Cell Biochem Biophys (2015) 71:271–278 DOI 10.1007/s12013-014-0195-y
ORIGINAL PAPER
Genetic Polymorphisms in miRNAs and Susceptibility to Colorectal Cancer Yu Wu • Xiaoxiong Hao • Zili Feng Yunsheng Liu
•
Published online: 12 September 2014 Ó Springer Science+Business Media New York 2014
Abstract The associations of SNPs rs11614913, rs229 2832, and rs2910164 in miRNAs have been exploded in several independent studies and meta-analyses, but the small sample sizes and incomplete data precluded well-defined roles of the miRNA SNPs in the development of CRC. The aim of this study was to combine all available data to comprehensively assess the unclear association. A meta-analysis of nine studies included 2,209 cancers and 2,803 controls, 2,349 cases and 2,663 controls, and 1,409 cases and 1,115 controls for SNP rs11614913, SNP rs2910164, and SNP rs2292832, respectively. The true effect size was estimated by an odds ratio (OR) and 95 % confidence intervals (CI) with the fixed
effects model. For SNP rs11614913, the risk of CRC was more pronounced in the C allele carriers as compared with the T allele carriers among the subjects of Asian decent (CC vs. TT: OR = 1.18, 95 % CI 1.01–1.38, P = 0.734; CC vs. TC ? TT: OR = 1.18, 95 % CI 1.02–1.36, P = 0.573; C vs. T: OR = 1.08, 95 % CI 1.00–1.17, P = 0.775). SNP rs2910164 and SNP rs2292832 were not found to be significantly associated with CRC risk. This meta-analysis reveals that SNP rs11614913, but not SNP rs2910164 and SNP rs2292832, may contribute to susceptibility to CRC in an Asian-specific manner. Keywords
miRNA SNPs CRC
Yu Wu and Xiaoxiong Hao contributed equally to this study. Y. Wu Y. Liu (&) Department of Military Medical Geography, College of High Altitude Military Medicine, Third Military Medical University, Chongqing 400038, China e-mail: [email protected] Y. Wu Y. Liu Key Laboratory of High Altitude Medicine, Third Military Medical University, Ministry of Education, Chongqing 400038, China Y. Wu Y. Liu The Key Laboratory of High Altitude Medicine, PLA, Chongqing 400038, China X. Hao Department of Health Service, Medical Training Base, Third Military Medical University, Chongqing, People’s Republic of China Z. Feng Battalion 19 of Department of Biomedical Engineering, Third Military Medical University, Gaotanyan Street, Chongqing 400038, China
Introduction MicroRNAs (miRNAs) are a subset of endogenous noncoding small RNAs with nearly 22 nucleotides which regulate gene expression at the post-transcriptional level through suppressing translation or degrading the stability of miRNAs [1, 2]. Deregulation of miRNAs induces tumorigenesis, as these molecules function as oncogenes or tumor suppressor genes [3]. Hundreds of miRNA genes have been found to participate in a wide range of physiologic and pathologic processes [4, 5], including cellular proliferation, apoptosis, and metabolism [6–8]. However, little is known about the biological mechanism of the miRNAs underlying cancers. miRNAs are important for the etiology of more human diseases than currently appreciated, and such an etiology can largely attribute to miRNA-related genetic variations. Several lines of evidence has claimed genetic alternation in miRNA target sites involved with disorders ranging from Parkinson’s disease to cancer [9]. Single nucleotide
123
272
polymorphisms (SNPs) or mutations functionally affect binding of miRNA to mRNA critical for regulating the expression level of mRNA, which may confer phenotypic differences among humans [10, 11]. A number of SNPs in miRNA have been explored in various cancers [12–16]. Among them, the most intensive and concentrated SNPs in recent years are miR-196a2 (rs11614913), miR-149 (rs2292832), and miR-146a (rs2910164), particularly their associations with the development of colorectal cancer (CRC), which has been a major public health problem worldwide [17]. While many investigators have been devoted to identify the roles of the three SNPs in CRC risk [18–22], the progress made in this community appears to be ponderous. One explanation is that the SNPs are independently investigated in relatively small-sized studies among diverse populations. Another possibility may be due to less concentration on the new field of miRNA, leading to scattered information in a small quantity of investigations. In this study, we performed a systematical and comprehensive meta-analysis of all publications to date in an attempt to precisely estimate the risk of CRC associated with the most widely replicated SNPs rs11614913, rs2292832, and rs2910164 in miRNAs.
Materials and Methods Study Selection Studies published up to December 9, 2013 concerning the connections between SNPs rs11614913, rs2292832, and rs2910164 in miRNAs and CRC risk were identified in the databases of PubMed, Embase, Web of Science, and China National Knowledge Infrastructure (CNKI) using the following keywords: ‘‘MicroRNA’’ or ‘‘miRNAs,’’ ‘‘miR196a2’’ or ‘‘miR-149’’ or ‘‘miR-146a,’’ ‘‘rs11614913’’ or ‘‘rs2292832’’ or ‘‘rs2910164,’’ ‘‘polymorphism’’ or ‘‘polymorphisms’’ or ‘‘variant,’’ and ‘‘colorectal cancer’’ or ‘‘colon cancer’’ or ‘‘rectal cancer.’’ No restrictions were designed for the search except that the studies must be conducted in humans. Identification of additional relevant articles or data available for the meta-analysis was done through hand searches in the reference lists of eligible studies, meta-analyses, and review articles.
Cell Biochem Biophys (2015) 71:271–278
odds ratio (OR) along with 95 % confidence interval (CI). Accordingly, the articles were excluded if they were a duplication of a follow-up of larger publication or published as an abstract, comment, review, and editorial or case report. Data Extraction The data were independently extracted by two investigators from each study including the name of the first author, year of publication, country where the study was completed, ethnicity, source of controls, genotyping methods, total number of cases and controls, and genotype data of CC, CT, and TT for SNPs rs11614913, rs2292832; CC, GC, and GG for SNP rs2910164. Meta-analysis OR along with 95 % CI was calculated to evaluate the strength of associations between SNPs rs11614913, rs2292832, and rs2910164 and CRC risk. The significance of the pooled ORs was examined by the Z test with a significant level at P \ 0.10. The effect size was estimated for SNPs rs11614913, rs2292832 under CC versus TT, CC ? TC vs. TT, CC versus TC ? TT, C versus T, and TC versus TT; for SNP rs2910164 under GG versus CC, GG ? GC versus CC, GG versus GC ? CC, G versus C, and GC versus CC. Subgroup analyses were carried out by ethnicity for SNPs rs11614913 and rs2910164, respectively. For each SNP, heterogeneity was measured with the I2 of Higgins and Thompson [23], and I2 [ 50 % was the indication of significant inconsistency between studies. Values from single studies were combined using fixed effects model (Mantel–Haenszel method) in the absence of heterogeneity; otherwise, the random effects model (DerSimonian and Laird method) was more appropriate [24, 25]. Whether the single studies included in the meta-analysis significantly altered the ORs was examined using sensitivity analyses by excluding each study and rechecking the corresponding results. Publication bias was determined by Begg’s funnel plot and Egger’s test [26]. Hardy–Weinberg equilibrium (HWE) was tested by Chi-square test in each control group. All analyses were performed with STATA software, version 12.0 (Stata Corporation, College Station, TX).
Inclusion and Exclusion Criteria
Results
To be eligible for this meta-analysis, the publications were required to fulfill the following criteria: (1) a case–control study evaluating at least one of the SNPs rs11614913, rs2292832, and rs2910164 and CRC risk; (2) published as a full-text paper with sufficient genotype data to estimate an
Characteristics of Eligible Studies
123
The process of study selection is presented in full detail in Fig. 1. The output of the initial screening was up to 61 articles. A total of nine articles [18–22, 27–30] were finally identified
Cell Biochem Biophys (2015) 71:271–278
273
TC ? TT genotypes (CC versus TT: OR = 1.18, 95 % CI 1.01–1.38, P = 0.734; CC versus TC ? TT: OR = 1.18, 95 % CI 1.02–1.36, P = 0.573) in the subjects of Asian decent. Analysis of the allele model also revealed a marginally elevated risk in the subgroup of Asians (C versus T: OR = 1.08, 95 % CI 1.00–1.17, P = 0.775). None of the genetic models showed evidence for modified risk of CRC in the Caucasian subgroup. Meta-analysis of the Association Between SNP rs2910164 and Susceptibility to CRC
Fig. 1 Flow diagram of search strategy and study selection
after excluding the studies regarding other SNPs or cancers, not providing genotype information of reference group [31], and review articles about miRNA SNPs [32, 33]. Of the included articles, the studies by Min et al. [18, 20, 21, 29] reported at least two of the studied SNPs in this meta-analysis; thus, they are retrieved separately and classified into the databases of SNPs rs11614913, rs2292832, or rs2910164. Therefore, there were 7 comparisons for SNP rs11614913, 3 for SNP rs2292832, and 5 for SNP rs2910164. HWE of genotype distribution in the controls was tested in 15 comparisons, and three did not respect the HWE [20, 21, 30] (Table 1) (Fig. 2). Meta-analysis of the Association Between SNP rs11614913 and Susceptibility to CRC Seven independent studies involving 2,209 cancers and 2,803 controls were sufficient enough for a meta-analysis to assess the association between the SNP rs11614913 and susceptibility to CRC. The result of this meta-analysis is shown in Table 2. With the heterogeneity test under all of the genetic models revealing no significant heterogeneity between studies, the fixed effects model was used to pool ORs. Generally, no statistically significant association was observed between the analyzed genetic models and CRC susceptibility. However, a tendency to an increased risk of CRC was found in the CC versus TT, CC versus TC ? TT contrast models (OR = 1.13, 95 % CI 0.98–1.29, P = 0.733; OR = 1.11, 95 % CI 0.98–1.26, P = 0.414, respectively), and the C versus T allele model (OR = 1.06, 95 % CI 0.99–1.14, P = 0.727). In the stratified analysis according to ethnicity, the CC genotype demonstrated a significant increase in risk of developing CRC as compared with the TT genotype and
Combining all of the five studies together resulted in 2,349 cases and 2,663 controls for the meta-analysis of the relationship between the SNP rs2910164 and CRC risk. Meta-analysis was carried out with the fixed effects model (on the basis of heterogeneity test) to estimate the pooled ORs. The summarized data did not show any increased risk in the general analysis. The null association remained in both Asians and Caucasians when performing stratification analysis based on racial decent (Table 2). Meta-analysis of the Association Between SNP rs2292832 and Susceptibility to CRC As shown in Table 2, accumulative data from three studies (1,409 cases and 1,115 controls) on the risk of CRC in correlation with the SNP rs2292832 failed to show an elevated or a reduced CRC risk in all of the contrast models. Considering the limited sample size, we did not carry out further stratified analysis. Heterogeneity Test and Sensitivity Analysis Of the three SNPs, no obvious heterogeneity was observed with the exception of the SNP rs2910164 in Asian population (Table 2). Subgroup analysis together with sensitivity analyses identified the study by Chae et al. [22] was the main source of the indicated significant heterogeneity. The exclusion of this study did increase the homogeneity evidently (P value increased from 0.052 to 0.967 for GG vs. CC), while the corresponding ORs were not affected significantly (data not shown). Publication Bias Begg’s funnel plot and Egger’s test were undertaken to detect the potential publication bias of the studies available for this meta-analysis. The symmetrical funnel plots for total studies (P = 0.921) and SNP rs11614913 (P = 1.000) (Fig. 3) under the dominant model did not suggest significant publication bias. Then Egger’s test was performed and
123
123
Publication year
2012
2012
2012
2012
2013
Min1
Chen
Zhu
Hezova1
Zhang1
Vinci1
2013
Vinci2
2012
2013
2013
2013
Hezova2
Vinci3
Chae
Ma
China
Korea
Italy
Czech
Korea
Country of study
Italy
China
Korea
Italy
China
Czech
China
China
Korea
China
Country of study
Asian
Asian
Caucasian
Caucasian
Asian
Ethnicity
Caucasian
Asian
Asian
Caucasian
Asian
Caucasian
Asian
Asian
Asian
Asian
Ethnicity
Hospital
Hospital
Hospital
Hospital
Population
Source mof control
Hospital
Population
Population
Hospital
Population
Hospital
Hospital
Hospital
Population
Hospital
Source of control
RT-PCR
PCR-RFLP
RT-PCR
TaqMan
PCR-RFLP
Genotyping method
RT-PCR
PCR-RFLP
PCR-RFLP
RT-PCR
PCR-RFLP
TaqMan
TaqMan
PCR-LDR
PCR-RFLP
PCR-RFLP
Genotyping method CC
T
C
56 64
27
446
1147
399
160
197
151
169
156
17
12
233
534
182
57
70
444
61
86
115
62
GG
79
50
48
62
79
82
140
240
872
494
91
94
535
C
104
596
619
110
548
141
563
134
451
264
1422
304
229
300
357
G
216
290
273
210
362
253
583
118
441
543
178
435
502
178
463
212
588
407
502
1203
568
178
212
502
Total
CG
58
190
177
86
204
89
303
68 120
Total
CC
128 201
Control
23
203
221
12
172
26
130
35
125
Case
160
443
446
160
455
197
573
126
446
252
Total
TC
Total
TT
Control
Case
192
165
13
9
188
CC
17
187
232
11
185
22
172
107
148
163
TT
614
282
65
79
245
CG
75
202
219
84
197
103
295
206
254
267
TC
1, 2, 3
one study contains several polymorphisms
PCR polymerase chain reaction, PCR-RFLP PCR-restriction fragment length polymorphism, PCR-LDR PCR-ligation detection reaction, RT-PCR real time-PCR, HWE Hardy-Weinberg equilibrium
2011
Min3
miR-146a rs2910164
Publication year
2012
Zhang2
First author
2011
Min2
mir-149 rs2292832
2011
2011
Zhan
miR-196a2 rs11614913
First author
Table 1 Principle characteristics of the studies included in the meta-analysis
397
121
100
124
69
GG
86
46
51
83
81
87
121
94
100
113
CC
998
612
91
97
621
C
109
576
683
106
567
147
639
420
550
593
T
1408
524
265
327
383
G
247
294
321
250
359
277
537
394
454
493
C
0.08
0.98
0.59
0.41
0.44
HWE
0.91
0.43
0.95
0.09
0.03
0.30
0.79
0.79
0.63
0.85
HWE
274 Cell Biochem Biophys (2015) 71:271–278
Cell Biochem Biophys (2015) 71:271–278
275
Fig. 2 Overall estimates of miRNA polymorphisms examined for CRC under the homozygote genotypes (CC vs. TT for rs11614913 and rs2292832; GG vs. CC for rs2910164). The summary odds ratio (OR) is shown by the middle of a solid diamond whose left and right extremes represent the corresponding 95 % CI. Horizontal axis represents ORs which were calculated against controls for each study
revealed no evidence of obvious bias among the studies (Total: P = 0.615, SNP rs11614913: P = 0.959).
Discussion It is widely acceptable that evaluations of the true effect size based on overall averages are more powerful than that gained from an independent study with a small number of genotyped cases and controls, when an appropriate method is adopted. Meta-analysis is an efficient tool to estimate the associations of SNPs and cancers by analyzing a large dataset consisting of many small-sized single reports with inconsistent findings [34]. For the SNPs investigated in our meta-analysis, they have been more commonly studied than various other SNPs of miRNAs in CRC community, which facilitates a meta-analysis in this study. In this meta-analysis including 2,209 cancers and 2,803 controls, 2,349 cases and 2,663 controls, and 1,409 cases and 1,115 controls for SNP rs11614913, SNP rs2910164, and SNP rs2292832, respectively, we examined the relationship between the three well-characterized SNPs in miRNAs and risk of CRC. From the general analysis, we found that compared to the T allele of the SNP rs11614913, the C allele conferred a significantly increased risk of suffering CRC, especially in Asian populations. On the
contrary, analyses of the SNP rs2910164 and SNP rs2292832 showed that they did not appear to be risk factors for CRC development. Although the biological mechanism of SNPs in miRNAs is a novel topic in cancer research, a large quantity of metaanalyses has been engaged in the newly developed field in the past few years [35–38]. A study carried out in 21 studies including 10,441 cases and 12,353 controls suggested that the miR-196a2 C allele is a low-penetrant risk factor for the development of breast cancer and lung cancer [39]. A later smaller meta-analysis of the association between a genetic variant in miR-196a2 and digestive system cancer risks revealed that in addition to hepatocellular carcinoma, the miR-196a2 also increased the susceptibility of CRC [40]. These data, along with the results of the current meta-analysis, implicated that the variation of miR-196a2 polymorphism occurs highly in breast cancer, lung cancer, hepatocellular carcinoma, and CRC, resulting in an increase in the risk of developing these cancers. Different discoveries were also presented in a recent quantitative assessment of cancer risk in four genetic variants in miRNAs (SNPs rs11614913, rs2910164, rs3746444, rs2292832) [41]. It was concluded that SNPs rs11614913 and rs2910164 may protect against the pathogenesis of certain cancers, such as CRC, lung cancer,
123
276 Table 2 Summary of metaanalysis results for the association of miRNA polymorphisms and CRC risk
Cell Biochem Biophys (2015) 71:271–278
Contrast models
Studies (cases/controls)
OR (95% CI)
Model
Heterogeneity Ph
I2 (%)
miR-196a2 rs11614913 Total studies
7 (2,209/2,803)
CC vs. TT
1.13 (0.98, 1.29)
FEM
0.733
0.0
CC ? TC vs. TT
1.04 (0.96, 1.14)
FEM
0.979
0.0
CC vs. TC ? TT
1.11 (0.98, 1.26)
FEM
0.414
1.4
C vs. T
1.06 (0.99, 1.14)
FEM
0.727
0.0
TC vs. TT
1.04 (0.94, 1.16)
FEM
0.969
0.0
CC vs. TT
1.18 (1.01, 1.38)
FEM
0.734
0.0
CC ? TC vs. TT CC vs. TC ? TT
1.06 (0.96, 1.16) 1.18 (1.02, 1.36)
FEM FEM
0.950 0.537
0.0 0.0
C vs. T
1.08 (1.00, 1.17)
FEM
0.775
0.0
TC vs. TT
1.05 (0.94, 1.18)
FEM
0.914
0.0
Ethnicity Asian
Caucasian
5 (1,852/2,413)
2 (357/390)
CC vs. TT
0.95 (0.70, 1.28)
FEM
0.993
0.0
CC ? TC vs. TT
0.98 (0.79, 1.20)
FEM
0.993
0.0
CC vs. TC ? TT
0.93 (0.71, 1.21)
FEM
0.462
0.0
C vs. T
0.96 (0.82, 1.13)
FEM
0.759
0.0
TC vs. TT
0.96 (0.73, 1.27)
FEM
0.847
0.0
GG vs. CC
0.98 (0.86, 1.12)
FEM
0.203
32.8
GG ? GC vs. CC
0.99 (0.91, 1.07)
FEM
0.651
0.0
GG vs. GC ? CC
1.04 (0.93, 1.17)
FEM
0.122
45.1
G vs. C
1.00 (0.94, 1.07)
FEM
0.142
41.9
GC vs. CC Ethnicity
0.98 (0.88, 1.08)
FEM
0.723
0.0
miR-146a rs2910164 Total
Asian
5 (2,349/2,663)
3 (1,992/2,273)
GG vs. CC
0.99 (0.86, 1.14)
FEM
0.052
66.1
GG ? GC vs. CC
0.99 (0.90, 1.08)
FEM
0.296
17.9
GG vs. GC ? CC
1.06 (0.93, 1.21)
FEM
0.032
71.0
G vs. C
1.01 (0.94, 1.08)
FEM
0.035
70.2
0.98 (0.88, 1.10)
FEM
0.368
0.0
0.96 (0.74, 1.25)
FEM
0.914
0.0
GC vs. CC Caucasian
2 (357/390)
GG vs. CC GG ? GC vs. CC
0.97 (0.79, 1.20)
FEM
0.936
0.0
GG vs. GC ? CC
0.98 (0.77, 1.24)
FEM
0.863
0.0
G vs. C
0.98 (0.84, 1.14)
FEM
0.867
0.0
GC vs. CC
0.94 (0.68, 1.30)
FEM
0.932
0.0
mir-149 rs2292832 Total CC vs. TT OR odds ratio, 95 % CI 95 % confidence interval, FEM fixed effects model, Ph P values for heterogeneity analyses
123
3 (1,409/1,115) 0.97 (0.76, 1.24)
FEM
0.964
0.0
CC ? TC vs. TT
0.94 (0.82, 1.09)
FEM
0.997
0.0
CC vs. TC ? TT
1.05 (0.83, 1.32)
FEM
0.986
0.0
C vs. T
0.97 (0.86, 1.08)
FEM
0.993
0.0
TC vs. TT
0.92 (0.78, 1.08)
FEM
0.976
0.0
Cell Biochem Biophys (2015) 71:271–278
Fig. 3 Funnel plots for association studies for miRNA polymorphisms
prostate cancer, and hepatocellular carcinoma. By comparing the numbers of participants included in each metaanalysis, the inconsistencies in results can be attributable to the different sample sizes, and the smaller studies might lack sufficient strength and recognition to provide compelling evidence for the true association. To our knowledge, the associations of SNPs rs11614913, rs2910164 and rs2292832 and CRC risk were for the first time investigated in such a comprehensive meta-analysis. The analyses of sensitivity and publication bias suggested that it is highly unlikely that the findings of the present study may be derived due to chance. However, the results should be interpreted with caution for several limitations. First, heterogeneity is an inevitable problem when a meta-analysis was conducted, and significant heterogeneity between studies may mask the true association that the authors tried to clarify. In our study, while we detected obvious heterogeneity in SNP rs2910164, the overall results were not significantly altered with or without the study identified as the main source of the heterogeneity. Second, the effects of more confounding factors, such as cigarette smoking, alcohol, or tea drinking on CRC risk, were not able to be confirmed, since only one study [21] contained detained genotype data. In summary, this meta-analysis provides support for an association between SNP rs11614913 and an increased risk of CRC in the population of Asian descent. Further welldesigned studies with thousands of subjects in each subgroup along with tissue-specific biochemical and functional characteristics are necessary to validate our findings in future.
References 1. Lewis, B. P., et al. (2003). Prediction of mammalian microRNA targets. Cell, 115(7), 787–798.
277 2. Cai, Y., et al. (2009). A brief review on the mechanisms of miRNA regulation. Genomics Proteomics Bioinformatics, 7(4), 147–154. 3. Mocellin, S., Pasquali, S., & Pilati, P. (2009). Oncomirs: from tumor biology to molecularly targeted anticancer strategies. Mini Reviews in Medicinal Chemistry, 9(1), 70–80. 4. Ambros, V. (2004). The functions of animal microRNAs. Nature, 431(7006), 350–355. 5. Bartel, D. P. (2004). MicroRNAs: genomics, biogenesis, mechanism, and function. Cell, 116(2), 281–297. 6. Johnnidis, J. B., et al. (2008). Regulation of progenitor cell proliferation and granulocyte function by microRNA-223. Nature, 451(7182), 1125–1129. 7. Gong, J., et al. (2013). MicroRNA-125b promotes apoptosis by regulating the expression of Mcl-1, Bcl-w and IL-6R. Oncogene, 32(25), 3071–3079. 8. Aumiller, V., & Forstemann, K. (2008). Roles of microRNAs beyond development–metabolism and neural plasticity. Biochimica et Biophysica Acta, 1779(11), 692–696. 9. Sethupathy, P., & Collins, F. S. (2008). MicroRNA target site polymorphisms and human disease. Trends in Genetics, 24(10), 489–497. 10. Chen, K., et al. (2008). Polymorphisms in microRNA targets: a gold mine for molecular epidemiology. Carcinogenesis, 29(7), 1306–1311. 11. Saunders, M. A., Liang, H., & Li, W. H. (2007). Human polymorphism at microRNAs and microRNA target sites. Proceedings of the National Academy of Sciences of the Unite States of America, 104(9), 3300–3305. 12. Smith, R. A., et al. (2012). A genetic variant located in miR-423 is associated with reduced breast cancer risk. Cancer Genomics & Proteomics, 9(3), 115–118. 13. Hishida, A., et al. (2011). Combined effect of miR-146a rs2910164 G/C polymorphism and Toll-like receptor 4 ?3725 G/C polymorphism on the risk of severe gastric atrophy in Japanese. Digestive Diseases and Sciences, 56(4), 1131–1137. 14. Horikawa, Y., et al. (2008). Single nucleotide polymorphisms of microRNA machinery genes modify the risk of renal cell carcinoma. Clinical Cancer Research, 14(23), 7956–7962. 15. Lung, R. W., et al. (2013). A single nucleotide polymorphism in microRNA-146a is associated with the risk for nasopharyngeal carcinoma. Molecular Carcinogenesis, 52(S1), 28–38. 16. Bae, J. S., et al. (2012). Association study of genetic variations in microRNAs with the risk of hepatitis B-related liver diseases. Digestive and liver disease, 44(10), 849–854. 17. Garcia-Albeniz, X., & Chan, A. T. (2011). Aspirin for the prevention of colorectal cancer. Best Practice & Research Clinical Gastroenterology, 25(4–5), 461–472. 18. Min, K. T., et al. (2012). Association of the miR-146aC [ G, 149C [ T, 196a2C [ T, and 499A [ G polymorphisms with colorectal cancer in the Korean population. Molecular Carcinogenesis, 51(Suppl 1), E65–E73. 19. Chen, H., et al. (2012). A variant in microRNA-196a2 is not associated with susceptibility to and progression of colorectal cancer in Chinese. Internal Medicine Journal, 42(6), e115–e119. 20. Vinci, S., et al. (2013). Genetic and epigenetic factors in regulation of microRNA in colorectal cancers. Methods, 59(1), 138–146. 21. Zhang, M., Study on the Associations of Lifestyle-related Factors, Genetic Variants in miRNA Encoding Regions and miRNA Binding Sites with Colorectal Cancer Risk. Zhejiang University: Zhejiang, 2012. 22. Chae, Y. S., et al. (2013). A miR-146a polymorphism (rs2910164) Predicts Risk of and Survival from Colorectal Cancer. Anticancer Research, 33(8), 3233–3239.
123
278 23. Higgins, J. P., & Thompson, S. G. (2002). Quantifying heterogeneity in a meta-analysis. Statistics in Medicine, 21(11), 1539–1558. 24. DerSimonian, R., & Laird, N. (1986). Meta-analysis in clinical trials. Controlled Clinical Trials, 7(3), 177–188. 25. Mantel, N., & Haenszel, W. (1959). Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute, 22(4), 719–748. 26. Egger, M., et al. (1997). Bias in meta-analysis detected by a simple, graphical test. BMJ, 315(7109), 629–634. 27. Zhan, J. F., et al. (2011). A functional variant in microRNA196a2 is associated with susceptibility of colorectal cancer in a Chinese population. Archives of Medical Research, 42(2), 144–148. 28. Zhu, L., et al. (2012). A functional polymorphism in miRNA196a2 is associated with colorectal cancer risk in a Chinese population. DNA and Cell Biology, 31(3), 350–354. 29. Hezova, R., et al. (2012). Evaluation of SNPs in miR-196-a2, miR-27a and miR-146a as risk factors of colorectal cancer. World Journal of Gastroenterology, 18(22), 2827–2831. 30. Ma, L., et al. (2013). A genetic variant in miR-146a modifies colorectal cancer susceptibility in a Chinese population. Archives of Toxicology, 87(5), 825–833. 31. Jang, M. J., et al. (2011). Prognostic significance of microRNA gene polymorphisms in patients with surgically resected colorectal cancer. Experimental and Therapeutic Medicine, 2(6), 1127–1132. 32. Okayama, H., Schetter, A. J., & Harris, C. C. (2012). MicroRNAs and inflammation in the pathogenesis and progression of colon cancer. Digestive Diseases, 30(Suppl 2), 9–15.
123
Cell Biochem Biophys (2015) 71:271–278 33. Tokarz, P., & Blasiak, J. (2012). The role of microRNA in metastatic colorectal cancer and its significance in cancer prognosis and treatment. Acta Biochimica Polonica, 59(4), 467–474. 34. Berlin, J. A., & Kim, C. J. (2007). The use of meta-analysis in pharmacoepidemiology. In B. L. Strom (Ed.), pharmacoepidemiology (4th ed., pp. 681–707). New York: Wiley. 35. Wu, D., et al. (2013). The miR-146a rs2910164 G [ C polymorphism and susceptibility to digestive cancer in Chinese. Asian Pacific Journal of Cancer Prevention, 14(1), 399–403. 36. Xu, Y., et al. (2013). Different effects of three polymorphisms in MicroRNAs on cancer risk in Asian population: evidence from published literatures. PLoS ONE, 8(6), e65123. 37. Zhang, Y. G., et al. (2013). Association of mir-499 and mir-149 polymorphisms with cancer risk in the Chinese population: evidence from published studies. Asian Pacific Journal of Cancer Prevention, 14(4), 2337–2342. 38. Wang, F., et al. (2012). A genetic variant in microRNA-196a2 is associated with increased cancer risk: a meta-analysis. Molecular Biology Reports, 39(1), 269–275. 39. Qiu, L. X., et al. (2011). miR-196a2 C allele is a low-penetrant risk factor for cancer development. Cytokine, 56(3), 589–592. 40. Guo, J., et al. (2012). A genetic variant in miR-196a2 increased digestive system cancer risks: a meta-analysis of 15 case-control studies. PLoS ONE, 7(1), e30585. 41. He, B., et al. (2012). The association between four genetic variants in microRNAs (rs11614913, rs2910164, rs3746444, rs2292832) and cancer risk: evidence from published studies. PLoS ONE, 7(11), e49032.
ORIGINAL PAPER
Genetic Polymorphisms in miRNAs and Susceptibility to Colorectal Cancer Yu Wu • Xiaoxiong Hao • Zili Feng Yunsheng Liu
•
Published online: 12 September 2014 Ó Springer Science+Business Media New York 2014
Abstract The associations of SNPs rs11614913, rs229 2832, and rs2910164 in miRNAs have been exploded in several independent studies and meta-analyses, but the small sample sizes and incomplete data precluded well-defined roles of the miRNA SNPs in the development of CRC. The aim of this study was to combine all available data to comprehensively assess the unclear association. A meta-analysis of nine studies included 2,209 cancers and 2,803 controls, 2,349 cases and 2,663 controls, and 1,409 cases and 1,115 controls for SNP rs11614913, SNP rs2910164, and SNP rs2292832, respectively. The true effect size was estimated by an odds ratio (OR) and 95 % confidence intervals (CI) with the fixed
effects model. For SNP rs11614913, the risk of CRC was more pronounced in the C allele carriers as compared with the T allele carriers among the subjects of Asian decent (CC vs. TT: OR = 1.18, 95 % CI 1.01–1.38, P = 0.734; CC vs. TC ? TT: OR = 1.18, 95 % CI 1.02–1.36, P = 0.573; C vs. T: OR = 1.08, 95 % CI 1.00–1.17, P = 0.775). SNP rs2910164 and SNP rs2292832 were not found to be significantly associated with CRC risk. This meta-analysis reveals that SNP rs11614913, but not SNP rs2910164 and SNP rs2292832, may contribute to susceptibility to CRC in an Asian-specific manner. Keywords
miRNA SNPs CRC
Yu Wu and Xiaoxiong Hao contributed equally to this study. Y. Wu Y. Liu (&) Department of Military Medical Geography, College of High Altitude Military Medicine, Third Military Medical University, Chongqing 400038, China e-mail: [email protected] Y. Wu Y. Liu Key Laboratory of High Altitude Medicine, Third Military Medical University, Ministry of Education, Chongqing 400038, China Y. Wu Y. Liu The Key Laboratory of High Altitude Medicine, PLA, Chongqing 400038, China X. Hao Department of Health Service, Medical Training Base, Third Military Medical University, Chongqing, People’s Republic of China Z. Feng Battalion 19 of Department of Biomedical Engineering, Third Military Medical University, Gaotanyan Street, Chongqing 400038, China
Introduction MicroRNAs (miRNAs) are a subset of endogenous noncoding small RNAs with nearly 22 nucleotides which regulate gene expression at the post-transcriptional level through suppressing translation or degrading the stability of miRNAs [1, 2]. Deregulation of miRNAs induces tumorigenesis, as these molecules function as oncogenes or tumor suppressor genes [3]. Hundreds of miRNA genes have been found to participate in a wide range of physiologic and pathologic processes [4, 5], including cellular proliferation, apoptosis, and metabolism [6–8]. However, little is known about the biological mechanism of the miRNAs underlying cancers. miRNAs are important for the etiology of more human diseases than currently appreciated, and such an etiology can largely attribute to miRNA-related genetic variations. Several lines of evidence has claimed genetic alternation in miRNA target sites involved with disorders ranging from Parkinson’s disease to cancer [9]. Single nucleotide
123
272
polymorphisms (SNPs) or mutations functionally affect binding of miRNA to mRNA critical for regulating the expression level of mRNA, which may confer phenotypic differences among humans [10, 11]. A number of SNPs in miRNA have been explored in various cancers [12–16]. Among them, the most intensive and concentrated SNPs in recent years are miR-196a2 (rs11614913), miR-149 (rs2292832), and miR-146a (rs2910164), particularly their associations with the development of colorectal cancer (CRC), which has been a major public health problem worldwide [17]. While many investigators have been devoted to identify the roles of the three SNPs in CRC risk [18–22], the progress made in this community appears to be ponderous. One explanation is that the SNPs are independently investigated in relatively small-sized studies among diverse populations. Another possibility may be due to less concentration on the new field of miRNA, leading to scattered information in a small quantity of investigations. In this study, we performed a systematical and comprehensive meta-analysis of all publications to date in an attempt to precisely estimate the risk of CRC associated with the most widely replicated SNPs rs11614913, rs2292832, and rs2910164 in miRNAs.
Materials and Methods Study Selection Studies published up to December 9, 2013 concerning the connections between SNPs rs11614913, rs2292832, and rs2910164 in miRNAs and CRC risk were identified in the databases of PubMed, Embase, Web of Science, and China National Knowledge Infrastructure (CNKI) using the following keywords: ‘‘MicroRNA’’ or ‘‘miRNAs,’’ ‘‘miR196a2’’ or ‘‘miR-149’’ or ‘‘miR-146a,’’ ‘‘rs11614913’’ or ‘‘rs2292832’’ or ‘‘rs2910164,’’ ‘‘polymorphism’’ or ‘‘polymorphisms’’ or ‘‘variant,’’ and ‘‘colorectal cancer’’ or ‘‘colon cancer’’ or ‘‘rectal cancer.’’ No restrictions were designed for the search except that the studies must be conducted in humans. Identification of additional relevant articles or data available for the meta-analysis was done through hand searches in the reference lists of eligible studies, meta-analyses, and review articles.
Cell Biochem Biophys (2015) 71:271–278
odds ratio (OR) along with 95 % confidence interval (CI). Accordingly, the articles were excluded if they were a duplication of a follow-up of larger publication or published as an abstract, comment, review, and editorial or case report. Data Extraction The data were independently extracted by two investigators from each study including the name of the first author, year of publication, country where the study was completed, ethnicity, source of controls, genotyping methods, total number of cases and controls, and genotype data of CC, CT, and TT for SNPs rs11614913, rs2292832; CC, GC, and GG for SNP rs2910164. Meta-analysis OR along with 95 % CI was calculated to evaluate the strength of associations between SNPs rs11614913, rs2292832, and rs2910164 and CRC risk. The significance of the pooled ORs was examined by the Z test with a significant level at P \ 0.10. The effect size was estimated for SNPs rs11614913, rs2292832 under CC versus TT, CC ? TC vs. TT, CC versus TC ? TT, C versus T, and TC versus TT; for SNP rs2910164 under GG versus CC, GG ? GC versus CC, GG versus GC ? CC, G versus C, and GC versus CC. Subgroup analyses were carried out by ethnicity for SNPs rs11614913 and rs2910164, respectively. For each SNP, heterogeneity was measured with the I2 of Higgins and Thompson [23], and I2 [ 50 % was the indication of significant inconsistency between studies. Values from single studies were combined using fixed effects model (Mantel–Haenszel method) in the absence of heterogeneity; otherwise, the random effects model (DerSimonian and Laird method) was more appropriate [24, 25]. Whether the single studies included in the meta-analysis significantly altered the ORs was examined using sensitivity analyses by excluding each study and rechecking the corresponding results. Publication bias was determined by Begg’s funnel plot and Egger’s test [26]. Hardy–Weinberg equilibrium (HWE) was tested by Chi-square test in each control group. All analyses were performed with STATA software, version 12.0 (Stata Corporation, College Station, TX).
Inclusion and Exclusion Criteria
Results
To be eligible for this meta-analysis, the publications were required to fulfill the following criteria: (1) a case–control study evaluating at least one of the SNPs rs11614913, rs2292832, and rs2910164 and CRC risk; (2) published as a full-text paper with sufficient genotype data to estimate an
Characteristics of Eligible Studies
123
The process of study selection is presented in full detail in Fig. 1. The output of the initial screening was up to 61 articles. A total of nine articles [18–22, 27–30] were finally identified
Cell Biochem Biophys (2015) 71:271–278
273
TC ? TT genotypes (CC versus TT: OR = 1.18, 95 % CI 1.01–1.38, P = 0.734; CC versus TC ? TT: OR = 1.18, 95 % CI 1.02–1.36, P = 0.573) in the subjects of Asian decent. Analysis of the allele model also revealed a marginally elevated risk in the subgroup of Asians (C versus T: OR = 1.08, 95 % CI 1.00–1.17, P = 0.775). None of the genetic models showed evidence for modified risk of CRC in the Caucasian subgroup. Meta-analysis of the Association Between SNP rs2910164 and Susceptibility to CRC
Fig. 1 Flow diagram of search strategy and study selection
after excluding the studies regarding other SNPs or cancers, not providing genotype information of reference group [31], and review articles about miRNA SNPs [32, 33]. Of the included articles, the studies by Min et al. [18, 20, 21, 29] reported at least two of the studied SNPs in this meta-analysis; thus, they are retrieved separately and classified into the databases of SNPs rs11614913, rs2292832, or rs2910164. Therefore, there were 7 comparisons for SNP rs11614913, 3 for SNP rs2292832, and 5 for SNP rs2910164. HWE of genotype distribution in the controls was tested in 15 comparisons, and three did not respect the HWE [20, 21, 30] (Table 1) (Fig. 2). Meta-analysis of the Association Between SNP rs11614913 and Susceptibility to CRC Seven independent studies involving 2,209 cancers and 2,803 controls were sufficient enough for a meta-analysis to assess the association between the SNP rs11614913 and susceptibility to CRC. The result of this meta-analysis is shown in Table 2. With the heterogeneity test under all of the genetic models revealing no significant heterogeneity between studies, the fixed effects model was used to pool ORs. Generally, no statistically significant association was observed between the analyzed genetic models and CRC susceptibility. However, a tendency to an increased risk of CRC was found in the CC versus TT, CC versus TC ? TT contrast models (OR = 1.13, 95 % CI 0.98–1.29, P = 0.733; OR = 1.11, 95 % CI 0.98–1.26, P = 0.414, respectively), and the C versus T allele model (OR = 1.06, 95 % CI 0.99–1.14, P = 0.727). In the stratified analysis according to ethnicity, the CC genotype demonstrated a significant increase in risk of developing CRC as compared with the TT genotype and
Combining all of the five studies together resulted in 2,349 cases and 2,663 controls for the meta-analysis of the relationship between the SNP rs2910164 and CRC risk. Meta-analysis was carried out with the fixed effects model (on the basis of heterogeneity test) to estimate the pooled ORs. The summarized data did not show any increased risk in the general analysis. The null association remained in both Asians and Caucasians when performing stratification analysis based on racial decent (Table 2). Meta-analysis of the Association Between SNP rs2292832 and Susceptibility to CRC As shown in Table 2, accumulative data from three studies (1,409 cases and 1,115 controls) on the risk of CRC in correlation with the SNP rs2292832 failed to show an elevated or a reduced CRC risk in all of the contrast models. Considering the limited sample size, we did not carry out further stratified analysis. Heterogeneity Test and Sensitivity Analysis Of the three SNPs, no obvious heterogeneity was observed with the exception of the SNP rs2910164 in Asian population (Table 2). Subgroup analysis together with sensitivity analyses identified the study by Chae et al. [22] was the main source of the indicated significant heterogeneity. The exclusion of this study did increase the homogeneity evidently (P value increased from 0.052 to 0.967 for GG vs. CC), while the corresponding ORs were not affected significantly (data not shown). Publication Bias Begg’s funnel plot and Egger’s test were undertaken to detect the potential publication bias of the studies available for this meta-analysis. The symmetrical funnel plots for total studies (P = 0.921) and SNP rs11614913 (P = 1.000) (Fig. 3) under the dominant model did not suggest significant publication bias. Then Egger’s test was performed and
123
123
Publication year
2012
2012
2012
2012
2013
Min1
Chen
Zhu
Hezova1
Zhang1
Vinci1
2013
Vinci2
2012
2013
2013
2013
Hezova2
Vinci3
Chae
Ma
China
Korea
Italy
Czech
Korea
Country of study
Italy
China
Korea
Italy
China
Czech
China
China
Korea
China
Country of study
Asian
Asian
Caucasian
Caucasian
Asian
Ethnicity
Caucasian
Asian
Asian
Caucasian
Asian
Caucasian
Asian
Asian
Asian
Asian
Ethnicity
Hospital
Hospital
Hospital
Hospital
Population
Source mof control
Hospital
Population
Population
Hospital
Population
Hospital
Hospital
Hospital
Population
Hospital
Source of control
RT-PCR
PCR-RFLP
RT-PCR
TaqMan
PCR-RFLP
Genotyping method
RT-PCR
PCR-RFLP
PCR-RFLP
RT-PCR
PCR-RFLP
TaqMan
TaqMan
PCR-LDR
PCR-RFLP
PCR-RFLP
Genotyping method CC
T
C
56 64
27
446
1147
399
160
197
151
169
156
17
12
233
534
182
57
70
444
61
86
115
62
GG
79
50
48
62
79
82
140
240
872
494
91
94
535
C
104
596
619
110
548
141
563
134
451
264
1422
304
229
300
357
G
216
290
273
210
362
253
583
118
441
543
178
435
502
178
463
212
588
407
502
1203
568
178
212
502
Total
CG
58
190
177
86
204
89
303
68 120
Total
CC
128 201
Control
23
203
221
12
172
26
130
35
125
Case
160
443
446
160
455
197
573
126
446
252
Total
TC
Total
TT
Control
Case
192
165
13
9
188
CC
17
187
232
11
185
22
172
107
148
163
TT
614
282
65
79
245
CG
75
202
219
84
197
103
295
206
254
267
TC
1, 2, 3
one study contains several polymorphisms
PCR polymerase chain reaction, PCR-RFLP PCR-restriction fragment length polymorphism, PCR-LDR PCR-ligation detection reaction, RT-PCR real time-PCR, HWE Hardy-Weinberg equilibrium
2011
Min3
miR-146a rs2910164
Publication year
2012
Zhang2
First author
2011
Min2
mir-149 rs2292832
2011
2011
Zhan
miR-196a2 rs11614913
First author
Table 1 Principle characteristics of the studies included in the meta-analysis
397
121
100
124
69
GG
86
46
51
83
81
87
121
94
100
113
CC
998
612
91
97
621
C
109
576
683
106
567
147
639
420
550
593
T
1408
524
265
327
383
G
247
294
321
250
359
277
537
394
454
493
C
0.08
0.98
0.59
0.41
0.44
HWE
0.91
0.43
0.95
0.09
0.03
0.30
0.79
0.79
0.63
0.85
HWE
274 Cell Biochem Biophys (2015) 71:271–278
Cell Biochem Biophys (2015) 71:271–278
275
Fig. 2 Overall estimates of miRNA polymorphisms examined for CRC under the homozygote genotypes (CC vs. TT for rs11614913 and rs2292832; GG vs. CC for rs2910164). The summary odds ratio (OR) is shown by the middle of a solid diamond whose left and right extremes represent the corresponding 95 % CI. Horizontal axis represents ORs which were calculated against controls for each study
revealed no evidence of obvious bias among the studies (Total: P = 0.615, SNP rs11614913: P = 0.959).
Discussion It is widely acceptable that evaluations of the true effect size based on overall averages are more powerful than that gained from an independent study with a small number of genotyped cases and controls, when an appropriate method is adopted. Meta-analysis is an efficient tool to estimate the associations of SNPs and cancers by analyzing a large dataset consisting of many small-sized single reports with inconsistent findings [34]. For the SNPs investigated in our meta-analysis, they have been more commonly studied than various other SNPs of miRNAs in CRC community, which facilitates a meta-analysis in this study. In this meta-analysis including 2,209 cancers and 2,803 controls, 2,349 cases and 2,663 controls, and 1,409 cases and 1,115 controls for SNP rs11614913, SNP rs2910164, and SNP rs2292832, respectively, we examined the relationship between the three well-characterized SNPs in miRNAs and risk of CRC. From the general analysis, we found that compared to the T allele of the SNP rs11614913, the C allele conferred a significantly increased risk of suffering CRC, especially in Asian populations. On the
contrary, analyses of the SNP rs2910164 and SNP rs2292832 showed that they did not appear to be risk factors for CRC development. Although the biological mechanism of SNPs in miRNAs is a novel topic in cancer research, a large quantity of metaanalyses has been engaged in the newly developed field in the past few years [35–38]. A study carried out in 21 studies including 10,441 cases and 12,353 controls suggested that the miR-196a2 C allele is a low-penetrant risk factor for the development of breast cancer and lung cancer [39]. A later smaller meta-analysis of the association between a genetic variant in miR-196a2 and digestive system cancer risks revealed that in addition to hepatocellular carcinoma, the miR-196a2 also increased the susceptibility of CRC [40]. These data, along with the results of the current meta-analysis, implicated that the variation of miR-196a2 polymorphism occurs highly in breast cancer, lung cancer, hepatocellular carcinoma, and CRC, resulting in an increase in the risk of developing these cancers. Different discoveries were also presented in a recent quantitative assessment of cancer risk in four genetic variants in miRNAs (SNPs rs11614913, rs2910164, rs3746444, rs2292832) [41]. It was concluded that SNPs rs11614913 and rs2910164 may protect against the pathogenesis of certain cancers, such as CRC, lung cancer,
123
276 Table 2 Summary of metaanalysis results for the association of miRNA polymorphisms and CRC risk
Cell Biochem Biophys (2015) 71:271–278
Contrast models
Studies (cases/controls)
OR (95% CI)
Model
Heterogeneity Ph
I2 (%)
miR-196a2 rs11614913 Total studies
7 (2,209/2,803)
CC vs. TT
1.13 (0.98, 1.29)
FEM
0.733
0.0
CC ? TC vs. TT
1.04 (0.96, 1.14)
FEM
0.979
0.0
CC vs. TC ? TT
1.11 (0.98, 1.26)
FEM
0.414
1.4
C vs. T
1.06 (0.99, 1.14)
FEM
0.727
0.0
TC vs. TT
1.04 (0.94, 1.16)
FEM
0.969
0.0
CC vs. TT
1.18 (1.01, 1.38)
FEM
0.734
0.0
CC ? TC vs. TT CC vs. TC ? TT
1.06 (0.96, 1.16) 1.18 (1.02, 1.36)
FEM FEM
0.950 0.537
0.0 0.0
C vs. T
1.08 (1.00, 1.17)
FEM
0.775
0.0
TC vs. TT
1.05 (0.94, 1.18)
FEM
0.914
0.0
Ethnicity Asian
Caucasian
5 (1,852/2,413)
2 (357/390)
CC vs. TT
0.95 (0.70, 1.28)
FEM
0.993
0.0
CC ? TC vs. TT
0.98 (0.79, 1.20)
FEM
0.993
0.0
CC vs. TC ? TT
0.93 (0.71, 1.21)
FEM
0.462
0.0
C vs. T
0.96 (0.82, 1.13)
FEM
0.759
0.0
TC vs. TT
0.96 (0.73, 1.27)
FEM
0.847
0.0
GG vs. CC
0.98 (0.86, 1.12)
FEM
0.203
32.8
GG ? GC vs. CC
0.99 (0.91, 1.07)
FEM
0.651
0.0
GG vs. GC ? CC
1.04 (0.93, 1.17)
FEM
0.122
45.1
G vs. C
1.00 (0.94, 1.07)
FEM
0.142
41.9
GC vs. CC Ethnicity
0.98 (0.88, 1.08)
FEM
0.723
0.0
miR-146a rs2910164 Total
Asian
5 (2,349/2,663)
3 (1,992/2,273)
GG vs. CC
0.99 (0.86, 1.14)
FEM
0.052
66.1
GG ? GC vs. CC
0.99 (0.90, 1.08)
FEM
0.296
17.9
GG vs. GC ? CC
1.06 (0.93, 1.21)
FEM
0.032
71.0
G vs. C
1.01 (0.94, 1.08)
FEM
0.035
70.2
0.98 (0.88, 1.10)
FEM
0.368
0.0
0.96 (0.74, 1.25)
FEM
0.914
0.0
GC vs. CC Caucasian
2 (357/390)
GG vs. CC GG ? GC vs. CC
0.97 (0.79, 1.20)
FEM
0.936
0.0
GG vs. GC ? CC
0.98 (0.77, 1.24)
FEM
0.863
0.0
G vs. C
0.98 (0.84, 1.14)
FEM
0.867
0.0
GC vs. CC
0.94 (0.68, 1.30)
FEM
0.932
0.0
mir-149 rs2292832 Total CC vs. TT OR odds ratio, 95 % CI 95 % confidence interval, FEM fixed effects model, Ph P values for heterogeneity analyses
123
3 (1,409/1,115) 0.97 (0.76, 1.24)
FEM
0.964
0.0
CC ? TC vs. TT
0.94 (0.82, 1.09)
FEM
0.997
0.0
CC vs. TC ? TT
1.05 (0.83, 1.32)
FEM
0.986
0.0
C vs. T
0.97 (0.86, 1.08)
FEM
0.993
0.0
TC vs. TT
0.92 (0.78, 1.08)
FEM
0.976
0.0
Cell Biochem Biophys (2015) 71:271–278
Fig. 3 Funnel plots for association studies for miRNA polymorphisms
prostate cancer, and hepatocellular carcinoma. By comparing the numbers of participants included in each metaanalysis, the inconsistencies in results can be attributable to the different sample sizes, and the smaller studies might lack sufficient strength and recognition to provide compelling evidence for the true association. To our knowledge, the associations of SNPs rs11614913, rs2910164 and rs2292832 and CRC risk were for the first time investigated in such a comprehensive meta-analysis. The analyses of sensitivity and publication bias suggested that it is highly unlikely that the findings of the present study may be derived due to chance. However, the results should be interpreted with caution for several limitations. First, heterogeneity is an inevitable problem when a meta-analysis was conducted, and significant heterogeneity between studies may mask the true association that the authors tried to clarify. In our study, while we detected obvious heterogeneity in SNP rs2910164, the overall results were not significantly altered with or without the study identified as the main source of the heterogeneity. Second, the effects of more confounding factors, such as cigarette smoking, alcohol, or tea drinking on CRC risk, were not able to be confirmed, since only one study [21] contained detained genotype data. In summary, this meta-analysis provides support for an association between SNP rs11614913 and an increased risk of CRC in the population of Asian descent. Further welldesigned studies with thousands of subjects in each subgroup along with tissue-specific biochemical and functional characteristics are necessary to validate our findings in future.
References 1. Lewis, B. P., et al. (2003). Prediction of mammalian microRNA targets. Cell, 115(7), 787–798.
277 2. Cai, Y., et al. (2009). A brief review on the mechanisms of miRNA regulation. Genomics Proteomics Bioinformatics, 7(4), 147–154. 3. Mocellin, S., Pasquali, S., & Pilati, P. (2009). Oncomirs: from tumor biology to molecularly targeted anticancer strategies. Mini Reviews in Medicinal Chemistry, 9(1), 70–80. 4. Ambros, V. (2004). The functions of animal microRNAs. Nature, 431(7006), 350–355. 5. Bartel, D. P. (2004). MicroRNAs: genomics, biogenesis, mechanism, and function. Cell, 116(2), 281–297. 6. Johnnidis, J. B., et al. (2008). Regulation of progenitor cell proliferation and granulocyte function by microRNA-223. Nature, 451(7182), 1125–1129. 7. Gong, J., et al. (2013). MicroRNA-125b promotes apoptosis by regulating the expression of Mcl-1, Bcl-w and IL-6R. Oncogene, 32(25), 3071–3079. 8. Aumiller, V., & Forstemann, K. (2008). Roles of microRNAs beyond development–metabolism and neural plasticity. Biochimica et Biophysica Acta, 1779(11), 692–696. 9. Sethupathy, P., & Collins, F. S. (2008). MicroRNA target site polymorphisms and human disease. Trends in Genetics, 24(10), 489–497. 10. Chen, K., et al. (2008). Polymorphisms in microRNA targets: a gold mine for molecular epidemiology. Carcinogenesis, 29(7), 1306–1311. 11. Saunders, M. A., Liang, H., & Li, W. H. (2007). Human polymorphism at microRNAs and microRNA target sites. Proceedings of the National Academy of Sciences of the Unite States of America, 104(9), 3300–3305. 12. Smith, R. A., et al. (2012). A genetic variant located in miR-423 is associated with reduced breast cancer risk. Cancer Genomics & Proteomics, 9(3), 115–118. 13. Hishida, A., et al. (2011). Combined effect of miR-146a rs2910164 G/C polymorphism and Toll-like receptor 4 ?3725 G/C polymorphism on the risk of severe gastric atrophy in Japanese. Digestive Diseases and Sciences, 56(4), 1131–1137. 14. Horikawa, Y., et al. (2008). Single nucleotide polymorphisms of microRNA machinery genes modify the risk of renal cell carcinoma. Clinical Cancer Research, 14(23), 7956–7962. 15. Lung, R. W., et al. (2013). A single nucleotide polymorphism in microRNA-146a is associated with the risk for nasopharyngeal carcinoma. Molecular Carcinogenesis, 52(S1), 28–38. 16. Bae, J. S., et al. (2012). Association study of genetic variations in microRNAs with the risk of hepatitis B-related liver diseases. Digestive and liver disease, 44(10), 849–854. 17. Garcia-Albeniz, X., & Chan, A. T. (2011). Aspirin for the prevention of colorectal cancer. Best Practice & Research Clinical Gastroenterology, 25(4–5), 461–472. 18. Min, K. T., et al. (2012). Association of the miR-146aC [ G, 149C [ T, 196a2C [ T, and 499A [ G polymorphisms with colorectal cancer in the Korean population. Molecular Carcinogenesis, 51(Suppl 1), E65–E73. 19. Chen, H., et al. (2012). A variant in microRNA-196a2 is not associated with susceptibility to and progression of colorectal cancer in Chinese. Internal Medicine Journal, 42(6), e115–e119. 20. Vinci, S., et al. (2013). Genetic and epigenetic factors in regulation of microRNA in colorectal cancers. Methods, 59(1), 138–146. 21. Zhang, M., Study on the Associations of Lifestyle-related Factors, Genetic Variants in miRNA Encoding Regions and miRNA Binding Sites with Colorectal Cancer Risk. Zhejiang University: Zhejiang, 2012. 22. Chae, Y. S., et al. (2013). A miR-146a polymorphism (rs2910164) Predicts Risk of and Survival from Colorectal Cancer. Anticancer Research, 33(8), 3233–3239.
123
278 23. Higgins, J. P., & Thompson, S. G. (2002). Quantifying heterogeneity in a meta-analysis. Statistics in Medicine, 21(11), 1539–1558. 24. DerSimonian, R., & Laird, N. (1986). Meta-analysis in clinical trials. Controlled Clinical Trials, 7(3), 177–188. 25. Mantel, N., & Haenszel, W. (1959). Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute, 22(4), 719–748. 26. Egger, M., et al. (1997). Bias in meta-analysis detected by a simple, graphical test. BMJ, 315(7109), 629–634. 27. Zhan, J. F., et al. (2011). A functional variant in microRNA196a2 is associated with susceptibility of colorectal cancer in a Chinese population. Archives of Medical Research, 42(2), 144–148. 28. Zhu, L., et al. (2012). A functional polymorphism in miRNA196a2 is associated with colorectal cancer risk in a Chinese population. DNA and Cell Biology, 31(3), 350–354. 29. Hezova, R., et al. (2012). Evaluation of SNPs in miR-196-a2, miR-27a and miR-146a as risk factors of colorectal cancer. World Journal of Gastroenterology, 18(22), 2827–2831. 30. Ma, L., et al. (2013). A genetic variant in miR-146a modifies colorectal cancer susceptibility in a Chinese population. Archives of Toxicology, 87(5), 825–833. 31. Jang, M. J., et al. (2011). Prognostic significance of microRNA gene polymorphisms in patients with surgically resected colorectal cancer. Experimental and Therapeutic Medicine, 2(6), 1127–1132. 32. Okayama, H., Schetter, A. J., & Harris, C. C. (2012). MicroRNAs and inflammation in the pathogenesis and progression of colon cancer. Digestive Diseases, 30(Suppl 2), 9–15.
123
Cell Biochem Biophys (2015) 71:271–278 33. Tokarz, P., & Blasiak, J. (2012). The role of microRNA in metastatic colorectal cancer and its significance in cancer prognosis and treatment. Acta Biochimica Polonica, 59(4), 467–474. 34. Berlin, J. A., & Kim, C. J. (2007). The use of meta-analysis in pharmacoepidemiology. In B. L. Strom (Ed.), pharmacoepidemiology (4th ed., pp. 681–707). New York: Wiley. 35. Wu, D., et al. (2013). The miR-146a rs2910164 G [ C polymorphism and susceptibility to digestive cancer in Chinese. Asian Pacific Journal of Cancer Prevention, 14(1), 399–403. 36. Xu, Y., et al. (2013). Different effects of three polymorphisms in MicroRNAs on cancer risk in Asian population: evidence from published literatures. PLoS ONE, 8(6), e65123. 37. Zhang, Y. G., et al. (2013). Association of mir-499 and mir-149 polymorphisms with cancer risk in the Chinese population: evidence from published studies. Asian Pacific Journal of Cancer Prevention, 14(4), 2337–2342. 38. Wang, F., et al. (2012). A genetic variant in microRNA-196a2 is associated with increased cancer risk: a meta-analysis. Molecular Biology Reports, 39(1), 269–275. 39. Qiu, L. X., et al. (2011). miR-196a2 C allele is a low-penetrant risk factor for cancer development. Cytokine, 56(3), 589–592. 40. Guo, J., et al. (2012). A genetic variant in miR-196a2 increased digestive system cancer risks: a meta-analysis of 15 case-control studies. PLoS ONE, 7(1), e30585. 41. He, B., et al. (2012). The association between four genetic variants in microRNAs (rs11614913, rs2910164, rs3746444, rs2292832) and cancer risk: evidence from published studies. PLoS ONE, 7(11), e49032.
