Increased prevalence of the CVD-associated ANRIL allele in the Roma/Gypsy population in comparison with the majority Czech population

Abstract Cardiovascular disease (CVD) is a major cause of death around the world, with highest prevalence reported in minority Roma/Gypsy populations living in developed countries. Whether these differences are caused by unhealthy lifestyles or genetic factors remain unknown. The aim of our study was to examine the genotype frequencies of the rs10757274 polymorphism in the 9p.21 locus within ANRIL (antisense non-coding RNA in the INK4 locus), a long non-coding RNA located in the vicinity of the CDKN2A/2B inhibitors loci. ANRIL is understood to be the strongest genetic determinant of CVD in Caucasians. Using PCR-RFLP, we analysed the ANRIL rs10757274 polymorphism in 298 non-Roma (50% male) and 302 Roma/Gypsy (50% male) adult (39.5 ± 15.1 years and 39.2 ± 12.8 years, respectively) subjects. We found that frequencies of the ANRIL GG, GA and AA genotypes were 20.1%, 52.4% and 27.5% in the majority population and 32.9%, 47.9% and 19.2% in Roma/Gypsy subjects, respectively. The distribution of genotypes was deemed significantly different at P < 0.001. Within the Roma/Gypsy population, we detected increased prevalence of the CVD-associated GG genotype. Increased prevalence of CVD among Roma/Gypsies subjects may be significantly linked to genetic background.

Cardiovascular diseases and myocardial infarction (CVD, MI) are the major causes of death in industrial countries. It is widely acknowledged that, next to environmental and lifestyle risk factors (Bhatnagar, 2017), genetic predisposition plays an important role in CVD determination (Dainis and Ashley, 2018).
Although previous publications on this topic are sparse and typically characterised by low numbers of examined subjects and non-representative selections, they report that prevalence of CVD risk factors in Roma/Gypsy communities is higher than in majority populations (Dobranici et al., 2012). A Czech national survey (Antosova et al., 2014) reported that prevalence of heart disease in the Roma/Gypsy population living in the Czech Republic was more than double that of the majority population. The question remains whether increased prevalence of CVD is caused not just by unhealthy lifestyles but also by genetic factors, as genetic differences between Roma/Gypsy and non-Roma populations have been widely documented (Mydlárová Blaščáková et al., 2017;Hubacek et al., 2017Hubacek et al., , 2020Nagy et al., 2017;Dlouhá et al., 2020).
A number of genome-wide association studies (GWAS) have detected a comprehensive list of common polymorphisms associated with increased risk of MI (for a summary, see Erdmann et al., 2018). Among the strongest are the single-nucleotide polymorphisms (rs10757274, rs10965215, rs1333040 and rs4977574 are among the most commonly studied) within ANRIL (antisense non-coding RNA in the INK4 locus; OMIM acc. N. 613149). A long non-coding regulatory RNA located in the 9p.21 locus (Helgadottir et al., 2007), ANRIL is clustered within the cyclin-dependent kinase inhibitors CDKN2A and CDKN2B. ANRIL is understood to regulate -through a complex mechanism -a number of processes implicated in endothelial injury and subsequent development of atherosclerosis. That ANRIL is expressed in almost all human tissue highlights the extraordinary importance of this regulatory RNA.
At-risk ANRIL alleles are associated with an approximate 30-35% increased risk of MI per each allele (Palomaki et al., 2010). Interestingly, despite the homogeneous and strong association between ANRIL SNPs and CVD risk among different ethnicities (Hu et al., 2019), identical variants are not typically associated with traditional CVD risk factors such as smoking, dyslipidaemia, obesity and hypertension (with the exception of diabetes). Nevertheless, even after adjusting for diabetes, the highly significant association between ANRIL polymorphisms and CVD indicates the independent influence of this locus on these two major non-communicable diseases.
Two groups of adult populations (the non-Roma Czech majority and the Roma/Gypsy ethnic minority) inhabiting a Czech region South Bohemia (Table 1; Hubacek et al., 2020) have been included in the study. Snowball sampling (Hughes et al., 1995) has been used to the Czech Roma/ Gypsy subpopulation (N = 302) selection and quota sampling (Walter, 1989) for recruiting Czech Caucasians/Slavs (N = 298). All subjects were older than 18 years at the time of data and samples collection and theirs ethnicity was based on self-reported information (Adámková et al., 2015, Hubáček et al., 2018.
Ethics committee at South-Bohemia University approved the study protocol and all participants signed the informed consent with participation in the study.
DNA were isolated ("Xtreme DNA Isolation Kit"), from buccal cells collected through "DNA buccal swabs" (both Isohelix, Cell Projects Ltd, UK) according to manufacturer conditions.
The ANRIL rs10757274 genotype was analysed using the polymerase chain reaction-restriction fragment length polymorphism method, as described in detail elsewhere (Hubacek et al., 2016). Restriction fragments were separated using a 10% polyacrylamide gel. Fermentas International Inc. (Burlington, Ontario, Canada) provided all PCR chemicals, with PCRs performed on the MJ Research DYAD Disciple PCR device.
Deviance of genotype frequencies among the groups examined was analysed using the Hardy-Weinberg equilibrium principle (Court, 2005(Court, -2008. Differences in allele and genotype frequencies were compared using an online chisquare test (Stangroom). Comparisons were performed for dominant, co-dominant and recessive models.
The call rates achieved for the ANRIL rs10757274 SNP were 96.6% in the Czech majority population and 96.7% in the minority population. No significant gender differences in genotype frequencies were observed in either ethnic group (not presented in detail). Distributions of genotypes were in agreement with the Hardy-Weinberg equilibrium (P = 0.36 for the majority population and P = 0.70 for the minority population).
Within the minority Roma/Gypsy population, there were significantly more carriers of the CVD-associated GG genotype (P < 0.001; for further details and comparisons, see Table 2). Consequently, prevalence of the G allele was also significantly higher (56.8% vs. 46.4%; P < 0.0005).
Despite the relatively low numbers of subjects enrolled in this study, the genotype frequencies observed for the majority population are in agreement with the frequency (22% of ANRIL rs10757274 GG homozygosity) reported in our previous study (Hubacek et al., 2016), as well as in neighbouring German population (Scheffold et al., 2011). In contrast, detected frequencies of the G allele in the minority Roma/Gypsy population (57%); are among the highest in the world and even higher than in the Indian population (Kumar et al., 2011), which is understood to be the origin of the settled Roma/Gypsy populations in Europe (Mendizabal et al., 2011).
Although examining a sole SNP is evidently insufficient at drawing major conclusions, we speculate that the increased prevalence of at-risk ANRIL genotypes in the Roma/Gypsy population could be an important modifier associated with higher prevalence of CVD in this ethnic group. To extend our findings, a further case-control study involving CVD-focused tagging of ANRIL SNPs within this ethnic group is required.
We conclude that CVD-associated ANRIL genotypes are more common in the Roma/Gypsy population compared to the majority Caucasian population.

Acknowledgments
The study was supported by Ministry of Health of the Czech Republic, grant nr. NV18-01-00046, all rights reserved; projects number LD14114 implemented under the financial support of the Ministry of Education, Youth and Sports within COST (Cooperation On Scientific and Technical Research), entitled "Obesity and overweight in the Romany minority in the Region of South Bohemia", and by the project supported by Ministry of Health Czech Republic -conceptual development of research organisation ("Institute for Clinical and Experimental Medicine -IKEM, IN 00023001").

Conflicts of interest
The authors declare no conflicts of interest.

Author Contributions
JAH conceived the original idea, collected the funding, drafted the manuscript, performed genetical analyses and analysed data; LŠ collected the data, supervised the examinations; VO collected the data, supervised the examinations; VA conceived the original idea, supervised the examinations and the project; VT conceived the original idea, collected the funding, supervised the project, analysed the data. All authors contributed to the interpretation of the results, discussed the results, read and corrected the final version of the manuscript.