Long Non-Coding RNA FOXD1-AS1 Promotes the Progression and Glycolysis of Nasopharyngeal Carcinoma by Sustaining FOXD1 Expression

Zhanwang Wang Central South University Third Xiangya Hospital https://orcid.org/0000-0002-3655-0704 Dong He Hunan Provincial People's Hospital Yuxing Zhu Central South University Third Xiangya Hospital Xueying Hu Central South University Third Xiangya Hospital Yi Jin Central South University Third Xiangya Hospital Lian Gong Central South University Third Xiangya Hospital Mengqing Xiao Central South University Third Xiangya Hospital Liang Xiang Central South University Third Xiangya Hospital Qinghai Zeng Central South University Third Xiangya Hospital Jianye Liu Central South University Third Xiangya Hospital Xingyu Chen Central South University Third Xiangya Hospital Yeyu Zhang Central South University Third Xiangya Hospital Ke Cao (  csucaoke@163.com ) Central South University Third Xiangya Hospital

metastasis, and occurrence of multiple human cancers, including NPC [12][13][14][15][16], how lncRNAs function in patients with NPC remains unclear. Thus, further investigation into lncRNA regulatory mechanisms to provide prognostic biomarkers and therapeutic targets for NPC patients would be bene cial. FOXD1 antisense RNA1 (FOXD1-AS1) is a novel lncRNA that has been reported to function as a potential oncogenic biomarker in glioma [17]. Nevertheless, the pivotal role of FOXD1-AS1 in NPC remains to be elucidated.
In this study, we found that FOXD1-AS1 was highly overexpressed in NPC compared to non-NPC (normal) tissues, and overexpression of FOXD1-AS1 was a predictor for poor overall survival (OS) and disease-free survival (DFS). In vitro and in vivo functional experiments veri ed that FOXD1-AS1 promotes NPC cell proliferation, migration, metastasis, and glycolysis levels, and inhibits apoptosis of NPC cells by upregulating FOXD1 expression. Therefore, our study elucidates the regulatory mechanism and clinical signi cance of FOXD1-AS1 in NPC and provides a potential prognostic biomarker and therapeutic target for NPC.

Methods And Materials
Patients and specimens Nasopharyngeal carcinoma tissues and normal nasopharyngeal tissues were obtained from 10 NPC patients who received treatment at the Third Xiangya Hospital (Changsha, Hunan, China) from May 2017 to May 2019. The patients were diagnosed by pathological analysis and were not subjected to radiation therapy or chemotherapy. The Institutional Review Board of the Ethics Committee of Third Xiangya Hospital approved the consent procedure, and written informed consent was provided by all patients in the study.

MTT assays
Approximately 1×10 4 NPC cells were seeded in a 96-well plate and incubated for 0, 12, 24, 48, and 72 h. After the incubation period, 50 μL MTT (Beyotime, Shanghai, China) was added into each well and reincubated in a 5% CO 2 atmosphere at 37°C for 4 h. Subsequently, the MTT was discarded and 150 μL of dimethyl sulfoxide (DMSO) was added into each well to dissolve the MTT crystals. Finally, the optical density at 570 nm was measured.
Wound healing assays Wound healing assays were performed to assess the migration capacity of NPC cells as previously described. NPC cells (1×10 5 /mL) were cultivated in 6-well plates. Once the cells were grown to 90% con uence, a wound was scratched using a 200-μl pipette tip, after which the cells were washed twice with phosphate-buffered saline (PBS) to dislodge any debris. DMEM without FBS was added to the 6-well plates. The photographs of wound healing were taken via a microscope at 0 h and 48 h.

Transwell assays
Transwell assays were performed to assess the invasion of NPC cells. After transfection for 24 h, NPC cells were collected and washed twice with serum-free DMEM. A 200 μl (5×10 4 cells/mL) cell suspension was plated into the upper chamber, which was pre-coated with Matrigel Matrix (Corning, USA), while 600 μl DMEM containing 20% FBS was added to the lower transwell chamber. After incubation at 37°C for 24 h, cells that had invaded across the membrane were xed with 95% ethanol for 20 min and then stained with hematoxylin for 10 min. Images of stained cells were randomly photographed and counted under an inverted microscope.

Apoptosis via ow cytometry
The apoptosis rate of NPC cells was detected using an Annexin V-PE/7-AAD apoptosis detection kit (KA3809, Abnova) following the manufacturer's instructions. Brie y, approximately 3×10 5 cells were seeded into 6-well plates. After transfection with different shRNAs or plasmids for 48 h, the cells were fully digested and collected. Cells were suspended in a cold binding buffer to a concentration of 1×10 6 cells/mL, and then stained with 5 μl of Annexin V-PE and 5 μL of 7-AAD solution in the dark at room temperature. The results were observed via ow cytometry (Beckman Coulter, Inc.).

Measurement of lactate production and glucose consumption.
Lactate production and glucose consumption were measured as previously described [19]. For this, 5×10 5 cells were cultured and incubated in 6-well plates for 10 h, after which the medium was discarded and the cells were incubated with a fresh medium for 8 h. The levels of lactate and glucose were detected using an automatic biochemical analyzer (HITACHI, Tokyo, Japan). Consistent amounts of protein were used to normalize the relative levels of lactate production and glucose consumption.

Fluorescence in situ hybridization
Fluorescence in situ hybridization (FISH) assays were performed to verify the location and expression of FOXD1-AS1 in NPC cells. Blue DAPI stain and red FISH Probe were purchased from RiboBio company (Guangzhou, China). A FISH kit (RiboBio, Guangzhou, China) was used to perform these assays in accordance with the manufacturer's protocols.

Immunohistochemistry (IHC)
Para n-embedded NPC tissues were sliced, dewaxed, hydrated, and antigen-repaired. Endogenous peroxidase was blocked; anti-FOXD1 (Omnimabs, OM203273), anti-Ki67 (Genetex, GTX16667), and anti-Caspase3 were added to these sections and incubated together at 4 °C overnight. Polymer enhancers were incubated for 20 min at room temperature, then biotin-labeled secondary antibodies were added and incubated for 30 min at room temperature. Next, a diaminobenzidine staining solution was used to stain the sections, followed by counterstaining with hematoxylin, and then the sections were mounted in glycerol-vinyl-alcohol (GVA Mount, Zymed). Two independent pathologists, who were blinded to the data and histopathological characteristics of the patients, evaluated the IHC scores according to the scoring standards as previously described.
Stability and α-amanitin treatment CNE-1 and 5-8F cells stably expressing shRNA against FOXD1-AS1 or NC were seeded into 6-well culture plates. Cells were then treated with 5 μg/ml actinomycin D and were harvested at the indicated time points. Total RNA was extracted for cDNA synthesis and subjected to qRT-PCR after 0, 2, 4, 6, 8, and 12 h of treatment.

RNA pull-down assay
Pierce™ Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher, Rockford, IL, USA) was used to perform RNA pull-down assays. Biotinylated bio-NC and bio-FOXD1-AS1 were transfected into CNE-1 and 5-8F cells. After being transfected for 48 h, the cell lysates were mixed with magnetic beads to form proteinbio/RNA-magnetic bead complexes, and high-salt elution was used to obtain protein-bio/RNA complexes. Subsequently, western blotting was performed to measure the relative expression of FOXD1 in the protein-bio/RNA mixture.

Chromatin immunoprecipitation (ChIP)
ChIP assays were performed as previously described [11]. For this, protein-DNA complexes were obtained with the anti-FOXD1 antibody (1:50; Omnimabs, OM203273), followed by analysis of the resulting retrieved RNA subjected to qRT-PCR. The primers for qRT-PCR quanti cation were designed according to the promoter regions of FOXD1-AS1, lactate dehydrogenase A (LDHA), pyruvate kinase M2 (PKM) and enolase 1 (ENO1), which contain FOXD1 binding sites. The primer sequences used were as follows:

Xenograft mouse model
Female BALB/c nude mice, aged 6 weeks (20±2 g), were purchased from the Laboratory Animal Center of Central South University (Changsha, China) and subjected to tumor implantation. CNE-1 and 5-8F cells (5×10 6 cells/ml, 0.2 ml) were transfected with either shRNA targeting FOXD1-AS1, shRNA targeting FOXD1-AS1+FOXD1 overexpression plasmids, or vector plasmids, and were then subcutaneously injected into the mice. On day 25, the mice were euthanized to harvest tumors, and the volume of the tumors was calculated using the formula V= (L×W²)/2. Tumors obtained from the mice were examined using routine hematoxylin and eosin (H&E) staining and photographed under a light microscope. All experimental procedures were approved by the Ethics Committee of the Third Xiangya Hospital of Central South University (Changsha, China).

Statistical analyses
Data are presented as mean ± SD, and all experiments were repeated independently three times. GraphPad Prism 8.0 Software (San Diego, CA, USA) was used to analyze the data. Student's t-tests or one-way ANOVA were used to analyze differences between groups. P < 0.05 (two-sided) was considered statistically signi cant.

Results
Upregulated FOXD1-AS1 expression is negatively correlated with the prognosis of nasopharyngeal carcinoma Analysis of both GEPIA and GEO (GSE95166) databases revealed that FOXD1-AS1 was upregulated in multiple cancers (Figure S1A), including NPC (Fig. 1A, B). Using qRT-PCR, we further veri ed that the expression level of FOXD1-AS1 was higher in NPC tissues than in adjacent non-cancerous nasopharyngeal tissues (Fig. 1C). Patients with higher expression levels of FOXD1-AS1 had a shorter overall survival (OS; P = 0.017, Fig. 1D) and a shorter disease-free survival (DFS; P = 0.03, Fig. 1E) time. Taken together, these results suggest that upregulated FOXD1-AS1 expression may promote tumor progression and serve as a prognostic biomarker in patients with NPC.

FOXD1-AS1 upregulates FOXD1 expression
To examine the detailed mechanism by which FOXD1-AS1 functions in patients with NPC, we conducted an RNA FISH assay, which indicated that FOXD1-AS1 is located in both the cytoplasm and nucleus of CNE-1 and 5-8F cells (Fig 3A). The UCSC Genome Browser reveals that FOXD1-AS1 is a divergent transcript of FOXD1; thus, we investigated their relationship in patients with NPC. We found that FOXD1 mRNA expression was positively correlated with FOXD1-AS1 mRNA upon analyzing GEPIA ( Fig 3B) and ChIPBase (Fig S2A) databases. Furthermore, FOXD1 expression at the mRNA and protein levels decreased after FOXD1-AS1 depletion while increased under FOXD1-AS1 overexpression (Fig 3C, D). Therefore, we hypothesized that FOXD1-AS1 could upregulate the expression of FOXD1. First, CNE-1 and 5-8F cells were transfected with NC or sh-FOXD1-AS1, followed by treatment with α-amanitin (an inhibitor of RNA polymerase II), which could stop new transcript synthesis. The results showed that FOXD1-AS1 depletion signi cantly shortened the half-life of FOXD1 mRNA (Fig 3E). Subsequently, FOXD1-AS1 was knocked down in CNE-1 and 5-8F cells following treatment with cycloheximide (CHX), a protein synthesis inhibitor, which could inhibit the synthesis of proteins. The results revealed that FOXD1-AS1 depletion strikingly reduced the stability of FOXD1 in CNE-1 and 5-8F cells (Fig 3F). Furthermore, it was veri ed that FOXD1-AS1 could directly bind to the FOXD1 protein by conducting an RNA pulldown assay ( Fig 3G) and RIP assay (Fig 3H). Taken together, these results suggest that FOXD1-AS1 could upregulate the expression level of FOXD1.

Elevated FOXD1 promotes proliferation, migration, and invasion, and suppresses apoptosis in NPC cells
To study the function of FOXD1, stable FOXD1-depletion or FOXD1-overexpression cells were established, and the expression e ciency of FOXD1 was validated through qRT-PCR (Fig 4A, B). We then investigated the impact of FOXD1 on oncogenic potential, and the results revealed that FOXD1 signi cantly facilitated the proliferation (Fig 4C, D), migration (Fig 4F, G), and invasion (Fig 4H) of NPC cells. Furthermore, ow cytometry indicated that FOXD1-AS1 suppressed the apoptosis rate of CNE-1 and 5-8F cells (Fig 4E). Analysis of the GEPIA, Starbase, and GEO (GSE12452) databases revealed that FOXD1 expression was higher in NPC samples than in adjacent normal nasopharyngeal samples (Fig 5A-D), and higher levels of FOXD1 were correlated with shorter overall survival and disease-free survival times of patients with NPC (Fig 5E, F).

FOXD1 promotes the glycolytic level of NPC cells
To clarify the downstream pathway of FOXD1, KEGG analysis was performed using co-expression genes (top 200) of FOXD1 found in the GEPIA database, and the results showed that FOXD1 was associated with glycolysis (data not shown). Among multiple glycolysis-related genes found in the GEPIA and ChIPBase databases, we found that FOXD1 has a strong positive association with lactate dehydrogenase A (LDHA), pyruvate kinase M2 (PKM) and enolase 1 (ENO1) genes (Fig 5H and Fig S2B-D), which play vital role in glycolysis [20][21][22]. The protein levels of LDHA, PKM, and ENO1 signi cantly decreased after FOXD1 knockdown, whereas the levels increased under FOXD1 overexpression ( Fig 5J). As a transcription factor, FOXD1 regulates the expression of speci c genes. Analysis of UCSC and JASPAR databases revealed that FOXD1 could bind to the promoters of LDHA, PKM, ENO1, and FOXD1-AS1. We designed the primers according to the binding sites of these genes and conducted a CHIP assay, which veri ed that FOXD1 could in fact bind to the promoters of LDHA, PKM, and ENO1 (Fig 5I), and positively regulated their expression levels (Fig 5J), whereas FOXD1 could not bind to the promoter of FOXD1-AS1. To investigate the role of FOXD1 in anaerobic glycolysis, we transfected CNE-1 and 5-8F with FOXD1 silencing shRNA or FOXD1 overexpression plasmids, and biochemically detected ATP levels, glucose consumption, and lactic acid production. Knockdown of FOXD1 expression decreased ATP levels, glucose consumption, and lactic acid production in CNE-1 and 5-8F cells, and overexpression of FOXD1 increased the glycolytic metabolism of CNE-1 and 5-8F cells (Fig 5K). These results indicated that FOXD1 increases NPC cell glycolytic metabolism through transcriptionally regulated related glycolytic genes.

FOXD1-AS1 promotes progression and glycolysis of NPC cells by sustaining FOXD1 expression
To verify whether FOXD1-AS1 plays an oncogenic role in NPC by increasing FOXD1 expression, we transfected CNE-1 and 5-8F cells with FOXD1-AS1 silencing shRNA or FOXD1-AS1 silencing shRNA together with a FOXD1 overexpression plasmid. The depletion of FOXD1-AS1 decreased the protein levels of FOXD1, LDHA, PKM, and ENO1, as well as ATP levels, glucose consumption, and lactic acid production, whereas the overexpression of FOXD1 signi cantly reversed the effects of FOXD1-AS1 knockdown of proteins and glycolytic metabolism levels in CNE-1 and 5-8F cells (Fig 6A, B). MTT, ow cytometry, wound healing, and transwell assays all demonstrated that knockdown of FOXD1-AS1 expression in CNE-1 and 5-8F cells inhibited proliferation, migration, and invasion, and induced apoptosis, whereas the overexpression of FOXD1 reversed this effect (Fig 6C-F).

Effect of the FOXD1-AS1/FOXD1 axis on NPC cell progression in vivo
Compared with normal tissues, FOXD1 was signi cantly increased in NPC tissues (Fig 7A), which was further veri ed by immunohistochemistry (Fig 7B). To con rm the effect of the FOXD1-AS1/FOXD1 axis in vivo, we employed a xenograft mouse model to validate it. The depletion of FOXD1-AS1 using the sh-FOXD1-AS1 plasmid reduced the tumorigenic ability of CNE-1 cells, whereas overexpression of FOXD1 reversed this effect (Fig 7C). Immunohistochemistry and western blot assays revealed that Ki67, caspase3, FOXD1, LDHA, PKM, and ENO1 protein levels were signi cantly decreased after knockdown of FOXD1-AS1 in CNE-1 cells, while these protein levels were restored by upregulating FOXD1 overexpression (Fig 7D, E).

Discussion
With great improvements technology, the therapeutic effect of radiotherapy on nasopharyngeal carcinoma becomes increasingly prominent, which signi cantly improves the prognosis and outcome of patients with NPC, 70-80% of whom have a 5-year overall survival rate [23]. However, some patients with NPC receiving radiotherapy develop radio-resistance, leading to local recurrence and distant metastasis. Simply increasing the radiation dose alone may cause serious post-treatment adverse reactions. Therefore, further investigation of the critical molecules involved in the progression of NPC may provide a novel biomarker for improving the e cacy of radiotherapy in patients with NPC.
Accumulating evidence has shown that dysregulated lncRNAs play an important role in the progression of various human cancers, including NPC. For example, upregulation of FAM225A indicated poor survival of NPC and promoted tumorigenesis and metastasis by sponging miR-1275/miR-590-3p and upregulating ITGB3 expression [24]. AATBC facilitated migration, invasion, and metastasis of NPC cells in vitro and in vivo through the miR-1237-3p-PNN-ZEB1 axis [25]. In this study, we rst reported that the lncRNA FOXD1-AS1 was obviously overexpressed in NPC tissues and cell lines and was closely correlated with poorer prognoses in patients with NPC. Therefore, it could be recognized as a prognostic biomarker for NPC. Functionally, the knockdown of FOXD1-AS1 inhibited cell proliferation, migration, and invasion, as well as promoted the apoptosis of NPC cells, revealing the oncogenic role of FOXD1-AS1. Similarly, the overexpression of FOXD1-AS1 contributed to cell growth, migration, and metastasis, and inhibited the apoptosis of NPC cells. Additionally, Gao et al. [17] reported that FOXD1-AS1 promotes cell proliferation and migration and decreases apoptosis by regulating eIF5a protein levels in gliomas.
Besides FOXD1-AS1 upregulation, FOXD1 expression level is also signi cantly increased in NPC tissues and is associated with a worse OS and DFS in patients with NPC, implying that FOXD1 may be involved in NPC development and progression. As a transcription factor, the FOXD1 protein possesses a highly conserved DNA-binding domain composed of approximately 100 amino acids in a helix-turn-helix structure that could bind with speci c DNA sequences on the promoter region of target genes, which promotes or inhibits the targeted gene expression [26][27]. Several studies have revealed that FOXD1 contributes to the progression of multiple cancers. For instance, increased FOXD1 expression facilitates cell proliferation and inhibits apoptosis in colorectal cancer by positively regulating Plk2, and a high expression of FOXD1 indicates a worse prognosis in these patients [28,29]. Similarly, FOXD1 can function as a regulator of VEGF-A and promote angiogenesis in colorectal cancer [30]. Furthermore, FOXD1 promotes breast cancer cell growth by accelerating the transition from G1 phase to S phase [31].
In our research, we demonstrated that ectopic expression of FOXD1 could promote cellular proliferation, migration, and invasion, and inhibit the apoptosis of NPC cells, suggesting that FOXD1 may serve as a potential therapeutic target for NPC.
To our knowledge, partial or natural antisense lncRNAs are generated from the antisense strand of a protein-coding gene and are involved in regulating the expression of this gene through various mechanisms [32][33][34]. The UCSC genome browser shows that FOXD1-AS1 is localized in the antisense strand of the FOXD1 gene, which prompted us to wonder whether FOXD1-AS1 might regulate FOXD1 expression to promote the progression and development of NPC. To verify this hypothesis, we examined the expression levels of FOXD1-AS1 and FOXD1, and we found that FOXD1-AS1 is concordantly overexpressed with FOXD1 in NPC cells and tissues. Furthermore, our results indicated that the silencing of FOXD1-AS1 contributed to the reduction of FOXD1 mRNA and protein levels, as well as to mRNA stability and protein degradation of FOXD1 in NPC cells. As such, FOXD1-AS1 and FOXD1 can form a socalled "RNA-RNA duplex" structure that protects FOXD1 mRNA from RNase degradation; moreover, we speculate that the mechanism via which FOXD1-AS1 stabilizes FOXD1 protein is that FOXD1-AS1 recruits deubiquitinating enzymes to avoid FOXD1 protein degradation by proteasomes, a hypothesis which needs to be further explored and validated in our future studies. Together, these results suggest that FOXD1-AS1 can stabilize FOXD1 mRNA and protein, thereby leading to the upregulation of FOXD1 expression at both RNA and protein levels. Similarly, several researchers also reported that antisense lncRNAs can regulate their related genes in cancer progression. For instance, EZR-AS1 promotes the mobility and invasiveness of esophageal squamous cell carcinoma via enhancing SMYD3-dependent H3K4 methylation which plays an important role in enhancing transcription of the EZR gene [35]. NAMPT-AS promotes tumor progression and metastasis by serving as a sponge and recruiting transcription factors to epigenetically activate NAMPT in triple negative breast cancer [36]. The role of FOXD1-AS1 on the tumorigenic ability of NPC cells and the regulatory effect of FOXD1-AS1 on FOXD1 was also veri ed with in vivo experiments. Our ndings suggest a potential molecular mechanism of FOXD1 upregulation in NPC.
Hypoxia is a fundamental characteristic of cancers and plays a vital role in the progression and development of human cancers [37]. Therefore, the glucose metabolism of tumor cells is different from that of normal cells. Tumor cells switch their metabolism mainly to glycolysis in an anaerobic environment, or even an environment with abundant oxygen (Warburg effect) [38]. The Warburg effect is also known as aerobic glycolysis, which explains the acceleration of tumor growth via increasing glucose uptake, ATP, and lactic acid production during the occurrence and progression of various human cancers [39]. Several studies have revealed that certain genes, including glucose transporter (GLUT), lactate dehydrogenase A (LDHA), hexokinase 2 (HK2), pyruvate kinase M2 (PKM), and enolase 1 (ENO1), are involved in glycolysis [40,41]. In this study, we found that FOXD1-AS1 could regulate the glycolysis of NPC cells by increasing the expression of LDHA, PKM, and ENO1 through the FOXD1-AS1-FOXD1 axis.
These three glycolysis genes also indicated poor survival rates of NPC patients.

Conclusions
In summary, we demonstrated that FOXD1-AS1 is overexpressed in NPC cells and tissues, which is closely associated with the progression and poor prognosis in NPC patients. Furthermore, mechanistic research revealed that FOXD1-AS1 plays an oncogenic role in progression and promotes the glycolysis of NPC by sustaining the expression of FOXD1. The identi ed FOXD1-AS1 may serve as a potential  Review of the GEPIA database showed that FOXD1-AS1 is upregulated in multiple cancers. As seen in the GEPIA database (A) and the GSE95166 dataset (B), the expression level of FOXD1-AS1 is higher in tumor tissues than it is in normal tissues. (C) Further qRT-PCR testing con rms that FOXD1-AS1 was overexpressed in 10 pairs of NPC tissues and their adjacent normal tissues. Review of the GEPIA database showed that upregulation of FOXD1-AS1 is correlated with a shorter overall survival time (D) and disease-free survival time (E). (F) The transfection e ciency of FOXD1-AS1 shRNA and overexpression plasmid. Bars represent the standard deviation, ns P > 0.05, *P < 0.01, **P < 0.001, ***P < 0.001.  RNA pull-down assay suggesting that FOXD1 binds with FOXD1-AS1. (H) RIP assay to verify that FOXD1-AS1 binds to FOXD1. Bars represent standard deviation, ns P > 0.05, *P < 0.01, **P < 0.001, ***P < 0.001.

Figure 4
Page 22/27 FOXD1 promotes cell proliferation, migration, invasion, and induces apoptosis in NPC cells. (A) CNE-1 and 5-8F cells were transfected with FOXD1-target shRNA and FOXD1 scramble control shRNA (NC-F), respectively, and the expression level of FOXD1 was detected by qRT-PCR. (B) CNE-1 and 5-8F cells were transfected with FOXD1 overexpression plasmids or control (blank) plasmids (Ctr-F) for 48 hours, and qRT-PCR was used to analyze their e ciency. (C-D) MTT assay showing that FOXD1 knockdown inhibits cell growth, while FOXD1 overexpression accelerates cell growth in CNE-1 and 5-8F cells. (E) The effects of FOXD1 knockdown and the contribution of FOXD1-AS1 overexpression to the apoptosis of NPC cells were analyzed via ow cytometry assay. The migration (F, G) and invasion (H) abilities were signi cantly inhibited by FOXD1 silencing, whereas they were signi cantly increased by transfection of FOXD1 overexpression plasmids. Bars represent standard deviation, ns P > 0.05, *P < 0.01, **P < 0.001, ***P < 0.001. Figure 5 Upregulation of FOXD1 indicates a poor prognosis in patients with NPC and promotes the glycolysis of NPC cells. Using the GEPIA database (A), the Starbase database (B), and the GSE12452 dataset (C), we found that the expression level of FOXD1 is obviously overexpressed in tumorous tissues than in normal tissues. (D) Further qRT-PCR assay con rming that FOXD1 was upregulated in 10 pairs of NPC tissues than what was seen in adjacent normal tissues. (E) A higher expression level of FOXD1 was correlated Page 24/27 with a shorter overall survival time and disease-free survival time, as seen in the GEPIA database. (F) Overall survival and disease-free survival times for FOXD1 were analyzed via a Kaplan Meier-plotter database. (G) Gene Set Enrichment Analysis (GSEA) exhibiting that FOXD1 is signi cantly associated with hypoxia. (H) FOXD1 expression is positively correlated with glycolysis-related genes including LDHA, PKM and ENO1. (I) ChIP assay con rms that the enrichment of LDHA, PKM, and ENO1 promoters increases obviously after overexpression of FOXD1, but upregulation of FOXD1 has no effect on the enrichment of FOXD1-AS1 promoters. (J) FOXD1 knockdown induces the protein level of LDHA, PKM, and ENO1, while FOXD1 overexpression increases the expression level of these genes in CNE-1 and 5-8F cells.