Aberrantly Glycosylated GLUT1 as a Poor Prognosis Marker in Aggressive Bladder Cancer

Abstract

Muscle-invasive bladder cancer (MIBC) remains a pressing health concern due to conventional treatment failure and significant molecular heterogeneity, hampering the development of novel targeted therapeutics. In our quest for novel targetable markers, recent glycoproteomics and bioinformatics data have pinpointed (glucose transporter 1) GLUT1 as a potential biomarker due to its increased expression in tumours compared to healthy tissues. This study explores this hypothesis in more detail, with emphasis on GLUT1 glycosylation patterns and cancer specificity. Immunohistochemistry analysis across a diverse set of human bladder tumours representing all disease stages revealed increasing GLUT1 expression with lesion severity, extending to metastasis, while remaining undetectable in healthy urothelium. In line with this, GLUT1 emerged as a marker of reduced overall survival. Revisiting nanoLC-EThcD-MS/MS data targeting immature O-glycosylation on muscle-invasive tumours identified GLUT1 as a carrier of short glycosylation associated with invasive disease. Precise glycosite mapping uncovered significant heterogeneity between patient samples, but also common glycopatterns that could provide the molecular basis for targeted solutions. Immature O-glycosylation conferred cancer specificity to GLUT1, laying the molecular groundwork for enhanced targeted therapeutics in bladder cancer. Future studies should focus on a comprehensive mapping of GLUT1 glycosites for highly specific cancer-targeted therapy development for bladder cancer.

Keywords: GLUT1; bladder cancer; glycomics; glycoproteomics; targetable biomarkers.

Figure 1

(a) Expression of GLUT1 (red), T (green), and ST (green) antigens in 5637 and T24 cell lines. Cells nuclei were stained with DAPI (blue). From left to right, the first panel shows GLUT1 at the cell membrane and, to some extent, in the cytoplasm of both cell lines. The second panel strongly suggests that cell lines did not express the T antigens, while overexpressing ST antigens at the cell membrane as outlined in the third panel. (b) Western blot for GLUT1 and ST antigens in GLUT1 immunoprecipitants from 5637 and T24 cell lines. Both 5637 and T24 cell lines demonstrated similar patterns. Namely, the GLUT1 blots show a band above 50 kDa consistent with the molecular weight of canonical GLUT1 forms (55 kDa). It also shows several bands below 37 kDa. The blot for ST antigens supports that lower-molecular-weight isoforms (under 37 kDa) carry ST antigens. Chemiluminescence was detected using a ChemiDoc Imager (Bio-rad, Hercules, CA, USA).

Figure 2

(a) Expression of GLUT1 across all stages of BLCA, from healthy urothelium to Ta, T1, T2, T3, and T4 tumours (n = 104), lymph node metastases, and distant metastases (n = 10). GLUT 1 was not expressed by the healthy urothelium, being increasingly expressed across disease stages and metastases. (b) GLUT1 expression in relevant healthy tissues. Thyroid, liver, testis, lung, stomach, pancreas, colon, and small intestine tissue samples were found negative for GLUT1, except for resident leukocyte populations. (c) Extension and intensity of GLUT1 expression across disease stages. GLUT1 expression increased with the severity of disease, being overrepresented in muscle-invasive disease compared to non-invasive lesions and the healthy urothelium. Comparisons were performed using an ordinary one-way ANOVA. Statistical significance was considered when p < 0.05. (d) Ten-year overall survival (OS) of BLCA patients carrying tumours overexpressing GLUT1 vs. patients carrying tumours with GLUT1 underrepresentation. GLUT1 overexpression in BLCA determined decreased 10-year OS of patients. Comparison of estimates was carried out using log-rank tests. Statistical significance was considered when p < 0.05.

Figure 3

(a) Co-localization of STn, ST, and GLUT1 in muscle-invasive BLCA. Co-localization of GLUT1 with STn and ST antigens in the same tumour area suggests that GLUT1 carries these PTMs. (b) GLUT1 glycopeptide presenting STn and ST antigens, identified by nanoLC-EThcD-MS/MS. The presence of product ions for HexNAc (Tn; m/z 204.08) and HexNAcHex (T; m/z 366.14) modifications (highlighted in the MS/MS spectra) strongly suggests the presence of STn and ST antigens in GLUT1 peptides. Discriminatory fragmentations are also highlighted. (c) GLUT1 glycosites identified by nanoLC-EThcD-MS/MS having its canonical isoform as reference. Twenty glycosites were identified for GLUT1, seven of which were observed in two or more samples (S95, T234, T238, S248, S4, S5, T9, T459, S465, and S473), supporting common glycopatterns between patients. STn antigen glycosylation of T234 was found in all samples. * Represents the samples to which the glycosite was assigned.