Discussion
HCMV is known to be a master regulator of host immunity, achieving lifelong persistence in infected individuals by utilising a wide range of strategies to modulate host protein expression. These include the deployment of proteins to target host factors for degradation. Here, we provide a searchable database that systematically details the route of degradation of cellular proteins during the establishment of a productive HCMV infection. Furthermore, this data can be used to predict molecules of key importance in antiviral immunity to HCMV on the basis of their degradation.
MG132 is a less selective proteasomal inhibitor than bortezomib, having previously been reported to inhibit lysosomal degradation pathways via inhibition of calpains and cathespsins (Kisselev and Goldberg, 2001), in addition to the proteasome. In our previous publication, 75% of proteins rescued by leupeptin at 12h of infection were also rescued by MG132. The usefulness of comparing this broad proteasomal/lysosomal inhibitor with the specific proteasomal inhibitor bortezomib is the identification that 62-85% of proteins rescued by MG132 were also rescued by bortezomib, suggesting that the proteasome is the predominant route for early protein degradation during HCMV infection. Overall, of all downregulated proteins, 44% were rescued by at least one of MG132 or bortezomib. It is possible that in order to downregulate certain proteins, HCMV must employ degradative pathways in order to achieve sufficiently rapid change in protein abundance.
We and others have previously shown that membrane proteins are targeted for lysosomal degradation during HCMV infection (Weekes et al., 2013;Fielding et al., 2014;Hsu et al., 2015;Fielding et al., 2017), and data here identified that all proteins rescued by MG132 but not bortezomib had a membrane origin. Certain proteins were exclusively degraded by a non-proteasomal route, including GLG1 and CHST14. Extension of these inhibitor studies to examining membrane-enriched samples, for example samples enriched for plasma membrane proteins (Weekes et al., 2013;Weekes et al., 2014) would therefore be of substantial interest, and may identify a distinct degradative route for proteins originating from these compartments.
Comparison of data from this study with our previous transcriptional analysis of host gene expression during infection suggested that 54% of the 81 proteins with MG132 and bortezomib rescue ratios <1.5 were more than 1.5-fold transcriptionally downregulated, which would be expected to be a major mechanism of protein downregulation in the absence of degradation. The fold change cutoff of 1.5 for both downregulation by HCMV, and rescue by either inhibitor was based on a significance threshold of p<0.01, however had the effect of excluding proteins with ‘borderline’ rescue ratios of >1.25 but <1.5. 39/81 proteins with MG132 and bortezomib rescue ratios <1.5 exhibited rescue ratios for MG132 or bortezomib or both that were nevertheless >1.25, suggesting that this group of proteins included some candidates that downregulated by degradation, at least in part.
Overall, this analysis of host protein degradation during HCMV infection has not only identified proteasomal degradation as a key mechanism subverted by the virus early during infection, but has also generated a shortlist of proteasomally degraded proteins enriched in known HCMV restriction factors. Further investigation into the role of the other proteins in this shortlist is warranted to determine if they also have restrictive capabilities. Identification of HCMV restriction factors, understanding the mechanism by which they restrict infection and identification of viral antagonists that target these factors for degradation are of fundamental interest due to the potential for therapeutic intervention.