Deep Intact Glycoproteomic Analysis Reveals Multi-Level Glycosylation Heterogeneity in Renal Tumors

Extensive changes in glycosylation are observed in many malignancies, but limited information is available about the modified proteins and sites. We used intact glycoproteomics to map N-glycosites in 32 surgically resected renal tumors and paired adjacent normal tissues. We identified 853,340 total and 54,354 unique N-linked glycopeptides from 1,940 glycoproteins containing 220 glycan compositions. The Shannon diversity index was used to characterize glycosylation diversity and identified tumor-associated differences enriched in glycoproteins mapping to cell-adhesion and immunity processes. Although more glycopeptides were identified in tumors, whole-glycoproteome Shannon diversity did not differ significantly between tissue types; tumor-associated differences were concentrated at specific glycoproteins and glycosites. . In tumors, complex fucosylated and complex sialylated glycans were more common. RNA-seq and uniLASSO-based integration nominated glycosylation-related transcripts associated with site-specific glycan-category changes. Changes in total protein abundance were moderately associated with changes in Shannon diversity at the protein (R2 = 0.290) and glycosite (R2 = 0.279) levels, but most of the linear variation was not accounted for by protein abundance alone. Diversity changes were also concentrated in functional and structural protein domains. The results provide prioritized candidates and hypotheses for independent validation and mechanistic studies of protein glycosylation changes in renal tumors.