Bortezomib (PS-341) br To further explore the PTM burden
To further explore the PTM burden, we investigated the Bortezomib (PS-341) changes in genes associated with different PTMs:
Figure 6. Correlation between Protein Secretory Burden and Gene Expression Fold Change
Cancer types with a significantly (p < 0.05) negative correlation are colored blue, significantly positive cancers are colored red, and those with an insig-nificant correlation are colored gray. See also Figures S5 and S6.
N- and O-linked glycosylation, and protein disulfide bond oxida-tion and reduction (Figure S5C). Nearly half of the studied cancer types exhibited a significant coordinated expression increase in genes associated with glycosylation and/or disulfide bond for-mation, suggesting an additional effort to reduce secretory stress. The opposite behavior was observed for CHOL, which exhibited significant expression decreases associated with di-sulfide redox and N-linked glycosylation. All of the cancer types that did not show a coordinated expression increase associated with these PTMs were those exhibiting a significant decrease in their tissue-specific secretome, providing additional support for this relief strategy.
Additional Contributors to the UPR
Although HNSC and rectum adenocarcinoma (READ) exhibited a coordinated expression decrease in tissue-specific SP genes, as well as a positive correlation between SB score and gene expression fold change, these cancer types still show evidence of an activated UPR, unlike CHOL and THCA. Because the UPR can be triggered by sources of stress other than an overbur-dened secretory pathway (e.g., genome instability, hypoxia, nutrient deprivation) (Corazzari et al., 2017), it is possible that one or more of these alternative sources are contributing to UPR activation in HNSC and READ cells, despite their modified secretory profile. We therefore compared genome instability among the different cancer types using mutation profiles from TCGA whole-exome sequencing datasets. HNSC and READ samples exhibited similar mutation burdens (median of 134 and 127 somatic mutations per sample, respectively), which were >2-fold greater than CHOL (63 median mutations per sam-ple) and >10-fold greater than THCA (12 mutations per sample) (all p < 10 6, one-sided Wilcoxon rank-sum test) (Figure S6). These results support the possibility that other sources of stress beyond those directly involving the ER and secretory pathway
could be responsible for elevated UPR activation in HNSC and READ.
The secretome is regarded as an attractive reservoir of disease biomarkers, as its extracellular nature offers the potential to evaluate physiological status through easily accessible biofluids (Kulasingam and Diamandis, 2008; Schaaij-Visser et al., 2013; Stastna and Van Eyk, 2012). Furthermore, there are many protein biomarkers in use for the diagnosis or monitoring of different cancer types based on their abundance in serum, plasma, or urine, such as PSA, CA-125, CA19-9, and NuMA for prostate, ovarian, pancreatic, and bladder cancer, respectively (Fuze€´ry et al., 2013).
Beyond its potential as a reservoir of biomarker candidates, the cancer secretome is known to play a crucial role in tumor development and invasion. We sought to evaluate cancer-asso-ciated shifts in secretome expression with regard to the function of the encoded proteins. The majority of shared pan-cancer changes in secretome expression were decreases and included proteins associated with functions such as cell-cell and cell-ma-trix adhesion, tumor suppressors with anti-proliferative or anti-migratory activities, and immune response. These proteins har-bor potential therapeutic opportunities, either by targeting the factors driving their expression decrease or through direct use of the tumor suppressor as a therapeutic peptide (Bonin-Debs et al., 2004; Guo et al., 2014; Oricchio et al., 2011). For example, ANGPTL1, which was among the top 1% core decreased secre-tome proteins, has been demonstrated to suppress cell migra-tion, invasion, angiogenesis, metastasis, and/or therapy resis-tance in hepatocellular carcinoma (Chen et al., 2016; Yan et al., 2017), colorectal cancer (Chen et al., 2017), and lung and breast cancers (Kuo et al., 2013).