miR-375-3p suppresses tumorigenesis and partially reverses chemoresistance by targeting YAP1 and SP1 in colorectal cancer cells

MiR-375 expression is genetically downregulated in colorectal cancer and negatively correlated with chemoresistance, malignancy and poor prognosis

To demonstrate the association of miR-375 with the response of CRC cells to chemotherapeutics, we established two drug-resistant cell lines, HCT116/FU and HCT8/FU, by long-term culture of their corresponding parental cell lines, HCT116 and HCT8, in vitro with escalating 5FU concentrations. The resistance of parental and resistant cell lines to 5FU was examined by treating them with different concentrations of 5FU. As shown from the growth inhibition curves (Supplementary Figure 1A, 1B), the inhibitory rates of resistant cells were significantly decreased compared with their parental cells. The IC50 of 5FU in parental cells was 22.88 ± 0.14μg/ml and 25.59 ± 0.16 μg/ml, respectively, indicating more potency compared with that of resistant cells (146.14±15.06 μg/ml and 140.22±10.40 μg/ml (Supplementary Table 1). To further determine the relationship between miR-375 and chemoresistance, we first analyzed miR-375 expression in parental cell lines HCT116 and HCT8 and established corresponding 5FU-resistant sublines HCT116/FU and HCT8/FU. The results showed that miR-375 was significantly decreased in both of the 5FU-resistant cell lines (Figure 1A). Then, we analyzed miR-375 expression by qRT-PCR and found that miR-375 was lower to different degrees in CRC cell lines than in colonic mucosal epithelial cells (FHC) (Figure 1B). Moreover, clinical samples of patients who relapsed after 5FU-based chemotherapy (the 5FU-resistant group) were compared with those of patients who did not (the 5FU-sensitive group). The results showed that miR-375 expression was much lower in the 5FU-resistant group (n=30) than in the 5FU-sensitive group (n=30), indicating that miR-375 expression was associated negatively with chemoresistance in CRC tissues (Figure 1C). In addition, we compared the expression of miR-375 in 40 paired CRC and their adjacent normal tissues and found that CRC patients generally had downregulated miR-375 in CRC tissues (Figure 1D). Similar results were obtained in 450 CRC and 8 normal specimens downloaded from the Starbase database (Figure 1E). Afterwards, we divided clinical specimens into two groups based on the miR-375 expression value to explore its correlation with clinicopathological variables. A chi-square test showed that the miR-375 expression level was notably correlated with tumor size (p=0.038), tumor invasion depth (p=0.020) and TNM stage (p=0.002) (Table 1). Likewise, the TCGA cohort showed that lower expression of miR-375 was associated with advanced stage (n=198) (Figure 1F). We then investigated the relationship between miR-375 expression and CRC prognosis. Kaplan–Meier survival analysis revealed that patients with lower miR-375 levels had a much poorer overall survival than those with higher miR-375 levels (Figure 1G), which was also identified in the TCGA database (Figure 1H). We next performed univariate and multivariate survival analyses to test the correlation of miR-375 expression with overall survival in CRC patients. As shown in Table 2, univariate analysis validated three prognostic factors: miR-375 expression (p<0.001), tumor invasion depth (p = 0.034) and TNM stage (p<0.001). Multivariate survival analysis indicated that miR-375 was an independent prognostic factor (hazard ratio [HR] =2.54; 95% confidence interval [CI] = 1.18–4.33, P = 0.001) for CRC patients. These results suggest that miR-375 may play a critical role in the progression and drug resistance of CRC.

MiR-375 inhibits proliferation and chemoresistance and promotes 5FU-induced apoptosis of CRC cells in vitro

Parental cell lines (HCT116 and HCT8) and 5FU-resistant cell lines (HCT116/FU and HCT8/FU) were lipotransfected with miR-375 mimics and inhibitors (Supplementary Figure 1C, 1D). The cell viability experiment demonstrated that overexpression of miR-375 significantly reduced CRC-resistant cell viability, especially when combined with 5FU treatment. While miR-375 inhibitors remarkably increased CRC cell viability and markedly decreased 5FU sensitivity (Figure 2A, Supplementary Figure 2A). A colony formation assay further verified that miR-375 inhibited cell proliferation and promoted 5FU sensitivity, and miR-375 combined with 5FU treatment led to little cell growth in HCT116/FU and HCT8/FU cells. In contrast, miR-375 inhibitors notably increased cell proliferation and enhanced 5FU resistance in HCT116 and HCT8 cells (Figure 2B, 2C, Supplementary Figure 2B, 2C). MTT analysis showed that miR-375 significantly decreased the tolerance of 5FU-resistant cells to 5FU, whereas miR-375 inhibitors distinctly enhanced 5FU resistance in CRC parental cells (Figure 2D, Supplementary Figure 2D). Dramatically, the same observation was found when miR-375 mimics or inhibitors were combined with various anticancer drugs, including capecitabine, oxaliplatin and irinotecan (Figure 2E, Supplementary Figure 2E). The weak drug tolerance level extended to other CRC cells transfected with miR-375, such as DLD1, HT29 and SW620 (Figure 2F). In addition, in CRC parental and resistant cells with 5FU regimens, miR-375 significantly increased the cell apoptosis rate, whereas miR-375 inhibitors decreased the cell apoptosis rate (Figure 2G, Supplementary Figure 2F).

MiR-375 suppresses tumorigenesis and reverses chemoresistance to 5FU of CRC in vivo

To determine the relationship between miR-375 and therapeutic effects, we used transfected HCT116 cells to transplant subcutaneously into nude mice, followed by subsequent intratumor injection of antagomiR-375 or antagomiR-NC after 7 days (Figure 3A). miR-375 expression was detected in tumors from two groups of nude mice using qRT-PCR. As expected, the miR-375 expression level was notably reduced in tumors from the antagomiR-375 group compared with that in the antagomiR-NC group (Figure 3B). In HCT116-Xenograft, tumor growth was significantly inhibited in the antagomiR-375 group compared with the antagomiR-NC group (Figure 3C). Immunohistochemical (IHC) analysis of tumor tissues collected from the antagomiR-375 group exhibited higher Ki67-positive compared with the control group, which implied that cell proliferation was increased when inhibition of miR-375 (Figure 3D). On the other hand, we conducted agomiR-375 and a relatively low dose of 5FU to explore whether the treatment could result in a synergistic effect in vivo. Transfected HCT116/FU cells were inoculated subcutaneously into nude mice to form tumors. All mice were divided into 4 groups with intratumor injections of agomiR-375/NC or 5FU every 3 days (Figure 3E). Consequently, the miR-375 expression level increased by approximately 25-fold compared with the agomiR-NC group (Figure 3F). tumors from the agomiR-375 group were extremely smaller than those from the agomiR-NC group. In addition, tumors combined with agomiR-375 and 5FU treatments presented greater anti-tumor effects than either agomiR-375 or 5FU alone. (Figure 3G). In addition, Ki67 staining showed that Ki67-positive cells were significantly reduced by the combinational therapy (Figure 3H). These results indicated that miR-375 can suppress tumor growth and enhance the chemosensitivity of 5FU in vivo.

YAP1 and SP1 are inversely associated with miR-375 and chemosensitivity

To identify miR-375 target genes involved in miR-375-mediated suppression of cell proliferation and chemoresistance, three different online miRNA target prediction bioinformatics databases (TargetScan7, PicTar and microT-CDS) were initially applied. Then, 28 potential genes of miR-375 were predicted through integrated analysis of three databases (Figure 4A). Finally, 13 potential candidate target genes that were associated with tumorigenesis and development were screened, and qRT-PCR was carried out later. As shown in Figure 4B and Supplementary Figure 3A, the mRNA expression levels of four genes (YAP1, SP1, PRKD1 and HAS2) among 13 candidate genes were significantly downregulated by miR-375 in HCT116/FU and HCT8/FU cells transfected with miR-375 mimics. Thereafter, we analyzed the mRNA levels of the four genes in CRC clinical specimens and their matched adjacent normal tissues (n=40, respectively) and found that only YAP1 and SP1 were significantly upregulated in CRC tissues (Figure 4C, 4D, Supplementary Figure 3B, 3C). Intriguingly, YAP1 and SP1 expression levels were strikingly higher in 5FU-resistant patients than in 5FU-sensitive patients (n=30, respectively), which revealed a reverse relationship between YAP1/SP1 expression and 5FU sensitivity (n=40, respectively) (Figure 4E, 4F). Moreover, a significantly negative correlation between YAP1 and SP1 expression and miR-375 was found in CRC tissues (Figure 4G, 4H).

MiR-375 directly targets YAP1 and SP1

To verify whether YAP1 and SP1 were direct targets of miR-375, we conducted a dual-luciferase assay. The putative binding sites of miR-375 on the YAP1 and SP1 3’UTR are shown in Figure 5A. After wild-type and mutant-type YAP1 and SP1 3’-UTR sequences were cloned into the luciferase reporter vector, respectively, we found that miR-375 significantly reduced the luciferase activities of the wild-type reporters of YAP1/SP1 3’-UTR both in HCT116/FU and HCT8/FU cells, whereas no obvious decrease was observed in binding the mutant-type reporters of YAP1/SP1 3′-UTR (Figure 5B). The results strongly demonstrated that YAP1 and SP1 were the direct targets of miR-375 in CRC. In particular, overexpression of miR-375 significantly decreased the mRNA expression of YAP1 and SP1, while inhibition of miR-375 remarkably elevated the mRNA expression of YAP1 and SP1 in both parental and 5FU-resistant cell lines (Figure 5C, 5D). Moreover, the total protein expression levels of YAP1 and SP1 were both observably higher in 5FU-resistant cells when compared with their corresponding parental cells (Figure 5E), which further signified that YAP1 and SP1 had a reverse association with miR-375 and 5FU sensitivity. Meanwhile, the total protein expression of YAP1 and SP1 was significantly attenuated by miR-375 in 5FU-resistant cells, whereas notably enhanced by miR-375 inhibitors in parental cells. These effects were more obvious when treated with 5FU. (Figure 5F). Additionally, we subcutaneously injected HCT116/FU and HCT116 cells into BALB/c nude mice and analyzed the levels of YAP1 and SP1 in tumors by IHC assays (Figure 5G, Supplementary Figure 3D), which also confirmed the relationship between YAP1/SP1 and miR-375, drug resistance. These results undeniably indicated that miR-375 negatively regulated YAP1 and SP1 by directly binding YAP1 and SP1 in CRC.

MiR-375 suppresses cell proliferation and 5FU resistance of CRC cells by repressing YAP1 and SP1

To determine the relationship between YAP1/SP1 expression and cell proliferation or resistance to 5FU in CRC cells, we knocked down YAP1/SP1 by siYAP1/siSP1 or overexpressed them using special plasmids (Supplementary Figure 3E, 3F). The MTT assay demonstrated that upregulation of YAP1/SP1 significantly induced proliferation and enhanced 5FU resistance in CRC parental cells, whereas downregulation of YAP1/SP1 relatively reduced cell proliferation and drug resistance in CRC 5FU-resistant cells (Figure 6A, 6B). Moreover, when miR-375 inhibitors were cotransfected with siYAP1/siSP1, a cell viability assay indicated that silencing YAP1/SP1 substantially reversed the resistance of miR-375 inhibitors to 5FU with or without drug administration (Figure 6C). In contrast, when miR-375 was cotransfected with YAP1/SP1 into the 5FU-resistant cells, cell viability was obviously decreased by miR-375 with or without 5FU-induced inhibition; nevertheless, the inhibitory effect on cell viability was partially abolished by YAP1/SP1 overexpression (Figure 6D). Moreover, flow cytometry analysis showed that the reduction of 5FU-induced apoptosis by miR-375 inhibitors was entirely reversed by siYAP1 or siSP1 in HCT116/FU cells (Figure 6E), whereas the escalation of 5FU-induced apoptosis by miR-375 was practically undermined by YAP1 or SP1 overexpression in HCT8/FU cells (Figure 6F). Taken together, miR-375 attenuated CRC cell resistance to 5FU by repressing YAP1 and SP1.

MiR-375 downregulates YAP1 expression via activating the Hippo-YAP1 signaling pathway

As mentioned in Figure 2G, the ectopic expression of miR-375 could significantly promote cell apoptosis in chemoresistant cells, especially when combined with 5FU treatment, and the remedy assay on proliferation and apoptosis implied that YAP1 can reverse the effect of miR-375 in CRC-resistant cells (Figure 6D, 6F). Thus, the potential resistance mechanism of miR-375 associated with YAP1 in CRC cells was further explored in our study.

Yes-associated protein 1 (YAP1), also known as YAP, is a component of the transcriptional co-activator YAP/TAZ downstream effector inducing target gene expression in the Hippo pathway [29]. According to previous studies [29–31], we selected YAP1 as well as its target downstream proteins related to cell growth, apoptosis and resistance involved in the Hippo pathway, including CTGF, cyclin D1 and BIRC5, and then detected them by western blotting. By overexpressing miR-375 in HCT116/FU cells, we found that total YAP1 expression was significantly downregulated and that its downstream genes CTGF, cyclin D1 and BIRC5 also declined. Surprisingly, compared to nontreatment with 5FU, these protein levels slightly decreased when combined with 5FU cytotoxicity (Figure 7A), which might imply that 5FU inhibited the expression of downstream genes of the Hippo pathway. Similar reductions in these genes were observed in HCT116/FU cells transfected with siYAP1 or siSP1 (Figure 7A). In contrast, miR-375 inhibitors obviously upregulated YAP1 and its downstream targets CTGF, cyclin D1 and BIRC5 in HCT116 cells (Figure 7B). Moreover, miR-375-induced lower levels of protein expression in HCT116/FU cells were observably reversed by YAP1 overexpression (Figure 7C), whereas miR-375 inhibitor-induced higher levels of protein expression in HCT116 cells were moderately counteracted by YAP1 knockdown (Figure 7D).

Meanwhile, apoptosis-related proteins, including cleaved caspase 3 and cleaved PARP1, were increased by miR-375, whereas the expression level of the anti-apoptosis protein Bcl-2 was markedly inhibited in two resistant cells (Figure 7E, Supplementary Figure 4A), similar to the effect of silencing YAP1 or SP1 (Figure 7F, Supplementary Figure 4B). In contrast, miR-375 inhibitors could significantly downregulate the expression of cleaved caspase 3, cleaved PARP1 and elevate the level of Bcl-2 expression in two parental cells (Figure 7G, Supplementary Figure 4C), and a similar result was observed in HCT116 and HCT8 cells overexpressing YAP1 or SP1 vectors (Figure 7H, Supplementary Figure 4D). In addition, IHC revealed that the level of cleaved caspase 3 was significantly increased in xenografts formed from miR-375 overexpression with or without 5FU treatment, while it was decreased from miR-375 inhibition (Supplementary Figure 4E). In summary, miR-375 promoted apoptosis by inhibiting YAP1/SP1 expression. To speak of miR-375-mediated YAP1, miR-375 suppressed YAP1, thereby decreasing the expression of Hippo-YAP1 downstream target genes in CRC cells, which ultimately led to the promotion of apoptosis-related protein expression.

Dual-luciferase reporter assay

Based on the wild-type and mutant 3’UTRs of target genes (YAP1/ SP1) binding to miR-375-3p (Figure 5A), the wt and mut vectors were constructed and subcloned into the pmirGLO-basic luciferase reporter (GeneCreat, China). In 1×10⁴ HCT116/FU or HCT8/FU cells per well in a 96-well plate cultured for 12 h, 50 ng of the wt or mut vector was cotransfected with either miR-375-3p mimics or NC mimics (50 nM) into the two cells using Lipofectamine 2000 according to the manufacturer’s protocol (Invitrogen, CA, USA). After 48 h of transfection, cells were harvested, and luciferase activity was measured by the dual-luciferase reporter assay system (Promega, WI, USA). Renilla luciferase activity was used as an internal reference. Each transfectant was performed in triplicate.

Cell viability assay and colony formation

The short-term effects of anticancer drugs on cell growth were measured by a cell viability assay. First, 1.5×10⁴ cells/ml were harvested and seeded into 96-well plates for 12 h of cultivation. Then, drug regimens were added into designated columns. After 72 h, cell viability assays were detected by MTT assays. “Relative cell viability” = the viability of cells in drug-containing medium/the viability of cells in drug-free medium. “Relative cell viability” was further fitted to a dose-response curve to estimate the IC50 by GraphPad Prism 7 software. The long-term effects of anticancer drugs on cell growth were assessed with a colony formation assay. After 36 h of transfection, a density of 200 cells/well was plated into 6-well plates and incubated for 48 h. Afterwards, different concentrations of 5FU (10 μg/ml and 50 μg/ml) were added to the designated wells. Cells were cultured for approximately 14 days. The colonies were fixed with methanol for 5 minutes and stained with 0.1% crystal violet for 15 minutes at room temperature. Finally, colonies were photographed and counted using ImageJ k 1.45.

Analysis of apoptosis

For apoptosis analysis, cells were seeded into 6-well plates (1.5×10⁵ cells/well) for 12 h incubation. MiR-375-3p mimics and NC mimics were transfected into HCT116/FU and HCT8/FU cells, while miR-375-3p inhibitors and NC inhibitors were transfected into HCT116 and HCT8 cells. After 12 h of transfection, each group was treated with PBS and different concentrations of 5FU (inhibitors: 10 μg/ml, 25 μg/ml; mimics: 150 μg/ml, 200 μg/ml) for 48 h. In addition, HCT116/FU cells were cotransfected with miR-375-3p inhibitors and YAP1 or SP1 small interfering RNAs (siYAP1 or siSP1, respectively) (miR-375-3p inhibitors 100 nM; siYAP1/siSP1 50 nM). Meanwhile, HCT8/FU cells were cotransfected with miR-375-3p and YAP1- or SP1-expressing plasmids (miR-375-3p 50 nM; YAP1/SP1 50 nM). After 12 h of transfection, the cell culture medium was changed, and the cells were cultured for 48 h. Cell apoptosis was detected with the Annexin V-FITC/Propidium Iodide (PI) Apoptosis Detection Kit (BD, Biosciences, CA, USA) and analyzed by fluorescence-activated cell sorting using a FACScan (BD Biosciences, San Jose, CA, USA).

In vivo tumorigenicity study

All animal experiments were approved by the Animal Care Committee of Nanjing First Hospital, Nanjing Medial University (acceptance No. SYXK 20160006). For HCT116 xenograft animal models, 4-week-old male BALB/c nude mice were randomly divided into two groups (n=5 per group). HCT116 cells transfected with antagomiR-375-3p or antagomiR-NC (a total of 5 × 10⁶ cells in 0.2 ml PBS) were subcutaneously injected into the right flank region of each nude mouse. After 8 days of tumor formation, 2 nmol antagomiR-375-3p/NC was injected intratumorally every 4 days for seven cycles. For HCT116/FU xenograft animal models, HCT116/FU cells transfected with agomiR-375-3p or agomiR-NC (a total of 5 × 10⁶ cells in 0.2 ml PBS) were subcutaneously injected into the right flank region of each nude mouse. Five days later, all types of xenografts were randomly divided into four groups and treated with (I) agomiR-NC+PBS (10 mg/kg); (II) agomiR-375-3p (2 nmol per mouse)+PBS, (III) agomiR-NC+5FU (10 mg/kg), or (IV) agomiR-375-3p+5FU. AgomiR-375-3p/NC and 5FU were administered via intratumoral injection. The corresponding treatment regimens were repeated every 3 days for eight cycles. Xenografts were measured every 3 days with digital calipers, and tumor volumes were calculated using the formula V =1/2 (L × W²). The curve of tumor growth was made based on tumor volume and corresponding time (days) after treatment. Tumor weights were measured using an electronic scale after mice were sacrificed. AgomiR-375-3p, antagomiR-375-3p and their negative controls agomiR-NC and antagomiR-NC were obtained from GenePharma (Shanghai, China).