Glycoproteomics identifies HOMER3 as a potentially targetable biomarker triggered by hypoxia and glucose deprivation in bladder cancer

Abstract

Background Muscle invasive bladder cancer (MIBC) remains amongst the deadliest genitourinary malignancies due to treatment failure and extensive molecular heterogeneity, delaying effective targeted therapeutics. Hypoxia and nutrient deprivation, oversialylation and O-glycans shortening are salient features of aggressive tumours, creating cell surface glycoproteome fingerprints with theranostics potential. Methods A glycomics guided glycoproteomics workflow was employed to identify potentially targetable biomarkers using invasive bladder cancer cell models. The 5637 and T24 cells O-glycome was characterized by mass spectrometry (MS), and the obtained information was used to guide glycoproteomics experiments, combining sialidase, lectin affinity and bottom-up protein identification by nanoLC-ESI-MS/MS. Data was curated by a bioinformatics approach developed in-house, sorting clinically relevant molecular signatures based on Human Protein Atlas insights. Top-ranked targets and glycoforms were validated in cell models, bladder tumours and metastases by MS and immunoassays. Cells grown under hypoxia and glucose deprivation disclosed the contribution of tumour microenvironment to the expression of relevant biomarkers. Cancer-specificity was validated in healthy tissues by immunohistochemistry and MS in 20 types of tissues/cells of different individuals. Results Sialylated T (ST) antigens were found to be the most abundant glycans in cell lines and over 900 glycoproteins were identified potentially carrying these glycans. HOMER3, typically a cytosolic protein, emerged as a top-ranked targetable glycoprotein at the cell surface carrying short-chain O-glycans. Plasma membrane HOMER3 was observed in more aggressive primary tumours and distant metastases, being an independent predictor of worst prognosis. This phenotype was triggered by nutrient deprivation and concomitant to increased cellular invasion. T24 HOMER3 knockdown significantly decreased proliferation and, to some extent, invasion in normoxia and hypoxia; whereas HOMER3 knock-in increased its membrane expression, which was more pronounced under glucose deprivation. HOMER3 overexpression was associated with increased cell proliferation in normoxia and potentiated invasion under hypoxia. Finally, the mapping of HOMER3-glycosites by EThcD-MS/MS in bladder tumours revealed potentially targetable domains not detected in healthy tissues. Conclusion HOMER3-glycoforms allow the identification of patients’ subsets facing worst prognosis, holding potential to address more aggressive hypoxic cells with limited off-target effects. The molecular rationale for identifying novel bladder cancer molecular targets has been established. Graphical abstract

The Effect of Hypoxia and Hypoxia-Associated Pathways in the Regulation of Antitumor Response: Friends or Foes?

“…In particular, the highly hypoxic muscle-invasive bladder cancer (MIBC) overexpresses the cancer-associated carbohydrate antigen sialyl-Tn (STn), which has been reported to be at least in part due to a HIF-1α-dependent cell survival strategy that favors cell migration and invasion ( 133 ). Recently, through the application of glycoproteomics in bladder cancer, the same group demonstrated cell surface expression of an ordinarily intracellular protein, homer homolog 3 (HOMER3), carrying short-chain O-glycans that are characteristic of membrane proteins ( 134 ). They further reported that under glucose deprivation and hypoxic conditions, HOMER3 contributed to the tumor cell’s invasive capacity.…”

“…Of interest, cell-surface expression of this protein was associated with significantly worse survival in MIBC patients. Furthermore, while HOMER3 expression was not cancer-specific, STn and HOMER3 did not co-express in healthy tissue, suggesting that HOMER3-STn could be a tumor-specific biomarker that can be used to target the more aggressive cancer cell populations residing in the hypoxic TME ( 134 ). Similarly, some cancer-specific glycoconjugates, like N-glycolyl (NeuGc) GM3 gangliosides, are promising therapeutic targets, as their increased expression is characteristic in tumors and almost not present in healthy human tissues ( 135 ).…”

A roadmap for translational cancer glycoimmunology at single cell resolution

“…Several tumour microenvironmental factors have been suggested to be upstream of glycan-associated phenotypic changes, including variations in oxygen [ 11 ], nutrients [ 24 ], and inflammatory cytokines levels [ 25 ]. The underlying structural alterations contribute to all accepted cancer hallmarks [ 9 ], highlighting the critical functional implications of glycans in cancer cells behaviour and fate [ 11 , 26 , 27 ], with considerable negative impact in clinical outcomes [ 10 , 28 , 29 ]. The most well-known glycome alterations result from premature stop in protein O -GalNAc glycosylation (occurring in Ser/Thr residues of proteins), yielding immature glycans, such as the Tn, sialyl-Tn (STn) and T antigens, rather than more extended and complex glycosidic chains.…”

“…The sialylated Tn antigen, or sialyl-Tn (STn), is also frequently expressed in tumour tissues, including breast, gastrointestinal, lung, prostate, oesophagus, and bladder cancers [ 43 ]. It modulates key aspects of cancer progression, being an independent predictor of poor prognosis [ 14 , 29 , 43 ] frequently found in glycoproteins intimately linked to cancer aggressiveness [ 10 , 27 ]. Numerous reports describe STn as a driver of decreased cell adhesion [ 11 , 44 ], increased migratory and invasive [ 11 , 45 ] capacity of tumour cells, and decreased chemotherapy-induced apoptosis [ 46 ].…”

“…Namely, overexpression of ST3Gal-I , responsible for ST biosynthesis, is a biomarker of metastatic potential [ 70 ], while ST antigen was demonstrated to impact on tumour cell proliferation, migration, and apoptosis. Of note, ST antigen has also been identified in several key cancer glycoproteins, some of which displaying prognostic potential [ 27 , 69 ].…”