Glucose-6-phosphate Dehydrogenase (G6PD): the Role in Tumor Progression and Immunotherapy

doi:10.21203/rs.3.rs-5034307/v1

Background: Numerous studies have shown that glucose-6-phosphate dehydrogenase (G6PD) is a tumor-promoting factor in a variety of malignancies. However, it is not entirely clear the role and the potential molecular mechanism of G6PDH in the pathogenesis or clinical prognosis of different tumors.

Methods: This study first investigated the pan-tumoral expression of G6PD, then G6PD gene expression were studied in cancers, survival prognosis, tumor immunity, immunosuppressive cell infiltration, DNA methylation, gene alteration assay, and response to immunotherapy. We also investigated the function of G6PD in the development and prognosis of various cancers.

Results: Our results suggest that G6PD expression was higher in tumor tissues than in normal tissues and was related to tumor stage, metastasis, and prognosis in most cancers and subtypes of the Cancer Genome Atlas. High G6PD expression is protective in a small number of cancers, including paraganglioma, pheochromocytoma, and ovarian serous cystadenocarcinoma. However, it is a risk factor for the majority of cancers. The prognosis for progression-free survival was better in people with G6PD alterations than in those without them. G6PD and immune cell infiltration were significantly positively correlated in prostate cancer, pancreatic adenocarcinoma, liver cancer, and low-grade glioma of the brain. Additionally, the degree of G6PD methylation was shown to inversely correlate with mRNA expression. The PGD, GCLM, SRXN1, TRIM16L, and TXNRD1 genes all showed significant positive correlation with G6PD expression level. The major genetic alterations were missense mutations in G6PD, and mutations at the R192C/S locus were detected in cutaneous melanoma, uterine cancer, and thyroid carcinoma. In several malignancies, G6PD expression is associated with immunological and chemotherapeutic outcomes.

Conclusions: According to the study, patients who expressed more G6PD generally had better therapeutic outcomes. Our study highlights the role of G6PD in oncogenesis, detection, prognosis, and treatment planning.

G6PD

Pan-tumor

DNA methylation

Oncogenic

Immunotherapy

It is thought that both the genome and epigenetics may change with the tumors start and get worse, and that these changes may affect each other [1–3]. Genetic changes in tumors have been linked to epigenetic modifications, which have been shown to contribute to genome instability and mutations [4–6]. In addition to tumor cells, endothelial cells, tumor-associated fibroblasts, and immune cells, the tumor microenvironment (TME) has a noncancerous cellular matrix that contains a large number of different peptide components, such as growth factors, chemokines, cytokines, and antibodies) [7]. The dynamic Changes of TME paly an important impact in tumor formation. According to several studies, the efficiency of immunotherapy often depends on the interaction between tumor cells and immune modulation inside the TME [8, 9]. Because of the connection between the human immune system and malignancies, immune checkpoint blockade (ICB) medicines have quickly developed as an effective tumor therapy in the last few years [10].

Glucose-6-phosphate dehydrogenase (G6PD) is the first and rate-limiting enzyme in the catalysis of the pentose phosphate pathway (PPP), which it is responsible for the majority of the cellular NADPH generated by most tumor cells [11]. G6PD has oncogenic activity and is important for cell survival and embryonic development [12]. The potential oncogenic activity of G6PD may stem from the fact that it is an angiogenic factor mediating endothelial cell growth, migration, and capillary formation[13]. It has also been shown that G6PD is a pro-carcinogenic factor in tumors including colorectal cancer, hepatocellular carcinoma, and breast cancer[14]. Meanwhile, G6PD can also affect the efficacy oxaliplatin, cisplatin and other tumor drugs[15]. In addition, G6PD may be a promising therapeutic target for treating small-cell lung cancer, glioma, and clear-cell renal cell carcinoma [16–18]. However, it was reported that G6PD was not strictly required for tumorigenesis and at most moderately promoted disease progression in the K-Ras breast cancer model [19]. Therefore, more in-depth experimental research is needed for the above phenomenon.

In this study, we utilized the TCGA project and other datasets to do a pan-tumor analysis of G6PD. We also looked at gene expression, survival prognosis, tumor immune and immunosuppressive cell infiltration, G6PD-related gene enrichment analysis, DNA methylation, genetic change analysis, and immunotherapy response to explore the role and potential molecular mechanism of G6PDH in the pathogenesis or clinical prognosis of different types of tumors.

G6PD protein structure, location and relevance to disease

According to previous reports [47], use the "Start PROTTER" module in the PROTTER (https://wlab.ethz.ch/protter/start/) website to get the human source G6PD protein topology. Using the Human protein atlas (https://www.proteinatlas.org/), the location of G6PD in cells and tissues and the expression of G6PD in different cell cycles were obtained. At the same time, the expression of G6PD in different organizations was obtained through the website of Genecard (www.genecards.org). Then, enter the GPS-Prot (http://gpsprot.org/index.php) [48] and OPENTARGET platform (https://platform.opentargets.org/) pages to get the protein and disease association maps related to G6PD. A flow chart of the research methodology and design is shown in Fig. 1.

Gene expression analysis

The expression difference of G6PD in tumors and adjacent normal tissues was analyzed between tumor and adjacent normal tissues across the Tumor IMmune Estimation Resource (TIMER2.0) (http://timer.cistrome.org/). Table 1 shows the abbreviations and full names of the 33 tumors involved in this study. The UALCAN portal (http://ualcan.path.uab.edu/analysis-prot.html), an interactive web resource for analyzing tumor Omics data [49], was used to analyze the protein expression levels of G6PD in tumors. Furthermore, to evaluate the differential expression of G6PD among tumor stages, we used the expression DIY module of GEPIA2 (http://gepia.cancer-pku.cn/) to make pathological stages (stage I, stage II, stage III, and stage IV) of all TCGA tumors. In addition, differential expression analysis of metastasis in various tumors, expressed as log2 TMP + 1, was performed using the Kaplan-Meier (KM) plotter. Differential expression was considered statistically significant, P < 0.05, P < 0.01, and P < 0.001.

Table 1

Abbreviations of tumor names in TCGA data
Abbreviation	Full name
ACC	Adrenocortical carcinoma
BLCA	Bladder urothelial carcinoma
BRCA	Breast invasive carcinoma
CESC	Cervical squamous cell carcinoma and endocervical adenocarcinoma
CHOL	Cholangiocarcinoma
COAD	Colon adenocarcinoma
DLBC	Lymphoid neoplasm diffuse large B-cell lymphoma
ESCA	Esophageal carcinoma
GBM	Glioblastoma multiforme
HNSC	Head and neck squamous cell carcinoma
KICH	Kidney chromophobe
KIRC	Kidney renal clear cell carcinoma
KIRP	Kidney renal papillary cel carcinoma
LAML	Acute mveloid leukemia
LGG	Brain lower grade glioma
UHC	Liver hepatocellular carcinoma
LUAD	Lung adenocarcinoma
LUSC	Lung squamous cell carcinoma
MESO	Mesothelioma
OV	Ovarian serous cystadenocarcinoma
PAAD	Pancreatic adenocarcinoma
PCPG	Pheochromocytoma and paraganglioma
PRAD	Prostate adenocarcinoma
READ	Rectum adenocarcinoma
SARC	Sarcoma
SKCM	Skin cutaneous melanoma
STAD	Stomach adenocarcinoma
TGCT	Testicular germ cell tumors
THCA	Thyroid carcinoma
THYM	Thymoma
UCEC	Uterine corpus endometrial carcinoma
UCS	Uterine carcinosarcoma
UVM	Uveal melanoma

Survival prognosis analysis

To investigate the time-dependent prognostic value of G6PD in 33 tumors, we performed a univariate Cox regression analysis using the "survival package" in the R software. Survival outcomes include overall survival (OS), disease-free survival (DFS), disease-specific survival (DSS), and progression-free survival (PFS). When the hazard ratio exceeded 1 (HR > 1), this indicated that the exposure factor (G6PD expression) was a contributor to the positive event (death). P < 0.05 was considered significant.

Tumor immune and immunosuppressive cell infiltration

We used TIMER2.0 [50] to analyze correlations between G6PD expression and the infiltration of six immune cells, including B cells, CD8⁺ T cells, CD4⁺ T cells, macrophages, neutrophils, and DCs. In addition, we also analyzed correlations between the expression of G6PD in TCGA tumors and the tumor invasion of four immunosuppressive cell types, including MDSCs, CAFs, tM2-TAMs, and regulatory T (Treg) cells. The TIDE, CIBERSORT, CIBERSORT-ABS, QUANTISEQ, and XCELL algorithms were applied to the calculation of immune infiltration. The purity-corrected part of Spearman's rho value and statistical significance (P < 0.05) were used for correlation analysis. Using the TIDE database (http://tide.dfci.harvard.edu/), G6PD was compared with other biomarkers to predict the efficacy of ICB therapy. The heatmap shows the number of immune cell infiltration in 33 TCGA tumors.

Epigenetic methylation analysis

We used the TCGA methylation module [51] of the UALCAN (http://ualcan.path.uab.edu/index.html) interactive web resource to analyze differential methylation levels of G6PD in normal vs. carcinoma tissues in TCGA tumor types.

Genetic alteration analysis

We used the cBioportal website (www.cbioportal.org/) to obtain the G6PD alteration frequencies of various tumor types, the frequency of co-mutation with other genes, and the potential correlation between G6PD mutation status and OS, DSS, DFS, and PFS. In addition, we also obtained the 3D structure of the G6PD protein and the specific mutations in each domain.

G6PD-related gene enrichment analysis

We used the STRING website (https://string-db.org/) [52] to obtain 50 genes related to G6PD expression and made a protein interaction network diagram. Then, the "Similar Gene Detection" module in GEPIA2 was used to obtain the top 100 G6PD-related targeted genes based on the data set of all TCGA tumor tissues. Next, we also applied the "Correlation Analysis" module of GEPIA2 to perform a paired gene Pearson correlation analysis on G6PD and the selected first five genes. The dot plot based on log2 TPM shows the P-value and the correlation coefficient (R). At the same time, we use the "Gene_Corr" module in the TIMER2 database to display the heatmap data of these five genes, including the partial correlation (cor) and P-value in the Spearman rank correlation test with purity adjustment. In addition, we used the 100 genes related to G6PD to use the "ggplot2" package and "clusterProfiler" package in R (version 4.1.1). KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis and GO (Gene Ontology) enrichment analysis. Two-tailed t-test was considered statistically significant.

Analyze immunotherapy response

To analyze the effect of G6PD on therapeutic responses, the ROC plotter server (http://www.rocplot.org/) was used to analyze associations of G6PD expression levels with immunotherapy therapeutic responses in breast cancer, glioblastoma, colorectal, and ovarian tumor patients [23]. In addition, we used the TIDE (tide.dfci.harvard.edu) website to obtain the correlation between the expression of G6PD in tumors and ICB therapy (PD-1 or PD-L1), as well as the correlation with cytotoxic T cell levels (CTL) invasion. Also, we looked at the response to immunotherapy for G6PD in three different datasets (GSE78220, GSE67501, and IMvigor210).

Patient tissue samples and cell culture

Human lung adenocarcinoma, colorectal carcinoma, and ovarian carcinoma tissues and corresponding normal tissues were obtained from patients admitted to Shanghai Tongji Hospital between January 2021 and December 2022. All participants signed a written informed consent form, and the study was approved by the Ethics Committee of Shanghai Tongji Hospital (2021-KYSB-061). Lung adenocarcinoma cell (A549) and human lung epithelial cell (Beas-2B), colon cancer cell (HCT-116) and normal colon epithelial cell (FHC), ovarian cancer cell (A2780) and normal ovarian surface epithelial cell (IOSE80), bladder cancer cell(BIU-87) and normal bladder epithelial cell (SV-HUC-1) were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai). All cells contained 10% fetal bovine serum (FBS) (Gibco, USA) and were in the presence of 5% CO₂.

Western blot

Cells and tissues were lysed in a Radio Immunoprecipitation Assay (RIPA, Thermo Fisher, USA). The protein was quantified by bicinchoninic acid (BCA) analysis (Beyotime, China). Total protein was separated by SDS-PAGE gels and transferred to polyvinylidene fluoride (PVDF) (Millipore, USA). The membranes were incubated with the primary antibodies against G6PD (1:1000, ab210702, Abcam) and actin (1:1000, ab8226, Abcam) overnight at 4^°C. Then the prepared membranes were incubated with a secondary antibody (1:2000, Santa Cruz Biotechnology) for 1 h. Finally, the blots were visualized with Immobilon™ Western Chemiluminescent HRP Substrate (Millipore) and related data was analyzed by Image Lab Software.

G6PD localization, single-cell variations, and expression in cells

First, we revealed the G6PD protein topology and intracellular membrane localization as shown in Fig. 2A. Then, we assessed the distribution of G6PD within the endoplasmic reticulum (ER), microtubules, and nucleus of A-431, U-2 osteosarcoma (U-2OS), and U-251 MG cells using an indirect immunofluorescence assay to characterize the intracellular localization of G6PD. G6PD exhibited overlap with the ER and microtubules, suggesting the subcellular localization of G6PD in the cytoplasm rather than the nucleus (Fig. 2B). This was consistent with G6PD being a marker enzyme of the endoplasmic reticulum. In addition, we also observed the localization of G6PD in tissues, such as liver, kidney, colon, etc. (Fig. 2C). We also found variations in protein and transcript expression do not correlate with the cell cycle by analyzing the G6PD RNA and protein in the U-2 OS FUCCI cell line following cell cycle changes, (Fig. 2D). In addition, the Protein-Protein Interaction Network and the gene and disease network interaction analysis revealed that G6PD was associated with anemia, glucose phosphate dehydrogenase deficiency, low-grade glioma, and breast adenocarcinoma (Fig. 2E, F).

G6PD expression in pan-tumor

We analyzed G6PD expression levels in the TCGA pan-tumor database to explore its oncogenic role. The results showed that the expression levels of G6PD were higher in tumor tissues of BLCA, BRCA, CESC, CHOL, COAD, ESCA, HNSC, KICH, KIRP, LIHC, LUAD, LUSC, READ, STAD, and UCEC than their corresponding adjacent normal tissues according to the TIMER database (Fig. 3A). We checked the expression levels of G6PD in tissues and cells of lung adenocarcinoma, colorectal cancer, ovarian cancer and bladder cancer by WB, and the results were consistent with the database analysis (Fig. 3B).The results of the Clinical Proteomic Tumor Analysis Consortium (CPTAC) dataset showed that the G6PD protein expression was significantly higher in tumor tissues of colon cancer and clear cell RCC versus in adjacent normal tissues, but decreased in LUAD, OV, and BRCA (Fig. 3C). In addition, we observed a correlation between G6PD expression and the pathological stages of tumors, including KIRC, KIRP, LIHC, OV, PAAD, and THCA (Fig. 3D, all P < 0.05). From the tumor metastasis data, we drew a trend that the expression of G6PD was positively correlated with the degree of tumor metastasis. (Fig. 3E). We also investigated the correlation between the expression level of G6PD and prognosis of patients with different tumors. According to the forest plots, high expressed G6PD was linked to poor prognosis of OS for tumors of BLCA (P = 0.005), BRCA (P = 0.03), HNSC (P = 0.016), KIRC (P < 0.001), KIRP (P = 0.001), LAML (P < 0.001), LGG (P < 0.001), LIHC (P < 0.001) and MESO (P = 0.017) (Fig. 4A). However, low expression of the G6PD was related to the poor prognosis of PCPG OS (P = 0.028). Regarding G6PD and DFS, a significant positive association was found in LIHC (P < 0.001), MESO (P = 0.034), and PRAD (P < 0.001), while a negative association was observed in OV (P = 0.001) (Fig. 4B). In terms of DSS, G6PD expression had a protective effect in PCPG (P = 0.012), whereas, it seemed to be a risk factor in BLCA (P < 0.001), BRCA (P = 0.006), HNSC (P = 0.039), KIRC (P < 0.001), KIRP (P < 0.001), LGG (P < 0.001), LIHC (P < 0.001), MESO (P = 0.05), PRAD (P = 0.008), and THCA (P = 0.023) (Fig. 4C). Moreover, the PFS forest plot confirmed the protective role of G6PD expression in OV (P = 0.015) and PCPG (P = 0.044). However, it was found to be a risk factor in BLCA (P = 0.014), GBM (P = 0.044), KIRC (P < 0.001), KIRP (P < 0.001), LGG (P < 0.001), LIHC (P < 0.001), MESO (P = 0.03) and PRAD (P = 0.02) (Fig. 4D). Generally, high expression of G6PD is protective in a few tumors (OV and PCPG), but is a risk factor in most tumors, such as BLCA, KIRC, and KIRP.

The correlation between G6PD expression levels and immune checkpoint blockade therapy

In order to investigate the correlation between G6PD expression levels and immune checkpoint blocking therapy, we first evaluated tumor immune infiltration in TCGA tumor types and subtypes. We found a significant positive correlation between the expression of G6PD in four tumor species (LGG, LIHC, PAAD, and PRAD) and the infiltration of six immune cell types, including B cells, CD8⁺ T cells, CD4⁺ T cells, macrophages, neutrophils, and DCs (all P < 0.05). However, LUSC showed significant negative associations (cor < 0, p < 0.05) between G6PD expression levels and tumor infiltration of the six immune cell types. Immune cells infiltration have also been demonstrated in all other tumor types (Fig. 5A). We also analyzed the correlations of G6PD expression levels in tumors with the infiltration of four immunosuppressive cells via the TIMER2.0 website, including myeloid-derived suppressor cells (MDSCs), cancer-associated fibroblasts (CAFs), the M2 subtype of tumor-associated macrophages (M2-TAMs), and regulatory T (Treg) cells. We observed a positive correlation between G6PD expression and tumor infiltration in MSDCs in BLCA, BRCA, BRCA-LumA, CHOL, ESCA, KIRP, LIHC, LUAD, LUSC, MESO, OV, SKCM-Primary, STAD, and UCEC, but a negative correlation in DLBC, PCPG, and UVM. The tumor infiltration of the other three immunosuppressive cells was also shown in Fig. 5B. A portion of the scatterplot data ( cor >0.4 and P < 0.05) of the above tumors were produced using one algorithm presented in Fig. 5C. For example, the G6PD expression level in LIHC was positively correlated with the infiltration level of MDSC (cor = 0.567, P = 9.73e-31) based on the TIDE algorithm. We compared the predictive ability of G6PD with other standardized biomarkers (TIDE, MSI. score, TMB, CD274, CD8, IFNG, T. Clonality, B. Clonality, and Merck18) for the efficacy of ICB. Interestingly, we found that the AUC for G6PD in 15 ICB subgroups (25 in total) was more significant than 0.5. In two ICB sub-cohorts (Zhao2019-PD1-Glioblastoma-Post Pos = 6, Neg = 3, and Chen2016-CTLA4-Melanoma-Nanostring Pos = 5, Neg = 11), the AUC could reach 0.89 and 0.87, respectively. (Fig. 5D).

Methylation analysis

The promoter methylation status analysis revealed that G6PD was hypermethylated in tumor tissues of PRAD (P = 1.81×10^− 9) and UCEC (P = 2.96×10^− 4). However, the methylation levels in tumor tissues were lower in BLCA (P = 0.028), LIHC (P = 0.018), KIRP (P = 2.63×10^− 3), BRCA (P = 1.62×10^− 12), KIRC (P = 0.042), and TGCT (P = 1.62×10^− 12) than in control paracancerous tissues (Fig. 6). Only the data with differences was shown here.

G6PD genetic alteration analysis

We queried the genetic alteration status of G6PD across the various tumor types in the TCGA cohorts, and the three-dimensional (3D) structure of the G6PD mutation protein was shown in Fig. 7A. The result showed that gene amplification was the most common genetic alteration in G6PD, followed by mutation. The tumor with the highest frequency of "amplification" and "deep deletion" was BLCA, and the tumor with the highest frequency of "mutation" was UCEC (Fig. 7B). The types, sites, and number of cases of G6PD gene alteration were shown in Fig. 7C. Missense mutations in G6PD were the main types of genetic alterations, and mutations at the R192C/S locus were found in THCA, UCEC, and SKCM. In addition, we explored the potential association of G6PD alterations with OS, DFS, DSS, and PFS in the cohort. The data in Fig. 7 (D-G) showed that UCEC cases with G6PD alterations had a better prognosis in terms of PFS (P = 0.0257) than subjects without G6PD alterations. Still, there was no significant difference in DSS (P = 0.0731), OS (P = 0.0861), and DFS (P = 0.338). The frequency of genes (PLXNA3, F8, FLNA, SLC10A3, IKBKG, FAM3A, GDI1, GAB3, MPP1 and TKTL1) that co-alternated with G6PD were shown in Fig. 7H.

G6PD-related gene enrichment analysis

To investigate the molecular mechanisms of G6PD protein in tumorigenesis, we screened targeted G6PD binding proteins and the G6PD expression related genes for GO and KEGG pathway enrichment analyses. We obtained 50 G6PD-binding proteins and demonstrated their interaction network by the STRING tool (Fig. 8A). Subsequently, we used the GEPIA2 tool to combine all tumor expression data from TCGA to obtain the top 100 genes associated with G6PD expression. As shown in Fig. 8B, the expression level of G6PD was positively correlated with PGD (R = 0.7), GCLM (R = 0.67), SRXN1 (R = 0.66), TRIM16L (R = 0.61), and TXNRD1 (R = 0.61). Genes were associated (all P < 0.001). There was a positive correlation between SND1 and the above five genes in most tumors (Fig. 8C). Meanwhile, we performed KEGG and GO enrichment analyses on these 100 genes. The resuts showed that the main enrichment pathways were in the ribosome, mRNA 3'UTR AU-rich region binding, etc (Fig. 8D).

Immunotherapeutic Response

We analyzed associations between G6PD expression level and clinical immunotherapy. We found that the expression level of G6PD was higher in the non-responsive group in colorectal and breast tumors, indicating that patients received immunotherapy less effectively. In contrast, glioblastoma and ovarian cancer patients with higher G6PD expression responded better to immunotherapy (Fig. 9A). Furthermore, we found that higher levels of G6PD expression in melanoma, glioblastoma and renal cancer favored ICB therapy with more prolonged survival than cohorts with lower G6PD expression levels, such as programmed death-receptor 1 (PD-1). Conversely, higher G6PD expression was unfavorable for ICB clinical treatment in bladder cancer patients with shorter survival, such as programmed death-ligand 1 (PD-L1) (Fig. 9B, upper panel). In the renal cancer cohort, G6PD expression levels were negatively correlated with CTL levels, suggesting an interaction with T-cell rejection (Fig. 9B, lower panel). In Fig. 9C, there was no significant difference in G6PD expression between the responding and non-responding groups in all three independent cohorts. However, a trend observed was that patients in the higher G6PD expression group responded better to immunotherapy.

In this study, pan-tumor analysis of G6PD expression level was performed, which correlated with G6PD tumor stage, metastasis, prognosis, molecular characterization of gene expression, gene alteration, DNA methylation, immune cell infiltration, immune evasion, and drug sensitivity of TCGA tumor types and subtypes. We also found that mRNA levels of G6PD in tumor tissues of BLCA, BRCA, CESC, CHOL, COAD, ESCA, HNSC, KICH, KIRP, LIHC, LUAD, LUSC, READ, STAD, and UCEC were higher than their corresponding adjacent normal tissues. Furthermore, we integrated the network of functional gene partners of G6PD, including PAK4 [20], CAT [21], and HSPB1 [22]. However, genetic heterogeneity can occur in different types and subtypes of tumors. In protein analysis, the expression levels of G6PD was significantly increased in tumor tissues versus adjacent normal tissues in colon cancer and clear cell RCC. Still, it was decreased in LUAD, OV, and BRCA. Interestingly, we found that the protein and mRNA of G6PD were expressed at opposite levels in BRCA and LUAD. In particular, in breast cancer, the expression levels of G6PD protein was lower in the cancer compared to normal samples, while the expression levels of G6PD mRNA was higher in the tumor. According to the tumor metastasis and pathological staging data we found that the expression levels of G6PD was positively correlated with the degree and stage of tumor metastasis. We also investigated the correlation between the expression levels of G6PD and the prognosis of patients with different tumors. From the OS, DFS, DSS, and PFS survival prognostic data, we can conclude that high expression of G6PD was protective in a few tumors (OV and PCPG) and a risk factor in most tumors, such as BLCA, KIRC, and KIRP. These results suggest that G6PD is an oncogenic molecule for tumor progression, metastasis, and poor prognosis. Our findings were consistent with the results of previous clinical trials, in which overexpression of G6PD can significantly reverse the inhibitory effect of miR-1 on tumor cell growth in pituitary tumors [23]. G6PD could be a marker predicting glioma risk, prognosis, and chemo-sensitivity [24]. Similar reports have been found in kidney, breast, and colorectal cancers [25–27]. Therefore, G6PD can be used as a biomarker for early tumor detection and follow-up.

Tumor immune escape is associated with the efficacy and prognosis of tumor immunotherapy. Therefore, a comprehensive analysis of tumor-infiltrating immune cells may reveal the mechanisms of tumor immune escape, and then provide opportunities for developing new therapeutic strategies [28–31]. There were two different immune escape hypotheses here. On the one hand, tumor-infiltrating immune cells suppress T-cell function or produce a dysfunctional T-cell phenotype, which promotes tumor escape from the host immune system [32, 33]. For example, macrophages can mediate T cell activation by producing IL-12 and expressing co-stimulatory molecules such as CD86, while suppressing T cell function by expressing T cell suppressor molecules and secreting immunosuppressive cytokines[34, 35]. Our results showed that G6PD is positively correlation with tumor immune cell infiltration in LGG, LIHC, PAAD, and PRAD. It is suggested that G6PD may enhance tumor immune escape by enhancing immune cell infiltration and thus inhibiting T cell function. On the other hand, previous studies have shown that several stromal cell types, such as CAFs, Tregs, M2-TAMs, and MDSCs, prevent T cells from reaching the tumor nest, ultimately limiting the efficacy of ICB [36, 37]. Recently, some essential studies have shown that G6PD expression levels in many malignancies are linked to immunological and chemotherapeutic outcomes, such as G6PD-TGF-β signaling [38], G6PD-macrophage polarization[39], and G6PD-cell adhesion molecules expression [40, 41]. We found that the expression levels of G6PD were proportional to the number of the four immunosuppressive cells in most tumors, which suggests that G6PD can promote tumorigenesis by promoting immunosuppressive cell infiltration. The AUC values of G6PD in the two ICB subgroups could reach 0.89 and 0.87, respectively, which indicates that G6PD has a good clinical predictive value for immunotherapy.

The protein interaction network analysis showed that G6PD protein was highly correlated with oxidative stress-related proteins TXNRD1 and PGD, which was consistent with previously reported literature [42, 43].

DNA methylation can inhibit gene expression by altering chromatin structure, DNA stability, and DNA conformation[44]. A relationship between DNA methylation and tumors has been reported in recent years. Hypermethylation of promoter regions often leads to silencing or inactivating tumor suppressor genes [44, 45]. Our results showed that G6PD methylation levels were lower in BLCA, LIHC, KIRP, BRCA, KIRC, and TGCT tumor tissues than in control paraneoplastic tissues, which were inversely correlated with the gene expression levels.

It was reported that genetic alterations are closely related to the occurrence and development of various human diseases[46]. In this study, we identified gene amplification as the most common genetic alteration in G6PD, followed by mutation. Our data showed that G6PD alterations were associated with a better prognosis in terms of PFS than subjects without G6PD alterations. However, these results do not suggest that alterations in the G6PD gene contribute to tumor prognosis, as tumor progression is a complex multifactorial process, and such results may be due to a combination of reasons. Missense mutations in G6PD were the main types of genetic alterations, and mutations at the R192C/S locus were found in THCA, UCEC, and SKCM. In addition, we found that there was very significantly correlations between the G6PD gene expression levels and the expression levels of PLXNA3, F8, FLNA, SLC10A3, IKBKG, FAM3A, GDI1, GAB3, MPP1, and TKTL1 in tumors. Subsequently, we obtained the top 100 genes associated with G6PD expression and performed KEGG and GO enrichment analyses. The results showed that the main enrichment pathways were in the ribosome and mRNA 3'UTR AU-rich region binding, etc.

Immunotherapy targeting blockade of PD-1/PD-L1 can induce robust and durable responses in patients with a variety of tumors[36]. We found that higher levels of G6PD expression in melanoma, glioblastoma, and renal cancer favored ICB (PD-1) therapy with more prolonged survival than cohorts with lower G6PD expression levels. Conversely, higher G6PD expression was unfavorable for ICB (PD-L1) clinical treatment in bladder cancer patients with shorter survival. We also analyzed the immunotherapeutic response to G6PD in the GSE78220, GSE67501, and IMvigor210 datasets. There were no significant differences in G6PD expression between the responder and non-responder groups in all three independent cohorts. However, a trend observed was that patients in the higher G6PD expression group had a better response to immunotherapy. This indicated that patients with higher G6PD expression levels were more effective for immunotherapy treatment, which provides a new idea for immunotherapy of tumors.

In conclusion, the present study investigated the role of the G6PD expression profile in pan-tumor comprehensively. We studied G6PD gene expression, survival prognosis analysis, tumor immune and immunosuppressive cell infiltration, G6PD-related gene enrichment analysis, DNA methylation, genetic alteration analysis, and immunotherapy response. We also investigated the potential molecular mechanism of G6PD in the pathogenesis or clinical prognosis of different tumors. Our study indicates that G6PD expression is higher in tumor tissues compared to normal tissues and is associated with tumor stage, metastasis, and prognosis across various cancers and subtypes in the Cancer Genome Atlas. High G6PD expression offers protective effects in a few cancers such as paraganglioma, pheochromocytoma, and ovarian serous cystadenocarcinoma. However, it poses a risk factor in the majority of cancers. Patients with G6PD alterations show better progression-free survival outcomes than those without. We found a significant positive correlation between G6PD and immune cell infiltration in prostate cancer, pancreatic adenocarcinoma, liver cancer, and low-grade glioma. Moreover, G6PD methylation levels inversely correlate with mRNA expression. Genes like PGD, GCLM, SRXN1, TRIM16L, and TXNRD1 demonstrate a significant positive correlation with G6PD expression. Missense mutations in G6PD, particularly at the R192C/S locus, are common in cutaneous melanoma, uterine cancer, and thyroid carcinoma. G6PD expression is also linked to immunological and chemotherapeutic outcomes in several malignancies.Our pan-tumor analysis provides a deep understanding of the functions of G6PD in oncogenesis, detection, prognosis, and therapeutic design in different tumors. According to the study, patients who expressed more G6PD generally had better therapeutic outcomes. Our study highlights the role of G6PD in oncogenesis, detection, prognosis, and treatment planning.

Although we performed a pan-tumor analysis of many aspects of G6PD using multiple databases, there is still a room for improvement in our study. First, our pan-tumor analysis involved too many aspects and therefore was not explored in sufficient depth. Second, the experiments performed in this study were limited, and we should perform more experiments in vivo or in vitro. Third, although the analysis in this study showed that G6PD expression was associated with clinical immunotherapy in tumors, we cannot yet infer through what specific mechanism by which G6PD affects tumor immunotherapy, so more in-depth studies are needed.

G6PD Glucose-6-phosphate dehydrogenase

PPP Pentose phosphate pathway

TCGA The Cancer Genome Atlas

AUC Area under curve

TME Tumor microenvironment

U-2 OS U-2 osteosarcoma

PPI Protein-protein interaction

PD-1 Programmed death-1

PD-L1 Programmed death-ligand 1

OS Overall survival

DFS Disease-free survival

DSS Disease-specific survival

PFS Progression-free survival

DCs Dendritic cells

CAFs Cancer-associated fibroblasts

MDSCs Myeloid-derived suppressor cells

M2-TAMs M2 subtype of tumor-associated macrophages

ICB Immune checkpoint blockade

KEGG Kyoto Encyclopedia of Genes;

GO Gene Ontology.

Ethics approval and consent to participate

All participants provided written informed consent, and the study was approved by the Ethics Committee of Shanghai Tongji Hospital (Approval No. 2021-KYSB-061).

Consent for publication

Not applicable.

Availability of data and materials

The G6PD protein topology was obtained using PROTTER (https://wlab.ethz.ch/protter/start/), while location and expression data were sourced from the Human Protein Atlas (https://www.proteinatlas.org/) and GeneCards (https://www.genecards.org/). Protein-disease associations were accessed via GPS-Prot (http://gpsprot.org/index.php) and OpenTargets (https://platform.opentargets.org/).

Gene and protein expression data, including tumor stages, were analyzed using TIMER2.0 (http://timer.cistrome.org/), UALCAN (http://ualcan.path.uab.edu/analysis-prot.html), and GEPIA2 (http://gepia.cancer-pku.cn/). Immune cell infiltration and ICB therapy predictions were evaluated using TIMER2.0, TIDE (http://tide.dfci.harvard.edu/), and other algorithms. Epigenetic methylation data were analyzed via UALCAN, and genetic alterations from cBioPortal (http://www.cbioportal.org/). Protein interaction networks and G6PD-related gene analyses were performed using STRING (https://string-db.org/), GEPIA2, and TIMER2.0, with gene enrichment conducted in R. Immunotherapy response was assessed via the ROC plotter (http://www.rocplot.org/) and TIDE, using datasets GSE78220, GSE67501, and IMvigor210.

Competing interests

The authors declare no competing interests.

Funding

This work was supported by the National Natural Science Foundation of China (82272630, 82203664, 82304365); the Postdoctoral Science Foundation of China (2023M732653); Shanghai Pujiang Programme (23PJD083); Shanghai Public Health System Construction Three-Year Action Plan (2022-2024) Key Disciplines (SHDC2022CRD040); Shanghai Hospital Development Center Foundation (SHDC22023303); Clinical Research Project of Tongji Hospital of Tongji University [ITJ(ZD)2309]

Authors' contributions

The work present here was carried out in collaboration among all authors.

DL conceived and jointly designed the study with ZS. YG and XW conducted this study and drafted this manuscript. JL, YL and RS analyzed the data. YG and XW performed experiments. ZS revised the manuscript. All authors read and approved the fnal manuscript.

Acknowledgements

Not applicable.

Dawson MA, Kouzarides T. Cancer epigenetics: from mechanism to therapy. Cell. 2012;150(1):12-27.
Coppedè F, Lopomo A, Spisni R, et al. Genetic and epigenetic biomarkers for diagnosis, prognosis and treatment of colorectal cancer. World J Gastroenterol. 2014;20(4):943-56.
Wu J, Chen ZP, Shang AQ, et al. Systemic bioinformatics analysis of recurrent aphthous stomatitis gene expression profiles. Oncotarget. 2017;8(67):111064-72.
Nadeu F, Diaz-Navarro A, Delgado J, et al. Genomic and Epigenomic Alterations in Chronic Lymphocytic Leukemia. Annu Rev Pathol. 2020;15:149-77.
Zhou S, Treloar AE, Lupien M. Emergence of the Noncoding Cancer Genome: A Target of Genetic and Epigenetic Alterations. Cancer Discov. 2016;6(11):1215-29.
Wu J, Lu WY, Cui LL. Clinical significance of STAT3 and MAPK phosphorylation, and the protein expression of cyclin D1 in skin squamous cell carcinoma tissues. Mol Med Rep. 2015;12(6):8129-34.
Lyssiotis CA, Kimmelman AC. Metabolic Interactions in the Tumor Microenvironment. Trends Cell Biol. 2017;27(11):863-75.
Bocchialini G, Lagrasta C, Madeddu D, et al. Spatial architecture of tumour-infiltrating lymphocytes as a prognostic parameter in resected non-small-cell lung cancer. Eur J Cardiothorac Surg. 2020;58(3):619-28.
Schlotter CM, Tietze L, Vogt U, et al. Ki67 and lymphocytes in the pretherapeutic core biopsy of primary invasive breast cancer: positive markers of therapy response prediction and superior survival. Horm Mol Biol Clin Investig. 2017;32(2).
Ribas A, Wolchok JD. Cancer immunotherapy using checkpoint blockade. Science. 2018;359(6382):1350-5.
Chen L, Zhang Z, Hoshino A, et al. NADPH production by the oxidative pentose-phosphate pathway supports folate metabolism. Nat Metab. 2019;1:404-15.
Yang HC, Wu YH, Liu HY, et al. What has passed is prolog: new cellular and physiological roles of G6PD. Free Radic Res. 2016;50(10):1047-64.
Leopold JA, Walker J, Scribner AW, et al. Glucose-6-phosphate dehydrogenase modulates vascular endothelial growth factor-mediated angiogenesis. J Biol Chem. 2003;278(34):32100-6.
Chen EI, Hewel J, Krueger JS, et al. Adaptation of energy metabolism in breast cancer brain metastases. Cancer Res. 2007;67(4):1472-86.
Zhang R, Tao F, Ruan S, et al. The TGFβ1-FOXM1-HMGA1-TGFβ1 positive feedback loop increases the cisplatin resistance of non-small cell lung cancer by inducing G6PD expression. Am J Transl Res. 2019;11(11):6860-76.
Wang S, Zeng F, Liang S, et al. lncRNA Linc00173 modulates glucosemetabolism and multidrug chemoresistancein SCLC: Potential molecular panel for targeted therapy. Mol Ther. 2021.
Ni Y, Yang Z, Agbana YL, et al. Silent information regulator 2 promotes clear cell renal cell carcinoma progression through deacetylation and small ubiquitin-related modifier 1 modification of glucose 6-phosphate dehydrogenase. Cancer Sci. 2021;112(10):4075-86.
Sacks D, Baxter B, Campbell B, et al. Multisociety Consensus Quality Improvement Revised Consensus Statement for Endovascular Therapy of Acute Ischemic Stroke. Int J Stroke. 2018;13(6):612-32.
Ghergurovich JM, Esposito M, Chen Z, et al. Glucose-6-Phosphate Dehydrogenase Is Not Essential for K-Ras-Driven Tumor Growth or Metastasis. Cancer Res. 2020;80(18):3820-9.
Zhang X, Zhang X, Li Y, et al. PAK4 regulates G6PD activity by p53 degradation involving colon cancer cell growth. Cell Death Dis. 2017;8(5):e2820.
Doğan S, Özcan T, Doğan M, et al. The effects on antioxidant enzymes of PMMA/hydroxyapatite nanocomposites/composites. Enzyme Microb Technol. 2020;142:109676.
Ye H, Huang H, Cao F, et al. HSPB1 Enhances SIRT2-Mediated G6PD Activation and Promotes Glioma Cell Proliferation. PLoS One. 2016;11(10):e0164285.
Fekete JT, Győrffy B. ROCplot.org: Validating predictive biomarkers of chemotherapy/hormonal therapy/anti-HER2 therapy using transcriptomic data of 3,104 breast cancer patients. Int J Cancer. 2019;145(11):3140-51.
Jones JG. Hepatic glucose and lipid metabolism. Diabetologia. 2016;59(6):1098-103.
Spencer NY, Stanton RC. Glucose 6-phosphate dehydrogenase and the kidney. Curr Opin Nephrol Hypertens. 2017;26(1):43-9.
Benito A, Polat IH, Noé V, et al. Glucose-6-phosphate dehydrogenase and transketolase modulate breast cancer cell metabolic reprogramming and correlate with poor patient outcome. Oncotarget. 2017;8(63):106693-706.
Dore MP, Davoli A, Longo N, et al. Glucose-6-phosphate dehydrogenase deficiency and risk of colorectal cancer in Northern Sardinia: A retrospective observational study. Medicine (Baltimore). 2016;95(44):e5254.
Lawal B, Lin LC, Lee JC, et al. Multi-Omics Data Analysis of Gene Expressions and Alterations, Cancer-Associated Fibroblast and Immune Infiltrations, Reveals the Onco-Immune Prognostic Relevance of STAT3/CDK2/4/6 in Human Malignancies. Cancers (Basel). 2021;13(5).
Man YG, Stojadinovic A, Mason J, et al. Tumor-infiltrating immune cells promoting tumor invasion and metastasis: existing theories. J Cancer. 2013;4(1):84-95.
Zhang Y, Zhang Z. The history and advances in cancer immunotherapy: understanding the characteristics of tumor-infiltrating immune cells and their therapeutic implications. Cell Mol Immunol. 2020;17(8):807-21.
Wang L, Wu J, Song S, et al. Plasma Exosome-Derived Sentrin SUMO-Specific Protease 1: A Prognostic Biomarker in Patients With Osteosarcoma. Front Oncol. 2021;11:625109.
Taube JM, Klein A, Brahmer JR, et al. Association of PD-1, PD-1 ligands, and other features of the tumor immune microenvironment with response to anti-PD-1 therapy. Clin Cancer Res. 2014;20(19):5064-74.
Yu GP, Chiang D, Song SJ, et al. Regulatory T cell dysfunction in subjects with common variable immunodeficiency complicated by autoimmune disease. Clin Immunol. 2009;131(2):240-53.
Weyand CM, Watanabe R, Zhang H, et al. Cytokines, growth factors and proteases in medium and large vessel vasculitis. Clin Immunol. 2019;206:33-41.
Albini A, Gallazzi M, Palano MT, et al. TIMP1 and TIMP2 Downregulate TGFβ Induced Decidual-like Phenotype in Natural Killer Cells. Cancers (Basel). 2021;13(19).
Mariathasan S, Turley SJ, Nickles D, et al. TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature. 2018;554(7693):544-8.
Peranzoni E, Lemoine J, Vimeux L, et al. Macrophages impede CD8 T cells from reaching tumor cells and limit the efficacy of anti-PD-1 treatment. Proc Natl Acad Sci U S A. 2018;115(17):E4041-4041E4050.
Parsanathan R, Jain SK. Glucose-6-Phosphate Dehydrogenase Deficiency Activates Endothelial Cell and Leukocyte Adhesion Mediated via the TGFβ/NADPH Oxidases/ROS Signaling Pathway. Int J Mol Sci. 2020;21(20).
Parsanathan R, Jain SK. G6PD deficiency shifts polarization of monocytes/macrophages towards a proinflammatory and profibrotic phenotype. Cell Mol Immunol. 2021;18(3):770-2.
Parsanathan R, Jain SK. Glucose-6-phosphate dehydrogenase deficiency increases cell adhesion molecules and activates human monocyte-endothelial cell adhesion: Protective role of l-cysteine. Arch Biochem Biophys. 2019;663:11-21.
Parsanathan R, Jain SK. L-Cysteine in vitro can restore cellular glutathione and inhibits the expression of cell adhesion molecules in G6PD-deficient monocytes. Amino Acids. 2018;50(7):909-21.
Killoy KM, Pehar M, Harlan BA, et al. Altered expression of clock and clock-controlled genes in a hSOD1-linked amyotrophic lateral sclerosis mouse model. FASEB J. 2021;35(2):e21343.
Hagiwara M, Fushimi A, Yamashita N, et al. MUC1-C activates the PBAF chromatin remodeling complex in integrating redox balance with progression of human prostate cancer stem cells. Oncogene. 2021;40(30):4930-40.
Wang M, Ngo V, Wang W. Deciphering the genetic code of DNA methylation. Brief Bioinform. 2021;22(5).
Mehdi A, Rabbani SA. Role of Methylation in Pro- and Anti-Cancer Immunity. Cancers (Basel). 2021;13(3).
Lin X. Genomic Variation Prediction: A Summary From Different Views. Front Cell Dev Biol. 2021;9:795883.
Wu SY, Lin KC, Lawal B, et al. MXD3 as an onco-immunological biomarker encompassing the tumor microenvironment, disease staging, prognoses, and therapeutic responses in multiple cancer types. Comput Struct Biotechnol J. 2021;19:4970-83.
Fahey ME, Bennett MJ, Mahon C, et al. GPS-Prot: a web-based visualization platform for integrating host-pathogen interaction data. BMC Bioinformatics. 2011;12:298.
Chen F, Chandrashekar DS, Varambally S, et al. Pan-cancer molecular subtypes revealed by mass-spectrometry-based proteomic characterization of more than 500 human cancers. Nat Commun. 2019;10(1):5679.
Li T, Fu J, Zeng Z, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48(W1):W509-509W514.
Shinawi T, Hill VK, Krex D, et al. DNA methylation profiles of long- and short-term glioblastoma survivors. Epigenetics. 2013;8(2):149-56.
Cui X, Zhang X, Liu M, et al. A pan-cancer analysis of the oncogenic role of staphylococcal nuclease domain-containing protein 1 (SND1) in human tumors. Genomics. 2020;112(6):3958-67.

No competing interests reported.

Glucose-6-phosphate Dehydrogenase (G6PD): the Role in Tumor Progression and Immunotherapy

Status:

Version 1

Abstract

Figures

Background

Materials and methods

G6PD protein structure, location and relevance to disease

Gene expression analysis

Survival prognosis analysis

Tumor immune and immunosuppressive cell infiltration

Epigenetic methylation analysis

Genetic alteration analysis

G6PD-related gene enrichment analysis

Analyze immunotherapy response

Patient tissue samples and cell culture

Western blot

RESULTS

G6PD localization, single-cell variations, and expression in cells

G6PD expression in pan-tumor

The correlation between G6PD expression levels and immune checkpoint blockade therapy

Methylation analysis

G6PD genetic alteration analysis

G6PD-related gene enrichment analysis

Immunotherapeutic Response

Discussion

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1