Human Cytomegalovirus gH/gL/gO Promotes the Fusion Step of Entry into All Cell Types, whereas gH/gL/UL128-131 Broadens Virus Tropism through a Distinct Mechanism

Abstract

Interaction between gH/gL and the fusion protein gB is likely a conserved feature of the entry mechanism for all herpesviruses. Human cytomegalovirus (HCMV) gH/gL can be bound by gO or by the set of proteins UL128, UL130, and UL131, forming gH/gL/gO and gH/gL/UL128-131. The mechanisms by which these complexes facilitate entry are poorly understood. Mutants lacking UL128-131 replicate well on fibroblasts but fail to enter epithelial/endothelial cells, and this has led to the general assumption that gH/gL/UL128-131 promotes gB-mediated fusion on epithelial/endothelial cells whereas gH/gL/gO provides this function on fibroblasts. This was challenged by observations that gO-null mutants were defective on all of these cell types, suggesting that entry into epithelial/endothelial cells requires both of the gH/gL complexes, but the severe replication defect of the gO mutants precluded detailed analysis. We previously reported that the ratio of gH/gL/gO and gH/gL/UL128-131 in the virion envelope varied dramatically among HCMV strains. Here, we show that strains not only differ in the ratio, but also vary in the total amount of gH/gL in the virion. Cell-type-specific particle-to-PFU ratios of HCMV strains that contained different amounts of gH/gL/gO and gH/gL/UL128-131 were determined. Infection of both fibroblasts and epithelial cells was generally correlated with the abundance of gH/gL/gO, but not with that of gH/gL/UL128-131. The low infectivity of virions rich in gH/gL/UL128-131 but low in gH/gL/gO could be overcome by treatment with the chemical fusogen polyethylene glycol (PEG), strongly arguing that gH/gL/gO provides the conserved herpesvirus gH/gL entry function of promoting gB-mediated fusion for entry into all cell types, whereas gH/gL/UL128-131 acts through a distinct mechanism to allow infection of select cell types. IMPORTANCE The functions of HCMV gH/gL complexes in entry are unclear. Unlike the well-studied Epstein-Barr virus (EBV), where gH/gL and gH/gL/gp42 complexes both seem capable of promoting gB fusion during entry into different cell types, our studies here suggest that for HCMV, gH/gL/gO promotes gB fusion on all cell types, whereas gH/gL/UL128-131 broadens virus tropism through a distinct, as yet unknown mechanism. To our knowledge, this is the first suggestion of a herpesvirus gH/gL that does not act by promoting gB fusion, which might make HCMV a useful model to study the fundamental mechanisms by which herpesvirus gH/gL regulates gB fusion. Moreover, gH/gL/UL128-131 is a candidate vaccine target. Our findings help to explain the cell-type-dependent virus neutralization exhibited by anti-gH/gL/UL128-131 antibodies and underscore the importance of gH/gL/gO as another important part of vaccine or therapeutic strategies.