Ectopic expression of miR‐944 impairs colorectal cancer cell proliferation and invasion by targeting GATA binding protein 6

Abstract miR‐944 is a microRNA that has been reported to play different important roles in the progression of cancer. Colorectal cancer (CRC) is a common cancer worldwide. A recent study has confirmed that miR‐944 plays a tumour suppressive role in CRC. However, biological functions and the mechanism of miR‐944 in CRC are poorly understood. Real‐time reverse transcription polymerase chain reaction of 100 CRC tissues showed that miR‐944 expression is frequently downregulated and is negatively associated with the T is the primary tumor, N is the lymph node, and M is the distant metastasis (TNM) stage (P = 0.009), depth of invasion (P = 0.001), and lymph node status (P = 0.002). Overexpression of mir‐944 significantly impaired the functions of proliferation, migration and invasion in CRC cells, while these functions increased in knockdown experiments. GATA binding protein 6 (GATA6) knockdown can reverse the CRC cells functions induced by miR‐944 inhibitor. Mechanistically, a Dual‐Luciferase Reporter Assay showed that miR‐944 is structurally combined with GATA6 and interacts with downstream proteins (CRT and p‐AKT) in CRC cells. In conclusion, these findings indicated that miR‐944 may be a tumour suppressor and could likely be used as a prognostic predictor and novel therapeutic target for CRC.

roles in human malignancies. In breast cancer, Ali Flores-Pérez et al 11 found that miR-944 acts as an inhibitor by targeting SIAH1 and PTP4A1. However, in cervical cancer studies, miR-944 plays an oncogenic role by targeting HWCW2 and S100PBP. 12 A recent study showed that the expression level of miR-944 in tumour tissues is less than that in adjacent non-tumour tissues in CRC. 13 Another study showed that the expression level of miR-944 was frequently low in recurrent CRC patients. 14 Dysregulation of GATA transcription factor 6 (GATA6) has been reported to affect the initiation and progression of tumours. In breast cancer, GATA6 can stimulate EMT by inhibiting the E-cadherin protein. 15 Strong expression of GATA6 is associated with liver metastasis and poor patient survival in CRC. 16 However, the role of miR-944 and its correlation with GATA6 in CRC have not been reported clearly. Therefore, we evaluated the expression and potential mechanism of signalling pathways in CRC. with 10% fetal bovine serum (FBS; Hyclone, Logan, UT)and 100 U/mL penicillin-streptomycin, and human 293 cells were cultured in DMEM (BI) with 10% FBS and 100 U/mL penicillin-streptomycin. The cells were all cultured in a humidified incubator with 5% CO 2 at 37°C. for 45 seconds. Here, U6 snRNA was used as an internal control.

| RNA isolation and quantitative real-time PCR
Fold changes (2 −ΔΔCt ) were used to analyse the relative expression of miR-944. Each experiment was replicated three times.

| Western blot analysis and coimmunoprecipitation
Total protein was extracted using Radio-Immunoprecipitation Assay lysis buffer containing 1% phenylindole sulfonylfluoride (PMSF) from two infected CRC cell lines (HCT116 and SW480). Harvested cell proteins were added to a 10% sodium dodecyl sulfate-polyacrylamide gel and transferred onto polyvinylidene fluoride membranes (Millipore, Bedford, MA). The membranes were incubated with rabbit anti-GATA6 (Abcam, Cambridge, UK), mouse anti-CRT (Abcam), and mouse anti-p-AKT (Proteintech, Chicago, IL) primary antibodies at a dilution of 1:1000 and with a mouse anti-GAPDH (Proteintech) primary antibody at a dilution of 1:3000 overnight at 4°C. The secondary antibodies were incubated with the membranes for 2 hours on the following day. The protein bands were analysed using an ECL detection kit (Thermo Fisher Scientific, Rockford, IL).
Co-immunoprecipitation was performed with SW480 and HCT-116 cell lines. After 48 hours, we collected the cells and added RIPA lysis buffer (composed of 20 mmol/L Tris/HCl (pH 7.4), 1.0% NP-40, 150 mmol/L NaCl, 1 mmol/L ethylenediaminetetraacetic acid, 10 μg/mL leupeptin, and 50 μg/mL PMSF). Supernatants containing proteins were then collected and incubated with anti-AKT (Proteintech) antibodies for 1 hour, followed by incubation with magnetic beads at 4°C for 18 hours. The samples were then subjected to SDS-PAGE for separation and detection. Co-precipitation was used to detect the CRT (Abcam) and AKT (Proteintech) antibodies, and endogenous coprecipitation was used to detect CRT protein bands. The experiment was repeated three times.

| Immunohistochemistry
Colorectal cancer tissues embedded in paraffin were cut into 4-μmthick continuous sections. After being deparaffinized at 65°C for 2 hours, antigen repair was performed at a high pressure for 2.5 minutes, and then slides were cooled at room temperature for 1 hour.
The sections were incubated with H 2 O 2 (3%) for 30 minutes to block endogenous peroxidase activity. Ten percent normal goat serum was added for 30 minutes to reduce nonspecific binding, and the sections were incubated with rabbit monoclonal anti-GATA6 (Abcam) overnight at 4°C. On the following day, after adding a secondary antibody for 20 minutes, the sections were incubated with streptavidin-peroxidase reagent for 20 minutes. 3,3′-Diaminobenzidine (DAB) was added for signal reactions. The stained sections were imaged at ×20 magnification. Immunohistochemistry (IHC) scoring was performed as described previously 17 . The staining range score was defined as 1 (0%-25%), 2 (25%-50%), 3 (50%-75%), and 4 (75%-100%) according to the positively colored range of the slides. Staining intensity scores were defined as 0 (negative), 1 (weak), 2 (moderate) and 3 F I G U R E 1 miR-944 expression is frequently downregulated in human colorectal cancer (CRC). A, The relative expression of miR-944 in 100 pairs of CRC tissues. miR-944 expression was analysed by quantitative real time polymerase chain reaction (qRT-PCR) (normalized to U6 snRNA). Data are shown as -ΔΔCT (tumour/normal). B, The relative expression of miR-944 in four CRC cell lines and CCC-HIE-2 cells. Data are shown as 2 -ΔΔct , n = 3, *P < 0.05 and **P < 0.01 by t test compared with CCC-HIE-2 cells The mean of the relative miR-944 expression is showed as fold change and 95% confidence interval (CI). CRC, colorectal cancer; miR-944, microRNA-944. *P < 0.05, **P < 0.01.

| Transfection
An miR-944 mimic and its corresponding negative control (referred as mimics-NC) and an miR-944 inhibitor and its corresponding NC (referred as inhibitor-NC) were all synthesized at GenePharma. GATA6 siRNA (referred to as GATA6-siRNA) and control siRNA (referred to as siRNA-control) were synthesized at HANBIO (Shanghai, China).
Transfection was performed using Lipofectamine 3000 (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions.

| Cell migration and invasion assay
HCT-116 and SW480 cells were infected as described previously.
The experiment was performed using 24-well Transwell plates and

| Dual-luciferase reporter assay
For the luciferase assays, the 3′-UTR of the GATA6 mRNA and mutant Each experiment was performed three times independently. was analysed by Spearman's correlation test. The results were considered to be statistically significant at P < 0.05 in all experiments.  Figure 1B). Furthermore, we analysed the correlation between clinical pathology and the expression of miR-944. We found that miR-944 expression was negatively associated with the TNM stage (P = 0.009), depth of invasion (P = 0.001), and lymph node status (P = 0.002). However, there was no significant correlation with age, gender, tumour size, the degree of tumour differentiation and metastasis (Table 1).

| miR-944 inhibits CRC cell proliferation, migration, and invasion
The

| miR-944 targets the GATA6 gene in CRC cells
To explore the potential mechanism by which miR-944 affects CRC cell function, we used two databases that predict the miRNA targets (microRNA and TargetScan) and found that GATA6 may be a potential target. We transfected miR-944 mimics into SW480 and HCT116 cells, and the expression of GATA6 was downregulated ( Figure 4A). Furthermore, the expression level of GATA6 increased in the miR-944 knockdown experiment ( Figure 4B). Moreover, we found that the cancer tissue with high expression levels of miR-944 showed GATA6 protein expression negativity (14/16; Figure 4E c,d, Table S1). Low expression levels of miR-944 showed GATA6 positivity (16/24; Figure 4E a,b, Table S1) in the IHC of 40 pairs of CRC patients collected recently (P = 0.001) ( Table 2). In conclusion, we used GATA6 as the target of miR-944.
To further confirm the combination, the GATA6 3′-UTR-binding sequence (wild type and mutant type), miR-944 was inserted into the pmiR-REPORT plasmid. Then, we cotransfected two pmiR-REPORT plasmids (wild-type and mutant-type) with a mimics-NC and an miR-944 mimic into HEK293 cells. As shown in Figure 4D, we found that transfection of the miR-944 mimics significantly reduced the luciferase activity compared with NC in cells infected with the wild-type 3′-UTR, while there was no significance in cells infected with the mutant-type 3′-UTR pmiR-REPORT plasmid. In conclusion, these results showed that GATA6 is a direct target of miR-944.

| miR-944 affects GATA6 and its downstream proteins in CRC cell lines
GATA6 is a transcription factor that binds to the promoter of CRT.
Our study also showed that miR-944 mediated the downregulation of CRT protein expression by inhibiting GATA6 protein expression ( Figure 4A,B). Immunoprecipitation showed that CRT has a close interaction with AKT ( Figure 4C). The p-AKT protein expression level was decreased in the miR-944 mimic group compared with the corresponding NC group and was increased in the miR-944 inhibitor group compared with the NC group in two CRC cell lines without changing the expression of total AKT proteins ( Figure 4A,B). In this study, we also examined other proliferation-, migration-and invasion-related proteins, TA B L E 2 Expression association between miR-944 and GATA6 (n = 40) in primary CRC F I G U R E 5 No differential protein expression levels were observed in the overexpression and knockdown experiments. The expression of GATA3, E-ca, ERK and P21 was not affected by the overexpression or knockdown of miR-944. E-ca: E-cadherin such as GATA3, ERK, P21 and E-ca ( Figure 5). There was no significant difference between the overexpression and knockdown experiments.
Overall, miR-944 inhibits CRC cell proliferation, migration and invasion by interacting with GATA6 and its downstream proteins.

| GATA6 knockdown reversed the impacts of silencing miR-944 in both cell lines
To verify the relationship between miR-944 and GATA/CRT/p-AKT in CRC, we performed an anti-sense experiment. First, we found that the expression level of GATA6 was significantly reduced when transfecting  is associated with better prognosis in human cancers, such as gastric cancer, bladder cancer and non-small cell lung cancer. 14,20,21 In this study, we analysed the expression of miR-944 in 100 pairs of human Our subsequent experiments showed that GATA6 is the target of miR-944 that was not reported previously to our knowledge.

| D ISCUSS I ON
In the 40 CRC tissues, there was a negative association between miR-944 expression and GATA6 expression. GATA transcription factors are a series of zinc finger proteins that can determine the consensus DNA sequence WGATAA. 22 The GATA family consists of six members (GATA1-6), 23  shown to suppress cancer cell invasion. 16 However, the mechanism by which overexpression of GATA6 inhibits the initiation and progression of CRC is not known. Our subsequent study aimed to determine the potential target of GATA6.
GATA6 has been experimentally identified as the transcription factor that binds to the promoter of calreticulin (CRT) by reporter gene and ChiP analyses in mice and humans. 26 However, no reports were definitely verified the relationship between GATA6 and CRT in cancers to our knowledge. In this study, we also found that the CRT protein changes with the level of GATA6 protein expression in CRC cell lines. Therefore, we concluded that miR-944 mediates CRT protein expression by inhibiting GATA6 protein expression.
In recent years, CRT expression has been reported to be associated with more invasive and advanced malignant processes as well as a poorer prognosis and is positively associated with the UICC stage and lymph node metastasis in several cancers, 27,28 such as breast cancer and pancreatic cancer. Several pathways have been reported to be related to malignant tumour function, such as the modulation of Slug/E-cadherin, 31 PI3K-Akt 32 and MEK-ERK. 33 Because the PI3k-AKT pathways were activated and upregulated by CRT and the expression of E-cadherin and ERK was not impacted by miR-944 in this study, we performed an immunoprecipitation experiment in HCT116 and SW480 cells and found that CRT is structurally associated with AKT. However, the Western blot results showed that miR-944 overexpression or downregulation of miR-944 expression can affect p-AKT protein expression without changing the expression levels of total AKT. The PI3K/AKT signalling pathway participates in the progression of cell proliferation, migration and apoptosis in CRC, 34 and it may be the underlying mechanism by which miR-944 affects CRC cell functions. In fact, the mechanism of miR-944 in CRC is complicated. miR-944 may affect other functions of CRC cells, and perhaps there are many cross-interactions of factors and signalling pathways that need to be further investigated.
In conclusion, our present study provided evidence that miR-944 acts as a tumour suppressor that is significantly downregulated in CRC tissues. For the first time, we showed that miR-944 inhibits CRC cell proliferation, invasion and migration by regulating GATA6 and its downstream proteins. Moreover, the expression of miR-944 was shown to be significantly associated with the TNM stage, depth of invasion, and lymph node status. The interaction between miR-944 and the GATA6 pathway might provide new insight for demonstrating malignant biological behaviour and providing new gene-targeted therapies for CRC.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.