Figure 2. Identification of proteins targeted for degradation by HCMV using an inhibitor-based proteomic screen. (A) Results of the inhibitor-based screen. All 145 proteins downregulated >1.5 fold are plotted, with down-regulated proteins divided into 4 groups using rescue ratios of >1.5 as cut-offs. The table shows the number of proteins in each group. For rescue ratios, the denominator (mock with drug)/mock was limited to a minimum of 1 to prevent artificial ratio inflation. (B) Examples of positive controls from the existing literature that were validated in this screen. (C) Examples of degraded proteins rescued by both inhibitors (top panels), MG132 only (middle panels), or bortezomib only (bottom panels).
Certain proteins exhibited a greater degree of rescue with MG132 compared to bortezomib (Figure 2B, yellow dots ). Of the 21 proteins only rescued >1.5 fold by MG132, 13 (62%) exhibited bortezomib rescue ratios of >1.25 and <1.5, suggesting that many of this group of proteins may nevertheless be proteasomally degraded. These included the PDZ domain containing protein 11 (PDZD11) and transcriptional repressor BEN Domain Containing 3 (BEND3) (Figure S2, Tables S1-2). In contrast, 8/21 proteins appeared genuinely to be selectively rescued by MG132 but not bortezomib (bortezomib rescue ratio <1.25), including the fibroblast growth factor receptor Golgi Glycoprotein 1 (GLG1), E3 ligase Neural Precursor Cell Expressed, Developmentally Down-Regulated 4 (NEDD4) and carbohydrate sulfotransferase 14 (CHST14) (Figure 2C middle panel, Table S1 ). Similar data were obtained for treatments with 500 nM and 1 µM bortezomib, validating these findings (Figure 3 ). Interestingly, Gene Ontology annotation of all 8 proteins indicated an association with either the cell membrane, the Golgi apparatus, or vesicle secretion. Furthermore, comparison of data with our previous study examining protein rescue by MG132 or leupeptin indicated that GLG1, NEDD4 and CHST14 were also significantly rescued by treatment with the lysosomal protease inhibitor leupeptin (Figure 3 ), suggesting that a proportion of the proteins rescued by MG132 alone are degraded lysosomally.
Of proteins exhibiting a greater degree of rescue with bortezomib compared to MG132 (Figure 2A, purple dots ), 8/9 (89%) exhibited MG132 rescue ratios >1.25 but <1.5 (examples in Figure S2, 2C bottom panel ), suggesting that the majority of all proteins in this class were in fact rescued by both inhibitors. The one exception was LIM domain-containing protein AJUBA, whose MG132 rescue ratio was 1.16 in this data (Figure 2C bottom panel ), but neared significance in our previous study (Table S2 ); these differences may reflect relatively poor quantitation by only two or one peptides respectively.