Introduction
Biliary Tract Cancers (BTCs) are a heterogeneous group of aggressive solid tumors arising from the malignant transformation of the epithelial cells facing the lumen of the biliary tree. BTCs includes a cluster of different biliary tumors classified into three subtypes, according to their anatomic location: intrahepatic, perihilar and distal cholangiocarcinoma (CCA) and gallbladder cancer (GBC) [
1]. BTCs represent the second most common primary hepatic malignancy after hepatocellular carcinoma, accounting for 10–15% of all liver cancers and approximately 3% of all gastrointestinal tumors [
2,
3]. Their incidence is increasing globally mainly as a result of a rise in intrahepatic CCA cases. For resectable cases, surgery followed by 6-month adjuvant chemotherapy is the current standard of care endorsed by international guidelines [
4], while for the advanced disease the association of PDL1 inhibitors to cisplatin/gemcitabine have changed the first-line treatment paradigm [
4]. Although surgery is a potential curative option, only approximately 30% of all BTC patients is suitable for this therapeutic approach. Conventional chemotherapy is generally ineffective, and patients develop resistance to palliative treatments [
4].
In the past few years, the advent of high-throughput technologies namely next-generation sequencing has enabled the identification of several druggable molecular vulnerabilities in BTC. Among them, mutations in isocitrate dehydrogenase 1 gene (IDH1) [
5] and fusions/rearrangements of the fibroblast growth factor receptor 2 gene (FGFR2) [
6,
7,
8] have been the most successfully exploited in clinical trials with currently three targeted agents approved for pretreated advanced CCA harboring these genomic aberrations. Despite latest genomic-driven progress, only a minority of patients may benefit from targeted therapies and even in responding patients acquired resistance almost invariably occurs leading to a meager overall survival in advanced BTC.
These observations highlighted the urgent unmet need for a better understanding of tumor biology, particularly of non-genetic molecular mechanisms in order to identify novel biomarkers predicting prognosis and treatment efficacy as well as novel therapeutic targets.
During the last decade, many studies have focused on the potential regulatory role of non-coding RNA (ncRNA) in human cancer [
9,
10]. MicroRNAs are endogenous small non-coding RNAs that play a pivotal role in regulating gene expression. MiRNAs may function as either oncogenes or tumor suppressors depending on the tumor type. Recent evidence shown that dysregulated miRNAs can affect the hallmarks of cancer, including proliferation, survival, invasion and metastasis, angiogenesis and drug resistance [
11]. Additionally, an increasing number of studies have identified miRNAs as potential new biomarkers for human cancer. Compelling evidence have demonstrated that miRNAs are dysregulated in BTC and promote cholangiocarcinogenesis and tumor progression [
12].
The miR-181 family belongs to a conserved group of miRNAs regulating cell proliferation, apoptosis, autophagy, mitochondrial function, and immune response [
13,
14]. The miR-181 family is composed of four members: miR-181a, miR-181b, miR-181c and miR-181d, located into genetic clusters on chromosomes 1, 9 and 19. Several studies have demonstrated that miR-181 family members have a fundamental role in tumor progression and drug resistance [
15,
16]. In fact, several evidence have shown that miR-181a and miR-181b are able to modulate different relevant biological processes by regulating targets involved in cancer-associated pathways in several human cancers [
17,
18]. However, little is known regarding the other two members, the miR-181c and miR-181d [
19,
20].
In this study, an exhaustively analysis of miRNA expression profiles of BTC patients and cell lines led us to identify miR-181c and -181d as significantly dysregulated in both human samples and in vitro models of BTC. We found that low miR-181c/d expression was associated with poor prognosis and lack of efficacy of treatments in BTC patients. Our study revealed the potential role of miR-181c/d as tumor suppressor in BTC and support the hypothesis that this might be exploited for new therapeutic approaches. Indeed, overexpression of miR-181c/d reduced cell growth and increased sensitivity to chemotherapy. We demonstrated that the miR-181c/d functional role is determined by binding to their target SIRT1 (Sirtuin 1). Based on the proposed integrative network, the interplay between miR-181c/d and SIRT1 had a negative regulatory effect on several important metabolic-related pathways modulating drug resistance in BTC. Overexpression of both miR-181c and miR-181d may represent innovative therapeutic tools to ameliorate the clinical management of BTC patients. Furthermore, we suggest that expression levels of miR-181c/d may be a useful biomarker to monitor and predict response to treatment in BTC patients.
Materials and methods
Human samples
The human BTC tissues were collected under approval of the Ethical Committee for Clinical Research at Azienda Ospedaliera Universitaria, Modena, Italy (ID 465/2018/SPER/AOUMO). The study protocols conformed to the ethical guidelines of the 1975 Declaration of Helsinki, as per ethical approval given by the institutional review board. Written informed consent was obtained from all participants. Diagnosis of BTC was established on radiological findings and was pathologically proven in all patients. Inclusion Criteria, Subject Demographics, Sex and Biological Variables and Clinicopathological characteristics were shown in Table
S1.
Disease recurrence was defined as the presence of imaging‐proven disease.
Total RNA was extracted from the FFPE (Formalin-Fixed-Paraffinn-Embedded) samples from 62 tumor and the matched nontumor component after microscopic dissection (Table
S1). MiRNAs were extracted from FFPE tissues before the beginning of therapy and at disease progression through Maxwell® RSC RNA FFPE Kit (Promega, Madison, WI, USA) and processed with Maxwell® RSC Instruments (Promega, Madison, WI, USA) according to the manufacturer’s instructions [
21,
22].
Cell culture and transfections
EGI-1, RBE and HUCCT1 cells were obtained from DSMZ (Braunschweig, Germany) and American Type Culture Collection (ATCC) (Manassas, VA, USA) and cultured in DMEM and RPMI 1640 medium (Invitrogen, Karlsruhe, Germany) with L-Glutamine, 10% fetal bovine serum (FBS; Invitrogen, Carlsbad, CA, USA), 100 U/mL penicillin and 50 µg of streptomycin, at 37 ◦C in humidified 5% CO2 atmosphere. H69 cells (obtained by Lonza, Siena, IT), an SV40-transformed (i.e. immortalized) normal human cholangiocyte cell line were grown in media containing DMEM/F12 (Sigma-Aldrich, St. Louis, MO) supplemented with fetal bovine serum (CellGro, Manassas, VA), 10% fetal bovine serum (FBS; Invitrogen, Carlsbad, CA, USA), 100 U/mL penicillin and 50 µg of streptomycin, at 37 ◦C in humidified 5% CO2 atmosphere. Gemcitabine- and Cisplatin-resistant cells (EGI-DR) were obtained by growing EGI cell lines in medium containing increased concentrations of Gemcitabine and Cisplatin for 4 weeks. EGI cell lines resistant to Gemcitabine 0,3 µM plus Cisplatin 3 µM were selected. Gemcitabine and Cisplatin were purchased from Sigma-Merck KGaA (Darmstadt, Germany). Authentication of cell lines was done by Eurofins (Milan, Italy) and confirmed by online STR-matching analysis (
www.dsmz.de/fp/cgi-bin/str.html). SiRNA against SIRT1, MiR-181c and -181d mimics, inhibitors and negative controls were purchased from Dharmacon (Lafayette, CO, USA). Transfection with 40 nM mimics or inhibitors and/or negative control mimics was performed with Lipofectamine 2000 reagent (Invitrogen, Karlsruhe, Germany).
RNA extraction and quantitative reverse transcription PCR
Total RNA was extracted from the cell samples using Trizol reagent (Invitrogen, California, CA, USA) or in alternative Maxwell RSC simplyRNA Cells (Glomax Discover System, Promega, Madison, WI, USA) and processed with Maxwell RSC Instrument (Glomax Discover System, Promega, Madison, WI, USA) following manufacturer’s instructions [
21]. Single-stranded complementary DNA (cDNA) was generated using High Capacity cDNA Reverse Transcription Kit (Life Technologies, Paisley, UK). Quantitative Reverse Transcription PCR (qRT-PCR) was performed in LightCycler 96 (Roche, Penzberg, Germany) using LightCycler FastStart DNA Master SYBR Green I (Roche, Penzberg, Germany) and each validated primer. Validated qRT-PCR primers of SIRT1, GAPDH and ACTB were purchased from Eurofins (Milan, Italy).
QRT-PCR for miRNA was performed using a TaqMan MicroRNA assay kit (Applied Biosystems, Foster City, CA, USA) and specific primer sets for U6 snRNA (Assay ID: 001973), mature miR-181c (Assay ID: 000482), miR-181d (Assay ID: 001098), (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions [
22]. Reverse transcription was performed with TaqMan microRNA reverse transcription kit (Life Technologies, Paisley, UK), and microRNA expression assessed by qPCR with TaqMan assay (Life Technologies, Paisley, UK). We used ACTB and U6 snRNA as internal normalizers for mRNA and miRNA, respectively.
The primer sequences are as follows:
SIRT1: REV CTGCCACAAGAACTAGAGGATAAG; For AGTGGCAAAGGAGCAGATTAG.
GAPDH: REV AGGGGTGCTAAGCAGTTGGT; FOR ATGTTCGTCATGGGTGTGAA
ACTINB: REV GACTCCATGCCCAGGAGGG; FOR AAGAGCTACGAGCTGCCTGA
Proliferation assay
MTS assays were performed using tetrazolium-based CellTiter 96 AQueous One Solution Cell Proliferation assay (Promega, Madison, WI, USA). Cells were seeded in 96-well plates at 10,000 cells per well for overnight incubation. Following adhesion of cells to the well, cells were treated with the experimental treatments indicated. Control groups were exposed to the same concentration of DMSO (Dimethyl sulfoxide; Sigma-Merck KGaA, Darmstadt, Germany). At the designated time-points, plates were read at the absorbance of 492 nm on a microplate reader (Glomax Discover System, Promega, Madison, WI, USA). Relative cell viability of an individual sample was calculated by normalizing their absorbance to that of the corresponding control. All experiments were done in triplicate. The GI50 was determined from the regression of a plot of the logarithm of the concentration versus percent inhibition by Prism GraphPad (GraphPad Software 8.0, La Jolla, CA, USA) using the Dose–Response One-Site Model.
The cells were seeded onto 24-well plates (2000 cells/well) and treated with indicated compounds or vehicle control (Dimethyl sulfoxide, DMSO; Sigma-Merck KGaA, Darmstadt, Germany) for 7–10 days. After washing and fixation, the cells were stained with 0.5% Crystal Violet (Bio Basic Inc., Markham, Canada) in 25% methanol for 10 min. Cell colonies were then photographed and counted.
Annexin V-FITC/PI staining assay
Cells were treated with indicated compounds and DMSO control for 48 h, and the cell apoptosis was measured by using Annexin V-FITC/PI Apoptosis Detection Kit (BD Biosciences, San Diego, CA, USA). Cells were suspended in binding buffer, stained with Annexin V-FITC and PI for 15 min at room temperature in dark. Apoptotic cells were analyzed using a BD Accuri C6 Analyzer (BD Biosciences, San Diego, CA, USA).
Western blot analysis
The cells were lysed with RIPA buffer containing protease and phosphatase inhibitors (Selleckchem). The protein concentration was tested with a BCA kit, and appropriate amounts of protein were prepared for SDS-PAGE and then transferred to PVDF membrane (Millipore, MA, USA). The membranes were blocked for 1 h with 5% non-fat dry milk and then incubated with rabbit anti-SIRT1 (1:1000; #9475; Cell Signaling Technology Europe, Netherlands); TFAM (1:1000; #8076; Cell Signaling Technology Europe, Netherlands); CATALASE (1:1000; #12,980; Cell Signaling Technology Europe, Netherlands); SOD2 (1:1000; #MA1-106; Invitrogen); UQCRC2 (1:1000; #ab14745; Abcam); MTCO2 (1:1000; #ab110258; Abcam); ASCL4 (1:500; #sc-365230; Santacruz); Actin-b (1:5000, A5441; Sigma) mABs were used as loading controls. The results were imaged using a gel image analysis system (Bio-Rad, California, USA) according to the manufacturer’s instructions.
RNAseq and small-RNAseq library preparation and deep sequencing
For RNA-seq analysis, libraries were prepared according to the manufacturer’s instructions (QuantSeq 3’ mRNA-Seq Library Prep Kit FWD for Illumina, Lexogen GmbH, Wien, Austria) starting from 250 ng of total RNA [
16,
21]. Quality control of library templates was performed using a High Sensitivity D1000 Screen Tape (Agilent Technologies, Santa Clara, CA, USA) on a TapeStation 4200 (Agilent Technologies, Santa Clara, CA, USA). The Qubit quantification platform (Qubit 2.0 Fluorometer, Life Technologies/Thermo Fisher Scientific, MA, USA) was used to normalize samples for the library preparation. Using multiplexing, up to 87 samples were combined into a single lane to yield sufficient coverage. The sequencing was carried out in collaboration with Next Generation Diagnostic S.r.l. (Pozzuoli, Naples, Italy). Libraries were sequenced by single-end chemistry on an NovaSeq6000 platform (SP 100 cycles; Illumina, Cambridge, UK). Each library was loaded at a concentration of 250 pM, which was previously established as optimal. An average yield of ~ 4.5 Mb was obtained per sample.
Small RNA-Seq was performed by using a Small RNA-Seq Library Prep Kit (Lexogen, GmbH, Wien, Austria), according to manufacturer protocols. To verify the quantity and quality of RNA extracted before the library construction, the samples were tested using 2100 Bioanalyzer “total RNA pico bioanalyzer kit”(Agilent, Santa Clara, CA, USA) and the concentrations of all RNA solutions were determined using a Qubit 2.0 fluorometer (Life Technologies/Thermo Fisher Scientific, MA, USA). The RNA samples were used to produce cDNA libraries using the Small RNASeq Library Prep kit (Lexogen GmbH, Wien, Austria) according to the user manual. Input RNA was primarily ligated to a 30 adapter then, after removing excess 30 adapter by column purification, it was ligated to 50 adapter and the excess was removed. In the second step the RNA, flanked by 50 and 30 adapters, was converted into cDNA during the PCR amplification step through the adjunction of multiplexing indices. These indices are used to distinguish the single samples after the pooling phase. The library product was cleaned-up and concentrated with gel-based purification protocol. This step removes linker-linker artefacts (120 bp) and long library fragments (> 200 bp). The presence of linker-linker artefacts in the library may reduce the power of amplification and the consequent sequencing results. A size selection using a 6% Tris–Glycine-SDS Precast Polyacrylamide Gel (Life Technologies/Thermo Fisher Scientific, MA, USA) was used to achieve this goal according to the Library prep instruction manual and using a Gel extraction module compatible with PAGE gel purification (Lexogen GmbH, Wien, Austria). The sequencing step was performed with NGS technologies using Illumina Novaseq 6000 SP kit (100 cycles) produced by Illumina (Illumina, Cambridge, UK) [
16].
Computational analysis of deep sequencing data
A data analysis was performed using the pipeline already established at the Bioinformatics and Statistics Core Facility at TIGEM [
16]. Briefly, the reads were trimmed to remove adapter sequences and low-quality ends and reads mapping to contaminating sequences (e.g., ribosomal RNA, phIX control) were filtered out. Reads were aligned and assigned to Human ENSEMBLE transcripts and genes (hg38 reference) by using RSEM version 1.2.25 with standard parameters. The threshold for statistical significance chosen was False Discovery Rate (FDR) < 0.05. The Gene set enrichment analysis (GSEA) was then performed restricting the output to the collection of “hallmark”, “Biocarta”, “Wikipathway”, “Reactome” and “KEGG” gene sets part of the Molecular Signatures Database (MSigDB v7.0). The threshold for statistical significance chosen in the GSEA was False Discovery Rate (FDR) < 0.25. The expression of differentially induced/suppressed genes (FDR < 0.05) was validated by RT-PCR. RNA-seq and smallRNA-seq data are included within this paper and its Supplementary Information files. Data have been deposited in NCBIs Gene Expression Omnibus (GEO) and a Provisional GEO accession number has been requested. The accession number will be provided from the corresponding author upon request.
Integrated analysis of miRNA and mRNA
Correlation analysis of mir181c/d and predicted target genes based on the mRNA:miRNA interaction network was performed by considering the expression profile in BTC patients. The network was constructed by considering the negative statistically significantly correlation (Pearson’s correlation) between miRNA and mRNA expression, which included mir181c/d and 352 predicted target mRNAs. To create the interaction network, we converted the read counts into counts-per-million (CPM) values for both mRNA and miRNA data. Cytoscape tool was used to build network of interaction (Cytoscape 3.8.2 version,
https://cytoscape.org). Targetscan 7.2 (
http://www.targetscan.org/vert_72/) and miRDB were used to identify predicted targets. To explore the correlation of candidate target genes with miRNAs, Cytoscape software (Cytoscape 3.8.2 version,
https://cytoscape.org) was employed to construct and analyzed the miRNA-hub gene network.
Prediction of miRNA-mRNA regulatory network
The list of targeted genes included in this analysis was derived from the functional gene set analysis performed with the up-regulated genes identified by the meta-analysis. Briefly, we extracted the list of genes associated with the functional gene ontology terms and biological pathways predicted by ClueGO. The list of conserved miRNAs predicted to target these genes was identified by Targetscan 7.2 (
http://www.targetscan.org/vert_72/). To reconstruct the miRNA-mRNA regulatory network, the expression intensity of each miRNA and their respective targets was extracted from the normalized expression (RNAseq and small-RNAseq) data of the BTC cohort. Pairwise correlation was computed between miRNA and their targets. MiRNAs that have a negative correlation (R equal or less than − 0.5) with at least one target were selected for the network reconstruction. Network visualization was performed in Cytoscape software (Cytoscape 3.8.2 version,
https://cytoscape.org) considering miRNA and mRNA as source and target nodes respectively.
Reporter assay
To prepare the reporter constructs, the 3′UTR of target genes containing the putative miR-181c/d oligonucleotides containing SIRT1 3’UTR binding sites were designed and purchased to Eurofins (Milan, Italy). The oligonucleotides were cloned at the downstream of the luciferase gene in the pGL3-luc vector (Promega, Madison, WI, USA) as schematically depicted in Fig.
5 [
16]. For generation of the mutant reporters, oligonucleotides containing mutated binding sites were used. Oligonucleotides sequences were reported below. H69 cells were co-transfected with reporter plasmid (250 ng), pRL-CMV-Renilla plasmid (25 ng) and 40 nM mimic-miRNA and/or negative control mimics (Dharmacon, Lafayette, CO, USA) in 96-well plates using Lipofectamine 2000 (Invitrogen, California, CA, USA) according to the manufacturer’s instructions [
16]. After 48 h of transfection, luciferase activity was measured using a Dual Luciferase Reporter Assay system (Promega, Madison, WI, USA) according to the manufacturer’s instruction. Firefly luciferase activity was normalized to Renilla luciferase activity.
SIRT1 3UTR_s: 5’pCTAGACCACCAGCATTAGGAACTTTAGCATGTCAAAATGAATGTTTACTTGTGAACTCGATAGAGCAAGGAAACC T-3′.
SIRT1 3UTR_anti: 5’p CTAGAGGTTTCCTTGCTCTATCGAGTTCACAAGTAAACATTCATTTTGACATGCTAAAGTTCCTAATGCTGGTGG T-3′
SIRT1 3UTR-MUT_s:
5’pCTAGACCACCAGCATTAGGAACTTTAGCATGTCAATATTTTTTTTTACTTGTGAACTCGATAGAGCAAGGAAACC T-3′.
SIRT1 3UTR-MUT_anti: 5’p CTAGAGGTTTCCTTGCTCTATCGAGTTCACAAGTAAAAAAAAATTTTGACATGCTAAAGTTCCTAATGCTGGTGG T-3′
Statistical analysis
Statistical tests including Student’s t test and one-way ANOVA with Bonferroni’s multiple comparison test were performed with Prism GraphPad (GraphPad Software 8.0, La Jolla, CA, USA). Data are presented as mean +/− SD. Results were considered statistically significant if p < 0.05. The calculation of the GI50 values were performed with Prism GraphPad Prism (GraphPad Software 8.0, La Jolla California, USA) and followed a nonlinear regression model applied to the sigmoidal dose–response curves of the cell viability data. The values were log transformed before fitting the model. 4.13. For survival data, Kaplan–Meier curves were plotted and compared using a logrank test. All tests were two-sided. To evaluate the effect of mir-181b and -181c on survivall in BTC patients, we performed a log-rank test between high and low-mi181a/b risk groups. We stratified patients into high- and low-risk subtypes with median risk score. Kaplan–Meier curve was used to compare the OS (overall survival) and PFS (Progression free survival) between the high- and low risk groups. All analysis was performed with R software (version 4.0.2) using survival R package. A p-value < 0.05 was considered statistically significant.
Discussion
In this study, using both patient-derived material and preclinical models, we demonstrated for the first time that miR-181c and -181d, act as tumour suppressive miRNAs in BTC. Indeed, we showed that their downregulation is associated with advanced tumour stage and poor clinical outcome. Mechanistically, the forced overexpression of miR-181c and -181d reduced tumour proliferation and counteracted chemoresistance in in vitro BTC models by targeting histone deacetylase SIRT1. Our findings support the translational potential of these miRNAs as predictive biomarkers and therapeutic tools in BTC.
The miR-181 family consists of four highly conserved members: miR-181a, miR-181b, miR-181c, and miR-181d [
53]. The sequence homology and difference among miR-181a-d and their gene loci on different chromosomes were elucidated by Indrieri et al. [
53]. The miR-181 family members are evolutionarily conserved among the vertebrate lineage with high homology implicating their functional redundancy [
53]. Notably, miR-181s are aberrantly expressed in tumor tissues and exhibit oncogenic or tumor-suppressive properties in a cancer-specific manner. The involvement of several members miRNA-181 family in cancer and drug-resistance has been demonstrated in a widely range of tumors [
53,
54]. Human miR-181c and -181d are transcribed in clusters at genomic locus on chromosomes 19 and are involved in the regulation of multiple cellular functions [
38]. MiR-181c and -181d are emerging as onco-suppressors in multiple cancer types, including non-small cell lung cancer (NSCLC) [
55], breast cancer [
56], hepatocellular carcinoma [
57], bladder cancer [
58], esophageal squamous cell carcinoma [
59,
60], brain cancers [
61]. However, miR-181d have also been shown to exhibit oncogenic properties in colorectal cancer (CRC) [
62], lung adenocarcinoma [
63], glioma [
64]. Several other microRNAs presented dual-opposite oncogenic or suppressive functions which may be attributed to organ-specific actions or different cellular contexts of tumors.
Numerous studies evidenced that miRNAs regulate several mechanisms driving drug-resistance [
65]. Moreover, therapeutic strategies aiming to overexpress oncosuppressor-miRNAs or inhibit onco-miRs may represent efficient approaches to overcome drug-resistance or increasing the efficacy of therapeutic regimens. The advent of precision medicine has paved the way for the introduction of miRNAs as biomarkers to predict therapeutic responses and cancer patient survival [
6,
65,
66]. In recent years, many studies have proposed miRNAs as potential biomarkers as diagnostic, prognostic, and predictive tools in BTC [
12,
67,
68]. Respect to other class of molecules, miRNA possesses several advantages: i) miRNA-isolation and detection in liquid biopsies (from blood, urine, and other bodily fluids) can be easily performed; ii) miRNAs possess a high specificity and sensitivity for the tissue or cell type of origin; iii) miRNA detection present high time- and cost-effectivenes in comparison to other available biomarkers; iv) the detection can be multi-plexed: a multi-miRNA profile (miRNA signature) provides a non-invasive method for the diagnosis and prediction of disease progression and treatment efficacy [
12].
Based on their biological relevance, members of the miR-181 family have been investigated as prognostic and predictive biomarkers in CRC [
69], oral [
70], esophageal [
71], endometrial [
72], and NSCLC [
54], breast [
73], ovarian [
74] and pancreatic [
75] cancers. Recently, our group shown that miR-181a and -181b may represent new biomarkers for therapeutic intervention in refractory melanomas [
16].
We found here that the downregulation of both miR-181c and -181d is a commonly occurring feature within the tumour area of BTC and that was correlated with a significantly shorter survival in our clinically-annotated patient cohort.
Furthermore, the increased expression of both miRNA levels was associated with a higher likelihood of achieving a response to standard chemotherapy treatment in the advanced-disease setting. We can speculate that the improved sensitivity to cytotoxics driven by miR-181c and -181d is a key factor for the better survival outcome experienced by BTC with upregulation of these miRNAs.
Through functional experiments we indeed showed that the forced expression of both miRNAs increased sensitivity to gem/cis in resistant BTC cell models, markedly suppressing cell viability and colony formation and inducing cell death. Although preliminary and in need of validation, this finding strongly suggests potential use of miR-181c and -181d as therapeutic tools. Potential applications of multidrug resistance-related miRNAs, including miR-181a/b have been already investigated in cancer [
16,
53,
76‐
80].
Here, we report the first evidence that SIRT1 is regulated by miR-181c/d in BTC and to a greater extent in cancer. Our experiments, suggest that miR-181c and -181d directly interact with the SIRT1 3′UTR to regulate SIRT1 expression. In support of this hypothesis, the expression level of miR-181c/d was inversely correlated with the SIRT1 transcript level in BTC specimens. Furthermore, the suppressive effect of miR-181c/d on SIRT1 expression sensitizes BTC cells to gem/cis treatment. Finally, high expression of miR-181c/d (and, thus, lower SIRT1) was associated with improved PFS in BTC patients.
SIRT1 is a nicotinamide adenosine dinucleotide (NAD/NADH)-dependent histone deacetylase (HDAC) that has been linked to control of longevity, gene silencing, cell-cycle progression, apoptosis, and energy homeostasis [
43]. Its expression has been shown to be altered in cancer cells, and it targets both histone and non-histone proteins for deacetylation and thereby alters metabolic programs. Interestingly, many of the metabolic pathways that are influenced by SIRT1 are also altered in tumor development. Not only does SIRT1 have the potential to regulate oncogenic factors, it also orchestrates many aspects of metabolism and lipid regulation and recent reports are beginning to connect these areas. SIRT1 influences pathways that provide an alternative means of deriving energy (such as fatty acid oxidation and gluconeogenesis) when a cell encounters nutritive stress and can therefore lead to altered lipid metabolism in various pathophysiological contexts [
43]. Recent studies have revealed the significance of SIRT1 as oncogene in BTC indicating that overexpression of SIRT1 is correlated with cell proliferation and progression of BTC [
42]. The same authors proposed the use of SIRT1 inhibitors as effective therapeutic approach against BTC [
42,
81]. Importantly, increasing evidence suggests that SIRT1 is a major player in cancer drug resistance [
41]. Our results provide evidence of a negative correlation between miR-181c/d expression and their target SIRT1 in patients with a positive outcome and responding to targeted therapies, thus strongly suggesting that miR-181c/d exert their tumor-suppressive action by targeting SIRT1 and thus silencing its pro-tumorigenic functions. Numerous SIRT1 small molecule inhibitors have been developed or are under development [
82]. However, they shown encouraging anti-cancer effect against cancers, but a limited specificity and potency [
82]. Our results proposed miR-181b and -miR181c as new generation of SIRT1 inhibitors with promising effects on overcoming cancer drug resistance and improve therapeutic outcomes of cancer treatment. Interesting, other member of miR-181 family were reported to mitochondrial-related metabolism, and antioxidant response [
16,
53,
83].
Finally, our analysis of the functional role of miR-181c/d also investigated their regulatory network, thus identifying key signaling pathway regulating BTC progression and resistance. In particular, a network of genes, involved in transcriptional regulation of metabolic processes and cell death, has been found to be negatively enriched in BTC with high expression levels of miR-181c/d. Metabolic reprogramming is a hallmark of cancer and allows tumour cells to meet the increased energy demands required for rapid proliferation, invasion, and metastasis [
84]. Targeting altered tumour metabolism is an emerging therapeutic strategy for cancer treatment, including BTC. We reported that miR-181c and -181d might regulate several key-metabolic related fundamental functions, notably associated to a tumor suppressive role and to a positive response to therapeutic agents in BTC
81. Our results strongly suggest the potential use of miR-181c/d in targeting SIRT1-mediated metabolism in BTC.
In summary, we demonstrated here for the first time that miRNA-181c and -181d play a tumour suppressive role in BTC by targeting the histone deacetylase SIRT1. As such, they are downregulated in tumour areas and in higher disease stages. Interestingly, we showed that their expression confers a better prognosis and more favourable response to standard chemotherapy both in the patient cohort and preclinical models. Collectively, these data support the transformative clinical application of miRNA-181c and -181d as novel predictive biomarkers and therapeutic tools in BTC.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.