quantities of HSV-tk expressing cells. In addition, the patchy focal necrosis and regeneration in the TK3 liver could be explained by the possibility that this founder mouse may be a mosaic as approximately 5–10% of founder mice showed mosaicism of some sort (either multiple integration sites or patchy cellular distribution). Boucher et al.  compared the efficacy of the HSV-tk/GCV system in two human carcinoma cell lines after exposure to GCV and found that the killing effect depended on the concentration of the tk enzyme, the number of cells expressing HSV-tk, different cell types and the overall confluence of the HSV-tk expressing-cells. These results emphasise the importance of cell-specific metabolism in HSV-tk/GCV-mediated cytotoxicity. In conclusion, the killing of cells with HSV-tk/GCV is a complex interactive sequence of biochemical and cellular events involving incorporation and accumulation of the monophosphorylate derivative of GCV into DNA, disruption and inhibition of the cell cycle, gap junction metabolite transfer, and apoptosis. Thus, we conclude that the exact mechanisms may differ in: (a) different cell types; (b) different species; (c) the concentration of GCV used; (d) the quantity of cells expressing HSV-tk; and (e) the distribution of cells expressing HSV-tk.
It is of interest that male HSV-tk mice (including the founder TK3 and male offspring of TK5) generated in this study were not able to reproduce when mated with wild-type female mice. RT-PCR showed that HSV-tk was expressed both in the liver tissue and ectopically in the testis even though a heptocyte-specific promoter was used in generating HSV-tk transgenic mouse. The results suggest that male infertility may result from ectopic expression of HSV-tk in the testis. It has been reported that the natural HSV-tk gene contains a cryptic internal testis-specific promoter [28–31], and pathology has revealed that testicular development was immature and almost no sperm was produced in these mice (data not shown). When we treated female transgenic mice with GCV liver damage was induced showing that the liver damage in the HSV-tk transgenic mice may be independent of the HSV-tk expression in testis (data not shown).
In summary, transgenic mice specifically expressing HSV-tk in the liver were generated. When transgenic mice were treated with GCV, morphological, clinical, and biochemical characteristics indicative of hepatocellular disease developed. These HSV-tk mice could be an alternative model for the study of inducible liver-specific disease, and may be useful in the study of the pathogenesis of liver diseases and potential therapies.
Plasmid construction and generation of transgenic mice
Total DNA was extracted from KM mouse blood for PCR amplification of murine albumin promoter (−310 bp to +25 bp) and enhancer (−9192 bp to −11 250 bp). The primers for albumin promoter and enhancer are: pro1, 5′-CTTAGGTACCTCCATGCCAAGGCCCACA-3′; pro2, 5′-CTTGCTCACCATGGTGGCGACCGGTAGTGGGGTTGATAGGAAAGG-3′; en1, 5′-ACGAGTCTAGAGTGGAGCTTACTTCTTTGATTTGA-3′; en2, 5′-CCGCGTCGACGGAAAAGCGCCTCCCCTAC-3′; The 1800 bp of the HSV-tk coding sequence were also amplified by PCR from the pTK-neo plasmid and the consensus Kozak sequence GCCACC was introduced in front of the translation start codon ATG by the primers tk1, 5′-CGTATACCGGTGCCACCATGGCTTCGTACCCCGGC-3′ and tk2, 5′-CCGCGTCGACGGAAAAGCGCCTCCCCTAC-3′) . Recombinant plasmid pLLTK was obtained by inserting all three of the amplified fragments into the multiple cloning sites of pcDNA3.1(+)/zeo (Invitrogen, Carlsbad, CA, USA) by cohesive-blunt end ligation. Then pLLTK was digested with HindIII and the 4200 bp fragment of LLTK (Fig. 7) was obtained with the QIAquick gel extraction kit (Qiagen, Valencia, CA, USA). After purification with S & S Elutip minicolumns (Schleicher & Schuell, Keen, NH, USA), the DNA fragment was microinjected into the male pronuclei of the KM mouse fertilized eggs and transgenic mice were generated.
GCV-induced cytotoxicity in cultured transfected cells
Human hepatic cell line Hep-G2 and mouse breast epithelia cell line HC-11 cells were seeded in 24-well plates and grown in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal bovine serum and penicillium/streptomycin. The cells were then transfected with pLLTK by using LipofectamineTM 2000 (Invitrogen). In a parallel setting, control cells were transfected with the positive pCMV-TK and pcDNA3.1(+)/zeo. Hep-G2 cells were treated with 40 µmol·L−1 GCV (Roche, Indianapolis, IN, USA) 24 h after transfection. The morphology of the cells was examined by using an Olympus IX 70 inverted microscope (Hamburg, Germany) each day; cell survival rates were measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (Sigma Aldrich, St Louis, MO, USA) staining at 7 days later. The units of absorption were measured by an Elx800 plate reader (Bio-Tek Instruments, Winooski, VT, USA) at 570 nm. Differences in survival rates among different transfected Hep-G2 cells post-GCV treatment were analysed statistically using Student's t-test (SAS Software). A P-value < 0.05 was considered significant.
PCR analysis for the integration of the transgene
The transgene in the founder animals and their progeny was identified by PCR analysis of genomic DNA obtained from tail biopsies. PCR analysis was performed in 25 µL reaction mixtures. The primers (stk1, 5′-GTATACCGGTATGCCCACGCTACTGCGG-3′; SH552: 5′-GCACTCGAGACCCGTGCGTTTTATTCTGTCT-3′) for HSV-tk were designed to amplify a 390 bp region. Amplification was performed on a thermocycler for 30 cycles of: 1 min at 94 °C, 1 min at 59 °C and 30 s at 72 °C. PCR products were then separated electrophoretically on 2% agarose gel and visualized after ethidium bromide staining.
Chromosomal localization of transgene integration sites by using FISH following G-banding
Chromosome preparation was performed following the reported methods with modifications [33–35]. FISH was carried out according to the previous study with some modifications [36,37]. The DNA fragment LLTK was used as probe and labelled with the DIG-Nick translation mix (Roche) according to the manufacturer's protocol. Finally, slides were counter-stained with propidium iodide antifading solution (Sigma-Aldrich), and examined on a fluorescent microscope (Leica DM RXA2, Wetzlar, Germany). The nuclei were red and the hybridization signals were yellow-green. Previously photographed G-banded metaphases were relocated, and re-photographed. Chromosomal localization of the transgene integration site was determined by combining FISH hybridization signals and G-banding results on the same metaphases. The mice examined included TK5-F1-455♂, TK5-F2-325♀. The standard idiograms of mouse chromosomes were obtained from the web site: http://www.ncbi.nlm.nih.gov/genome/guide/mouse/.
RT-PCR of HSV-tk transcription
The 60-day-old transgenic mouse TK5-F1-455 was killed under pentobarbital anaesthesia (60 mg·kg−1) with all possible measures taken to ensure minimum pain and discomfort. Animal experiments were performed according to the National Institute of Health Guidelines for Care and Use of Laboratory Animals. The tissues of heart, liver, spleen, kidney, brain, intestine, pancreas, skin, testis and blood were powdered over an ice bath, total RNA was extracted using the Trizol reagent (Gibco/BRL). Primers stk1 and SH552 were used to amplify the 390-bp region HSV-tk. The 190 bp β-actin fragment amplified using primers MA2 (5′-CCACAGGCATTGTGATGGA-3′) and MA3 (5′-GCTGTGGTGGTGAAGCTGTA-3′) was used as an internal control.
Immunohistochemical analysis of HSV-tk expression
Immunohistochemical staining was performed to detect HSV-tk expression in hepatocytes. Tissues fixed in 4% (v/v) phosphate-buffered formalin were embedded in paraffin and 5 µm-thick sections were stained. Briefly, paraffin-embedded tissue sections were dewaxed, rehydrated, and permeated before blotting, the slides were then incubated with the rabbit polyclonal anti-(HSV-tk) Ig diluted 1 : 250 in NaCl/Tris for 1 h at 37 °C in a humidified chamber. The slides were then washed three times and incubated with biotinylated goat anti-rabbit immunoglobulins (DAKO) at a dilution of 1 : 300 in NaCl/Tris for 1 h at room temperature with protection against light. After washing three times with NaCl/Tris, peroxidase-conjugated streptavidin (DAKO) diluted 1 : 300 was added for 1 h at room temperature and washed three times with NaCl/Tris. Finally, the signal was visualized by incubating the slides with 3–3′-diaminobenzidine (DAKO). The examined mice included TK5-F1-455, TK5-F2-325 and TK3.
Induction of liver damage
For the present study, mice were housed individually at 22 °C using a 12 h light/12 h dark photoperiod. Three transgenic mice of TK5-F1 (including three ♂), TK5-F2 (including two ♂ and one ♀) and TK3 aged 8–14 weeks, received tail vein injections of sodium GCV (10 mg·kg−1) at 48 h intervals through a 29-gauge needle on 10 occasions, meanwhile three nontransgenic mice served as the control. GCV was dissolved in NaCl/Pi and filter-sterilized before administration.
Twenty-one days after GCV treatment the mice were killed under pentobarbital anaesthesia. The tissues were fixed with 4% (v/v) phosphate-buffered formalin and paraffin-embedded sections were stained using hematoxylin and eosin as described previously [38,39].
Biochemical analysis of blood
The physiological function of the liver was examined by determining biochemical serum values. Before the experiment, physiological function of the liver was determined to be normal by examination of serum ALT, AST, total bilirubin, albumin, globulin creatinine, total protein, lactate dehydrogenase and others. Peripheral blood was extracted from the tail vein once a week for 3 weeks for analysis of ALT, AST, total bilirubin and creatinine. Student's t-test (SAS software) was performed to compare the differences in serum values in transgenic mice and controls after GCV administration. P < 0.05 was considered significant.
We thank Drs Yi-Ping Hu (Department of Cell Biology, Second Military Medical University of PLA, Shanghai, China), Willams Summers (Yale University, USA), Zheng-Hong Yuan (Key Laboratory of Medical Molecular Virology, Fudan University, Shanghai, China) and Hynes (Friedrich Miescher Institute of Switzerland, Basel, Switzerland) for pTK-neo plasmid, polyclonal rabbit anti-(HSV-tk), Hep-G2 and HC-11 cell lines, respectively. We would like to thank Drs Zhao-Rui Ren, Jing-Zhi Zhang (Institute of Medical Genetics, School of Medicine, Shanghai Jiao Tong University, China) and Dr Xiao-Kang Li (the National Institute for Child and Development, Japan) for their very helpful discussion in this work. We also thank Yi-Wen Zhu, Wen-Ying Huang, Xiu-Li Gong (Institute of Medical Genetics, School of Medicine, Shanghai Jiao Tong University, China) for their performing the FISH and microinjections. The study was supported by: the Chinese National Program for High Technology Research and Development (No.2002AA216091).