Down‐regulation of MTHFD2 inhibits NSCLC progression by suppressing cycle‐related genes

Abstract Methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) is a bifunctional enzyme located in the mitochondria. It has been reported to be overexpressed in several malignancies. However, the relationship between the expression of MTHFD2 and non‐small cell lung cancer (NSCLC) remains largely unknown. In this study, we found that MTHFD2 was significantly overexpressed in NSCLC tissues and cell lines. Knockdown of MTHFD2 resulted in reduced cell growth and tumorigenicity in vitro and in vivo. Besides, the mRNA and protein expression level of cell cycle genes, such as CCNA2, MCM7 and SKP2, was decreased in MTHFD2 knockdown H1299 cells. Our results indicate that the inhibitory effect of MTHFD2 knockdown on NSCLC may be mediated via suppressing cell cycle‐related genes. These findings delineate the role of MTHFD2 in the development of NSCLC and may have potential applications in the treatment of NSCLC.


| INTRODUC TI ON
Lung cancer is a leading cause of cancer-related death worldwide.
More than 220 000 cases are expected to be newly diagnosed in the United States in recent years, and the 5-year overall survival rate remains lower than most other types of cancer. 1,2 Non-small cell lung cancer (NSCLC) is the most common histological type of lung cancer in the clinic. Although advances in the surgical techniques, systemic chemotherapy and immunotherapy have improved the clinical outcome for NSCLC patients, the prognosis still remains unsatisfactory. 3,4 Therefore, it is of great clinical value to elucidate the molecular mechanism underlying the progression of NSCLC, so as to identify more specific therapeutic targets and develop new modalities of treatment.
Mutation in the expression pattern of cancer-related genes, which includes oncogenes and tumour suppressors, plays important roles in the tumorigenesis or tumour progression. 5,6 The microarray technology is a high-throughput platform to analyse the gene expression profiling. 7 Through using bioinformatics analysis, the microarray has been an effective strategy to obtain gene signature during tumorigenesis and identify molecular biomarkers for cancer patients. 8 Therefore, to explore and identify new molecular signature of NSCLC using microarray-based gene expression have a great appeal.
Mitochondrial methylenetetrahydrofolate dehydrogenase 2 (MTHFD2), a mitochondrial enzyme involved in the metabolism of folate, [9][10][11] is a bifunctional enzyme located in the mitochondria with methylene dehydrogenase and cyclohydrolase activities. 12,13 More recently, studies showed that MTHFD2 could confer redox homeostasis and promote caners cell growth, metastasis and correlate with poor survival. 14 MTHFD2 suppression decreased tumour burden and prolonged survival. 15,16 To date, the relation-

| Patient tissues
With the approval of the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University and informed consent, human NSCLC tissues and their adjacent tissues were obtained from the First Affiliated Hospital of Wenzhou Medical University. Fresh tissues were immediately snap-frozen and stored at −80°C, or fixed and embedded in paraffin.
Cells were cultured in a humidified atmosphere with 5% CO 2 at 37°C.

| Establishment of stable MTHFD2 knockdown cell lines
The short hairpin RNAs (shRNAs) targeting the mRNA sequence of MTHFD2 (shMTHFD2) and a negative control shRNA (shCtrl) were generated. The sequence of shMTHFD2 was AATGTGTTTGGATCAGTAT.
A549 and H1299 cell lines were infected with the lentivirus knocking down MTHFD2 (LV-shMTHFD2) or negative control (LV-shCtrl). The lentivirus was packaged and purchased from GENECHEM using above corresponding sequences. Stably transfected cell lines were isolated based on the puromycin selection.

| Cell proliferation assay
After being transfected, the cells were seeded into 96-well plates for further incubation. Cells were counted daily using the Celigo Imaging Cytometer (Nexcelom Bioscience), and each experiment was performed in triplicates.

| MTT assay
MTT assay was utilized to measure cell viability. Briefly, cells were seeded into 96-well plates and cultured overnight. MTT solution (20 μL) was added to each well. After 4 hours additional incubation, 150 μL DMSO was added. Absorbance was measured at 490 nm with an Enzyme mark instrument (M2009PR, Tecan infinite).
Data were analysed by flow cytometer (Millipore).

| Colony formation assay
The bottom agar layer was added to each well by 0.5% agar and media solution until it is semi-solid. And the top agar layer was made of 0.3% agar and media solution. Each cell line was seeded at 1000 cells/well on 6-well plates at 37°C in 5% CO 2 atmosphere overnight.
The culture medium was replaced by the fresh medium every two days to keep cells growing for 2 weeks. After 2 weeks, the colonies were stained with GIEMSA and photographed.

| RNA extraction and quantitative realtime PCR
Total RNA was extracted by using the TRIzol reagent (Invitrogen).
The cDNA of mRNA was synthesized using the M-MLV Reverse Transcriptase (Promega). Real-time PCR was performed with the Stratagene Mx3000P qPCR system (Stratagene-Agilent) using SYBR Green mix (Takara). During the cDNA preparation, the mRNA level was normalized to GAPDH expression. Each sample was tested in triplicate. The primers used were summarized in Table S1.

| RNA-seq and bioinformatics analysis
Total RNA was extracted respectively from 15 pairs of lung cancer tissues and the surrounding normal tissues or H1299 cells by using Trizol reagent (Invitrogen). An Agilent 2100 instrument was used to control the quality of total RNA. Amplified RNA was obtained via a GeneChip

| Western blot analysis
Total proteins were extracted from cells using ice-cold Radio Immunoprecipitation Assay (RIPA) lysis buffer containing protease inhibitors. BCA Protein Assay Kit (Takara) was used to quantify protein concentration. Approximately 20 μg of total protein lysate was separated by 10% SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Membranes were blocked with 5% skimmed milk for 2 hours and then incubated overnight with primary antibodies. After three rinses, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies at room temperature for 1 hour. Signals were visualized with Enhanced Chemiluminescence Detection Kit (Pierce Biotechnology).

| Statistical analysis
GraphPad Prism 7 (Graph Pad Software Inc) was used to analyse data. Data were expressed as mean ± SD of three independent experiments. Student's t test was used to compare two groups. Pvalue < .05 was regarded as statistically significant.

| Identification of MTHFD2 as a potential oncogene in human NSCLC
To identify dysregulated genes in NSCLC, the Human Gene Expression Array was employed to compare differentially expressed genes (DEGs) among 15 pairs of clinical NSCLC tissues and the corresponding adjacent non-tumour tissues (Table S2). As shown in Figure 1A, global transcriptional states of theses tumour samples were distinct from the corresponding normal tissues that were highly consistent in distribution within normal groups. Additionally, the microarray similarity in each group is greater than that between tumour and normal samples ( Figure 1B). Through using fold change > 1.8 and P < .05 as the threshold cut-offs, in total, 1588 genes including 600 up-regulated genes and 988 down-regulated genes showed statistically significant differential expression in NSCLC samples ( Figure 1C). As shown in Figure 1D, the supervised clustering of these DEGs identified was also exhibited.
According to the literature search and functional prediction of these significantly up-regulated DEGs, 20 highly abundant candidates in clinical NSCLC samples were chosen for further screening in H1299 cells using qRT-PCR, and up-regulated MTHFD2 which is a mitochondrial enzyme was ultimately selected for subsequent investigations ( Figure 1E).

| MTHFD2 is up-regulated in NSCLC tissues and cells
To assess the roles of MTHFD2 in NSCLC, we first evaluated the protein expression of MTHFD2 in a cohort of 150 clinical samples F I G U R E 1 MTHFD2 is selected as a potential oncogene in human NSCLC. A, Pearson's correlation plot with hierarchical clustering of 15 pairs of NSCLC tissues (red) and corresponding adjacent non-tumour tissues (black). B, Three-dimensional principal component analysis for total transcriptional landscapes of 15 pairs of NSCLC tissues and corresponding adjacent non-tumour tissues. C, Volcano plot of DEGs between NSCLC tissues and corresponding adjacent non-tumour tissues. Red, significantly DEGs. Fold change > 1.8 and P < .05 were considered significant. D, Supervised clustering of genes identified from NSCLC tissues and corresponding adjacent non-tumour tissues. E, qRT-PCR was used to detect the mRNA expression of 20 candidates in H1299 cells. The mRNA level of each sample is normalized to that of GAPDH by the 2-ΔCt method before comparative analysis consisted of 100 primary NSCLC tissues and 50 surrounding normal tissues (Table S3). As shown in Figure 2A,B, the protein level of MTHFD2 was significantly up-regulated in NSCLC specimens when compared with normal lung tissues. Next, we measured the mRNA level of MTHFD2 by qRT-PCR in NSCLC cells ( Figure 3A). The mRNA level of each sample was normalized to that of GAPDH prior to comparative analysis using the 2-ΔCt method. The ΔCt is equal to the difference between the ΔCt of MTHFD2 and GAPDH. The absolute value of ΔCt less than 12 is regarded to be expressive of high abundance, collectively, our data suggest that MTHFD2 is significantly up-regulated in NSCLC. We also examined MTHFD2 protein expression in lung cancer cells and lung normal epithelial cells. The result showed that MTHFD2 protein level was significantly overexpressed in lung cancer cells compared with that in lung normal epithelial cell Beas-2B ( Figure 3B).

| Knockdown of MTHFD2 inhibits the cell growth of NSCLC in vitro
To explore the functional role of MTHFD2 in NSCLC, we trans-

| Knockdown of MTHFD2 inhibits the tumorigenicity of NSCLC
The ability to form colonies in an anchorage-independent manner in soft agar cultures is one of the most significant characteristics of cancer cells. To evaluate the effect of MTHFD2 on colony formation F I G U R E 2 MTHFD2 is overexpressed in NSCLC tissue. A, Representative immunohistochemical staining of MTHFD2 on a tissue array containing100 clinical NSCLC tissues and 50 non-tumour tissues. The cytoplasmic staining results were evaluated based on the percentage of positive cells and the intensity of staining. B, Quantification of MTHFD2 protein expression. The values are expressed as mean ± SD (n = 50 for noncancerous group and n = 100 for NSCLC group). *Significantly different from normal tissues, ***P < .001 of NSCLC, soft agar assay was performed with H1299 and A549 cells treated with MTHFD2 knockdown. Colony formation assay showed that MTHFD2 silencing significantly decreased the number of colonies when compared with control in H1299 and A549 cells

| mRNA profiling reveals downregulation of cell cycle-related genes in H1299 cells with MTHFD2 knockdown
To explore the molecular mechanisms of MTHFD2 inhibiting cell proliferation and tumour growth, a genome-wide mRNA screening was employed to compare gene expression profiles between MTHFD2 silencing group and control in H1299 cells. Global transcriptional state of cells with MTHFD2 silencing was largely distinct with shCtrl-treated cells ( Figure 6A). Through using the criteria of fold change > 1.5 and P-value < .05, 501 DEGs were identified. Among these DEGs, 101 genes were up-regulated, and 400 genes were down-regulated ( Figure 6B and Table S4), and the supervised clustering of these DEGs was also exhibited ( Figure 6C). Furthermore, F I G U R E 3 Knockdown of MTHFD2 inhibits NSCLC cell growth in vitro. A, qRT-PCR was used to detect the mRNA expression of MTHFD2 in four NSCLC cell lines. The mRNA level of each sample is normalized to that of GAPDH by the 2-ΔCt method before comparative analysis. B, MTHFD2 protein level was significantly overexpressed in NSCLC cells compared with that in lung normal epithelial cell Beas-2B. C, Western blot analysis of MTHFD2 expression in MTHFD2-silenced cells. Cells were respectively transfected with shMTHFD2 at a low MOI (shMTHFD2-L) and high MOI (shMTHFD2-H) for 48 h. GAPDH was used as a loading control. D, qRT-PCR was used to detect the mRNA expression of MTHFD2 after transfection with shMTHFD2 at a high MOI for 48 h. The mRNA level is normalized to GAPDH by the 2 −ΔCt method before comparative analysis. E, Celigo cell counting analysis of cells transfected with shMTHFD2. Cell growth was measured using fluorescent photomicrographs every day for 5 d to capture the cells with green fluorescence. Growth curve was plotted by algorithms of the raw data of images (shCtrl vs shMTHFD2). The values are expressed as mean ± SD.*Significantly different from shCtrl, ***P < .001, ****P < .0001 the diseases and functional analysis showed that the terms of cellular growth, proliferation and cancer were largely suppressed ( Figure 6D). Altogether, our results indicate that the inhibitory effect of MTHFD2 knockdown on NSCLC may be mediated via suppressing cell cycle-related genes.

| Ingenuity pathway analysis of DEGs from H1299 cells with MTHFD2 knockdown
To further explore the regulatory pathways affected by these DEGs, we next performed the IPA functional analysis. Herein, canonical pathway analysis identified the S-phase entry pathway (Z score < 2) was the most significantly suppressed signalling ( Figure 7A). 17 qRT-PCR was used to verify the down-regulation of 3 cell cycle-related genes in H1299 cells after silencing MTHFD2 ( Figure 7B). Besides, MTHFD2 overexpression increased the protein levels of CCNA2, MCM7 and SKP2 ( Figure 7C). Conversely, the protein expression of CCNA2, MCM7 and SKP2 was significantly suppressed after MTHFD2 silencing ( Figure 7D). Further, we assessed the protein levels of cycle-related genes in tumour xenograft tissues. In accordance with our above results, Western blot analysis revealed that CCNA2, MCM7 and SKP2 expression were decreased in the MTHFD2 knockdown tumour tissues compared with shCtrl tumour tissues ( Figure 7E). Altogether, these results suggest that MTHFD2 knockdown may inhibit cell proliferation and tumour growth via regulating cell cycle-related genes.

F I G U R E 4 Knockdown of MTHFD2
inhibits proliferation and promotes apoptosis in NSCLC cells. A, Cell proliferation was determined by MTT assay in cells transfected with shMTHFD2 and shCtrl for 5 d. B, Cell apoptosis was measured using Annexin V staining and flow cytometry analysis in two groups of cells stably silencing MTHFD2. The horizontal coordinate was the signal value of Annexin V-APC, and the vertical coordinate was the green fluorescence of the target gene virus infected in cells. C, Representative images of decreased colonies formation in monolayer culture induced by MTHFD2 silenced in NSCLC cells. The values are expressed as mean ± SD.*Significantly different from shCtrl, **P < .01, ***P < .001, ****P < .0001. D, Western blot analysis of MTHFD2 protein expression in A549 cells following MTHFD2 overexpression. E, Representative results of colony formation assay in MTHFD2-overexpressing A549 cells. The quantitative numbers of colonies are shown at the right panel. The experiment was performed in triplicate wells in three independent experiments. *Significantly different from vector, ***P < .001

| D ISCUSS I ON
Non-small cell lung cancer is a widespread malignancy with increasing incidence rate which demands intensive investigation. 4 Herein, we initially identify MTHFD2 is significantly up-regulated in NSCLC as a potential oncogene. We also demonstrate that MTHFD2 plays an essential role in the development of NSCLC.
Methylenetetrahydrofolate dehydrogenase 2 was found to be co-expressed with cell cycle proteins to progress cancer cell proliferation. 13,18 To explore the molecular mechanisms underlying the oncogenic role of MTHFD2, mRNA profiling was employed to obtain F I G U R E 5 Knockdown of MTHFD2 inhibits the tumorigenicity of NSCLC. A, Images of the subcutaneous tumours formed in the nude mice after injection of shMTHFD2 and shCtrl transfected H1299 cells. B, Tumour growth curves revealed that xenograft tumour growth in nude mice was significantly slower in shMTHFD2-treated group than that of shCtrl. C, Mean tumour weights 48 d after transplantation was shown. D, Western blot analysis of MTHFD2 in tumour xenograft tissues. The values are expressed as mean ± SD.*Significantly different from shCtrl, *P < .05, **P < .01 F I G U R E 6 Deregulated genes in MTHFD2 knockdown H1299 cells. A, Pearson's correlation plot with hierarchical clustering of H1299 cells transfected with shMTHFD2 and shCtrl. B, Volcano plot of DEGs between shMTHFD2 group and shCtrl group. Red, significantly DEGs. Fold change > 1.5 and P < .05 were considered significant. C, Supervised clustering of genes identified from shMTHFD2 group and shCtrl group. D, Diseases and functional analysis of DEGs between shMTHFD2 and shCtrl by IPA software. a: cell cycle; b: DNA replication, recombination and repair; c: cancer; d: organismal injury and abnormalities; e: reproducitive system disease; f: neurological disease; g: cellular assembly and organization; h: organismal survival; i: gastrointestinal disease; j: cell death and survival; k: connective tissue disorder; l: cellular growth and proliferation; m: cellular development; n: endocrine disorders; o: reproductive system development and function potential genes and pathways regulated by MTHFD2 via comparing DEGs between MTHFD silencing and empty virus-transfected H1299 cells. By functional analysis, we identified cellular growth and proliferation to be highly suppressed, which may lead to the inhibitory effects of MTHFD2 silencing on NSCLC. Besides, we found that MTHFD2 knockdown significantly down-regulated the mRNA and protein expression of cell cycle genes such as CCNA2, MCM7 and SKP2. Thus, our study found the cooperation between MTHFD2 and cell cycle genes in NSCLC development. 19 Methylenetetrahydrofolate dehydrogenase 2 is a mitochondrial methylenetetrahydrofolate dehydrogenase and cyclohydrolase involved in one-carbon metabolism. MTHFD2 plays a critical role in controlling N6-methyladenosine (m6A) methylation of HIF-2α levels and the oxidation of methylene-THF to 10-formyl-THF in mitochondria, which results in promoted metabolic reprograming and tumour growth. 20 In addition, MTHFD2-dependent glycine synthesis is a prerequisite for angiogenesis. 10 The exact mechanistic role of MTHFD2 in cancer is still a topic in the future. Given such important roles in the cancer cells proliferation, MTHFD2 has been recently considered to be a promising target for multiple types of cancer. [21][22][23] During the treatments of acute myeloid leukaemia and colorectal cancer, targeting MTHFD2 can markedly suppress the tumour progression both in vitro and in vivo. 16,21 Importantly, our findings here extend the therapeutic function of MTHFD2 to NSCLC, which targeting MTHFD2 can be a potentially valuable approach in the clinic.
Our findings also suggest that this enzyme may represent a novel therapeutic target for NSCLC treatment.
This finding provides additional evidence that MTHFD2 contributes to malignancy in NSCLC. It is possible that MTHFD2 plays important roles for conferring drug resistance in NSCLC. 6 Moreover, MTHFD2mediated lung cancer cells resistance to gefitinib. 24 KRAS mutation status is associated with the expression of MTHFD2 in lung cancer. 25 MTHFD2 was also identified as a miR-9 target gene that affects cell proliferation. 26 Previously published studies have relied on shRNA or small molecule inhibitors directed suppression of MTHFD2. [14][15][16]21,24 In future studies, it may also be informative to development of selective MTHFD2 inhibitors testing their effects in preclinical trials and the combinatorial effects with clinical chemotherapy drugs.

F I G U R E 7
Identification of several genes as target of MTHFD2 in NSCLC. A, IPA canonical pathway analysis of DEGs between shMTHFD2 and shCtrl. On the horizontal axis, most significantly overrepresented pathways identified are exhibited, whereas the vertical axis shows the −Log 10 of the p-value calculated based on the Fisher exact test. The ratio reported as orange points represent the numbers of genes in a given pathway that meet cut-off criteria divided by the total numbers of genes that make up that pathway. a: cell cycle control of chromosomal replication; b: 5-aminoimidazole ribonucleotide biosynthesis I; c: purine nucleotides de novo biosynthesis II; d: oestrogen-mediated S-phase entry; e: cleavage and polyadenylation of pre-mRNA; f: antiproliferation role of TOB in T-cell signalling; g: Myo-inositol biosynthesis; h: p53 signalling; i: tetrahydrofolate salvage from 5,10-methenyltetrahydrofolate; j: D-myo-inositol(1,4,5)trisphosphate degradation; k: protein kinase A signalling; l: Wnt/β-catenin signalling; m: aryl hydrocarbon receptor signalling; n: pyrimidine deoxyribonucleotides de novo biosynthesis; o: histidine degradation III. B, qRT-PCR was used to detect the mRNA expression of 3 selected genes after transfection with shMTHFD2 and shCtrl in H1299 cells. The values are expressed as mean ± SD. *Significantly different from shCtrl, *P < .05, **P < .01. C, Western blot analysis of MTHFD2, CCNA2, MCM7 and SKP2 in MTHFD2-overexpressing H1299 cells. D, Western blot analysis of CCNA2, MCM7 and SKP2 after transfection with shMTHFD2 and shCtrl in H1299 cells. E, Western blot analysis of CCNA2, MCM7 and SKP2 in tumour xenograft tissues In summary, our preliminary study demonstrates that MTHFD2 is up-regulated in NSCLC and plays important roles in the cell growth of NSCLC via promoting cell cycle genes expression. Our study provides a basis for utilizing MTHFD2 as a new diagnostic and therapeutic target in NSCLC.

ACK N OWLED G EM ENTS
This work was financially supported by the Medical Scientific Research Fund of Zhejiang Province (2019322308), Wenzhou science and technology project (Y20170280 and Y20190179) and Science and Technology Innovation Activity Plan for College Students of Zhejiang Province (2019R413083).

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.