Distribution of Six Polymorphisms in Two Communities with a Historical High Incidence of Diabetes and Obesity in Yucatan, Mexico

Aims: Maya ancestry populations from Yucatan have exhibited a high prevalence of diabetes and obesity; consequently, the aim of this study was to determine the allelic and genotype frequencies of six polymorphisms associated with diabetes and obesity in two Maya populations. Methodology: Healthy individuals with Maya ancestry were recruited in small rural and urban communities from Yucatan. Six polymorphisms present in five genes ( PPARGC1A , NRF1 , SLC30A8 , ADRA2A and UCP3 ) were genotyped using TaqMan assays. Linkage disequilibrium analysis was performed for rs13266634 and rs11558471 ( SLC30A8 ). Results: The observed frequencies in the small rural community (SRC) and Merida were in Hardy Weinberg equilibrium. Frequencies of five polymorphisms (rs8192678, rs1882095, rs13266634, rs11558471 and rs3781907) correlated with 1000 genomes project data, furthermore, statistical analysis did not reveal a significant difference between genotype frequencies of the SRC and Merida populations. Contrary, frequencies from Mexicans living in Los Angeles compared with frequencies obtained in Yucatan (SRC and Merida) indicated significant difference on genotype frequencies in the ADRA2A gene (P-value= .017). The polymorphisms rs13266634 and rs11558471 in the SLC30A8 gene displayed strong linkage disequilibrium (D’= 0.96), displaying frequencies of 0.725, 0.255, and 0.015 for haplotypes C-A , T-G , and C-G respectively. Conclusion: Distribution of A allele in rs553668 ( ADRA2A ) in the Yucatan populations was higher than the frequency reported for Mexicans from LA, Americans, Europeans, and Africans. This finding could be related to blood pressure levels in the Maya populations. Additionally, a high frequency of C-A haplotype (rs13266634 and rs11558471) in the SLC30A8 gene could be associated with an increased risk of diabetes and obesity in these populations.


INTRODUCTION
The state of Yucatan is known for its Maya heritage, with the ancient Pre-Hispanic cultures having settled in this region more than 3,500 years ago. Past studies have emphasized certain physical and genetic characteristics of Maya descendants, which are thought are playing a major role in their physical constitution and risk of metabolic disease [1][2][3][4]. Currently, Yucatan is in the top five states in the nation with the highest rates of diabetes and hypertension. The National Health and Nutrition survey of 2012 registered 7,210 and 7,070 cases of diabetes and hypertension, respectively, in Yucatan [5,6]. Projections for the next 25 years estimate an increase from 44.3 to 60.5 million cases of Type 2 diabetes (T2D) [7]. Reports from 2011 show that the prevalence of T2D in Yucatan is 8.2-9.2% [5]. Despite this modest average, communities in the northeast coast of Yucatan display a higher prevalence of T2D [3,8]. Specifically, there is a case of a small rural community (SRC) with less than 5,000 habitants where the prevalence of T2D rises up to 22% (unpublished data). Additionally, this population exhibits a strong Maya ancestry, as well as a high prevalence of obesity [3,8].

Function of Genes Involved in T2D of this Study
Peroxisome proliferator-activated receptor gamma coactivator 1-alpha encoded by PPARGC1A gene is a transcriptional coactivator of nuclear receptors and other transcription factors that regulate metabolic processes, including mitochondrial biogenesis and respiration, and hepatic gluconeogenesis [22,23]. This protein may be also involved in controlling blood pressure, regulating cellular cholesterol homeostasis, and obesity development [24]. The rs8192678 variant has been associated with diabetes related traits such as insulin resistance [ Mitochondrial uncoupling proteins encoded by the UCP3 gene are members of the larger family of mitochondrial anion carrier proteins (MACP). It is primarily expressed in skeletal muscle [22]. UCPs facilitate the transfer of anions from the inner to the outer mitochondrial membrane and the return transfer of protons from the outer to the inner mitochondrial membrane. Expression levels of this gene increase when fatty acid supplies to mitochondria exceed their oxidation capacity and the protein enables the export of fatty acids from mitochondria. The haplotype that includes the variants rs3781907, rs11235972, and rs1800849 has been associated with high LDL-cholesterol in a Finnish population [35].
The two Maya populations of this study exhibit particular genetic characteristics inherent to Maya ancestry, which makes them unique for distribution studies as an initial approach for research in complex diseases. Therefore, the aim of this study was to describe the distribution of six polymorphisms associated with T2D and obesity in two communities with high prevalence of these pathologies.

Subjects
Individuals were recruited in the SRC and Merida (Yucatan, Mexico). The study included 181 healthy, contemporary Mexicans with Maya ancestry, between ages of 20 to 90 years. Participants had at least three generations of ancestors born at the place of sampling and their last names were of Maya origin. The study was performed according to the Helsinki Declaration and Institutional Review Board authorizations were obtained from Centro de Investigaciones Regionales "Dr. Hideyo Noguchi", Universidad Autónoma de Yucatán, and informed consent was obtained from all participants.

DNA Isolation
DNA was isolated from peripheral blood using either a commercial kit Invisorb® Spin Blood Midi Kit (STRATEC Molecular GmbH, Berlin, Germany), as described in the manual and/or using the automated extraction system Chemagic Prepito® (PerkinElmer Inc., Waltham, MA, USA).

Statistical Analysis
Allelic and genotype frequencies, and Hardy Weinberg equilibrium (HWE) were evaluated with SNPstat software (Barcelona, Spain). Analysis of Fisher's exact test or the Chi square test were performed with R package version 3.2.3, (Vienna, Austria). Linkage disequilibrium and haplotype frequency in variants located in the SLC30A8 gene (rs13266634 and rs11558471) were assessed using SNPstat. Haplotype was verified using Haploview version 4.1 (Cambridge, MA, USA), values of D' and r 2 in both SNPs were determined.

Genotype and Allelic Frequencies
In this study were included 181 individuals from two different populations, 93 subjects were from Merida and 88 from the SRC. The genotype and allelic frequencies are shown in Fig. 1 (Fig. 1E).

Linkage Disequilibrium Analysis
Polymorphisms rs13266634 and rs11558471 in the SLC30A8 gene showed a strong linkage disequilibrium of D'= 0.96 (Fig. 2), displaying haplotypes C-A, T-G, and C-G with frequencies of 0.725, 0.255 and 0.015, respectively.

DISCUSSION
The Maya populations of this study have been very important historically. The two populations are located near each other (with 53 km of distance between them); however, there are notable differences, for example the SRC is an isolated population in a rural community, meanwhile Merida is a cosmopolitan city in an urban area with genetic and ancestry differences that could influence studies of complex diseases [3,46].   [46]. Physiologically, individuals of Maya populations exhibit different metabolic rates, heart rates, blood pressure, and incidence of hyperthyroidism [47,48], making this population a target of study in the context of its genetic variation and susceptibility for multifactorial diseases, such as obesity and diabetes. Genotype frequency and distribution studies of the SLC16A11 gene among different Maya populations, for example, could give us insight of how genomic variants may influence the incidence of multifactorial diseases in Mexicans [49,50]. This is the first study reporting frequencies of five polymorphisms (rs8192678, rs1882095, rs11558471, rs553668 and rs3781907) located in five genes (PPARGC1A, NRF1, SLC30A8, ADRA2A and UCP3) in two contemporary Maya populations with a history of high prevalence of obesity and diabetes [8,50]. Genotype and allelic frequencies of these communities were analyzed and compared with those of African, European, South and East Asian, American and Mexican living in LA populations reported in the 1000 GP. Mexican populations are genetically shaped by an admixture of different ethnicities; hence, allelic and genotype frequencies cannot be considered the same for Mayas and for Mexican-Americans. Volunteers from LA genotyped on the 1000 GP are considered globally as a representative sample of Mexican-Mestizos, who present a high rate of admixture owing to their Spanish and Native-American genetic background and, to a minor extent, to Asian and African genetic background [51]. According to the International HapMap Project, Mexicans that are from Los Angeles have 45%, 49, and 5% of Native American, European, and African ancestry respectively (http://hapmap.org); however, evaluation of ancestry in Mexico City using trios, estimated proportions were 65%, 31%, and 3% in Native American, European and African respectively, that is different to HapMap analysis [52].
The frequencies obtained from PPARGC1A, SLC30A8, and UCP3 are in agreement to those reported by the 1000 GP in Mexican-Mestizos from LA, unlike the frequencies of NRF1 and ADRA2A, with 5.7% and 14.3%, respectively. The results of NRF1 frequencies are similar to those reported in African and South Asians, despite their evolutionary distance. Furthermore, in contrast with data reported from individuals living in LA with Mexican ancestry, whose frequencies are used as a reference of Mexican-Mestizos, a high frequency of the A allele of ADRA2A was found in the populations from the SRC and Merida. In a previous study, Kurnik et al. [53] associated homozygous individuals containing A allele with a greater hypotensive response to Dexmedetomidine (agonist of alpha receptors 2). The variant rs553668 has been also associated to overexpression of ADRA2A which contributes to T2D [34] and may be favoring a low blood pressure and increase risk to T2D.
The Maya, African and South Asian populations have high risk to T2D including migrants from South Asia [54,55] and Africa [56][57][58] living in other countries, however our study showed that the frequencies of rs1882095 (NRF1) in the Maya, Asian, and African populations are similar. Nevertheless, there were no significant differences with the 1000 GP. Consequently, it will be important to increase the sample size in future studies to determine if the difference in frequency of rs1882095 in Maya populations are statistically significant.
As we mentioned most of the SNPs frequencies were similar, even though the SRC may have a different genetic background due to a poor rate of migration inducing less genetic admixture with other communities. In addition, the analyzed frequencies of variants rs1155847 and rs13266634 located in the SLC30A8 gene displayed similar frequencies compared to those reported for Maya populations living in communities of Yucatan [50]. Furthermore, linkage disequilibrium of these SNPs resulted in a D'=0.96 indicating a high linkage disequilibrium; the fact that these SNPs are 950 bp apart from each other on chromosome 8q24.1 explains this D' value. The analysis also showed high haplotype frequency with 72.5% of C-A haplotype, 25.5% of T-G haplotype and 1.5% of C-G. This is important because the C allele (rs13266634) and A allele (rs1155847) from C-A haplotype have been reported as risk alleles in other populations [39,40,42].
In previous reports haplotype C-G in rs13266634 and rs1995222 were associated with a higher risk of T2D in all ethnicities (OR = 1.67, P = 8.6 × 10 -5 ) and in Indian subjects (OR = 1.93, P = .001) [59]. Moreover, rs11558471, rs2466295 and rs4876703 exhibited the A-C-A haplotype, which was associated with a high risk of T2D development in a Chinese Han population [60]. These results are remarkable because frequencies in SNP rs13266634 in Africans are quite contrasting. In this African population, the homozygous T allele shows a frequency of 0.5%, meanwhile the frequency of the same SNP in Mexicans in LA, SRC and Merida, present frequencies of 7.8%, 3.4% and 9%, respectively, without having a significant difference among them.
Results obtained in this work could help researchers around the world in selecting variants to develop and replicate T2D casecontrol studies in Mexican populations including Maya. Furthermore, the high difference of 14.3% for ADRA2A in the Maya populations of this study may explain the risk to T2D in Mexicans and also might help in determine susceptibility to this disease in the future.

CONCLUSION
The distribution of rs553668 (ADRA2A gene) has a higher frequency in our samples compared to European, Asian and African populations, which could be related to other studies that report lower blood pressures in populations living in the Yucatan Peninsula. Additionally, according to our significant findings regarding ADRA2A, we suggest to take into consideration the C-A (rs13266634 and rs11558471) haplotype for future case-control diabetes studies, and related traits.