SIRT3 Inhibits Cholangiocarcinoma Cell Proliferation by Targeting HIF1α to regulate the EMT Signaling Pathway

Background Cholangiocarcinoma (CCA), is a rare biliary adenocarcinoma associated with poor outcomes. Deacetylase Sirtuin ‐ 3 (SIRT3), a histone deacetylase (HDAC), has been considered to be associated with various cancers and can be a potential new target for CCA.We intended to identify the target of SIRT3 and explore the mechanism of SIRT3 in CCA. Methods The expression levels of SIRT3 and hypoxia ‐ inducible factor ‐ 1α (HIF1α) in CCA tissues and cell lines were examined by RT-qPCR. CCK-8 and EdU methods were used to detect cell proliferation in CCA. To assess the levels of proteins related to cell proliferation and epithelial-mesenchymal transition (EMT) process, western blot analysis was conducted. Co-Immunoprecipitation and deacetylation assays were used to explore HIF1α protein acetylation, stability and the relationship between SIRT3 and HIF1α in CCA cells.


Introduction
Cholangiocarcinoma (CCA) is a primary liver malignancy, initially occurring in the intrahepatic biliary tract 1 . According to the existing statistic, the morbidity and mortality of CCA are increasing sharply each year due to its late clinical presentation, high conversion rate and poor prognosis 2 . CCA is often diagnosed at an advanced stage due to limited diagnostic methods 3,4 . Surgery is the choice for most patients, however, because of the rapid progression and metastasis of CCA tumors, surgical treatment is not appropriate for all patients 5 . In addition, radiotherapy or chemotherapy is still not ideal for improving the prognosis of patients with CCA 6,7 . Thus, new therapeutic targets or strategies are urgently needed 8 .
Sirtuin-3 (SIRT3), is a nicotinamide adenine dinucleotide (NAD + )-dependent deacetylase in mitochondria 9 . SIRT3 regulates cell survival, oxidative stress, apoptosis, metabolism, and through deacetylation 10 . SIRT3 plays an important role in tumor promotion or inhibition in numerous cancers [11][12][13][14] . However, the possible biological role of SIRT3 in CCA and its potential molecular mechanism were not available in previous studies. Although previous study has reported the anti-Warburg role of SIRT3 and has con rmed that SIRT3 inhibits the CCA progression by regulating the downstream pathway HIF1α/PDK1/PDHA1 15 . The Warburg effect represents the process of increased glucose uptake and aerobic glycolysis, which leads to the production of lactic acid, which becomes an early phenotype of cancer cells 16,17 . The speci c underlying regulatory mechanism of SIRT3 on CCA inhibition remains unclear.
Hypoxia inducible factor 1α, (HIF1α), has been found to regulate the expression of glycolytic-related genes 18 . Recently increasing evidence shows that HIF1α is involved in malignant disease progression 19,20 . For example, increased HIF-1α levels are related to a poor prognosis in breast cancer 21 . HIF1α protein was highly expressed in prostate, lung, and colon cancer 22 . In cervical cancer, the deacetylation of HIF1α is increased by SIRT2, which reduces HIF1α degradation and increases HIF1α stability by further ubiquitination 23 . Thus, in this study, we hypothesized that HIF1α was modi ed by SIRT3 deacetylation to make it more susceptible to degradation and inhibit the proliferation of CCA. To test our hypothesis, we further explored the molecular mechanism and functional signi cance of SIRT3 to provide a novel research strategy for CCA treatment.

Materials And Methods
Cell culture The CCA cell lines (HuCCT1, RBE and HCCC-9810) and a normal intrahepatic biliary epithelial cell (HIBEpiC) were used in this study (Chulabhorn Research Institute, Thailand). These CCA cell lines were maintained in RPMI-1640 supplemented with 10% FBS (Gibco, Grand Island, USA). A humidi ed incubator with 5% CO 2 was used to culture the cells at 37°C.

Cell transfection
pcDNA3.1 targeting SIRT3, HIF1α, and the negative control pcDNA3.1 were purchased from RiboBio, Guangzhou, China. The SIRT3 and HIF1α plasmids were from obtained from the Zhao Lab of Fudan patient has given informed consent according to the established plan approved by the Ethics Committee of The First A liated Hospital of Anhui Medical University.
Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analysis TRIzol reagent (Invitrogen, USA) was used to extract Total RNA from cells or tissues. Total RNA was reverse transcribed to cDNA by Reverse Transcription Kit (Takara, Dalian, China). Three replicates of each sample were ampli ed and analyzed with a Roche Light-Cycler (Roche, Basel, Switzerland). Results were normalized using GAPDH expression. The 2 −ΔΔCt method was used to calculate relative expression level.
Cell counting Kit-8 (CCK-8) assay CCA cells were seeded into 96-well plates at the density of 2 × 10 3 cells/well. Next, cells were cultured for 24, 48, or 72 h in an incubator with 5% CO 2 . After 10 μL of CCK-8 solution at 37°C for 2.5 h was added to each well, a microplate reader was applied to determine sample absorbance at 450 nm. The optical density (OD) value was measured, and the experiments were repeated three times.
Co-Immunoprecipitation assay CCA cells were collected and dissolved in 500 μl co-IP buffer containing a a mixture of protease inhibitors (Sigma-Aldrich). Cells were prepared for extracting total proteins. Next, 1-2 μg HIF1α antibody was added into 300 μL cell lysates at 4°C for 30 minutes. Then, proteins were incubated with the beads at 4°C overnight. On the following day, the beads were washed with for three times, and then, and the proteins were prepared for western blotting.
Deacetylation assay HIF-1α and GFP proteins were expressed in HuCCT1 cells. In the deacetylation assay, the acetylated HIF-1α or SIRT3 immunoprecipitates were resuspended in 100 μl deacetylation buffer and 1 mM NAD. The Reactions were pretreated with 10 mM nicotinamide for 10 min and incubated for 2 h at 30 °C. Next, beads were placed on ice for 15 min and boiled with sample buffer. The supernatant was analyzed by Western blotting with anti-acetyllysine antibody.

Statistical analysis
Statistical analyses was performed with SPSS 19.0 software. The results presented as means ± standard deviation (SD). One-way analysis of variance (ANOVA) followed by LSD test was performed for multiple comparisons, and the student's t test was used for comparisons between two groups The data were followed by Tukey's post hoc test. p < 0.05 was considered statistically signi cant.

SIRT3 inhibits cell proliferation and EMT process
To investigate the role of SIRT3 in CCA, we rst used RT-qPCR to analyze the expression of SIRT3 in 20 pairs of CCA tissues and 20 adjacent normal tissues, which showed that SIRT3 expression was lower in CCA tissues than in adjacent normal tissues (Fig. 1A). Decreased expression of SIRT3 was observed in HuCCT1, RBE, HCCC9810 cells, compared with HIBEpiC cells (Fig. 1B). Then we overexpressed SIRT3 and found elevated expression level of SIRT3 in HuCCT1 and RBE cells via RT-qPCR analysis (Fig. 1C).
Declined cell viability induced by overexpressed SIRT3 was observed using a CCK-8 assay (Fig. 1D). The EdU incorporation assays also demonstrated that overexpression of SIRT3 reduced HuCCT1and RBE cell proliferation (Fig. 1E). Similarly, SIRT3 overexpression reduced the protein levels of PCNA, Ki67, CDK2, which were associated with cells proliferation in CCA cells (Fig. 1F). Western blotting also revealed that the SIRT3 overexpression increased E-cadherin protein level. On the contrary, proteins like N-cadherin, Slug, Twist with strong EMT capacity were suppressed (Fig. 1G). Based on the above ndings, it can be concluded that SIRT3 was downregulated in cholangiocarcinoma, and SIRT3 overexpression inhibited cell proliferation and EMT process.

SIRT3 directly interacts with HIF1α in CCA cells
To explore the interaction between SIRT3 and HIF1α, co-immunoprecipitation (Co-IP) assays were performed. We found that SIRT3 and HIF1α precipitated each other as revealed by co-immunoprecipitation (Co-IP) assays, and the correlation between SIRT3 and HIF1α was initially identi ed ( Fig. 2A). Protein level of HIF1α in lysed protein samples was observed, and we found that the protein level of HIF1α was reduced in SIRT3 overexpression group, while the protein level of HIF1α was increased in SIRT3 knockdown group in CCA cells (Fig. 2B). We analyzed the expression of HIF1α in 20 pairs of CCA tissues and 20 pairs of adjacent normal tissues by RT-qPCR, and the results showed that HIF1α was highly expressed in CCA tissues compared with adjacent normal tissues (Fig. 2C). The expression protein levels of HIF1α were also upregulated in CCA cell lines (Fig. 2D). Furthermore, HIF1α and SIRT3 expression showed a signi cant negative correlation in CCA tissues (Fig. 2E). Thus, these data suggested that SIRT3 contributed to proliferation by directly modulating HIF1α in CCA.

SIRT3 induces degradation of HIF1α via deacetylation
To further detect the relationship between SIRT3 and HIF1α, HIF1α was deacetylated by SIRT3 in HuCCT1 cells. SIRT3 knockdown showed high protein level of HIF1α as well as the HIF1α acetylation level, which indicated that SIRT3 induced HIF1α deacetylation (Fig. 3A). Then, we suppressed HIF1α synthesis by CHX (Cycloheximide) and HIF1α proteasomal degradation by MG132. In HuCCT1 cells, HIF1α degraded faster after SIRT3 overexpression, and the degradation of HIF1α was blocked by MG132 (Fig. 3B). Similarly, SIRT3 knockdown decreased the HIF1α-wt and HIF1α-mut protein stability, while sh-SIRT3 had no effect on the GFP and K709A (Fig. 3C). The results taken together indicated that HIF1α was more easily to be degraded and showed a reduction in stability through deacetylation via SIRT3 knockdown.

Overexpression of HIF1α reverses the inhibition of SIRT3 on cell proliferation and EMT process
To test the overexpression e ciency, we overexpressed HIF1α and found elevated expression levels of HIF1α in CCA HuCCT1 and RBE cells by RT-qPCR (Fig. 4A). The CCK-8 cell viability assay showed that HIF1α overexpression signi cantly reversed the inhibitory effects of SIRT3 overexpression on cell proliferation in CCA cells (Fig. 4B-C). EdU assay also showed that overexpression of HIF1α partially reversed the reduced proliferation of CCA cells induced by SIRT3 overexpression (Fig. 4D). The overexpression of HIF1α reversed the decreased levels of proliferation-related proteins (PCNA, Ki67, CDK2) induced by SIRT3 overexpression (Fig. 4E). In addition, western Blotting also showed that HIF1α overexpression reversed the promotive effect of SIRT3 overexpression on the protein and expression levels of E-cadherin, while HIF1α upregulation also reversed the suppressive effect of SIRT3 overexpression on the protein and expression levels of N-cadherin, Slug and Twist (Fig. 4F). These results further suggested that HIF1α overexpression reversed the inhibitory SIRT3 on cell proliferation and EMT process.

Discussion
Sirtuin 3 (SIRT3) was found to have the most powerful mitochondrial deacetylase activity and is able to target key proteins for the normal function and metabolism of organelles 24 . Previous studies have found that SIRT3 acted as a tumor promoter in the development of some types of cancer, such as esophageal cancer 25 , breast cancer 26 , gastric cancer 27 , colon cancer 28 . On the contrary, SIRT3 has also been described as a tumor suppressor in some cancers 29,30 . However, there are very few studies on SIRT3 in CCA. In our study, SIRT3 was lowly expressed in CCA tissues and cells. In addition, SIRT3 overexpression inhibited cell proliferation and EMT process. Next, we began to study the speci c regulation mechanism of SIRT3 in CCA.
Previously, studies have shown that SIRT3 expression is closely associated with the expression of various HIF1α-dependent genes in breast cancer patients 31 . In our study, we further con rmed the correlation between SIRT3 and HIF1α in CCA. HIF1α expression was negatively correlated with SIRT3 expression in CCA tissues, suggesting that SIRT3 contributed to cell proliferation by regulating HIF1α.
Acetylation is a key posttranslational modi cation and is also considered to be a key metabolic enzyme 32,33 . Notably, in cervical cancer, SIRT2 increases HIF1α deacetylation, which reduces HIF1α degradation and increases HIF1α stability through further ubiquitination 23 . Deacetylases have also been investigated as potential therapeutic targets for CCA 34 . Our study con rmed that SIRT3 can indeed control glycolytic metabolism by modifying the stability and activity of HIF1α via deacetylation. We found that HIF1α modi ed with deacetylation was more easily degraded, which downregulated the expression level of HIF1α and decreased the stability of HIF1α. These results suggested that SIRT3 directly inhibited the function of HIF1α. Subsequently, we performed a rescue assay in which HIF1α overexpression reversed the suppressive effect of SIRT3 overexpression on cell proliferation and EMT process. Therefore, the decrease of HIF1α stability can enhance the inhibitory effect of SIRT3 on CCA cell proliferation and EMT process. In subsequent experiments, we will explore the effects of deacetylation and ubiquitination on the SIRT3/HIF1α axis and the mechanism of SIRT3 in various stages of tumor development.
In conclusion, SIRT3 exerted its inhibitory role on the proliferation and EMT process of CCA cells via regulating HIF1α. This nding may provide new research strategy for the CCA treatment.

Declarations
Samples were obtained only after the patient has given informed consent according to the established plan approved by the Ethics Committee of The First A liated Hospital of Anhui Medical University.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used during the current study are available from the corresponding author on reasonable request.  Representative images of immunoblotting and quantitative analysis of PCNA, Ki67 and CDK2 were examined by western blot and RT-qPCR. (G) Expression and protein levels of E-cadherin, N-cadherin, Slug, and Twist were examined by western blot and RT-qPCR. *p < 0.05, ** p < 0.01, *** p < 0.001.