LncRNAs H19 and HULC, activated by oxidative stress, promote cell migration and invasion in cholangiocarcinoma through a ceRNA manner

QBC939 human cholangiocarcinoma cells were obtained from Shuguang Wang (The Third Military Medical University, China). SK-cha-1 human cholangiocarcinoma cells were kindly provided by Dr. Chundong Yu (Xiamen University, China). RBE human cholangiocarcinoma cells and HEK293T human embryonic kidney cells were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). All CCA cell lines were cultured in RPMI-1640 medium with 10 % fetal calf serum. HEK-293 T cells were cultured in DMEM medium with 10 % fetal calf serum. All cells were maintained in a 37 °C humidified incubator with 5 % CO₂.

Cells were collected in EP tubes. Then, total RNA was extracted from tissue or cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instructions. RNA was reverse-transcribed into cDNA with the ReverTra Ace® qPCR RT Kit (TOYOBO, Japan) or PrimeScript® RT reagent Kit with gDNA Eraser (Takara, Japan). Real-time PCR for lncRNA was performed using the SYBR Premix ExTaq real-time PCR kit (Takara, Japan) according to the manufacturer’s instructions with GAPDH as normalization control. The expression level for each lncRNA and mRNA was determined using the 2^−△△Ct method. The specificity and reliability of the PCR were confirmed by sequencing the PCR product fragments. All primers are shown in Additional file 1: Table S1.

CCA cells were treated with 1 μM H₂O₂ for 1 h and 1 μM glucose oxidase for 24 h for short-term and long-term oxide incubation in vitro, respectively.

Full-length H19-pcDNA 3.1 vectors were purchased from Integrated Biotech Solutions (Shanghai, China). The pcDNA 3.1 vector was used for lncRNA and mRNA overexpression, and the psiCHECK-2 vector was used for the luciferase target assay. Primers and oligonucleotides for full-length HULC, IL-6, and CXCR4 amplification and 3′-UTR segments of IL-6 and CXCR4 construction are listed in Additional file 1: Table S1.

We plated 1.6 × 10⁴ RBE cells or 5 × 10⁴ HEK-293 T cells in 48-well plates. Then, 500 ng psiCHECK-2-derived reporter vectors (Promega, Madison, WI, USA), 500 ng pcDNA3.1 vectors overexpressing lncRNAs (Invitrogen Corporation, Carlsbad, CA, USA), and 30 nM miRNA mimics (GenePharma, Shanghai, China) were co-transfected with Lipofectamine® LTX (Invitrogen Corporation, Carlsbad, CA, USA) and Lipofectamine® 3000 (Invitrogen Corporation, Carlsbad, CA, USA). Finally, 48 h after transfection, cells were harvested for the dual luciferase reporter assay (Promega, Madison, WI, USA). Each experiment was repeated at least three times.

As previously described, 72 h after transfection, the RBE cells were harvested, and total protein was extracted from the cells using RIPA lysis buffer (Beyotime Biotechnology, Shanghai, China) with 1× complete ULTRA (Roche, Nutley, USA). The proteins were detected with antibodies against CXCR4 (ab2074, Abcam, USA), IL-6 (SC-7920, Stantacruz, USA), GAPDH (Proteintech Technology, Manchester, UK), and β-tubulin (32−2600, Invitrogen, USA). The protein levels were normalized to GAPDH or β-tubulin.

In the assay, the enrichment of miRNAs binding with AGO2, a key component of the microRNA-containing RISC complex, were used as the positive controls, and U6 as the negative. The AGO2 (FB261-S2, Abnova, Taiwan) monoclonal primary antibody was rip grade, and the procedure of rip experiment was followed to the manual of Magan rip Kit (17−700, Millipore, USA).

Transfected RBE cells were cultured in 24-well plates for 24 h in standard conditions until 80−100 % confluency was reached. Linear wound tracks were generated with sterile, 10-μl pipettes and maintained under standard conditions. The scratched cells were rinsed twice with PBS to remove non-adherent cells, and fresh RPMI-1640 medium (with no fetal calf serum) was added. Photographs of the centers of the gaps were taken using a 10× phase-contrast microscope (Zeiss). Cell migration at 0 and 24 h after scratching was evaluated by determining the wound distance at two random wound gap locations. Three independent scratch wound experiments were used for calculations.

Migration and invasion assays were conducted using Transwell chambers (8 μm, Corning Costar Co., Cambridge, USA) according to the manufacturer’s instructions. For the migration assays, 8 × 10⁴ RBE cells transfected after 24 h were plated in the top chambers lined with a non-coated membrane. For the invasion assay, the chamber inserts were coated with a 1:9 deliquation of Cultrex® Basement Membrane Extract (Trevigen, USA). Then, 1.2 × 10⁵ cells were plated in the top chamber, while 600 μl RPMI-1640 with 10 % fetal bovine serum was added to the lower chamber as a chemoattractant. After a 24-h incubation at 37 °C, cells located in the lower chamber were fixed with 100 % methanol and stained with 0.1 % crystal violet. Cells were counted in ten fields for triplicate membranes at 10× magnification using a microscope (Zeiss). Five random sights in each sample were selected to analyze cell count, and the mean of triplicate experiments was calculated.

In human bile duct cells, inflammatory responses were usually triggered by oxidative stress, which was induced by HBV, HCV, or excretory-secretory proteins released by C. sinensis or O. riverine [6, 8]. We speculated that the lncRNAs stimulated by oxidative stress may trigger inflammation in CCA. To identify whether any lncRNA related with inflammation is regulated by oxidative stress, we firstly reanalyzed several GEO data, GSE76743 [29], GSE57539 [30], GSE67106 [31], GSE55146 [32], and GSE70544 [33], and identified 40 lncRNAs responding to oxidative stress or inflammatory response (Fig. 1a). Among these, XIST, CDKN2B-AS1, CRNDE, TUG1, SOX2-OT, etc. have been reported playing crucial roles in inflammation pathway of cancer progression. Importantly, two lncRNAs, HULC and H19, showed higher expressed among all of the five selected datasets, suggesting that they may involve in inflammation pathway that is regulated by oxidative stress. To validate the expression patterns of HULC and H19 under oxidative stress, we then used three human CCA cell lines, QBC939, SK-cha-1, and RBE, for short-term or long-term oxide incubation using H₂O₂ and glucose as described in previous studies [34] and measured the expression levels of HULC and H19 by qPCR. The results showed that both lncRNAs were significantly up-regulated in three cell lines (Fig. 1b), suggesting a potential function in acute oxidative stress responses. Similarly, HULC and H19 were also significantly up-regulated in at least two cell lines after long-term oxidative stress induced by glucose oxidase (Fig. 1c), and these two lncRNAs might be more likely to contribute to the pathogenesis of CCA through chronic inflammation caused by long-term oxidative stress. It is worth noting that HULC and H19 were up-regulated after both short- and long-term oxidative stress, implying their pivotal roles in inflammation promotion and CCA pathogenesis. Additionally, we have also investigated the heme oxygenase-1 (HO-1) gene expression, which were reported to have the cytoprotective effect under various pathophysiological stimulating conditions, such as oxidative stress [35]. As a result, the expression of HO-1 had increased significantly under the oxidative stress condition (Additional file 1: Figure S1), suggesting that the HO-1 may also protect cholangiocarcinoma from oxidative stress [36, 37].

Combined with previous observations, we speculated that H19 and HULC are stimulated by oxidative stress and regulate inflammation pathways involved in CCA pathogenesis. To validate this speculation, we investigated the function of H19 and HULC in CCA cells. Several inflammation pathways are involved in CCA by modifying migration and invasion; therefore, we used scratch wound healing assays to further demonstrate the function of H19 and HULC in migration potency. After 24 h in culture, CCA cells with forced expression of H19 or HULC (Additional file 1: Figure S2A) dramatically promoted wound closure compared to the control (Fig. 2a), suggesting that H19 and HULC have migration regulating roles. We also performed migration assays and invasion assays to validate the migration and invasion abilities of H19 and HULC. Consistence with the scratch wound healing assays, H19 and HULC significantly expedited migration (2.0-fold for H19 and 3.3-fold for HULC, Fig. 2b) and invasion (1.77-fold for H19 and 1.64-fold for HULC, Fig. 2c), which further strengthened the migration and invasion roles of H19 and HULC in CCA. Furthermore, we also investigated the migration and invasion abilities of RBE cells when knocking down H19 or HULC. The results showed that the migration (Additional file 1: Figure S3A) and invasion (Additional file 1: Figure S3B) abilities of RBE cells were decreased when H19 or HULC was knocked down (Additional file 1: Figure S2B). These data imply that H19 and HULC may regulate the inflammation factors regulating migration and invasion after stimulating oxidative stress.

We further investigated the potential target genes of H19 and HULC and the underlying regulatory mechanisms. Previous studies showed that both H19 and HULC are similar to ceRNAs [38] in human embryonic kidney cells and human embryonal carcinoma cells by sponging let-7 and miR-372, respectively [27, 39]. To determine whether H19 and HULC have similar roles in CCA, we focused on the potential functions of the let-7 and miR-372 families in CCA cells (Additional file 1: Figure S4A). We used bio-informational predicting and found the let-7 might target IL-6, one of the most important inflammatory factors in the pathogenesis, growth and migration of CCA [40], and miR-372 and miR-373 might target CXCR4, a chemokine receptor involved in CCA progression and metastasis [41] (Fig. 3a). Subsequently, a dual luciferase reporter system (Additional file 1: Figure S4B) and a western bolt assay (Fig. 3b) in RBE CCA cells had identified the direct regulation of IL-6 and CXCR4 by let-7a/let-7b and miR-372/miR-373, respectively. We also performed miRNA down-regulation experiment by transfecting the miRNA inhibitors into RBE cells under oxidative stress and found that the expression levels of IL-6 and CXCR4 increased significantly (Fig. 3c, d, Additional file 1: Figure S5). These results revealed that the activation of these mRNAs over oxidative stress is miRNA-dependent.

To further confirm whether H19 and HULC regulate inflammation and CCA pathogenesis through ceRNA patterns, we co-transfected miRNAs, 3′-UTR, and lncRNAs in HEK-293 T and RBE cells. Compared with empty vectors, we found that overexpressed lncRNAs rescued the luciferase activity suppressed by miRNAs (Fig. 3e in RBE cells and 3f in HEK-293 T cells), suggesting that H19 up-regulates IL-6 by suppressing let-7a and let-7b, and HULC up-regulates CXCR4 by suppressing miR-372 and miR-373. Moreover, we further performed a RNA immunoprecipitation (RIP) experiment to investigate the direct binding of H19, HULC to their certain miRNAs. In the assay, the enrichment of miRNAs binding with AGO2, a key component of the microRNA-containing RISC complex, were used as the positive controls and U6 as the negative [42]. The RIP results revealed that H19 and HULC were associated with the AGO2 protein in CAA cells (Fig. 4a, b). Moreover, we detected the AGO2-immunoprecipitated lncRNAs with or without miRNAs that knocked down by inhibitors. The results showed that the enrichment of AGO2-immunoprecipitated H19 and HULC was decreased in the let-7a/7b or miR-372/373 inhibitors transfected into RBE cells (Fig. 4c), indicating that H19 and HULC are directly binding to the certain miRNAs, let-7a/7b, miR-372/373, and function as ceRNA manner in CAA cells. The protein levels were also measured and overexpressed H19 or HULC in RBE cells increased the abundance of IL-6 and CXCR4 proteins, respectively (Fig. 4d, e). Additionally, we further testified that both of their target genes IL-6 and CXCR4 are down-regulated when H19 and HULC are knocked down, respectively, in RBE cells under the oxidative stress condition (Fig. 4f, g). Functional validation also indicated that forced expression of IL-6 and CXCR4 plays a similar role to H19 and HULC in promoting cell migration and invasion in CCA cells (Additional file 1: Figure S6 and S7). These results suggested that H19 and HULC regulate cell migration and invasion by increasing IL-6 and CXCR4 levels through sponging let-7a/let-7b and miR-372/miR-373, respectively.

In conclusion, our results suggest that two inflammation-associated lncRNAs in CCA, H19 and HULC, target inflammatory genes, including IL-6 and CXCR4, using ceRNA methods and primarily regulated cell migration and invasion (Fig. 5). These lncRNAs might serve as key regulators in CCA.