This article is mentioned in:

Introduction

Melanoma is the most lethal skin cancer; it is responsible for the vast majority of cutaneous cancer deaths globally, even though it only accounts for ≤5% of all cutaneous carcinomas (1). In the early stages, melanoma can be treated with surgical resection; however, once metastasis occurs, it is resistant to conventional radio- and chemotherapy, and is extremely difficult to treat (2). Therefore, it is urgent to acquire an improved understanding of the properties of melanoma in order to develop effective treatment regimens.

Alterations in cellular metabolism have been recognized as hallmarks of malignant tumors (3,4). Aerobic glycolysis, also termed the Warburg effect, is one of the most important hallmarks in the reprogramming of cancer metabolism via upregulated glycolytic enzymes and activated regulatory factors, including oncogenes (p53, c-Myc and K-Ras), essential signaling pathways [PI3K/Akt, liver kinase B1/AMP kinase and hypoxia-inducible factor 1 (HIF1) signaling pathways] and epigenetic regulators such as sirtuins (SIRTs) (5-8). Importantly, aerobic glycolysis has been demonstrated to provide a specific microenvironment to promote unconstrained proliferation and invasion (9). Therefore, controlling cellular metabolism may be a potential targeted therapeutic strategy to treat malignant tumors such as melanoma.

Forkhead box O (FOXO) transcription factors, including FOXO1, FOXO3a, FOXO4 and FOXO6, which are conserved from Caenorhabditis elegans to mammals, serve pivotal roles in multiple cellular processes, such as cell cycle progression, apoptotic cell death, DNA repair, oxidative stress, epithelial-mesenchymal transition and cellular metabolism (10-13). FOXO3a, an important member of the FOXO family, participates into the modulation of cell growth in multiple tumors, including glioblastoma (14), prostate cancer (15), lung adeno-carcinoma (16), ovarian cancer (17), colorectal cancer (18) and Hodgkin's lymphoma (19). It was reported that FOXO3a is also an important regulator of cellular metabolism in tumors; for example, FOXO3a regulates reactive oxygen metabolism by inhibiting mitochondrial gene expression in colon cancer (20). Additionally, FOXO3a has been shown to regulate multiple cellular process, including cell survival, apoptosis (21-23), migration and invasion (24) in melanoma. However, the role of FOXO3a in the regulation of cellular metabolism in melanoma has never been explored.

The present study aimed to elucidate the role of the FOXO3a-SIRT6 axis in the interplay between cellular metabolism and tumor progression, thereby providing novel insight into potential melanoma treatment strategies. In the present study, it was observed that FOXO3a inhibited aerobic glycolysis by targeting the promoter of SIRT6 and promoting its transcription, thereby inhibiting the expression of a cluster of glycolysis-associated genes.

Materials and methods

Cell lines and reagents

The MV3 melanoma cell line was obtained from the Third Military Medical University, and cultured in RPMI-1640 (Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin (P/S; Invitrogen; Thermo Fisher Scientific, Inc.). PIG1 normal melanocytes, and SK-MEL-28 and A375 melanoma cell lines were purchased from the American Type Culture Collection (ATCC) and cultured in Dulbecco's Modified Eagle's minimum essential medium (DMEM, Thermo Fisher Scientific, Inc.) supplemented with 10% FBS and 1% P/S. All cells were cultured at 37°C in a 5% CO2 incubator (Sanyo). 2-Deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4-yl) amino]-D-glucose (2-NBDG; cat. no. N13195) was purchased from BD Biosciences. MTT (cat. no. M2128) and DMSO (cat. no. D2650) were purchased from Sigma-Aldrich (Merck KGaA).

Reverse transcription-quantitative PCR (RT-qPCR)

RNA was extracted from cells following specific treatments using RNAiso Plus (Takara Bio, Inc.), trichloromethane (Sigma-Aldrich; Merck KGaA), isopropanol (Shanghai Dingguo Biological Technology Co., Ltd.) and 75% ethanol (Shanghai Dingguo Biological Technology Co., Ltd.) according to the manufacturer's protocol. cDNA was obtained from 2 µg RNA/sample using a GoScript™ Reverse Transcriptase kit (cat. no. A5001; Promega Corporation) according to the manufacturer's protocols. Then, RT-qPCR was performed to analyze the mRNA expression of genes using a LightCycler® 96 Instrument (Roche Diagnostics). Promega GoTaq® qPCR Master Mix (cat. no. A6001; Promega Corporation) was used. The PCR reaction conditions were as follows: 95°C pre- denaturation for 10 min; then, 45 cycles of 95°C for 15 sec, 60°C for 30 sec and 72°C for 30 sec; then, 95°C for 10 sec, 60°C for 1 min, 97°C for 1 sec and 37°C for 30 sec. Results were calculated via the 2-∆∆Cq method (25) with ACTB expression used as the internal control (Cq value was used instead of Ct value in this study). The primers, which were also used in a previous study (14), were presented in Table I.

Table I Primers used for reverse transcription-quantitative PCR.

Table I

Primers used for reverse transcription-quantitative PCR.

Gene	Sequence (5′-3′)
FOXO3A	F: ACGTCTTCAGGTCCTCCTGTT
	R: GGGGAAGCACCAAAGAAGAGAG
SIRT6	F: CTCGAAGTGGAGCTGGACC
	R: TCCTCGGGGATCATGGAGTC
GLUT1	F: TGTGTATGCCACCATTGGCT
	R: CTAGCGCGATGGTCATGAGT
GLUT4	F: GGACAGCCAGCCTACGCCACCATA
	R: GGACAGCCAGCCTACGCCACCATA
HK1	F: GCACGTTTGCACCATTGTCT
	R: TTGTGGAAACGCCGGGAATA
HK2	F: GAATGGGAAGTGGGGTGGAG
	R: GAGGAGGATGCTCTCGTCCA
HK3	F: TTCCCATGTAGGCAGCTTGG
	R: ATGAGGCCTATCTCGCAACG
GAPDH	F: CTCTGCTCCTCCTGTTCGAC
	R: GCGCCCAATACGACCAAATC
PFK1	F: CTGCCCCTCATGGAATGTGT
	R: ATACCGGGGGTCTGACATGA
PKM2	F: AATGCAGTCCTGGATGGAGC
	R: ACTGCAGCACTTGAAGGAGG
LDHA1	F: GGTCCTTGGGGAACATGGAG
	R: TAGCCCAGGATGTGTAGCCT
LDHA2	F: AGCTGTTCCACTTAAGGCCC
	R: AGGAATCGGGAATGCACGTC
ACTB	F: CGTCTTCCCCTCCATCGTG
	R: TCGATGGGGTACTTCAGGGT

[i] F, forward; R, reverse; FOXO3a, forkhead box O3; SIRT6, sirtuin 6; GLUT1/4, glucose transporter 1/4; HK1-3, hexokinase 1-3; PFK1, phosphofructokinase 1; PKM2, pyruvate kinase isozyme M2; LDHA1/2, lactate dehydrogenase A1/A2; ACTB, β-actin.

Vector construction and stable transfection

Short hairpin RNA (shRNA) sequences were designed using siRNAext (http://sirna.wi.mit.edu/), and then synthesized by BGI and cloned into a lentiviral pLKO.1 vector (cat. no. 10878; Addgene, Inc.). The sequences in FOXO3a and SIRT6 targeted by the shRNAs were presented in Table II. Human full-length SIRT6 (GenBank no. CR457200.1) cDNA was from MV3 cells via PCR; PrimeSTAR® Max DNA Polymerase (Takara Bio, Inc.) was used. Thermocycling conditions were as follows: 98°C pre-denaturation for 5 min; then, 28 cycles of 98°C for 30 sec, 60°C for 30 sec and 72°C for 20 sec; then, 72°C for 10 sec. The products were constructed into a lentiviral pCDH-CMV-MCS-EF1-Puro vector (cat. no. CD510B; System Biosciences, LLC). The primers were listed in Table II. HA-FOXO3a WT plasmid (cat. no. 1787; Addgene, Inc.) was obtained from Addgene and then cloned into the pCDH-CMV-MCS-EF1-Puro vector. Plasmids were packaged into lentivirus as previously described (26). Briefly, 293FT cells (ATCC) were cultured in DMEM (Thermo Fisher Scientific, Inc.) with 10% FBS, 1% P/S and 0.5 mg/ml gene-ticin (Thermo Fisher Scientific, Inc.), which was replaced with lentiviral culture medium prior to transfection with plasmids, which was comprised of DMEM, 10% FBS, 2 mM L-glutamine (Invitrogen; Thermo Fisher Scientific, Inc.), 0.1 mM non-essential amino acid (Invitrogen; Thermo Fisher Scientific, Inc.) and 1 mM sodium pyruvate (Invitrogen; Thermo Fisher Scientific, Inc.). 293FT cells at 100% confluence were transfected with 0.625 µg of the plasmid of interest, plus the packaging plasmids pLP1, pLP2 and pLP/VSVG (Nova Lifetech, Inc.), using Opti-MEM™ medium (Gibco; Thermo Fisher Scientific, Inc.) and Lipofectamine™ 2000 (Thermo Fisher Scientific, Inc.). At 3 days later, the viral supernatant was aspirated with a syringe, filtered through a 0.45-µm filter membrane and collected in a 1.5-ml centrifuge tube. Fresh lentivirus culture medium was added to the 293FT cell culture wells, which were cultured for a further 48 h before collecting the second viral supernatant. Then, 40,000 MV3 cells in a 60-mm dish were infected with 2 ml lentivirus containing >107 TU/ml using 4 µg/ml polybrene (Sigma-Aldrich; Merck KGaA) and incubated at 37°C with 5% CO2 for 48 h. Then, the cells were re-seeded into a 100-mm petri dish and cultured in standard medium. After 3 days, 2 µg/ml puromycin was used to continuously screen the cells for ≥72 h. RT-qPCR or western blotting was performed to verify the expression of the target genes.

Table II Primers or targeted sequences for vector construction.

Table II

Primers or targeted sequences for vector construction.

Primer/target site	Sequence (5′-3′)
shFOXO3a#1	AATGTGACATGGAGTCCATTAT
shFOXO3a#2	GGACAATAGCAACAAGTATACC
shSIRT6	AAGAATGTGCCAAGTGTAAGA
Scramble	ATCCGTCCGAACGTAAGTCAA
SIRT6	Forward (EcoRI): CCGGAATTCATGTCGGTGAATTACGCGGCGGC Reverse (BamHI): CGCGGATCCTT AACTGGGGACCGCCTTGG

[i] Underlined regions indicate recognition sites for the specified restriction enzyme. sh, short hairpin RNA; SIRT6, sirtuin 6; FOXO3a, forkhead box O3.

Western blot assay

Western blotting was conducted to analyze the expression of proteins. Briefly, cells at 80% confluence were trypsinized and collected in 5-ml tubes. Then, cells were centrifuged at 600 x g for 5 min at 4°C, the supernatant was removed, and the cell pellet was washed three times with PBS. Protein was extracted from the cell pellet using RIPA lysis buffer (cat. no. P0013B; Beyotime Institute of Biotechnology), and the mixture was allowed to stand on ice for 30 min. The mixture was centrifuged at 12,000 x g at 4°C for 15 min, and the protein concentration was determined by using an Enhanced BCA Protein Assay Kit (cat. no. P0009; Beyotime Institute of Biotechnology). Protein (100 µg/lane) was separated via 10% SDS-PAGE. Then, protein was transferred to PVDF membranes (cat. no. IPVH00010; Merck KGaA). Membranes were blocked with 5% BSA (Fraction V; cat. no. ST023; Beyotime Institute of Biotechnology) at room temperature for 2 h. Primary antibodies, including rabbit anti-FOXO3a (1:1,000; cat. no. 2497; Cell Signaling Technology, Inc.), rabbit anti-SIRT6 (1:800; cat. no. 12486; Cell Signaling Technology, Inc.) and anti-α-Tubulin Antibody (1:200; cat. no. 2144; Cell Signaling Technology, Inc.) were incubated at 4°C overnight. Then, horseradish peroxidase-conjugated goat anti-mouse (1:20,000; cat. no. ab205719; Abcam) or goat anti-rabbit IgG (1:20,000; cat. no. ab205718; Abcam) was incubated at room temperature for 2 h. ChemiSignal™ Plus ECL (Clinx Science Instruments Co., Ltd.) was used to visualize bands, which were imaged using a GenoSens 2000 Touch gel imaging system (Clinx Science Instruments Co., Ltd.).

MTT assay

Cells (1,000/well) were cultured in 96-well plates at 37°C in a CO2 incubator, and MTT assays were performed as previously described (27) at indicated times (0, 2, 4 and 6 days).

Glucose uptake assay

Cells (2×105/well) were cultured at 37°C in a CO2 incubator in glucose-free RPMI-1640 (Procell Life Science & Technology Co., Ltd.) with FBS and P/S in 6-well plates for 120 min, and the medium was removed. A fluorescent glucose analogue, 2-NBDG (100 µM), was dissolved in Kerbs-Ringer bicarbonate (KRB) buffer (129 mM NaCl, 4.8 mM KCl, 5 mM NaHCO3, 1.2 mM MgSO4, 2 mM CaCl2 and 10 mM HEPES) and added to the plates prior to incubation at 37°C for a further 120 min. The cells were collected using trypsin and washed with KRB buffer, and the fluorescence of 2-NBDG in the cells was detected via flow cytometry (Acurri C6; BD Biosciences) and analyzed by using FlowJo 7.6.1 (FlowJo LLC).

Glucose consumption, lactate and lactate dehydrogenase (LDH) assays

Cells (2×105/well) were cultured in 6-well plates at 37°C for 48 h. The glucose content in the medium was detected using a Glucose Assay kit (cat. no. GAGO20; Sigma-Aldrich; Merck KGaA), the lactate content in the medium was detected using a Lactate Assay kit (cat. no. MAK064; Sigma-Aldrich; Merck KGaA) and the lactate dehydrogenase activity of cells was detected using a lactate dehydrogenase assay kit (cat. no. MAK066; Sigma-Aldrich; Merck KGaA), all according to the manufacturer's protocols. Samples were analyzed using a SYNERGY HTX multi-mode reader (Biotek Instruments, Inc.). The rates of glucose consumption, lactate production and relative LDH activity were calculated according to the standard curve line and OD value of each sample. Then, these values were also normalized by cell numbers determined by using a blood cell counting chamber.

Glucose stress flux test

A glucose stress flux test was conducted as previously reported (14). In brief, 40,000 cells were seeded into XF96 cell culture microplates and cultured at 37°C for 24 h. The medium was then replaced with Seahorse XF DMEM (Agilent Technologies, Inc.) containing 2 µM glutamine, and the microplates were maintained in a non-CO2 incubator at 37°C for 60 min. Then, a Seahorse XF glycolytic stress test was performed using a Seahorse XFp analyzer (Agilent Technologies, Inc.). In this test, a final concentration of 10 µM glucose (Sangon Biotech Co., Ltd.), 1 µg/ml oligomycin (cat. no. 495455; Sigma-Aldrich; Merck KGaA) and 50 µM 2-deoxyglucose (Sangon Biotech Co., Ltd.) were used.

Chromatin immunoprecipitation (ChIP) assay

ChIP assays were performed in MV3 cells by using the EZ CHIP™ kit (cat. no. 17-371; Merck KGaA) according to the manufacturer's protocol. 293FT cells were cultivated in a 100-mm dish; when the cells reached 80% confluence, 1% fresh formaldehyde (Sangon Biotech Co., Ltd.) was added and the cells were cultured in a 37°C incubator for 10 min to crosslink protein and DNA. The cells were washed with PBS buffer containing 1 mM phenylmethylsulfonylfluoride (PMSF; cat. no. ST505; Beyotime Institute of Biotechnology), and then collected in a 1.5-ml tube for centrifugation at 1,000 x g for 2 min at 4°C. The centrifuged cell pellet was resuspended in 200 µl SDS lysis buffer (cat. no. P0013G; Beyotime Institute of Biotechnology) containing 1 mM PMSF on ice for 10 min, following which it was subjected to ultrasonic vibration with 15 sec ON, 30 sec OFF for 8 cycles on ice to break the genomic DNA into <1,000-bp fragments. The sample was then resuspended in sodium chloride, incubated at 65°C for 4 h, and then mixed with Tris-balanced phenol and centrifuged at 12,000 x g at 4°C for 5 min. Finally, 200 µl chloroform was added to the pellet, which was then centrifugated at 4°C at 12,000 x g for 5 min.

The resulting supernatant was moved into an ice-cold centrifuge tube, and ChIP dilution buffer containing 1 mM PMSF was added to a final volume of 2 ml, of which 20 µl was collected to use as the input control. Then, 70 µl of Protein A + G Agarose (containing salmon sperm DNA) was added to the remaining sample, which was incubated slowly at 4°C on a shaker for 30 min. The solution was then centrifuged at 1,000 x g for 1 min at 4°C. and the supernatant was collected. Then, 2 µg FOXO3a primary antibody (cat. no. ab12162; Abcam) or rabbit IgG (cat. no. A7016, Beyotime Institute of Biotechnology) as a blank control was added, along with 60 µl Protein A + G Agarose containing salmon sperm DNA; the mixture was incubated slowly at 4°C on a shaker for 60 min and subsequently centrifuged at 1,000 x g for 1 min at 4°C. The supernatant was removed, and the pellet was centrifuged once at 1,000 x g for 1 min at 4°C with 1 ml Low Salt Immune Complex Wash Buffer, High Salt Immune Complex Wash Buffer and LiCl Immune Complex Wash Buffer, followed by centrifugation under the same conditions with 1 ml TE Buffer twice. Elution Buffer was subsequently added prior to centrifugation at 1,000 x g for 1 min at 4°C, following which this step was repeated. The resulting supernatant was collected, the sample was recovered and concentrated using a AxyPrep DNA Gel Extraction kit (Axygen Bioscience, Inc.), and then subjected to qPCR analysis using the primers in Table III.

Table III SIRT6 primers used for chromatin immunoprecipitation and luciferase assays.

Table III

SIRT6 primers used for chromatin immunoprecipitation and luciferase assays.

Primer	Sequence (5′-3′)
SIRT6-p	F1: AAGACAATCCGTGGGCTTGG
	R1: GAGCTACCCAGGTACCCTG F2: TGGCTAGGACTCAGCACG R2:
	TAGGGGAGGAAGGAGGTGG
SIRT6-p-0.1k	F (NheI): CCGGCTAGCGCCCGGCTCACTCACTTTTTAG
SIRT6-p-0.2k	F (NheI): CCGGCTAGCCTGCCTTGGCCTCCCAAAGT
SIRT6-p-0.6k	F (NheI): CCGGCTAGCCTATCATCACTGGACTGATTTCAGTTTC
SIRT6-p-1.2k	F (NheI): CCGGCTAGCGGGTAATAAGACACCCAACAGAGG
SIRT6-p- (all)	R (XhoI): CCGCTCGAGGTAATGGTGACATGGTGTGGTTG
SIRT6-p-N0.9k	F (NheI): CCGGCTAGCCTGGTCACATGTTTGTGTCCAC
SIRT6-p-N0.9k	R (XhoI): CCGCTCGAGAAAGTTTCCCTTGTTGAGGCCG

[i] Underlined regions indicate recognition sites for the specified restriction enzyme. SIRT6, sirtuin 6; F, forward; R, reverse.

Dual-luciferase assay and promoter analysis

To analyze the similarity of the promotor region of human and mouse SIRT6, the promoter region (-2,500 to 0 bp) of mouse SIRT6 and the promoter region (−2,500 to 0 bp) of human SIRT6 were compared using the online Bl2seq tool in the SilkDB (http://www.silkdb.org/silkdb/). The sequences in the promoter regions, which were termed 0.1 k (−1,100 to −939 bp), 0.2 k (−1,147 to −939 bp), 0.6 k (−1,538 to −939 bp), 1.2 k (−2,152 to −939 bp) and N0.9 k (−938 to −25 bp), were obtained from genomic DNA extracted from MV3 cells via PCR by using primers in Table III. The mutant (Mut) sequence of SIRT6 promoter region was synthesized by Sangon Biotech Co., Ltd. These sequences were cloned into pGL3-basic luciferase reporter vectors (cat. no. E1751; Promega Corporation). Then, 1 µg pGL3 vector and 1 µg pRL-TK expressing Renilla luciferase (Youbio, Inc.) were co-transfected into 20,000 MV3 cells/well in a 24-well plate with X-tremeGENE™ HP DNA Transfection reagent (cat. no. 6366546001; Roche Diagnostics), and the luciferase assay was performed as previously described using a Dual-Lumi™ Luciferase Reporter Gene assay kit (cat. no. RG088S; Beyotime Institute of Biotechnology) (28). The promotor activity was normalized to Renilla luciferase activity.

The profile of FOXO3a in Mus musculus and Homo sapiens was downloaded from the JASPER database (version 5.0_ ALPHA; http://jaspar.binf.ku.dk/) and then compared with the promoter region of human SIRT6 to find the potential binding site of FOXO3a.

Tumor xenografts

The animal experiments in the current study were approved and supervised by the Institutional Animal Care and Use Committees of the Southwest University (permit no. IACUC-20190402-02) and the Experimental Animal Care and Use Committees of the Institute of Sericulture and Systems Biology. The study was performed according to the Laboratory Animal Management Regulations and the Measures of Chongqing Municipality on the Management of Experimental Animals. A total of nine 4-week-old female mice (weight, 18-20 g; BALB/c-nu; Beijing Huafukang Bioscience Co. Ltd.) were purchased and housed in a specific pathogen-free room to acclimate for ~1 week. The animals were house at 22°C with 40-60% humidity under a 12:12-h light/dark cycle. Mice were provided ad libitum access to food and water. Then, MV3 cells (1×106) in 100 µl PBS were subcutaneously injected into the left flank of mice. Every group contained ≥3 mice. Then, 10 days later, the first measurements of the length and width of tumors were made by caliper, and tumor growth was measured for 25 days after this point. The tumor volume was calculated with the following formula: Volume = tumor length x width2 x π/6. At the end of the experiment, animals was sacrificed with CO2 in a 10-l volume chamber with a flow rate of 2 l/min and a displacement rate of 20% volume/min, and then tumors were removed and weighed. The maximum tumor diameter observed was 1.37 cm.

Bioinformatics analysis

The Cancer Genome Atlas (TCGA) Ocular melanoma (project no. TCGA-UVM) dataset was downloaded from University of California Santa Cruz Xena (http://xena.ucsc.edu/). Other clinical databases, including Tumor Melanoma Metastatic Bhardwaj-44-MAS5.0-u133p2, Mixed Melanoma Briggs-70-MAS5.0-u133a, Mixed Melanoma (Metastasis) Hynes-83-MAS5.0-u133a, Exp Cellline Melanoma-Exosome McMasters-8-MAS5.0-u133p2, Exp Melanoma Augustine-50-MAS5.0-u133p2, Tumor Melanoma Jönsson-214-custom-ilmnht12v4 and Tumor Melanoma (Metastatic) Matta-87-MAS5.0-u133p2 were downloaded from the public R2 platform (https://hgserver1.amc.nl/). Genes in certain datasets were further analyzed by using alternative probes provided by the databases. All data were analyzed by using the software GraphPad Prism 6 (GraphPad Software, Inc.).

Statistical analysis

All the experiments were repeated 3 times and the data collected were analyzed by using GraphPad Prism 6. Data were presented as the mean ± SD. Unpaired two-tailed Student's t-test was applied to determine significant differences between two groups. The scan cutoff modus was used to separate high- and low-expression groups for Kaplan-Meier analysis, with the exception of the analysis of data from the Cancer Genome Atlas (TCGA) Ocular melanoma (project no. TCGA-UVM) dataset, for which the median cutoff modus was used. Log-rank (Mentel-Cox) tests were conducted to determine significance for survival analysis. A Bonferroni correction was applied after the log-rank test to control for multiple comparisons [P<0.00833 (0.05/6) was considered to indicate a statistically significant different for this analysis]. Pearson correlation coefficient was used to analyze the correlation of the expression levels of 2 genes. One-way ANOVA followed by Dunnett's test was performed to compare the mean of each experiment group with the control group in datasets containing multiple comparisons. One-way ANOVA followed by Tukey's test was performed to compare the mean of each group with the mean of every other group when performing multiple comparisons. P<0.05 was considered to indicate a statistically significant difference.

Results

High FOXO3A expression predicts improved prognosis of patients with melanoma

To elucidate the relationship between FOXO3A mRNA expression and the prognosis of patients with melanoma, its expression was analyzed in the database termed Tumor Melanoma Metastatic Bhardwaj-44-MAS5.0-u133p2 from the R2 platform. The results showed that high FOXO3A expression predicted improved overall survival and metastasis-free survival in this cohort (Fig. 1A and B). Furthermore, FOXO3A expression was negatively associated with the survival rate of patients with metastatic melanoma (Figs. 1C and S1A-D). Additionally, FOXO3A expression was lower in patients with stage IV melanoma than those with stage III melanoma (Figs. 1D, and S1E and F). Importantly, FOXO3A expression was lower in nevus compared to normal tissues, and FOXO3A expression was further decreased in melanoma compared with nevus (Fig. 1E). Next, in the database termed Mixed Melanoma (Metastasis) Hynes-83-MAS5.0-u133a, it was found that FOXO3A expression was lower in melanoma metastasis compared with the primary tumors (Fig. 1F). In an experimental database termed Exp Cellline Melanoma-Exosome McMasters-8-MAS5.0-u133p2, it was demonstrated that FOXO3A expression was lower in the exosomes of A375 melanoma cells compared to those of HeMa-LP normal melanocytes (Fig. 1G). N-Ras mutations arise in 15-20% of all melanomas, and have been shown to be associated with aggressive clinical behavior and poor prognosis (29). Of note, FOXO3A expression was lower in N-Ras mutant melanoma cells compared with wild-type cells (Fig. 1H). These results implied that FOXO3a may act as a tumor suppressor in melanoma.

Figure 1 High FOXO3A expression predicts improved prognosis for patients with melanoma. (A and B) Relationship between FOXO3A expression, and overall and metastasis-free survival in the database termed Tumor Melanoma Metastatic Bhardwaj-44-MAS5.0-u133p2 from the R2 platform. (C) Relationship between FOXO3A expression and the survival rate following metastasis in the database termed Tumor Melanoma Metastatic Bhardwaj-44-MAS5.0-u133p2. (D) Expression of FOXO3A in stage III and IV melanoma in the Tumor Melanoma Metastatic Bhardwaj-44-MAS5.0-u133p2 database. (E) Expression of FOXO3A in normal tissues, nevus and melanoma in the database termed Mixed Melanoma Briggs-70-MAS5.0-u133a. (F) Expression of FOXO3A in primary melanoma and melanoma metastases in Mixed Melanoma (Metastasis) Hynes-83-MAS5.0-u133a database. (G) Expression of FOXO3A in the exosomes of MeLa-LP normal melanocytes and A375 melanoma cells in the database termed Exp Cellline Melanoma-Exosome McMasters-8-MAS5.0-u133p2. (H) Expression of FOXO3A in WT melanoma cells and N-Ras MT melanoma cells in the database termed Exp Melanoma Augustine-50-MAS5.0-u133p2. FOXO3A, forkhead box O3; WT, wild-type; MT, mutant.

FOXO3A transcriptionally promotes the expression of SIRT6 in melanoma

It was previously shown that the FOXO3a genotype was strongly associated with human longevity (30). As a deacetylase, SIRT6 also was shown to be related to human longevity (31). Additionally, FOXO3a and SIRT6 are both regulators of hepatic sterol regulatory element-binding protein 2 and cholesterol biosynthesis, as well as low-density lipoprotein-cholesterol homeostasis (32,33). These findings suggested that FOXO3a and SIRT6 are highly associated; however, their relationship had not been elucidated. In the present study, it was found that FOXO3A mRNA expression was positively correlated with SIRT6 mRNA expression (detected by probe 219613_s_at or 233179_x_at) in a melanoma cohort (Fig. 2A and B). Additionally, mRNA expression levels of FOXO3A and SIRT6 were correlated with each other in PIG1 melanocytes and 3 melanoma cell lines (Fig. 2C-E). Furthermore, FOXO3a expression was silenced in MV3 melanoma cells via virus-mediated transfection, and the results showed that both the mRNA and protein levels of SIRT6 were downregulated after FOXO3a silencing (Fig. 2F and G). Consistently, FOXO3a overexpression also induced upregulation of SIRT6 expression in MV3 cells (Fig. 2H and I).

Figure 2 FOXO3a transcriptionally regulates the expression of SIRT6 in melanoma. (A and B) Correlation of FOXO3A expression and SIRT6 expression in a melanoma cohort detected using different SIRT6 probes. (C) Western blotting was performed to detect the protein expression of FOXO3a and SIRT6 in PIG1 normal melanocytes and 3 different melanoma cell lines. (D) RT-qPCR was used to detect the mRNA expression of FOXO3A and SIRT6 in 4 different cell lines. One-way ANOVA followed by Dunnett's test was performed to compare the mean of melanoma cell group with PIG1 melanocytes. *P<0.05, **P<0.01, ***P<0.001 vs. PIG1. (E) Correlation of FOXO3A mRNA expression and SIRT6 expression in 4 different cell lines. (F) Expression of FOXO3a and SIRT6 as detected via western blotting in MV3 cells after FOXO3a silencing. (G) Relative expression of FOXO3A and SIRT6 detected via RT-qPCR in MV3 cells after FOXO3a silencing. *P<0.05, **P<0.01 vs. Scramble. (H) Expression of FOXO3a and SIRT6 detected via western blotting in MV3 cells after FOXO3a overexpression. (I) Relative expression of FOXO3A and SIRT6 detected via RT-qPCR in MV3 cells after FOXO3a overexpression. **P<0.01 vs. Vector. (J) Overview of insert fragments in pGL3 vectors used in dual-luciferase assays and ChIP primers designed based on the promoter of the human SIRT6 gene. Firefly luciferase/Renilla luciferase (F-luc/R-luc) was shown. SIRT6 promoter (different regions) activity was detected by dual-luciferase assays in MV3 cells. One-way ANOVA followed by Tukey's test was performed to compare groups. ***P<0.001. (K and L) CHIP RT-qPCR assay was used to detect the enrichment of FOXO3a in the promoter region of SIRT6 in MV3 cells. **P<0.01 vs. IgG. n.s., not significant. (M) Logo for FOXO3a in Mus musculus and Homo sapiens was downloaded from the JASPER website. (N) Prediction of FOXO3a binding site in the −2,083 to −1,859 bp region of the SIRT6 promoter. Then, WT and Mut versions of this sequence were inserted into luciferase vectors. (O) SIRT6 promoter (−2,128 to −914, −2,083 to −1,859 WT and −2,083 to −1,859 Mut) activity was detected by dual-luciferase assay in MV3 cells. ***P<0.001 vs. −2128 to −914. FOXO3a, forkhead box O3; SIRT6, sirtuin 6; RT-qPCR, reverse transcription-quantitative PCR; ChIP, chromatin immunoprecipitation; F-luc, firefly luciferase; R-luc, Renilla luciferase; sh, short hairpin (RNA); WT, wild-type; Mut, mutant

A previous study reported that FOXO3a regulated the transcription of SIRT6 by binding and activating nuclear respiratory factor 1 (NRF1) in the mouse (34). However, the promoter of human SIRT6 was not similar with the promoter in mouse SIRT6. No NRF1-binding sites (5′AGG GCG CAT GCG CCC TC3′) were identified in the promoter regions (−2,500 to 0 bp) of human SIRT6, implying that FOXO3a may regulate the transcription of SIRT6 through another mode of action (data not shown). The promoter region (−2,500 to 0 bp) of mouse SIRT6 and the promoter region (−2,500 to 0 bp) of human SIRT6 were analyzed by using the online Bl2seq tool in the SilkDB. The results showed that there were several similar DNA sequences in both promoters (Fig. 2J). It was hypothesized that the binding sites of FOXO3a in the promoter region of SIRT6 may be conserved in mammals. Therefore, these DNA sequences may be candidate binding sites. Therefore, several regions (0.1, 0.2, 0.6, 1.2 and N0.9 k) were cloned from the promoter of human SIRT6 and constructed into pGL3 vectors (Fig. 2J). A dual-luciferase assay showed that only the −2,128 to −1,514 bp region exhibited significant activity compared with other regions (Fig. 2J). As there were four candidate binding sites for FOXO3a in this region, ChIP RT-qPCR assays were used to detect the precise binding site. The results showed that FOXO3a exhibited significant enrichment in the −2,083 to −1,859 bp region, with no enrichment in the −1,852 to −1,746 bp region (Fig. 2K and L). Then, the −2,083 to −1,859 bp region was analyzed in the JASPER website. The profile of FOXO3a in Mus musculus and Homo sapiens (Fig. 2M) was used to find the specific binding site in the −2,083 to −1,859 bp region of the SIRT6 promoter. It was shown that there was a predicted site sequence (5′GGTAAATA3′) that was highly similar to the FOXO3a binding profile (Fig. 2N). Then, wild-type (WT) and Mut sequences of this region were synthesized and cloned into a pGL3 vector (Fig. 2N). A luciferase activity assay revealed that −2,083 to −1,859 WT showed a similar level of promoter activity as the −2,128 to −914 bp region, whereas −2,083 to −1,859 Mut significantly decreased promoter activity (Fig. 2O). These results indicated that FOXO3a regulated SIRT6 expression in human melanoma cells via a transcriptional manner that is distinct from that in the mouse.

FOXO3a is negatively correlated with aerobic glycolytic genes in melanoma cohorts

SIRT6 is a major regulator of aerobic glycolysis (26), which is an important cause of tumor progression. Whether FOXO3a also contributed to the aerobic glycolysis in melanoma was explored. By analyzing data from clinical databases, the results showed that FOXO3A mRNA expression was negatively correlated with the expression of a cluster of genes that participate in aerobic glycolysis, such as hexokinase 1 (HK1), HK3, phosphofructokinase (PFK) muscle, PFK fructobiphosphatase 3, pyruvate kinase isozyme (PKM) and LDHA (Figs. 3A-F and S2A-O). Notably, the majority of these genes are also targets of SIRT6 (35). These results implied that FOXO3a-SIRT6 may be a major regulator controlling the expression of these glycolytic genes in melanoma.

Figure 3 FOXO3a is negatively correlated with aerobic glycolysis-associated genes in melanoma cohorts. Correlation of FOXO3A expression and the expression of glycolysis genes, including (A) HK1, (B) HK3, (C) PFKM, (D) PFKFB3, (E) PKM and (F) LDHA in melanoma cohorts. HK, hexokinase; PFK, phosphofructokinase; PFKM, PFK muscle; PFKFB3, PFK fructobiphosphatase 3; PKM, pyruvate kinase isozyme; LDHA, lactate dehydrogenase A.

SIRT6 overexpression rescues FOXO3a deficiency-induced upregulation of aerobic glycolysis

To validate the hypothesis that FOXO3a-SIRT may regulate glycolysis in melanoma cells, SIRT6 was overexpressed using a SIRT6 overexpression vector (Fig. 4A and B) in FOXO3a-silenced MV3 cells (Fig. 4C and D). The results revealed that FOXO3a silencing induced upregulation of a cluster of glycolytic genes, including glucose transporter 4 (GLUT4), GLUT1, HK1, HK2, HK3, GAPDH, PFK1, PKM2, lactate dehydrogenase A1 (LDHA1) and LDHA2 (Fig. 4D). However, this effect was rescued by overexpression of SIRT6 (Fig. 4D). Additionally, FOXO3a silencing markedly promoted glucose uptake in MV3 cells (Fig. 4E). Consistently, FOXO3a-induced glucose uptake was also recovered by SIRT6 overexpression (Fig. 4E). Then, the glucose consumption, lactate production and LDH activity of MV3 cells were detected. The results showed that FOXO3a silencing upregulated aerobic glycolysis, whereas SIRT6 overexpression rescued this effect (Fig. 4F-H). The glycolytic flux test was performed using a Seahorse XFp analyzer in FOXO3a-silenced MV3 cells after SIRT6 restoration, with the results showing that glycolytic flux was notably increased after FOXO3a silencing, whereas SIRT6 overexpression returned glycolytic flux to the levels of the control group (Fig. 4I). These results indicated that the FOXO3a-SIRT6 axis serves an important role in controlling the aerobic glycolysis of melanoma cells.

Figure 4 SIRT6 overexpression rescues FOXO3a deficiency-induced upregulation of aerobic glycolysis. SIRT6 overexpression was confirmed via (A) western blot and (B) RT-qPCR analyses in MV3 cells infected with SIRT6 overexpression vector. (C) Western blotting was used to detect the protein expression of FOXO3a and SIRT6 in FOXO3a-silenced MV3 cells after SIRT6 restoration. (D) Relative expression of SIRT6 target glycolysis-associated genes as determined via RT-qPCR in FOXO3a-silenced MV3 cells following SIRT6 restoration. (E) Glucose uptake detected by flow cytometry in FOXO3a-silenced MV3 cells after SIRT6 restoration. (F) Glucose consumption detected in FOXO3a-silenced MV3 cells after SIRT6 restoration. (G) Lactate production detected using a lactate assay kit in FOXO3a-silenced MV3 cells after SIRT6 restoration. (H) LDH activity detected using an LDH assay kit in FOXO3a-silenced MV3 cells after SIRT6 restoration. (I) Glycolytic stress flux test was conducted by using a Seahorse XF analyzer in FOXO3a-silenced MV3 cells after SIRT6 restoration. *P<0.05, **P<0.01, ***P<0.001. FOXO3a, forkhead box O3; SIRT6, sirtuin 6; RT-qPCR, reverse transcription-quantitative PCR; HK, hexokinase; GLUT, glucose transporter; PFK1, phosphofructokinase 1; PKM2, pyruvate kinase isozyme 2; LDHA, lactate dehydrogenase A; LDH, lactate dehydrogenase; sh, short hairpin (RNA); 2-NBDG, 2-Deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]-D-glucose; ECAR, extracellular acidification rate; 2-DG, 2-deoxy-D-glucose.

SIRT6 downregulation retrieves FOXO3a overexpression-induced decrease of aerobic glycolysis

To further validate these results, SIRT6 was knocked down using a SIRT6 shRNA (Fig. 5A and B) in FOXO3a-overexpressing MV3 cells (Fig 5C and D). The results showed that SIRT6 silencing rescued the FOXO3a overexpression-induced downregulation of glycolytic genes, decrease in glucose uptake, decline in glucose consumption, reduction of lactate production, attenuation of LDH activity and upregulation of glycolytic flux (Fig. 5E-I). These results further indicated that the FOXO3a-SIRT6 axis was a major regulator of cellular metabolism in melanoma cells.

Figure 5 SIRT6 downregulation rescues FOXO3a overexpression-induced downregulation of aerobic glycolysis. SIRT6 knockdown was confirmed via (A) western blotting and (B) RT-qPCR in MV3 cells infected with shSIRT6. (C) Western blotting was used to detect the protein expression of FOXO3a and SIRT6 in FOXO3a-overexpressing MV3 cells after SIRT6 silencing. (D) Relative expression of SIRT6 target glycolysis-associated genes detected by RT-qPCR in FOXO3a-overexpressing MV3 cells after SIRT6 silencing. (E) Glucose uptake detected by flow cytometry in the 2-NBDG-treated FOXO3a-overexpressing MV3 cells after SIRT6 silencing. (F) Glucose consumption detected in FOXO3a-overexpressing MV3 cells after SIRT6 silencing. (G) Lactate production detected using a lactate assay kit in FOXO3a-overexpressing MV3 cells after SIRT6 silencing. (H) LDH activity detected using an LDH assay kit in FOXO3a-overexpressing MV3 cells after SIRT6 silencing. (I) Glycolytic stress flux test was conducted by using a Seahorse XF analyzer in FOXO3a-overexpressing MV3 cells after SIRT6 silencing. *P<0.05, **P<0.01, ***P<0.001. n.s., not significant. FOXO3a, forkhead box O3; SIRT6, sirtuin 6; RT-qPCR, reverse transcription-quantitative PCR; HK, hexokinase; GLUT, glucose transporter; PFK1, phosphofructokinase 1; PKM2, pyruvate kinase isozyme 2; LDHA, lactate dehydrogenase A; LDH, lactate dehydrogenase; sh, short hairpin (RNA); 2-NBDG, 2-Deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4-yl) amino]-D-glucose; ECAR, extracellular acidification rate; 2-DG, 2-deoxy-D-glucose.

FOXO3a-SIRT6 axis contributes to melanoma cell viability in vitro and tumorigenicity in vivo

The effect of this regulatory axis on the viability of melanoma cells was subsequently investigated in vitro. MTT assays showed that FOXO3a silencing promoted cell viability, whereas FOXO3a overexpression inhibited cell viability (Fig. 6A and B). Furthermore, SIRT6 overexpression and silencing rescued the effects on cell viability induced by FOXO3a silencing and overexpression, respectively (Fig. 6A and B). Subsequently, a xenograft nude mouse model was employed to confirm the effect of the FOXO3a-SIRT6 axis on the growth of MV3 cells in vivo. The results showed that FOXO3a silencing promoted tumor growth in vivo, whereas SIRT6 restoration rescued the effect of FOXO3a silencing (Fig. 6C and D). Then, the glucose and lactate contents of tumors were detected, and the results showed that the contents of glucose and lactate were significantly increased in FOXO3a-silenced tumors; however, this effect was attenuated by SIRT6 restoration (Fig. 6E and F). To further validate the role of the FOXO3a-SIRT6 axis using clinical data, a cohort from the TCGA melanoma database was analyzed. The results showed that FOXO3a high/SIRT6 high co-expression predicted the best overall survival rate, whereas the FOXO3a low/SIRT6 low subgroup exhibited the worst prognosis (Fig. 6G). These results indicated that the FOXO3a-SIRT6 regulatory axis is an important regulator in cellular metabolism and tumor growth of melanoma cells both in vitro and in vivo.

Figure 6 FOXO3a-SIRT6 axis contributes to melanoma cell viability in vitro and tumorigenicity in vivo. (A) Cell viability was determined by MTT assays in FOXO3a-silenced MV3 cells after SIRT6 overexpression. (B) Cell viability was determined by MTT assays in FOXO3a-overexpressing MV3 cells after SIRT6 silencing. (C) Tumor volume across 25 days after formation of visible tumors in BALB/c-nu mice injected MV3 cells with FOXO3a silencing and SIRT6 overexpression. The first measurements were made 10 days after injection. Scale bar, 1 cm. (D) Tumor weight at 35 days after FOXO3a-silenced MV3 cells with SIRT6 overexpression were injected into BALB/c-nu mice. (E) Relative glucose content was detected in tumors formed by FOXO3a-silenced MV3 cells with SIRT6 restoration. (F) Relative lactate content was detected using a lactate assay kit in tumors formed by FOXO3a-silenced MV3 cells with SIRT6 restoration. (G) Kaplan-Meier analysis of FOXO3A and SIRT6 expression in the TCGA Ocular melanomas (project no. TCGA-UVM) dataset downloaded from the UCSC Xena (http://xena.ucsc.edu/). Groups were separated based on the median cutoff modus; a Bonferroni correction was used after the log-rank test to determine significant differences between two groups. P<0.00833 (0.05/6) was considered significant in this panel. Significant comparisons were marked as red. *P<0.05, **P<0.01, ***P<0.001. FOXO3a, forkhead box O3; SIRT6, sirtuin 6; sh, short hairpin (RNA).

Discussion

Previous studies have reported that FOXO3a is a downstream factor of PI3K/AKT that inhibits the survival, growth, migration and invasion of uveal melanoma cells, as well as inducing cell cycle arrest at G1 phase and apoptosis by transcriptionally regulating the expression of its downstream genes, including Bcl-2-like protein 11, cyclin-dependent kinase inhibitor 1B, survivin and cyclin D1 (21,23,24,36). Furthermore, FOXO3a triple mutant overexpression sensitized melanoma cells to apoptosis induced by temozolomide (22). These previous indicated that FOXO3a plays important roles in the development of melanoma.

In the present study, it was found that high FOXO3a expression predicted improved prognosis for patients with melanoma. Additionally, FOXO3a expression was associated with the malignancy of melanoma. These results were consistent with previous reports. However, in addition to transcriptionally regulated genes related to cell cycle and apoptosis (21,23,24,36), there was an alternative mechanism for FOXO3a in the development of melanoma; it was observed that FOXO3a regulated aerobic glycolysis by regulating the expression of SIRT6, which is recognized as a major regulator of cellular metabolism in cancer (8).

Aerobic glycolysis is an important feature of melanoma; in normoxia, melanoma cells with varying heterogeneity typically display highly glycolytic phenotypes, in which 60-80% of glucose is metabolized into lactate (37). BRAF(V600E) oncogene, a major driver in the tumorigenesis of melanoma, has been shown to promote aerobic glycolysis (38). This highly glycolytic phenotype is characterized by high expression of glycolysis-associated genes that encode glucose transporters and enzymes involved in aerobic glycolysis. For example, in melanoma, high expression levels of glycolytic proteins such as GAPDH and PKM2 were associated with worse clinical outcome in stage III melanoma (39,40). Additionally, GLUT1 was highly expressed in melanoma tissues and revealed to enhance the metastasis of malignant melanoma cells (41). In stage IV melanomas with high serum LDH, glycolysis is the principle source of energy (42). These findings indicate that modulating aerobic glycolysis may be a promising approach to treat melanoma. In the present study, it was shown that FOXO3A expression in several cohorts of patient with melanoma was negatively correlated with the expression of a cluster of glycolysis-associated genes, including HK1, HK3, PFKM, PFKFB3, PKM and LDHA.

It was revealed that FOXO3a was positively correlated with the expression of a major glycolysis regulator, SIRT6, a histone deacetylase. SIRT6 can regulate the expression of various glycolysis-associated genes, including GLUT1, PDK4, PDK1, ALDOC, PFK1, LDHB, LDHA, TPI5 and GAPDH by directly deacetylating histone 3 lysine 9; meanwhile, SIRT6 also represses the transcriptional activity of HIF1α and Myc, transcription factors that also regulate genes associated with glycolysis (35,43,44). The critical function of SIRT6 in the modulation of aerobic glycolysis has been reported in numerous tumors, such as breast cancer (45), urothelial carcinoma (46) and hepatocellular carcinoma (47). However, to our knowledge, the function of SIRT6-regulated glycolysis had not previously been explored in melanoma.

In the present study, it was found that FOXO3a could transcriptionally promote the expression of SIRT6. The detailed mechanism of FOXO3-regulated SIRT6 expression in humans varies from that previously reported in mice. In mice, the transcription factor NRF1 regulates the transcription of SIRT6, whereas FOXO3a only functions as a co-promoter (34). However, no binding sites for NRF1 were observed in the promoter region of the human SIRT6 promoter. Conversely, it was demonstrated that FOXO3a functioned as a direct transcription factor that targeted the promoter of SIRT6. Recently, a similar mechanism has also been reported in a study in colon cancer (48).

To elucidate the critical function of the FOXO3a-SIRT6 axis in the regulation of glycolysis and tumor growth in melanoma, SIRT6 was overexpressed in FOXO3a-silenced MV3 cells and SIRT6 was knocked down in FOXO3a-overexpressing MV3 cells. Then, glucose uptake and consumption, lactate production, LDH activity, glycolytic gene expression and cell viability were evaluated in vitro, and tumor growth was investigated in vivo. The results showed that the effects of altered FOXO3a expression could be rescued by inverse manipulations of SIRT6 expression, indicating that the FOXO3a-SIRT6 axis played a pivotal role in the modulation of cell metabolism and tumor growth. However, the function of this axis may vary in different types of cancers and under different conditions. For example, GLUT1 expression is reduced in the absence of FOXO3a in glioma cells under serum starvation (49), implying that FOXO3a has a more complex role in the regulation of cell metabolism and cell survival.

In conclusion, the present study identified a novel mechanism for FOXO3a in the suppression of melanoma development. FOXO3a transcriptionally promoted the expression of SIRT6, which subsequently suppressed the expression of a number of glycolysis-associated genes (Fig. 7). The FOXO3a-SIRT6 regulatory axis serves an important role in modulating cellular metabolism, thereby affecting cancer development in vitro and in vivo. The present findings indicated that the FOXO3a-SIRT6 axis may be a therapeutic target for the treatment of melanoma.

Figure 7 Model of action for the FOXO3a-SIRT6 regulatory axis in the modulation of cell metabolism in melanoma cells. Briefly, FOXO3a transcriptionally promotes the expression of SIRT6, which subsequently deacetylates H3K9 in the promoters of a cluster of glycolysis-associated genes, transcriptionally suppressing their expression in association with HIF1α or Myc. FOXO3a, forkhead box O3; SIRT6, sirtuin 6; HIF1α, hypoxia-inducible factor 1α; Ac, acetyl; HK, hexokinase; GLUT, glucose transporter; PFK1, phosphofructokinase 1; PKM2, pyruvate kinase isozyme 2; LDHA, lactate dehydrogenase A.

Supplementary Data

Funding

This work was supported by the Project Funded by Chongqing Special Postdoctoral Science Foundation (grant no. XmT2018080), the National Key Research and Development Program of China (grant nos. 2017YFC1308600 and 2016YFC1302204), the National Natural Science Foundation of China (grant nos. 81672502, 31672496 and 81902664), the Fundamental Research Funds for the Central Universities (grant no. XDJK2019C013), and the Research and Innovation Project of Graduate Students in Chongqing (grant no. CYS19136).

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

ZD performed the experiments, acquired the data and drafted the manuscript. JY, LL, LT and PS performed molecular biological experiments. JZ, XZ and LG analyzed the data and performed statistical analysis. ZW and HC conceived and designed the study, and reviewed/revised the manuscript, figures and tables. All authors read and approved the final manuscript.

Ethics approval and consent to participate

The animal experiments in the current study were approved and supervised by the Institutional Animal Care and Use Committees of the Southwest University (permit no. IACUC-20190402-02) and the Experimental Animal Care and Use Committees of the Institute of Sericulture and Systems Biology. The study was performed according to the Laboratory Animal Management Regulations and the Measures of Chongqing Municipality on the Management of Experimental Animals.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Acknowledgments

We thank Mr. Gaichao Zhao (State Key Laboratory of Silkworm Genome Biology, Southwest University) for his contributions to the revision of this manuscript.

References