Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Circular RNA AKT3 upregulates PIK3R1 to enhance cisplatin resistance in gastric cancer via miR-198 suppression

Circular RNA AKT3 upregulates PIK3R1 to enhance cisplatin resistance in gastric cancer via... Background: Cisplatin (CDDP) treatment is one of the most predominant chemotherapeutic strategies for patients with gastric cancer (GC). A better understanding of the mechanisms of CDDP resistance can greatly improve therapeutic efficacy in patients with GC. Circular RNAs (circRNAs) are a class of noncoding RNAs whose functions are related to the pathogenesis of cancer, but, in CDDP resistance of GC remains unknown. Methods: circAKT3 (hsa_circ_0000199, a circRNA originating from exons 8, 9, 10, and 11 of the AKT3 gene) was identified by RNA sequencing and verified by quantitative reverse transcription PCR. The role of circAKT3 in CDDP resistance in GC was assessed both in vitro and in vivo. Luciferase reporter assay, biotin-coupled RNA pull-down and fluorescence in situ hybridization (FISH) were conducted to evaluate the interaction between circAKT3 and miR-198. Functional experiments were measured by western blotting, a cytotoxicity assay, clonogenic assay and flow cytometry. Results: The expression of circAKT3 was higher in CDDP-resistant GC tissues and cells than in CDDP-sensitive samples. The upregulation of circAKT3 in GC patients receiving CDDP therapy was significantly associated with aggressive characteristics and was an independent risk factor for disease-free survival (DFS). Our data indicated that circAKT3 promotes DNA damage repair and inhibits the apoptosis of GC cells in vivo and in vitro. Mechanistically, we verified that circAKT3 could promote PIK3R1 expression by sponging miR-198. Conclusions: circAKT3 plays an important role in the resistance of GC to CDDP. Thus, our results highlight the potential of circAKT3 as a therapeutic target for GC patients receiving CDDP therapy. Keywords: Cisplatin resistance, Gastric cancer, circAKT3, Circular RNA, miR-198, PIK3R1 Background months [3]. In patients with histologically confirmed Gastric cancer (GC) is the most common malignant advanced GC and who are chemotherapy-naïve, cisplatin tumor of the digestive tract in East Asia and the third (CDDP) and fluorouracil-based chemotherapies were leading cause of cancer-related death worldwide [1, 2]. deemed as first-line treatments [4]. However, patients At present, the main treatments for advanced GC are always acquired drug resistance after several cycles of systemic chemotherapy and palliative surgery, but the CDDP-based treatment. Thus, chemotherapy resistance has overall median survival after treatment is only 8 to 11 limited overall clinical efficacy in patients [5, 6]. To improve GC patient survival, illuminating the molecular mechanism * Correspondence: xuzekuan@njmu.edu.cn underlying CDDP resistance in GC is essential. Xiaoxu Huang, Zheng Li, Qiang Zhang and Weizhi Wang contributed The cytotoxicity of CDDP is mediated by its inter- equally to this work. 1 action with DNA to form DNA adducts. Intracellular Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, Jiangsu province, CDDP primarily binds to nuclear DNA with high affinity China and can physically interact with mitochondrial DNA Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Huang et al. Molecular Cancer (2019) 18:71 Page 2 of 20 (mtDNA), which is involved in the activation of several recurrence during CDDP-based therapy; both definitions signaling pathways and apoptosis [7–9]. In recent years, followed standard CDDP response definitions published studies have shown that the PI3K/AKT signaling path- elsewhere [14]. Forty-four samples (Cohort 1) were used way could suppress cell apoptosis and facilitate cell for circRNAs validation, and another 105 samples (Cohort survival. This PI3K/AKT signaling function is crucial in 2) were used to quantify circAKT3 levels and to analyze the regulation of chemotherapy resistance of cancer cells the correlation between circAKT3 expression and [10, 11]. Activated PI3K/AKT signaling promotes the outcomes after R0 excision in patients undergoing phosphorylation of caspase-3 and prevents the activation CDDP-based chemotherapy. The samples from cohorts 1 of caspase-3 and the inhibition of apoptosis [12]. and 2 were obtained in 2013–2016 and 2007–2011, Circular RNAs (circRNAs), a category of noncoding respectively. The grouping of the ROC curve was based RNAs (ncRNAs), play a crucial role in the process of on the median relative expression of circAKT3. Detailed transcriptional and posttranscriptional gene expression information is listed in Additional file 1: Table S1. [13]. Recently, circRNAs were found to function as com- petitive endogenous RNAs (ceRNAs) to sponge micro- Cell culture RNAs (miRNAs) and then suppress their functions, The CDDP-sensitive cell lines SGC7901 and BGC823 as indicating a novel mechanism for regulating miRNA well as their CDDP-resistant strains (SGC7901CDDP activity and providing a promising mode of action for and BGC823CDDP, respectively) were maintained in circRNAs. As miRNAs regulate a series of biological RPMI 1640 medium (Wisent, Shanghai, China) supple- processes, circRNA sponge activity will affect these mented with 10% fetal bovine serum (FBS) (Wisent, biological behaviors as well [13]. miRNAs are a large Biocenter, China) (Additional file 2: Figure S1A). 293 T class of short (~ 22 nt) ncRNAs that posttranscriptionally cells were cultured in DMEM with high glucose (Gib- regulate gene expression through direct base pairing to co-BRL, Carlsbad, CA, USA) supplemented with 10% target sites within mRNAs. circRNAs can affect miRNA FBS. 293 T, SGC7901CDDP, BGC823 and SGC7901 cells activities by competing for miRNA-binding sites [13]. were purchased from the Cell Bank of Type Culture Col- However, the function of circRNAs as miRNA sponges lection of Chinese Academy of Sciences, and has not been clearly elucidated in GC resistance to BGC823CDDP cells were established as previously de- CDDP. scribed [15]. To investigate the potential roles of circRNAs in the regulation of CDDP resistance in GC, we performed RNA miRNA targets prediction of circAKT3 sequencing (RNA-Seq) and verified thousands of distinct To predict the miRNA-binding sites of circAKT3, we circRNAs in CDDP-sensitive and CDDP-resistant GC used the bioinformatic databases miRanda, PITA and cells from humans. Through functional gain and loss RNAhybrid. Filtering restrictions were as follows: (1) experiments, we further observed that hsa_circ_0000199, total score ≥ 140, total energy < 17 kcal/mol; (2) com- which originates from exons 8, 9, 10, and 11 of the AKT3 bined interaction energy (△△G) < 10; and (3) minimum gene and is termed circAKT3, was significantly free energy (MFE) ≤ 20 kcal/mol. Detailed information is upregulated in both CDDP-resistant GC tissues and listed in Additional 3: Dataset S1. CDDP-resistant cells. Furthermore, we found that circAKT3 modulates CDDP sensitivity by sponging miR-198 that suppresses PIK3R1 expression via activation RNA preparation, treatment with RNase R, and PCR of the PI3K/AKT pathway in GC. Total RNA was extracted from GC cells or tissues using TRIzol Reagent (Invitrogen, 15,596,018). RNase R treat- Methods ment was carried out for 15 min at 37 °C using 3 U/mg Patients and samples RNase R (Epicenter). For Quantitative real-time PCR In total, 149 GC tissues (cohorts 1, 2) were obtained from (RT-PCR), 500 ng of treated RNA was directly reverse the First Affiliated Hospital of Nanjing Medical University. transcribed using Prime Script RT Master Mix (Takara, All samples were collected in accordance with HIPAA Japan) and either random or oligo(dT) primers. Reverse guidelines and approved institutional protocols. Patients transcription of miRNA was performed using a New received treatment with standard CDDP-based thera- Poly(A) Tailing Kit (ThermoFisher Scientific, China). peutic regimens after surgery. Disease-free survival (DFS) mRNA was reverse transcribed into cDNA with a Prime- was defined as the time interval between gastrectomy (R0 Script RT Master Mix Kit (Takara, RR036A, Japan). excision) and the time of either disease recurrence or cDNA was amplified using Universal SYBR Green Mas- disease-associated death. CDDP resistance was defined as ter Mix (4,913,914,001, Roche, Shanghai, China). The tumor relapse during CDDP-based chemotherapy after R0 CT value was measured during the exponential growth excision, and CDDP sensitivity was defined as no tumor phase. Relative gene expression levels were determined Huang et al. Molecular Cancer (2019) 18:71 Page 3 of 20 Fig. 1 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 4 of 20 (See figure on previous page.) Fig. 1 circAKT3 expression is increased in CDDP-resistant GC cells and tissues. a Validated expression of 10 circRNAs in the tissues from 44 GC patients using RT-qPCR. b Expression levels of circAKT3 in CDDP-resistant and their matched sensitive parental cell lines (SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823) normalized to GAPDH expression. c The existence of circAKT3 was validated by Sanger sequencing. The red arrow shows the “head-to-tail” splicing sites of circAKT3. d The existence of circAKT3 was validated in SGC7901CDDP and BGC823CDDP cell lines by RT-PCR. Divergent primers amplified circAKT3 in cDNA but not in genomic DNA (gDNA). GAPDH served as a negative control. e RNA from SGC7901CDDP and BGC823CDDP cells was treated with or without RNase R for RT-qPCR. The relative levels of circAKT3 and AKT3 mRNA were normalized to the values measured in the mock-treated cells. f Levels of small nucleolar RNA (U6, as a positive control for the nuclear fraction), GAPDH (positive control for cytoplasmic fraction), AKT3 mRNA and circRNAs from the nuclear and cytoplasmic fractions of SGC7901CDDP cells. g RNA stability of circular and linear transcripts of AKT3 and of 18S rRNA in SGC7901CDDP cells. h Representative images of RNA FISH of circAKT3 expression in SGC7901CDDP cells, which show that circAKT3 is predominantly localized to the cytoplasm. Nuclei were stained with DAPI. Scale bar, 10 μm. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 -△△CT using the 2 method. The primers used are listed in MgCl2, 1 mM CaCl and 250 mM sucrose) supple- Additional file 1: Table S2. mented with protease inhibitors. After centrifugation at 600 × g for 5 min at 4 °C, the resultant supernatant was Isolation of nuclear and cytoplasmic fractions collected as the cytoplasmic fraction and mixed with an SGC7901CDDP and BGC823CDDP cells were lysed on equal volume of TRIsure reagent. After the pellet was ice for 10 min in 0.3% NP-40/NIB-250 buffer (15 mM washed with NIB-250, the nuclei were lysed in TRIsure Tris–HCl (pH 7.5), 60 mM KCl, 15 mM NaCl, 5 mM reagent. Fig. 2 The circAKT3 expression level is correlated with poor prognosis in GC patients receiving CDDP therapy. a Expression levels of circAKT3 in tissue samples of 105 GC patients (CDDP-resistant and CDDP-sensitive groups) normalized to GAPDH expression. b Kaplan-Meier survival curves of DFS for patients with high (n = 53) or low (n = 52) circAKT3 expression. The median circAKT3 expression value was used as the cutoff. c ROC curves of circAKT3. d Multivariate analyses of hazard ratios for DFS. The results are presented as the mean ± SEM Huang et al. Molecular Cancer (2019) 18:71 Page 5 of 20 Table 1 Correlation of relative circAKT3 expression with the Oligonucleotide transfection clinicopathological characteristics of 105 patients accepted siRNA and miRNA mimics and inhibitors were synthe- cisplatin-based chemotherapy with gastric cancer sized by GenePharma (Shanghai, China). The sequences Characteristics Number No. of patients P value used are listed in Additional file 1: Tables S3 and S4. high low circAKT3 circAKT3 Transfection was carried out using Lipofectamine RNAi- MAX (Life Technologies) according to the manufac- Age(y) turer’s instructions. < 60 62 34 28 0.283 ≥ 60 43 19 24 RNA pull-down Gender A pull-down assay was performed as described Male 80 37 43 0.121 previously [17, 18]. The biotin-labeled circAKT3 probe Female 25 16 9 was synthesized by RiboBio (Guangzhou, China). In brief, 1 × 10 circAKT3-overexpressing GC cells were Tumor size(cm) harvested, lysed, and sonicated. The circAKT3 or oligo < 3.5 56 21 35 0.004** probe was incubated with streptavidin-coupled ≥ 3.5 49 32 17 Dynabeads (Invitrogen) at 30 °C overnight to generate Histological grade probe-bound Dynabeads. After the treated beads were Well-moderately 33 11 22 0.017* washed with wash buffer, the RNA complexes bound Poorly-signet 72 42 30 to the beads were eluted and disrupted with lysis buf- fer and proteinase K prior to RT-PCR or RT-qPCR. Clinical stage Biotinylated probes sequences used in this study (see II 39 13 26 0.007** Additional file 1:Table S5). III 66 40 26 T classification Luciferase reporter assay T1-T2 31 9 22 0.004** 293 T, SGC7901CDDP and BGC823CDDP cells were T3-T4 74 44 30 seeded in 24-well plates and cotransfected with corre- sponding plasmids and miRNA mimics in triplicate. At N classification 48 h after transfection, luciferase reporter assays were N0 14 5 9 0.232 conducted using a dual-luciferase reporter assay system N1-N3 91 48 43 (Promega, Madison, WI) according to the manufac- Cisplatin chemosensitivity turer’s instructions. Relative luciferase activity was Sensitive 92 41 51 0.001** normalized to Renilla luciferase activity. Resistant 13 12 1 Fluorescence in situ hybridization (FISH) Vector construction The double FISH assay was performed in The method for overexpressing circRNAs was reported SGC7901CDDP cells and GC tissues as previously de- previously [16]. For the construction of circAKT3 scribed [16, 19]. Biotin-labeled probes specific to cir- overexpression plasmids, human circAKT3 cDNA was cAKT3 and Dig-labeled locked nucleic acid miR-198 amplified using PrimerSTAR Max DNA Polymerase Mix probes were used in the hybridization (Exiqon, Vedbaek, (Takara, RR036A, Japan) and inserted into the pCD5-ciR Denmark). The sequences are listed in Additional file 1: vector (Greenseed Biotech Co, Guangzhou, China). The Table S6, FISH probes sequences used in this study. The pCD5-ciR vector contains a front circular frame and a signals of the biotin-labeled probes were detected using back circular frame. Transfection was carried out using Cy5-conjugated streptavidin (Life Technologies), and the Lipofectamine 2000 (Invitrogen) according to the manu- signals of the Dig-labeled miR-198 probes were detected facturer’s instructions. The luciferase reporter containing using a tyramide-conjugated Alexa 488 fluorochrome the circAKT3 sequence in the 3′-UTR was constructed by TSA kit. Nuclei were counterstained with 4,6-diamidi- subcloning the circAKT3 fragment into the region directly no-2-phenylindole. Images were acquired on a Leica downstream of a cytomegalovirus promoter-driven firefly TCS SP2 AOBS confocal microscope (Leica Microsys- luciferase (FL) cassette in a pCDNA3.0 vector. Mutations tems, Mannheim, Germany). CircAKT3 and miR-198 ex- of each miRNA-binding site in the circAKT3 sequence pression levels were evaluated by the proportions and were created using a Mut Express II Fast Mutagenesis Kit intensities of the positive cells detected within 5 fields of (Vazyme, Nanjing, China). The mutations were introduced view on every slide (400-fold magnification). Proportion in both the circAKT3-expressing vector and the luciferase scores were assigned as follows: < 10% = 0, 10–25% = 1, reporter containing the circAKT3 sequence. 26–50% = 2, 51–75% = 3 and > 75% = 4. Intensity scores Huang et al. Molecular Cancer (2019) 18:71 Page 6 of 20 Fig. 3 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 7 of 20 (See figure on previous page.) Fig. 3 Downregulation of circAKT3 facilitates cisplatin sensitivity of CDDP-resistant GC cells in vitro. a Illustration showing the siRNA targeting the back-splice junction (si-circ-1 and si-circ-2). b RT-qPCR results for circular and linear transcripts of AKT3 in SGC7901CDDP cells treated with or without siRNA (NC, negative control; si-NC, control oligonucleotides with scramble sequence; si-circ-1 and si-circ-2, oligonucleotides targeting the back-splice junction). c Relative cell viability of NC, si-NC- transfected or si-circ-1-transfected SGC7901CDDP cells in the presence of CDDP at the indicated concentrations for 48 h. d Colony-forming ability of the NC, si-NC-transfected and si-circ-1-transfected SGC7901CDDP cells in the absence (Vehicle) or presence of CDDP (6 μM) for 48 h. e The apoptotic rates of NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 in the absence (Vehicle) or presence CDDP (6 μM) for 48 h were visualized by flow cytometry. f Western blot analysis shows apoptotic proteins in NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 upon CDDP treatment (6 μM) for 48 h (GAPDH was used as the loading control). g Immunofluorescence staining of γH2AX foci in NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 at 2 h after CDDP treatment (6 μM). Scale bars, 10 μm. h Percentage of cells containing > 10 γH2AX foci in NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 at 0 to 8 h after CDDP treatment (6 μM) removal. i Western blot analysis of γH2AX and BRCA1 expression in NC SGC7901CDDP and BGC823CDDP cells and cells transfected with si-NC or si-circ-1 after CDDP treatment (6 μM) removal. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 were assigned as follows: 0 = no staining, 1 = weak, 2 = Apoptosis assay moderate, 3 = strong and 4 = significantly strong. Cell apoptosis was detected using a PI/Annexin V-FITC Apoptosis Detection Kit (BD Pharmingen, 556,547) according to the manufacturer’s instructions. Briefly, Western blot analysis after GC cells were treated with CDDP at the indicated For western blot analysis, cells were extracted using a concentrations for 48 h in 6-well plates, they were har- protein extraction kit (Key Gene, KGP9100). Lipid pro- vested and resuspended in 300 ml of binding buffer. teins were added into 8, 10, 12% or 15% gels, subjected Next, 5 μl of Annexin V-FITC and 5 μl of PI were added to 120 V to promote migration, and then transferred to the suspensions, and the cells were incubated in the onto nitrocellulose membranes. The membranes were dark at 4 °C for 15 min. The samples were subsequently blocked with 5% BSA in TBST buffer and incubated with analyzed with a flow cytometer (Gallios, Beckman, specific primary antibodies at 4 °C overnight. The next USA). day, membranes were washed 3 times for 15 min in TBST and incubated with secondary antibodies for 2 h at room temperature. HRP substrate (WBKL0100, Milli- Actinomycin D assay pore, USA) was used to detect the protein bands (Mo- The Actinomycin D assay was performed as previously lecular Imager, ChemiDoc XRS+, BIO-RAD, USA), and described [16]. SGC7901CDDP and BGC823CDDP cells the band intensities were quantified using Image-Pro were seeded in 5 wells in 24-well plates (5 × 10 cells per Plus software (Mediacy, USA). Detailed information of well). Twenty-four hours later, the cells were exposed to antibody used in this study (see Additional file 1: Actinomycin D (2 μg/ml, Abcam, ab141058) for 0 h, 6 h, Table S7). 12 h, 18 h and 24 h. The cells were then harvested, and the relative RNA levels of circAKT3 and AKT3 mRNA were analyzed by RT-qPCR and normalized to the values Cytotoxicity assay measured in the group in the 0 h group (mock The cytotoxicity assay was performed as previously treatment). described [15]. Cell viability was measured using Cell Counting Kit-8(CCK8)following the manufacturer’s directions (Dojindo, Kumamoto, Japan). Immunofluorescence staining Cells seeded onto coated cover slips growth for 24 h, Clonogenic assay then treated with CDDP, and harvested the cells at 0, 2, A clonogenic assay was performed as previously de- and 8 h. The cells were fixed with 4% paraformaldehyde scribed [15]. At 48 h after transfection, BGC823CDDP, at room temperature for 15 min and then permeabilized SGC7901CDDP, BGC823 and SGC7901 cells were cul- with PBS containing 0.25% Triton X-100 for 10 min. tured with CDDP at the indicated concentrations for 3 Next, the cells were blocked with 1% BSA for 20 min h. Then, the cells were harvested, seeded into six-well before incubation with primary antibodies at room plates (500 cells per well) and cultured for an additional temperature for 2 h. After the cells were washed with 2 (BGC823CDDP and SGC7901CDDP cells) or 3 weeks PBS, they were incubated with appropriate secondary (BGC823 and SGC7901 cells). For scoring the antibodies (FITC-conjugated goat anti-rabbit, Molecular colony-forming units, we fixed cells in 1 ml of methanol Probes, USA) at room temperature for 2 h. Following a for 10 min and then stained the cells with crystal violet final wash with PBS, cells were mounted with antifading for 15 min. mounting medium containing DAPI. The images were Huang et al. Molecular Cancer (2019) 18:71 Page 8 of 20 Fig. 4 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 9 of 20 (See figure on previous page.) Fig. 4 Overexpression of circAKT3 increases the resistance of CDDP-sensitive GC cells to cisplatin in vitro. a The expression levels of circAKT3 and AKT3 mRNA in SGC7901 and BGC823 cells after stable transfection of circAKT3 or empty vector plasmids were detected by RT-qPCR. b Relative cell viability of circAKT3 or vector-transfected SGC7901 cells exposed to CDDP at the indicated concentrations for 48 h. c Colony-forming ability of SGC7901 cells transfected with circAKT3 or vector in the absence (Vehicle) or presence of CDDP (1.5 μM) for 48 h. d The apoptosis rates of SGC7901 cells transfected with circAKT3 or vector after CDDP (1.5 μM) treatment for 48 h were detected by flow cytometry. e Western blot analysis of apoptotic proteins in SGC7901 cells transfected with circAKT3 or vector upon CDDP (1.5 μM) treatment for 48 h (GAPDH was used as the loading control). f Immunofluorescence staining of γH2AX foci in SGC7901 cells transfected with circAKT3 or vector at 2 h after CDDP treatment (1.5 μM). Scale bars, 10 μm. g Percentage of cells containing > 10 γH2AX foci in SGC7901 cells transfected with circAKT3 or vector at 0 to 8 h after CDDP (1.5 μM) treatment. h Western blot analysis of γH2AX and BRCA1 expression in SGC7901 cells transfected with circAKT3 or vector after CDDP treatment (1.5 μM) removal. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 captured with a Leica DMI3000B (Germany) fluores- Statistical analysis cence microscope. All experiments were performed in triplicate. Data were analyzed with SPSS 19.0 software (IBM, USA) and presented as the mean ± SEM. The statistical significance Transduction with lentivirus of the results was calculated using an unpaired Student’s SGC7901CDDP cells stably expressing circAKT3 siRNA t-test. DFS analysis was performed using the Kaplan-Meier (si-circ-1) and its negative control siRNA (si-NC) were method and log-rank test. Clinicopathological features were generated by infection with lentiviruses as previously analyzed by a χ test. A Cox proportional hazards regres- described [20]. Transfection was carried out according to sion model was used to identify independent prognostic the manufacturer’s instructions. The lentiviral expressing factors associated with DFS. Linear correlation analyzes vectors were purchased from HanBio Co. Ltd. (Shanghai, were performed to determine correlations between China). circAKT3, miR-198 and PIK3R1 expression levels. A P value< 0.05 was defined as statistically significant. Nude mouse xenograft model Results Six-week-old female BALB/c nude mice were purchased Ectopic circAKT3 expression levels are observed in CDDP- from the Laboratory Animal Center of Nanjing Medical resistant GC cells and tissues and are correlated with University and maintained under pathogen-free condi- poor prognosis in GC patients receiving CDDP therapy tions. A total of 5 × 10 SGC7901CDDP cells infected To characterize circular RNA transcripts, we conducted with lentivirus containing si-circ-1 or si-NC (2 μlof 10 RNA-Seq analysis of CDDP-resistant SGC7901 and − 1 viral genomes μl , HanBio) in 100 μl of PBS were sub- BGC823 cells (i.e., SGC7901CDDP and BGC823CDDP) cutaneously injected into different sides of the groin of and their corresponding parental strains (i.e., SGC7901 each mouse. One week after injection, we intraperitone- and BGC823), which are sensitive to CDDP. The ally injected mice with cisplatin (5 mg/kg) in PBS or PBS sequencing statistics are not shown. The analysis indi- alone three times per week. The xenograft tumors were cated that a series of circRNAs were differentially harvested after 5 weeks. The entire experimental proto- expressed in CDDP-resistant GC cells compared with col was conducted in accordance with the guidelines of the sensitive parental GC cells. We then chose the top the local institutional animal care and use committee. 20 significantly upregulated circRNAs and verify their expression levels. Detailed information of 20 candidate circRNAs in Additional file 1: Table S8 (including loca- Immunohistochemical staining (IHC) tion, genomic and spliced length). Using divergent Xenografts and GC tissues exposed to the indicated con- primers to specifically target the circular junction as well centrations of CDDP were prepared for IHC as previ- as combined quantitative reverse transcription PCR ously described [21]. Sections were identified by IHC (RT-qPCR) analysis and sequencing validation, we found Imager (DM4000B, LEIKA, Germany), and target pro- that only 10 of these circRNAs had confirmed differ- tein expression levels were evaluated by the proportions ences in expression and that circAKT3 was the most ob- and intensities of positive cells detected within 5 fields viously upregulated circRNA in CDDP-resistant patients of view on every slide (400-fold magnification). Propor- of cohort 1 (Fig. 1a and Additional file 2: Figure S1b-c). tion scores were assigned as follows: < 10% = 0, 10–25% circAKT3 (hsa_circ_0000199) has been mapped to exons =1, 26–50% = 2, 51–75% = 3 and > 75% = 4. Intensity 8, 9, 10, and 11 of the AKT3 gene (555 bp) (Additional file scores were assigned as follows: 0 = no staining, 1 = 2: Figure S1d). Consistent with the RNA-Seq results, the weak, 2 = moderate, 3 = strong and 4 = significantly expression of circAKT3 was obviously increased in strong. CDDP-resistant GC cells (Fig. 1b). Subsequently, we Huang et al. Molecular Cancer (2019) 18:71 Page 10 of 20 Fig. 5 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 11 of 20 (See figure on previous page.) Fig. 5 circAKT3 exerts its function by sponging miR-198. a & b Schematic illustration showing the overlap of the target miRNAs of circAKT3 predicted by miRanda, PITA and RNAhybrid. c & d Lysates prepared from SGC7901CDDP and BGC823CDDP cells stably transfected with circAKT3 or vector were subjected to RNA pull-down and tested by RT-PCR (C) and RT-qPCR (D). The relative level of circAKT3 was normalized to the input. GAPDH served as a negative control. e & f The relative levels of 11 miRNA candidates in SGC7901CDDP and BGC823CDDP lysates were detected by RT-qPCR. Multiple miRNAs were pulled down by circAKT3, and miR-198 was pulled down by circAKT3 in both cell lines. g Schematic illustration showing the 3′UTR of luciferase reporters containing the complete circAKT3 sequence (luc-wt) or the circAKT3 sequence with deletions of miR-198 (luc-m1-m8) binding sites. h Reporter assays showing the luciferase activity of luc-wt and luc-m1-m8 in 293 T cells cotransfected with miR-198 mimics or a scrambled oligonucleotide (control). i FISH showing the colocalization of circAKT3 and miR-198 in SGC7901CDDP cells. Nuclei were stained with DAPI. Scale bar, 10 μm. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01 verified the head-to-tail splicing of the RT-PCR product of grade, clinical stage and circAKT3 expression level (Add- circAKT3 by Sanger sequencing (Fig. 1c). Meanwhile, to itional file 1: Table S9). Subsequently, multivariate analysis exclude possibilities such as genomic rearrangements or indicated that circAKT3 expression, along with tumor size trans-splicing, several experiments were employed. First, and clinical stage, was an independent risk factor for DFS we designed convergent primers to amplify AKT3 mRNA (Additional file 1: Table S9 and Fig. 2d). and divergent primers to amplify circAKT3. Using cDNA and genomic DNA (gDNA) from SGC7901CDDP and circAKT3 facilitates CDDP resistance in vitro BGC823CDDP cell lines as templates, the circAKT3 amp- First, we designed two siRNA oligonucleotides (si-circ-1 lification product was only observed in cDNA by diver- and si-circ-2) to target the unique back-splice junction of gent primers but not in gDNA (Fig. 1d). In addition, the circAKT3 (Fig. 3a); si-circ-1 successfully knocked down fragment of the linear form of AKT3 was digested by circAKT3 expression but had no effect on the levels of en- RNase R, but circAKT3 remained after RNase R treatment dogenous linear AKT3 transcript in SGC7901CDDP and (Fig. 1e). Then, the relative expression levels of cir- BGC823CDDP cells (Fig. 3b and Additional file 4:Figure cAKT3 were detected in the cytoplasm and nucleus of S2a). Additionally, to further assess the role of circAKT3, SGC7901CDDP and BGC823CDDP cells (Fig. 1f and Add- circAKT3 was overexpressed in SGC7901 and BGC823 itional file 2: Figure S1e). The RT-qPCR results demon- cells via transfection of pCD5-ciR-AKT3 e8–11 (Fig. 4a). strated that circAKT3 was enriched in the cytoplasm. Importantly, elevated expression of circAKT3 had no ef- Moreover, we used Actinomycin D to suppress transcrip- fect on the levels of linear AKT3 mRNA, as confirmed by tion and measure the half-life of circAKT3 in RT-qPCR (Fig. 4a). circAKT3 inhibition reduced the via- SGC7901CDDP and BGC823CDDP cells; we found that bility of SGC7901CDDP and BGC823CDDP cells (Fig. 3c circAKT3 was more stable than AKT3 mRNA (Fig. 1g and Additional file 4: Figure S2b). In addition, knockdown and Additional file 2: Figure S1f). Additionally, the FISH of circAKT3 significantly decreased the number of cell results displayed a dominantly cytoplasmic distribution colonies (Fig. 3d and Additional file 4: Figure S2c) and of circAKT3 (Fig. 1h). promoted apoptosis (Fig. 3e and Additional file 4:Figure Next, we detected the expression level of circAKT3 in S2d). The phosphorylated histone family member X tissues of patients from cohort 2. Consistent with the (γH2AX) forms discrete nuclear foci and acts as a plat- RNA-Seq results, circAKT3 was significantly more highly form to recruit additional factors and enhance the DNA expressed in the CDDP-resistant GC tissues than in the repair pathway [22]. Meanwhile, circAKT3-knockdown sensitive tissues (Fig. 2a). Compared with GC patients ex- cells showed significantly more γH2AX foci than the con- pressing low levels of circAKT3, GC patients receiving trol cells at 2 h after CDDP treatment (Fig. 3g and Add- CDDP therapy and exhibiting upregulation of circAKT3 itional file 4: Figure S2e). circAKT3-knockdown cells had showed a significant association with decreased five-year a higherpercentageofactivefocirelativetothatincontrol DFS (Fig. 2b). To further verify that circAKT3 may be a cells from 0 to 8 h after CDDP treatment (Fig. 3hand therapeutic target for CDDP-resistant patients, we calcu- Additional file 4: Figure S2f). However, compared with the lated the area under the receiver operating characteristic negative control, ectopic circAKT3 expression signifi- curve (AUC) using the expression levels of circAKT3. The cantly increased cell viability and the number of cell area under the curve is 91% (Fig. 2c), suggesting that the colonies and inhibited apoptosis and the formation of expression level of circAKT3 is a good predictive bio- γH2AX foci in SGC7901 and BGC823 cells (Fig. 4b-g and marker of CDDP resistance for GC patients. Analysis of Additional file 5: Figure S3a-e). We used western blotting the clinicopathological characteristics in cohort 2 showed to investigate the underlying mechanism of these activ- that circAKT3 expression was positively related to tumor ities. In the presence of CDDP, knockdown of circAKT3 size, histological grade, clinical stage, T classification and in SGC7901CDDP and BGC823CDDP cells increased CDDP resistance (Table 1). A univariate analysis showed cleaved caspase-3 protein levels, while the levels of the that DFS was obviously related to tumor size, histological inactivated form of caspase-3 protein was decreased (Fig. Huang et al. Molecular Cancer (2019) 18:71 Page 12 of 20 Fig. 6 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 13 of 20 (See figure on previous page.) Fig. 6 PIK3R1 is a direct target of miR-198. a mRNA microarray data of the top 20 upregulated genes in SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 cells are presented as heatmaps. b Venn diagram showing 4 genes that are putative miR-198 targets computationally predicted by four algorithms (miRanda, RNAhybrid, miRWalk and TargetScan) among the top 20 upregulated genes. c Schematic of PIK3R1 3’UTR wild-type (WT) and mutant (Mut) luciferase reporter vectors is shown. d The relative luciferase activities were analyzed in 293 T cells cotransfected with miR- 198 mimics or miR-NC and luciferase reporter vectors PIK3R1 3’UTR (WT) or PIK3R1 3’UTR (Mut). e & f The expression of PIK3R1 was analyzed using RT-qPCR in tissues of cohort 1 (E) and in SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 cells (F). g The expression of PIK3R1 was analyzed using western blot in SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 cells. h & i The expression of PIK3R1 was analyzed by RT- qPCR (G) and western blot (H). SGC7901CDDP and BGC823CDDP cells were transfected with miR-198 mimic or cotransfected with the indicated vectors. j The IC50 of miR-198 was analyzed by the CCK8 assay. SGC7901CDDP cells were transfected with miR-198 mimic alone or cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 h. k The apoptosis rates of SGC7901CDDP cells transfected with miR-198 mimic alone or cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 h. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01 3f). In contrast, cleaved and inactivated caspase-3 protein RNA FISH assays revealed that circAKT3 and miR-198 levels were observed when circAKT3 was overexpressed were colocalized in the cytoplasm (Fig. 5i). (Fig. 4e). These data are consistent with a previous study reporting that CDDP-induced increases in Breast cancer type 1 susceptibility protein(BRCA1) expression leads to PIK3R1 is a direct target of miR-198 enhanced DNA damage repair (DDR) in breast cancer A microarray assay was further performed with cells [23]. After CDDP treatment, knockdown of SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 circAKT3 in SGC7901CDDP and BGC823CDDP cells cells to validate the results of the ceRNA analysis. We increased γH2AX but decreased BRCA1 protein levels. analyzed the top 20 upregulated genes according to four circAKT3 overexpression also inhibited γH2AX and pro- algorithms (miRanda, RNAhybrid, miRWalk and moted BRCA1 protein levels compared with the levels in TargetScan) prediction, and miR-198 could target the the controls (Figs. 3iand 4h). 3’UTRs of PIK3R1, CHRM3, HIPK2 and MAFB (Add- itional file 3: Dataset S3). We performed luciferase reporter assays to determine whether miR-198 directly circAKT3 exerts its function by sponging miR-198 targets these 4 genes in 293 T cells (Fig. 6c and Additional To address whether circAKT3 could sponge miRNAs in file 6: Figure S4c and e). In 293 T, SGC7901CDDP and GC cells, we selected 11 candidate miRNAs by overlap- BGC823CDDP cells cotransfected with miR-198 mimic, ping the prediction results of the miRNA recognition ele- reporter constructs containing wild-type miR-198 binding ments in the circAKT3 sequence using miRanda, PITA, sites at the PIK3R1 3′UTR exhibited decreased luciferase and RNAhybrid (Fig. 5a-b). Next, we investigated whether activity relative to that of reported constructs with candidate miRNAs could directly bind circAKT3. A mutated binding sites (Fig. 6d and Additional file 6:Figure biotin-labeled circAKT3 probe was designed and verified S4f and g). PIK3R1 protein (p85α, encoded by PIK3R1) is to pull down circAKT3 in SGC7901CDDP and the regulatory subunit of PI3K. A functional study BGC823CDDP cell lines, and the pull-down efficiency was demonstrated that PIK3R1 was highly expressed in significantly enhanced in cells with stable circAKT3 over- CDDP-resistant ovarian cancer cells, and downregu- expression (Fig. 5c-d). The miRNAs were extracted after lated PIK3R1 resensitized the abovementioned cells to pull-down, and the levels of the 11 candidate miRNAs platinum-based treatment, which reveals the promising were detected by RT-qPCR. As shown in Fig. 5e-f, in both involvement of p85α in secondary CDDP resistance [24]. SGC7901CDDP and BGC823CDDP cells, miR-198 was Compared with parental CDDP-sensitive cells, CDDP-re- abundantly pulled down by circAKT3. Furthermore, using sistant cells showed obvious increases in the expression of the RNAhybrid bioinformatics prediction tool, we calcu- PIK3R1 mRNA and protein levels (Fig. 6f and g). lated the secondary conformation of circAKT3 and Furthermore, we found that miR-198 mimics significantly miR-198 and found that there were 8 predicted binding inhibited PIK3R1 mRNA and protein levels and that domains (largest combined with a MFE > − 20 kcal/mol) ectopic PIK3R1 expression abolished the influence caused (Additional file 3: Dataset S2). Next, the results of the lu- by miR-198 overexpression (Fig. 6h and i). Subsequently, ciferase reporter assays showed that miR-198 expression the data showed that overexpression of miR-198 inhibited significantly reduced the luciferase activity of the reporter cell viability and induced apoptosis in SGC7901CDDP containing the complete circAKT3 sequence appended to and BGC823CDDP cells. However, cotransfection of the 3′-UTR of luciferase (luc-wt) compared to that of the PIK3R1 and miR-198 abrogated these effects (Fig. 6jand k reporter containing circAKT3 with mutated miR-198 and Additional file 6: Figure S4h-j and Additional file 7: binding sites (luc-m1, m2 and m8) (Fig. 5g-h). Moreover, Figure S5a-b). Huang et al. Molecular Cancer (2019) 18:71 Page 14 of 20 Fig. 7 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 15 of 20 (See figure on previous page.) Fig. 7 circAKT3 regulates PIK3R1 expression, induces cisplatin resistance and activates the PI3K/AKT signaling cascade by targeting miR-198 in vitro. a & b The expression levels of PIK3R1 were analyzed using RT-qPCR. SGC7901CDDP cells were transfected with indicated vectors alone or cotransfected the inhibitors, and SGC7901 cells were transfected with the indicated vectors and miR-198 mimics. c The IC50 of circAKT3 was analyzed by the CCK8 assay. SGC7901CDDP cells were transfected with inhibitor alone or cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 h. d The apoptosis rates of SGC7901CDDP cells transfected with indicated vectors alone or cotransfected the inhibitors upon CDDP exposure (6 μM) for 48 h. e The expression levels of PIK3R1, apoptosis markers, γH2AX, BRCA1 and PI3K/AKT signaling molecules were determined using western blotting in SGC7901CDDP cells transfected with indicated vectors alone or cotransfected with the inhibitor after CDDP treatment (6 μM). f The expression levels of PIK3R1, apoptosis markers, γH2AX, BRCA1 and PI3K/AKT signaling molecules were determined using western blotting in SGC7901 cells transfected with the indicated vectors and miR-198 mimics after CDDP treatment (1.5 μM). Proteins were isolated from the indicated cells. g The expression levels of PIK3R1, apoptosis makers, γH2AX, BRCA1 and PI3K/AKT signaling molecules were determined using western blotting in SGC7901 cells transfected with the indicated vectors and BKM20 after CDDP treatment (1.5 μM). The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01 circAKT3 regulates PIK3R1 expression, activates the PI3K/ employed a specific p110α inhibitor, BKM20, and per- AKT signaling pathway and ultimately facilitates CDDP formed western blotting to determine whether deactiva- resistance by targeting miR-198 in vitro tion of PI3K/AKT signaling can overcome the changes Cotransfection of si-circ-1 and anti-miR-198 could caused by circAKT3 overexpression. The results indi- counteract the si-circ-1-induced downregulation of cated that si-PIK3R1 and BKM20 significantly inhibited PIK3R1 in SGC7901CDDP cells (Fig. 7a). Notably, PIK3R1 and p110α levels, respectively as well as inhib- cotransfection of circAKT3 and miR-198 attenuated the ited p-AKT, reduced BRCA1 levels, increased cleaved expression of PIK3R1 compared to transfection of caspase-3 levels, and promoted γH2AX (Fig. 7g). These circAKT3 alone in SGC7901 cells (Fig. 7b). The CCK8 results suggest that circAKT3 functions by targeting and flow cytometry analyses indicated that transfection miR-198 as a ceRNA to regulate PIK3R1 expression, ac- with si-circ-1 inhibited cell viability and promoted apop- tivate the PI3K/AKT signaling cascade and facilitate tosis after CDDP treatment while cotransfection with CDDP resistance. si-circ-1 and anti-miR-198 significantly promoted cell viability and inhibited apoptosis compared with the con- circAKT3 promotes CDDP resistance of GC cells in vivo trols (Fig. 7c and d, Additional file 7: Figure S5d). To investigate the potential clinical relevance of cir- Furthermore, upregulation of circAKT3 led to enhanced cAKT3 in vivo, we subcutaneously injected cell viability, but this effect could be significantly abol- SGC7901CDDP cells with or without stable circAKT3 ished by ectopic expression of miR-198 (Additional file knockdown (Additional file 8: Figure S6e) into the dorsal 8: Figure S6a). Additionally, overexpression of circAKT3 flanks of female BALB/c nude mice and allowed the cells induced a reduction in apoptosis. However, cotransfec- to proliferate for 5 weeks. Tumor xenograft data indi- tion of circAKT3 and miR-198 mimics led to enhanced cated that circAKT3 inhibition in CDDP-resistant cells apoptosis (Additional file 8: Figure S6b). Notably, cir- can significantly decrease xenograft tumor growth and cAKT3 upregulation could inhibit γH2AX expression, as sensitize cells to CDDP treatment (Fig. 8a and b). IHC indicated by the reduced fluorescence of γH2AX, analysis of tumor xenograft samples further indicated cotransfection of circAKT3 and miR-198 mimics led to that the protein levels of γH2AX and cleaved caspase-3 abolish this effect (Additional file 8: Figure S6c and d). were notably increased, but BRCA1 was decreased upon Transfection of si-circ-1 significantly reduced PIK3R1 circAKT3 inhibition (Fig. 8c). FISH showed the colocali- expression and inhibited the expression of canonical zation of circAKT3 and miR-198 in tissues from patients PI3K/AKT signaling molecules, as shown by western with CDDP-resistant or CDDP-sensitive GC (Fig. 8d). blot, and downregulating of both circAKT3 and The FISH score data showed that the expression of cir- miR-198 abrogated these effects in SGC7901CDDP cells cAKT3 was significantly higher in CDDP-resistant GC (Fig. 7e). Similar results are presented in Additional file tissues than CDDP-sensitive GC tissues; however, 9: Figure S7a-f. Meanwhile, Transfection of circAKT3 miR-198 expression showed the opposite result (Fig. 8d). significantly increased PIK3R1 expression and induced Similarly, IHC scores and western blot analyses indicated the expression of canonical PI3K/AKT signaling mole- that PIK3R1 protein expression was obviously increased cules, as shown by western blot, and concomitant in CDDP-resistant GC tissues compared to overexpression of both circAKT3 and miR-198 abro- CDDP-sensitive GC tissues (Fig. 8e and Additional file 10: gated these effects in SGC7901 cells (Fig. 7f). Transfec- Figure S8a and b). Furthermore, correlations were identi- tion of si-PIK3R1 significantly inhibited PIK3R1 and fied between circAKT3 and miR-198 expression levels and levels of phosphorylated PI3K/AKT signaling molecules PIK3R1 protein levels in these 44 GC tissue samples (Fig. in SGC7901 cells with circAKT3 overexpression. We 8f). Based on these data, we conclude that circAKT3 Huang et al. Molecular Cancer (2019) 18:71 Page 16 of 20 Fig. 8 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 17 of 20 (See figure on previous page.) Fig. 8 circAKT3 promotes cisplatin resistance of GC cells in vivo. a Xenograft tumors of sacrificed mice with or without CDDP treatment (3 mg/kg, three times a week) at the end of the experiment. b Growth curves of subcutaneous xenograft tumors. c PIK3R1, γH2AX, cleaved caspase-3 and BRCA1 expression levels are shown in representative xenograft tumors by IHC (Left) (400x magnification, scale bars = 50 μm). Quantification of the IHC scores of PIK3R1, γH2AX, cleaved caspase-3 and BRCA1 expression levels (Right). d FISH showing the colocalization of circAKT3 and miR-198 in CDDP-resistant or CDDP-sensitive GC tissues from patients. FISH scores of circAKT3 and miR-198 were further calculated in 14 CDDP-resistant and 30 CDDP-sensitive patient tissues. Nuclei were stained with DAPI. Scale bar, 10 μm. e IHC staining of PIK3R1 in CDDP-resistant or CDDP- sensitive GC tissues from patients (400x magnification, scale bars = 50 μm). The IHC scores of PIK3R1 were further determined in 14 CDDP- resistant and 30 CDDP-sensitive patient tissues. f Three-dimensional scatter plot of circAKT3, miR-198 and PIK3R1 levels in 14 CDDP-resistant and 30 CDDP-sensitive GC tissues from patients. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 increases the tolerance of GC cells to CDDP by targeting regulates PI3K signaling in CDDP resistance. Because PIK3R1 through miR-198. the PI3K pathway is a critical player in tumorigenesis and the ubiquitously hyperactivated signaling pathway in Discussion neoplasms, its inhibition both pharmacologically and CDDP treatment is one of the most predominant genetically is considered to be the most promising strat- chemotherapeutic strategies for patients with GC [25]. egy for targeted cancer treatment [38]. The PI3K path- In this study, using RNA-Seq analysis, we determined way has been revealed as a mediator of platinum that circRNA expression is associated with CDDP resist- resistance [39, 40] For instance, AKT activation medi- ance in GC. We found a novel circular RNA termed ated resistance to caspase-independent CDDP-induced circAKT3 that was upregulated in tissue samples from apoptosis through inhibiting the apoptosis-inducing patients with CDDP-resistant GC and in CDDP-resistant factor-associated pathway [40]. One study reported that cell lines and was correlated with five-year DFS. More- PI3K/AKT activation induced the upregulation of over, circAKT3 was expressed at higher levels than other BRCA1 in tamoxifen-resistant breast cancer cells and candidate circRNAs in CDDP-resistant GC patients, resensitized them to CDDP treatment [23]. Nevertheless, which meant that it may play a more important role the underlying mechanism of how PIK3R1 mediates re- than other circRNAs in GC. sistance to chemotherapy and whether this involves ca- DNA is the recognized target for CDDP cytotoxicity in nonical PI3K signaling and downstream BRCA1 activity cancer therapy. The resultant biological process in remains to be investigated in GC. response to CDDP and other DNA-damaging therapies Based on the abovementioned studies, we conducted a is the activation of apoptosis and the destruction of ma- series of experiments and demonstrated that circAKT3 lignant cells. The most favorable evidence is the hyper- reduced CDDP-induced activation of caspase-3 and sensitivity of both eukaryotic and prokaryotic cells apoptosis, leading to enhanced DDR and resistance to deficient in DNA repair to CDDP [26]. An enhanced CDDP chemotherapy. Mechanistically, circAKT3 func- DNA repair mechanism can induce the survival of tions as a ceRNA by sponging miR-198 to abolish the damaged or mutated tumor cells, resulting in resistance suppressive effect of this miRNA on its target gene and subsequent tumor recurrence [27]. BRCA1, a PIK3R1, which activated the PI3K/AKT signaling path- tumor-suppressor gene, is widely involved in cellular way in GC cells. The study showed that PI3K/AKT path- metabolism [28], transcriptional regulation [29, 30], and way activation contributes to upregulation of the DNA epigenetic modification [31]. A growing body of repair molecule BRCA1 and leads to resistance to evidence revealed that BRCA1 has a great effect on the CDDP-based DNA-damaging chemotherapy [23]. In the modulation of CDDP resistance [32–34]. Alterative current study, circAKT3 influenced DDR in GC cells, BRCA1 expression can regulate the mitochondrial fis- implying that circAKT3 might enhance CDDP resist- sion program, which could modulate CDDP sensitivity ance through the PI3K/AKT pathway and DDR mech- [32]. Some microRNAs contribute to DNA repair and anisms in GC cells. However, there are some CDDP sensitivity through BRCA1 deregulation, includ- drawbacks in this study, and we have not yet identi- ing miR-9 [35] and miR-638 [36]. A previous study dem- fied the specific mechanism by which BRCA1 regu- onstrated that BRCA1 mRNA levels were negatively lates DDR in GC. At the same time, the reasons for associated with CDDP sensitivity in GC [37]. Currently, circAKT3 shear formation and the upstream regula- there is a lack of research on circRNAs that regulate the tory mechanism were not discussed. For the in vivo BRCA1 gene in tumor cells, including GC. The under- experiments, we did not use animal models of GC in lying mechanisms require further exploration. situ, nor did we fully simulate the process of drug re- PIK3R1 encodes the regulatory subunit of PI3K sistance. These approaches should be further pursued (p85α). Some studies have reported that p85α positively in subsequent studies. Huang et al. Molecular Cancer (2019) 18:71 Page 18 of 20 Conclusions with circAKT3 or vector upon CDDP (1.5 μM) for 48 h by flow cytometry. In conclusion, we show that circAKT3 is upregulated in d Immunofluorescence staining of γH2AX foci in BGC823 cells transfected with circAKT3 or vector at 2 h after CDDP treatment (1.5 μM). Scale bars, human GC and that it can efficiently sponge miR-198 to 10 μm. e Percentage of cells containing >10 γH2AX foci in BGC823 cells restore PIK3R1 expression. We also demonstrate that transfected with circAKT3 or vector at 0 to 8 h after CDDP treatment (1.5 downregulation of circAKT3 can effectively promote μM) removal. The results are presented as the mean ± SEM. *P<0.05, CDDP sensitivity in GC cells by targeting the miR-198/ **P<0.01, ***P<0.001. (TIF 3594 kb) PIK3R1 axis. Our results provide novel evidence that cir- Additional file 6: Figure S4. a Predicted secondary structure of circAKT3 using the Vienna RNA package. b The expression of miR-198 cRNAs function as “microRNA sponges” and highlight a was analyzed using RT-qPCR in tissues of cohort 1. c, d, e The relative lu- promising therapeutic target for the CDDP resistance of ciferase activities were analyzed in 293T cells cotransfected with miR-198 GC patients. mimics or miR-NC and luciferase reporter vectors containing the WT or Mut 3’UTR of CHRM3 (C), HIPK2 (D), and MAFB (E). f & g The relative lucif- erase activities were analyzed in SGC7901CDDP(f) and BGC823CDDP(g) Additional files cells cotransfected with miR-198 mimics or miR-NC and luciferase re- porter vectors PIK3R1 3’UTR (WT) or PIK3R1 3’UTR (Mut). h & i The expres- Additional file 1: Table S1. Detailed information of cohort 1 and 2 is sion levels of PIK3R1 (h) and miR-198 (i) in SGC7901CDDP and listed. Table S2. Primers and RNA sequences used in this study. BGC823CDDP cells after transfection of PIK3R1 plasmids or miR-198 Table S3. SiRNA sequences used in this study. Table S4. MiRNA mimics, and inhibitors sequences used in this study. Table S5. Biotinylated probes mimics were detected by RT-qPCR. j The IC50 of miR-198 was analyzed sequences used in this study. Table S6. Fish probes sequences used in by the CCK8 assay. BGC823CDDP cells were transfected with miR-198 this study. Table S7. Detailed information of antibody used in this mimic alone or cotransfected with the indicated vectors upon CDDP ex- study. Table S8. Detailed information of 20 candidate circRNAs posure (15 μM) for 48 h. The results are presented as the mean ± SEM. (including location, genomic and spliced length). Table S9. Univariate *P<0.05, **P<0.01, ***P<0.001. (TIF 1650 kb) and multivariate analysis of cohort 2. (DOCX 31 kb) Additional file 7: Figure S5. a Apoptotic flow cytometry. SGC7901CDDP Additional file 2: Figure S1. a SGC7901CDDP/SGC7901 and cells were transfected with miR-198 mimic alone or cotransfected with the BGC823CDDP/BGC823 cell viability in response to different indicated vectors upon CDDP exposure (6 μM) for 48 h b The apoptosis rate concentrations of cisplatin. b RT-PCR products with divergent primers was analyzed by flow cytometry. BGC823CDDP cells were transfected with showing a single, distinct product of the expected size. c Melting curves miR-198 mimic alone or cotransfected with the indicated vectors upon of RT-qPCR product of verified circRNAs, indicating the specificity of RT- CDDP exposure (15 μM) for 48 h. c The expression levels of miR-198 in qPCR products with no primer dimers or nonspecific amplified products. d Schematic illustrating that circAKT3 (hsa_circ_0000199) is derived from SGC7901CDDP and BGC823CDDP cells after transfection of anti-miR-198 exons 8, 9, 10, and 11 of the AKT3 gene (555 bp). e Levels of small nucle- were detected by RT-qPCR. d Apoptotic flow cytometry. SGC7901CDDP cells olar RNA (U6, as a positive control for the nuclear fraction), GAPDH (posi- transfected with indicated vectors alone or cotransfected the inhibitors tive control for the cytoplasmic fraction), AKT3 mRNA and circRNAs from upon CDDP exposure (6 μM) for 48 h. The results are presented as the mean nuclear and cytoplasmic fractions of BGC823CDDP cells. f RNA stability of ±SEM. *P<0.05, **P<0.01, ***P<0.001. (TIF 2595 kb) the circular and linear transcripts of AKT3 and 18S rRNA in BGC823CDDP Additional file 8: Figure S6. a The IC50 was analyzed by CCK8 assay, cells. The results are presented as the mean ± SEM. *P<0.05, **P<0.01, SGC7901 cells were transfected with inhibitor alone or cotransfected with ***P<0.001. (TIF 2497 kb) the indicated vectors upon CDDP exposure (1.5 μM) for 48 h. b The Additional file 3: Additional Dataset (1) miRNA target prediction of apoptosis rate was analyzed by flow cytometry. SGC7901 cells were circAKT3. Additional Dataset (2) Eight miR-198 binding sites mutation of transfected with inhibitor alone or cotransfected with the indicated circAKT3. Additional Dataset (3) mRNA target prediction of miR-198. vectors upon CDDP exposure (1.5 μM) for 48 h. c Immunofluorescence (ZIP 1800 kb) staining of γH2AX foci in SGC7901 cells transfected with inhibitor alone Additional file 4: Figure S2. a RT-qPCR results for the circular and linear or cotransfected with the indicated vectors at 2 h after CDDP treatment transcripts of AKT3 in BGC823CDDP cells treated with or without siRNA (1.5 μM). Scale bars =10 μm. d Percentage of cells containing >10 γH2AX (NC, negative control; si-NC, control oligonucleotides with scramble se- foci in SGC7901 cells transfected with inhibitor alone or cotransfected quence; si-circ-1 and si-circ-2, oligonucleotides targeting the back-splice with the indicated vectors at 0 to 8 after CDDP treatment (1.5 μM) junction). b Relative cell viability of NC BGC823CDDP cells and removal. e SGC7901CDDP cells stably expressing circAKT3 siRNA (si-circ-1) BGC823CDDP cells transfected with si-NC- or si-circ-1 after CDDP treat- and its negative control siRNA (si-NC) were generated by infection with ment at the indicated concentrations for 48 h. c Colony-forming ability of lentiviruses, the expression levels of cricAKT3 were detected by RT-qPCR. the NC BGC823CDDP cells and si-NC- or si-circ-1-transfected The results are presented as the mean ± SEM. *P<0.05, **P<0.01, BGC823CDDP cells in the absence (Vehicle) or presence of CDDP (15 μM) ***P<0.001. (TIF 2160 kb) for 48 h. d The apoptosis rates of NC BGC823CDDP cells and Additional file 9: Figure S7. a The expression levels of PIK3R1 were BGC823CDDP cells transfected with si-NC or si-circ-1 in the absence (Ve- analyzed using RT-qPCR. SGC7901CDDP cells were cotransfected with inhibi- hicle) or upon CDDP (15 μM) for 48 h by flow cytometry. e Immunofluor- tors alone or the indicated vectors. b The IC50 of circAKT3 was analyzed by escence staining of γH2AX foci in NC BGC823CDDP cells and the CCK8 assay. SGC7901CDDP cells were transfected with inhibitor alone or BGC823CDDP cells transfected with si-NC or si-circ-1 at 2 h after CDDP cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 treatment (15 μM). Scale bars, 10 μm. f Percentage of cells containing h. c The apoptosis rates of SGC7901CDDP cells transfected with inhibitor >10 γH2AX foci in NC BGC823CDDP cells and BGC823CDDP cells trans- alone or cotransfected with the indicated vectors upon CDDP exposure (6 fected with si-NC or si-circ-1 at 0 to 8 h after CDDP treatment (15 μM) re- moval. The results are presented as the mean ± SEM. *P<0.05, **P<0.01, μM) for 48 h. d Immunofluorescence staining of γH2AX foci in ***P<0.001. (TIF 5287 kb) SGC7901CDDP cells transfected with inhibitor alone or cotransfected with the indicated vectors at 2 h after CDDP treatment (6 μM). Scale bars, 10 μm. Additional file 5: Figure S3. a Relative cell viability of circAKT3- or e Percentage of cells containing >10 γH2AX foci in SGC7901CDDP cells vector-transfected BGC823 cells with CDDP treatment at the indicated transfected with inhibitor alone or cotransfected with the indicated vectors concentrations for 48 h. b Colony-forming ability of BGC823 cells trans- at 0 to 8 after CDDP treatment (6 μM) removal. f The expression levels of fected with circAKT3 or vector in the absence (Vehicle) or presence of PIK3R1, apoptosis markers, γH2AX, BRCA1 and PI3K/AKT signaling molecules CDDP (1.5 μM) for 48 h. c The apoptosis rates of BGC823 cells transfected Huang et al. Molecular Cancer (2019) 18:71 Page 19 of 20 Hospital of Wannan Medical College, Wuhu, Anhui, China. Department of were determined using western blotting in SGC7901CDDP cells transfected Neurosurgery, The First Affiliated Hospital of Nanjing Medical University, with inhibitor alone or cotransfected with the indicated vectors after CDDP Nanjing, China. treatment (6 μM). The results are presented as the mean ± SEM. *P<0.05, **P<0.01, ***P<0.001. (TIF 2258 kb) Received: 21 November 2018 Accepted: 21 February 2019 Additional file 10: Figure S8. a The expression of PIK3R1 were analyzed using western blot in tissue of cohort 1. b The expression of PIK3R1 were further determined in tissues of 14 CDDP-resistant and 30 References CDDP-sensitive patients (Relative to GAPDH). (TIF 604 kb) 1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin. 2015;65(2):87–108. Abbreviations 2. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, Parkin DM, AUC: Area under the receiver operating characteristic curve; BRCA1: Breast Forman D, Bray F. Cancer incidence and mortality worldwide: sources, cancer type 1 susceptibility protein; CCK8: Cell Counting Kit-8; methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5): CDDP: Cisplatin; ceRNAs: Competitive endogenous RNAs; circRNAs: Circular E359–86. RNAs; DDR: DNA damage repair; DFS: Disease-free survival; FBS: Fetal bovine 3. Zhao TT, Xu H, Xu HM, Wang ZN, Xu YY, Song YX, Yin SC, Liu XY, Miao ZF. serum; FISH: Fluorescence in situ hybridization; FL: Firefly luciferase; The efficacy and safety of targeted therapy with or without chemotherapy GC: Gastric cancer; gDNA: Genomic DNA; IHC: Immunohistochemical in advanced gastric cancer treatment: a network meta-analysis of well- staining; MFE: Minimum free energy; miRNAs: MicroRNAs; designed randomized controlled trials. Gastric Cancer. 2018;21(3):361–71. mtDNA: Mitochondrial DNA; ncRNAs: Noncoding RNAs; RNA-Seq: RNA 4. Koizumi W, Narahara H, Hara T, Takagane A, Akiya T, Takagi M, Miyashita K, sequencing; RT-PCR: Quantitative real-time PCR; RT-qPCR: Quantitative Nishizaki T, Kobayashi O, Takiyama W, et al. S-1 plus cisplatin versus S-1 reverse transcription PCR; γH2AX: Phosphorylated histone family member X alone for first-line treatment of advanced gastric cancer (SPIRITS trial): a phase III trial. Lancet Oncol. 2008;9(3):215–21. 5. Longley DB, Harkin DP, Johnston PG. 5-fluorouracil: mechanisms of action Acknowledgments and clinical strategies. Nat Rev Cancer. 2003;3(5):330–8. Jiangsu Key Lab of Cancer Biomarkers, Prevention and Treatment, 6. Shah MA. Update on metastatic gastric and esophageal cancers. J Clin Collaborative Innovation Center for Cancer Personalized Medicine, Nanjing Oncol. 2015;33(16):1760–9. Medical University. 7. Yang Z, Schumaker LM, Egorin MJ, Zuhowski EG, Guo Z, Cullen KJ. Cisplatin preferentially binds mitochondrial DNA and voltage-dependent anion Funding channel protein in the mitochondrial membrane of head and neck This work was supported by the National Natural Science Foundation of squamous cell carcinoma: possible role in apoptosis. Clin Cancer Res. 2006; China (81572362); the National Natural Science Foundation Project of 12(19):5817–25. International Cooperation (NSFC-NIH, 81361120398); the Primary Research & 8. Galluzzi L, Senovilla L, Vitale I, Michels J, Martins I, Kepp O, Castedo M, Development Plan of Jiangsu Province (BE2016786); the Program for Kroemer G. Molecular mechanisms of cisplatin resistance. Oncogene. 2012; Development of Innovative Research Team in the First Affiliated Hospital of 31(15):1869–83. NJMU; the Priority Academic Program Development of Jiangsu Higher 9. Galluzzi L, Vitale I, Michels J, Brenner C, Szabadkai G, Harel-Bellan A, Castedo Education Institutions (PAPD, JX10231801); 333 Project of Jiangsu Province M, Kroemer G. Systems biology of cisplatin resistance: past, present and (BRA2015474); Jiangsu Key Medical Discipline (General future. Cell Death Dis. 2014;5:e1257. Surgery)(ZDXKA2016005); University Natural Science Research Project of 10. Yu HG, Ai YW, Yu LL, Zhou XD, Liu J, Li JH, Xu XM, Liu S, Chen J, Liu F, et al. Anhui Province(KJ2018A0247); Wannan Medical College Cultivation Fund of Phosphoinositide 3-kinase/Akt pathway plays an important role in the Key Scientific Project(WK2017ZF10). chemoresistance of gastric cancer cells against etoposide and doxorubicin induced cell death. Int J Cancer. 2008;122(2):433–43. Availability of data and materials 11. Lee S, Choi EJ, Jin C, Kim DH. Activation of PI3K/Akt pathway by PTEN The datasets supporting the conclusions of this article are included within reduction and PIK3CA mRNA amplification contributes to cisplatin the article and its Additional files. resistance in an ovarian cancer cell line. Gynecol Oncol. 2005;97(1):26–34. 12. Siddik ZH. Cisplatin: mode of cytotoxic action and molecular basis of Authors’ contributions resistance. Oncogene. 2003;22(47):7265–79. ZKX, ZL and WZW conceived of the study and XXH carried out its design. 13. Kulcheski FR, Christoff AP, Margis R. Circular RNAs are miRNA sponges and XXH, QZ, BWL, LW, ZPX and XZ performed the experiments. XXH, XZ and can be used as a new class of biomarker. J Biotechnol. 2016;238:42–51. ZYH collected clinical samples. XXH, ALZ, QL, GLS and QL analyzed the data 14. Bagrodia A, Lee BH, Lee W, Cha EK, Sfakianos JP, Iyer G, Pietzak EJ, Gao SP, and wrote the paper. LJW, LZ and HX revised the paper. All authors read and Zabor EC, Ostrovnaya I, et al. Genetic determinants of cisplatin resistance in approved the final manuscript. patients with advanced germ cell tumors. J Clin Oncol. 2016;34(33):4000–7. 15. Li B, Wang W, Li Z, Chen Z, Zhi X, Xu J, Li Q, Wang L, Huang X, Wang L, et Ethics approval and consent to participate al. MicroRNA-148a-3p enhances cisplatin cytotoxicity in gastric cancer The present study was approved by the Hospital’s Protection of Human through mitochondrial fission induction and cyto-protective autophagy Subjects Committee. suppression. Cancer Lett. 2017;410:212–27. 16. Zheng Q, Bao C, Guo W, Li S, Chen J, Chen B, Luo Y, Lyu D, Li Y, Shi G, et al. Consent for publication Circular RNA profiling reveals an abundant circHIPK3 that regulates cell We have received consents from individual patients who have participated growth by sponging multiple miRNAs. Nat Commun. 2016;7:11215. in this study. The consent forms will be provided upon request. 17. Wang K, Long B, Liu F, Wang JX, Liu CY, Zhao B, Zhou LY, Sun T, Wang M, Yu T, et al. A circular RNA protects the heart from pathological hypertrophy Competing interests and heart failure by targeting miR-223. Eur Heart J. 2016;37(33):2602–11. The authors declare that they have no competing interests. 18. Hsiao KY, Lin YC, Gupta SK, Chang N, Yen L, Sun HS, Tsai SJ. Noncoding effects of circular RNA CCDC66 promote Colon Cancer growth and Publisher’sNote metastasis. Cancer Res. 2017;77(9):2339–50. Springer Nature remains neutral with regard to jurisdictional claims in 19. Han D, Li J, Wang H, Su X, Hou J, Gu Y, Qian C, Lin Y, Liu X, Huang M, et al. published maps and institutional affiliations. Circular RNA circMTO1 acts as the sponge of microRNA-9 to suppress hepatocellular carcinoma progression. Hepatology. 2017;66(4):1151–64. Author details 20. Huang R, Zhang Y, Han B, Bai Y, Zhou R, Gan G, Chao J, Hu G, Yao H. Department of General Surgery, The First Affiliated Hospital of Nanjing Circular RNA HIPK2 regulates astrocyte activation via cooperation of Medical University, 300 Guangzhou Road, Nanjing 210029, Jiangsu province, autophagy and ER stress by targeting MIR124-2HG. Autophagy. 2017;13(10): China. Department of Gastrointestinal Surgery, The First Affiliated Yijishan 1722–41. Huang et al. Molecular Cancer (2019) 18:71 Page 20 of 20 21. Lu Y, Xiao L, Liu Y, Wang H, Li H, Zhou Q, Pan J, Lei B, Huang A, Qi S. MIR517C inhibits autophagy and the epithelial-to-mesenchymal (−like) transition phenotype in human glioblastoma through KPNA2-dependent disruption of TP53 nuclear translocation. Autophagy. 2015;11(12):2213–32. 22. Kinner A, Wu W, Staudt C, Iliakis G. Gamma-H2AX in recognition and signaling of DNA double-strand breaks in the context of chromatin. Nucleic Acids Res. 2008;36(17):5678–94. 23. Zhu Y, Liu Y, Zhang C, Chu J, Wu Y, Li Y, Liu J, Li Q, Li S, Shi Q, et al. Tamoxifen- resistant breast cancer cells are resistant to DNA-damaging chemotherapy because of upregulated BARD1 and BRCA1. Nat Commun. 2018;9(1):1595. 24. Cheung LW, Mills GB. Targeting therapeutic liabilities engendered by PIK3R1 mutations for cancer treatment. Pharmacogenomics. 2016;17(3):297–307. 25. Mo D, Fang H, Niu K, Liu J, Wu M, Li S, Zhu T, Aleskandarany MA, Arora A, Lobo DN, et al. Human helicase RECQL4 drives cisplatin resistance in gastric Cancer by activating an AKT-YB1-MDR1 signaling pathway. Cancer Res. 2016;76(10):3057–66. 26. Fraval HN, Rawlings CJ, Roberts JJ. Increased sensitivity of UV-repair- deficient human cells to DNA bound platinum products which unlike thymine dimers are not recognized by an endonuclease extracted from Micrococcus luteus. Mutat Res. 1978;51(1):121–32. 27. Turner KM, Sun Y, Ji P, Granberg KJ, Bernard B, Hu L, Cogdell DE, Zhou X, Yli-Harja O, Nykter M, et al. Genomically amplified Akt3 activates DNA repair pathway and promotes glioma progression. Proc Natl Acad Sci U S A. 2015; 112(11):3421–6. 28. Li D, Chen NN, Cao JM, Sun WP, Zhou YM, Li CY, Wang XX. BRCA1 as a nicotinamide adenine dinucleotide (NAD)-dependent metabolic switch in ovarian cancer. Cell Cycle. 2014;13(16):2564–71. 29. Dacheux E, Vincent A, Nazaret N, Combet C, Wierinckx A, Mazoyer S, Diaz JJ, Lachuer J, Venezia ND. BRCA1-dependent translational regulation in breast Cancer cells. PLoS One. 2013;8(6):e67313. 30. Li D, Bi FF, Chen NN, Cao JM, Sun WP, Zhou YM, Cao C, Li CY, Yang Q. Epigenetic repression of phosphatidylethanolamine N-methyltransferase (PEMT) in BRCA1-mutated breast cancer. Oncotarget. 2014;5(5):1315–25. 31. Li D, Bi FF, Cao JM, Cao C, Liu B, Yang Q. Regulation of DNA methyltransferase 1 transcription in BRCA1-mutated breast cancer: a novel crosstalk between E2F1 motif hypermethylation and loss of histone H3 lysine 9 acetylation. Mol Cancer. 2014;13:26. 32. Fan S, Liu B, Sun L, Lv XB, Lin Z, Chen W, Chen W, Tang Q, Wang Y, Su Y, et al. Mitochondrial fission determines cisplatin sensitivity in tongue squamous cell carcinoma through the BRCA1-miR-593-5p-MFF axis. Oncotarget. 2015; 6(17):14885–904. 33. Paul BT, Blanchard Z, Ridgway M, ElShamy WM. BRCA1-IRIS inactivation sensitizes ovarian tumors to cisplatin. Oncogene. 2015;34(23):3036–52. 34. Shin DH, Choi YJ, Park JW. SIRT1 and AMPK mediate hypoxia-induced resistance of non-small cell lung cancers to cisplatin and doxorubicin. Cancer Res. 2014;74(1):298–308. 35. Sun C, Li N, Yang Z, Zhou B, He Y, Weng D, Fang Y, Wu P, Chen P, Yang X, et al. miR-9 regulation of BRCA1 and ovarian cancer sensitivity to cisplatin and PARP inhibition. J Natl Cancer Inst. 2013;105(22):1750–8. 36. Tan X, Peng J, Fu Y, An S, Rezaei K, Tabbara S, Teal CB, Man YG, Brem RF. Fu SW: miR-638 mediated regulation of BRCA1 affects DNA repair and sensitivity to UV and cisplatin in triple-negative breast cancer. Breast Cancer Res. 2014;16(5):435. 37. Wang L, Wei J, Qian X, Yin H, Zhao Y, Yu L, Wang T, Liu B. ERCC1 and BRCA1 mRNA expression levels in metastatic malignant effusions is associated with chemosensitivity to cisplatin and/or docetaxel. BMC Cancer. 2008;8:97. 38. Hennessy BT, Smith DL, Ram PT, Lu Y, Mills GB. Exploiting the PI3K/AKT pathway for cancer drug discovery. Nat Rev Drug Discov. 2005;4(12):988– 39. Hui RC, Francis RE, Guest SK, Costa JR, Gomes AR, Myatt SS, Brosens JJ, Lam EW. Doxorubicin activates FOXO3a to induce the expression of multidrug resistance gene ABCB1 (MDR1) in K562 leukemic cells. Mol Cancer Ther. 2008;7(3):670–8. 40. Yang X, Fraser M, Abedini MR, Bai T, Tsang BK. Regulation of apoptosis- inducing factor-mediated, cisplatin-induced apoptosis by Akt. Br J Cancer. 2008;98(4):803–8. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Molecular Cancer Springer Journals

Circular RNA AKT3 upregulates PIK3R1 to enhance cisplatin resistance in gastric cancer via miR-198 suppression

Loading next page...
 
/lp/springer-journals/circular-rna-akt3-upregulates-pik3r1-to-enhance-cisplatin-resistance-NBEniwFdLX

References (43)

Publisher
Springer Journals
Copyright
Copyright © 2019 by The Author(s).
Subject
Biomedicine; Cancer Research; Oncology
eISSN
1476-4598
DOI
10.1186/s12943-019-0969-3
Publisher site
See Article on Publisher Site

Abstract

Background: Cisplatin (CDDP) treatment is one of the most predominant chemotherapeutic strategies for patients with gastric cancer (GC). A better understanding of the mechanisms of CDDP resistance can greatly improve therapeutic efficacy in patients with GC. Circular RNAs (circRNAs) are a class of noncoding RNAs whose functions are related to the pathogenesis of cancer, but, in CDDP resistance of GC remains unknown. Methods: circAKT3 (hsa_circ_0000199, a circRNA originating from exons 8, 9, 10, and 11 of the AKT3 gene) was identified by RNA sequencing and verified by quantitative reverse transcription PCR. The role of circAKT3 in CDDP resistance in GC was assessed both in vitro and in vivo. Luciferase reporter assay, biotin-coupled RNA pull-down and fluorescence in situ hybridization (FISH) were conducted to evaluate the interaction between circAKT3 and miR-198. Functional experiments were measured by western blotting, a cytotoxicity assay, clonogenic assay and flow cytometry. Results: The expression of circAKT3 was higher in CDDP-resistant GC tissues and cells than in CDDP-sensitive samples. The upregulation of circAKT3 in GC patients receiving CDDP therapy was significantly associated with aggressive characteristics and was an independent risk factor for disease-free survival (DFS). Our data indicated that circAKT3 promotes DNA damage repair and inhibits the apoptosis of GC cells in vivo and in vitro. Mechanistically, we verified that circAKT3 could promote PIK3R1 expression by sponging miR-198. Conclusions: circAKT3 plays an important role in the resistance of GC to CDDP. Thus, our results highlight the potential of circAKT3 as a therapeutic target for GC patients receiving CDDP therapy. Keywords: Cisplatin resistance, Gastric cancer, circAKT3, Circular RNA, miR-198, PIK3R1 Background months [3]. In patients with histologically confirmed Gastric cancer (GC) is the most common malignant advanced GC and who are chemotherapy-naïve, cisplatin tumor of the digestive tract in East Asia and the third (CDDP) and fluorouracil-based chemotherapies were leading cause of cancer-related death worldwide [1, 2]. deemed as first-line treatments [4]. However, patients At present, the main treatments for advanced GC are always acquired drug resistance after several cycles of systemic chemotherapy and palliative surgery, but the CDDP-based treatment. Thus, chemotherapy resistance has overall median survival after treatment is only 8 to 11 limited overall clinical efficacy in patients [5, 6]. To improve GC patient survival, illuminating the molecular mechanism * Correspondence: xuzekuan@njmu.edu.cn underlying CDDP resistance in GC is essential. Xiaoxu Huang, Zheng Li, Qiang Zhang and Weizhi Wang contributed The cytotoxicity of CDDP is mediated by its inter- equally to this work. 1 action with DNA to form DNA adducts. Intracellular Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, Jiangsu province, CDDP primarily binds to nuclear DNA with high affinity China and can physically interact with mitochondrial DNA Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Huang et al. Molecular Cancer (2019) 18:71 Page 2 of 20 (mtDNA), which is involved in the activation of several recurrence during CDDP-based therapy; both definitions signaling pathways and apoptosis [7–9]. In recent years, followed standard CDDP response definitions published studies have shown that the PI3K/AKT signaling path- elsewhere [14]. Forty-four samples (Cohort 1) were used way could suppress cell apoptosis and facilitate cell for circRNAs validation, and another 105 samples (Cohort survival. This PI3K/AKT signaling function is crucial in 2) were used to quantify circAKT3 levels and to analyze the regulation of chemotherapy resistance of cancer cells the correlation between circAKT3 expression and [10, 11]. Activated PI3K/AKT signaling promotes the outcomes after R0 excision in patients undergoing phosphorylation of caspase-3 and prevents the activation CDDP-based chemotherapy. The samples from cohorts 1 of caspase-3 and the inhibition of apoptosis [12]. and 2 were obtained in 2013–2016 and 2007–2011, Circular RNAs (circRNAs), a category of noncoding respectively. The grouping of the ROC curve was based RNAs (ncRNAs), play a crucial role in the process of on the median relative expression of circAKT3. Detailed transcriptional and posttranscriptional gene expression information is listed in Additional file 1: Table S1. [13]. Recently, circRNAs were found to function as com- petitive endogenous RNAs (ceRNAs) to sponge micro- Cell culture RNAs (miRNAs) and then suppress their functions, The CDDP-sensitive cell lines SGC7901 and BGC823 as indicating a novel mechanism for regulating miRNA well as their CDDP-resistant strains (SGC7901CDDP activity and providing a promising mode of action for and BGC823CDDP, respectively) were maintained in circRNAs. As miRNAs regulate a series of biological RPMI 1640 medium (Wisent, Shanghai, China) supple- processes, circRNA sponge activity will affect these mented with 10% fetal bovine serum (FBS) (Wisent, biological behaviors as well [13]. miRNAs are a large Biocenter, China) (Additional file 2: Figure S1A). 293 T class of short (~ 22 nt) ncRNAs that posttranscriptionally cells were cultured in DMEM with high glucose (Gib- regulate gene expression through direct base pairing to co-BRL, Carlsbad, CA, USA) supplemented with 10% target sites within mRNAs. circRNAs can affect miRNA FBS. 293 T, SGC7901CDDP, BGC823 and SGC7901 cells activities by competing for miRNA-binding sites [13]. were purchased from the Cell Bank of Type Culture Col- However, the function of circRNAs as miRNA sponges lection of Chinese Academy of Sciences, and has not been clearly elucidated in GC resistance to BGC823CDDP cells were established as previously de- CDDP. scribed [15]. To investigate the potential roles of circRNAs in the regulation of CDDP resistance in GC, we performed RNA miRNA targets prediction of circAKT3 sequencing (RNA-Seq) and verified thousands of distinct To predict the miRNA-binding sites of circAKT3, we circRNAs in CDDP-sensitive and CDDP-resistant GC used the bioinformatic databases miRanda, PITA and cells from humans. Through functional gain and loss RNAhybrid. Filtering restrictions were as follows: (1) experiments, we further observed that hsa_circ_0000199, total score ≥ 140, total energy < 17 kcal/mol; (2) com- which originates from exons 8, 9, 10, and 11 of the AKT3 bined interaction energy (△△G) < 10; and (3) minimum gene and is termed circAKT3, was significantly free energy (MFE) ≤ 20 kcal/mol. Detailed information is upregulated in both CDDP-resistant GC tissues and listed in Additional 3: Dataset S1. CDDP-resistant cells. Furthermore, we found that circAKT3 modulates CDDP sensitivity by sponging miR-198 that suppresses PIK3R1 expression via activation RNA preparation, treatment with RNase R, and PCR of the PI3K/AKT pathway in GC. Total RNA was extracted from GC cells or tissues using TRIzol Reagent (Invitrogen, 15,596,018). RNase R treat- Methods ment was carried out for 15 min at 37 °C using 3 U/mg Patients and samples RNase R (Epicenter). For Quantitative real-time PCR In total, 149 GC tissues (cohorts 1, 2) were obtained from (RT-PCR), 500 ng of treated RNA was directly reverse the First Affiliated Hospital of Nanjing Medical University. transcribed using Prime Script RT Master Mix (Takara, All samples were collected in accordance with HIPAA Japan) and either random or oligo(dT) primers. Reverse guidelines and approved institutional protocols. Patients transcription of miRNA was performed using a New received treatment with standard CDDP-based thera- Poly(A) Tailing Kit (ThermoFisher Scientific, China). peutic regimens after surgery. Disease-free survival (DFS) mRNA was reverse transcribed into cDNA with a Prime- was defined as the time interval between gastrectomy (R0 Script RT Master Mix Kit (Takara, RR036A, Japan). excision) and the time of either disease recurrence or cDNA was amplified using Universal SYBR Green Mas- disease-associated death. CDDP resistance was defined as ter Mix (4,913,914,001, Roche, Shanghai, China). The tumor relapse during CDDP-based chemotherapy after R0 CT value was measured during the exponential growth excision, and CDDP sensitivity was defined as no tumor phase. Relative gene expression levels were determined Huang et al. Molecular Cancer (2019) 18:71 Page 3 of 20 Fig. 1 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 4 of 20 (See figure on previous page.) Fig. 1 circAKT3 expression is increased in CDDP-resistant GC cells and tissues. a Validated expression of 10 circRNAs in the tissues from 44 GC patients using RT-qPCR. b Expression levels of circAKT3 in CDDP-resistant and their matched sensitive parental cell lines (SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823) normalized to GAPDH expression. c The existence of circAKT3 was validated by Sanger sequencing. The red arrow shows the “head-to-tail” splicing sites of circAKT3. d The existence of circAKT3 was validated in SGC7901CDDP and BGC823CDDP cell lines by RT-PCR. Divergent primers amplified circAKT3 in cDNA but not in genomic DNA (gDNA). GAPDH served as a negative control. e RNA from SGC7901CDDP and BGC823CDDP cells was treated with or without RNase R for RT-qPCR. The relative levels of circAKT3 and AKT3 mRNA were normalized to the values measured in the mock-treated cells. f Levels of small nucleolar RNA (U6, as a positive control for the nuclear fraction), GAPDH (positive control for cytoplasmic fraction), AKT3 mRNA and circRNAs from the nuclear and cytoplasmic fractions of SGC7901CDDP cells. g RNA stability of circular and linear transcripts of AKT3 and of 18S rRNA in SGC7901CDDP cells. h Representative images of RNA FISH of circAKT3 expression in SGC7901CDDP cells, which show that circAKT3 is predominantly localized to the cytoplasm. Nuclei were stained with DAPI. Scale bar, 10 μm. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 -△△CT using the 2 method. The primers used are listed in MgCl2, 1 mM CaCl and 250 mM sucrose) supple- Additional file 1: Table S2. mented with protease inhibitors. After centrifugation at 600 × g for 5 min at 4 °C, the resultant supernatant was Isolation of nuclear and cytoplasmic fractions collected as the cytoplasmic fraction and mixed with an SGC7901CDDP and BGC823CDDP cells were lysed on equal volume of TRIsure reagent. After the pellet was ice for 10 min in 0.3% NP-40/NIB-250 buffer (15 mM washed with NIB-250, the nuclei were lysed in TRIsure Tris–HCl (pH 7.5), 60 mM KCl, 15 mM NaCl, 5 mM reagent. Fig. 2 The circAKT3 expression level is correlated with poor prognosis in GC patients receiving CDDP therapy. a Expression levels of circAKT3 in tissue samples of 105 GC patients (CDDP-resistant and CDDP-sensitive groups) normalized to GAPDH expression. b Kaplan-Meier survival curves of DFS for patients with high (n = 53) or low (n = 52) circAKT3 expression. The median circAKT3 expression value was used as the cutoff. c ROC curves of circAKT3. d Multivariate analyses of hazard ratios for DFS. The results are presented as the mean ± SEM Huang et al. Molecular Cancer (2019) 18:71 Page 5 of 20 Table 1 Correlation of relative circAKT3 expression with the Oligonucleotide transfection clinicopathological characteristics of 105 patients accepted siRNA and miRNA mimics and inhibitors were synthe- cisplatin-based chemotherapy with gastric cancer sized by GenePharma (Shanghai, China). The sequences Characteristics Number No. of patients P value used are listed in Additional file 1: Tables S3 and S4. high low circAKT3 circAKT3 Transfection was carried out using Lipofectamine RNAi- MAX (Life Technologies) according to the manufac- Age(y) turer’s instructions. < 60 62 34 28 0.283 ≥ 60 43 19 24 RNA pull-down Gender A pull-down assay was performed as described Male 80 37 43 0.121 previously [17, 18]. The biotin-labeled circAKT3 probe Female 25 16 9 was synthesized by RiboBio (Guangzhou, China). In brief, 1 × 10 circAKT3-overexpressing GC cells were Tumor size(cm) harvested, lysed, and sonicated. The circAKT3 or oligo < 3.5 56 21 35 0.004** probe was incubated with streptavidin-coupled ≥ 3.5 49 32 17 Dynabeads (Invitrogen) at 30 °C overnight to generate Histological grade probe-bound Dynabeads. After the treated beads were Well-moderately 33 11 22 0.017* washed with wash buffer, the RNA complexes bound Poorly-signet 72 42 30 to the beads were eluted and disrupted with lysis buf- fer and proteinase K prior to RT-PCR or RT-qPCR. Clinical stage Biotinylated probes sequences used in this study (see II 39 13 26 0.007** Additional file 1:Table S5). III 66 40 26 T classification Luciferase reporter assay T1-T2 31 9 22 0.004** 293 T, SGC7901CDDP and BGC823CDDP cells were T3-T4 74 44 30 seeded in 24-well plates and cotransfected with corre- sponding plasmids and miRNA mimics in triplicate. At N classification 48 h after transfection, luciferase reporter assays were N0 14 5 9 0.232 conducted using a dual-luciferase reporter assay system N1-N3 91 48 43 (Promega, Madison, WI) according to the manufac- Cisplatin chemosensitivity turer’s instructions. Relative luciferase activity was Sensitive 92 41 51 0.001** normalized to Renilla luciferase activity. Resistant 13 12 1 Fluorescence in situ hybridization (FISH) Vector construction The double FISH assay was performed in The method for overexpressing circRNAs was reported SGC7901CDDP cells and GC tissues as previously de- previously [16]. For the construction of circAKT3 scribed [16, 19]. Biotin-labeled probes specific to cir- overexpression plasmids, human circAKT3 cDNA was cAKT3 and Dig-labeled locked nucleic acid miR-198 amplified using PrimerSTAR Max DNA Polymerase Mix probes were used in the hybridization (Exiqon, Vedbaek, (Takara, RR036A, Japan) and inserted into the pCD5-ciR Denmark). The sequences are listed in Additional file 1: vector (Greenseed Biotech Co, Guangzhou, China). The Table S6, FISH probes sequences used in this study. The pCD5-ciR vector contains a front circular frame and a signals of the biotin-labeled probes were detected using back circular frame. Transfection was carried out using Cy5-conjugated streptavidin (Life Technologies), and the Lipofectamine 2000 (Invitrogen) according to the manu- signals of the Dig-labeled miR-198 probes were detected facturer’s instructions. The luciferase reporter containing using a tyramide-conjugated Alexa 488 fluorochrome the circAKT3 sequence in the 3′-UTR was constructed by TSA kit. Nuclei were counterstained with 4,6-diamidi- subcloning the circAKT3 fragment into the region directly no-2-phenylindole. Images were acquired on a Leica downstream of a cytomegalovirus promoter-driven firefly TCS SP2 AOBS confocal microscope (Leica Microsys- luciferase (FL) cassette in a pCDNA3.0 vector. Mutations tems, Mannheim, Germany). CircAKT3 and miR-198 ex- of each miRNA-binding site in the circAKT3 sequence pression levels were evaluated by the proportions and were created using a Mut Express II Fast Mutagenesis Kit intensities of the positive cells detected within 5 fields of (Vazyme, Nanjing, China). The mutations were introduced view on every slide (400-fold magnification). Proportion in both the circAKT3-expressing vector and the luciferase scores were assigned as follows: < 10% = 0, 10–25% = 1, reporter containing the circAKT3 sequence. 26–50% = 2, 51–75% = 3 and > 75% = 4. Intensity scores Huang et al. Molecular Cancer (2019) 18:71 Page 6 of 20 Fig. 3 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 7 of 20 (See figure on previous page.) Fig. 3 Downregulation of circAKT3 facilitates cisplatin sensitivity of CDDP-resistant GC cells in vitro. a Illustration showing the siRNA targeting the back-splice junction (si-circ-1 and si-circ-2). b RT-qPCR results for circular and linear transcripts of AKT3 in SGC7901CDDP cells treated with or without siRNA (NC, negative control; si-NC, control oligonucleotides with scramble sequence; si-circ-1 and si-circ-2, oligonucleotides targeting the back-splice junction). c Relative cell viability of NC, si-NC- transfected or si-circ-1-transfected SGC7901CDDP cells in the presence of CDDP at the indicated concentrations for 48 h. d Colony-forming ability of the NC, si-NC-transfected and si-circ-1-transfected SGC7901CDDP cells in the absence (Vehicle) or presence of CDDP (6 μM) for 48 h. e The apoptotic rates of NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 in the absence (Vehicle) or presence CDDP (6 μM) for 48 h were visualized by flow cytometry. f Western blot analysis shows apoptotic proteins in NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 upon CDDP treatment (6 μM) for 48 h (GAPDH was used as the loading control). g Immunofluorescence staining of γH2AX foci in NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 at 2 h after CDDP treatment (6 μM). Scale bars, 10 μm. h Percentage of cells containing > 10 γH2AX foci in NC SGC7901CDDP cells and SGC7901CDDP cells transfected with si-NC or si-circ-1 at 0 to 8 h after CDDP treatment (6 μM) removal. i Western blot analysis of γH2AX and BRCA1 expression in NC SGC7901CDDP and BGC823CDDP cells and cells transfected with si-NC or si-circ-1 after CDDP treatment (6 μM) removal. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 were assigned as follows: 0 = no staining, 1 = weak, 2 = Apoptosis assay moderate, 3 = strong and 4 = significantly strong. Cell apoptosis was detected using a PI/Annexin V-FITC Apoptosis Detection Kit (BD Pharmingen, 556,547) according to the manufacturer’s instructions. Briefly, Western blot analysis after GC cells were treated with CDDP at the indicated For western blot analysis, cells were extracted using a concentrations for 48 h in 6-well plates, they were har- protein extraction kit (Key Gene, KGP9100). Lipid pro- vested and resuspended in 300 ml of binding buffer. teins were added into 8, 10, 12% or 15% gels, subjected Next, 5 μl of Annexin V-FITC and 5 μl of PI were added to 120 V to promote migration, and then transferred to the suspensions, and the cells were incubated in the onto nitrocellulose membranes. The membranes were dark at 4 °C for 15 min. The samples were subsequently blocked with 5% BSA in TBST buffer and incubated with analyzed with a flow cytometer (Gallios, Beckman, specific primary antibodies at 4 °C overnight. The next USA). day, membranes were washed 3 times for 15 min in TBST and incubated with secondary antibodies for 2 h at room temperature. HRP substrate (WBKL0100, Milli- Actinomycin D assay pore, USA) was used to detect the protein bands (Mo- The Actinomycin D assay was performed as previously lecular Imager, ChemiDoc XRS+, BIO-RAD, USA), and described [16]. SGC7901CDDP and BGC823CDDP cells the band intensities were quantified using Image-Pro were seeded in 5 wells in 24-well plates (5 × 10 cells per Plus software (Mediacy, USA). Detailed information of well). Twenty-four hours later, the cells were exposed to antibody used in this study (see Additional file 1: Actinomycin D (2 μg/ml, Abcam, ab141058) for 0 h, 6 h, Table S7). 12 h, 18 h and 24 h. The cells were then harvested, and the relative RNA levels of circAKT3 and AKT3 mRNA were analyzed by RT-qPCR and normalized to the values Cytotoxicity assay measured in the group in the 0 h group (mock The cytotoxicity assay was performed as previously treatment). described [15]. Cell viability was measured using Cell Counting Kit-8(CCK8)following the manufacturer’s directions (Dojindo, Kumamoto, Japan). Immunofluorescence staining Cells seeded onto coated cover slips growth for 24 h, Clonogenic assay then treated with CDDP, and harvested the cells at 0, 2, A clonogenic assay was performed as previously de- and 8 h. The cells were fixed with 4% paraformaldehyde scribed [15]. At 48 h after transfection, BGC823CDDP, at room temperature for 15 min and then permeabilized SGC7901CDDP, BGC823 and SGC7901 cells were cul- with PBS containing 0.25% Triton X-100 for 10 min. tured with CDDP at the indicated concentrations for 3 Next, the cells were blocked with 1% BSA for 20 min h. Then, the cells were harvested, seeded into six-well before incubation with primary antibodies at room plates (500 cells per well) and cultured for an additional temperature for 2 h. After the cells were washed with 2 (BGC823CDDP and SGC7901CDDP cells) or 3 weeks PBS, they were incubated with appropriate secondary (BGC823 and SGC7901 cells). For scoring the antibodies (FITC-conjugated goat anti-rabbit, Molecular colony-forming units, we fixed cells in 1 ml of methanol Probes, USA) at room temperature for 2 h. Following a for 10 min and then stained the cells with crystal violet final wash with PBS, cells were mounted with antifading for 15 min. mounting medium containing DAPI. The images were Huang et al. Molecular Cancer (2019) 18:71 Page 8 of 20 Fig. 4 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 9 of 20 (See figure on previous page.) Fig. 4 Overexpression of circAKT3 increases the resistance of CDDP-sensitive GC cells to cisplatin in vitro. a The expression levels of circAKT3 and AKT3 mRNA in SGC7901 and BGC823 cells after stable transfection of circAKT3 or empty vector plasmids were detected by RT-qPCR. b Relative cell viability of circAKT3 or vector-transfected SGC7901 cells exposed to CDDP at the indicated concentrations for 48 h. c Colony-forming ability of SGC7901 cells transfected with circAKT3 or vector in the absence (Vehicle) or presence of CDDP (1.5 μM) for 48 h. d The apoptosis rates of SGC7901 cells transfected with circAKT3 or vector after CDDP (1.5 μM) treatment for 48 h were detected by flow cytometry. e Western blot analysis of apoptotic proteins in SGC7901 cells transfected with circAKT3 or vector upon CDDP (1.5 μM) treatment for 48 h (GAPDH was used as the loading control). f Immunofluorescence staining of γH2AX foci in SGC7901 cells transfected with circAKT3 or vector at 2 h after CDDP treatment (1.5 μM). Scale bars, 10 μm. g Percentage of cells containing > 10 γH2AX foci in SGC7901 cells transfected with circAKT3 or vector at 0 to 8 h after CDDP (1.5 μM) treatment. h Western blot analysis of γH2AX and BRCA1 expression in SGC7901 cells transfected with circAKT3 or vector after CDDP treatment (1.5 μM) removal. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 captured with a Leica DMI3000B (Germany) fluores- Statistical analysis cence microscope. All experiments were performed in triplicate. Data were analyzed with SPSS 19.0 software (IBM, USA) and presented as the mean ± SEM. The statistical significance Transduction with lentivirus of the results was calculated using an unpaired Student’s SGC7901CDDP cells stably expressing circAKT3 siRNA t-test. DFS analysis was performed using the Kaplan-Meier (si-circ-1) and its negative control siRNA (si-NC) were method and log-rank test. Clinicopathological features were generated by infection with lentiviruses as previously analyzed by a χ test. A Cox proportional hazards regres- described [20]. Transfection was carried out according to sion model was used to identify independent prognostic the manufacturer’s instructions. The lentiviral expressing factors associated with DFS. Linear correlation analyzes vectors were purchased from HanBio Co. Ltd. (Shanghai, were performed to determine correlations between China). circAKT3, miR-198 and PIK3R1 expression levels. A P value< 0.05 was defined as statistically significant. Nude mouse xenograft model Results Six-week-old female BALB/c nude mice were purchased Ectopic circAKT3 expression levels are observed in CDDP- from the Laboratory Animal Center of Nanjing Medical resistant GC cells and tissues and are correlated with University and maintained under pathogen-free condi- poor prognosis in GC patients receiving CDDP therapy tions. A total of 5 × 10 SGC7901CDDP cells infected To characterize circular RNA transcripts, we conducted with lentivirus containing si-circ-1 or si-NC (2 μlof 10 RNA-Seq analysis of CDDP-resistant SGC7901 and − 1 viral genomes μl , HanBio) in 100 μl of PBS were sub- BGC823 cells (i.e., SGC7901CDDP and BGC823CDDP) cutaneously injected into different sides of the groin of and their corresponding parental strains (i.e., SGC7901 each mouse. One week after injection, we intraperitone- and BGC823), which are sensitive to CDDP. The ally injected mice with cisplatin (5 mg/kg) in PBS or PBS sequencing statistics are not shown. The analysis indi- alone three times per week. The xenograft tumors were cated that a series of circRNAs were differentially harvested after 5 weeks. The entire experimental proto- expressed in CDDP-resistant GC cells compared with col was conducted in accordance with the guidelines of the sensitive parental GC cells. We then chose the top the local institutional animal care and use committee. 20 significantly upregulated circRNAs and verify their expression levels. Detailed information of 20 candidate circRNAs in Additional file 1: Table S8 (including loca- Immunohistochemical staining (IHC) tion, genomic and spliced length). Using divergent Xenografts and GC tissues exposed to the indicated con- primers to specifically target the circular junction as well centrations of CDDP were prepared for IHC as previ- as combined quantitative reverse transcription PCR ously described [21]. Sections were identified by IHC (RT-qPCR) analysis and sequencing validation, we found Imager (DM4000B, LEIKA, Germany), and target pro- that only 10 of these circRNAs had confirmed differ- tein expression levels were evaluated by the proportions ences in expression and that circAKT3 was the most ob- and intensities of positive cells detected within 5 fields viously upregulated circRNA in CDDP-resistant patients of view on every slide (400-fold magnification). Propor- of cohort 1 (Fig. 1a and Additional file 2: Figure S1b-c). tion scores were assigned as follows: < 10% = 0, 10–25% circAKT3 (hsa_circ_0000199) has been mapped to exons =1, 26–50% = 2, 51–75% = 3 and > 75% = 4. Intensity 8, 9, 10, and 11 of the AKT3 gene (555 bp) (Additional file scores were assigned as follows: 0 = no staining, 1 = 2: Figure S1d). Consistent with the RNA-Seq results, the weak, 2 = moderate, 3 = strong and 4 = significantly expression of circAKT3 was obviously increased in strong. CDDP-resistant GC cells (Fig. 1b). Subsequently, we Huang et al. Molecular Cancer (2019) 18:71 Page 10 of 20 Fig. 5 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 11 of 20 (See figure on previous page.) Fig. 5 circAKT3 exerts its function by sponging miR-198. a & b Schematic illustration showing the overlap of the target miRNAs of circAKT3 predicted by miRanda, PITA and RNAhybrid. c & d Lysates prepared from SGC7901CDDP and BGC823CDDP cells stably transfected with circAKT3 or vector were subjected to RNA pull-down and tested by RT-PCR (C) and RT-qPCR (D). The relative level of circAKT3 was normalized to the input. GAPDH served as a negative control. e & f The relative levels of 11 miRNA candidates in SGC7901CDDP and BGC823CDDP lysates were detected by RT-qPCR. Multiple miRNAs were pulled down by circAKT3, and miR-198 was pulled down by circAKT3 in both cell lines. g Schematic illustration showing the 3′UTR of luciferase reporters containing the complete circAKT3 sequence (luc-wt) or the circAKT3 sequence with deletions of miR-198 (luc-m1-m8) binding sites. h Reporter assays showing the luciferase activity of luc-wt and luc-m1-m8 in 293 T cells cotransfected with miR-198 mimics or a scrambled oligonucleotide (control). i FISH showing the colocalization of circAKT3 and miR-198 in SGC7901CDDP cells. Nuclei were stained with DAPI. Scale bar, 10 μm. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01 verified the head-to-tail splicing of the RT-PCR product of grade, clinical stage and circAKT3 expression level (Add- circAKT3 by Sanger sequencing (Fig. 1c). Meanwhile, to itional file 1: Table S9). Subsequently, multivariate analysis exclude possibilities such as genomic rearrangements or indicated that circAKT3 expression, along with tumor size trans-splicing, several experiments were employed. First, and clinical stage, was an independent risk factor for DFS we designed convergent primers to amplify AKT3 mRNA (Additional file 1: Table S9 and Fig. 2d). and divergent primers to amplify circAKT3. Using cDNA and genomic DNA (gDNA) from SGC7901CDDP and circAKT3 facilitates CDDP resistance in vitro BGC823CDDP cell lines as templates, the circAKT3 amp- First, we designed two siRNA oligonucleotides (si-circ-1 lification product was only observed in cDNA by diver- and si-circ-2) to target the unique back-splice junction of gent primers but not in gDNA (Fig. 1d). In addition, the circAKT3 (Fig. 3a); si-circ-1 successfully knocked down fragment of the linear form of AKT3 was digested by circAKT3 expression but had no effect on the levels of en- RNase R, but circAKT3 remained after RNase R treatment dogenous linear AKT3 transcript in SGC7901CDDP and (Fig. 1e). Then, the relative expression levels of cir- BGC823CDDP cells (Fig. 3b and Additional file 4:Figure cAKT3 were detected in the cytoplasm and nucleus of S2a). Additionally, to further assess the role of circAKT3, SGC7901CDDP and BGC823CDDP cells (Fig. 1f and Add- circAKT3 was overexpressed in SGC7901 and BGC823 itional file 2: Figure S1e). The RT-qPCR results demon- cells via transfection of pCD5-ciR-AKT3 e8–11 (Fig. 4a). strated that circAKT3 was enriched in the cytoplasm. Importantly, elevated expression of circAKT3 had no ef- Moreover, we used Actinomycin D to suppress transcrip- fect on the levels of linear AKT3 mRNA, as confirmed by tion and measure the half-life of circAKT3 in RT-qPCR (Fig. 4a). circAKT3 inhibition reduced the via- SGC7901CDDP and BGC823CDDP cells; we found that bility of SGC7901CDDP and BGC823CDDP cells (Fig. 3c circAKT3 was more stable than AKT3 mRNA (Fig. 1g and Additional file 4: Figure S2b). In addition, knockdown and Additional file 2: Figure S1f). Additionally, the FISH of circAKT3 significantly decreased the number of cell results displayed a dominantly cytoplasmic distribution colonies (Fig. 3d and Additional file 4: Figure S2c) and of circAKT3 (Fig. 1h). promoted apoptosis (Fig. 3e and Additional file 4:Figure Next, we detected the expression level of circAKT3 in S2d). The phosphorylated histone family member X tissues of patients from cohort 2. Consistent with the (γH2AX) forms discrete nuclear foci and acts as a plat- RNA-Seq results, circAKT3 was significantly more highly form to recruit additional factors and enhance the DNA expressed in the CDDP-resistant GC tissues than in the repair pathway [22]. Meanwhile, circAKT3-knockdown sensitive tissues (Fig. 2a). Compared with GC patients ex- cells showed significantly more γH2AX foci than the con- pressing low levels of circAKT3, GC patients receiving trol cells at 2 h after CDDP treatment (Fig. 3g and Add- CDDP therapy and exhibiting upregulation of circAKT3 itional file 4: Figure S2e). circAKT3-knockdown cells had showed a significant association with decreased five-year a higherpercentageofactivefocirelativetothatincontrol DFS (Fig. 2b). To further verify that circAKT3 may be a cells from 0 to 8 h after CDDP treatment (Fig. 3hand therapeutic target for CDDP-resistant patients, we calcu- Additional file 4: Figure S2f). However, compared with the lated the area under the receiver operating characteristic negative control, ectopic circAKT3 expression signifi- curve (AUC) using the expression levels of circAKT3. The cantly increased cell viability and the number of cell area under the curve is 91% (Fig. 2c), suggesting that the colonies and inhibited apoptosis and the formation of expression level of circAKT3 is a good predictive bio- γH2AX foci in SGC7901 and BGC823 cells (Fig. 4b-g and marker of CDDP resistance for GC patients. Analysis of Additional file 5: Figure S3a-e). We used western blotting the clinicopathological characteristics in cohort 2 showed to investigate the underlying mechanism of these activ- that circAKT3 expression was positively related to tumor ities. In the presence of CDDP, knockdown of circAKT3 size, histological grade, clinical stage, T classification and in SGC7901CDDP and BGC823CDDP cells increased CDDP resistance (Table 1). A univariate analysis showed cleaved caspase-3 protein levels, while the levels of the that DFS was obviously related to tumor size, histological inactivated form of caspase-3 protein was decreased (Fig. Huang et al. Molecular Cancer (2019) 18:71 Page 12 of 20 Fig. 6 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 13 of 20 (See figure on previous page.) Fig. 6 PIK3R1 is a direct target of miR-198. a mRNA microarray data of the top 20 upregulated genes in SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 cells are presented as heatmaps. b Venn diagram showing 4 genes that are putative miR-198 targets computationally predicted by four algorithms (miRanda, RNAhybrid, miRWalk and TargetScan) among the top 20 upregulated genes. c Schematic of PIK3R1 3’UTR wild-type (WT) and mutant (Mut) luciferase reporter vectors is shown. d The relative luciferase activities were analyzed in 293 T cells cotransfected with miR- 198 mimics or miR-NC and luciferase reporter vectors PIK3R1 3’UTR (WT) or PIK3R1 3’UTR (Mut). e & f The expression of PIK3R1 was analyzed using RT-qPCR in tissues of cohort 1 (E) and in SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 cells (F). g The expression of PIK3R1 was analyzed using western blot in SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 cells. h & i The expression of PIK3R1 was analyzed by RT- qPCR (G) and western blot (H). SGC7901CDDP and BGC823CDDP cells were transfected with miR-198 mimic or cotransfected with the indicated vectors. j The IC50 of miR-198 was analyzed by the CCK8 assay. SGC7901CDDP cells were transfected with miR-198 mimic alone or cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 h. k The apoptosis rates of SGC7901CDDP cells transfected with miR-198 mimic alone or cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 h. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01 3f). In contrast, cleaved and inactivated caspase-3 protein RNA FISH assays revealed that circAKT3 and miR-198 levels were observed when circAKT3 was overexpressed were colocalized in the cytoplasm (Fig. 5i). (Fig. 4e). These data are consistent with a previous study reporting that CDDP-induced increases in Breast cancer type 1 susceptibility protein(BRCA1) expression leads to PIK3R1 is a direct target of miR-198 enhanced DNA damage repair (DDR) in breast cancer A microarray assay was further performed with cells [23]. After CDDP treatment, knockdown of SGC7901CDDP, BGC823CDDP, SGC7901 and BGC823 circAKT3 in SGC7901CDDP and BGC823CDDP cells cells to validate the results of the ceRNA analysis. We increased γH2AX but decreased BRCA1 protein levels. analyzed the top 20 upregulated genes according to four circAKT3 overexpression also inhibited γH2AX and pro- algorithms (miRanda, RNAhybrid, miRWalk and moted BRCA1 protein levels compared with the levels in TargetScan) prediction, and miR-198 could target the the controls (Figs. 3iand 4h). 3’UTRs of PIK3R1, CHRM3, HIPK2 and MAFB (Add- itional file 3: Dataset S3). We performed luciferase reporter assays to determine whether miR-198 directly circAKT3 exerts its function by sponging miR-198 targets these 4 genes in 293 T cells (Fig. 6c and Additional To address whether circAKT3 could sponge miRNAs in file 6: Figure S4c and e). In 293 T, SGC7901CDDP and GC cells, we selected 11 candidate miRNAs by overlap- BGC823CDDP cells cotransfected with miR-198 mimic, ping the prediction results of the miRNA recognition ele- reporter constructs containing wild-type miR-198 binding ments in the circAKT3 sequence using miRanda, PITA, sites at the PIK3R1 3′UTR exhibited decreased luciferase and RNAhybrid (Fig. 5a-b). Next, we investigated whether activity relative to that of reported constructs with candidate miRNAs could directly bind circAKT3. A mutated binding sites (Fig. 6d and Additional file 6:Figure biotin-labeled circAKT3 probe was designed and verified S4f and g). PIK3R1 protein (p85α, encoded by PIK3R1) is to pull down circAKT3 in SGC7901CDDP and the regulatory subunit of PI3K. A functional study BGC823CDDP cell lines, and the pull-down efficiency was demonstrated that PIK3R1 was highly expressed in significantly enhanced in cells with stable circAKT3 over- CDDP-resistant ovarian cancer cells, and downregu- expression (Fig. 5c-d). The miRNAs were extracted after lated PIK3R1 resensitized the abovementioned cells to pull-down, and the levels of the 11 candidate miRNAs platinum-based treatment, which reveals the promising were detected by RT-qPCR. As shown in Fig. 5e-f, in both involvement of p85α in secondary CDDP resistance [24]. SGC7901CDDP and BGC823CDDP cells, miR-198 was Compared with parental CDDP-sensitive cells, CDDP-re- abundantly pulled down by circAKT3. Furthermore, using sistant cells showed obvious increases in the expression of the RNAhybrid bioinformatics prediction tool, we calcu- PIK3R1 mRNA and protein levels (Fig. 6f and g). lated the secondary conformation of circAKT3 and Furthermore, we found that miR-198 mimics significantly miR-198 and found that there were 8 predicted binding inhibited PIK3R1 mRNA and protein levels and that domains (largest combined with a MFE > − 20 kcal/mol) ectopic PIK3R1 expression abolished the influence caused (Additional file 3: Dataset S2). Next, the results of the lu- by miR-198 overexpression (Fig. 6h and i). Subsequently, ciferase reporter assays showed that miR-198 expression the data showed that overexpression of miR-198 inhibited significantly reduced the luciferase activity of the reporter cell viability and induced apoptosis in SGC7901CDDP containing the complete circAKT3 sequence appended to and BGC823CDDP cells. However, cotransfection of the 3′-UTR of luciferase (luc-wt) compared to that of the PIK3R1 and miR-198 abrogated these effects (Fig. 6jand k reporter containing circAKT3 with mutated miR-198 and Additional file 6: Figure S4h-j and Additional file 7: binding sites (luc-m1, m2 and m8) (Fig. 5g-h). Moreover, Figure S5a-b). Huang et al. Molecular Cancer (2019) 18:71 Page 14 of 20 Fig. 7 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 15 of 20 (See figure on previous page.) Fig. 7 circAKT3 regulates PIK3R1 expression, induces cisplatin resistance and activates the PI3K/AKT signaling cascade by targeting miR-198 in vitro. a & b The expression levels of PIK3R1 were analyzed using RT-qPCR. SGC7901CDDP cells were transfected with indicated vectors alone or cotransfected the inhibitors, and SGC7901 cells were transfected with the indicated vectors and miR-198 mimics. c The IC50 of circAKT3 was analyzed by the CCK8 assay. SGC7901CDDP cells were transfected with inhibitor alone or cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 h. d The apoptosis rates of SGC7901CDDP cells transfected with indicated vectors alone or cotransfected the inhibitors upon CDDP exposure (6 μM) for 48 h. e The expression levels of PIK3R1, apoptosis markers, γH2AX, BRCA1 and PI3K/AKT signaling molecules were determined using western blotting in SGC7901CDDP cells transfected with indicated vectors alone or cotransfected with the inhibitor after CDDP treatment (6 μM). f The expression levels of PIK3R1, apoptosis markers, γH2AX, BRCA1 and PI3K/AKT signaling molecules were determined using western blotting in SGC7901 cells transfected with the indicated vectors and miR-198 mimics after CDDP treatment (1.5 μM). Proteins were isolated from the indicated cells. g The expression levels of PIK3R1, apoptosis makers, γH2AX, BRCA1 and PI3K/AKT signaling molecules were determined using western blotting in SGC7901 cells transfected with the indicated vectors and BKM20 after CDDP treatment (1.5 μM). The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01 circAKT3 regulates PIK3R1 expression, activates the PI3K/ employed a specific p110α inhibitor, BKM20, and per- AKT signaling pathway and ultimately facilitates CDDP formed western blotting to determine whether deactiva- resistance by targeting miR-198 in vitro tion of PI3K/AKT signaling can overcome the changes Cotransfection of si-circ-1 and anti-miR-198 could caused by circAKT3 overexpression. The results indi- counteract the si-circ-1-induced downregulation of cated that si-PIK3R1 and BKM20 significantly inhibited PIK3R1 in SGC7901CDDP cells (Fig. 7a). Notably, PIK3R1 and p110α levels, respectively as well as inhib- cotransfection of circAKT3 and miR-198 attenuated the ited p-AKT, reduced BRCA1 levels, increased cleaved expression of PIK3R1 compared to transfection of caspase-3 levels, and promoted γH2AX (Fig. 7g). These circAKT3 alone in SGC7901 cells (Fig. 7b). The CCK8 results suggest that circAKT3 functions by targeting and flow cytometry analyses indicated that transfection miR-198 as a ceRNA to regulate PIK3R1 expression, ac- with si-circ-1 inhibited cell viability and promoted apop- tivate the PI3K/AKT signaling cascade and facilitate tosis after CDDP treatment while cotransfection with CDDP resistance. si-circ-1 and anti-miR-198 significantly promoted cell viability and inhibited apoptosis compared with the con- circAKT3 promotes CDDP resistance of GC cells in vivo trols (Fig. 7c and d, Additional file 7: Figure S5d). To investigate the potential clinical relevance of cir- Furthermore, upregulation of circAKT3 led to enhanced cAKT3 in vivo, we subcutaneously injected cell viability, but this effect could be significantly abol- SGC7901CDDP cells with or without stable circAKT3 ished by ectopic expression of miR-198 (Additional file knockdown (Additional file 8: Figure S6e) into the dorsal 8: Figure S6a). Additionally, overexpression of circAKT3 flanks of female BALB/c nude mice and allowed the cells induced a reduction in apoptosis. However, cotransfec- to proliferate for 5 weeks. Tumor xenograft data indi- tion of circAKT3 and miR-198 mimics led to enhanced cated that circAKT3 inhibition in CDDP-resistant cells apoptosis (Additional file 8: Figure S6b). Notably, cir- can significantly decrease xenograft tumor growth and cAKT3 upregulation could inhibit γH2AX expression, as sensitize cells to CDDP treatment (Fig. 8a and b). IHC indicated by the reduced fluorescence of γH2AX, analysis of tumor xenograft samples further indicated cotransfection of circAKT3 and miR-198 mimics led to that the protein levels of γH2AX and cleaved caspase-3 abolish this effect (Additional file 8: Figure S6c and d). were notably increased, but BRCA1 was decreased upon Transfection of si-circ-1 significantly reduced PIK3R1 circAKT3 inhibition (Fig. 8c). FISH showed the colocali- expression and inhibited the expression of canonical zation of circAKT3 and miR-198 in tissues from patients PI3K/AKT signaling molecules, as shown by western with CDDP-resistant or CDDP-sensitive GC (Fig. 8d). blot, and downregulating of both circAKT3 and The FISH score data showed that the expression of cir- miR-198 abrogated these effects in SGC7901CDDP cells cAKT3 was significantly higher in CDDP-resistant GC (Fig. 7e). Similar results are presented in Additional file tissues than CDDP-sensitive GC tissues; however, 9: Figure S7a-f. Meanwhile, Transfection of circAKT3 miR-198 expression showed the opposite result (Fig. 8d). significantly increased PIK3R1 expression and induced Similarly, IHC scores and western blot analyses indicated the expression of canonical PI3K/AKT signaling mole- that PIK3R1 protein expression was obviously increased cules, as shown by western blot, and concomitant in CDDP-resistant GC tissues compared to overexpression of both circAKT3 and miR-198 abro- CDDP-sensitive GC tissues (Fig. 8e and Additional file 10: gated these effects in SGC7901 cells (Fig. 7f). Transfec- Figure S8a and b). Furthermore, correlations were identi- tion of si-PIK3R1 significantly inhibited PIK3R1 and fied between circAKT3 and miR-198 expression levels and levels of phosphorylated PI3K/AKT signaling molecules PIK3R1 protein levels in these 44 GC tissue samples (Fig. in SGC7901 cells with circAKT3 overexpression. We 8f). Based on these data, we conclude that circAKT3 Huang et al. Molecular Cancer (2019) 18:71 Page 16 of 20 Fig. 8 (See legend on next page.) Huang et al. Molecular Cancer (2019) 18:71 Page 17 of 20 (See figure on previous page.) Fig. 8 circAKT3 promotes cisplatin resistance of GC cells in vivo. a Xenograft tumors of sacrificed mice with or without CDDP treatment (3 mg/kg, three times a week) at the end of the experiment. b Growth curves of subcutaneous xenograft tumors. c PIK3R1, γH2AX, cleaved caspase-3 and BRCA1 expression levels are shown in representative xenograft tumors by IHC (Left) (400x magnification, scale bars = 50 μm). Quantification of the IHC scores of PIK3R1, γH2AX, cleaved caspase-3 and BRCA1 expression levels (Right). d FISH showing the colocalization of circAKT3 and miR-198 in CDDP-resistant or CDDP-sensitive GC tissues from patients. FISH scores of circAKT3 and miR-198 were further calculated in 14 CDDP-resistant and 30 CDDP-sensitive patient tissues. Nuclei were stained with DAPI. Scale bar, 10 μm. e IHC staining of PIK3R1 in CDDP-resistant or CDDP- sensitive GC tissues from patients (400x magnification, scale bars = 50 μm). The IHC scores of PIK3R1 were further determined in 14 CDDP- resistant and 30 CDDP-sensitive patient tissues. f Three-dimensional scatter plot of circAKT3, miR-198 and PIK3R1 levels in 14 CDDP-resistant and 30 CDDP-sensitive GC tissues from patients. The results are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 increases the tolerance of GC cells to CDDP by targeting regulates PI3K signaling in CDDP resistance. Because PIK3R1 through miR-198. the PI3K pathway is a critical player in tumorigenesis and the ubiquitously hyperactivated signaling pathway in Discussion neoplasms, its inhibition both pharmacologically and CDDP treatment is one of the most predominant genetically is considered to be the most promising strat- chemotherapeutic strategies for patients with GC [25]. egy for targeted cancer treatment [38]. The PI3K path- In this study, using RNA-Seq analysis, we determined way has been revealed as a mediator of platinum that circRNA expression is associated with CDDP resist- resistance [39, 40] For instance, AKT activation medi- ance in GC. We found a novel circular RNA termed ated resistance to caspase-independent CDDP-induced circAKT3 that was upregulated in tissue samples from apoptosis through inhibiting the apoptosis-inducing patients with CDDP-resistant GC and in CDDP-resistant factor-associated pathway [40]. One study reported that cell lines and was correlated with five-year DFS. More- PI3K/AKT activation induced the upregulation of over, circAKT3 was expressed at higher levels than other BRCA1 in tamoxifen-resistant breast cancer cells and candidate circRNAs in CDDP-resistant GC patients, resensitized them to CDDP treatment [23]. Nevertheless, which meant that it may play a more important role the underlying mechanism of how PIK3R1 mediates re- than other circRNAs in GC. sistance to chemotherapy and whether this involves ca- DNA is the recognized target for CDDP cytotoxicity in nonical PI3K signaling and downstream BRCA1 activity cancer therapy. The resultant biological process in remains to be investigated in GC. response to CDDP and other DNA-damaging therapies Based on the abovementioned studies, we conducted a is the activation of apoptosis and the destruction of ma- series of experiments and demonstrated that circAKT3 lignant cells. The most favorable evidence is the hyper- reduced CDDP-induced activation of caspase-3 and sensitivity of both eukaryotic and prokaryotic cells apoptosis, leading to enhanced DDR and resistance to deficient in DNA repair to CDDP [26]. An enhanced CDDP chemotherapy. Mechanistically, circAKT3 func- DNA repair mechanism can induce the survival of tions as a ceRNA by sponging miR-198 to abolish the damaged or mutated tumor cells, resulting in resistance suppressive effect of this miRNA on its target gene and subsequent tumor recurrence [27]. BRCA1, a PIK3R1, which activated the PI3K/AKT signaling path- tumor-suppressor gene, is widely involved in cellular way in GC cells. The study showed that PI3K/AKT path- metabolism [28], transcriptional regulation [29, 30], and way activation contributes to upregulation of the DNA epigenetic modification [31]. A growing body of repair molecule BRCA1 and leads to resistance to evidence revealed that BRCA1 has a great effect on the CDDP-based DNA-damaging chemotherapy [23]. In the modulation of CDDP resistance [32–34]. Alterative current study, circAKT3 influenced DDR in GC cells, BRCA1 expression can regulate the mitochondrial fis- implying that circAKT3 might enhance CDDP resist- sion program, which could modulate CDDP sensitivity ance through the PI3K/AKT pathway and DDR mech- [32]. Some microRNAs contribute to DNA repair and anisms in GC cells. However, there are some CDDP sensitivity through BRCA1 deregulation, includ- drawbacks in this study, and we have not yet identi- ing miR-9 [35] and miR-638 [36]. A previous study dem- fied the specific mechanism by which BRCA1 regu- onstrated that BRCA1 mRNA levels were negatively lates DDR in GC. At the same time, the reasons for associated with CDDP sensitivity in GC [37]. Currently, circAKT3 shear formation and the upstream regula- there is a lack of research on circRNAs that regulate the tory mechanism were not discussed. For the in vivo BRCA1 gene in tumor cells, including GC. The under- experiments, we did not use animal models of GC in lying mechanisms require further exploration. situ, nor did we fully simulate the process of drug re- PIK3R1 encodes the regulatory subunit of PI3K sistance. These approaches should be further pursued (p85α). Some studies have reported that p85α positively in subsequent studies. Huang et al. Molecular Cancer (2019) 18:71 Page 18 of 20 Conclusions with circAKT3 or vector upon CDDP (1.5 μM) for 48 h by flow cytometry. In conclusion, we show that circAKT3 is upregulated in d Immunofluorescence staining of γH2AX foci in BGC823 cells transfected with circAKT3 or vector at 2 h after CDDP treatment (1.5 μM). Scale bars, human GC and that it can efficiently sponge miR-198 to 10 μm. e Percentage of cells containing >10 γH2AX foci in BGC823 cells restore PIK3R1 expression. We also demonstrate that transfected with circAKT3 or vector at 0 to 8 h after CDDP treatment (1.5 downregulation of circAKT3 can effectively promote μM) removal. The results are presented as the mean ± SEM. *P<0.05, CDDP sensitivity in GC cells by targeting the miR-198/ **P<0.01, ***P<0.001. (TIF 3594 kb) PIK3R1 axis. Our results provide novel evidence that cir- Additional file 6: Figure S4. a Predicted secondary structure of circAKT3 using the Vienna RNA package. b The expression of miR-198 cRNAs function as “microRNA sponges” and highlight a was analyzed using RT-qPCR in tissues of cohort 1. c, d, e The relative lu- promising therapeutic target for the CDDP resistance of ciferase activities were analyzed in 293T cells cotransfected with miR-198 GC patients. mimics or miR-NC and luciferase reporter vectors containing the WT or Mut 3’UTR of CHRM3 (C), HIPK2 (D), and MAFB (E). f & g The relative lucif- erase activities were analyzed in SGC7901CDDP(f) and BGC823CDDP(g) Additional files cells cotransfected with miR-198 mimics or miR-NC and luciferase re- porter vectors PIK3R1 3’UTR (WT) or PIK3R1 3’UTR (Mut). h & i The expres- Additional file 1: Table S1. Detailed information of cohort 1 and 2 is sion levels of PIK3R1 (h) and miR-198 (i) in SGC7901CDDP and listed. Table S2. Primers and RNA sequences used in this study. BGC823CDDP cells after transfection of PIK3R1 plasmids or miR-198 Table S3. SiRNA sequences used in this study. Table S4. MiRNA mimics, and inhibitors sequences used in this study. Table S5. Biotinylated probes mimics were detected by RT-qPCR. j The IC50 of miR-198 was analyzed sequences used in this study. Table S6. Fish probes sequences used in by the CCK8 assay. BGC823CDDP cells were transfected with miR-198 this study. Table S7. Detailed information of antibody used in this mimic alone or cotransfected with the indicated vectors upon CDDP ex- study. Table S8. Detailed information of 20 candidate circRNAs posure (15 μM) for 48 h. The results are presented as the mean ± SEM. (including location, genomic and spliced length). Table S9. Univariate *P<0.05, **P<0.01, ***P<0.001. (TIF 1650 kb) and multivariate analysis of cohort 2. (DOCX 31 kb) Additional file 7: Figure S5. a Apoptotic flow cytometry. SGC7901CDDP Additional file 2: Figure S1. a SGC7901CDDP/SGC7901 and cells were transfected with miR-198 mimic alone or cotransfected with the BGC823CDDP/BGC823 cell viability in response to different indicated vectors upon CDDP exposure (6 μM) for 48 h b The apoptosis rate concentrations of cisplatin. b RT-PCR products with divergent primers was analyzed by flow cytometry. BGC823CDDP cells were transfected with showing a single, distinct product of the expected size. c Melting curves miR-198 mimic alone or cotransfected with the indicated vectors upon of RT-qPCR product of verified circRNAs, indicating the specificity of RT- CDDP exposure (15 μM) for 48 h. c The expression levels of miR-198 in qPCR products with no primer dimers or nonspecific amplified products. d Schematic illustrating that circAKT3 (hsa_circ_0000199) is derived from SGC7901CDDP and BGC823CDDP cells after transfection of anti-miR-198 exons 8, 9, 10, and 11 of the AKT3 gene (555 bp). e Levels of small nucle- were detected by RT-qPCR. d Apoptotic flow cytometry. SGC7901CDDP cells olar RNA (U6, as a positive control for the nuclear fraction), GAPDH (posi- transfected with indicated vectors alone or cotransfected the inhibitors tive control for the cytoplasmic fraction), AKT3 mRNA and circRNAs from upon CDDP exposure (6 μM) for 48 h. The results are presented as the mean nuclear and cytoplasmic fractions of BGC823CDDP cells. f RNA stability of ±SEM. *P<0.05, **P<0.01, ***P<0.001. (TIF 2595 kb) the circular and linear transcripts of AKT3 and 18S rRNA in BGC823CDDP Additional file 8: Figure S6. a The IC50 was analyzed by CCK8 assay, cells. The results are presented as the mean ± SEM. *P<0.05, **P<0.01, SGC7901 cells were transfected with inhibitor alone or cotransfected with ***P<0.001. (TIF 2497 kb) the indicated vectors upon CDDP exposure (1.5 μM) for 48 h. b The Additional file 3: Additional Dataset (1) miRNA target prediction of apoptosis rate was analyzed by flow cytometry. SGC7901 cells were circAKT3. Additional Dataset (2) Eight miR-198 binding sites mutation of transfected with inhibitor alone or cotransfected with the indicated circAKT3. Additional Dataset (3) mRNA target prediction of miR-198. vectors upon CDDP exposure (1.5 μM) for 48 h. c Immunofluorescence (ZIP 1800 kb) staining of γH2AX foci in SGC7901 cells transfected with inhibitor alone Additional file 4: Figure S2. a RT-qPCR results for the circular and linear or cotransfected with the indicated vectors at 2 h after CDDP treatment transcripts of AKT3 in BGC823CDDP cells treated with or without siRNA (1.5 μM). Scale bars =10 μm. d Percentage of cells containing >10 γH2AX (NC, negative control; si-NC, control oligonucleotides with scramble se- foci in SGC7901 cells transfected with inhibitor alone or cotransfected quence; si-circ-1 and si-circ-2, oligonucleotides targeting the back-splice with the indicated vectors at 0 to 8 after CDDP treatment (1.5 μM) junction). b Relative cell viability of NC BGC823CDDP cells and removal. e SGC7901CDDP cells stably expressing circAKT3 siRNA (si-circ-1) BGC823CDDP cells transfected with si-NC- or si-circ-1 after CDDP treat- and its negative control siRNA (si-NC) were generated by infection with ment at the indicated concentrations for 48 h. c Colony-forming ability of lentiviruses, the expression levels of cricAKT3 were detected by RT-qPCR. the NC BGC823CDDP cells and si-NC- or si-circ-1-transfected The results are presented as the mean ± SEM. *P<0.05, **P<0.01, BGC823CDDP cells in the absence (Vehicle) or presence of CDDP (15 μM) ***P<0.001. (TIF 2160 kb) for 48 h. d The apoptosis rates of NC BGC823CDDP cells and Additional file 9: Figure S7. a The expression levels of PIK3R1 were BGC823CDDP cells transfected with si-NC or si-circ-1 in the absence (Ve- analyzed using RT-qPCR. SGC7901CDDP cells were cotransfected with inhibi- hicle) or upon CDDP (15 μM) for 48 h by flow cytometry. e Immunofluor- tors alone or the indicated vectors. b The IC50 of circAKT3 was analyzed by escence staining of γH2AX foci in NC BGC823CDDP cells and the CCK8 assay. SGC7901CDDP cells were transfected with inhibitor alone or BGC823CDDP cells transfected with si-NC or si-circ-1 at 2 h after CDDP cotransfected with the indicated vectors upon CDDP exposure (6 μM) for 48 treatment (15 μM). Scale bars, 10 μm. f Percentage of cells containing h. c The apoptosis rates of SGC7901CDDP cells transfected with inhibitor >10 γH2AX foci in NC BGC823CDDP cells and BGC823CDDP cells trans- alone or cotransfected with the indicated vectors upon CDDP exposure (6 fected with si-NC or si-circ-1 at 0 to 8 h after CDDP treatment (15 μM) re- moval. The results are presented as the mean ± SEM. *P<0.05, **P<0.01, μM) for 48 h. d Immunofluorescence staining of γH2AX foci in ***P<0.001. (TIF 5287 kb) SGC7901CDDP cells transfected with inhibitor alone or cotransfected with the indicated vectors at 2 h after CDDP treatment (6 μM). Scale bars, 10 μm. Additional file 5: Figure S3. a Relative cell viability of circAKT3- or e Percentage of cells containing >10 γH2AX foci in SGC7901CDDP cells vector-transfected BGC823 cells with CDDP treatment at the indicated transfected with inhibitor alone or cotransfected with the indicated vectors concentrations for 48 h. b Colony-forming ability of BGC823 cells trans- at 0 to 8 after CDDP treatment (6 μM) removal. f The expression levels of fected with circAKT3 or vector in the absence (Vehicle) or presence of PIK3R1, apoptosis markers, γH2AX, BRCA1 and PI3K/AKT signaling molecules CDDP (1.5 μM) for 48 h. c The apoptosis rates of BGC823 cells transfected Huang et al. Molecular Cancer (2019) 18:71 Page 19 of 20 Hospital of Wannan Medical College, Wuhu, Anhui, China. Department of were determined using western blotting in SGC7901CDDP cells transfected Neurosurgery, The First Affiliated Hospital of Nanjing Medical University, with inhibitor alone or cotransfected with the indicated vectors after CDDP Nanjing, China. treatment (6 μM). The results are presented as the mean ± SEM. *P<0.05, **P<0.01, ***P<0.001. (TIF 2258 kb) Received: 21 November 2018 Accepted: 21 February 2019 Additional file 10: Figure S8. a The expression of PIK3R1 were analyzed using western blot in tissue of cohort 1. b The expression of PIK3R1 were further determined in tissues of 14 CDDP-resistant and 30 References CDDP-sensitive patients (Relative to GAPDH). (TIF 604 kb) 1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin. 2015;65(2):87–108. Abbreviations 2. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, Parkin DM, AUC: Area under the receiver operating characteristic curve; BRCA1: Breast Forman D, Bray F. Cancer incidence and mortality worldwide: sources, cancer type 1 susceptibility protein; CCK8: Cell Counting Kit-8; methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5): CDDP: Cisplatin; ceRNAs: Competitive endogenous RNAs; circRNAs: Circular E359–86. RNAs; DDR: DNA damage repair; DFS: Disease-free survival; FBS: Fetal bovine 3. Zhao TT, Xu H, Xu HM, Wang ZN, Xu YY, Song YX, Yin SC, Liu XY, Miao ZF. serum; FISH: Fluorescence in situ hybridization; FL: Firefly luciferase; The efficacy and safety of targeted therapy with or without chemotherapy GC: Gastric cancer; gDNA: Genomic DNA; IHC: Immunohistochemical in advanced gastric cancer treatment: a network meta-analysis of well- staining; MFE: Minimum free energy; miRNAs: MicroRNAs; designed randomized controlled trials. Gastric Cancer. 2018;21(3):361–71. mtDNA: Mitochondrial DNA; ncRNAs: Noncoding RNAs; RNA-Seq: RNA 4. Koizumi W, Narahara H, Hara T, Takagane A, Akiya T, Takagi M, Miyashita K, sequencing; RT-PCR: Quantitative real-time PCR; RT-qPCR: Quantitative Nishizaki T, Kobayashi O, Takiyama W, et al. S-1 plus cisplatin versus S-1 reverse transcription PCR; γH2AX: Phosphorylated histone family member X alone for first-line treatment of advanced gastric cancer (SPIRITS trial): a phase III trial. Lancet Oncol. 2008;9(3):215–21. 5. Longley DB, Harkin DP, Johnston PG. 5-fluorouracil: mechanisms of action Acknowledgments and clinical strategies. Nat Rev Cancer. 2003;3(5):330–8. Jiangsu Key Lab of Cancer Biomarkers, Prevention and Treatment, 6. Shah MA. Update on metastatic gastric and esophageal cancers. J Clin Collaborative Innovation Center for Cancer Personalized Medicine, Nanjing Oncol. 2015;33(16):1760–9. Medical University. 7. Yang Z, Schumaker LM, Egorin MJ, Zuhowski EG, Guo Z, Cullen KJ. Cisplatin preferentially binds mitochondrial DNA and voltage-dependent anion Funding channel protein in the mitochondrial membrane of head and neck This work was supported by the National Natural Science Foundation of squamous cell carcinoma: possible role in apoptosis. Clin Cancer Res. 2006; China (81572362); the National Natural Science Foundation Project of 12(19):5817–25. International Cooperation (NSFC-NIH, 81361120398); the Primary Research & 8. Galluzzi L, Senovilla L, Vitale I, Michels J, Martins I, Kepp O, Castedo M, Development Plan of Jiangsu Province (BE2016786); the Program for Kroemer G. Molecular mechanisms of cisplatin resistance. Oncogene. 2012; Development of Innovative Research Team in the First Affiliated Hospital of 31(15):1869–83. NJMU; the Priority Academic Program Development of Jiangsu Higher 9. Galluzzi L, Vitale I, Michels J, Brenner C, Szabadkai G, Harel-Bellan A, Castedo Education Institutions (PAPD, JX10231801); 333 Project of Jiangsu Province M, Kroemer G. Systems biology of cisplatin resistance: past, present and (BRA2015474); Jiangsu Key Medical Discipline (General future. Cell Death Dis. 2014;5:e1257. Surgery)(ZDXKA2016005); University Natural Science Research Project of 10. Yu HG, Ai YW, Yu LL, Zhou XD, Liu J, Li JH, Xu XM, Liu S, Chen J, Liu F, et al. Anhui Province(KJ2018A0247); Wannan Medical College Cultivation Fund of Phosphoinositide 3-kinase/Akt pathway plays an important role in the Key Scientific Project(WK2017ZF10). chemoresistance of gastric cancer cells against etoposide and doxorubicin induced cell death. Int J Cancer. 2008;122(2):433–43. Availability of data and materials 11. Lee S, Choi EJ, Jin C, Kim DH. Activation of PI3K/Akt pathway by PTEN The datasets supporting the conclusions of this article are included within reduction and PIK3CA mRNA amplification contributes to cisplatin the article and its Additional files. resistance in an ovarian cancer cell line. Gynecol Oncol. 2005;97(1):26–34. 12. Siddik ZH. Cisplatin: mode of cytotoxic action and molecular basis of Authors’ contributions resistance. Oncogene. 2003;22(47):7265–79. ZKX, ZL and WZW conceived of the study and XXH carried out its design. 13. Kulcheski FR, Christoff AP, Margis R. Circular RNAs are miRNA sponges and XXH, QZ, BWL, LW, ZPX and XZ performed the experiments. XXH, XZ and can be used as a new class of biomarker. J Biotechnol. 2016;238:42–51. ZYH collected clinical samples. XXH, ALZ, QL, GLS and QL analyzed the data 14. Bagrodia A, Lee BH, Lee W, Cha EK, Sfakianos JP, Iyer G, Pietzak EJ, Gao SP, and wrote the paper. LJW, LZ and HX revised the paper. All authors read and Zabor EC, Ostrovnaya I, et al. Genetic determinants of cisplatin resistance in approved the final manuscript. patients with advanced germ cell tumors. J Clin Oncol. 2016;34(33):4000–7. 15. Li B, Wang W, Li Z, Chen Z, Zhi X, Xu J, Li Q, Wang L, Huang X, Wang L, et Ethics approval and consent to participate al. MicroRNA-148a-3p enhances cisplatin cytotoxicity in gastric cancer The present study was approved by the Hospital’s Protection of Human through mitochondrial fission induction and cyto-protective autophagy Subjects Committee. suppression. Cancer Lett. 2017;410:212–27. 16. Zheng Q, Bao C, Guo W, Li S, Chen J, Chen B, Luo Y, Lyu D, Li Y, Shi G, et al. Consent for publication Circular RNA profiling reveals an abundant circHIPK3 that regulates cell We have received consents from individual patients who have participated growth by sponging multiple miRNAs. Nat Commun. 2016;7:11215. in this study. The consent forms will be provided upon request. 17. Wang K, Long B, Liu F, Wang JX, Liu CY, Zhao B, Zhou LY, Sun T, Wang M, Yu T, et al. A circular RNA protects the heart from pathological hypertrophy Competing interests and heart failure by targeting miR-223. Eur Heart J. 2016;37(33):2602–11. The authors declare that they have no competing interests. 18. Hsiao KY, Lin YC, Gupta SK, Chang N, Yen L, Sun HS, Tsai SJ. Noncoding effects of circular RNA CCDC66 promote Colon Cancer growth and Publisher’sNote metastasis. Cancer Res. 2017;77(9):2339–50. Springer Nature remains neutral with regard to jurisdictional claims in 19. Han D, Li J, Wang H, Su X, Hou J, Gu Y, Qian C, Lin Y, Liu X, Huang M, et al. published maps and institutional affiliations. Circular RNA circMTO1 acts as the sponge of microRNA-9 to suppress hepatocellular carcinoma progression. Hepatology. 2017;66(4):1151–64. Author details 20. Huang R, Zhang Y, Han B, Bai Y, Zhou R, Gan G, Chao J, Hu G, Yao H. Department of General Surgery, The First Affiliated Hospital of Nanjing Circular RNA HIPK2 regulates astrocyte activation via cooperation of Medical University, 300 Guangzhou Road, Nanjing 210029, Jiangsu province, autophagy and ER stress by targeting MIR124-2HG. Autophagy. 2017;13(10): China. Department of Gastrointestinal Surgery, The First Affiliated Yijishan 1722–41. Huang et al. Molecular Cancer (2019) 18:71 Page 20 of 20 21. Lu Y, Xiao L, Liu Y, Wang H, Li H, Zhou Q, Pan J, Lei B, Huang A, Qi S. MIR517C inhibits autophagy and the epithelial-to-mesenchymal (−like) transition phenotype in human glioblastoma through KPNA2-dependent disruption of TP53 nuclear translocation. Autophagy. 2015;11(12):2213–32. 22. Kinner A, Wu W, Staudt C, Iliakis G. Gamma-H2AX in recognition and signaling of DNA double-strand breaks in the context of chromatin. Nucleic Acids Res. 2008;36(17):5678–94. 23. Zhu Y, Liu Y, Zhang C, Chu J, Wu Y, Li Y, Liu J, Li Q, Li S, Shi Q, et al. Tamoxifen- resistant breast cancer cells are resistant to DNA-damaging chemotherapy because of upregulated BARD1 and BRCA1. Nat Commun. 2018;9(1):1595. 24. Cheung LW, Mills GB. Targeting therapeutic liabilities engendered by PIK3R1 mutations for cancer treatment. Pharmacogenomics. 2016;17(3):297–307. 25. Mo D, Fang H, Niu K, Liu J, Wu M, Li S, Zhu T, Aleskandarany MA, Arora A, Lobo DN, et al. Human helicase RECQL4 drives cisplatin resistance in gastric Cancer by activating an AKT-YB1-MDR1 signaling pathway. Cancer Res. 2016;76(10):3057–66. 26. Fraval HN, Rawlings CJ, Roberts JJ. Increased sensitivity of UV-repair- deficient human cells to DNA bound platinum products which unlike thymine dimers are not recognized by an endonuclease extracted from Micrococcus luteus. Mutat Res. 1978;51(1):121–32. 27. Turner KM, Sun Y, Ji P, Granberg KJ, Bernard B, Hu L, Cogdell DE, Zhou X, Yli-Harja O, Nykter M, et al. Genomically amplified Akt3 activates DNA repair pathway and promotes glioma progression. Proc Natl Acad Sci U S A. 2015; 112(11):3421–6. 28. Li D, Chen NN, Cao JM, Sun WP, Zhou YM, Li CY, Wang XX. BRCA1 as a nicotinamide adenine dinucleotide (NAD)-dependent metabolic switch in ovarian cancer. Cell Cycle. 2014;13(16):2564–71. 29. Dacheux E, Vincent A, Nazaret N, Combet C, Wierinckx A, Mazoyer S, Diaz JJ, Lachuer J, Venezia ND. BRCA1-dependent translational regulation in breast Cancer cells. PLoS One. 2013;8(6):e67313. 30. Li D, Bi FF, Chen NN, Cao JM, Sun WP, Zhou YM, Cao C, Li CY, Yang Q. Epigenetic repression of phosphatidylethanolamine N-methyltransferase (PEMT) in BRCA1-mutated breast cancer. Oncotarget. 2014;5(5):1315–25. 31. Li D, Bi FF, Cao JM, Cao C, Liu B, Yang Q. Regulation of DNA methyltransferase 1 transcription in BRCA1-mutated breast cancer: a novel crosstalk between E2F1 motif hypermethylation and loss of histone H3 lysine 9 acetylation. Mol Cancer. 2014;13:26. 32. Fan S, Liu B, Sun L, Lv XB, Lin Z, Chen W, Chen W, Tang Q, Wang Y, Su Y, et al. Mitochondrial fission determines cisplatin sensitivity in tongue squamous cell carcinoma through the BRCA1-miR-593-5p-MFF axis. Oncotarget. 2015; 6(17):14885–904. 33. Paul BT, Blanchard Z, Ridgway M, ElShamy WM. BRCA1-IRIS inactivation sensitizes ovarian tumors to cisplatin. Oncogene. 2015;34(23):3036–52. 34. Shin DH, Choi YJ, Park JW. SIRT1 and AMPK mediate hypoxia-induced resistance of non-small cell lung cancers to cisplatin and doxorubicin. Cancer Res. 2014;74(1):298–308. 35. Sun C, Li N, Yang Z, Zhou B, He Y, Weng D, Fang Y, Wu P, Chen P, Yang X, et al. miR-9 regulation of BRCA1 and ovarian cancer sensitivity to cisplatin and PARP inhibition. J Natl Cancer Inst. 2013;105(22):1750–8. 36. Tan X, Peng J, Fu Y, An S, Rezaei K, Tabbara S, Teal CB, Man YG, Brem RF. Fu SW: miR-638 mediated regulation of BRCA1 affects DNA repair and sensitivity to UV and cisplatin in triple-negative breast cancer. Breast Cancer Res. 2014;16(5):435. 37. Wang L, Wei J, Qian X, Yin H, Zhao Y, Yu L, Wang T, Liu B. ERCC1 and BRCA1 mRNA expression levels in metastatic malignant effusions is associated with chemosensitivity to cisplatin and/or docetaxel. BMC Cancer. 2008;8:97. 38. Hennessy BT, Smith DL, Ram PT, Lu Y, Mills GB. Exploiting the PI3K/AKT pathway for cancer drug discovery. Nat Rev Drug Discov. 2005;4(12):988– 39. Hui RC, Francis RE, Guest SK, Costa JR, Gomes AR, Myatt SS, Brosens JJ, Lam EW. Doxorubicin activates FOXO3a to induce the expression of multidrug resistance gene ABCB1 (MDR1) in K562 leukemic cells. Mol Cancer Ther. 2008;7(3):670–8. 40. Yang X, Fraser M, Abedini MR, Bai T, Tsang BK. Regulation of apoptosis- inducing factor-mediated, cisplatin-induced apoptosis by Akt. Br J Cancer. 2008;98(4):803–8.

Journal

Molecular CancerSpringer Journals

Published: Mar 30, 2019

There are no references for this article.