Die Rolle von Glykanen für die durch das Glykoprotein O vermittelte Infektion des humanen Cytomegalievirus
Zusammenfassung der Projektergebnisse
The human cytomegalovirus (HCMV) is the most common infectious cause of congenital birth defects. It is particularly challenging for immunization strategies that HCMV can cause congenital disease also in children of mothers that have antibodies against the virus. The gHgLgO entry glycoprotein complex, and particularly the gH subunit, is a major target of neutralizing antibodies. It had been shown that viruses lacking gO are not only much less infectious, but also more sensitive to neutralization by anti-gH antibodies, indicating that the gO component of the gHgLgO complex interferes with neutralization. There are 8 genetic variants of gO that display heterogeneity in up to 19 predicted glycosylation site. For other viruses such as HIV and SARS-COV2 it has been shown that glycans can shield functionally important structures from neutralizing antibodies. In addition, it is known that glycans can also help the virus to attach via glycan binding structures on the cell surface. Therefore, this project tested the hypotheses that i) gO glycosylation shields gHgLgO from neutralizing antibodies and that ii) glycosylation of gO directly impacts HCMV entry. A variety of different methods was established to address these questions. Recombinant viruses were generated to test the impact of genotypic differences in gO. Soluble gHgLgO complexes were used to establish methods that allow direct investigation of the glycosylation site occupancy of gHgLgO as well as the interactions of gHgLgO with the known gO receptor PDGFRα, neutralizing antibodies and the surface of host cells. To study the role of gO glycans in the context of viral infection, the natural diversity of gO genotypes was utilized. The gOs of strain TR and strain Towne (TN) belong to different genotypes. Their protein sequence is about 30% different. Accordingly, they are also predicted to differ regarding glycosylation at 5 positions, 3 of which were confirmed by mass spectrometry in this project. It had been found that a recombinant virus in which the endogenous gO of strain TR was replaced by that of TN was partially resistant to the neutralizing anti-gH antibody 14-4b. To dissect the impact of differences between the genotypes regarding glycosylation from the overall different protein sequence, a set of gO mutants was generated that introduce the glycosylation pattern of TNgO into TRgO. Comparing these virus mutants to TR wild type demonstrated that none of the five glycosylation sites had a significant impact on antibody sensitivity. This indicates that structural differences between the gO genotypes, rather than differences in glycosylation are the reason for the reduced sensitivity of TN. Using ELISA and biolayer interferometry it was tested if the binding affinity of 14-4b to the trimer was altered upon replacing TRgO with TNgO. The results obtained by both methods demonstrated that binding of the antibody to gH was independent of which gO subunit was present. This contrasts with the observed difference in neutralization and suggested that the 14-4b epitope is differentially oriented depending on the gO genotype, thereby changing the effect of antibody binding rather than binding itself. Even full enzymatic deglycosylation of the trimer did not change the affinity of 14-4b significantly, demonstrating that the accessibility of its epitope is unaltered by the presence or absence of glycans on the protein. Taken together, this work demonstrated that glycosylation of gO does not influence binding of the neutralizing antibody 14-4b. A similar approach was taken to study the impact of gO glycosylation on receptor binding, revealing that genotypic differences between TNgO and TRgO, including strain-specific glycosylation, do not affect binding to PDGFRα and that gHgLgO dependent attachment is unlikely to be mediated by glycan binding proteins. Additionally, a conserved site in gO was identified that is dispensable for PDGFRα binding but crucial for attachment to fibroblasts and epithelial cells. Future projects will utilize the established methods to follow up on the impact of the trimer on epithelial cell infection and a novel panel of neutralizing anti-gHgLgO antibodies will be used to further investigate the impact of gO glycosylation on neutralization via different epitopes.
Projektbezogene Publikationen (Auswahl)
- Polymorphisms in Human Cytomegalovirus Glycoprotein O (gO) Exert Epistatic Influences on Cell-Free and Cell-to-Cell Spread and Antibody Neutralization on gH Epitopes. Journal of Virology. 2020;94:2051-2070
Day LZ, Stegmann C, Schultz EP, Lanchy JM, Yu Q, Ryckman BJ
(Siehe online unter https://doi.org/10.1128/jvi.02051-19) - Mutagenesis of Human Cytomegalovirus Glycoprotein L Disproportionately Disrupts gH/gL/gO over gH/gL/pUL128-131. Journal of Virology. 2021;95(17):612-633
Schultz EP, Yu Q, Stegmann C, et al.
(Siehe online unter https://doi.org/10.1128/jvi.00612-21)
