Abstract 1593 Neonatal Infectious Disease Platform, Monday, 5/3

Since beta herpesviruses exhibit a remarkable degree of species specificity, studies of the pathogenesis of infection with human cytomegalovirus (HCMV) cannot be performed in animal models. The guinea pit cytomegalovirus (GPCMV) is useful for studies of the pathogenesis of congenital infection since, in contrast to the cytomegaloviruses of other small mammals, GPCMV crosses the placenta, causing infection in utero. However, the lack of information about the molecular biology of GPCMV and technical difficulties inherent to the generation of recombinant cytomegaloviruses has rendered detailed viral genetic analyses problematic. We report the utilization of the E. coli guanosine phosphoribosyltransferase (gpt) gene in the development of a technique to enable the positive selection of recombinant GPCMV. Guinea pig fibroblast lung (GPFL) cells were co-transfected with GPCMV DNA and a plasmid containing a double cassette carrying both the gpt gene (under control of the HSV thymidine kinase promoter) and the green fluorescent protein (GFP) gene (under control of the beta actin promoter). Homologous flanking sequences for recombination corresponding to the Hind III "N" region of the GPCMV genome were included. This region was selected for insertional mutagenesis because it is co-linear with regions of the HCMV genome known to be dispensable. We sequenced the Hind III "N" region to analyze for possible coding regions: although we identified transcriptional activity from this region by Northern blot, the function of any gene(s) mapping to this region is currently unclear. Selection of gpt positive recombinant virus was achieved by inclusion in the tissue culture media of the de novo purine synthesis pathway inhibitor, mycophenolic acid, and the gpt enzyme substrate, xanthine. Following serial passage in selective media and cloning by limiting dilution, a pure population of recombinant GPCMV was generated. Southern blot analysis was used to confirm the clonality of the recombinant. In a one-step growth curve analysis in GPFL cells, replication kinetics of the recombinant virus were identical to wild type. These data suggested that the locus of insertion in the Hind III "N" region is not essential for viral replication. We next evaluated the pathogenicity of the recombinant virus in strain 2 guinea pigs. At seven days following intraperitoneal inoculation of 1×106 pfu, recombinant virus could be isolated from blood culture, suggesting that the recombinant virus retains the potential for virulence in vivo. The development of this mutagenesis system allows for analysis of the role of specific viral genes in pathogenesis studies, and the presence of the GFP marker in the GPCMV genome should prove useful for monitoring viral replication in vivo. Further studies to define tissue distribution and pathogenicity of recombinant virus in guinea pigs are in progress.