Genetic association of ANRIL with susceptibility to Ischemic stroke: A comprehensive meta-analysis.

BACKGROUND: Ischemic stroke (IS) is a complex polygenic disease with a strong genetic background. The relationship between the ANRIL (antisense non-coding RNA in the INK4 locus) in chromosome 9p21 region and IS has been reported across populations worldwide; however, these studies have yielded inconsistent results. The aim of this study is to clarify the types of single-nucleotide polymorphisms on the ANRIL locus associated with susceptibility to IS using meta-analysis and comprehensively assess the strength of the association.
METHODS: Relevant studies were identified by comprehensive and systematic literature searches. The quality of each study was assessed using the Newcastle-Ottawa Scale. Allele and genotype frequencies were extracted from each of the included studies. Odds ratios with corresponding 95% confidence intervals of combined analyses were calculated under three genetic models (allele frequency comparison, dominant model, and recessive model) using a random-effects or fixed-effects model. Heterogeneity was tested using the chi-square test based on the Cochran Q statistic and I2 metric, and subgroup analyses and a meta-regression model were used to explore sources of heterogeneity. The correction for multiple testing used the false discovery rate method proposed by Benjamini and Hochberg. The assessment of publication bias employed funnel plots and Egger's test.
RESULTS: We identified 25 studies (15 SNPs, involving a total of 11,527 cases and 12,216 controls maximum) and performed a meta-analysis. Eight SNPs (rs10757274, rs10757278, rs2383206, rs1333040, rs1333049, rs1537378, rs4977574, and rs1004638) in ANRIL were significantly associated with IS risk. Six of these SNPs (rs10757274, rs10757278, rs2383206, rs1333040, rs1537378, and rs4977574) had a significant relationship to the large artery atherosclerosis subtype of IS. Two SNPs (rs2383206 and rs4977574) were associated with IS mainly in Asians, and three SNPs (rs10757274, rs1333040, and rs1333049) were associated with susceptibility to IS mainly in Caucasians. Sensitivity analyses confirmed the reliability of the original results. Ethnicity and individual studies may be the main sources of heterogeneity in ANRIL.
CONCLUSIONS: Our results suggest that some single-nucleotide polymorphisms on the ANRIL locus may be associated with IS risk. Future studies with larger sample numbers are necessary to confirm this result. Additional functional analyses of causal effects of these polymorphisms on IS subtypes are also essential.

Introduction

Stroke is the second leading cause of death in the world [1] and the first leading cause of death in China [2]. In 2017, the National Epidemiological Survey of Stroke in China (NESS-China) from 31 provinces reported that the incidence and mortality rates of stroke were 246.8 and 114.8 per 100,000 person-years, respectively, and it is estimated that about 3.4 million new stroke cases occur each year [3]. Stroke warrants some of the highest medical costs in China, costing nearly 75.6 billion yuan (RMB) in direct medical costs [4]. Hospitalization expenses are projected to increase significantly with the expected improvement in people’s living standards [5]. Ischemic stroke (IS) accounted for 43.7%–78.9% of all stroke cases in China [6]. IS is a complex disorder with a strong genetic component [7]. Thrombosis of brain arteries secondary to atherosclerosis is considered one of the major pathophysiological mechanisms of IS [8]. Thus, studies into genetic susceptibility to atherosclerosis have attracted a lot of attention.

ANRIL (antisense non-coding RNA in the INK4 locus), which belongs to the long non-coding RNA family, was found to have a strong association with the risk for cardio-metabolic diseases [9], playing a key role in atherosclerotic diseases such as IS. A number of studies have explored the relationship between ANRIL and IS across populations worldwide. However, most of these studies used small sample sizes and the findings were inconclusive. Data from linkage and association studies showed that susceptible locus for common diseases had only minimal effects. Meta-analysis is a powerful tool that allows the detection and validation of minimal biological effects in human genetic association studies [10]. Researchers have investigated the role of a few single-nucleotide polymorphisms (SNPs) on the ANRIL locus in IS across different populations by meta-analysis. However, the association of other genetic variants and other SNPs in ANRIL with IS deserves further analyses. In addition, some recently published studies across ethnicities were found in the literature search. In this study, we conducted an updated meta-analysis on all available association study data to comprehensively evaluate the contribution of ANRIL to the risk of IS.

Materials and methods

Study design

This research was conducted according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) statement and the guidelines presented in Systematic Reviews of Genetic Association Studies by Sagoo et al. [10]. ANRIL polymorphism was used as the exposure and IS as an outcome. This work did not require the approval of an ethics committee and was not registered in any database. The completed PRISMA checklist and Meta-analysis on Genetic Association Studies Checklist are given in S1, S2 Appendices.

Data collection

All studies involving the relationship between ANRIL gene polymorphisms and stroke were identified independently by three investigators (Bai N, Liu W, and Zhou Q) by searching the following databases until August 2021: PubMed (from 1966), EMBASE (from 1966), the Cochrane Library (from 2003), ProQuest Dissertations & Theses Database (from 1980), Biosis Preview (from 1990), Web of Science (from 1990), China National Knowledge Infrastructure (CNKI, from 194), and Wanfang Database (including journal articles, dissertations or theses, and conferences literature, from 1990). We used the following keywords or their combinations in search strategies: “ANRIL”, “CDKN2B-AS1”, “antisense non-coding RNA in the INK4 locus”, or “9p21” and “stroke”, “cerebral infarction”, or “cerebrovascular disease”. We limited the search to only human studies. Examples of the keywords search strategy in PubMed are: (“ANRIL”[All Fields] OR “CDKN2B-AS1”[All Fields] OR “antisense non-coding RNA in the INK4 locus”[All Fields] OR “9p21”[All Fields]) AND (“stroke”[All Fields] OR “cerebral infarction”[All Fields] OR “cerebrovascular disease”[All Fields]).

The references listed in the retrieved articles and in review articles as well as abstracts from recent conferences were also searched for possible eligible studies. Only the most recent or complete reports were selected for analysis if the same or a similar patient cohort was included in several publications. There were no restrictions on the source of the control group, and studies in which the control groups were not in Hardy-Weinberg equilibrium were excluded [11].

Studies meeting the following criteria were included for meta-analysis: 1) genetic association studies of the ANRIL polymorphisms with IS were performed using a population (hospital)-based, case-control, nested case-control, or cohort design; 2) IS was diagnosed using a standard that has been widely accepted; 3) control subjects were unrelated individuals, with no symptomatic vascular disease as confirmed by physicians; 4) genotype or allele frequencies were reported in both patients with IS and in controls or could be calculated successfully; and 5) a genetic variant of ANRIL was included in at least two of the studies. Case-only studies, family-based studies, and review articles were excluded. The quality of included studies was assessed based on the published study [12] and the Newcastle-Ottawa Scale (NOS) [13]. A NOS score ≥7 was considered high quality [13].

Data extraction

Data were carefully extracted from all eligible studies independently by two authors (Liu W, Xiang T), and any disagreements were resolved by discussion. The following information was extracted: first author’s surname, year of publication, country of origin, study design, sex composition of the case and control groups, ethnicity of the subjects studied, total number of subjects, definition and characteristics of cases and controls, genetic variants associated with IS, genotyping methods, distribution of genotypes and alleles, IS subtype (if reported), information on additional genetic variants, as well as gene–gene and gene–environment interactions (if investigated). Genotype frequencies were calculated where possible.

For studies that included subjects from different ethnic groups, data were extracted separately for each ethnic group. When some of the information was not available, we contacted the corresponding author by email for additional information.

Statistical analyses

Odds ratios (ORs) and pooled ORs with corresponding 95% confidence intervals (CIs) were calculated using the fixed-effects or random-effects model. For the chi-square test based on Cochran Q statistic, p-values <0.10 were considered to be statistically significant [14]. The I2 metric was used to evaluate the heterogeneity among studies [15].

Hardy-Weinberg equilibrium was tested in the control groups using the chi-square test. Three genetic models were used to examine the association of ANRIL polymorphisms and risk of IS: (1) allele contrast (AC) (effect of each additional risk allele), (2) dominant model (DM), and (3) recessive model (RM). Multiple testing correction was conducted using the false discovery rate (FDR) method proposed by Benjamini and Hochberg. Inverted funnel plots and Egger’s test were performed to detect publication bias in the analyses involving different genetic variants. Publication bias was considered to be present if the inverted funnel plot was asymmetric and/or Egger’s test result was significant (p <0.10).

Sub-population analyses were conducted for ethnicity [16], and subgroup analyses for IS subtype, age, or sex (if available) were also performed [17]. A sensitivity analysis was performed with the exclusion of specific studies [18], such as poor-quality studies (NOS <7) or studies where no ANRIL genetic variants were found in either cases or controls. All statistical analyses were performed with the Cochrane Review Manager (RevMan, version 5.4) and STATA 16.0 package. A probability value of p<0.05 (two-tailed) was considered significant unless indicated otherwise.

Results

Study selection and characteristics of eligible datasets

We found 856 records by primary searches in the databases and six additional records were identified from other sources, including 113 articles from English-language databases and 749 items from Chinese-language databases. Initially, 115 potentially relevant articles (16 in Chinese and 99 in English) were initially selected after reading the titles and abstracts. After reading the full text of these articles, 90 articles were excluded because of duplicates, reviews, mixed samples (transient ischemic attack or hemorrhagic stroke were not excluded), insufficient data, irrelevant content, genetic variants beyond the scope of this study, or ineligible study design. Finally, 25 articles (2 in Chinese and 23 in English) [19-43] involving 15 SNPs (rs2383207, rs10757274, rs10757278, rs2383206, rs1333040, rs1333049, rs1537378, rs4977574, rs1004638, rs7865618, rs10965227, rs1333042, rs7044859, rs10116277, and rs10757269) were found to be eligible for the meta-analysis after applying all the inclusion and exclusion criteria described above. The results of the systematic literature search and article selection are summarized in Fig 1. The excluded articles and the reasons for excluding each article are given in S3 Appendix.

Fig 1

Flowchart of the literature search and article selection for the meta-analysis.

Twenty-three of the included articles were full-length reports published in peer-reviewed journals [19–27, 29–33, 35–43], and two were Master degree thesis [28, 34]. The characteristics of these studies and the ANRIL polymorphisms involved in the meta-analysis are summarized in Table 1. A summary of the total number of studies on different ANRIL SNPs is provided in Table 2.

Table 1

Characteristics of included studies and ANRIL polymorphisms for meta-analysis.

Studies (Year)	Countries Population	Variants	Samples Selection/Characteristics		NOS Score
			Cases	Controls
Zee RY2007 [19]	USWhiteCaucasians	rs10757274rs2383206	Entire IS. N = 254, age: 61.0±0.3. Men alone.	N = 254, PB, age: 60.8±0.3. Men alone.	9
Helgadottir A2008 [20]	IcelandSwedenCaucasians	rs10757278	IS (LAA and CE). N = 491. No description of age and gender.	N = 14993, PB. No description of age and gender.	5
Smith JG2009 [21]	SwedenCaucasians	rs2383207rs10757274rs1333049rs1333040	Entire IS.LSR, N = 1837, age: 73.4±12.0. F: 992(54%)MDC, N = 888, age: 62.9±6.6. F: 488(55%).	LSR, N = 947, age: 73.2±11.9. F: 540 (57%);MDC, N = 893, age: 62.9±6.6. F: 482 (54%).	8
Hu WL2009 [22]	ChinaAsians	rs10757274rs2383206	Entire IS. N = 355, age: 58.72±10.87.F: 95 (26.8%).	N = 430, HB, age: 60.4±10.91. F: 130 (30.2%).	7
Gschwendtner A2009 [23]		rs7044859rs7865618rs1537378rs2383207rs10757278			7
	Munich(Germany)Caucasians		N = 1090, age: 65.4±13.5. F: 418 (38.3%).Data of IS subtypes available.	N = 1244, PB, age: 62.4±10.9. F: 471 (37.9%).
	London(UK)Caucasians		N = 758, age: 66±13.2. F: 314 (41.4%).Data of IS subtypes available.	N = 872, PB, age: 65.3±8.8. F: 374 (42.9%).
	Baltimore(USA)MixedPopulations		N = 652, age: 41.1±7.3. F: 301 (46.2%).Data of IS subtypes available. White: 327,Black: 275, Other ethnicity: 50.	N = 718, PB, age: 39±7.1. F: 373 (51.9%).White: 384, Black: 271, other ethnicity: 63.
	Jacksonville(USA)MixedPopulations		ISGSN = 603, age: 64.6±13.8. F: 347 (57.5%).Data of IS subtypes available. White: 445,Black: 139, other ethnicity: 19.	ISGSN = 435, age: 60.7±14.9. F: 272 (62.5%).White: 314, Black: 106, other ethnicity: 15.
	Boston(USA)Mixedpopulations		N = 608, age: 65.2±15.7. F: 274 (45%).Data of IS subtypes available. White: 549,Black: 22, other ethnicity: 37.	N = 519, PB, age: 66.8±9.3. F: 270 (52%).White: 498, Black: 8, other ethnicity: 13.
	Aberdeen(UK)Caucasians		N = 607, age: 69.6±12.2. F: 273 (45%).Data of IS subtypes available.	N = 517, PB, age: 67.1±9.0. F: 252 (48.7%).
Ding H2009 [24]	ChinaAsians	rs2383206rs1004638rs10757278	N1 = 558, age: 61.0±9.8. F: 196 (35.2%).N2 = 442, age: 63.8±10.6. F: 143 (32.3%).Entire IS.	N1 = 557, age: 62.3±9.3. F: 211 (37.9%).N2 = 502, age: 62.5±8.7. F: 237 (47.2%).PB/HB. F: 448 (42.3%)	8
Yamagishi K2009 [25]	USACaucasiansAfricans	rs10757274	N = 13380 (African-Americans: 3499, Whites: 9881) for rs10757274. IS = 524 (African-Americans: 218, Whites: 306). No description of age and gender.	N = 11528 (African-Americans: 2804, Whites: 8724) for rs10757274.No description of age and gender.	7
		rs2383206	N = 12888 (African-Americans: 3399, Whites: 9489) for rs2383206. IS = 516 (African-Americans: 212, Whites: 304). No description of age and gender.	N = 11078 (African-Americans: 2719, Whites: 8359) for rs2383206. No description of age and gender.Notes: the controls is not in HWE in Whites for rs2383206.	7
Luke MM2009 [26]	AustriaCaucasians	rs10757274	Entire IS. N = 562, age: 66.0±14. F: 236 (42%).	N = 815, PB, age: 58.8±8.5. F: 418 (51.3%).	6
Olsson S2011 [27]	SwedenCaucasians	rs10965227rs1333040rs10757278rs1537378	Entire IS. N = 844, age: 56±11. F: 290(34%).	N = 668, PB, age: 56±10. F: 276 (41%).	8
Yue XY2011 [28]	ChinaAsians	rs10757274rs10757278rs2383206rs2383207rs1004638rs1333049rs1537378	Entire IS. N = 769, age: 59.91±13.11. F:257 (33.4%).	N = 682, PB, age: 59.37±11.53. F: 254 (37.2%)	8
Lin HF2011 [29]	TaiwanChinaAsians	rs1333040rs2383207rs1333049	Entire IS. N = 687, age: 64.4±12.4. F: 249(36.2%).	N = 1377, PB, age: 55.1±12.4.F: 742 (53.9%).	5
Zhang WL2012 [30]	ChinaAsians	rs10757274rs2383206rs2383207rs10757278	N(LAA) = 724, age: 61.5± 9.1. F: 263(36.3%)N(SVO) = 466, age: 61.0±8.5.F: 169(36.3%).	N = 1664, PB, age: 59.8±8.2. F: 689 (41.4%).	8
Wang C2012 [31]	ChinaAsians	rs1333040	Entire IS. N = 286, age: 60.37±7.71.Female alone.	N = 831, PB, age: 57.94±8.75.Female alone.	8
Heckman MG2013 [32]	USACaucasiansAfricans	rs1333040rs4977574rs1333042rs2383207	N = 264, age: 72±12. F: 117 (44.32%). For SWISS Caucasians, entire IS.N = 449, age: 71±15. F: 184 (40.98%). ForISGS Caucasians, entire IS.N = 166, age: 61±13. F: 84 (50.60%).For ISGS African American, entire IS.Notes: The data of ISGS was removed from last analyses because of its being from the Gschwendtner’s study in rs2383207 andrs1333040	N = 374, PB, age: 72±11. F: 169 (45.19%).For SWISS Caucasians.N = 334, PB, age: 67±15. F: 165 (49.40%).For ISGS Caucasians.N = 117, PB, age: 59±14. F: 69 (58.97%).For ISGS African American.Notes: The controls are not in HWE in Caucasians, SWISS.	6
Lovkvist H2013 [33]	SwedenCaucasiansLSRMDCSAHLSIS	rs4977574	N = 3986, age: 70. F: 1775 (44.5%).LAA, SAA and CE.	N = 2459, PB, age: 68. F: 1069 (43.5%).	7
Zhang T2014 [34]	ChinaAsians	rs10757274	LAA alone. N = 229, age: 59.36±11.15.F: 104 (45.41%).	N = 233, PB, age: 58.88±8.17. F: 113 (48.5%).	8
Lu Z2015 [35]	ChinaAsians	rs10757278rs1333049rs2383206rs1537378rs4977574rs2383207	N = 153 (Entire IS without carotid plaque),age: 56.56±7.6. F: 57 (37.25%).	N = 258, PB, age: 56.34±7.85. F: 131(50.78%).	6
Bi JJ2015 [36]	ChinaAsians	rs10757278rs1537378	LAA alone. N = 116, age: 53.74±12.32.F: 26 (22.41%).	N = 118, PB, age: 53.52±11.98. F: 33 (27.97%).	7
Cao XL2016 [37]	ChinaAsians	rs1333040rs1333042rs4977574	Entire IS. N = 569, age: 62.53±11.92.F: 173 (30.4%). Including LAA and SAA	N = 541, PB/HB, age: 6139±11.41. F: 195 (36%)	8
Akinyemi R2017 [38]	Nigeria and GhanaAfricans	rs2383207	Entire IS. N = 429, age: 61.34±12.83.F: 231 (53.85%)	N = 483, PB, age: 60.26±12.56.F: 247 (51.14%).	7
Yang JL2018 [39]	ChinaAsians	rs1333049rs2383207	Entire IS. N = 550, age: 70.10 ± 8.82.F: 244 (44.4%)	N = 550, HB, age: 69.23 ± 9.68. F: 257 (46.7%).	6
Xiong L2018 [40]	ChinaAsians	rs10757278rs1004638rs1333040rs1333049rs1537375rs1537378rs2383206rs2383207rs7044859rs7865618rs10116277rs10757269rs10757274	LAA alone. N = 200, age: 59.12±8.65.F: 77 (38.5%).	N = 205, PB, age: 56.87±7.87. F: 94 (45.85%).	8
Ferreira LE2019 [41]	BrazilCaucasians	rs2383207	LAA alone. N = 195, age: 66.9±11.6.F: 72 (36.9%).	N = 249, PB, age: 61.6±10.7. F: 138 (55.4%).	8
Han XM2020 [42]	ChinaAsians	rs10757278	Entire IS. N = 505, age: 59.9±10.9.F: 180 (35.6%)	N = 652, HB, age: 59.0±11.9. F: 253 (38.8%)	7
Wang Q2021 [43]	ChinaAsians	rs2383207rs4977574	N = 567, age: 61.72±10.17. F: 203 (35.8%).Including LAA and SVO.	N = 552, HB, age: 61.9±9.52. F: 204 (37%).	7

Notes: NOS: Newcastle-Ottawa Scale; HWE: Hardy-Weinberg equilibrium; F: female; PB: population-based; HB: hospital-based; IS: ischemic stroke; LAA: large-artery atherosclerosis; SVO and SAA: small-vessel occlusion; CE: cardioembolism; SWISS: siblings with ischemic stroke study; ISGS: ischemic stroke genetics study; LSR: Lund Stroke Register; MDC: Malmo Diet and cancer Study; SAHLSIS: Sahlgrenska Academy study on ischemic stroke.

Table 2

ANRIL SNPs included in the meta-analysis.

SNPs	Studies (n)	Cases (n)	Controls (n)	Composition of studies n (%)
				Caucasians	Asians	Afirican	Mixed populations
rs2383207	12	11,527	12,216	3(25.0%)	7(58.3%)	1(8.3%)	1(8.3%)
rs10757274	10	7,059	18,784	4(40.0%)	5(500.%)	0	1(10.0%)
rs10757278	10	9352	24552	2(20.0%)	7(70.0%)	0	1(10.0%)
rs2383206	9	4,431	8,423	2(22.0%)	6(67.0%)	1(11%)	0
rs1333040	9	6,581	8,379	4(44.0%)	5(56.0%)	0	0
rs1333049	7	5,351	6,061	2(29.0%)	5(71.0%)	0	0
rs1537378	6	6,166	6,129	1(16.0%)	4(67.0%)	0	1(16.0%)
rs4977574	5	6,083	4,593	1(20.0%)	3(60.0%)	0	1(20.0%)
rs1004638	3	1,959	1,941	0(0.0%)	3(100.0%)	0	0
rs7865618	2	4,303	4,477	1(50.0%)	0	0	1(50.0%)
rs10965227	2	1,395	1,223	1(50.0%)	1(50.0%)	0	0
rs1333042	2	1,281	1,220	1(50.0%)	1(50.0%)	0	0
rs7044859	2	4,322	4,461	1(50.0%)	0	0	1(50.0%)
rs10116277	2	512	1371	1(50.0%)	1(50.0%)	0	0
rs10757269	2	754	752	0(0.0%)	2(100.0%)	0	0

Most of the included studies had NOS scores of 7–9, four studies had NOS scores of 6 [26, 32, 35, 39], and two studies had NOS scores of 5 [20, 29].

Genetic association of 15 ANRIL SNPs with IS

SNP rs2383207

The association of rs2383207 with IS risk was investigated in 12 studies [21, 23, 28–30, 32, 35, 38–41, 43] involving 11, 527 cases and 12, 216 controls.

No significant association of rs2383207 with IS was found under three genetic models in whole studied population, sub-populations, and IS subtypes. High heterogeneity was detected in the whole studied population (AC: I2 = 82%, p <0.001; DM: I2 = 71.6%, p <0.001; RM: I2 = 74.5%, p <0.001) and in large-artery atherosclerosis (LAA) subtypes (AC: I2 = 85.7%, p<0.001; DM: I2 = 77.6%, p<0.001; RM: I2 = 76.9%, p<0.001) with all three models; however, the heterogeneity disappeared when the Caucasian studies were excluded, suggesting that ethnicity (Caucasian) may be the source of heterogeneity. Meta-regression analysis to identify different sources of heterogeneity indicated that ethnicity may be linked to heterogeneity (p = 0.085), but this finding had no statistical significance.

The sensitivity analysis excluding the poor-quality studies [29, 32, 35, 39] gave similar overall results, confirming that the results were stable and reliable. We did not find publication bias for this SNP using the funnel plots and Egger’s test (p = 0.167 in the allelic comparison model).

SNP rs10757274

Ten articles [19, 21–23, 25, 26, 28, 30, 34, 40] explored the relationship of SNP rs10757274 (7,059 cases and 18,784 controls) to IS. The G allele was found to have a significant relationship to IS risk in the whole studied population (OR = 1.11, 95%CI: 1.06–1.16, FDR-corrected p (p-FDR) <0.001) (Fig 2) and in the Caucasian studies (OR = 1.13, 95%CI: 1.06–1.20, p-FDR<0.001). The AA genotype conferred a protective effect in the whole studied population (OR = 0.90, 95%CI: 0.83–0.98, p-FDR = 0.0255).

Fig 2

Forest plot of rs10757274 allele frequency (G vs. A) associated with IS in the whole studied population.

In the IS subtype analyses, the G allele and GG genotype conferred susceptibility to LAA in the whole studied population (G allele: OR = 1.18, 95%CI: 1.08–1.30, p-FDR <0.001; GG genotype: OR = 1.31, 95%CI: 1.13–1.52, p-FDR <0.001), but mainly in Asians (G allele: OR = 1.18, 95%CI: 1.06–1.31, p-FDR = 0.003; GG genotype: OR = 1.33, 95%CI: 1.12–1.57, p-FDR = 0.003). In contrast, the AA genotype had a protective role in LAA only in the whole studied population (OR = 0.84, 95%CI = 0.73–0.96, p-FDR = 0.014). Sex had no effect in any of the comparisons.

Significant heterogeneity among studies was detected only in the recessive model (GG/(AA+AG)) in the whole studied population (I2 = 54.8%, p = 0.018) and in the Caucasians studies (I2 = 78.3%, p = 0.003). The heterogeneity disappeared in the whole studied population (I2 = 40%, p = 0.11) and in Caucasians (I2 = 47%, p = 0.15) after excluding the study by Yamagishi et al. [25]. The sensitivity analyses after removing the one study with NOS <7 [26] did not alter the final results in any of the genetic comparisons in the whole studied population or in Caucasians, further confirming the reliability of the results. No significant publication bias was detected in all three genetic models.

SNP rs10757278

The role of rs10757278 in IS was analyzed in 10 studies [20, 23, 24, 27, 28, 30, 35, 36, 40, 42] involving 9,352 cases and 2, 4552 controls. A positive association was found in the whole studied population, and in Asians and Caucasians with IS using the combined results. The G allele and GG genotype increased the susceptibility to IS in the whole studied population (G allele: OR = 1.11, 95%CI: 1.04–1.20, p-FDR = 0.006); GG genotype: OR = 1.19, 95%CI: 1.06–1.34, p-FDR = 0.006) (Figs 3 and 4), in Asians (G allele: OR = 1.16, 95%CI: 1.04–1.30, p-FDR = 0.0135; GG genotype: OR = 1.25, 95%CI: 1.07–1.48, p-FDR = 0.0135), and in Caucasians (G allele: OD = 1.12, 95%CI: 1.04–1.20, p-FDR = 0.006; GG genotype: OR = 1.18, 95%CI: 1.05–1.33, p-FDR = 0.007. The AA genotype played a protective role against IS in the whole studied population (OR = 0. 94, 95%CI: 0.88–1.00, p-FDR = 0.04), in Asians (OR = 0.91, 95%CI: 0.82–1.00, p-FDR = 0.04), and in Caucasians (OR = 0.88, 95%CI: 0.78–0.98 p-FDR = 0.021).

Fig 3

Forest plot of rs10757278 allele frequency (G vs. A) in the whole studied population.

Fig 4

Forest plot of rs10757278 genotype frequency (GG vs. (AA+GA)) in the whole studied population.

Significant heterogeneity was found in both the allelic comparison and recessive model (GG vs. (AA+GA)) in the whole studied population (AC: I2 = 62.2%, p = 0.005; RM: I2 = 57.7%, p = 0.011), but mainly in Asians (AC: I2 = 63.2%, p = 0.012; RM: I2 = 52.3%, p = 0.05), which suggested that Asians may be the source of heterogeneity.

For IS subtypes, the G allele or GG genotype increased the risk for LAA alone in the whole studied population (G allele: OR = 1.16, 95%CI: 1.01–1.33, p-FDR = 0.038; GG genotype: OR = 1.29, 95%CI: 1.15–1.45), p-FDR = 0.000), in Asians (G allele: OR = 1.28, 95%CI: 1.05–1.56, p-FDR-0.0255; GG genotype: OR = 1.44, 95%CI: 1.22–1.71), p-FDR = 0.000), and in Caucasians (G allele: OR = 1.12, 95%CI: 1.02–1.24, p-FDR = 0.04; GG genotype: OR = 1.19, 95%CI: 1.02–1.39, p-FDR = 0.04). In contrast, the AA genotype had a protective effect on LAA in the whole studied population (OR = 0.87, 95%CI: 0.78–0.98, p-FDR = 0.0375) and in Asians (OR = 0.83, 95%CI: 0.70–0.99, p-FDR = 0.042). No heterogeneity was detected in any of the comparisons for IS subtypes. Additionally, no age difference was found in the three genetic models. The sensitivity analyses excluding the low-quality studies (NOS <7) [20, 35] did not affect the stability of the original results. We found a publication bias in the allelic comparison in the whole studied population (p = 0.019, Egger’s test) (Fig 5), indicating that more studies are needed to verify the conclusion.

Fig 5

Funnel plot of rs10757278 studies in the allelic comparison in the whole studied population.

SNP rs2383206

The role of rs2383206 in IS was investigated in nine studies involving 4,431 cases and 8,423 controls) [19, 22–25, 28, 30, 35, 40]. The G allele and GG genotype increased the IS risk in the whole studied population (G allele: OR = 1.08, 95%CI: 1.02–1.14, p-FDR = 0.0075; GG genotype: OR = 1.15, 95%CI: 1.05–1.26, p-FDR = 0.0075) (Figs 6 and 7) and in Asians (G allele: OR = 1.09, 95%CI: 1.03–1.16, p-FDR = 0.015; GG genotype: OR = 1.15, 95%CI: 1.03–1.28, p-FDR = 0.015). Three studies analyzed rs2383206 in IS subtypes, and the pooled results showed that carriers with G and GG had increased risk for the LAA subtype (G allele: OR = 1.17, 95%CI: 1.06–1.29, p-FDR = 0.0015; GG genotype: OR = 1.30, 95%CI: 1.11–1.51, p = FDR = 0.0015). In contrast, the AA genotype decreased susceptibility to LAA (OR = 0.85, 95%CI: 0.73–0.99, p-FDR = 0.039). No significant association with IS was detected in the age subgroup (<45 vs. ≥45 years old). There was no heterogeneity in any of the comparisons.

Fig 6

Forest plot of rs2383206 allele frequency (G vs. A) in the whole studied population.

Fig 7

Forest plot of rs2383206 genotype frequency (GG vs. (AA+AG)) in the whole studied population.

The sensitivity analyses after excluding the poor-quality study [35]) gave similar overall results, confirming the stability of the results. There was no publication bias under the three genetic models in the whole studied population (Egger’s test for AC p = 0.978, for DM p = 0.572, for RM p = 0.569).

SNP rs1333040

The role of rs1333040 in IS was analyzed in nine studies [21, 23, 24, 27, 29, 31, 32, 37, 40] involving 6,581 cases and 8,379 controls.

The combined results showed that the TT genotype conferred increased risk (OR = 1.09, 95%CI: 1.00–1.19, p-FDR = 0.044) (Fig 8), and the C allele or CC genotype played a protective role in IS in the whole studied population (C allele: OR = 0.92, 95%CI: 0.88–0.97, p-FDR = 0.003; CC genotype: OR = 0.83, 95%CI: 0.73–0.94, p-FDR = 0.006). In contrast, in the sub-population analyses, the C allele showed a protective effect on IS, but only in in Caucasians (OR = 0.92, 95%CI: 0.86–0.98, p-FDR = 0.018).

Fig 8

Forest plot of rs1333040 genotype frequency (TT vs. (CC+CT)) in the whole studied population.

No significant relationship of rs1333040 with LAA was found in the whole studied population; however, an association with LAA risk was found in Caucasians. Patients with the C allele and CC genotype had a lower possibility of developing LAA (C allele: OR = 0.86, 95%CI: 0.76–0.96, p-FDR = 0.03; CC genotype: OR = 0.78, 95%CI: 0.63–0.98, p-FDR = 0.037). In contrast, patents with the TT genotype seemed to be more predisposed to LAA risk (OR = 1.20, 95%CI: 1.01,1.42, P-FDR = 0.037). No sex difference was found for IS in any of the comparisons. There was no significant heterogeneity among the studies.

The sensitivity analyses after excluding low-quality studies (NOS <7) [29, 32] did not alter the final results. No publication bias was detected in the three genetic models in the whole studied population (Egger’s test for AC p = 0.772, for DM p = 0.502, for RM p = 0.875).

SNP rs1333049

The role of rs1333049 in IS was analyzed in seven studies involving 5,351 cases and 6,061 controls [21, 23, 28, 29, 35, 39, 40]. Pooled analyses showed that the C allele increased the susceptibility to IS (OR = 1.09, 95%CI: 1.03–1.15, p-FDR = 0.009) in the whole studied population (Fig 9) and in Caucasians (OR = 1.15, 95%CI: 1.06–1.24, p-FDR = 0.001). No significant association was found in Asians, LAA subtype, or age subgroup (<45 vs. ≥45 years old). No heterogeneity was detected in any of the genetic comparisons.

Fig 9

Forest plot of rs1333049 allele frequency (C vs. G) in the whole studied population.

The sensitivity analyses after removing low-quality studies (NOS <7) [29, 35, 39] remained unchanged in the three models in the whole studied population. No publication bias was found in three genetic models (Egger’s test for AC p = 0.845, for DM p = 0.854, for RM p = 0.187).

SNP rs1537378

The role of rs1537378 in IS was analyzed in six studies [23, 27, 28, 35, 36, 40] involving 6,166 cases and 6,129 controls. The CC genotype was found to increase the risk for IS in the whole studied population (OR = 1.18, 95%CI: 1.09–1.27, p-FDR = 0.000) (Fig 10), in Asians (OR = 1.43, 95%CI: .1.20–1.71, p-FDR = 0.000), and in Caucasians (OR = 1.21, 95%CI: 1.07–1.37, p-FDR = 0.003). In contrast, the T allele and TT genotype had a protective effect on IS in the whole studied population (T allele: OR = 0.80, 95%CI: 0.70–0.92, p-FDR = 0.001; TT genotype: OR = 0.83, 95%CI: 0.74–0.93, p-FDR = 0.001), in Asians (T allele: OR = 0.70, 95%CI: 0.60–0.82, p-FDR = 0.000; TT genotype: OR = 0.49, 95%CI: 0.30–0.80, p-FDR = 0.005), and in Caucasians (T allele: OR = 0.85, 95%CI: 0.78–0.93, p-FDR = 0.000; TT genotype: OR = 0.79, 95%CI; 0.67–0.93, p-FDR = 0.006).

Fig 10

Forest plot of rs1537378 genotype frequency (CC vs. (CT+TT)) and susceptibility to all types of IS in the whole studied population.

In the IS subtype analyses, a significant relationship was found only in LAA. The LAA risk was higher in carriers with the CC genotype, and patients carrying the T allele and TT genotype had lower risk for LAA in the whole studied population, in Asians, and in Caucasians. In patients who were ≥45 years old, the CC genotype was also associated with higher risk for all types of IS, and only T allele had a protective role.

Significant heterogeneity among studies was found in the T allele (T/C) and CC genotype comparisons (CC vs. (CT+TT)) only in the whole studied population. The heterogeneity disappeared after removing the study by Bi et al. [36], which suggested it may be a source of heterogeneity; however, the final results remained unchanged. The sensitivity analyses after excluding the study with NOS = 6 [35] did not alter any of the results, indicating the reliability and stability of the original results. The funnel plot was asymmetric in all three genetic comparisons in the whole studied population (Egger’s test for AC p = 0.019; for DM p = 0.033; for RM p = 0.046) (Fig 11), which suggested there might be some publication bias. The trim and fill method was used to identify and correct the bias, and the combined effect was found to be unchanged, indicating that the possible publication bias had little effect on the results.

Fig 11

Funnel plot of rs1537378 allele frequency (T vs. C) and IS susceptibility in the whole studied population.

SNP rs4977574

The role of rs4977574 in IS was analyzed in five studies [32, 33, 35, 37, 43] involving 6,083 cases and 4,593 controls that included three Asian, one Caucasian, and one mixed populations.

The pooled results indicated that rs4977574 was strongly associated with IS. It was found that The G allele and GG genotype conferred susceptibility IS in the whole studied population (G allele: OR = 1.11, 95%CI: 1.05–1.17, p-FDR = 0.000; GG genotype: OR = 1.13, 95%CI: 1.03–1.24, p-FDR = 0.011) (Figs 12 and 13). In contrast, the AA genotype decreased the risk of IS in the whole studied population (OR = 0.86, 95%CI: 0.79–0.94, p-FDR = 0.0015).

Fig 12

Forest plot of rs4977574 allelic frequency (G vs. A) and IS susceptibility in the whole studied population.

Fig 13

Forest plot of rs4977574 genotype frequency (GG vs. (GA+AA)) and IS susceptibility in the whole studied population.

The G allele and AA genotype had significant association with IS risk only in Asians (G allele: OR = 1.20, 95%CI: 1.06–1.36, p-FDR = 0.0004; AA genotype: OR = 0.75, 95%CI: 0.63–0.89, p-FDR = 0.0015). Significant heterogeneity was found only in the allelic comparison model (I2 = 64.1%, p = 0.062) in Caucasians; however, the heterogeneity disappeared (I2 = 0%, p = 0.714) after removing the study by Lovkvist et al. [33].

The IS subtype analysis showed that the G allele and GG genotype were risk factors for LAA in the whole studied population (G allele: OR = 1.22, 95%CI:1.09–1.37, p-FDR = 0.003; GG genotype: OR = 1.26, 95%CI:1.05–1.52, p-FDR-0.015) and in Caucasians (G allele: OR = 1.26, 95%CI:1.09–1.46, p-FDR = 0.006; GG genotype: OR = 1.35, 95%CI:1.07–1.71, p-FDR = 0.014). In contrast, the AA genotype had a protective role in the whole studied population (OR = 0.75, 95%CI:0.62–0.90, p-FDR = 0.003) and in Caucasians (OR = 0.74, 95%CI:0.58–0.94, p-FDR = 0.014). Heterogeneity was detected in the small-vessel occlusion and cardioembolism subtypes; however, the source of the heterogeneity was not analyzed because of the small number of included studies.

The sensitivity analysis after omitting two poor-quality studies [32, 35] showed that the final pooled results were not affected. No publication bias was detected by the funnel plots or Egger’s test in the three genetic models in the whole studied population.

SNP rs1004638

The role of rs1004638 in IS was analyzed in three studies [24, 28, 40] comprising only subjects with 1,959 cases and 1,941 controls. Significant associations of this SNP with IS were found in all genetic comparisons (AC: OR = 1.15, 95%CI: 1.04–1.26, p-FDR = 0.015; RM: OR = 1.21, 95%CI: 1.03–1.43, p-FDR = 0.024; DM: OR = 0.85, 95%CI: 0.73–0.98, p-FDR = 0.024) without no heterogeneity among the studies. The A allele and AA genotype increased susceptibility to IS, whereas the TT genotype had a protective role. Sensitivity analysis and publication bias were not performed because of the small number of included studies.

Other SNPs

For each of the remaining six SNPs, rs7865618, rs10965227, rs1333042, rs7044859, rs10116277, and rs10757269, only two studies with from 512 to 4,322 cases and from 752 to 4,477 controls, were included for meta-analyses. No significant association was found in any of the comparisons. Heterogeneity between studies, sensitivity analysis, and publication bias were not explored because of the small number of studies for each SNP.

Discussion

The meta-analysis results showed that eight SNPs (rs10757274, rs10757278, rs2383206, rs1333040, rs1333049, rs1537378, rs4977574, and rs1004638) in ANRIL were significantly associated with IS risk, and six of these SNPs (rs10757274, rs10757278, rs2383206, rs1333040, rs1537378, and rs4977574) were also found to be related to the LAA subtype of IS. Two of the SNPs (rs2383206 and rs4977574) were associated with IS mainly in Asians, and three SNPs (rs10757274, rs1333040, and rs1333049) were associated with susceptibility to IS mainly in Caucasians.

The locus close to a cluster of cell-cycle regulating genes in chromosome 9p21, such as CDKN2A and CDKN2B, regulates vascular remodeling pathways. The proteins encoded by these genes affect cell-cycle progression, resulting in an antiproliferative effect on arterial smooth muscle. In human white blood cells, the homozygous carriers of the 9p21 risk allele are associated with down-regulation of CDKN2B expression and up-regulation of genes involved in cellular proliferation. Markedly decreased expression of CDKN2A and CDKN2B was reported in mutant mice and doubling of the proliferative capacity of mutant aortic smooth muscle cells in culture was detected, a cellular phenotype relevant to atherosclerosis [44].

ANRIL encodes a large antisense long non-coding RNA in which the first exon is located in the CDKN2A promoter and overlaps with the two exons of CDKN2B. Expression of ANRIL co-clustered mainly with p14/ARF under both physiologic and pathologic conditions. The 9p21 region may promote atherosclerosis by regulating the expression of ANRIL, which in turn is associated with altered expression of genes that control cellular proliferation pathways [9].

ANRIL was recently shown to be expressed in human atheromatous vessels, including both abdominal aortic aneurysm and carotid endarterectomy samples, as well as in isolated vascular endothelial cells, monocyte-derived macrophages, and coronary smooth muscle cells. Moreover, ANRIL expression was significantly associated with the alteration in function of vascular endothelial cells and vascular smooth muscle cells in both human or animal models [45]. Together, these findings indicate that ANRIL has a direct effect on the pathobiology of atherosclerosis. Therefore, ANRIL is considered a good candidate for atherosclerotic disease risk, such as coronary artery disease (CAD) and IS [46, 47].

Studies have shown that different ANRIL transcripts exhibit disease-specific expression patterns in CAD, which further supports the hypothesis that ANRIL is the causative gene at the 9p21 CAD susceptibility locus [48]. Recently, a few meta-analyses using SNPs also indicated a significant association of ANRIL with CAD [49-55]. IS is known to share common pathophysiological mechanisms with CAD, and CAD and IS seem to have common susceptibility locus. A comprehensive review indicated that increased ANRIL expression was associated with IS risk in animal models by promoting angiogenesis and regulating inflammation [56], and patients with IS were also found to have significantly higher serum ANRIL levels in clinical practice [57, 58].

Some studies have explored the functional effect of SNPs in ANRIL. The rs1333049 risk allele (C allele) was found to influence ANRIL expression levels in vascular smooth muscle cells, which was associated with elevated levels of these cells in atherosclerosis plaques involved in the pathogenesis of atherosclerosis [59]. Rs1333040 is located in an intronic enhancer region that was found to influence the activity of the enhancer and ANRIL expression. Rs10757274 showed high linkage disequilibrium with myocardial infraction-associated SNPs, including rs1537373, rs4977575, and rs10757272, and contributed to the activation or inhibition of the expression of the related genes [55]. A few SNPs were found to have a significant relationship to vascular risk factors. Patients carrying mutant alleles of rs1333049 and rs4977574 had elevated total cholesterol, triglycerides, and low-density lipoprotein cholesterol levels [60-62]. The risk allele of rs4977574 was also found to be related to carotid plaque formation in patients with acute IS [63] or type 2 diabetes [64]. All of these factors may lead to the progression of atherosclerotic vascular diseases or IS.

A few meta-analyses have reported the association of ANRIL with IS; however, these meta-analyses have some limitations, such as failure to include all eligible studies [34, 43, 52, 65–68], no comprehensive analyses [66-68], confounding cases (patients with transient ischemic attack or other types of stroke were included in the IS samples) [34, 52, 65, 67], as well as wrong SNP loci [65] or errors in extracting and analyzing data [34, 65, 67], which could have influenced the overall results. Two previous genome-wide association studies (GWAS) [69, 70] explored the relationship of ANRIL SNPs and IS in a Caucasian cohort with European ancestry, but only one SNP (rs2383207) was found to be association with LAA. Ethnicity may partly explain the discrepancy between the GWAS results and the results of the present meta-analysis, which included more Asians.

The potential biological mechanisms, including how ANRIL is strongly associated with the risk for cardio-metabolic diseases, are still unknown. Recent reports have found that the N4-acetylcytidine modification of RNA, which regulated gene expression, and microRNA-mediated gene expression and immuno-deficiency in the gut microbiome, were key to cardiometabolic diseases, including IS [71-78]. However, the few studies that have investigated the role of ANRIL SNP loci in the N4-acetylcytidine regulatory pathway failed to find definite effects of RNA modification or immuno-deficiency on the development of IS.

Our meta-analysis has some limitations. Firstly, there is language bias because we only searched studies of ANRIL polymorphisms on IS reported in Chinese and English, and therefore may have missed studies published in other languages. Secondly, the number of studies included in this meta-analysis was moderate, and seven of the SNPs (rs1004638, rs7865618, rs10965227, rs1333042, rs7044859, rs10116277, and rs10757269) were involved in three or less studies. Therefore, some results could be influenced by random error and/or publication bias. Thirdly, the presence of potential confounders between studies or between cases and controls within each study, such as age, sex, or ethnic admixture, were unadjusted that may have influenced the results. Fourthly, it is well known that it is very important to conduct causal inference analysis to determine if the associated genetic polymorphisms are causally triggering the development of IS by mediating the expression of this gene in specific tissues [79-82]. Although, this meta-analysis aimed to discuss the association of ANRIL with IS using SNPs as genetic marker, no causal genetic effects of ANRIL on IS can be established. Fifthly, machine learning is considered a useful tool for the classification and prediction of diseases based on biomarkers [83-86] that we have yet to use to analyze the role of ANRIL in susceptibility to IS. Sixthly, GWAS, case-only studies, and family-based studies were not included because of differences in study design, but they could be useful for meta-analysis in the future. Finally, the inter-study heterogeneity in the pooled analyses may have affected the results for several SNPs.

In summary, our accumulated pooled analyses indicate that ANRIL has a significant association with IS risk in Asian populations. The causal effects of the ANRIL SNPs associated with IS can be explored by Mendelian randomization analysis in the future.

PRISMA 2009 checklist used in this meta-analysis.

(DOCX)

Click here for additional data file.

Meta-analysis on genetic association studies checklist.

(DOCX)

Click here for additional data file.

The excluded articles and the reasons for exclusion of each article.

(DOCX)

Click here for additional data file.

13 Sep 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Mingqing Xu

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

Additional Editor Comments:

The title should have a clear, precise scientific meaning and should not contain a colon. Where possible, the title should be read as one concise sentence. Please re-write the title ensuring that it is informative and appropriate.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In the manuscript entitled “The Research on Association of the ANRIL Gene with Ischemic Stroke: the Evidence From a Comprehensive Meta-Analysis”, the authors explored whether f the All the data used in the published GWASA papers should be included, if cannot, please state the reasons.

The following papers can be cited and followed for the meta-analytic procedures to improve the quality. (if the data is not enough available, at least DISCUSSION should be added as the LIMITATION of this study with enough citation to support the viewpoints):

Ref 1: Wu Y, et al. Multi-trait analysis for genome-wide association study of five psychiatric disorders. Transl Psychiatry. 2020 Jun 30;10(1):209.

Ref 2: Jiang L, et al. Sex-Specific Association of Circulating Ferritin Level and Risk of Type 2 Diabetes: A Dose-Response Meta-Analysis of Prospective Studies..J Clin Endocrinol Metab. 2019 Oct 1;104(10):4539-4551.

Ref 3: Xu M, Lin Z. Genetic influences of dopamine transport gene on alcohol dependence: a pooled analysis of 13 studies with 2483 cases and 1753 controls.Prog Neuropsychopharmacol Biol Psychiatry. 2011 Jul 1;35(5):1255-60.

Ref 4: Xu M, Sham P, Ye Z, Lindpaintner K, He L. A1166C genetic variation of the angiotensin II type I receptor gene and susceptibility to coronary heart disease: collaborative of 53 studies with 20,435 cases and 23,674 controls.Atherosclerosis. 2010 Nov;213(1):191-9.

Ref 5: Xu M, et al. Quantitative assessment of the effect of angiotensinogen gene polymorphisms on the risk of coronary heart disease. Circulation. 2007 Sep 18;116(12):1356-66

Trans-ethnitic meta-analysis can be referred to Ref 1.

Subgroup analyses based on sex, age, race, gene dosage can be referred to References 2. and Quality assessment score can also to referred to Refences 3-5.

I am wondering if the authors may integrate genotype data with eQTL from GTEX or pQTLs is to explore the causality of the genetic variant in the development of stroke. But I strongly suggest to do causal inference analysis to see if the associated Genetic Polymorphisms in this gene are causally triggering the development of stroke through mediating the expression of this gene in specific tissues. If cannot, please discuss the limitations in the Discussion in detail with additional citations to support the viewpoints. For these reasons, the following papers regarding causal inference between genetic varients,inter-mediator phenotype and disease outcome can be cited and followed.

Reference 1: Fuquan Zhang, Ancha Baranova, Chao Zhou, et al. Causal influences of neuroticism on mental health and cardiovascular disease. Human Genetics. 2021 May 1

Reference 2:Fuquan Zhang, et al. Genetic evidence suggests posttraumatic stress disorder as a subtype of major depressive disorder. Journal of Clinical Investigation. 2021 May 30

Reference 3:Wang, X. et al. Genetic support of a causal relationship between iron status and type 2 diabetes: a Mendelian randomization study. The Journal of clinical endocrinology and metabolism, doi:10.1210/clinem/dgab454 (2021).

I am not sure if the genetic polymorphism can be used for predict stroke based on a machine learning model. In the PRECISION MEDICINE era, deep learning or machine learning is a hot topic in classification and prediction of diseases based on biomarkers. The authors may discuss the possibility to use the genetic variants related to stroke for the prediction or early diagnosis of stroke. The authors may cite the following papers for discussion or follow the analytic procedures to construct machine learning prediction models.

Reference 1:Yu H, et al. LEPR hypomethylation is significantly associated with gastric cancer in males.Exp Mol Pathol. 2020 Oct;116:104493.

Reference 2:Liu M, et al. A multi-model deep convolutional neural network for automatic hippocampus segmentation and classification in Alzheimer's disease.Neuroimage. 2020 Mar;208:11645

I suggest one paragraph in the DISCUSSION section to elucidate the potential biological regulation mechanisms regarding how the genetic variant to affect the stroke outcome. The following papers clearly disclosed some genes indications whose abnormal expressions are mediated through mRNA modifications. The recent progress in N4-Acetylcytidine on RNA expression is also playing key role on the human diseases. I suggest the authors discussing this mRNA modifications/ N4-Acetylcytidine with their findings in the DISCUSSION section because the knowledge needs to be updated.

Reference1: Jin G, et al. The Processing, Gene Regulation, Biological Functions, and Clinical Relevance of N4-Acetylcytidine on RNA: A Systematic Review. Mol Ther Nucleic Acids. 2020. PMID: 32171170

Reviewer #2: 1. I am not sure if the 15 SNPs in ANRIL Gene are tagSNPs or not. Only tagSNPs that are unrelated with each other are informative, and it is better that the tagSNPs may cover this whole gene region.

2. Multiple test correction should be conducted .

3. Statiscial power should be calculaed with proper method to ensure less type 2 errors.

3. subgroup analysis based on sex, age should be conducted.

4. I would suggest collecting published GWAS data to redo the meta-analysis if it is possible.

5. For the genetic varients with significant association with the phenotype, deep discussion about the potential biological mechanisms involved in the phenotype is need. Especially, you had better see if this kind of SNP is causal or not.. Especially the potential causal effects for the SNPs with strong association signals should be explored.

All the figures are not clear and need to be well tailored for publishing.

The language should be polished further.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

14 Jan 2022

Dear editors,

Thank you for these valuable comments, we have had a detailed answer to some problems. However, we had some difficult in dealing with other suggestions the reviewers mede. We hope that these faults will not affect the final decision the journal made. Please see this as follows:

The title should have a clear, precise scientific meaning and should not contain a colon. Where possible, the title should be read as one concise sentence. Please re-write the title ensuring that it is informative and appropriate.

Answer: the new title: Genetic association of ANRIL with susceptibility to Ischemic Stroke: a comprehensive meta-analysis

Reviewer #1:

1. I am not sure if the 15 SNPs in ANRIL Gene are tagSNPs or not. Only tagSNPs that are unrelated with each other are informative, and it is better that the tagSNPs may cover this whole gene region.

Answer: the objectives of the manuscript was all SNPs in ANRIL gene involving relationship to ischemic stroke, and aimed to investigate the role of ANRIL gene using SNPs as genetic marker in ischemic stroke risk. The SNPs were included in the meta-analysis only based on the SNP locus being studied in at least two articles. And so, it is not important whether the SNPs are tagSNPs.

2. Multiple test correction should be conducted .

Answer: We have discussed the focus in “Statistical analyses” section: “The false discovery rate (FDR) method by Benjamini–Hochberg was used for multiple testing correction”.

3. Statiscial power should be calculaed with proper method to ensure less type 2 errors.

Answer: In this meta-analysis, pooled ORs with corresponding 95% CI were calculated by the fixed-effects or the random-effects model based on three genetic models. This is a most common methods, and it is not important to calculate the statistical power again.

3. subgroup analysis based on sex, age should be conducted.

Answer: we have clearly described the proplem. Please see “Sub-population analyses were conducted for ethnicity, and meanwhile, subgroup analyses for IS subtype, age or gender (if available) were also performed” paragraph in “Statistical analyses” section.

4. I would suggest collecting published GWAS data to redo the meta-analysis if it is possible.

Answer: we only studied the SNPs in ANRIL gene in current manuscript. The GWAS data, case-only studies and family-based studies were not included because of different study design, as well as different analysis methods might be needed in meta-analysis based on GWAS data.

5. For the genetic varients with significant association with the phenotype, deep discussion about the potential biological mechanisms involved in the phenotype is need. Especially, you had better see if this kind of SNP is causal or not.. Especially the potential causal effects for the SNPs with strong association signals should be explored.

Answer: we have tried our best to make deep discussion on the potential biological mechanisms of these risk SNPs loci, however, some biological mechanisms for a few SNPs still is not available due to too few articles involving the loci.

6, All the figures are not clear and need to be well tailored for publishing.

Answer: we have revised the figures again.

7,The language should be polished further.

Answer: we have reviewed the language problems with the help of native English speaker.

Reviewer #2:

In the manuscript entitled “The Research on Association of the ANRIL Gene with Ischemic Stroke: the Evidence From a Comprehensive Meta-Analysis”, the authors explored whether f the All the data used in the published GWASA papers should be included, if cannot, please state the reasons.

The following papers can be cited and followed for the meta-analytic procedures to improve the quality. (if the data is not enough available, at least DISCUSSION should be added as the LIMITATION of this study with enough citation to support the viewpoints):

I am not sure if the genetic polymorphism can be used for predict stroke based on a machine learning model. In the PRECISION MEDICINE era, deep learning or machine learning is a hot topic in classification and prediction of diseases based on biomarkers. The authors may discuss the possibility to use the genetic variants related to stroke for the prediction or early diagnosis of stroke. The authors may cite the following papers for discussion or follow the analytic procedures to construct machine learning prediction models.

I suggest one paragraph in the DISCUSSION section to elucidate the potential biological regulation mechanisms regarding how the genetic variant to affect the stroke outcome. The following papers clearly disclosed some genes indications whose abnormal expressions are mediated through mRNA modifications. The recent progress in N4-Acetylcytidine on RNA expression is also playing key role on the human diseases. I suggest the authors discussing this mRNA modifications/ N4-Acetylcytidine with their findings in the DISCUSSION section because the knowledge needs to be updated.

Answer: the manuscript aimed to discuss the relationship of the ANRIL gene using SNPs as genetic marker in Ischemic stroke employing the meta analyses methods. The conclusions might be a association of the targeted gene with Ischemic stroke being detected or not, however, the results did not make us find some causal genes.

The GWAS data, case-only studies and family-based studies were not included because of different study design, as well as different analysis methods might be needed in meta-analysis based on GWAS data.

we have tried our best to make deep discussion on the potential biological mechanisms of these risk SNPs loci, however, some biological mechanisms for a few SNPs still is not available due to too few articles involving the loci.

Regards,

Hua Liu

Submitted filename: the second reply 1.14.docx

Click here for additional data file.

20 Jan 2022

Genetic association of ANRIL with susceptibility to Ischemic Stroke: a comprehensive meta-analysis

PONE-D-21-27176R1

Dear Dr. Liu,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Mingqing Xu

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

This paper can be accepted for publication now.

Reviewers' comments:

28 Feb 2022

PONE-D-21-27176R1

Genetic association of ANRIL with susceptibility to Ischemic Stroke: a comprehensive meta-analysis

Dear Dr. Liu:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Mingqing Xu

Academic Editor

PLOS ONE