JOURNAL OF VIROLOGY Apr 2003 p 4972 4978 Vol 77 No 8 0022 538X 03 08 00H110010 DOI 10 1128 JVI 77 8 4972 4978 2003 Copyright 2003 American Society for Microbiology All Rights Reserved Murine Coronavirus Induced Hepatitis JHM Genetic Background Eliminates A59 Spike Determined Hepatotropism Sonia Navas and Susan R Weiss Department of Microbiology University of Pennsylvania School of Medicine Philadelphia Pennsylvania 19104 6076 Received 10 December 2002 Accepted 28 January 2003 Recombinant murine coronaviruses differing only in the spike gene and containing the strain A59 mod erately hepatotropic and JHM neurotropic spike genes in the background of the JHM genome were compared for the ability to replicate in the liver and induce hepatitis in weanling C57BL 6 mice Interestingly expression of the A59 spike glycoprotein within the background of the neurotropic JHM strain does not reproduce the A59 hepatotropic phenotype Thus the JHM genetic background plays a dominant role over the spike in the determination of hepatotropism The murine coronavirus mouse hepatitis virus MHV has a single stranded positive sense RNA genome of approximately 31 kb 35 MHV infection of the central nervous system CNS has been used as a model for the study of chronic demyelinating diseases 3 6 9 12 19 23 26 32 Some MHV strains cause a wide range of liver injuries that range from minimal changes to fulminant hepatitis 5 11 13 20 The development of targeted RNA recombination 14 24 25 38 has allowed the generation of chimeric recombinant viruses differing only in the spike gene 22 Targeted RNA recombi nation was developed by using the MHV A59 strain 24 25 and thus until recently only the A59 genetic background has been used for generation of recombinant viruses Using chi meric A59 recombinant viruses we have previously demon strated that the spike protein of MHV is a major determinant of neurovirulence 33 34 demyelination 4 and hepatotro pism 27 We further demonstrated that the ability to induce hepatitis is largely determined by the spike gene when the rest of the viral genes are derived from A59 27 We compared three isogenic recombinant viruses differing only in the spike gene and expressing the spike protein of A59 RA59 formerly named S A59 R13 MHV 2 Penn 98 1 or MHV JHM SJHM RA59 formerly named S 4 R21 all in the background of the A59 genome We found that after intrahepatic inoculation with 500 PFU per mouse the spike gene determined the viral load in the liver and that the amounts of antigen staining and ne crosis in the liver correlated with the viral load 27 Our results were not surprising since the spike glycoprotein is re sponsible for viral receptor attachment entry and cell to cell fusion 7 8 39 Thus the spike would be expected to play a crucial role in initiation of infection as well as in virus spread ing 8 Our goal in this study was to explore the role of the viral genetic background in the development of murine coro navirus induced hepatitis Our data demonstrate that the pres ence of background genes from the neurotropic JHM strain of MHV eliminates A59 spike hepatotropism even after inocula tion with 10 6 PFU directly into the liver We have previously described the targeted recombination technology used to select recombinant viruses in the A59 background differing only in the spike gene including the wild type A59 recombinant RA59 and the A59 recombinant virus expressing the JHM spike SJHM RA59 27 33 In this study we have constructed two new recombinant viruses Briefly recombinant viruses were selected by using fMHV JHM clone B3b 29 an interspecies chimeric JHM recom binant virus encoding the ectodomain of the spike gene of feline infectious peritonitis virus 15 and synthetic capped RNA transcribed from pJHM a plasmid containing the HE gene through the 3H11032 end of the JHM genome in which spike genes can be switched Here we have changed our previous virus terminology in order to make it easier to identify the source of the spike and the source of the other viral genes Thus recombinant parental strains are named RA59 and RJHM and chimeric recombinant viruses are named after the source of the spike followed by the type of viral genetic back ground Thus RJHM is a wild type recombinant virus express ing the JHM spike protein whereas SA59 RJHM expresses the A59 spike in the JHM background For each genotype at least two independent recombinant viruses with the same spike gene sequence were evaluated in order to minimize the possible interference of spurious mutations outside of the spike gene Note that independent recombinant viruses of the same geno type are not distinguished by name We initially compared the in vitro replication characteristics of the recombinant viruses We and others have previously reported that MHV A59 replicates to a high titer in a number of cell lines in vitro whereas MHV JHM replicates to a lower titer and displays higher levels of fusion and cytotoxicity 7 8 Here to determine whether non spike genes affect replication properties we analyzed the kinetics of both released and cell associated virus production for the recombinant viruses SJHM RA59 and SA59 RJHM compared to those of RA59 and RJHM in L2 cells Fig 1 In quantifying released virus we found that recombinant viruses containing the JHM spike gene SJHM RA59 RJHM and parental JHM data not shown all replicated with slower kinetics and to a lower final titer than RA59 confirming our previous results that in cell culture the spike gene of JHM conferred an alteration in in vitro replication 33 Interestingly JHM recombinants ex Corresponding author Mailing address Department of Microbi ology University of Pennsylvania School of Medicine 36th St and Hamilton Walk Philadelphia PA 19104 6076 Phone 215 898 8013 Fax 215 573 4858 E mail weisssr mail med upenn edu 4972 on April 11 2015 by guest http jvi asm org Downloaded from pressing the A59 spike gene SA59 RJHM displayed the same pattern of low virus production as RJHM and SJHM RA59 Thus introduction of the A59 spike gene into the JHM back ground did not alter the slower kinetics or the final extent of replication of JHM Comparison of recombinant viruses with the same spike gene and differing only in non spike genes SJHM RA59 versus RJHM and SA59 RJHM versus RA59 confirms that JHM genes other than the spike gene play a role in determining the extent of in vitro virus replication When cell associated virus production was quantified however the difference between RA59 and the other strains was much less dramatic This suggests that the presence of high titers of RA59 in the supernatant may be due to factors such as virus stability rather than replication itself RA59 is in fact more stable than SJHM RA59 37 Strikingly SA59 RJHM exhib ited a small plaque phenotype H110211 mm like JHM data not shown demonstrating that the A59 spike gene in the JHM background does not confer the large plaque phenotype H110222 mm characteristic of A59 Thus in cell culture JHM back ground genes contribute to plaque morphology as well as replication kinetics In order to determine whether replacement of the JHM spike gene with the A59 spike gene alters the in vivo JHM phenotype SA59 RJHM we evaluated virulence by both in tracranial and intrahepatic inoculations of 4 week old male C57BL 6 mice NCI We also evaluated the virulence of RA59 RJHM and SJHM RA59 Table 1 Infected mice were observed for mortality and the 50 lethal dose LD 50 was calculated as previously described 27 We confirmed our pre vious results demonstrating that the spike gene of the highly neurovirulent JHM strain is sufficient to confer high neuro virulence in the A59 background In contrast substitution of the spike gene of neuroattenuated A59 for the JHM spike gene SA59 RJHM had lesser effects on JHM neurovirulence sug gesting that genes other than the spike gene contribute signif icantly to JHM neurovirulence When virulence was assessed by direct inoculation into the livers of mice which eliminates the CNS disease 10 the LD 50 s of the SJHM RA59 RJHM and SA59 RJHM viruses were dramatically higher while the virulence of RA59 was the same as that obtained by intracra nial inoculation Table 1 This attenuation following intrahe patic inoculation suggests that RJHM SJHM RA59 and SA59 RJHM are very poorly hepatotropic These data confirm that the JHM spike gene does confer a dramatic increase in neurovirulence within the A59 background 33 and further demonstrate that genes other than the spike gene may contrib ute to JHM virulence Hepatotropism and pathogenesis in the liver of recombinant viruses RA59 RJHM SJHM RA59 and SA59 RJHM were evaluated by direct inoculation of the livers of 4 week old male C57BL 6 mice that were sacrificed at day 5 postinfection p i the peak of replication in the liver 27 After perfusion with phosphate buffered saline livers were homogenized and the virus titer was determined on L2 cells as previously described 27 We first inoculated mice with 500 PFU the same dose used in our previous study 27 RA59 replicated efficiently whereas the A59 recombinant virus expressing the spike gene of JHM SJHM RA59 replicated to a minimal extent above the level of detection RA59 versus SJHM RA59 P H11021 0 01 Wilcoxon rank sum test confirming our previous data 27 Fig 2A Remarkably after inoculation with 500 PFU no infectious virus was recovered from livers infected with the FIG 1 Time course of released A and cell associated B virus production in L2 cell cultures L2 cells were infected in duplicate with RA59 F SJHM RA59 E RJHM H12135 or SA59 RJHM ata multiplicity of infection of 1 PFU cell The data shown represent the mean titers of duplicate samples In the case of SA59 RJHM two independent recombinant viruses were analyzed At the indicated times virus titers were determined in cells and culture supernatants by plaque assay in L2 cells TABLE 1 Virulence of recombinant viruses after intracranial and intrahepatic inoculations Virus Log 10 LD 50 Intracranial Intrahepatic RA59 3 8 3 8 RJHM 0 7 H110226 0 SJHM RA59 1 0 H110225 0 SA59 RJHM 2 0 H110226 0 VOL 77 2003 NOTES 4973 on April 11 2015 by guest http jvi asm org Downloaded from JHM recombinant virus expressing the A59 spike gene SA59 RJHM or RJHM Fig 2A While a negative or very low burden of RJHM virus can be expected the absence of SA59 RJHM virus production in the liver was surprising with respect to our previous report in which we demonstrated that the spike gene is a major determinant of hepatotropism 27 This finding prompted us to further analyze liver tropism following inoculation with larger viral doses 10 5 and 10 6 PFU Fig 2B Strikingly SA59 RJHM like RJHM itself was still unable to induce a productive infection in the liver only a few mice showed minimal replication just above the limit of detection even after inoculation with 10 6 PFU Fig 2B These data demonstrate that substitution of the A59 spike gene for the JHM spike gene does not alter the low hepatotropism pheno type of JHM suggesting that background genes do play a significant role in liver tropism Supporting these data we also found that after inoculation with 10 5 PFU the A59 recombi nant virus expressing the JHM spike gene SJHM RA59 was able to replicate to an extent similar to that of RA59 Fig 2B This fact further supports a contribution of background genes to hepatotropism These data may seem to contradict our pre vious report demonstrating that the spike is a major determi nant of hepatotropism 27 However in the present study we confirm our previous finding that at a low virus dose 500 PFU the extent of hepatitis is determined by the spike gene within the A59 background at this dose the A59 recombinant virus expressing the JHM spike gene SJHM RA59 replicates to a minimal extent while RA59 replicates efficiently To further evaluate hepatitis induced by these viruses his tological assessment of liver sections was performed Formalin fixed paraffin embedded liver tissue was sectioned and stained with hematoxylin and eosin for histopathological diagnosis Hepatitis was graded blindly and scored on a scale of 0 to 4 as previously described 1 Immunohistochemical analysis was performed on replicate sections as previously reported 27 by using a monoclonal antibody MAb against the nucleocapsid protein of MHV JHM MAb clone 1 16 1 kindly provided by J L Leibowitz TexasA Vector was performed on all sections in order to enhance virus localization Appropriate controls were per formed 27 and slides were read in a blinded manner The resulting data are shown in Fig 3 4 and 5 After inoculation with 500 PFU hepatitis induced by RA59 was mainly characterized by moderate hepatocellular damage although some mice developed mild or even severe hepatitis This range of liver lesions has been previously observed for RA59 21 27 The A59 recombinant virus expressing the JHM spike gene SJHM RA59 induced min imal changes to mild hepatitis RA59 versus SJHM RA59 P H11021 0 01 Wilcoxon rank sum test confirming our previous report 27 Interestingly the JHM recombinant virus ex pressing the A59 spike gene SA59 RJHM like RJHM itself induced minimal changes in the liver RJHM or SA59RJHM versus A59 P H11021 0 001 Wilcoxon rank sum test After inoculation with a larger viral dose 10 5 PFU RA59 induced the same extent of moderate hepatitis as did inoculation with 500 PFU and remarkably the A59 recom binant virus expressing the JHM spike gene SJHM RA59 caused an exacerbation in the degree of hepatitis inducing liver lesions to the same extent as RA59 does Fig 3B and 4C Strikingly after inoculation of 10 6 PFU both the RJHM and JHM recombinant virus expressing the A59 spike gene SA59 RJHM caused mild hepatitis Fig 3B Thus JHM recombinant viruses irrespective of the spike gene caused very minimal after 500 PFU to mild after 10 6 PFU hepatitis whereas the A59 recombinant virus express ing the JHM spike gene like RA59 itself induced moderate hepatitis but only after inoculation with a large viral dose 10 5 PFU JHM background versus A59 background P H11021 0 01 Wilcoxon rank sum test We found a correlation between the viral replication titers and the degree of hepa titis induced by each virus Fig 3C The extent of hepato cellular injury was also monitored by determining the num ber of nonconfluent necrotic foci per random field three liver sections per sample eight random fields per section FIG 2 Viral loads in livers of C57BL 6 mice after intrahepatic inoculation with 500 A or 10 5 to 10 6 B PFU of recombinant virus RA59 SJHM RA59 RJHM or SA59 RJHM at 5 day p i Viral titers were determined by plaque assay and are presented as the log 10 PFU per gram of liver Each point represents viral titer of individual mice and bars represent the log 10 of the mean in each group The limit of detection was 200 PFU g of liver The numbers of mice examined per viral dose were as follows RA59 SJHM RA59 and RJHM 5 SA59 RJHM 10 RJHM and SA59 RJHM did not induce a productive infection in the liver irrespective of the viral dose P H11021 0 01 Wilcoxon rank sum test RA59 replicated efficiently irrespective of the viral dose whereas SJHM RA59 replicated to a minimal extent after inoc ulation with 500 PFU RA59 versus SJHM RA59 P H11021 0 01 Wilcoxon rank sum test SJHM RA59 was able to replicate to an extent similar to that of RA59 only after inoculation of large viral doses 4974 NOTES J VIROL on April 11 2015 by guest http jvi asm org Downloaded from Fig 3D and this analysis showed that recombinant viruses with the JHM background induced one to four necrotic or inflammatory foci per field whereas recombinant viruses with the A59 background showed more than eight foci per field or even confluent necrosis Figure 4 shows liver histopathology at day 5 after intra hepatic inoculation with 10 5 RA59 and SJHM RA59 10 6 RJHM and SA59 RJHM or 500 SJHM RA59 PFU The MHV A59 and MHV JHM strains are very different in liver histopathology We found that our recombinant wild type viruses RA59 and RJHM exhibited the same phenotype in the liver as the corresponding parental viruses data not shown Thus RA59 caused moderate necrosis with inflam mation located in the portal areas as well as in the lobular parenchyma Fig 4A whereas RJHM induced scattered inflammatory foci with occasional spotty necrosis even after inoculation with 10 6 PFU Fig 4B The A59 recombinant virus expressing the JHM spike gene SJHM RA59 caused noticeable hepatocellular damage with increased inflamma tion after inoculation with 10 5 PFU Fig 4C this is in contrast to the occasional necrosis found after inoculation of 500 PFU Fig 4E Strikingly the JHM recombinant expressing the A59 spike gene SA59 RJHM like RJHM itself caused minimal changes to mild hepatitis even after 10 6 PFU with histopathological changes characterized by small scattered inflammatory foci with some focal necrosis Fig 4B and D Figure 5 shows immunohistochemical staining of liver sec tions from mice inoculated directly in the liver with 10 5 RA59 and SJHM RA59 10 6 RJHM and SA59 RJHM or 500 SJHM RA59 PFU by day 5 p i In general viral antigen immunolabeling always colocalized with areas of hepatocellu lar degeneration and necrosis as well as with individual or small clusters of hepatocytes It is noticeable that the RA59 and SJHM RA59 viruses exhibited differences in viral spread ing in the liver thus SJHM RA59 is not able to spread as efficiently as RA59 does Fig 5A versus C Viral spreading was very limited for both RJHM and SA59 RJHM showing isolated viral staining Fig 5B versus D Overall our data suggest that although the spike gene is critical in the development of MHV liver pathogenesis other genes do contribute to hepatitis development These findings have led us to speculate that genes other the spike gene greatly influence postentry events that determine the outcome of in fection that is the ability of the virus to replicate and spread within the liver Thus the JHM spike gene is likely inefficient at mediating entry into one or more liver cell types but once entry is achieved the A59 background genes allow efficient replication Thus at large virus doses enough virus can enter the cells to allow efficient replication in the context of the A59 background genes Conversely A59 spike gene mediated entry into liver cells may be efficient but the JHM background genes eliminate replication and hepatitis development There are many potential mechanisms involving both nonstructural and structural proteins by which genes other than the spike gene may influence pathogenesis Thus less efficient JHM replication in one or more cell types in the liver may account for the phenotype of SA59 RJHM It has been previously proposed that the hepatotropism of a given MHV strain may be determined by its ability to replicate in nonparenchymal cells in the liver 30 31 36 Differences in replication would be due to the ability of the replicase or other nonstructural proteins to interact with cell type spe cific factors needed for replication Structural genes other than the spike gene may also play a role in the development FIG 3 Extent of hepatitis scored as minimal 1 mild 2 mod erate 3 or severe 4 as described in the text and titers log 10 PFU per gram of recombinant viruses in livers of C57BL 6 mice after intrahepatic inoculation with 500 A or 10 5 to 10 6 B C and D PFU of RA59 SJHM RA59 SA59 RJHM or RJHM virus The results are shown as percentages of mice with minimal mild moderate and severe hepatitis The numbers of mice examined per viral dose were as follows RA59 SJHM RA59 and RJHM 5 SA59 RJHM 10 A Af ter inoculation with 500 PFU RA59 induced moderate hepatitis whereas SJHM RA59 like RJHM and SA59RJHM caused minimal hepatocellular damage RA59 versus SJHM RA59 P H11021 0 01 RA59 versus RJHM or SA59 RJHM P H11021 0 001 Wilcoxon rank sum test B After inoculation with 10 5 to 10 6 PFU RJHM and SA59 RJHM induced mild hepatitis whereas SJHM RA59 caused the same extent of hepatitis as RA59 JHM background versus A59 background P H11021