Long Noncoding RNA GCASPC, a Target of miR-17-3p, Negatively Regulates Pyruvate Carboxylase–Dependent Cell Proliferation in Gallbladder Cancer

Gallbladder cancer is the most common biliary tract cancer and the fifth most common gastrointestinal malignancy worldwide (1). The prognosis of gallbladder cancer remains extremely poor despite recent advances in gallbladder cancer treatment, with a median survival time of 9.2 months for suspected carcinomas and 26.5 months for incidental gallbladder cancer (2, 3). Although great efforts have been put into clarifying the pathophysiologic mechanisms contributing to the progression of gallbladder cancer, much of it remains unknown (4, 5). Thus, it is vital to reveal the molecular mechanisms of gallbladder carcinogenesis to facilitate development of novel cancer biomarkers and appropriate therapeutic strategies.

Long noncoding RNAs (lncRNA), a subgroup of noncdoing RNAs (ncRNA), are longer than 200 nucleotides in length and with little protein-coding potential (6, 7). lncRNAs are abundantly expressed in mammalian cells, and a number of them have been identified as critical regulators in a diverse array of cellular processes via controlling multiple levels of the gene expression, including carcinogenesis (8–11). A growing volume of literature has demonstrated that lncRNAs expression profiling may facilitate the diagnosis of human cancers (12–14). They have the potential to serve as prognostic indicators and therapeutic targets. Although over 95,000 human lncRNAs have been annotated (15), only a few of them have been functionally characterized. In our previous studies, we have identified several dysregulated lncRNAs in gallbladder cancer (16–18). Yet, there have been no systematic profiling studies of lncRNAs in gallbladder cancer up until now.

Although emerging evidence has shown the paramount role of lncRNAs in tumor development, only a small portion of them, such as HOX transcript antisense RNA (HOTAIR) and metastasis-associated lung adenocarcinoma transcript 1 (MALAT1), have been well characterized in various carcinomas (6). Unlike the well-established molecular mechanism of miRNAs action (19), which is based on seed sequence base pairing, the action mode of lncRNAs remains to be explored. Studies revealed that lncRNAs may interact with DNA, RNA, or protein and regulate a large number of genes with different mechanisms, thereby affecting a variety of cellular pathways (9–15). Thus, the molecular mechanisms of lncRNAs action can be diversified and require intensive investigations.

In the present study, through transcriptome microarray analysis, we found a number of lncRNAs dysregulated in gallbladder cancer compared with paired nontumoral tissues. Among the downregulated lncRNAs, we further characterized the clinicopathologic relevance of a novel gallbladder cancer–associated suppressor of pyruvate carboxylase lncRNA (lncRNA GCASPC) in gallbladder cancer progression. GCASPC interacted with pyruvate carboxylase (PC) protein in gallbladder cancer cells, and the antiproliferative functions of GCASPC can be neutralized by PC. We provided in vitro and in vivo data to demonstrate that GCASPC, which is a target of miR-17-3p, suppressed cell proliferation in gallbladder cancer by destabilization of PC protein.

Briefly, samples (five gallbladder cancer tissues and five corresponding nontumor tissues; Supplementary Table S1) were used to synthesize double-stranded complementary DNA (cDNA), and double-stranded cDNA was labeled and hybridized to Human Gene 2.0 arrays (Affymetrix) according to the manufacturer's protocol, and Affymetrix Expression Console Software (version 1.3.1) was used for microarray analysis. Raw data (CEL files) were normalized at the transcript level using the robust multiaverage method (RMA workflow). Median summarization of transcript expression was calculated. The random variance model (RVM) t test was used to identify differentially expressed genes between the gallbladder cancer and nontumoral groups. A P value was calculated using the paired t test. The threshold set for dysregulated genes was a fold change >2.0 and a P value <0.05. Hierarchical clustering (Cluster3.0) and TreeView analysis (Stanford University, Stanford, CA) were performed based on the results of differentially expressed genes. Data are available via Gene Expression Omnibus (GEO) GSE62335.

The human specimens in this study were sanctioned by the local ethics committee at the Shanghai Jiao Tong University School of Medicine, Xinhua Hospital (Shanghai, China). Two independent cohorts involving 131 gallbladder cancer patients were enrolled in this study. Forty-two fresh gallbladder cancer tissue pairs were collected from patients at Xinhua Hospital (Shanghai Jiao Tong University School of Medicine) from April 2008 to May 2013. Another 89 fresh gallbladder cancer tissue pairs were collected from Eastern Hospital of Hepatobiliary (Second Military Medical College, Shanghai, China) from August 2007 to September 2014 and used for further validation. Tissue samples were collected in the operating room and processed immediately within 15 minutes. Patients' clinical information is listed in Supplementary Tables S2 and S3. The data do not contain any information that could identify patients. None of the patients received preoperative treatment, including chemotherapy or radiotherapy. The nontumorous samples were taken at a distance of at least 5 cm from the tumor, and all tissues were examined histologically.

Four human gallbladder cancer cell lines (GBC-SD, SGC-996, NOZ, and OCUG-1) were used in this study. GBC-SD, SGC-996, NOZ, OCUG-1, and the nontumorigenic human intrahepatic biliary epithelial cell line H69 were purchased from the Health Science Research Resources Bank on July 2013 where they were characterized by Mycoplasma detection, DNA fingerprinting, isozyme detection, and cell vitality detection. The last cell characterization with the above methods was performed on March 2015. These cell lines were immediately expanded and frozen such that they could be restarted every 3 to 4 months from a frozen vial of the same batch of cells. Cells were cultured at 37°C in an atmosphere of 5% CO₂ in DMEM (Gibco BRL) supplemented with 10% FBS, penicillin, and streptomycin (Thermo Scientific). The passage numbers for GBC-SD, SGC-996, NOZ, OCUG-1, and H69 were 16, 11, 9, 23, and 10, respectively. All cell lines have been passaged for fewer than 6 months in our laboratory after resuscitation.

Rapid amplification of cDNA ends (RACE) analysis and subcellular fractionation analysis were performed as described previously (18). We determined the protein-coding potential of transcript using an in vitro translation assay and a combination of protein-coding potential assessment software.

Detailed descriptions of plasmid construction, lentiviral vector construction, and cell transfections can be found in the Supplementary Materials and Methods.

RNA preparation, quantitative real-time PCR (qRT-PCR), and Western blot analysis were performed as described previously (18).

Cell proliferation was determined with the Cell Counting Kit-8 (CCK-8). Cell cycle was determined with flow cytometric analysis.

Nude mice (age 4∽5 weeks) were purchased from the Shanghai Experimental Animal Center of the Chinese Academy of Sciences, Shanghai, and housed in a pathogen-free facility in the Experimental Animal Centre of Xinhua hospital. All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the US NIH (NIH publication number 85-23, revised 1996). Stably overexpressing or silencing gallbladder cancer cells diluted to a concentration of 1 × 10⁷ cells/mL in physiologic saline. Mice were subcutaneously injected with 0.1 mL of the suspension into either side of flank area. Tumor volumes were measured (0.5 × length × width²) in mice every 5 days. After 30 days, mice were sacrificed, and tumors were weighed, exercised, and subjected to immunohistochemical analysis of Ki67.

The experiments were performed as described by Li and colleagues (12).

All statistical analyses were performed using SPSS version 17.0 software. All data are presented as the mean ± SD. Unless otherwise noted, the differences between two groups were analyzed using the Student t test. The Kaplan–Meier method was used to calculate survival, and significance was determined by the log-rank test. Multivariate logistic regression was performed to identify the independent factors related to gallbladder cancer prognosis. The relationship between GCASPC expression levels and clinical parameters was assessed with the nonparametric Mann–Whitney–Wilcoxon test. Risk score analysis was performed to investigate the effectiveness of the GCASPC for prediction. Correlations between GCASPC and miR-17-3p were analyzed by Spearman rank correlation. P values were two-sided, and a value of <0.05 was considered to be statistically significant. One asterisk and two asterisks indicate P < 0.05 and P < 0.01, respectively.

To identify transcripts that potentially drive gallbladder tumorigenesis, lncRNAs and mRNAs expression profiles were determined by microarray analysis. A hierarchical clustering analysis showed systematic variations in transcript expression levels between gallbladder cancer tissues and paired adjacent nontumor tissues from 5 gallbladder cancer patients (Fig. 1A). To validate our microarray findings, we randomly selected differentially expressed transcripts (10 lncRNAs and 8 mRNAs) and analyzed their expression, using qRT-PCR, in 15 pairs of randomly selected gallbladder cancer and corresponding nontumor tissues from cohort 1 (Supplementary Figs. S1 and S2). Thus, qRT-PCR analysis confirmed our microarray findings, indicating that a set of lncRNAs are frequently aberrantly expressed in gallbladder cancer tissues.

In the present study, we focused on the lncRNAs that are significantly downregulated in gallbladder cancer tissues. We identified a modestly conserved GCASPC (lnc-SOD2-1:1, LNCipedia annotation; NONHSAT115853, NONCODE v4; RP1-56L9.7-001, GENCODE v13) on human chromosome 6;160060339-160061133 as one of the top ranked candidates with a significant P value (P = 0.003; P < 0.05). We noted that GCASPC was located within the intron of insulin-like growth factor 2 receptor (IGF2R; Supplementary Fig. S3A). To explore the potential relationship of the GCASPC and IGF2R transcripts, we first examined the expression levels in 27 gallbladder cancer tissues (cohort 1). The results showed that no correlation (r² = 0.031; P = 0.377) existed between the transcript levels of GCASPC and IGF2R (Supplementary Fig. S3B). Furthermore, GCASPC was statistically unchanged in GBC-SD and SGC-996 cells transfected with two different siRNAs (designated as si-1 and si-2) against IGF2R, despite significant reduction in IGF2R messenger RNA expression (Supplementary Fig. S3C). IGF2R was not significantly changed in GBC-SD cells with two different shRNAs (shRNAs, designated as sh-1 and sh-2) against GCASPC (Supplementary Fig. S3D). GCASPC is polyadenylated (Supplementary Fig. S3E). GCASPC was composed of two exons and spanned nearly 740 base pairs (bp), identifying it as a modestly conserved locus (Supplementary Fig. S3F). The sequence of full-length GCASPC is presented in Supplementary Fig. S4. We verified that GCASPC was indeed a noncoding RNA with an in vitro translation assay (Supplementary Fig. S5A) and online protein-coding potential assessment softwares (Supplementary Fig. S5B and S5C). Subcellular fractionation analysis revealed GCASPC is mainly located in the cytoplasm of gallbladder cancer cells (Supplementary Fig. S5D). The RNAfold image is presented in Supplementary Fig. S5E and S5F. Expression levels of GCASPC were significantly downregulated in gallbladder cancer cell lines compared with nontumorigenic human intrahepatic biliary epithelial cell line H69, and GCASPC copy number in gallbladder cancer cells varied from 40 to 130 copies per cell (Supplementary Fig. S6A). The expression of GCASPC is comparable with H19, HOTAIR, CCAT1, and MALAT1 (Supplementary Fig. S6B). We did not find any changes in GCASPC expression levels in gallbladder cancer cells treated with DNA methylation inhibitor 5-azacytidine (Supplementary Fig. S6C). We determined that GCASPC was upregulated by the histone deacetylase inhibitor trichostatin A (TSA) in SGC-996 cells (Supplementary Fig. S6D). These results indicate that GCASPC expression in gallbladder cancer is likely to be regulated by histone acetylation.

To further investigate the role of GCASPC in gallbladder cancer, we examined 42 paired gallbladder cancer/nontumor tissue specimens (Supporting Table 2, cohort 1). The transcript levels of GCASPC were significantly lower in gallbladder cancer tissues, after normalizing to U6 expression (P < 0.001; Fig. 1B). Furthermore, receiver operating characteristic (ROC) curves were determined to evaluate the sensitivity and specificity of GCASPC expression in predicting gallbladder cancer tissues from normal tissues. Notably, GCASPC displayed considerable predictive significance, with an area under the curve (AUC) of 0.697 [95% confidence interval (CI), 0.584–0.810; P = 0.002; Fig. 1C].

According to the median ratio of relative GCASPC expression in tumor tissues, the gallbladder cancer patients were classified into two groups: High-GCASPC group: GCASPC expression ratio ≥ median ratio; and Low-GCASPC group: GCASPC expression ratio ≤ median ratio. To determine whether GCASPC expression level in gallbladder cancer was associated with specific clinicopathologic characteristics, we measured GCASPC expression levels in tumor tissues from another 89 gallbladder cancer patients independent from 42 gallbladder cancer patients of cohort 1 (Supplementary Table S3, cohort 2) by qRT-PCR. As demonstrated in Table 1, a lower GCASPC expression level was significantly more frequent in tissues with increased tumor size (P = 0.039; P < 0.05) and advanced American Joint Committee on Cancer (AJCC) tumor stage (P = 0.011; P < 0.05) in cohort 2. These associations were confirmed by analysis of samples from cohort 1 (Supplementary Table S4). Furthermore, Kaplan–Meier and log-rank test analyses suggested a correlation between low tumoral GCASPC expression and reduced overall survival (OS) and disease-free survival (DFS) rates (P < 0.001 for both OS and DFS; Fig. 1E and F). In addition, multivariate analysis showed that GCASPC expression (95% CI, 1.421–5.034; P = 0.006) and local invasion status (95% CI, 1.342–5.579; P = 0.005) were independent factors that affected the OS of gallbladder cancer patients after radical gallbladder cancer resection (Supplementary Table S5).

To evaluate the biological effects of GCASPC on development of gallbladder cancer, we performed gain- and loss-of-function studies in gallbladder cancer cells. After detecting the expression levels of GCASPC in a variety of gallbladder cancer cell lines (Supplementary Fig. S6A), we constructed cell lines with stable GCASPC overexpression and downregulation (Supplementary Fig. S7A–S7D). Cell-counting kit-8 assays indicated that exogenous expression of GCASPC decreased the proliferative capacity of SGC-996 (Fig. 2A) and NOZ cells (Supplementary Fig. S8A), compared with that of parallel stable cell lines containing the empty vector. Consistent with decreased cell proliferation, GCASPC-overexpressing SGC-996 (Fig. 2A) and NOZ (Supplementary Fig. S8A) cells exhibited lower levels of proliferating cell nuclear antigen (PCNA) expression. Conversely, cell proliferation and PCNA expression were increased in GBC-SD (Fig. 2B) and OCUG-1 (Supplementary Fig. S8B) cells when GCASPC expression was knocked down. These data suggest that GCASPC plays a physiologic role in regulating gallbladder cancer cell proliferation. Next, we examined differences in cell-cycle distributions following GCASPC overexpression or silencing by FACS analysis of propidium-iodide–stained cells. GCASPC overexpression resulted in significant G₁–S arrest in SGC-996 and NOZ cells, whereas cell-cycle progression beyond the G₁–S transition was observed in GCASPC-knockdown GBC-SD and OCUG-1 cells (Fig. 2C and D; Supplementary Fig. S8C and S8D).

The growth-suppressive effect of GCASPC was confirmed by in vivo tumor growth assays. Our results showed that the growth of tumors from GCASPC-overexpressed xenografts was significantly inhibited as demonstrated by decreased mean volumes and weights as well as slower tumor growth rates, and the growth of tumors from GCASPC-downregulated xenografts was significantly promoted, compared with that of tumors formed from control xenografts. Moreover, immunohistochemical staining of tumor tissues indicated a decrease in Ki67 in GCASPC-upregulated xenografts versus vector-transduced xenografts. In contrast, GCASPC-knockdown xenografts showed stronger staining for Ki67 (Fig. 2E and F; Supplementary Fig. S8E and S8F). In summary, these data suggest that GCASPC suppressed gallbladder cancer cell proliferation.

We sought to explore the molecular mechanisms by which GCASPC exerts its effects on gallbladder cancer cell proliferation. As lncRNAs have been reported to exert cis-regulatory effects on nearby genes (20), we examined whether manipulation of GCASPC expression levels would affect the mRNA levels of its in cis genes. As demonstrated in Supplementary Fig. S9A, no statistical changes in the transcript levels of neighboring genes were observed in SGC-996 cells with GCASPC overexpression, as represented by SOD2, MAS1, loc729603, AIRN, and SLC22A1. It suggests that GCASPC may act in trans. Recent studies have suggested that lncRNAs participate in molecular regulation pathways through interacting with proteins (9, 12). Thus, we hypothesized that GCASPC might function through a similar mechanism. To test this hypothesis, we performed RNA pull-down assays to identify proteins associated with GCASPC RNA in NOZ cells as previously described (9). RNA-associated proteins were analyzed by SDS/PAGE and silver staining (Fig. 3A). Three distinct bands specific to GCASPC were excised and subjected to mass spectrometry (Supplementary Table S6). PC was detected by Western blotting from three independent RNA pull-down assays in cell extracts from SGC-996 and NOZ cells (Fig. 3B). The specificity of this interaction was further verified with RNA immunoprecipitation (RIP; Fig. 3C). Notably, deletion-mapping analyses identified that 3′-end segment (472–741 nt) of GCASPC is required for the association with PC (Fig. 3D). RNA folding analyses (21) of this 3prime; region indicated a stable stem-loop structure (Supplementary Fig. S9B), which might provide the necessary spatial conformation for the interaction. In addition, we found that PC was significantly upregulated in gallbladder cancer tissues compared with adjacent nontumor tissues (Supplementary Fig. S9C). As mentioned above (Supplementary Fig. S5D), GCASPC is mainly located in the cytoplasm of gallbladder cancer cells. We further detected the subcellular fractionation of GCASPC because PC is a mitochondria protein. We isolated pure mitochondria fractionation via standard cellular fractionation methods and found that GCASPC was mainly located in the mitochondria fraction (Fig. 3E). Then, we analyzed the interaction between GCASPC and PC in the cytoplasmic (without mitochondria) and mitochondria fraction. The RIP analysis demonstrated that the interaction between GCASPC and PC specifically takes place in the mitochondria fraction (Fig. 3F).

Next, we sought to characterize the effects of GCASPC on PC. We detected a significant upregulation of the PC protein and PC activity upon GCASPC knockdown in both GBC-SD (Fig. 4A) and OCUG-1 cells (Supplementary Fig. S10A), and a downregulation of the PC protein and PC activity in GCASPC-overexpressing SGC-996 (Fig. 4A) and NOZ cells (Supplementary Fig. S10B), but we did not observe a significant change in PC mRNA levels (Supplementary Fig. S10C and S10D). As 3′-end segment (472–741 nt) of GCASPC is required for the association with PC, we overexpressed the truncated version (472–741 nt) and analyzed its impact on cell proliferation, PC protein expression, and activity. The data demonstrated that the truncated version could significantly suppress cell proliferation, PC expression, and activity (Supplementary Fig. S10E and S10F).

Based on this finding, we hypothesize that GCASPC binds to PC and affects its biological activity at the translational or posttranslational level. To identify these hypotheses, we firstly observed the expression of PC proteins in gallbladder cancer cells incubated with the protein synthesis inhibitor cycloheximide (CHX). As shown in Fig. 4B, CHX decreased the expression of PC proteins by inhibiting protein synthesis. However, knockdown GCASPC still induced the upregulation of PC protein levels under the treatment of CHX (Fig. 4B). These results suggest that GCASPC might promote the PC protein degradation. We used the proteasome inhibitor MG-132 to further clarify the possible mechanism. Firstly, ectopic expression of GCASPC downregulated the protein levels of PC (Fig. 4A), suggesting that GCASPC destabilized PC protein. As illustrated in Fig. 4C, MG-132 upregulated the protein levels of PC, suggesting that the inhibition of ubiquitination-proteasome pathway might ameliorate the degradation of PC. The last but not the least, MG-132 abolished the reduction of PC protein levels in GCASPC-overexpressing SGC-996 (Fig. 4C) and NOZ cells (Supplementary Fig. S11A). We further examined whether GCASPC affects PC protein stability by performing an ubiquitination assay and found that the PC ubiquitination level was significantly higher in cells that overexpressed GCASPC relative to control cells (Supplementary Fig. S11B). These data indicate that GCASPC downregulates PC protein abundance via the ubiquitination-proteasome pathway. Functionally, PC inhibition suppressed the proliferation of gallbladder cancer cells (Fig. 4D; Supplementary Fig. S11C), and PC overexpression promoted the proliferation of GBC-SD cells and abrogated the effect of GCASPC overexpression on suppressing cell proliferation (Fig. 4E), indicating that the function of GCASPC depends on PC. Collectively, these data suggest that GCASPC suppressed tumorigenesis by negatively regulating PC-dependent cell proliferation.

A competitive RNA (ceRNA) hypothesis has been proposed and recent studies have suggested the existence of the interaction between lncRNAs and miRNAs (10, 11, 16), imposing an additional level of posttranscriptional regulation. We performed a search for miRNAs that have complementary base pairing with GCASPC, using online software program miRDB (http://mirdb.org; ref. 22). The search results demonstrated that 30 miRNAs formed complementary base pairing with GCASPC (Supplementary Table S5). As miR-17-3p achieved the highest score according to miRDB and formed no complementary base pairing with PC mRNA according to Targetscan, we selected it for further studies. What is more, other miRNAs on the list shared no common interaction site with miR-17-3p (Supplementary Table S7). According to the prediction results, there was one putative miR-17-3p binding site in exon 2 of GCASPC (Fig. 5A). To confirm the direct binding between GCASPC and miR-17-3p, luciferase reporter constructs were generated. We observed that miR-17-3p mimics reduced the luciferase activities of wild-type (WT) GCASPC reporter vector, but not a mutant GCASPC, indicating that miR-17-3p binds to GCASPC in a sequence-specific manner. We further clarified the regulatory relationship between GCASPC and miR-17-3p. Overexpression of miR-17-3p significantly suppressed the expression of GCASPC in GBC-SD and OCUG-1 cells (Fig. 5B). In contrast, inhibition of miR-17-3p enhanced the expression of GCASPC in SGC-996 and NOZ cells (Supplementary Fig. S12A). However, there was no obvious difference in miR-17-3p level after overexpression or knockdown of GCASPC (Fig. 5B; Supplementary Fig. S12B). It suggests that GCASPC is targeted by miR-17-3p. To distinguish between a transcriptional and a posttranscriptional mechanism, we treated GBC-SD cells with alpha-amanitin, which blocked RNA Polymerase II transcription. This experiment revealed that overexpression of miR-17-3p decreased the GCASPC half-life (Fig. 5C). The miRNAs are known to bind their targets and cause translational repression and/or RNA degradation in an Ago2-dependent manner. We performed RNA pull-down experiments by using GCASPC probe and then examined Ago2 and miR-17-3p simultaneously as described previously (18) to determine whether GCASPC and miR-17-3p are in the same RISC complex. The in vitro RNA pull-down experiment was performed to confirm the direct physical association between GCASPC and Ago2. As a result, we detected Ago2 (Supplementary Fig. S12C). Furthermore, we detected miR-17-3p in the same pellet, supporting that miR-17-3p is bona fide GCASPC-targeting miRNA (Supplementary Fig. S12D). Furthermore, miR-17-3p knockdown suppressed the proliferation and expression levels of PC in GCASPC-knockdown GBC-SD cells (Fig. 5D and E). A statistically significant inverse correlation was observed between GCASPC and miR-17-3p transcript levels in 42 gallbladder cancer specimens (r = –0.498, P = 0.002, Fig. 5F). In general, these data suggest that miR-17-3p directly binds to GCASPC and negatively regulates GCASPC-mediated tumor-suppressive activity.

The molecular classification of gallbladder cancer has identified a number of protein-coding genes as valuable biomarkers and prognostic indicators (4, 5, 23). However, a poor overlap exists between these biomarkers of gallbladder cancer. Thus, it might be a better resolution to establish more-accurate prognostic gene signatures by using a combination of different types of transcripts (24). A growing volume of literature has demonstrated that noncoding RNAs, predominantly miRNAs, could serve as potential biomarkers of gallbladder cancer (3, 25). Given the fact that lncRNAs are more abundantly expressed in mammalian cells, it is plausible to speculate that lncRNAs, once regarded as “transcriptional noise,” may be potential prognostic indicators in gallbladder cancer. Although thousands of lncRNAs have been annotated (15), functional interpretation has just started.

In the present study, we revealed signatures of a small number of lncRNAs that are aberrantly expressed in human gallbladder cancer, compared with nontumor tissues. We identified a new lncRNA transcript (GCASPC), which was significantly downregulated in gallbladder cancer tissues from two cohorts of patients. We determined that the low expression level of GCASPC was significantly associated with numerous clinicopathologic characteristics, including tumor size, AJCC stage, frequent recurrence, and cancer-related death. A multivariate analysis revealed that GCASPC expression level was an independent risk factor for OS after surgery. These data suggest that GCASPC can be a potential prognostic indicator for gallbladder cancer.

By applying loss- and gain-of-function approaches, we identified that GCASPC plays a role in cell proliferation and cell-cycle progression. Although it has been suggested that some lncRNAs act in cis (neighboring genes) through transcriptional interference, the majority of lncRNAs primarily function in trans by targeting multiple chromatin regulatory proteins to distant genes (26). Here, we identified that GCASPC had no effect on the expression level of neighboring genes, suggesting that GCASPC functions in trans. The GCASPC transcript was found to associate with PC to promote PC protein degradation. However, the underlying mechanisms of how GCASPC regulates the ubiquitination and/or ubiquitination associated enzymes require further investigation. PC, an enzyme that converts pyruvate to oxaloacetate, has recently been demonstrated to play an important role in cancer cell metabolism and proliferation (27, 28). We also found that the GCASPC abrogated PC-mediated gallbladder cancer cell proliferation, indicating that GCASPC function in a PC-dependent manner. Enhanced glycolysis under aerobic conditions (the Warburg effect) has been a hallmark of cancer for many decades (29). However, accelerated glycolysis alone is insufficient to meet the total metabolic demands of proliferating cancer cells. The Krebs cycle is also a source of energy via the oxidation of pyruvate, fatty acids, and amino acids such as glutamine (27). Continued functioning of the Krebs cycle requires the replenishment of intermediates that are diverted for anabolic uses or glutathione synthesis, which was accomplished via two major pathways: carboxylation of pyruvate to oxaloacetate via ATP-dependent PC (27) and glutaminolysis (30). We presume that the inhibition of gallbladder cancer cell proliferation by PC suppression with GCASPC upregulation is accompanied by a decrease in anaplerotic input into the Krebs cycle, which has been verified in non-small cell lung cancer cells with PC suppression (27). However, whether it is true in this case requires further investigation.

A growing number of reports suggest the existence of a widespread interaction network involving ceRNAs, where ncRNAs could regulate modulatory RNA by binding and titrating them off their binding sites on protein coding messengers (31, 32). Inspired by the discoveries of the interaction between lncRNAs and miRNAs (33, 34), we sought to identify the role of miRNAs in the regulation of lncRNAs. Luciferase assays indicated that miR-17-3p directly binds to GCASPC. MiR-17-3p overexpression silenced GCASPC in gallbladder cancer cells. Furthermore, GCASPC transcript level was inversely correlated with miR-17-3p mRNA level in gallbladder cancer tissues. However, the alignment between the GCASPC and miR-17-3p is not very specific, as 30 miRNAs were predicted to form complementary base pairing with GCASPC. What is more, miR-17-3p may also act independently of miR-17-3p, as it shares homology with a number of protein-coding genes such as TIMP3 (35) and MDM2 (36). In addition to miRNAs, lncRNAs could also be regulated by typical transcriptional factor (37, 38), mRNA binding protein (39), promoter methylation (40), and histone acetylation levels (41). Our data revealed that GCASPC was upregulated by the histone deacetylase inhibitor TSA, suggesting that GCASPC expression in gallbladder cancer is likely to be regulated by histone acetylation. The precise molecular mechanism of the downregulation of GCASPC in gallbladder cancer calls for further research.

In summary, we showed the detailed mechanistic insight of miR-17-3p-GCASPC-PC axis in gallbladder cancer. This finding suggests that GCASPC may be the important target for tumor therapy.

No potential conflicts of interest were disclosed.

Conception and design: M.-z. Ma, Y. Zhang, Z.-y. Hou, W. Gong, X. Kong, J.-d. Wang, Z.-w. Quan

Development of methodology: M.-z. Ma

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M.-z. Ma, Y. Zhang, M.-z. Weng, S.-h. Wang, Z.-y. Hou, Y.-y. Qin, W. Gong, Y.-J Zhang, J.-d. Wang

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M.-z. Ma, Y. Zhang, M.-z. Weng, Y.-y. Qin, J.-d. Wang

Writing, review, and/or revision of the manuscript: M.-z. Ma, Z.-y. Hou, X. Kong, Z.-w. Quan

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Hu, Y.-J Zhang, X. Kong, J.-d. Wang, Z.-w. Quan

Study supervision: S.-h. Wang, Z.-y. Hou, Z.-w. Quan

This work was supported by the National Nature Science Foundation of China (grant nos. 81272747, 81401932, and 30571824, for Z.-w. Quan), National Nature Science Foundation of China (grant nos. 81372642 and 30972919 to J.-d. Wang), and the Key University Science Research Project of Anhui Province (KJ2016A738 to Y. Zhang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.