Metabolomic profiling of oral squamous cell carcinoma reveals glutaminolysis-related proteins: GLS1 and GLUD1 are potential diagnostic and prognostic factors

It is essential to explore potential molecular involved in oral squamous cell carcinoma (OSCC) malignant transformation and progression. The purpose of this study was to delineate the metabolic characteristics of OSCC patients. We firstly report GC-MS based metabolic profiling of OSCC tissues from 8 patients, including tumor tissues and its matched normal tissues, then another 81 OSCC patients with prognostic information were used to verify metabolomics results by exam key enzymes expression. A panel including ten metabolite biomarker were identified, which can distinguish OSCC tumor from normal controls. On the base of metabolites with significantly difference, we applied KEGG for further pathway analysis and found OSCC have highly active glutaminolysis metabolism. Therefore, glutaminolysis-related key enzymes GLS1 and GLUD1 were detected by immunohistochemical and results demonstrated that GLS1 and GLUD1 were highly expressed in OSCC. Moreover, Kaplan-Meier survival analysis also showed and GLS1 and GLUD1 expressions were correlated with poor prognosis and tumor progression in OSCC patients. In conclusion, our results indicated that OSCC have highly active glutaminolysis metabolism, and GLS1 and GLUD1 enzymes are potential diagnostic and prognostic factors for OSCC. Our study has shown that a GC-MS based metabolite analysis was able to identify biomarker metabolites which can significantly differentiate OSCC tissues from normal control tissues. In this study, we found OSCC have highly active glutaminolysis metabolism and key enzymes GLS1 and GLUD1 expression predict poor prognosis. All of these results propose that amino acid metabolism reprogramming may represent new potential for the treatment of OSCC. However, further studies are needed to elucidate the potential role of glutaminolysis in the carcinogenesis of OSCC.


Abstract
It is essential to explore potential molecular involved in oral squamous cell carcinoma (OSCC) malignant transformation and progression. The purpose of this study was to delineate the metabolic characteristics of OSCC patients. We firstly report GC-MS based metabolic profiling of OSCC tissues from 8 patients, including tumor tissues and its matched normal tissues, then another 81 OSCC patients with prognostic information were used to verify metabolomics results by exam key enzymes expression. A panel including ten metabolite biomarker were identified, which can distinguish OSCC tumor from normal controls. On the base of metabolites with significantly difference, we applied KEGG for further pathway analysis and found OSCC have highly active glutaminolysis metabolism.
Therefore, glutaminolysis-related key enzymes GLS1 and GLUD1 were detected by immunohistochemical and results demonstrated that GLS1 and GLUD1 were highly expressed in OSCC. Moreover, Kaplan-Meier survival analysis also showed and GLS1 and GLUD1 expressions were correlated with poor prognosis and tumor progression in OSCC patients. In conclusion, our results indicated that OSCC have highly active glutaminolysis metabolism, and GLS1 and GLUD1 enzymes are potential diagnostic and prognostic factors for OSCC.

Background
Oral squamous cell carcinoma (OSCC) is one of the most common cancers in head and neck region. There are 6 million deaths worldwide every year due to OSCC (1) and tobacco use (smoked or chewed), alcohol consumption and human papillomavirus infection are regarded as the most important risk factors for OSCC (2). The diagnostic methods of OSCC including physical examination, endoscopy, radiography, computed tomography, magnetic resonance imaging, serum and urine analyses, and histopathological examination of tissue biopsies. Although these managements for OSCC have been greatly improved, survival rate also remains poor due to regional and distant metastases (3).
Changes in metabolic pathways have been observed in almost all tumors, leading to dependence on specific nutrients or enzymes. The most striking feature of cancer cells is that they alter their metabolic pathways to meet cancer cell energy need. Changes in cell metabolism can contribute to transformation and tumor progression. Metabolic phenotypes can also be exploited to image tumors, provide diagnostic and prognostic information (4). Therefore, better understanding of the molecular and related key pathways that lead to the progression of OSCC is essential for improving diagnostic and prognostic predictors.
Lactate fermentation and aerobic respiration provide energy for cancer cell progression.
Warburg Effect indicated that cancer cells can obtain energy from lactic acid fermentation even oxygen is enough, namely aerobic glycolysis (5,6). In addition, cancer cells can also obtain energy from other metabolism pathway, for example amino acid and lipid metabolism, pentose phosphate pathway, glutaminolysis, and mitochondrial biogenesis (7)(8)(9). Glutaminolysis, deposing glutamine into glutamate, further alpha ketoglutarate for maintaining tricarboxylic acid cycle, replenish glucose metabolism and provide energy for cells, then leads to a glutamine addiction in some cancer cell. It has been reported that the c-myc can regulate genes which is necessary for glutamine catabolism (5). Thus, cancer cells metabolic reprogramming may be determined by their genetic profile.
Conclusively, it is very important to discover reliable biomarkers for OSCC early diagnosis or predict prognosis (10,11).
Metabolomic studies is a useful tool for identifying cancer therapeutic targets, tumor tissue and biological fluids samples were widely used in metabolomics research. However, metabolomics studies of oral squamous cell carcinoma tissues are very limited up to date, most studies have focused on distinguishing OSCC from normal groups use plasma or saliva samples (12). Our purpose was to explore tissue metabolite biomarkers of OSCC, which not only discriminate tumor from normal control, but also investigate a panel of potential prognostic and therapeutic marker of OSCC.

Tumor Tissue Specimens
Our study was reviewed and approved by the medical ethics committee of Stomatological

Metabolomics statistical analysis
After removed the artificial peaks due to derivatization, raw data files of GC-MS analysis determined by NIST 2014 standard mass spectral databases built-in Xcalibur 2.2 software (Thermo Scientific, Waltham, MA). Sandardized using internal parameters, the peak area of metabolites were calculated with Xcalibur 2.2 software. We use Metaboanalyst1 3.0 perform statistical analyses, which contain the R package of statistical computing software. Total spectral intensity and additionally Pareto scaled were used to normalized initial experimental data.
Univariate analysis was by ANOVA, such as fold change, T test, Volcano Plot. Multivariate analysis was via unsupervised Principal Component Analysis (PCA) followed by Partial Least Squares Discriminant Analysis (PLS-DA). Hierarchical cluster analysis was carried out using Dendrogram and Heatmap. We also applied KEGG for further metabolite pathway analysis. Immunohistochemical results were analyzed by SPSS 17.0 software package. The relationships between GLS1 protein expression and the clinicopathological parameters were determined by Chi-square tests. We estimated survival curves using the Kaplan-Meier method and compared them using a two-sided log-rank test p.

OSCC have Highly Active Glutaminolysis metabolism
Anaplerosis Tricarboxylic Acid Cycle Altered cell metabolism enables tumors to sustain their increased energetic and biosynthetic needs. We found OSCC have highly active glutaminolysis metabolism by metabolic pathway analysis ( Figure 3A). In order to elucidate the important role of energetics alteration in OSCC tissues, additional experiments were performed to validate the role of glutaminolysis in OSCC tumorigenesis. We found the intermediate product of glutaminolysis such as glutamine, glutamate and α-ketoglutarate were differentially elevated in OSCC tissues compared to normal adjacent tissues by metabolomic profiling analysis (Table3), glutamine, glutamate and α-ketoglutarate concentration in tumor tissues were significantly higher than that in normal tissues (Figure3B-D), suggesting glutaminolysis may be involved in the development of OSCC. Then we explored two key enzymes of glutaminolysis-GLS1 and GLUD1 and figured out their higher expression in OSCC tissues than normal tissues in mRNA level (Figure4 A, B). Meanwhile, at protein level, we discovered the same result as mRNA level by examing the expression of GLS1 and GLUD1 protein in 81 OSCC tissues using immunohistochemical ( Figure4C).. All these results demonstrated OSCC have highly active glutamine catabolism, and glutaminolysis was likely to play a remarkable role in OSCC tumorigenesis. But its mechanism of glutaminolysis promoting tumorigenesis was still unveiled. We hypothesized glutamine converted into glutamate and further α-ketoglutarate in order to capturing nutrients for TCA cycle to support cancer cell metabolic needs. Detailed mechanism hypothesis of glutamine supplying for Tricarboxylic Acid Cycle to product ATP in tumor showed in Figure 5.

Glutaminolysis-related proteins GLS1 and GLUD1
were predictor of poor prognosis in OSCC tissues Next, we further illuminated role of glutaminolysis-related proteins on OSCC progression.
We detected the expression of GLS1 and GLUD1 protein in 81 OSCC tissues, the relationship between GLS1 and GLUD1 protein expression and clinicopathological parameters and prognosis were evaluated.GLS1 expression were also related with TNM, LNM and invasive depth of tumor, but not correlated with age, sex, grade, smoking, WPOI, GLUD1 expression only correlated with invasive depth of tumor (Table 4). Our result validated GLS1 and GLUD1 expression correlated with poor prognosis of patients (Figure6).
The result further confirmed that glutaminolysis may promote OSCC tumorigenesis and clinical progress, the specific mechanism needs to be further studied.

Discussion
Cell metabolic pathway was a complex network which consists of key enzyme, regulatory genes and metabolite. Cancer cells undergo metabolic reprogramming, which regulatory networks were altered to adapt to the metabolic pressures and provide energy for cancer cell growth, metabolic reprogramming was one of the hallmarks of cancer (13)(14)(15). Compared to other 'omics' (such as proteome/genome/transcriptome), most metabolites are small molecular compounds, highly conservative and stable performance.
The statistical analysis of metabolomics data is more convenient, the results are easier to understand and more accurate.
The metabolomic profile of OSCC is an untapped resource for a cancer not only with recently increasing incidence but also that would highly benefit from advances for OSCC early diagnosis. Most of the samples for the metabolomics study of OSCC were biofluids (eg. plasma, urine, saliva) and cell lines (16)(17)(18)(19), tissue samples were rarely reported (20), the most important reason is that biofluids are readily available. Most studies are based on biofluids metabolomics to distinguish from cancer patients to healthy control people, we need to use tissue samples to further study the metabolomics of OSCC, which can not only distinguish the cancer tissue from the normal paired tissue, but also study the metabolic pathway leading to carcinogenesis, so as to find therapeutic targets for cancer in the field of metabolomics.
In this pilot study, we used GC-MS to analyze the metabolomics of OSCC tumor tissue and matched normal tissue samples, a total of 244 metabolites were identified, there were 85 known metabolites significant differences between the tumor and normal tissues (Fold change >1.5 or 0.667), 10 metabolites with important features were further distinguished by Volcano Plot. In our study, most amino acids, such as ornithine, L-cysteine, gammaaminobutyric acid, lysine, aspartic acid, tyrosine, serine, alanine, glycine, N-methyl alanine, proline, ornithine significantly increase in OSCC tumor tissues. Our results are consistent with the previous findings in which oral squamous cell carcinoma tissues had higher amino acid levels than normal tissues (21)(22). However, there were reports revealed lower relative concentration of amino acids as compared to healthy groups in some cancers, such as breast, pancreatic, oral, and colorectal cancers) (23)(24)(25)(26). This anomaly suggests that cancer cells build a second metabolic pathway to generate energy for rapid growth, which need more glucogenic amino acids. It can also revealed during the proliferation of cancer cells, the continuous use of amino acids leads to an increase in the concentration of amino acids at the time of biosynthesis, and a decrease in the concentration of amino acids at the end of biosynthesis. Altered cell metabolism enables tumors to sustain their increased energetic and biosynthetic needs.
We also found the intermediate product of glutaminolysis such as glutamine, glutamate and α-ketoglutarate were differentially elevated in OSCC tissues, demonstrated OSCC have highly active glutamine catabolism. We further validated that glutaminolysis-related proteins GLS1 and GLUD1 correlated with recurrence and poor prognosis in OSCC.
Therefore, it can be inferred that glutaminolysis was likely to play an important role in OSCC tumorigenesis, its mechanism of promoting tumorigenesis may be due to glutamine which converted into glutamate and α-ketoglutarate, and then into the tricarboxylic acid cycle produce energy for cancer cell (Fig 5). Regulatory molecular mechanism of GLS1 and GLUD1 promote OSCC tumorigenesis need to be further study. With the deepening of the research on the interaction mechanism between Cell metabolism and cancer development, we believe that GLS1 and GLUD1 may be a new target for cancer treatment.
Overall, the results from this based on GS-MS global metabolomic analysis between tumor tissues and matched normal tissues revealed a number of metabolic readouts, including changes in metabolites related to energetics. In PLS-DA and hierarchical clustering analysis (HCA), cancer tissues were well-separated from normal tissues. Glutaminolysisrelated key enzymes GLS1 and GLUD1 correlated with clinical progress and overall survival in OSCC, suggesting altered metabolites may be useful prognostic biomarkers of OSCC.

Conclusions
Our study has shown that a GC-MS based metabolite analysis was able to identify biomarker metabolites which can significantly differentiate OSCC tissues from normal control tissues. In this study, we found OSCC have highly active glutaminolysis metabolism and key enzymes GLS1 and GLUD1 expression predict poor prognosis. All of these results propose that amino acid metabolism reprogramming may represent new potential for the treatment of OSCC. However, further studies are needed to elucidate the potential role of glutaminolysis in the carcinogenesis of OSCC.

Consent for publication
If the paper is accepted, we agree to be published in BMC cancer journal.

Availability of data and material
All data and material are available online.