miR-144 suppresses cell proliferation and migration in colorectal cancer by targeting NRAS

2.1 CRC patient tissues

CRC and adjacent noncancerous tissues were obtained from 13 patients who underwent surgical resection of CRC at Jinling Hospital (Nanjing, China). Both the CRC specimens and adjacent normal specimens were confirmed by pathologic diagnosis. All patients’ samples in this study were obtained from consentient individuals and the collectors of these specimens complied with the protocols authorized by the Ethics Committee of Nanjing University. Clinical specimens were instantly frozen in nitrogen canister after resection and conserved at−80°C. The clinicopathological data of 13 CRC patients is shown in Table 1.

2.2 Cell culture

Human CRC cell lines (SW480, LoVo, and Caco2) were ordered from the SIBCB (Shanghai, China).

2.3 RNA isolation and real-time quantitative reverse transcription PCR (qRT-PCR)

Total RNA from CRC cells or tissue samples was isolated by using Trizol reagent (Invitrogen, CA). The procedure and analysis of qRT-PCR experiments are following methods described in a previous study.7

2.4 Antibodies

Antibodies used to be probed for NRAS and GAPDH were ordered from Abcam (ab77392; Abcam, CA) and Santa Cruz Biotechnology (sc-25778; Santa Cruz, CA), respectively.

2.5 RNA oligos and NRAS vector

MiR-144 mimic, miR-144 inhibitor, and NRAS siRNA (including negative control RNAs) were ordered from GenePharma (Shanghai, China). NRAS overexpression vector was constructed from Genescript (Nanjing, China). RNA and proteins were isolated 48 hours after transfection.

2.6 Luciferase assay

pMIR-REPORT vector was constructed to carry a 570-bp fragment of NRAS 3′UTR with presumed miR-144 target site from Genescript. To confirm combining selectivity, NRAS 3′UTR sequences binding to the miR-144 seed region were changed from ATACTGT to TATGACA, and then the mutant sequences were inserted into identical same types of vector (Genescript). β-galactosidase (β-gal) expression plasmid (Ambion) was applied to verifying the efficiency of transfection.

2.7 Cell proliferation assay

A total of 96-well plates were utilized to culture SW480 cells with 2 × 10⁴ cell density each well. Cell counting kit-8 (CCK-8) reagent (Beyotime, Nanjing, China) were used to assay the cell proliferation index at different time points (every 12 hours, until an endpoint 60 hours) after transfection. The mean value of absorbance in triplicate was recorded as a final result at 450 nm wavelength.

A total of 48-well plates (Corning) were used to culture SW480 cells for EdU assay. The proliferation rate of CRC cells was estimated by the aid of an EdU assay kit (RiBoBio, Guangzhou, China) at 80% cells confluency after transfection. The protocol of the EdU assay kit was abided by but the nucleus staining materials were changed to obtain more distinct images. Specifically, the nucleus dye was DAPI (Beyotime) instead of Hoechst 33342 (provided by kit). Then, all stained cells were photographed by microphotography (BX51 Olympus, Japan).

2.8 Cell migration assay

The migration phenomenon of CRC cells was detected in 24-well millipore plates and the bottom of transwell chambers were covered with an 8-μm pore membrane (Millipore). The chambers were inverted in Petri dishes and 1% gelatin was added to the bottom of the membrane. The transfected CRC cells were collected after 24 hours and then resuspended in RPMI-1640 without FBS. The upper chamber was added with the resuspended cells approximately 2 × 10⁴ cells each well and the lower compartment was loaded with 500 μL RPMI-1640 (20% FBS). The whole plates were placed in a common humid incubator with 5% CO₂ at 37°C for 1 day. Then, migrant cells were fixed by 4% paraformaldehyde for 25 minutes (RT), followed by staining using 0.1% crystal violet for 20 minutes at RT.

2.9 Mice experiments

Tumor xenografts were established with male SCID mice about 4-week-old. At the beginning ofestablishment of the mice model, SW480 cells were treated with different vector tools to upregulate miR-144 and/or NRAS. Specifically, lentivirus infected alone SW480 cells to elevate miR-144 level, and NRAS vector was transfected alone the CRC cells to increase expression of NRAS. The rest of SW480 cells were cotransfected with the LV-miR-144 and the NRAS vector. Subsequently, SCID mice were subcutaneously injected with the above SW480 cells infected and/or transfected and each mouse was carried 3 × 10⁴ cells. Five mice were divided into one group. The mice were killed to separate grafted tumors from them after 30 days and the weights of neoplasm were measured immediately after resection. Then, parts of the tissues were applied to extracting purpose RNA and protein, and the rest of samples were fixed for making pathological slides with 4% paraformaldehyde for 1 day The fixed tissues were stained with hematoxylin and eosin (H&E) or used to perform immunohistochemical staining for NRAS and Ki-67.

2.10 Statistical analysis

Representative images of the Western blot and cell migration experiments were selected from the acquired images that were obtained from at least three independent experiments. Triplicate experiments were strictly conducted, including RT-qPCR, luciferase reporter assay, and proliferation assay. Numerical data were demonstrated as the mean ± standard error of mean. For statistical significance of differences analysis, the Student t-test was used. P < .05 was identified as differences had statistical significance.

3.1 CRC tissues exhibit upregulated NRAS protein level

First, we investigated NRAS expression levels in 13 paired CRC tissues. Significantly increased NRAS protein level in CRC tissues (CRC) contrasted with normal adjacent tissues was observed (Figure 1A and 1B). Then, we analyzed the levels of NRAS mRNA expression in the same CRC samples and normal adjacent samples and identified irregular changes in the NRAS mRNA expression level between the tumor tissues and the adjacent nontumor tissues (Figure 1C). Pearson's correlation scatter plot was used to analyze the correlation between the differences in the levels of NRAS protein and mRNA expression. As shown in Figure 1D, there was little correlation between the protein and mRNA expression. The inconsistency between the changes in NRAS protein and mRNA expression suggested that NRAS expression was controlled by a posttranscriptional mechanism.

3.2 miR-144 potentially targets NRAS

Because of miRNA is an essential posttranscriptional modulator of gene expression, we supposed some miRNAs regulate NRAS protein levels in CRC tissues. Online software Targetscan was utilized to predict possible miRNAs that regulate NRAS. miR-144 was identified as a potential candidate. The presumed combination between miR-144 and the NRAS 3′UTR is shown in Figure 2A. miR-144 seed region (2-8 bases of the mature miRNA) interacted with the target site in accordance with perfect base-pairing. Besides the miR-144 target sequence at the NRAS 3′UTR was highly conserved in diverse species (Figure 2A).

Next, we investigated whether there exists inverse correlation between miR-144 and NRAS levels in CRC samples. It was discovered that miR-144 expression levels were consistently lower in CRC samples compared with paired normal samples (Figure 2B). Furthermore, we identified the changes of the miR-144 expression levels were inversely correlated with those of the NRAS by using Pearson's correlation analysis (Figure 2C). Thus, NRAS was a candidate target of miR-144 in CRC.

3.3 miR-144 directly regulates the NRAS expression posttranscriptionally

NRAS expression levels were tested in CRC cell line SW480 after miR-144 alteration (increase or decrease). Figure 3A shows the efficient alteration of the miR-144 expression levels in SW480 cells by miR-144 mimic and inhibitor. As anticipated, NRAS expression dramatically decreased after miR-144 upregulation, whereas miR-144 inhibitor promoted the NRAS expression in SW480 cells (Figure 3B and 3C). Two other CRC cell lines (LoVo and Caco2) were used to strengthen the accuracy of these results (Figure 3A-C).

Next, to explore whether miR-144 prohibits NRAS expression through binding to the target site in the NRAS 3′UTR, we constructed a reporter plasmid to connect the NRAS 3′UTR sequence including the hypothesized miR-144 target site with the downstream location of the firefly luciferase gene. Subsequently, SW480 cells were transfected with the plasmid accompanied by the miR-144 mimic, miR-144 inhibitor, control mimic, or control inhibitor. As hypothesized, the upregulation of miR-144 expression led to nearly 70% decrease in luciferase signal contrasted to cells dealt with negative control RNAs, whereas miR-144 knockdown led to an approximate 60% increment in reporter activity compared with control cells (Figure 3D). The point mutations were subsequently induced into the miR-144 target region of the NRAS 3′UTR in a mutant plasmid. The plasmid was constructed to abrogate miR-144 target ability. The activity of mutant reporter was not affected by the upregulation or downregulation of the miR-144 expression (Figure 3D). These results indicated that target region is responsible for the interaction of miRNA-mRNA.

3.4 miR-144 suppresses CRC cell proliferation by targeting NRAS

It was presumed that miR-144 inhibits the CRC tumorigenic process by suppressing NRAS expression. We first analyze the impact of miR-144 on CRC cell proliferation by utilizing CCK-8 and EdU assays. The upregulation of the miR-144 expression decreased SW480 proliferation; conversely, the downregulation of the miR-144 expression had the opposite influence SW480 proliferation (Figure 4A-C).

An important issue was whether the effects of miR-144 on CRC cell proliferation originated from NRAS suppression by miR-144 since a single miRNA could have different targets.26 Therefore, we examined the functions of NRAS on SW480 cell proliferation. NRAS protein expression was downregulated and upregulated using siRNA (si-NRAS) and an overexpression plasmid (NRAS vector), respectively. It is shown that effective downregulation and upregulation of NRAS expression in SW480 cells (Figure S1A and S1B). Consequently, NRAS overexpression promoted the multiplication of SW480 cells, whereas downregulation of NRAS expression inhibited on cell multiplication (Figure S1C-E). Therefore, miR-144 and NRAS affect cell proliferation in CRC. We subsequently verified whether the upregulation of NRAS expression is enough to remedy the antiproliferative functions of miR-144 in CRC cells. Some of the SW480 cells were transfected with the miR-144 mimic or NRAS vector and the others were cotransfected with the miR-144 mimic and NRAS vector. As supposed, SW480 cells cotransfected with the miR-144 mimic and NRAS vector had markedly higher proliferation rates compared with those cells dealt with the miR-144 mimic alone (Figure 4F).

3.5 miR-144 suppresses CRC cell migration by targeting NRAS

We also checked the effect of miR-144 on CRC cell migration by performing the Transwell assay. The number of migratory cells was notably lower in SW480 cells after the miR-144 mimic transfection and higher in cells after the miR-144 inhibitor transfection inhibitor in contrast to control cells (Figure 5A and 5B). Then we validated that cell migration repression by miR-144 was due to the inhibition of NRAS. SW480 cells transfected with si-NRAS and the NRAS vector showed reduced and enhanced migration percentages, respectively (Figure S1F and S1G), suggesting that NRAS promotes SW480 cell migration. Then, a rescue experiment was performed in a similar manner as the proliferation experiment. SW480 cells cotransfected with the miR-144 mimic and NRAS vector exhibited a significantly higher migration rate than the control group (Figure 5C and 5D). Thus, the recovery of the ability of miR-144 to suppress cell migration by restoring NRAS demonstrated that miR-144 inhibits SW480 cell migration by targeting NRAS.

3.6 miR-144 inhibits mice CRC growth by targeting NRAS

A key problem to be solved was whether these effects of miR-144 and NRAS on CRC work in vivo. miR-144 lentivirus (LV-miR-144) and NRAS vector are used solely or combined in SW480. By subcutaneously implanting these SW480 cells into SCID mice, the CRC xenograft model was constructed. After sacrificing the mice, we observed a markedly decreased tumor weight in miR-144 overexpression group compared with the control group, whereas NRAS overexpression increased tumor weights (Figure 6A and 6B). Moreover, NRAS overexpression could abolish the growth-suppressing effects of miR-144 (Figure 6A and 6B), indicating that miR-144 inhibits tumor growth by silencing NRAS, which was collaborated by the miR-144 and NRAS levels in tumors (Figure 6C-E). H&E staining and immunohistochemical assays further demonstrated the antitumor growth role of miR-144 by suppressing NRAS (Figure 6F-H). These results were consistent with the in vitro results, which clearly validated the anti-CRC role of miR-144 by suppressing NRAS.