Genetic alteration of UL75 gene of HCMV and its effect on pathogenicity of the virus

Abstract

Background: Human cytomegalovirus (HCMV) is the main cause of viral congenital infections and sensorineural hearing loss (SNHL). About 1% of all live births are affected by congenital human cytomegalovirus (cHCMV). Only about 10-15% of the infected babies at the time of birth show severe symptoms such as mental retardation and microcephaly, but around 10-15% of asymptomatic children will develop late-onset symptoms such as SNHL later in life. Prognostic identification of the outcomes in the infected infants is impossible due to the absence of reliable markers and lack of a deep understanding of the pathogenesis of the virus. Objective: The aim of the current study was to determine whether mutation in the envelope glycoprotein H of HCMV is associated with the pathogenicity of the virus and with outcomes relevant to congenitally infected infants. The hypothesis of this project is that known mutation of the UL75 gene that encodes the envelope glycoprotein H of HCMV alters the characteristics of the virus allowing it to become more or less pathogenic which may influence the outcome of infection in congenitally infected infants. Methodology: A range of assays was established in vitro to measure the phenotypic effects of HCMV, including rate of growth, pro-inflammatory cytokine response and glycosylation pattern. These assays were initially used to compare the characteristics of five laboratory strains of HCMV: AD169, Towne, Davis, Toledo and Merlin. Next, homologous recombination by allelic exchange to bacterial artificial chromosome BAC-RCMV1111 which carries the genome of strain Merlin was used to replace the wild-type gH2 gene in Merlin with the gH-1 gene excised from lab strain Davis. The recombinant virus was successfully grown in culture 18 and the phenotypic characteristics of this virus evaluated and compared to the control strain Merlin. Results: The growth of the five laboratory strains in cell culture showed that the DNA of the five laboratory strains was detectable in cells and supernatant indicating viral replication. However, the rate of the growth of each strain was different. AD169 had the shortest replicative cycle and the fastest growth in cell culture. While the slowest growth in culture occurred for Merlin and Davis. The CPE of the five laboratory strains appeared in cell culture. The strains were assessed against a panel of pro-inflammatory cytokines to see which were induced in the cells at 24 and 72 hours. In fact, very few of the cytokines were stimulated, with only MCP-1, IL-6, IL-8, IL-33, sTREM-1, sTNF-R1, sCD40L and CX3CL1 showing any quantifiable changes. AD169 stimulated the highest level of cytokines, in particular for MCP-1, IL-6, IL-8 and IL-33. The lectin assay gave a distinct result for Merlin as opposed to all other viral strains. Four lectins with different terminal sugar specificities were highly upregulated in Merlin but little response for any lectins was seen in any of the other lab strains. Success of the UL75 substitution in the BAC- RCMV111 was confirmed using Sanger sequencing to sequence the gene of interest. Also, Immunofluorescence-labelled antibodies against the viral gene phosphoprotein 65 (pp65) confirmed the active replication of the recombinant virus, and the structural integrity was confirmed by visualisation of the virus using electron microscopy. When the gH1 gene from Davis was substituted into Merlin it changed the characteristics of the virus considerably: it had a slower growth rate and replicated to lower titres. It apparently stimulated a much more pro-inflammatory response as measured by cytokine production than did Merlin, and in contrast did not show the same strong lectin binding pattern that was found for Merlin. 19

Conclusion: These data show that different laboratory strains of HCMV which carry different glycoprotein types, have quite different growth characteristics and that a single substitution of gene UL75 in strain Merlin with a mutated form of the gene produced a significant alteration in the phenotype of the virus. Together these data suggest that different glycoprotein types can have a significant impact on the characteristics of the virus which in vivo would undoubtedly impact on pathogenesis and may influence the outcome of infection in congenitally infected infants and other types of CMV disease.