ACSM3 suppresses the pathogenesis of high-grade serous ovarian carcinoma via promoting AMPK activity

Abstract

Purpose

Ovarian carcinoma is the fifth commonest malignancy in females and exhibits a high recurrence rate. High-grade serous ovarian carcinoma (HGSOC) is the main histologic subtype. It displays extensive genetic heterogeneity. Here, we aimed to identify potential therapeutic targets for HGSOC.

Methods

Both bioinformatic data from TCGA and 73 pairs of tumor and normal samples from patients were analyzed to reveal the expression level of ACSM3 in HGSOC. Next, cellular and animal experiments, including cell proliferation, colony formation and xenograft assays were performed to explore the suppressive function of ACSM3. Finally, biochemical methods, AMP/ATP ratio measurements and Western blotting were used to elucidate the mechanism underlying the ACSM3-AMPK axis in HGSOC.

Results

After analyzing transcriptome data of TCGA HGSOC samples, we found that ACSM3 is down-regulated in patient samples compared with normal controls. This observation was validated using data from primary clinical samples. Proliferation, soft agar colony formation and xenograft assays revealed that ACSM3 is able to suppress HGSOC tumor growth both in vitro and in vivo. Moreover, we found that ACSM3 overexpression increased the AMP/ATP ratio and the phosphorylation level of AMPK at threonine 172. In addition, we found that AMPK silencing in EFO21 and SKOV3 cells completely abolished the anti-oncogenic effect of ACSM3.

Conclusion

Our data indicate that the ACSM3-AMPK axis is involved in the pathogenesis of HGSOC and, as such, may act as a therapeutic target for this cancer.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13402-021-00658-1.

Introduction

The annual number of diagnosed cases of ovarian carcinoma has increased to ~239,000 worldwide, including ~152,000 deaths, making it the fifth most common deleterious malignancy in females [1]. Ovarian carcinoma is an extremely heterogeneous cancer type both at the genomic and the cellular level, including cell types originating from stromal, epithelial and even germ cells [2, 3]. According to histology, ovarian carcinomas can be divided into at least five subtypes, i.e., high-grade serous ovarian cancer (HGSOC), low-grade serous ovarian cancer, endometrioid cancer, clear cell cancer, mucinous ovarian cancer and carcinosarcoma [4]. As the most predominant subtype, HGSOC accounts for 65–70% of all ovarian carcinoma cases [1, 4]. DNA copy number and structural changes have been reported to mainly contribute to its genetic heterogeneity. It has been reported that the genomic instability of HGSOC is caused by loss of DNA repair functions, especially homologous recombination [2]. In addition, TP53 mutations have been considered as drivers of HGSOC tumor progression, which differs from that of low-grade serous carcinoma [5, 6]. Although several immunotherapies, including VEGF (vascular endothelial growth factor) and PARP (polyADP-ribose polymerase) inhibitors such as bevacizumab have been used in clinic trials, the high recurrence rate of about 70% still poses a great challenge for obtaining long term survival [7, 8]. This means that more detailed molecular information is required to improve the poor prognosis.

AMPK (AMP-activated protein kinase) is a well-known cellular metabolic sensor that plays a pivotal role in energy homeostasis [9, 10]. AMPK is a serine/threonine kinase, composed of one catalytic α and two regulatory β, γ subunits [11, 12]. In case of low ATP levels and a high AMP/ATP ratio, AMPK is activated after a conformational change of its three subunits and a subsequent phosphorylation at the threonine 172 residue [13, 14].

The ACS (Acyl-coenzyme A synthetase) family catalyzes the formation of acyl-CoA from coenzyme A, thereby mediating the activation of fatty acids [15]. Among the 26 members of the ACS family, 5 are responsible for the activation of medium-chain length fatty acids, which are classified as the ACSM (acyl-CoA medium-chain) subfamily [16, 17]. Up to now, 5 human ACSM genes have been identified, named ACSM1–5 [17]. Accumulating evidence indicates that ACSM3 is, among others, associated with muscular growth, functioning of the urinary system and metabolic diseases. It has been reported that aberrant expression of ACSM3 induced by oar-miR-655-3p causes abnormal muscle development in sheep [18] and that elevated ACSM3 levels are associated with muscular dystrophy [19]. ACSM3 has been found to be down-regulated in high-fat diet (HFD)-induced obesity in C57BL/6 J mice [20]. In the urinary system, it has been found that ACSM3 is down-regulated in cadmium-accumulated kidneys [21] and ulcerative colitis [22]. Additional studies have indicated that ACSM3 may also be involved in the development and metastasis of multiple cancers. Microarray-based studies have shown that decreased ACSM3 expression enhances metastasis and predicts a poor clinical outcome in patients with hepatocellular carcinoma [23]. A meta-analysis of mRNA profiles revealed that ACSM3 was largely down-regulated in hepatocellular carcinoma tissues and patients with severe clinical manifestations including high AFP (alpha-fetoprotein) levels, high ALT (alanine aminotransferase) levels and large tumor volumes [24]. Based these observations, we speculated that ACSM3 may also play a role in other tumors. Here, we analyzed ACSM3 expression levels in HGSOC tissues from the TCGA (The Cancer Genome Atlas) database and found that they were down-regulated. Subsequent molecular analyses revealed that ACSM3 may suppress the pathogenesis of HGSOC cells via promoting AMPK activation. The ACSM3-AMPK axis may, therefore, serve as a potential therapeutic target for HGSOC.

Materials and methods

Plasmid construction

The coding sequence of ACSM3 (NCBI reference sequence: NM_005622.4) was obtained from EFO21 cell-derived cDNA and ligated into a pLVX-PURO vector (Cat: #125839, Addgene) using T4 DNA ligase (Cat: #M0202L, NEB). shRNAs targeting ACSM3 and AMPK and sgRNAs targeting ACSM3 were ligated into a pLKO.1-PURO vector (Cat: #8453, Addgene) and a pX459 V2 vector (Cat: #108294, Addgene), respectively. The respective shRNA and sgRNA sequences are listed in Table 1.

Table 1.

shRNA and sgRNA sequences

Targets	Directions	Sequences (5′- > 3′)
scramble	Forward	CCGGCATGTTAAGTAAGCCGATATACTCGAGTATATCGGCTTACTTAACATGTTTTTC
	Reverse	AATTGAAAAACATGTTAAGTAAGCCGATATACTCGAGTATATCGGCTTACTTAACATG
shACSM3#1	Forward	CCGGCCTGATTACAAGTCACATGATCTCGAGATCATGTGACTTGTAATCAGGTTTTTC
	Reverse	AATTGAAAAACCTGATTACAAGTCACATGATCTCGAGATCATGTGACTTGTAATCAGG
shACSM3#2	Forward	CCGGGCTGGAAAGAAACCTTCAAATCTCGAGATTTGAAGGTTTCTTTCCAGCTTTTTC
	Reverse	AATTGAAAAAGCTGGAAAGAAACCTTCAAATCTCGAGATTTGAAGGTTTCTTTCCAGC
shAMPKα1#1	Forward	CCGGGTTGCCTACCATCTCATAATACTCGAGTATTATGAGATGGTAGGCAACTTTTTC
	Reverse	AATTGAAAAAGTTGCCTACCATCTCATAATACTCGAGTATTATGAGATGGTAGGCAAC
shAMPKα1#2	Forward	CCGGGTAGCTGTGAAGATACTCAATCTCGAGATTGAGTATCTTCACAGCTACTTTTTC
	Reverse	AATTGAAAAAGTAGCTGTGAAGATACTCAATCTCGAGATTGAGTATCTTCACAGCTAC
shAMPKα2#1	Forward	CCGGGTGGCTTATCATCTTATCATTCTCGAGAATGATAAGATGATAAGCCACTTTTTC
	Reverse	AATTGAAAAAGTGGCTTATCATCTTATCATTCTCGAGAATGATAAGATGATAAGCCAC
shAMPKα2#2	Forward	CCGGCCCACTGAAACGAGCAACTATCTCGAGATAGTTGCTCGTTTCAGTGGGTTTTTC
	Reverse	AATTGAAAAACCCACTGAAACGAGCAACTATCTCGAGATAGTTGCTCGTTTCAGTGGG
sg-ACSM3#1	Forward	CACCGATGGCCATATTCTTTACCAG
	Reverse	AAACCTGGTAAAGAATATGGCCATC
sg-ACSM3#2	Forward	CACCGATAGAAGTCGGCTCAAAACG
	Reverse	AAACCGTTTTGAGCCGACTTCTATC

Cell culture and cell line construction

The EFO21 (Cat: ACC 235) cell line was purchased from the Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures GmbH (Braunschweig, Germany) and cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS, Gibco, Cat: 10099141C), 2 mM L-glutamine, 1 × MEM non-essential amino acids, 1 mM sodium pyruvate and 50 μg/ml penicillin/streptomycin (P/S, Cat:10378016, ThermoFisher Scientific). The SKOV3 (Cat: ATCC HTB-77) cell line was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in ATCC-formulated McCoy’s 5a Medium Modified (Cat: 30–2007) supplemeted with 10% FBS and 50 μg/ml P/S. All cells were maintained in an incubator at 37 °C with 5% CO₂.

To construct stably transfected cell lines, lentiviral vectors psPAX2 (Cat: #12260, Addgene) and pMD2.G (Cat: #12259, Addgene) were co-transfected into HEK293T cells using PolyJet (SL100688, Signagen, USA). Forty-eight hours later, the supernatants were collected and filtered through 0.45 μm filters. Next, the pseudovirus particles and 10 μg/ml polybrene (Cat: TR-1003-G, Sigma, USA) were used to infect EFO21 and SKOV3 cells. After 2 μg/ml puromycin (Cat: A1113802, ThermoFisher Scientific, USA) selection, positive cells were used for further experiments.

OCR and ECAR measurements

Intracellular ATP production rates were measured using a XF24 extracellular flux analyzer (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instructions. 1 × 10⁵ cells were seeded in 24-well culture microplates. After an incubation period of 20 h, the basal oxygen consumption rate (OCR) and the extracellular acidification rate (ECAR) were determined. For the ECAR measurement, glucose (2.5 mM, at 20 min), oligomycin (2.5 mM, at 50 min) and 2-Deoxy-D-glucose (2-DG, 5 mM, at 80 min) were sequentially added to the culture medium. For the OCR measurement, oligomycin (2.5 mM, at 20 min), 4-(trifluoromethoxy) phenylhydrazone (FCCP, 300 nM, at 50 min) and rotenone (2.5 mM, at 80 min) were sequentially added to the culture medium. The OCR and ECAR values were normalized by the total cell numbers in each well, which were quantified using DAPI staining (Cat: E607303, Sangon Biotech, Shanghai).

Western blotting

After the preparation of tissue lysates, protein concentrations were determined using a BCA Protein Assay Kit (Cat: ab102536, Abcam, Cambridge, MA, USA). For the extraction of the protein samples, equal numbers of cells were seeded into each well and directly lysed using 1 × SDS loading buffer (50 mM Tris, pH 6.8, 100 mM DTT, 2% SDS (w/v), 10% glycerol (v/v), 0.1% Bromophenol blue (w/v)). Subsequently, the lysates were boiled at 100 °C for 10 min. After separation by 10% SDS-PAGE, the proteins were transferred to nitrocellulose membranes. Next, the membranes were blocked with 5% non-fat milk powder at room temperature for 1 h and incubated with primary antibody at 4 °C overnight, and then with horse raddish peroxidase (HRP)-conjugated secondary antibody at room temperature for 1 h. Signals were detected using Tanon High-sig ECL Western Blotting Substrate (Cat: 180–501, Shanghai, China). Primary antibodies directed against ACSM3 (Cat: ab113920, 1:1000), pAMPKα-T172 (Cat: ab133448, 1:1000) and AMPKα (Cat: ab80039, 1:2000) were purchased from Abcam (Cambridge, MA, USA), and those direcetd against ACTIN (Cat: #3700, 1:10000) and FLAG (Cat: #14793, 1:10000) from Cell Signaling Technology (Danvers, MA, USA).

Immunofluorescence assay

EFO21 cells were seeded onto cover glasses and transfected as indicated. Next, the cells were washed with PBS three times, fixed with 4% PFA (paraformaldehyde) Fix Solution (Cat: E672002, Sangon Biotech, Shanghai) for 15 min, and permeabilized with 0.2% Triton X-100 for 10 min. Anti-FLAG and anti-pAMPKα-T172 primary antibodies were incubated at 4 °C overnight with DAPI dye (Cat: E607303, Sangon Biotech, Shanghai, China), after which fluorophore-conjugated secondary antibodies were incubated at room temperature for 2 h. Images were captured using a Leica fluorescence microscope.

Cell proliferation assay

Cell proliferation rates were determined using CCK-8 (cell counting kit-8) (Cat: HY-K0301, MedChemExpress, Shanghai, China) according to the manufacturer’s instructions. Briefly, cells were seeded in 96-well plates and harvested at the indicated time points. After subsequent CCK-8 treatment at 37 °C for 2 h, absorbance was measured at 450 nm using an automatic microplate reader (Cat: DY001292, Nantong, China).

Colony formation assay

Soft agar colony assays were carried out as previously described [25]. Briefly, 1.0% agar and 2 × RPMI-1640 were evenly mixed to make a 0.5% base layer. Next, 2000 cells were seeded in a 0.35% top agar layer and the dishes were placed in a 37 °C incubator for 3 weeks, after which they were evaluated by crystal violet staining.

AMP/ATP ratio measurement

AMP/ATP ratios were measured using an AMP/ATP Ratio Assay Kit (ab65313, Abcam, Burlingame, CA, USA) that is based on the conversion of ATP by luciferase and then bioluminescent detection. All steps were performed according to the manufacturer’s instructions. Briefly, cells were treated with nucleotide releasing buffer for 10 min after which the ATP reaction mix was added to the buffer for a 2 min incubation. The concentration of ATP was determined using a luminescence plate reader. Next, the ADP reaction buffer was added to the same wells. After another incubation for 2 min, the ADP concentration was determined using a luminescence plate reader.

Animal xenografting

Cell line-derived xenograft experiments were conducted as previously described [26]. BALB/c nude mice were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. The mice were randomly divided with 12 mice in each group. Next, 1 × 10⁶ cells in sterile PBS were subcutaneously injected into both flanks of each mouse and tumor volumes were recorded every 4 days using a vernier caliper.

Patient specimens

Tumor and normal tissues from HGSOC patients (n = 73) were collected from the Chengdu Fifth People’s Hospital. The clinical parameters are presented in Table 2. All participants (patients or their families) signed written informed consent forms. The protocols involved in this study were approved by the patients and the Ethics Committee of Chengdu Fifth People’s Hospital. The samples were frozen in liquid nitrogen and RNAlater Stabilization Solution Invitrogen (Cat: AM7021, Waltham, MA, USA) and stored for further use. For the preparation of protein samples, tumor and normal tissues were lysed using a total tissue protein lysis buffer (Cat: C500028–0010, Sangon Biotech, Shanghai, China). The protein concentrations were quantified using a BCA Protein Assay Kit (Cat: PC0020, Solarbio, Beijing, China).

Table 2.

Clinical information of HGSOC patients involved in this study

Information	Number	Percentage
Total number	73
FIGO stages
I	6	8.2%
II	11	15.1%
> III	52	71.2%
Missing	4	5.5%
Familial status
Sporadic	48	65.8%
Familial	23	31.5%
BRCA1	11	47.8%
BRCA2	5	21.7%
non-BRCA1/2	7	30.4%
Missing	2	2.7%
Overall survival (months), median (95% CI)	35.7 (32.1–43.6)
Progession-free survival (months), median (96% CI)	18.1 (10.6–25.9)

CI confidence interval, FIGO International Federation of Gynecology and Obstetrics, HGSOC high-grade serous ovarian cancer

RNA extraction and RT-PCR

Tissue mRNAs were extracted using a TaKaRa MiniBEST Universal RNA Extraction Kit (Cat: 9767, Beijing, China) after which cDNAs were synthesized using a TaKaRa First-Strand cDNA synthesis kit (Cat: 6110A, Beijing, China). mRNA levels were measured using an ABI7500 system and a TaKaRa TB Green Fast qPCR Mix (Cat: RR430S, Beijing, China). The ACSM3 expression levels were calculated using 2^(−△△CT) and normalized by ACTB. The primers used were: ACSM3, Forward, 5′-AGG AAG ATG CTA CGT CAT GCC-3′, Reverse, 5’-ATC CCC AGT TTG AAG TCC TGT-3′; ACTB, Forward, 5’-CAT GTA CGT TGC TAT CCA GGC-3′, Reverse, 5’-CTC CTT AAT GTC ACG CAC GAT-3′.

TCGA data analysis

TCGA-OV (ovarian serous cystadenocarcinoma) expression profiles and clinical data were analyzed using GEPIA2 (Gene Expression Profiling Interactive Analysis) [27], a widely used online tool for gene expression analysis based on tumor and normal samples from the TCGA (The Cancer Genome Atlas project) and the GTEx (Genotype-Tissue Expression project) databases. The expression levels in tumor and normal tissues were transformed by log2(TPM + 1). For differential analysis, one-way ANOVA with a cutoff of log2FC >1 and a p < 0.01 were used for comparison of patients (n = 426) and normal (n = 88) cases. For overall survival analysis, the Mantel-Cox log-rank test was used, with a cutoff based on the median of ASCM3 expression (high-expression: samples with expression > median, low-expression: samples with expression < median).

Statistical analysis

All results are presented as mean ± SEM. GraphPad Prism 7 software (California, USA) was used for statistical analysis, in which student’s t test was performed for comparisons for two groups, and one-way ANOVA for more than two groups. p < 0.05 was considered statistically significant.

Results

ACSM3 is down-regulated in HGSOC patients

First, we analyzed the expression spectrum of all 26 ACS family genes in 21 different normal tissues. Interestingly, we found that ACSM3 was highly expressed in ovarian tissues, while the expression of the other 25 genes was relatively low (Fig. S1A). This result indicated that ACSM3 may play a distinct role in ovarian carcinoma. To test this, we turned to the TCGA (The Cancer Genome Atlas) database and found that the expression of the ACSM3 gene was decreased more than 3.8 folds in comparison with that in normal tissues (Fig. 1A). Next, we compared both the mRNA and protein levels of ACSM3 in 16 pairs of cancer tissues and its normal counterparts using RT-PCR and Western blotting, respectively. The mRNA and protein levels showed a similar pattern, i.e., the overall levels in cancer tissues were lower than those in normal tissues (Fig. 1B and C). We also checked the expression levels of all ACS family members in normal donor (n = 9) and ovarian cancer tissues (n = 8). We found that, compared with the normal tissues, the ACSM3 expression level was decreased by almost 15 fold (−3.91 cycles) in the ovarian cancer tissues, which was lower than that of ACSS3, ACSM1 and ACSF2 (Fig. S1B). In contrast, we found that S27A3 was significantly upregulated in the ovarian cancer tissues (Fig. S1B). These findings reinforce the idea that ACSM3 may play a distinct role in ovarian cancer development. In addition, we found that the ACSM3 expression levels were positively correlated with overall survival. During the same time range, patients of the high ACSM3 group exhibited a higher survival rate than those of the low ACSM3 group (Fig. 1D). From the online database analysis and samples available in our lab, we conclude that ACSM3 is highly expressed in normal ovarian tissues and down-regulated in HGSOC.

Fig. 1.

ACSM3 is downregulated in patients with HGSOC. A TCGA ACSM3 ovarian cancer expression levels; normal (n = 88; Con) and HGSOC (n = 426; Tumor). B Relative ACSM3 mRNA expression levels of HGSOC and paired normal tissue samples (n = 78). The ACSM3 levels were calculated as 2^(−△△CT) and normalized to ACTB. ***p < 0.001 vs control. C ACSM3 protein levels of HGSOC and paired normal tissue samples. β-Actin (ACTIN) was used as loading control. Left, representative Western blots of patient samples (n = 16); Right, relative ACSM3 protein levels (n = 78). ***p < 0.001 vs control. D Overall survival rates of HGSOC patients with different ACSM3 protein levels. Red and blue dotted lines indicate 95% confidence intervals

ACSM3 is involved in the pathogenesis of ovarian cancer cells

Next, we used ovary cancer cells to explore the role of ACSM3 in the pathogenesis of ovarian carcinoma. First, we checked the ACSM3 expression levels in cell lines derived from different types of ovarian carcinoma. We noted a large difference among these cell lines, both in ACSM3 mRNA (Fig. S2A) and protein (Fig. S2B) levels. Particularly, we found that the expression of ACSM3 was high in EFO21 and low in SKOV3, two adenocarcinoma cell lines. Therefore, we knocked down ACSM3 in EFO21 cells using shRNA (small hairpin RNA) and overexpressed ACSM3 in SKOV3 cells using a lentivirus vector, after which the expression levels were verified using both RT-PCR and Western blotting (Figs. 2A and S2C, Fig. 2B). Using cell proliferation and colony formation assays, we found that ACSM3 knockdown in EFO21 cells promoted their proliferation and colony forming capacities (Fig. 2A and C), while ACSM3 overexpression in SKOV3 cells decreased their proliferation and colony forming capacities (Fig. 2B and D). Based on these results, we next carried out cell line-derived xenograft experiments in nude mice. After subcutaneous injection of stably transfected cells, tumor formation was tracked every 4 days, after which the tumors were harvested on day 28. We found that the volumes and weights of SKOV3-ACSM3-OE-derived tumors were significantly decreased compared with those of SKOV3-control tumors (Fig. 2E and F). Similarly, we found that after ACSM3 knockdown in EFO21 cells, the growth and final weights of the tumors were increased slightly, but significantly, in comparison with those of wild-type cells (Fig. 2G and H). These data indicate that ACSM3 suppresses tumor development and progression in vivo. Collectively, our findings indicate that ACSM3 suppresses the proliferation and tumor development and progression of ovarian cancer cells.

Fig. 2.

ACSM3 is involved in the pathogenesis of ovarian carcinoma. A Cell proliferation assay (CCK-8) of ACSM3-KD EFO21 cells. Two different shRNA sequences targeting SCSM3 (#1, #2) were used to knock down ACSM3 in EFO21 cells. A two-way ANOVA test was used to compare the four groups. The mRNA levels of the ACSM3 constructed cells were validated by RT-PCR. N = 6, ***p < 0.001 vs control. B Cell proliferation assay (CCK-8) of ACSM3-OE SKOV3 cells. A two-way ANOVA test was used for statistical analysis. The expression level of ACSM3 was validated by Western blotting. N = 6, ***p < 0.001 vs control. C-D Colony formation assay of ACSM3-KD EFO21 and ACSM3-OE SKOV3 cells and statistical results. Cell construction and statistics were the same as under (A). Cells in A and B were used, respectively. N = 6, ***p < 0.001 by student’s t test. E Tumor volumes of ACSM3-OE SKOV3 derived xenografts in nude mice. Cell construction was the same as under (B). N = 12, **p < 0.01 vs control by two-way ANOVA. F Tumor sizes and weights of ACSM3-OE SKOV3 derived xenografts. Two thousand constructed cells were injected into nude mice and the tumors were measured on day 28. The picture is a representative of the final tumors. N = 12, **p < 0.01 vs control by student’s t test. G Tumor volumes of ACSM3-KD EFO21 derived xenografts in nude mice. Cell construction was the same as under (A). N = 12, ***p < 0.001 vs control by two-way ANOVA. H Tumor sizes and weights of ACSM3-KD EFO21 derived xenografts. Cells were injected into nude mice and the tumors were measured on day 28. The figure is a representative of the final tumors. N = 12, **p < 0.01 vs control by student’s t test

ACSM3 promotes AMPK activity by affecting ovarian cancer cell metabolism

Due to its catalytic role in the activation of fatty acids [15], we were curious about the metabolic role of ACSM3 in ovarian carcinoma. The CRISPR-Cas9 technology was used to knockout ACSM3 (sg-ACSM3#1 and sg-ACSM3#2) in EFO21 cells. We found that in two cell clones (EFO21-KO-ACSM3#1, EFO21-KO-ACSM3#5) ACSM3 was completely knocked out and in one cell clone (EFO21-KD-ACSM3#6) partially knocked down, as validated by Western blotting and Sanger sequencing (Fig. S3A and B). All three cell clones exhibited decreased AMP/ATP ratios (Fig. 3A). When ACSM3 was re-introduced into EFO21-KO cells, the AMP/ATP ratio increased again (Fig. 3B). Accordingly, we found that exogenous over-expression of ACSM3 in SKOV3 cells led to an elevated AMP/ATP ratio as well (Fig. 3C). These observations indicate that ACSM3 affects the metabolism in these cells. A high AMP/ATP ratio is a signal for AMPK activation [13, 14]. Therefore, we set out to determine AMPK activity by detecting its phosphorylation level at T172 (pAMPKα-T172) [13, 14]. Both ACSM3 re-introduction in EFO21-KO cells and exogenous over-expression in SKOV3 cells promoted AMPK activity, as determined by Western blotting (Fig. 3D and E) and immunofluorescence detection of pAMPKα-T172 (Figs. 3F, S3C). Since mitochondria are the main cellular organelles involved in metabolism, we also measured ECAR (Excellular Acidification Rate) and OCR (Oxygen Consumption Rate) as indicators of mitochondrial respiration and glycolysis, respectively. In all three EFO21-ACSM3-KO cell lines (#1, #5, #6), both ECAR and OCR were decreased compared with parental EFO21 WT cells (Fig. S3D and E). Together, these results indicate that ACSM3 affects cellular metabolism, resulting in an increased AMP/ATP ratio and in AMPK activation.

Fig. 3.

ACSM3 promotes AMPK activity by affecting metabolism in HGSOC cells. A AMP/ATP ratio of ACSM3-KO EFO21 cells. Different KO (#1, #5) or KD (#6) cell clones were generated by CRISPR-Cas9. KO#1, KO#5 and KD#6 cells were validated by Western blotting (Fig. S3A) and Sanger sequencing (Fig. S3B). N = 6, ***p < 0.001 vs. control by student’s t test. B AMP/ATP ratio of ACSM3-KO and ACSM3-restoration in EFO21#1 cells, and validation of ACSM3 expression by Western blotting (right panel). N = 6, **p < 0.01, ***p < 0.001 vs control by student’s t test. C AMP/ATP ratio of ACSM3-OE SKOV3 cells. N = 6, ***p < 0.001 vs. control by student’s t test. D Western blotting of pAMPKα-T172, pAMPKα and statistic results of pAMPKα-T172/pAMPKα ratio in ACSM3-KO and ACSM3-restoration EFO21#1 cells. The cells used were the same as under (B). N = 6, **p < 0.01, ***p < 0.001 vs control by student’s t test. E Western blot analysis of pAMPKα-T172 and pAMPKα expression, and statistic results of pAMPKα-T172/pAMPKα ratios in ACSM3-OE SKOV3 cells. The cells used were the same as under (C). N = 6, **p < 0.01 vs control. F Immunofluorescence detection of pAMPKα-T172 and ACSM3 in ACSM3-KO and ACSM3-restoration EFO21#1 cells. The cells used were the same as under (C). N = 6, *p < 0.5 and ***p < 0.001 vs control by student’s t test

AMPK mediates the anti-oncogenic effect of ACSM3 in ovarian carcinoma

To confirm that the anti-oncogenic effect of ACSM3 in ovarian carcinoma is mediated by AMPK, shRNAs were used to knock down AMPK expression in EFO21 and SKOV3 cells (AMPK-KD) (Fig. S4A). We found that ACSM3 over-expression (OE) (Fig. S4B, right panel) suppresses SKOV3 cell proliferation, while AMPK knockdown (KD) abolishes the suppressive effects, as indicated by cell proliferation and colony formation assays (Fig. 4A and C). Conversely, we found that ACSM3-KD in EFO21 cells (Fig. S4B, left panel) led to enhanced proliferation, whereas AMPK-KD abolished the effects (Fig. 4B and D). In cell line-derived xenografts in nude mice, ACSM3 over-expression (OE) in SKOV3 cells inhibited tumor development and progression, as indicated by tumor volume curves and final tumor weights on day 28 (Fig. 4E and F). Intriguingly, these effects of ACSM3 were largely blocked by AMPK knockdown in SKOV3 cells (Fig. 4E and F). In addition, we found that ACSM3 deletion in EFO21 cells promoted xenograft tumor growth, which was completely blocked by AMPK knockdown (Fig. 4G). In summary, we found that ACSM3 can suppress in vitro ovarian carcinoma cell proliferation and inhibit in vivo ovarian carcinoma tumor development and progression in mice (Figs. 2 and 4). These effects were completely abolished after AMPK expression knockdown, indicating that AMPK acts as a downstream effector of ACSM3 and that the anti-oncogenic effect of ACSM3 in ovarian carcinoma is mediated by AMPK.

Fig. 4.

AMPK mediates the anti-oncogenic activity of ACSM3. A Cell proliferation assay (CCK-8) of ACSM3-OE and AMPK-KD SKOV3 cells. Two different shRNA sequences targeting AMPK showed similar results. VEC: empty vector; ACSM3: ACSM3 overexpression in SKOV3 cells. N = 6, ***p < 0.001 vs. control by two-way ANOVA. B Cell proliferation assay (CCK-8) of ACSM3-KD and AMPK-KD EFO21 cells. VEC: empty vector, negative control; shACSM3: ACSM3 knockdown in EFO21 cells. N = 6, ***p < 0.001 vs. control by two-way ANOVA. C Colony formation assay of ACSM3-OE and AMPK-KD SKOV3 cells. The cells used were the same as under (A). N = 6, **p < 0.01 vs control by student’s t test. D Colony formation assay of ACSM3-KD and AMPK-KD EFO21 cells. The cells used were the same as under (B). N = 6, ***p < 0.001 vs. control. E Tumor volumes of ACSM3-OE and AMPK-KD SKOV3 derived xenografts in nude mice. The cells used were the same as under (A). N = 12, **p < 0.01 vs control by two-way ANOVA. F Tumors and tumor weights of ACSM3-OE and AMPK-KD SKOV3 derived xenografts. Cells were injected into nude mice and the tumors were measured on day 28. The tumors shown are representative of the tumors. N = 12, *p < 0.05 vs control by student’s t test. G Tumor volumes of ACSM3-KD and AMPK-KD EFO21 derived xenografts in nude mice. The cells were the same as under (B). The tumors in the picture are representative of all tumors recovered. N = 12, **p < 0.01 vs control by student’s t test. H Working model of ACSM3-AMPK in ovarian carcinoma. The ACSM3-AMPK axis inhibits the development of HGSOC. Loss of ACSM3 leads to a decrease in the AMP/ATP ratio, after which AMPK activity is suppressed, which leads to ovarian tumorigenesis

Discussion

A lack of effective early screening and detection methods contributes to the high mortality rate of ovarian carcinoma patients. Currently, surgery and chemotherapy, occasionally in combination with immunotherapies involving VEGF or PARP inhibitors, are applied as major approaches [7, 8, 28, 29]. However, recurrence is a major challenge for long term survival [30, 31]. Therefore, it is of utmost importance to better understand the molecular mechanisms underlying ovarian cancer develoment and progression. Here, we found that ACSM3 is highly expressed in the normal ovary, but not the other 25 ACS family members. In HGSOC, ACSM3 was found to be down-regulated and subsequent in vitro experiments indicated that it may play a role in ovarian carcinoma development. The ACSM3 level was lower in SKOV3 cells than in EFO21 cells, and we found that tumors derived from SKOV3-Control/Vector cells were larger than those of EFO21-Control/Scramble cells, indicating that the inhibitory role of ACSM3 in ovarian tumorigenesis may be universal. Concordantly, Yan et al. found that ACSM3 suppressed ovarian cancer progression through repressing cell proliferation and decreasing cell migration and invasion [32]. Both studies indicate that ACSM3 may be a key factor in HGSOC tumorigenesis. Different from Yan et al., reporting that ACSM3 inhibits the integrin β1/AKT signaling pathway [32], our data revealed a completely new molecular mechanism by which ACSM3 represses ovarian cancer development. We propose that ACSM3 decreases mitochondrial respiration and glycolysis and, consequently, increases the AMP/ATP ratio. As a result, AMPK becomes activated. The inhibitory effect of ACSM3 was found to be blocked after AMPK knockdown, which indicates that AMPK acts downstream of ACSM3 and that the anti-oncogenic effect of ACSM3 in ovarian carcinoma is mediated by AMPK. Regardless of ACSM3 expression, we found that AMPK knockdown promoted the in vitro colony forming capacity and the in vivo tumor forming capacity of ovarian carcinoma cells. Others have also found that AMPK may be involved in ovarian carcinoma development. An analysis of patient-derived tissue samples revealed a lower AMPK expression in higher grade tumors together with an adverse prognosis [33]. A decreased AMPK-β1 and low AMPK level was found to promote oncogenic capacity and tumor grade in ovarian carcinoma [34]. It has been reported that AMPK activation may result in autophagy in SKOV3 ovarian cancer cells [35], and that AMPK activation by metformin [36], glucose deprivation [37] and Honokiol [38] may induce ovarian carcinoma apoptosis and cell death in advanced tumor stages. Therefore, AMPK is considered to play a positive role in suppressing the development of ovarian carcinoma and that targeting of AMPK signaling may provide opportunities for combating ovarian carcinoma [39]. Here, we identified ACSM3 as an upstream factor that regulates AMPK activity in HGSOC. The ACSM3-AMPK axis may serve as a potential therapeutic target and, as such, may be instrumental for future ovarian cancer drug development.

Authors’ contributions

Study design: XY and KJL. Experiments and data collection: FY, GXW, QZ, XC, JL, QH, LY, CW, MH, YL, JC, LL and HYW. Data analysis and interpretation: FY, GXW, QZ, XC, JL, QH, LY, CW, MH, YL, JC, LL, HYW, XY and KJL. Manuscript writing: XY and KJL. All author(s) read and approved the final manuscript.

Funding

No

Availability of data and materials

Not applicable.

Declarations

Ethics approval

All animal experiments in this study were conducted in accordance with the guidelines of the animal ethical committee for animal experimentation in China. The experimental design was approved by the Chengdu University of Traditional Chinese Medicine.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.