Research On The Relationship Between Viral Infection And Induction Of Biliary Atresia In Animal Models

Backgrounds and Purpose:Biliary atresia (BA) is a progressive, inflammatory cholangiopathy of infancy that leads to obstruction of extra-and intrahepatic bile ducts with eventual biliary cirrhosis. Despite its importance in child health, the cause of biliary atresia remains unclear. Evidence accumulated from both human and animal studies indicated that one possible pathological mechanism is the host immune response triggered by perinatal viral infection. Among the suspected viral agents, cytomegalovirus (CMV) and rotavirus drew more attentions. High incidence of cytomegalovirus CMV infection was found in patients with BA, but no solid evidence supported that the infection was the causing factor of the biliary system obstruction. So one purpose of our study is to set up a perinatal CMV infection induced hepatobiliary injury model, analyze the pathological change in hepatobiliary system and systemic inflammatory or immune response in the host.The study would provide important information on understanding the role of CMV in the inducing of the bile injury or even obstruction. The rhesus rotavirus (RRV)-induced murine model shares many similarities to the disease process found in children affected with BA and were used as a vital tool in analyzing various aspect of the infectious process in biliary atresia. As rotavirus families are refractory to the reverse genetic technique, the studies aimed to understand the molecular mechanism that characterizes the rotavirus life cycle and the pathogenesis of viral infection was hampered. Take advantage of a well-known property of rotavirus that they undergo genetic reassortment after mixed infection in cell culture or in vivo, we aim to create virus with single gene segment reassortment. Administrate these reassortants to neonatal mice, characterize their disease phenotype, we will define the key gene segment which determine the viral virulent to biliary tract. We also aim to define the role of other gene segments in this model. Cell binding assay and infectivity analysis were also employed to verify the mechanism by which the gene segments or its protein products worked. The studies will contribute to the understanding the initial stage that virus attact target cell, shadow light on a potential mechanism by which biliary atresia occurs. Material and methods:A, Creation of perinatal CMV infection induced hepatobiliary injury model and injury evaluation:Guinea pig cytomegalovirus (gpCMV) were maintained in fetal guinea pig pulmonary fibroblast cells.The experiment has 3 groups:Prenatal group:Female guinea pigs at the 40th to 43th pregnant day (3rd trimester) were introperitoneally injected with virus supernates at a dose of 1×109TCID50 per dam. Two other subsets of dams severed as saline or blank control. Alive pups were sacrificed at different time point after birth. Neonatal group:A subset of healthy pups was introperitoneally injected with virus supernates within 24 hours of birth at a dose of 1×108 TCID50 per pup, Pups were sacrificed at different time point after inoculation. Infantile Group:Another subset of healthy pups was inoculated with a same dose of virus supernates at day 10 after birth. Pups were sacrificed at different time point after inoculation. Samples of the liver and proximal extrohepatic bile duct, blood, bile were collected. Weight gain, clinical sign of hepatobiliary injury and survival were recorded. Hepatic biochemical analysis was performed. A portion of the liver and extrohepatic bile duct were preserved in formalin. Serial sections underwent hematoxylin and eosin (HE) staining or in situ apoptosis analysis or immunohistochemical analysis for the presence of CD8+T cells. The expression of hepatic growth Factor (HGF), interleukin-6 (IL-6) in bile was examined with enzyme linked immunosorbent assay (ELISA). Other portion of liver and extrohepatic bile duct were iced and underwent in situ hybridization for the distribution of viral mRNA. B. Determine the specific RRV gene domain responsible for the injury bile duct injury in murine. The Virulent strain RRV was matched with avirulent strain EDIM (in vivo) or TCUH (in vitro). A subset of reassortants with double or triple gene segment reassortment was adapted to back cross with parental strain or cross with other progeny clones. MA 104 cells was infected with virus supernants generated from matching, the cell plates were overlayed with 0.2% agarose in EBSS, the plaque purified virus were freeze-thawed once and genotyped by polyacrylamide gel electrophoresis. Newborn balb/c mice were inoculated with reassortants, monitored for 21 days. A subset of pups was sacrificed at the 7th day after injection for histological analysis. The reassortant's binding ability was analyzed: immortalized cholangiocytes were inoculated with reassortant clones at 4℃for 1 hour; the amount of attached virus was expressed as a percentage of the total amount of virus used to inoculate the cells. In the infectivity essay, immortalized cholangiocytes were infected with reassortant at MOI of 1.the viral titer was measured after 48 hours. The extrohepatic bile ducts collected from mice at 7 days post inoculation were homogenized, tiered on the MA 104 cells.Results:A.In guinea pigs, perinatal inoculation of gpCMV induced signs of hepatobiliary system injury.1. Birth weight of pups in prenatal group was lower than saline or blank controls (P=0.0029, P=0027, respectively). These pups developed growth retardation versus that of saline and blank controls. Pups in neonatal and infantile group showed no sign of growth retardation.2. TB, DB, ALT, AST levels in Pups of prenatal group were higher at day 1 or day 10 after birth as compared to controls (P<0.05), while they were not significantly different between these two time points (P>0.05). Those parameters at day 20 decreased as compared to those at day 1 or day 10. A few pups in prenatal showed sign of jaundice within 10 days after birth, which were associated with increased TB and DB level. In neonatal group, a temporary increased AST level was observed at day 10-post injection.3. Histological analysis of the livers revealed a notable infiltration of mononuclear cells in 63.6%(14/22) of pups in prenatal group. Inflammatory cells were mainly found in portal tract area. There was interlobular portal tract damage and fibrosis, bile ductular proliferation. Hepatocytes balloon-like degeneration, necrosis and ductular cholestasis were often found. Micro-abscesses were seen (2/14). Pathological changes were also found in pups of neonatal group (3/8, but not as severe as those in prenatal group. No notable pathologic change was found in infantile group.4. Apoptosis signals were mainly seen in hepatocytes and mesenchymal cells in portal tract, coincident with the distribution of inflammatory cells, no definite signal was observed in vascular or ductular epithelial cells.5. Viral distribution:gpCMV mRNA signal were found in endothelial and epithelial cells and portal tract stoma.No signal was found on hepatocytes. Positive rate of gpCMV was 17.27%(4/22) in pups of prenatal group, significantly lower than those of neonatal or infantile groups (55%,50%respectively).6. HGF level increased in bile from pups of neonatal group and infantile group (P=0.0292, P=0.0461). No IL-6 was found in any sample with this method.7. CD8+T cells were often found in samples of prenatal group (8/14) and neonatal group (3/3), distributed mainly in liver parenchyma and portal area. B. In the murine biliary atresia model,1. Reassortant creation and pathogenecity assessment:21 reassortants were generated in the matching of RRV with EDIM, among of them only 4 were clones with single gene reassortment (4/170). Only one of them was on EDIM background. The reassortant D6/2(EDIM background strain with the 4th gene segment derived from RRV) caused signs of biliary injury that was similar with that of RRV. In the crossing of RRV with TUCH (back crossing or progeny crossing included), all the 22 single-segment reassortants were generated and selected to inoculate newborn pups. They were named as RTn (RRV background clone with the nth gene segment from TUCH) TRn (TUCH background clone with the nth gene segment from RRV).2.Clone RT4 did not induce disease, on the contrary, clone TR4 carried on signs of biliary obstruction in 94.12% of pups and produced a mortality rate of 88.24%, which is of no difference with that of pups injected with RRV (P=1.000＞0.05). An intermediate phenotype was seen in RT3 (RRV background strain with the 3rd gene segment deriving from TUCH) infected mice, although 58.33 percent of pups developed sign of biliary obstruction, the rate was significantly lower than that of mice injected with RRV (P=0.001＜0.05), so was the mortality rate (20.83%, P=0.000<0.05). Pups injected with other RRV background reassortants developed manifestation of biliary obstruction, which were similar with that in RRV induced model. Reassortant clone TR7as well as TR1 and TR2, did not elicit any sign of hepatobiliary injury nor did they cause mortality. Most of other TUCH background reassortants brought on temporary sign of hepatobiliary injury, and most of the pups survived.3. The histological appearances of the portal area as well as the extrohepatic bile ducts from different strain infected mice were consistent with the mice's symptoms.4. Viral binding ability to cholangiocyte was significantly changed by manipulating the 4th gene segment. The clone RT4's binding ratio (3.97%±0.93%) was much lower than that of the parental strain RRV (16.03%±1.31%) (P=0.003<0.05) and that of other RRV background strains (P＜0.05). Other RRV background reassortants'binding ratios were not significantly different from that of parental strains (P＜0.05). As we prospected, replacement of the 4th gene segment of non-pathogenic TUCH strain with corresponding gene from RRV enhanced the viral attachment, clone TR4 exhibited a significantly higher binding percentage (12.43%±2.25%) as compared with that of parental strain TUCH (5.33%±1.27%, p=0.001). Besides TR8 and TR2 other TUCH background clones showed a similar binding ratio as compared with that of TUCH (P>0.05)5.Consistent with the finding in binding assay, clone RT's titer in cholangiocytes as well as in bile duct at day 7 post inoculation was significantly lower than that of RRV(P= 0.013,0.000 accordingly).On the other hand, clone TR4's titer were definitely higher than that of TUCH both in vitro and in vivo (P=0.046,0.004 accordingly). On the cholangiocytes, all the other RRV background clones'titer were similar with that of parental strain RRV (P>0.05), but most of those (except for RT8, RT9) were lower than that of RRV in vivo (P<0.05). Consistent with the observation that mice injected with R1,R2 and R7 did not show any symptoms of the disease, the viral titer both in chlangiocytes and in bile duct were even lower than TUCH (P＜0.05). Alternatively, most of the other symptoms inducing strains titer were higher than that of parental strain TUCH (P<0.05) in vitro. Spearman correlation analysis indicated the mean titers were related to the morbidity of the disease among TUCH background reassortants.Conclusion:A.In the guinea pigs model,1 Introperitoneally inoculating the third trimester pregnant dams or neonatal pups with gpCMV could induce inflammation and injury in hepatobiliary system.2 Intra-hepatic infiltrated inflammatory cells were mainly mononuclear cells, parts of which were CD8+T cells. Those indicated Thl immune response played an important role in the phenotypic progression.3 Viral mRNA signals were mainly found in portal tract and endothelial system, the tropism of gpCMV to bile duct epithelial cells and the corresponding damage to bile duct system were proved.4 All injury phenotypes were from pups of prenatal and neonatal groups, closely related to the inoculation time, which indicated immature host immune system might play a key role in the process.B. In the murine biliary atresia model:1 Rhesus rotavirus gene segment 4 was the major determinants of pathogenecity. It also determined the RRV's specific tropism to cholangiocytes. Manipulation of gene segment 4 greatly changed the viral binding ability and infectivity, may in turn, and determined the viral phenotype. Expression of related cell surface receptor with binding site for VP4 (product of gene segment 4) might determine the susceptibility of changiocyte to viral insults.2 Gene segment 7 did not contribute to the viral pathogenecity. Replacement of the 7th segment did not affect the viral binding; either did it affect its replication on cholangiocytes. Indicated the gene (encoding VP7) did not contribute to the virulence, the cell surface binding receptor related to VP7 might not play an important role in cholangiocyte susceptibility.3 Gene segment 3 (encoding VP3) played an important role in RRV's pathogenecity, the intermediated virulent phenotype induced by reassortant TR3 probably was the result of interaction of multiple-factors, among them, host immune response may take an important role.4 Gene segment 1,2 did not contribute to the viral pathogenecity, but they might coordinate in the process of viral replication.