Abstract
Background: SNPs in the promoter region of miRNAs have been reported to be associated with cancer prognosis. Our previous study found that miR-146b had a strong correlation with the stage classification of gastric cancer and contributed to tumor progression. The current study was aimed at investigating whether an SNP located in the promoter region of miR-146b could affect the survival rate of gastric cancer.
Methods: Using bioinformatics tools, we identified one SNP (rs1536309) that is located in the miR-146b promoter. We genotyped this SNP site to assess its association with gastric cancer prognosis in 940 cases.
Results: We found that the dominant model of miR-146b rs1536309 was associated with a higher survival rate of gastric cancer. The association remained significant in the subgroup analysis by age (≤60), sex (male), tumor size (≤5 cm), histologic type (diffuse), lymph node metastasis (N0), distant metastasis (M0), and TNM stage (I/II).
Conclusions: Our results suggested that the miR-146b rs1536309 polymorphism may be a potential biomarker for the prognosis of gastric cancer.
Impact: This is the first evidence showing that patients carrying the miR-146b-5p rs1536309 CC/CT genotypes exhibited better survival than those carrying the TT genotype, suggesting the protective effect of the C allele in the prognosis of gastric cancer. Cancer Epidemiol Biomarkers Prev; 27(7); 822–8. ©2018 AACR.
Introduction
Gastric cancer is one of the most prevalent life-threatening cancers in the world, with high incidence rates in Eastern Asia (1). With improvements in surgical and adjuvant multimodal treatments, the mortality of gastric cancer has declined during the past decade. However, the prognosis of gastric cancer is still poor, and the 5-year survival rate is only approximately 20% (2). In China, gastric cancer remains the third leading cause of cancer-related death and accounted for approximately 498,000 new deaths in 2015 (3). Evidence has revealed that patients with the same tumor grade and pathologic stage who receive similar treatments may have different clinical outcomes, a finding that indicates the importance of individual variants influenced by genetic and environmental factors (4). Therefore, identifying genetic variations in key genes involved in tumor progression as biomarkers to predict the prognosis of gastric cancer patients is crucial and necessary. It may yield benefits for individualized therapy and consequently improve survival outcomes.
miRNAs are short, noncoding regulatory RNAs (18–25 nucleotides) that exert posttranscriptional regulatory functions by targeting the 3′-untranslated regions (3′UTR) of mRNAs for cleavage or transcriptional repression (5). Accumulating evidence has demonstrated that miRNAs are involved in the development of gastric cancer and can serve as a novel tool for predicting the prognosis of gastric cancer (6). An miRNA profiling assay has been used to explore the involvement of miRNAs in cancers. By using this method, our previous study identified the overexpression of miR-146b-5p in gastric cancer and showed a strong correlation of its expression with gastric cancer staging classification (7).
SNPs, a major type of genetic variant, in miRNA precursors and their promoter regions may affect miRNA expression levels, leading to alterations in a variety of biological processes and thereby influencing the survival of patients (8, 9). Therefore, it is rational to postulate that SNPs located in the promoter or the precursor of miR-146b-5p might exist and be associated with prognosis in gastric cancer patients. By using the 1000 Genomes Project database as a resource, we identified a common [minor allele frequency (MAF) > 0.05] SNP (rs1536309) in the miR-146b-5p promoter. Whether this site is related to the prognosis of gastric cancer is still unknown. In this study, we genotyped the rs1536309 SNP in miR-146b-5p in a follow-up study of 940 gastric cancer patients to evaluate its association with the survival of gastric cancer.
Methods and Materials
Ethics statement
The study was approved by the Institutional Review Board of Nanjing Medical University (Nanjing, China). Informed written consent was obtained from all subjects. The experimental protocol was carried out in accordance with International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS).
Study subjects
The details of the recruited subjects were described in the previous study (10). A total of 1,022 gastric cancer patients who underwent gastrectomy between January 1999 and December 2006 were recruited from Yixin People's Hospital, Yixin City, China; 78 of these patients (7.6%) were excluded due to a lack of adequate follow-up information. Four patients who did not have adenocarcinoma according to the pathologic diagnosis by senior pathologists were also excluded from this study. Finally, 940 gastric cancer cases with 100% R0 resection were involved in further analyses. The maximum follow-up time was 119.0 months, and the median follow-up time was 68.5 months. Survival time was estimated from the date of surgery to the date of death or last follow-up (March 2009). Clinical features such as tumor site, tumor size, histologic type, depth of invasion, lymph node metastasis, distant metastasis, and tumor–node–metastasis (TNM) classification were collocated from the medical records of the patients.
SNP selection
We focused on both the miR-146b gene and its promoter region (2 kb upstream from the transcription start site (TTS): chr10: 104194269-104196341) using the UCSC browser (http://genome.ucsc.edu/), in which a total of 9 SNPs resided. On the basis of the criteria of MAF >0.05 in the CHB [Han Chinese in Beijing, China (CHB)] and JPT (Japanese in Tokyo, Japan (JPT)] populations in the 1000 Genomes Project, only SNP rs1536309 were enrolled for further genotyping.
Genotyping
Genomic DNA was extracted from paraffin sections of tumor tissues from all 940 cases. miR-146b-5p rs1536309 (primer: forward: 5′-TGCCTGGATCGCCTTAGCT-3′; reverse: 5′-AGTCCAGTTTCTCATTTTGAAGCA-3′. probe G: 5′-CATTCCCCGTCATC-3′; probe A: 5′-CATTCCCCGTCATC-3′) was genotyped by the TaqMan SNP genotyping assay on the ABI 7900HT Real-time PCR System (Applied Biosystems). Approximately 10% of all samples were randomly selected for further confirmation, and the results were 100% concordant.
Real-time PCR
A total of 56 freshly frozen cancer tissues, which were included in the 940 cases, were subjected to total RNA extraction by the TRIzol Reagent (Invitrogen). The expression of miR-146b-5p was measured by real-time PCR (ABI 7300) with the SYBR Green assay (TaKaRa Biotechnology) after reverse transcription. The conditions for real-time PCR were the same as those in the previous study (7).
Cell culture
Two different gastric cancer cell lines (SGC-7901 and BGC-823) were purchased from the cell bank of Chinese Academy of Sciences (Beijing, China) in 2012 and authenticated by Beijing Microread Genetech Co; Ltd. by using STR Multi-amplification Kit (Microreader 21 ID System) and 3100 DNA Analyzer (Applied Biosystems) in 2014. Cells were cultured in DMEM/high glucose culture medium with 10% FBS, 10 mmol/L HEPES, 2 mmol/L l-glutamine, 1 mmol/L sodium pyruvate, 100 U/mL penicillin, and 100 μg/mL streptomycin. All cells were grown at 37°C in a humidified atmosphere with 5% CO2. Most reagents were obtained from GIBCO.
Construction of promoter-reporter plasmids and luciferase assays
The 5′ flanking region (−2 kb from the transcriptional start site of the miR-146b promoter region) fragments with different rs1536309 alleles were constructed and inserted into the pGL3-promoter vector. DNA sequencing was used to confirm the plasmids.
Cells were cultured in a 24-well plate for 24 hours. The cells were then transfected with the luciferase reporter constructs mentioned above and pRL-SV40 (as control) by using Lipofectamine 2000. Luciferase activity was evaluated with a Dual Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions.
Bioinformatic analysis
To explore the potential functions of the polymorphism, bioinformatic analysis was performed by using HaploReg v4.1 (http://www.broadinstitute.org/mammals/haploreg).
Statistical analysis
Survival time was calculated from the date of gastric cancer diagnosis to the date of death or last follow-up. Survival curves were assessed by Kaplan–Meier analyses. The associations between the survival time and rs1536309, patient characteristics, and clinical information were estimated with the log-rank test. The mean survival time was presented when the median survival time was not available. HRs and their 95% confidence intervals (CIs) with adjustments were calculated by univariate or multivariate Cox regression models, and the Schoenfeld residual was used to evaluate the proportional hazards assumption. The clinical features with significant association of gastric cancer survival served as adjustment variables in the multivariate model (Supplementary Table S1). To determine the predictive factors for the prognosis of gastric cancer, Cox stepwise regression analysis was used with P < 0.05 for entering and P > 0.10 for removing the individual explanatory variables. All statistical analyses were performed with SAS software (version 9.4; SAS Institute Inc.) with a two-sided P value.
Results
Patient characteristics
As shown in Supplementary Table S1, a total of 940 gastric cancer patients, including 724 males (77.0%) and 216 females (23.0%), were included in our study. All of the patients were treated with surgical resection, and 32.4% of them received adjuvant chemotherapy after surgery. Over a follow-up period of up to 119.0 months, 439 deaths were recorded. Histologic type, lymph node metastasis, distant metastasis, depth of invasion, and TNM stage showed a significant association with survival time (log-rank P < 0.05). Compared with patients with diffuse-type gastric cancer, those with intestinal-type histology had better survival (HR = 0.72; 95% CI, 0.59–0.87). Patients with lymph node metastasis or distant metastasis were found to have a higher risk of death (lymph node metastasis: HR = 1.86; 95% CI, 1.51–2.28; distant metastasis: HR = 1.67; 95% CI, 1.18–2.36) than those without lymph node metastasis or distant metastasis. In addition, the risk of death increased in a dose-response manner (log-rank P < 0.001) as the depth of invasion or the TNM stage increased.
Association between miR-146b-5p rs1536309 and gastric cancer prognosis
The log-rank test and Kaplan–Meier survival curves were used to evaluate the association between miR-146b-5p rs1536309 and gastric cancer survival in different genetic models (Table 1). In the codominant model, patients carrying the CT genotype had a better overall survival than those carrying the TT genotype (HR = 0.76; 95% CI, 0.61–0.96; Fig. 1A). In the dominant model, patients with the CT/CC genotypes had a reduced risk of death compared with patients with the TT genotype (HR = 0.78; 95% CI = 0.63–0.98; Fig. 1B). In addition, the Schoenfeld test showed the existence of proportional hazards assumption for dominant model of rs1536309 (P = 0.669) and even the global model (P = 0.497).
miR-146b-5p rs1536309 . | Patients/deaths . | MST (months) . | Log-rank P . | Unadjusted HR (95% CI) . | Adjusted HR (95% CI)a . |
---|---|---|---|---|---|
Codominant model | |||||
TT | 668/332 | 56.93 | 1.00 (reference) | 1.00 (reference) | |
CT | 251/97 | 62.24b | 0.003 | 0.71 (0.57–0.89) | 0.76 (0.61–0.96) |
CC | 21/10 | 26.02b | 0.594 | 1.09 (0.58–2.04) | 1.08 (0.58–2.03) |
Additive model | 0.010 | 0.78 (0.64–0.96) | 0.83 (0.68–1.01) | ||
Dominant model | |||||
TT | 668/332 | 56.93 | 1.00 (reference) | 1.00 (reference) | |
CC+CT | 272/107 | 61.51b | 0.005 | 0.73 (0.59–0.91) | 0.78 (0.63–0.98) |
Recessive model | |||||
CT+TT | 919/429 | 68.53 | 1.00 (reference) | 1.00 (reference) | |
CC | 21/10 | 26.02b | 0.594 | 1.19 (0.63–2.22) | 1.16 (0.62–2.17) |
miR-146b-5p rs1536309 . | Patients/deaths . | MST (months) . | Log-rank P . | Unadjusted HR (95% CI) . | Adjusted HR (95% CI)a . |
---|---|---|---|---|---|
Codominant model | |||||
TT | 668/332 | 56.93 | 1.00 (reference) | 1.00 (reference) | |
CT | 251/97 | 62.24b | 0.003 | 0.71 (0.57–0.89) | 0.76 (0.61–0.96) |
CC | 21/10 | 26.02b | 0.594 | 1.09 (0.58–2.04) | 1.08 (0.58–2.03) |
Additive model | 0.010 | 0.78 (0.64–0.96) | 0.83 (0.68–1.01) | ||
Dominant model | |||||
TT | 668/332 | 56.93 | 1.00 (reference) | 1.00 (reference) | |
CC+CT | 272/107 | 61.51b | 0.005 | 0.73 (0.59–0.91) | 0.78 (0.63–0.98) |
Recessive model | |||||
CT+TT | 919/429 | 68.53 | 1.00 (reference) | 1.00 (reference) | |
CC | 21/10 | 26.02b | 0.594 | 1.19 (0.63–2.22) | 1.16 (0.62–2.17) |
NOTE: Data in bold, P < 0.05.
Abbreviation: MST, median survival time.
aAdjusted for age, sex, tumor size, histological types, tumor site, and TNM stage.
bMean survival time was provided when MST could not be calculated.
Stratified analyses and stepwise Cox regression models for the prognosis of gastric cancer
Stratified analyses were further performed to evaluate the association between miR-146b-5p rs1536309 and gastric cancer survival. As shown in Table 2, better survival was observed in patients with CT/CC genotypes than in those with the TT genotype among the following subgroups: age > 60 (HR = 0.73; 95% CI, 0.54–0.98), male (0.73, 0.57–0.94), tumor size ≤5 cm (0.72, 0.54–0.96), intestinal histology (0.69, 0.48–0.99), T1 depth of invasion (0.44, 0.22–0.87), no lymph node metastasis (0.61, 0.41–0.91), no distant metastasis (0.79, 0.63–0.99), TNM stage I/II (0.65, 0.45–0.93) and patients without chemotherapy (0.65, 0.49–0.86). A stepwise Cox regression model was subsequently used to assess the association between the variables containing the selected characteristics of the patients (age and sex), clinical features (tumor size, tumor sites, histologic type, and TNM stage), miR-146b-5p rs1536309, and gastric cancer prognosis. Finally, two variables (TNM stage and rs1536309) were included in this model (TNM stage: P< 0.001; rs1536309: P = 0.022; Table 3).
. | rs1536309 (patients/deaths) . | . | . | . | |
---|---|---|---|---|---|
Variables . | TT . | CC+CT . | Log-rank P . | Unadjusted HR (95% CI) . | Adjusted HR (95% CI)a . |
Total | 668/332 | 272/107 | 0.005 | 0.73 (0.59–0.91) | 0.78 (0.63–0.97) |
Age (years) | |||||
≤60 | 318/153 | 123/49 | 0.113 | 0.77 (0.56–1.06) | 0.84 (0.61–1.17) |
>60 | 350/179 | 149/58 | 0.018 | 0.70 (0.52–0.94) | 0.73 (0.54–0.98) |
Sex | |||||
Male | 512/254 | 212/81 | 0.004 | 0.69 (0.54–0.89) | 0.73 (0.57–0.94) |
Female | 156/78 | 60/26 | 0.668 | 0.91 (0.58–1.42) | 0.93 (0.59–1.46) |
Tumor size | |||||
≤5 cm | 404/185 | 175/60 | 0.008 | 0.68 (0.51–0.91) | 0.72 (0.54–0.96) |
>5 cm | 264/147 | 97/47 | 0.366 | 0.86 (0.62–1.19) | 0.88 (0.63–1.23) |
Tumor sites | |||||
Cardia | 267/131 | 93/36 | 0.061 | 0.70 (0.49–1.02) | 0.74 (0.51–1.07) |
Noncardia | 401/201 | 179/71 | 0.029 | 0.74 (0.57–0.97) | 0.81 (0.62–1.06) |
Histologic type | |||||
Diffuse | 400/216 | 141/67 | 0.205 | 0.84 (0.64–1.10) | 0.84 (0.64–1.11) |
Intestinal | 268/116 | 131/40 | 0.013 | 0.64 (0.45–0.91) | 0.69 (0.48–0.99) |
Depth of invasion | |||||
T1 | 90/34 | 60/11 | 0.009 | 0.42 (0.21–0.82) | 0.44 (0.22–0.87) |
T2 | 140/57 | 60/26 | 0.812 | 1.06 (0.67–1.68) | 1.07 (0.67–1.70) |
T3 | 403/220 | 142/64 | 0.096 | 0.79 (0.60–1.04) | 0.82 (0.62–1.09) |
T4 | 35/21 | 10/6 | 0.814 | 0.90 (0.36–2.22) | 1.01 (0.37–2.79) |
Lymph node metastasis | |||||
N0 | 255/100 | 119/31 | 0.012 | 0.60 (0.40–0.90) | 0.61 (0.41–0.91) |
N1/N2/N3 | 413/232 | 153/76 | 0.216 | 0.85 (0.66–1.10) | 0.87 (0.67–1.13) |
Distant metastasis | |||||
M0 | 628/307 | 254/97 | 0.006 | 0.73 (0.58–0.91) | 0.79 (0.63–0.99) |
M1 | 40/25 | 18/10 | 0.484 | 0.77 (0.37–1.60) | 0.65 (0.30–1.40) |
TNM stages | |||||
I/II | 304/122 | 143/41 | 0.011 | 0.63 (0.45–0.90) | 0.65 (0.45–0.93) |
III/IV | 364/210 | 129/66 | 0.389 | 0.89 (0.67–1.17) | 0.89 (0.67–1.17) |
Postoperative chemotherapy | |||||
No | 457/238 | 178/63 | <0.001 | 0.60 (0.46–0.80) | 0.65 (0.49–0.86) |
Yes | 211/94 | 94/44 | 0.766 | 1.06 (0.74–1.51) | 1.10 (0.76–1.58) |
. | rs1536309 (patients/deaths) . | . | . | . | |
---|---|---|---|---|---|
Variables . | TT . | CC+CT . | Log-rank P . | Unadjusted HR (95% CI) . | Adjusted HR (95% CI)a . |
Total | 668/332 | 272/107 | 0.005 | 0.73 (0.59–0.91) | 0.78 (0.63–0.97) |
Age (years) | |||||
≤60 | 318/153 | 123/49 | 0.113 | 0.77 (0.56–1.06) | 0.84 (0.61–1.17) |
>60 | 350/179 | 149/58 | 0.018 | 0.70 (0.52–0.94) | 0.73 (0.54–0.98) |
Sex | |||||
Male | 512/254 | 212/81 | 0.004 | 0.69 (0.54–0.89) | 0.73 (0.57–0.94) |
Female | 156/78 | 60/26 | 0.668 | 0.91 (0.58–1.42) | 0.93 (0.59–1.46) |
Tumor size | |||||
≤5 cm | 404/185 | 175/60 | 0.008 | 0.68 (0.51–0.91) | 0.72 (0.54–0.96) |
>5 cm | 264/147 | 97/47 | 0.366 | 0.86 (0.62–1.19) | 0.88 (0.63–1.23) |
Tumor sites | |||||
Cardia | 267/131 | 93/36 | 0.061 | 0.70 (0.49–1.02) | 0.74 (0.51–1.07) |
Noncardia | 401/201 | 179/71 | 0.029 | 0.74 (0.57–0.97) | 0.81 (0.62–1.06) |
Histologic type | |||||
Diffuse | 400/216 | 141/67 | 0.205 | 0.84 (0.64–1.10) | 0.84 (0.64–1.11) |
Intestinal | 268/116 | 131/40 | 0.013 | 0.64 (0.45–0.91) | 0.69 (0.48–0.99) |
Depth of invasion | |||||
T1 | 90/34 | 60/11 | 0.009 | 0.42 (0.21–0.82) | 0.44 (0.22–0.87) |
T2 | 140/57 | 60/26 | 0.812 | 1.06 (0.67–1.68) | 1.07 (0.67–1.70) |
T3 | 403/220 | 142/64 | 0.096 | 0.79 (0.60–1.04) | 0.82 (0.62–1.09) |
T4 | 35/21 | 10/6 | 0.814 | 0.90 (0.36–2.22) | 1.01 (0.37–2.79) |
Lymph node metastasis | |||||
N0 | 255/100 | 119/31 | 0.012 | 0.60 (0.40–0.90) | 0.61 (0.41–0.91) |
N1/N2/N3 | 413/232 | 153/76 | 0.216 | 0.85 (0.66–1.10) | 0.87 (0.67–1.13) |
Distant metastasis | |||||
M0 | 628/307 | 254/97 | 0.006 | 0.73 (0.58–0.91) | 0.79 (0.63–0.99) |
M1 | 40/25 | 18/10 | 0.484 | 0.77 (0.37–1.60) | 0.65 (0.30–1.40) |
TNM stages | |||||
I/II | 304/122 | 143/41 | 0.011 | 0.63 (0.45–0.90) | 0.65 (0.45–0.93) |
III/IV | 364/210 | 129/66 | 0.389 | 0.89 (0.67–1.17) | 0.89 (0.67–1.17) |
Postoperative chemotherapy | |||||
No | 457/238 | 178/63 | <0.001 | 0.60 (0.46–0.80) | 0.65 (0.49–0.86) |
Yes | 211/94 | 94/44 | 0.766 | 1.06 (0.74–1.51) | 1.10 (0.76–1.58) |
NOTE: Data in bold, P < 0.05.
aAdjusted for age, sex, size, site, histologic type, and TNM stage in the Cox regression model.
Entered variables . | β . | SE . | HR (95% CI) . | P . |
---|---|---|---|---|
TNM stage (III/IV vs. I/II) | 0.60 | 0.10 | 1.82 (1.50–2.21) | <0.001 |
rs1536309 (CT/CC vs. TT) | −0.25 | 0.11 | 0.78 (0.62–0.97) | 0.022 |
Entered variables . | β . | SE . | HR (95% CI) . | P . |
---|---|---|---|---|
TNM stage (III/IV vs. I/II) | 0.60 | 0.10 | 1.82 (1.50–2.21) | <0.001 |
rs1536309 (CT/CC vs. TT) | −0.25 | 0.11 | 0.78 (0.62–0.97) | 0.022 |
NOTE: Data in bold, P < 0.05.
β, regression coefficient.
Association between rs1536309 genotype and the expression of miR-146b-5p
We analyzed the expression level of miR-146b-5p in gastric cancer tissues with the different rs1536309 genotypes in the current study. As shown in Fig. 2, patients carrying the CT/CC genotypes expressed lower levels of miR-146b-5p than those with the TT genotype.
Effect of miR-146b-5p rs1536309 on transcriptional activity in cell lines
To explore the biological functions of rs1536309 on miR-146b-5p expression, we constructed luciferase reporter plasmids with either the C or T allele of rs1536309 and transfected them into different cell lines to assess whether this SNP could alter the promoter activity. We observed that the transcriptional activity of the construct with the C allele was significantly lower than that of the construct with the T allele (Fig. 3), which indicated that the C allele could reduce the promoter activity of miR-146b-5p and influence its expression level.
The potential effect of miR-146b-5p rs1536309 on the binding affinity of transcription factors
Given that rs1536309 is located in the promoter region of miR-146b, this genetic variant may affect the promoter activity by altering the binding of transcription factors. Bioinformatic analysis by HaploReg v4.1 suggested that the rs1536309 C allele may significantly reduce the binding affinity of MZF1 (myeloid zinc finger 1; Table 4).
. | PWMb match score . | T allele: TGAGGCTTCCCCTAGCACCTCCATTCCCCATCATCCTGCTTCAAAATGAGAAACTGGAC . | |
---|---|---|---|
PWM IDa . | T allele . | C allele . | C allele: TGAGGCTTCCCCTAGCACCTCCATTCCCCGTCATCCTGCTTCAAAATGAGAAACTGGAC . |
Egr-1_disc4 | −8.9 | −10.4 | RACTACAWBTCCCRGMRKGCMYCGC |
MZF1::1-4_2 | 13.8 | 4.2 | KWBCCCMYMVHHM |
MZF1::1-4_3 | 11.9 | 4.7 | YYHCCCMTMV |
Znf143_disc3 | 13.8 | 13.1 | VCYVCVNBBCCCVSVVBSC |
. | PWMb match score . | T allele: TGAGGCTTCCCCTAGCACCTCCATTCCCCATCATCCTGCTTCAAAATGAGAAACTGGAC . | |
---|---|---|---|
PWM IDa . | T allele . | C allele . | C allele: TGAGGCTTCCCCTAGCACCTCCATTCCCCGTCATCCTGCTTCAAAATGAGAAACTGGAC . |
Egr-1_disc4 | −8.9 | −10.4 | RACTACAWBTCCCRGMRKGCMYCGC |
MZF1::1-4_2 | 13.8 | 4.2 | KWBCCCMYMVHHM |
MZF1::1-4_3 | 11.9 | 4.7 | YYHCCCMTMV |
Znf143_disc3 | 13.8 | 13.1 | VCYVCVNBBCCCVSVVBSC |
aKheradpour and Kellis, 2013.
bPWM, position weight matrix.
Discussion
miRNAs have been reported to play an important role in the occurrence and development of cancer by modulating the expression of cancer-related genes. In our previous miRNA profiling study, we identified the overexpression of miR-146b-5p in gastric cancer tissues, which also showed a strong correlation with the TNM stage (7). These findings indicated that miR-146b-5p may function as a proto-oncogene and may be associated with the prognosis of gastric cancer. Consistent with these findings, Yoon and colleagues reported that the elevation of miR-146b-5p was a strong risk factor for tumor relapse and the poor survival rate of gastric cancer through the inhibition of NOVA1 (11). miR-146b-5p was also observed to play a vital role in the migration and invasion of several tumor cells (12, 13).
Increasing evidence has demonstrated that genetic variants such as SNPs in miRNA promoters or precursor miRNAs are associated with cancer prognosis (9, 14, 15). These variants may affect the expression levels of mature miRNAs or change the binding between the miRNAs and their targets, resulting in a corresponding regulation of target mRNA translation and leading to multiple functional consequences (16–18). Therefore, investigating the association between miR-146b-5p–related SNPs and the prognosis of gastric cancer is necessary. In this study, we focused on SNPs located in the promoter and precursor regions of miR-146b-5p. By using the database of the 1000 Genomes Project, only one SNP (rs1536309) site, which was located 1066 bp upstream of the TTS of the miR-146b-5p precursor, was selected. The association between miR-146b-5p rs1536309 and gastric cancer survival was subsequently evaluated in a follow-up study of 940 GC patients. Our results showed that individuals carrying the C allele had better survival. In the stepwise Cox regression analysis, miR-146b-5p rs1536309 was identified as an independent prognostic factor for gastric cancer.
In particular, this protective effect was significantly represented in the subgroups of patients with age >60, male, tumor size ≤5 cm, and intestinal histology. Compared with elderly patients, the young patients with gastric cancer were more likely to be associated with higher grade, distant metastases, and adjacent organ invasion (19). Thereby, the protective effect of rs1536309 might be more easily observed in older patients. In a recent study, the upregulated miR-146b-5p was found correlated with E2 expression (20). Thus, the influence of this SNP site might be attenuated in female due to the higher level of miR-146b-5p. In our study, a worse survival was observed among patients with diffuse type or tumor size >5 cm. The effect of rs1536309 might be confounded by the bigger tumor size or diffuse type. The survival benefits of rs1536309 were also observed among patients with T1 depth of invasion, no lymph node metastasis, no distant metastasis, and TNM stage I/II. The TNM stage, which is identified by depth of invasion, lymph node metastasis, and distant metastasis, was considered as the independent prognostic factor in gastric cancer in this study. Therefore, the main factor for the prognosis in advanced gastric cancer might not be rs1536309. Moreover, no differences of rs1536309 genotypes were found in the prognosis of patients with postoperative chemotherapy. The relative sample size of the patients and the use of several different regimens might be the reason. Further studies are still needed to evaluate the effect of rs1536309 on patients with chemotherapy.
In addition, given that rs1536309 was genotyped in tumor tissues, we cannot exclude the possibility that the genotypes might include mutations specific to the cancer genome. However, our previous study demonstrated the high concordance rate between tumor tissues and the peripheral blood, suggesting that the majority of SNPs could be accurately genotyped using the DNA isolated from tumor tissues (21). These data suggested that miR-146b-5p rs1536309 may serve as a promising biomarker for the prognosis of gastric cancer patients.
SNPs in the promoters of miRNAs were reported to participate in the alteration of transcriptional activity and thereby affect the expression levels. In our study, patients carrying the rs1536309 C allele were found to have lower expression of miR-146b-5p. The luciferase assay also showed that the rs1536309 C allele significantly reduced the promoter activity. To further explore the potential function of rs1536309, we also searched the Haploreg database and found that the C allele may reduce the binding affinity of the MZF1 (myeloid zinc finger 1) transcription factor, which belongs to the SCAN zinc finger (SCAN-ZF) transcription factor family. MZF1 has been found to be involved in the etiology of several major solid tumors (22). In gastric cancer, MZF1 was found to facilitate the growth, invasion, metastasis, and angiogenesis of cancer cells through the activation of the MMP-14 promoter (23). Taken together, these data suggest that the rs1536309 C allele may reduce the binding affinity of MZF1, thus attenuating miR-146b-5p promoter activity, which ultimately results in the downregulation of miR-146b-5p and has a protective effect in the prognosis of gastric cancer. Further studies are needed to address these questions.
Some limitations of this study should be acknowledged. First, although we collected the data of overall survival, the data of disease-specific and recurrence-free survival were not available due to the incomplete death records. Considering that gastric cancer-related death was found in most of the patients, the overall survival could be used as a surrogate endpoint of disease specific survival. Second, Helicobacter pylori infection, which was regarded as an important risk factor in gastric cancer, was not analyzed due to the lack of information. Third, by using the luciferase assay and bioinformatic analyses, rs1536309 C allele was believed to be associated with the reduced binding affinity of MZF1. However, functional experiments, such as chromatin immunoprecipitation assay, are still needed for further validation.
In conclusion, our results provide the first evidence that the miR-146b-5p rs1536309 CC/CT genotypes are associated with better survival than the TT genotype, suggesting a protective effect of the C allele in the prognosis of gastric cancer. These results indicated that this SNP site may serve as a biomarker to predict and improve prognosis in patients with gastric cancer. However, further validation and functional studies are needed to clarify these findings.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Disclaimer
The funding agencies had no role in the study design, data collection and analysis, decision to publish, or the preparation of the manuscript.
Authors' Contributions
Conception and design: Q. Li, F. Li, Z. Xu, Z. Zhang
Development of methodology: Z. Li, B. Li, L. Wang, F. Li
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): F. Li, L. Yang, W. Gong
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): W. Wang, M. Du, L. Zhang, Q. Li
Writing, review, and/or revision of the manuscript: W. Wang
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): W. Wang, Z. Xu, D. Zhang, F. Qiang
Study supervision: Z. Li, Z. Xu, Z. Zhang
Other (conducted functional experiments): L. Zhang, Q. Li, H. Xu
Acknowledgments
This work was partially supported by the Natural Science Foundation of Jiangsu Province (SBK2017042797, W.Z. Wang); the National Natural Science Foundation of China (81572362, to Z.K. Xu; 81473049, to Z.D. Zhang; and 81230068, to Z.D. Zhang); the National Natural Science Foundation Project of International Cooperation (NSFC-NIH, 81361120398, to Z.K. Xu); the Primary Research & Development Plan of Jiangsu Province (BE2016786, to Z.K. Xu); 333 Project of Jiangsu Province (BRA2015474, to Z.K. Xu); the Priority Academic Program Development of Jiangsu Higher Education Institutions (JX10231801, to Z.K. Xu); the Key Medical Subjects of Jiangsu Province (General Surgery; ZDXKA2016005, to Z.K. Xu); the Program for Development of Innovative Research Team in the First Affiliated Hospital of NJMU (to Z.K. Xu); Jiangsu Key Lab of Cancer Biomarkers, Prevention and Treatment, Collaborative Innovation Center for Cancer Personalized Medicine, Nanjing Medical University (to Z.K. Xu and Z.D. 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.