• 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2020-03
  • 2020-07
  • 2020-08
  • 4μ8C br The trend of expression decreases among the


    The trend of expression decreases among the secretome was also observed in the cancer-specific analyses, in which liver-related cancers (LIHC and CHOL) exhibited a particularly strong decrease in the expression of liver-specific SP genes. This reduced expression of tissue-specific genes in hepatocellular carcinoma has been explored previously; the extent of expres-sion decrease was shown to negatively correlate with tumor grade or degree of dedifferentiation (Ge et al., 2005; Uhlen et al., 2017). We investigated this further, focusing on the subset of proteins targeted to the secretory pathway and spanning many different cancer types. Using tissue-specific gene classifi-cation from the HPA, this phenomenon of a significant decrease in expression of SP genes specific to the tissue of origin of the cancer was found to hold across the majority of examined can-cer types.
    Since UPR activation (Urra et al., 2016) and increased expres-sion of secretory pathway machinery (Dejeans et al., 2014) are common in many cancers, our results suggest a common pattern by which tumor 4μ8C modify their secretory profile to alle-viate ER stress by reducing the production of tissue-specific components in favor of tumorigenic factors. Consistent with this hypothesis, CHOL and THCA, which exhibited among the strongest decreases in their tissue-specific SP genes, were associated with the weakest UPR activation and displayed no 
    bias toward the increased expression of low-burden (shorter and with fewer PTMs) SP genes. Conversely, the few cancer types with an insignificant decrease in their tissue-specific SP gene expression (BLCA, ESCA, PRAD, and UCEC) exhibited increased expression associated with the UPR and displayed an apparent bias in expression toward lower-burden SP genes.
    Given that different tissues exhibit fine-tuned expression of their secretory machinery to accommodate their unique secre-tome profile (Feizi et al., 2017), it is reasonable to expect that a malignant cell could quickly overload this system and induce ER stress upon increasing the production of tumorigenic compo-nents without an accompanying decrease in other SP genes. A number of anti-cancer therapies that activate the UPR are under development or approved for clinical use, demonstrating the importance of this system in cancer treatment (Hetz et al., 2013). Although many cancers are known to leverage an acti-vated UPR for its cytoprotective and restorative effects, UPR-targeted therapies function by driving the response further to a pro-apoptotic regime. We reasoned that the strong decrease in tissue-specific SP gene expression observed in CHOL or THCA cells, coupled with the insignificant coordinated expres-sion increase in UPR-associated genes, could indicate a height-ened sensitivity of these cancers toward this form of stress. In support of this hypothesis, treatment of CHOL cells in vitro and in a subcutaneous transplantation mouse model with bortezo-mib, which activates the UPR via proteasome inhibition, was shown to inhibit proliferation and induce apoptosis (Vaeteewoot-tacharn et al., 2013). Furthermore, bortezomib has been found to induce apoptosis in THCA cell lines with half-maximal inhibitory concentration (IC50) values lower than those of other cancer types (e.g., glioma, colon, renal, ovarian, prostate) (Mitsiades et al., 2006), whereas bortezomib treatment of BLCA cell line 253JB-V did not result in significant apoptosis and could not inhibit 253JB-V tumor growth in mice unless combined with another therapy (gemcitabine) (Kamat et al., 2004).
    Overall, the functional diversity and close involvement of the secretome in a number of critical tumorigenic and metastatic processes highlights the importance of this group of proteins in cancer pathophysiology and presents a strong case for its tar-geting in anti-cancer therapeutic development. In addition, the ranked list of secretome biomarker candidates for each of the 32 different cancer types is expected to help facilitate the devel-opment of more accurate, less invasive diagnostic methods.