MTHFD1L-Mediated Redox Homeostasis Promotes Tumor Progression in Tongue Squamous Cell Carcinoma

Background: Routine changes in cell metabolism can drive tumor development, as the cellular program develops to promote glycolysis and redox homeostasis during tumor progression; however, the associated mechanisms in tongue squamous cell carcinoma (TSCC) remain unclear. Methods: We investigated methylenetetrahydrofolate dehydrogenase 1-like (MTHFD1L) expression, its clinical relevance, redox modification, and molecular mechanisms using TSCC cells and tissues. The anti-tumor effects of MTHFD1L knockdown on TSCC tumorigenesis were evaluated in vitro and in vivo. Kaplan-Meier curves and the log-rank test were used to analyze disease-free survival and overall survival. Results: TSCC patients with high expression levels of MTHFD1L had shorter overall survival (P < 0.05) and disease-free survival (P < 0.05). Knockdown of MTHFD1L reduced nicotinamide adenine dinucleotide phosphate (NADPH) levels and increased reactive oxygen species (ROS), which accelerated cell death under oxidative stress, such as hypoxia or glucose deprivation. Additionally, inhibition of MTHFD1L suppressed TSCC cell growth and delayed the cell cycle, including in xenograft experiments. Conclusions: MTHFD1L confers redox homeostasis and promotes TSCC cell growth, which provides a great opportunity to study tumor metabolism in head and neck cancer. The mTORC1-4EBP1-eIF4E axis may affect the expression of MTHFD1L in TSCC. Inhibition of the expression of MTHFD1L may be an actionable and effective therapeutic target in TSCC.


INTRODUCTION
As a subset of oral cancer, tongue squamous cell carcinoma (TSCC) constitutes more than half of oral SCCs in China and is characterized by its high incidence and poor prognosis (1)(2)(3). Despite advances in medical treatments, there are still many TSCC patients who are ineligible for surgical therapy and who experience recurrence, metastasis and radiochemotherapy resistance, making TSCC very difficult for doctors and miserable for patients. Therefore, the identification of a reliable prognostic biomarker or promising therapeutic target for TSCC is a pressing task. Changes in cell metabolism are known to drive tumor development. Numerous studies have confirmed that metabolic reprogramming occurs in tumorigenesis and in development (4,5). Our recent studies investigated the roles of metabolic reprogramming in promoting glycolysis and redox hemostasis (6)(7)(8); however, the regulation of NADP metabolism in TSCC remains unclear and is of interest.
When angiogenesis lags around tumor cells, ischemic and hypoxic microenvironments are formed in local tumor tissues, and a large amount of reactive oxygen species (ROS) are produced. ROS are molecules such as O 2 •-, H 2 O 2 , HO 2 •, and HO• that contain one or more unpaired electrons in their orbital that are can participate in a considerable amount of oxidation reactions (9). ROS are generated from several different sources, including NADPH oxidases, mitochondria, cytochrome P450, and xanthine oxidase (10)(11)(12)(13). In cells, high levels of ROS cause irreversible damage to cellular components, leading to cell-cycle arrest and apoptosis (14). Thus, to survive oxidative stress, cancer cells need increased anti-oxidant capacity (15,16). The folate cycle plays a central role in cell metabolism (17) and produces metabolites that are critical for cell growth, including nucleotides and the major cellular anti-oxidant source NADPH (18,19). One study demonstrated that the most overexpressed NADPH-generating enzyme in the folate cycle, methylene tetrahydrofolate dehydrogenase 2 (MTHFD2), was associated with colorectal cancer progression (20). MTHFD1L catalyzes 10-formyl-THF to formate, which is the final step in the flow of 1C units from mitochondria to cytoplasm (21), acting important roles in embryonic development. As well, MTHFD1L encodes formyltetrahydrofolate synthetase and is involved in the synthesis of tetrahydrofolate (THF) in the mitochondria, playing critical roles in folate cycle maintenance (18,22,23). It has been reported that MTHFD1L overexpressed in many types of cancers, and promoted tumorigenesis and tumor progression (18,(24)(25)(26)(27).
The clinical relevance, function, and underlying regulatory mechanisms of MTHFD1L were investigated in this study. Here, we hypothesized that MTHFD1L plays an important role in TSCC cell survival during oxidative stress, resulting in TSCC progression. We expect MTHFD1L to be a new diagnostic and therapeutic target for TSCC.

Ethics Statement
We obtained human samples from Sun Yat-sen University Cancer Center; the tumor specimens used for the research were   Figure 1A); then, we used "MTHFD1L" as a keyword in the Oncomine search with "Cancer vs. Normal Analysis" as the primary filter and "Head and Neck cancer" as the cancer type. The mRNA levels of MTHFD1L were upregulated in multiple data sets, including those of Peng, He, Ye, and Vasko. The MTHFD1L expression data were log-transformed and mediancentered for each array, and the standard deviation (SD) was normalized to one for each array.
To explore the differences in potential biological functions in the low-and high-expression sets of MTHFD1L gene, GSEA was used using the Molecular Signatures Database (MSigDB) of Hallmark gene sets (h.all.v6.1.symbols).
(Suzhou, China). We used established stable cell lines from the CAL-27 and SCC-15 cell lines by selection with 3 µg.mL −1 puromycin for 3 weeks. We purchased adenoviruses from GenePharma Co., Ltd.

Antibodies and Western Blot Analysis
We used 10% SDS-PAGE gels to separate equal amounts of protein and transferred the proteins onto polyvinylidene fluoride (PVDF) membranes for detection. The membranes were then sequentially incubated with specific anti-bodies at 4 • C overnight; the membranes were then incubated with secondary antibodies at room temperature for 1 h the protein bands were detected using enhanced chemiluminescence. Anti-MTHFD1L, anti-MTHFD2 were purchased from Abcam, anti p-S6RP were purchased from Cell Signaling Technology, and anti-β-Actin was purchased from Proteintech (Wuhan, China).

RNA Extraction and Quantitative RT-PCR (qRT-PCR)
We extracted total RNA from the indicated cells using a RaPure Total RNA Micro Kit (Magen, Guangzhou, China). Then, extracted RNA was reverse transcribed with a ReverTraAce qPCR RT Master Mix kit (ToYoBo, Shanghai, China). The qRT-PCR primers for MTHFD1L were purchased from Invitrogen (Shanghai); qRT-PCR was performed with the SYBR Green Real-time PCR Master Mix (ToYoBo, Shanghai, China). The quantification method was -ddCt method.

Cell Colony Formation
Cells with MTHFD1L knockdown were seeded into six-well plates (500 cells/well) and were incubated in a humidified 5% CO 2 incubator at 37 • C for 14 days. We used formalin to fix the cells for 10 min and then stained the cells with crystal violet for 15 min. Next, we captured images of the clones and counted the number of clones (colonies with >50 cells were counted) using Image-Pro Plus 6.0 software.

Cell Proliferation
We performed MTS assay (Promega Biotech Co., Ltd., Madison, WI, USA) to determine the cell viability. We seeded cells into 96-well plates (4,000 cells/well), cultured them overnight, then transfected them with the MTHFD1L shRNA or a negative control. Cell viability was analyzed 1 week after transfection.

ROS, NADPH and GSH Measurement, Cell Apoptosis and Cell Cycle Analysis
First, cells were cultured in glucose-deprived medium and were incubated with 10 mmol/L 2, 7-dichlorodihydrofluorescein diacetate (H2-DCFDA, Thermo Fisher Scientific, cat. no. D399) at 37 • C for 30 min. Then, we collected the cells and washed them twice with 4 • C PBS; PBS was used to resuspend the cells; FACScan Flow Cytometer (Beckman-Coulter) was immediately used to measure the fluorescence. The NADP/NADPH-Glo Kit (Promega, cat. no. G9081) was used to measure the intracellular levels of NADPH and for the total NADP measurement. The intracellular levels of GSH/GSSG was measured with a GSH/GSSG-Glo Assay kit (Promega, WI, USA). Annexin V-FITC and PI (4A Biotech Co. cat. no. FXP018) were used for cell apoptosis analysis with a flow cytometer. The cell cycle was analyzed by PI/RNase staining buffer (BD Pharmingen TM ) with a flow cytometer.

Animal Experiments
We obtained 30 4-week-old female BALB/c nude mice from Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China), we quarantined them for 1 week before the animal experiments in an animal laboratory center. We suspended the cells for the in vivo tumorigenesis study. MTHFD1L-SC or shMTHFD1L-1 and shMTHFD1L-2 (3 × 10 6 cells, in PBS) were injected into BALB/c mice. We measured the volume of the tumors and the weight of the mice every 2 days for nearly 1 month. The mice were sacrificed at the end of the experiment. Tumors were excised, photographed, and processed for immunohistochemical analyses. All animal-associated experimental procedures were approved by the Animal Care and Use Committee of Sun Yatsen University. We made every effort to reduce the suffering of the animals.

Immunohistochemistry (IHC)
A total of 106 formalin-fixed paraffin-embedded (FFPE) TSCC tissue sections were provided by the Sun Yat-sen University Cancer Center. These sections were incubated with anti-MTHFD1L primary anti-bodies and secondary anti-bodies. Scoring for both the percentage and the intensity of positively stained tumor cells was performed by two independent pathologists under double-blind conditions.

Statistical Analysis
Statistical analyses were performed using the SPSS statistical software package (version 24.0 SPSS Inc., Chicago, IL, USA) to evaluate the relationship between MTHFD1L expression and TSCC clinicopathological features. We used Kaplan-Meier curves and the log-rank test to analyze disease-free survival and overall survival. P < 0.05 was considered to be statistically significant.

MTHFD1L Is Highly Expressed in TSCC
A Pyeon Multicancer study on the Oncomine database demonstrated that the MTHFD1L expression level was 4.50 times higher in TSCC tissues (15 samples) than in normal head and neck tissues (14 samples) and in cervix and uteri tissue (8 samples) (Figure 1A) (18). MTHFD1L was also found to be expressed at higher levels in head and neck cancer tissues than in normal tissues (P < 0.05) in several other studies ( Figure 1B) (30)(31)(32)(33).
To verify these findings, we analyzed the levels of MTHFD1L mRNA and protein expression separately in TSCC tissues (T) and adjacent non-carcinoma tissues (ANTs). The qRT-PCR results showed that the expression of MTHFD1L was significantly higher in the cancerous tissues than in the ANTs in 48 pairs of TSCC patients' fresh frozen tissues (P < 0.0001; Figure 1C). Western blot analysis also showed that expression of MTHFD1L was much more higher in the cancerous tissues than in the ANTs in 10 pairs of TSCC patients' fresh frozen tissues (Figure 1D).

The Clinicopathological Features of MTHFD1L in TSCC
In order to further explore the expression of MTHFD1L in TSCC tissue specimens and its correlation with clinicopathological characteristics in TSCC, we performed immunohistochemical analysis of 106 paraffin-embedded TSCC tissues with matched ANTs to evaluate MTHFD1L expression levels in TSCC. In our study, the immunohistochemical results demonstrated that the MTHFD1L protein was expressed at much higher levels in the tumor tissues than in the matched ANTs (Figure 2A). Statistical analysis of MTHFD1L expression in TSCC and ANTs was showed in Figure 2B. Then, we analyzed the potential correlation between the associated clinicopathological features of TSCC patient tissues and MTHFD1L protein expression.
The level of MTHFD1L expression and the clinicopathological features of TSCC patients are described in Table 1. The expression level of MTHFD1L was correlated with age staging (P = 0.024), T classification (P < 0.001), N classification (P = 0.003), clinical TNM stage (P = 0.001), but not with relapse (P = 0.622), gender (P = 0.232), differentiation state (P = 0.476), and M classification (P = 0.754; Table 1). These results indicated that high MTHFD1L expression was associated with advanced TNM stage in TSCC.
By comparing the overall survival time(OS) and diseasefree survival time(DFS) with the MTHFD1L expression (low, medium, high) of TSCC patients, we found that high MTHFD1L expression was associated with decreased diseasefree survival time (P = 0.036) and with poor prognosis (P = 0.044) in Figure 2C. We further performed the Kaplan-Meier survival analysis in the MTHFD1L expression subgroup. The results indicated that patients with high expression of MTHFD1L had shorter OS and DFS than patients with low expression of MTHFD1L. (P = 0.017 for OS, P = 0.014 for DFS; Supplementary Table 1). These results indicated that expression of MTHFD1L was associated with shorter survival of TSCC patients.

MTHFD1L Is Essential for TSCC Cell Proliferation and Cell Cycling
Furthermore, we examined the expression of MTHFD1L at the protein level in a human normal tongue epithelial cell line (NOK) and in TSCC cell lines (CAL-27, Tca8113, SCC-15, SCC-9, SCC-25) by Western blotting (Figure 3A). Among the cell lines examined, the expression of MTHFD1L was higher in the TSCC cell lines than in the NOK cell lines.
MTHFD1L plays an important role in the folate cycle, and folate metabolism can affect nucleic acid formation, then influencing cell proliferation (22). Thus, to clearly identify the functions of MTHFD1L in the TSCC cell lines, we knocked down MTHFD1L expression in the CAL-27 and SCC-15 cell lines ( Figure 3B). As expected, TSCC cell viability and colony formation were inhibited in the MTHFD1L knockdown cell line. After knocking down MTHFD1L, the cell proliferation rate, as well as the relative colony numbers, were significantly lower than those of the negative control cells (Figure 3C). We also observed that the knockdown of MTHFD1L affected the cell cycle, as the cell cycle was delayed in MTHFD1L knockdown cell lines compared with the negative control cells, with knockdown cells mainly remaining in the G1 phase (Figure 3D, Right). This result is statistically significant (Figure 3D, Left).

MTHFD1L Inhibits TSCC Cell Apoptosis via Antioxidant Activity
We performed gene set enrichment analysis (GSEA), and the signature of ROS-related genes was more abundant in tumors with high expression of MTHFD1L (normalized enrichment score = 1.681, P = 0.0037; Figure 4A) (The list of the detected enriched pathways was presented in Supplementary Table 2), revealing MTHFD1L acts important roles in redox homeostasis. MTHFD1L plays critical roles in folate cycle maintenance, which is an important source of NADPH (18). Thus, we knocked the MTHFD1L expression in the CAL-27 and SCC-15 cell lines, finding that the NADPH/NADP + ratio was reduced (as well as GSH/GSSG, Supplementary Figure 1) and that the ROS levels increased (Figures 4A-C). We also found that the MTHFD2 expression reduced after MTHFD1L knockdown (Supplementary Figure 2). To analyze apoptosis, we generated an oxidative stress environment under glucose deprivation and added H 2 O 2 separately. We observed that the cell death rate of MTHFD1L-knockdown cells was higher than that of the negative control cells. Interestingly, whether glucose deprivation or added H 2 O 2 induced cell death could be rescued by adding the antioxidant N-acetyl-L-cysteine (NAC) (P < 0.01; Figures 4D,E). These data demonstrate that MTHFD1L is essential for the maintenance of redox homeostasis and facilitates TSCC cell survival during oxidative stress.

MTHFD1L Promotes TSCC Tumorigenesis in vivo
We then performed cell-based xenograft experiments to evaluate the relationship between MTHFD1L expression and how it influenced tumorigenic ability in nude mice. As expected, the MTHFD1L knockdown group had slower tumor growth and lower tumor weight than the negative control group (P < 001, for control vs. knockdown groups; Figures 5A-C). We observed that in the CAL-27 MTHFD1L knockdown tumor biopsies. Immunohistochemistry indicated that there were reduced cell proliferation indices based on Ki67 and increased cell-apoptosis-associated indices based on cleaved caspase-3 (P < 0.05 for control vs. knockdown groups; Figures 5D,E). These results highlight the crucial roles of MTHFD1L in TSCC tumorigenesis.

Signaling Pathway of MTHFD1L Expression in TSCC
In summary, we have revealed that MTHFD1L promotes the development of TSCC by regulating the redox homeostasis. To further analyze the molecular regulation mechanism that may be involved, GSEA analysis indicated that the MTHFD1L expression is positively correlated with mTORC1 signaling pathway in TSCC (normalized enrichment score= 2.645, P < 0.01, Figure 6A, the list of the detected enriched pathways was presented in Supplementary Table 2). The mTORC1 signaling pathway is a typical signaling pathway involved in protein synthesis and cell proliferation (34). In addition, mTORC1 can activate cells through metabolic reprogramming, which makes cells dependent on glucose and glutamine uptake, especially in metabolic stress environments (35).
In order to verify whether mTORC1 affects the expression of MTHFD1L in TSCC, we treated TSCC cells with mTORC1 specific inhibitor rapamycin, and found that the activity of MTHFD1L promoter activity decreased significantly (Figure 6B).
In addition, rapamycin can significantly inhibit the phosphorylation of downstream substrate 4EBP1 (Figure 6C). Phosphorylated 4EBP1 can promote the release of eIF4E, and phosphorylation of Thr70 is most important for the release of eIF4E (36). eIF4E is the most important eukaryotic translation initiation factor, which is the most effective rate-limiting regulator of mRNA translation (37), and the decrease in p-4EBP1 is accompanied by a decrease in the expression of MTHFD1L ( Figure 6C). Then, we also knocked down the expression of raptor, an important component of mTORC1 activity, in the CAL-27 cell line, and found that the phosphorylation of the downstream substrate 4EBP1 was inhibited ( Figure 6D). Conversely, we knocked down the expression of MTHFD1L and found that other readouts of mTORC1 activity such as p-S6RP didn't change their expression (Supplementary Figure 2). In consequence, we predict that the mTORC1-4EBP1-eIF4E axis may affect the expression of MTHFD1L, which indicating that mTORC1 has a positive regulation of MTHFD1L in TSCC.
At the same time, we found that there was no significant change in cell apoptosis after treatment of TSCC cells with rapamycin or knockdown of raptor expression in TSCC cells. Interestingly, apoptosis was increased after added H 2 O 2 ( Figure 6E).

DISCUSSION
The folate cycle is associated with the maintenance of epigenetic modifications, nucleotide synthesis and anti-oxidant production, these biological events are always associated with tumorigenesis and tumor development. However, we do not know much about its effects on TSCC, hence this topic attracted our interest. MTHFD1L is a crucial folate cycle component. In this study, MTHFD1L was highly expressed in TSCC tissues compared with adjacent non-carcinoma tissues (ANTs), and its high expression was also associated with advanced clinical TNM stage and poor prognosis, demonstrating that MTHFD1L could be a prognostic indicator in TSCC. To determine the function of MTHFD1L in TSCC cell lines, we established MTHFD1L knockdown cell lines that effectively decreased cell proliferation rates and caused a cell cycle delay. All of these effects were related to NADPH reduction and ROS accumulation, as well as cell apoptosis and TSCC growth inhibition in vivo. In our next study, we also want to figure out the regulation mechanism of MTHFD1L expression in TSCC lymphatic metastasis.   It is well-known that MTHFD1L is overexpressed in multiple solid cancers, demonstrating that the folate cycle plays an important role in tumor metabolism and in accelerating tumor growth. Many conventional chemotherapies and radiotherapies eradicate cancer cells through ROS induction (26). These findings will inspire us to find more effective ways to treat cancer, such as the sensitization of existing cancer therapies using interference with folate circulation.
Additionally, the regulatory mechanism of MTHFD1L in TSCC should be studied in more depth. Previous studies have demonstrated that MTHFD1L is transcriptionally regulated by the transcription factor nuclear factor (erythroid-derived 2)like 2 (NRF2) (18), which acts as the central regulator for redox homeostasis (38). The KEAP1/NRF2 pathway is the key pathway that provides defense against oxidative stress, and it has been indicated to be the most frequently mutated pathway in human HCC (39). Whether this pathway plays a role in TSCC metabolism will be explored further in our next study.
The epigenetic regulation of metabolism-related genes has also been reported in several studies that have attracted our attention. MTHFD1L silencing reduced proliferation and enhanced the apoptosis of non-small cell lung cancer by suppressing DNA methylation (40). MTHFD1L supports the Flow of Mitochondrial One-carbon Units into the Methyl Cycle in Embryos as well (22). These reports provide support for our future research directions and further confirm the possibility that metabolic genes could serve as targets for tumor treatment. In the present study, the MTHFD1L expression could be reduced by mTORC1 inhibition and knockdown of raptor. We further found that mTORC1 facilitated the expression of MTHFD1L by promoting the phosphorylation of 4EBP1 to release eIF4E, thereby affecting the development of TSCC.

CONCLUSION
The expression of MTHFD1L accelerates the tumorigenesis in TSCC. Inhibition of the expression of MTHFD1L may become an actionable and effective therapeutic target in TSCC. This finding may also provide inspiration for us to study more about tumor metabolism in head and neck cancer.

DATA AVAILABILITY STATEMENT
All of the authors declare that all the data of this study are available in this article.

ETHICS STATEMENT
TSCC tissues were collected from patients who first diagnosed in Sun Yat-sen University Cancer Center and underwent surgical resection in the Head and Neck Surgery Department (SYSUCC, Guangzhou, China). All patients signed consent letters, and any manipulation of the tissues was approved by the Ethics Committee of Sun Yat-sen University. All animal procedures were performed in accordance with the guidelines of the Institutional Animal Care and Use Committee and the guidelines of Guangzhou Medical University and Sun Yat-sen University.

AUTHOR CONTRIBUTIONS
HL, AY, and XF designed the study. TT, FY, and CW performed the in vitro and animal experiments. HL, XF, TT, and FY analyzed the data and wrote the manuscript. All of the authors read and approved the final manuscript.