Circular RNA circ-MTHFD1L induces HR repair to promote gemcitabine resistance via the miR-615-3p/RPN6 axis in pancreatic ductal adenocarcinoma

Abstract

Background: Chemoresistance of pancreatic cancer is the main reason for the poor treatment effect of pancreatic cancer patients. Exploring chemotherapy resistance-related genes has been a difficult and hot topic of oncology. Numerous studies implicate the key roles of circular RNAs (circRNAs) in the development of pancreatic cancer. However, the regulation of circRNAs in the process of pancreatic ductal adenocarcinoma (PDAC) chemotherapy resistance is not yet fully clear.

Methods: Based on the cross-analysis of the Gene Expression Omnibus (GEO) database and the data of our center, we explored a new molecule, hsa_circ_0078297 (circ-MTHFD1L), related to chemotherapy resistance. QRT-PCR was used to detect the expression of circRNAs, miRNAs, and mRNAs in human PDAC tissues and their matched normal tissues. The interaction between circ-MTHFD1L and miR-615-3p/RPN6 signal axis was confirmed by a series of experiments such as Dual-luciferase reporter assay, fluorescence in situ hybridization (FISH) RNA immunoprecipitation (RIP) assays.

Results: Circ-MTHFD1L was significantly increased in PDAC tissues and cells. And in PDAC patients, the higher the expression level of circ-MTHFD1L, the worse the prognosis. Mechanism analysis showed that circ-MTHFD1L, as an endogenous miR-615-3p sponge, upregulates the expression of RPN6, thereby promoting DNA damage repair and exerting its effect on enhancing gemcitabine chemotherapy resistance. More importantly, we also found that Silencing circ-MTHFD1L combined with olaparib can increase the sensitivity of pancreatic cancer to gemcitabine.

Conclusion: Circ-MTHFD1L maintains PDAC gemcitabine resistance through the miR-615-3p/RPN6 signal axis. Circ-MTHFD1L may be a molecular marker for the effective treatment of PDAC.

Keywords: Circ-MTHFD1L; Gemcitabine resistance; Pancreatic cancer; RPN6; miR-615-3p.

Fig. 1

Discovery and characteristics of a new gemcitabine resistance-related circRNA in pancreatic cancer. A-B Two gemcitabine-resistant PDAC cell lines (PANC-1-GR, BxPC-3-GR) were established and confirmed using the IC50 and Cell viability C Volcano plots showing the up-regulated (red) and down-regulated (blue) circRNAs in PDAC from GSE79634, GSE69362 and circRNA-seq of PANC-1 (WT vs GR) (fold change > 2 or < 0.5, P < 0.05). D The Venn diagram shows the intersection of upregulated circRNA among the three results. E Schematic diagram of chromosomal location and formation of circ-MTHFD1L. F PCR detection of circ-MTHFD1L and linear transcript of MTHFD1L by divergent and convergent primers in cDNA and genomic DNA (gDNA). G Relative RNA level of circ-MTHFD1L and linear MTHFD1L treated with RNase R. H Relative RNA level of circ-MTHFD1L and linear MTHFD1L treated with actinomycin D at the indicated time. I The expression levels of circ-MTHFD1L in PADC cells (8988, AsPC-1, BxPC-3, PANC-1, SW1990) and normal human pancreatic duct epithelial (HPDE) were analyzed by qRT-PCR. J The expression levels of circ-MTHFD1L in tumor and matched non-tumor tissues from 96 gemcitabine-treated PADC patients were analyzed by qRT-PCR. K-L Kaplan-Meier survival analysis showed that PADC patients with low and high circ-MTHFD1L expression of overall survival and PFS. The median circ-MTHFD1L expression was used as the cutoff value. M Relative RNA level of circ-MTHFD1L in PFS < 10 months group and PFS ≥ 10 months group were analyzed by qRT-PCR. N Localization of circ-MTHFD1L (red) in PANC-1-GR cells and BxPC-3-GR cells using fluorescence in situ hybridization(FISH). Cell nuclei were counterstained with DAPI (blue). Scale bar, 50 μm. Data are shown as mean ± SD. *P < 0.05; ** P < 0.01; *** P < 0.001, between the indicated groups

Fig. 2

Circ-MTHFD1L is crucial for maintaining gemcitabine resistance. A Quantitative RT-PCR (qRT-PCR) analysis of circ-MTHFD1L in the two gemcitabine-resistant cell lines compared with their parental cells. B Three knockdown sequences were designed to target the circ-MTHFD1L junction site C-D The overexpression and knockdown efficiency of circ-MTHFD1L were confirmed by qRT-PCR analysis of circ-MTHFD1L and MTHFD1L mRNA in PANC-1-GR. E-H The proliferative ability of stably transfected PANC-1-GR or BxPC-3-GR cells dealing with gemcitabine was investigated via colony formation assays (E, F) and CCK-8 assays (cell viability and IC50) (G, H). Representative colony formation images are shown (E), and the numbers of colonies were summarized (F). I Flow cytometry analysis of the cell cycle progression of stably transfected PANC-1-GR or BxPC-3-GR cells dealing with gemcitabine was performed. Representative images and quantification of the results are presented. Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001, between the indicated groups

Fig. 3

Circ-MTHFD1L acts as a miRNA sponge for miR-615-3p. A RIP assay showed the association between AGO2 and circ-MTHFD1L. Top, IP efficiency of the anti-AGO2 antibody in Western blots. Bottom, relative enrichment represents RNA associated with AGO2 relative to the input control. An IgG antibody served as a control. B The potential target miRNAs of circ-MTHFD1L were predicted in the ENCORI, circInteractome, circbank, and circAtlas databases. C QRT-PCR was utilized to determine the relative expression levels of miR-516b-5p and miR-615-3p in precipitates from PANC-1-GR and BxPC-3-GR cell lysate pulled down by the circ-MTHFD1L probe or oligo probe. D Schematic diagram of the circ-MTHFD1L-WT and circ-MTHFD1L-Mut luciferase vectors. E Relative luciferase activities in PANC-1-GR and BxPC-3-GR cells co-transfected with circ-MTHFD1L-WT or circ-MTHFD1L-Mut miR-615-3p mimic, inhibitor, or corresponding negative control. F RIP assay was carried out with anti-AGO2 antibodies or IgG in PANC-1-GR cells after transfection with the miR-615-3p mimic or mimic NC, and qRT-PCR was then performed to detect the enrichment of circ-MTHFD1L and miR-615-3p. G The co-localization of circ-MTHFD1L (red) and miR-615-3p (green) was observed by fluorescence in situ hybridization (FISH) in PANC-1-GR cells. Cell nuclei were counterstained with DAPI (blue). Scale bar: 50 μm. H Volcano plots showing the up-regulated (red) and down-regulated (blue) miRNAs in PDAC from GSE112264, GSE113486 (fold change > 2 or < 0.5, P < 0.05). I The expression levels of miR-615-3p in PADC cells (8988, AsPC-1, BxPC-3, PANC-1, SW1990) and normal human pancreatic duct epithelial (HPDE) were analyzed by qRT-PCR. J The expression levels of circ-MTHFD1L in tumor and matched non-tumor tissues from 96 gemcitabine-treated PADC patients were analyzed by qRT-PCR. K Relative RNA level of circ-MTHFD1L in PFS < 10 months group and PFS ≥ 10 months group were analyzed by qRT-PCR. L The correlation between circ-MTHFD1L and miR-615-3p in 96 gemcitabine-treated PDAC patients was analyzed by Pearson correlation analysis. M-N The relative expression of miR-615-3p in PANC-1-GR or BxPC-3-GR cells was analyzed by qRT-PCR after indicated transfection. Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001, between the indicated groups

Fig. 4

MiR-615-3p reverses gemcitabine resistance of circ-MTHFD1L. A Quantitative RT-PCR (qRT-PCR) analysis of miR-615-3p in the two gemcitabine-resistant cell lines compared with their parental cells. B-E The proliferative ability of the indicated cells dealing with gemcitabine was investigated via colony formation assays (B, C) and CCK-8 assays (cell viability and IC50) (D, E). Representative colony formation images are shown (B), and the numbers of colonies were summarized (C). F Flow cytometry analysis of the cell cycle progression of the indicated cells dealing with gemcitabine was performed. Representative images and quantification of the results are presented. Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001, between the indicated groups

Fig. 5

The target gene of miR-615-3p, RPN6, is upregulated in pancreatic cancer and is associated with a poor prognosis. A The potential target mRNAs of miR-615-3p were predicted in ENCORI, miRDB, mirDIP, miRTarBase, and TargetScan databases. B Schematic diagram of the RPN6-WT and RPN6-Mut luciferase vectors. C Relative luciferase activities in PANC-1-GR and BxPC-3-GR cells co-transfected with RPN6-WT or RPN6-Mut and the miR-615-3p mimic, inhibitor or corresponding negative control. D-E The relative expression of RPN6 in PANC-1-GR or BxPC-3-GR cells was analyzed by qRT-PCR and western blot after indicated transfection. F-G The expression levels of RPN6 in PADC cells (8988, AsPC-1, BxPC-3, PANC-1, SW1990) and normal human pancreatic duct epithelial (HPDE) were analyzed by qRT-PCR and western blot assays. H-I The expression levels of RPN6 in tumor and matched non-tumor tissues from 96 gemcitabine-treated PADC patients were analyzed by qRT-PCR (H). There are representative images of RPN6 protein levels in PDAC tumors and adjacent normal tissues by western blot assays (I). J-K Kaplan-Meier survival analysis showed that PADC patients with low and high circ-MTHFD1L expression of overall survival and PFS. The median circ-MTHFD1L expression was used as the cutoff value. L Relative RNA level of RPN6 in PFS < 10 months group and PFS ≥ 10 months group. M-N IHC staining shows the abundances of RPN6 in the indicated groups (M). The IHC scores of RPN6 were further quantified (N). O Box-plot of RPN6 expression in TCGA PDAC tumor and matched TCGA normal pancreatic tissues along with GTEx data. P-Q Overall survival and DFS of PDAC patients (N = 178) from TCGA project with high or low RPN6 expression levels. R-S The correlation between RPN6 and circ-MTHFD1L and miR-615-3p in PDAC was analyzed by Pearson correlation analysis. Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001, between the indicated groups

Fig. 6

RPN6 enhances gemcitabine resistance by promoting cell cycle progression and DNA damage repair. A-B The bubble chart shows the significant enrichment of Gene Ontology (GO) term and KEGG pathway via Gene Set Enrichment Analysis (GSEA) grouped by RPN6 expression level based on RNA-seq data of TCGA PDAC datasets. C Gene set enrichment analysis (GSEA) of TCGA datasets showed that higher RPN6 expression was significantly associated with cell cycle, double-strand break repair, and homologous PDAC. D QRT-PCR and Western blot analysis of circ-MTHFD1L in the two gemcitabine-resistant cell lines compared with their parental cells. E-F PANC-1-GR cells with stable RPN6 overexpression or knockdown were constructed. G-J The proliferative ability of stably transfected PANC-1-GR or BxPC-3-GR cells dealing with gemcitabine was investigated via colony formation assays (G, H) and CCK-8 assays (cell viability and IC50) (I, J). Representative colony formation images are shown (G), and the numbers of colonies were summarized (H). K Flow cytometry analysis of the cell cycle progression of stably transfected PANC-1-GR or BxPC-3-GR cells dealing with gemcitabine was performed. Representative images and quantification of the results are presented. L Western blot analysis was performed to assess the protein levels of RAD50, RAD51, BRCA1, BRCA2, PCNA, P53 and P21 after knocking down or overexpressing RPN6 in PANC-1-GR and BxPC-3-GR cells. GAPDH was used as a loading control. M Confocal microscopic analysis of γH2AX staining showed that RPN6 decreased the distribution of γ-H2AX foci. Representative images and quantification of the results are presented. N The comet assay confirmed that RPN6 was closely related to the DNA repair capacity. Representative images and quantification of the results are presented. Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001, between the indicated groups

Fig. 7

Circ-MTHFD1L maintains gemcitabine resistance through miR-615-3p/RPN6 axis. A-F The proliferative ability of the indicated cells dealing with gemcitabine was investigated via colony formation assays (A-F) and CCK-8 assays (cell viability and IC50) (E-F). Representative colony formation images are shown (A, C), and the numbers of colonies were summarized (B, D). G Flow cytometry analysis of the cell cycle progression of the indicated cells dealing with gemcitabine was performed. I Confocal microscopic analysis of γH2AX staining and the comet assay in the indicated cells was evaluated. The quantification of the results is presented. Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001, between the indicated groups. J The RAD50, RAD51, BRCA1, BRCA2, PCNA, P53 and P21 protein levels in PANC-1-GR and BxPC-3-GR cells from different groups were determined by western blot analysis. GAPDH was used as a loading control

Fig. 8

Pancreatic cancer CDXs with circ-MTHFD1L silencing is more sensitive to olaparib combined with gemcitabine therapy. A-D BALB/c nude mice (4–6 week-old males) were subcutaneously injected with stably transfected PANC-1-GR cells. Four weeks after implantation, the mice were treated with Gemcitabine (50 mg/kg/twice a week) by tail vein injection. Mice were sacrificed on week 8, and xenografts were isolated and measured. E-H BALB/c nude mice (4–6 week-old males) were subcutaneously injected with PANC-1-GR cells and treated with gemcitabine (50 mg/kg/twice a week) tail vein injection starting on week 4. Six weeks later, the most resistant xenograft was disaggregated and implanted subcutaneously into BALB/c nude mice (4–6 week-old males) as the second GR-CDX. Two weeks after implantation, the second GR-CDX generation mice were treated with sh-circ-MTHFD1L lentivirus (twice a week) and randomized into four groups on week 3. The four groups of mice were treated as follows: gemcitabine alone, olaparib alone, gemcitabine combined with olaparib, and negative control. Mice were sacrificed on week 6, and xenografts were isolated and measured. Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001, between the indicated groups. I Schematic illustration of the circ-MTHFD1L/miR-615-3p/RPN6 axis in PDAC cells