GLUT1 Transcription Level Increased in Most Types of Cancer
We assessed GLUT1 mRNA expression in cancer and its corresponding normal tissue with four independent cancer data mining databases. 1) we used the Oncomine database. GLUT1 upregulated in almost all cancers, including breast, lung, kidney, pancreatic, bladder, head and neck, colorectal, esophageal, gastric, ovarian cancer and leukemia and lymphoma (Fig 1a). The upregulation in breast cancer was the greatest with 10 up and 1 down datasets. 2) we examined the mRNA expression in 33 types of human cancer with a combined TCGA and GTEx database (Fig 1b). It was significantly high in 16 types and low in 2 types. 3) we used the GPL570 Platform (HG-U133_Plus_2) in the GENT database and found it overexpressed in certain cancers, including breast cancer (Fig 1c). 4) using the UALCAN web tool, we found that the expression was significantly high in most cancers, including the breast (Fig 1d).
GLUT1 mRNA and Protein Increased in Breast Cancer
GLUT1 mRNA level upregulated in 10 out of 11 datasets in breast cancer (Oncomine). In Curtis Breast dataset [23] (Table 1), it increased by 2.317 folds in invasive breast carcinoma, 2.083 folds in invasive ductal breast carcinoma, and 1.615 folds in invasive lobular breast carcinoma (Fig 2a). In The Cancer Genome Atlas dataset (TCGA) (Table 1), it also increased in invasive breast carcinoma (2.251 folds), invasive ductal breast carcinoma (2.557 folds), and invasive lobular breast carcinoma (1.628 folds) compared to their normal counterparts (Fig 2b). Similarly, augmented expression was observed in invasive breast carcinoma compared to normal tissue (1.643 folds) in Gluck Breast dataset [24] (Table 1). Zhao et al. [25] (Table 1) reported an increase in invasive ductal (2.8 folds) and in lobular (2.075 folds) breast carcinomas versus normal samples. The similar results were obtained with GEPIA (Fig 2c) and UALCAN (Fig 2d) web tool. In addition, GLUT1 protein expression upregulated significantly in breast cancer tissues compared to normal tissues (Fig 2e).
Relation of GLUT1 mRNA and Pathological Characteristics
Using the UALCAN web tool (Table 2), we found that GLUT1 significantly upregulated in breast cancer independence of race (Fig S1 a), age (Fig S1 b), nodal metastasis status (Fig S1 d) and menopause status (Figure S1 g). The expression also significantly increased in different stages except stage4 (Fig S1 c), probably because of the small sample size of stage4. GLUT1 increased in almost all major subclasses (with TNBC types) except TNBC-LAR and TNBC-MSL (Fig S1 e). Apart from metaplastic, it increased in all histology subtypes (Fig S1 f).
GLUT1 Associated with Poor Prognosis and Metastasis
GLUT1 expression was negatively correlated with OS (Fig 3a) (GENT database), metastasis free survival and lung metastasis free survival (Fig 3b, Table 3) (PROGgeneV2 database) of patients. High mRNA expression of GLUT1 was predicted to decrease DMFS in all patients (Fig 3c) with different subclasses (Fig 3d). It was also negatively correlated with OS, RFS, DMFS and PPS in lymph node positive breast cancer (Fig 3e). Thus, the Kaplan-Meier curve and log rank test analyses revealed that the increase was significantly associated with recurrence, distant metastasis and poor prognosis.
Mutations, Methylation and Copy Number Alterations Analysis
We further analyzed GLUT1 mutations, methylation and copy number alterations in breast cancer. Mutations were altered in about 2% of sequenced patients in 33 types of cancer (Fig 4a) and about 6% of sequenced patients in breast cancer (Fig 4b) (cBioPortal web tools). Methylation was negatively correlated with mRNA expression with spearman and pearson analyses (Fig 4c). In the CNAs analysis, amplification and gain were predominantly correlated with GLUT1 expression (Fig 4d). These results suggest that the overexpression of GLUT1 could partly result from alterations in methylation and CNAs. Heatmaps of top (1-25) genes with hyper (Fig 4e) and hypo (Fig 4f) methylated promoters were also examined with UALCAN web tools.
Relation of GLUT1 and JUN and their expression in human breast cancer tissues
We investigated the transcription factors that have the capacity to bind to the promoter of GLUT1 in GeneCard and PROMO databases and found that transcription factor c-Jun encoded by gene JUN can bind to the GLUT1 promoter. We predicted the binding site sequences (Table 4). Our results suggest that the binding may regulate the GLUT1 gene expression or mRNA stability.
We analyzed the expression and prognostic value of JUN. The mRNA expression (p < 1E-12) and protein phosphorylation level at T239S243 (P = 0.002) of JUN significantly decreased in breast cancer compared to normal tissue (UALCAN database) (Fig 5a, Fig 5b). JUN expression varied with tumor stages (Fig 5c), and was positively correlated with OS in all patients (Fig 5d). In addition, the decreased methylation level was associated with increased overall survival (p = 0.022, HR = 1.736), indicating that the low expression of JUN might be partly due to methylation (Fig 5e). By using the bc-GenExMiner web tool, we found that GLUT1 expression was negatively correlated with JUN (p = 0.0001, r = -0.13) (Fig 5f).
We further found that GLUT1 was upregulated in 38 paired human breast cancer tissues while JUN was downregulated by qRT-PCR (Fig 5g, Fig 5h). Through IHC, we verified that the protein level of GLUT1 was significantly increased in tumor samples(Fig 5i).
Pathway Analysis of Alterations in GLUT1 and JUN and their Co-expressed Genes
We examined genes that co-expressed with GLUT1 and JUN in breast cancer in Adorno cell line dataset [26] (Fig S2 a, Fig S2 b), and found that the transcription level of GLUT1 and JUN were closely corelated and might contribute to certain signaling pathways. Hence, we performed GO and KEGG analysis of genes associated with GLUT1 and JUN.
With GO analysis, we found both GLUT1 and JUN regulated three aspects, i.e. cellular components [proteinaceous extracellular matrix, extracellular matrix component, basement membrane and collagen trimer (Fig 6a)], biological processes [Extracellular structure organization, extracellular matrix organization and collagen-activated signaling pathway (Fig 6b)], and molecular functions [Transcription coactivator activity, transcription factor activity, RNA polymerase II transcription factor binding and extracellular matrix constituent (Fig 6c)]. These results are known to relate to cancer metastasis.
With KEGG analysis, we found GLUT1 and JUN are associated with seven breast cancer related pathways. Among them, focal adhesion and ECM-receptor interaction were involved in tumorigenesis and metastasis (Fig 6d, Fig S3).