(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



An endothelial regulatory module links blood pressure regulation with elite athletic performance [1]

['Kim Fegraeus', 'Department Of Medical Sciences', 'Science For Life Laboratory', 'Uppsala University', 'Maria K. Rosengren', 'Department Of Animal Biosciences', 'Swedish University Of Agricultural Sciences Uppsala', 'Rakan Naboulsi', 'Childhood Cancer Research Unit', 'Department Of Women S']

Date: 2024-07 

The control of transcription is crucial for homeostasis in mammals. A previous selective sweep analysis of horse racing performance revealed a 19.6 kb candidate regulatory region 50 kb downstream of the Endothelin3 (EDN3) gene. Here, the region was narrowed to a 5.5 kb span of 14 SNVs, with elite and sub-elite haplotypes analyzed for association to racing performance, blood pressure and plasma levels of EDN3 in Coldblooded trotters and Standardbreds. Comparative analysis of human HiCap data identified the span as an enhancer cluster active in endothelial cells, interacting with genes relevant to blood pressure regulation. Coldblooded trotters with the sub-elite haplotype had significantly higher blood pressure compared to horses with the elite performing haplotype during exercise. Alleles within the elite haplotype were part of the standing variation in pre-domestication horses, and have risen in frequency during the era of breed development and selection. These results advance our understanding of the molecular genetics of athletic performance and vascular traits in both horses and humans.

A previous study discovered that a genomic region close to the Endothelin3 gene was associated with harness racing performance. Here, careful phenotypic documentation of athletic performance and blood pressure measurements in horses, followed by state-of-the-art genomics, allowed us to identify a 5.5 kb regulatory region located approximately 50 kb 3’ of the EDN3 gene. A comparative analysis of the region using human HiCap data supported a regulatory role as, in endothelial cells, interaction was observed between the region and multiple genes relevant to blood pressure regulation and athletic performance. Long range cis-regulatory modules are critical for cooperatively controlling multiple genes located within transcriptionally active domains. We measured blood pressure in Coldblooded trotters during exercise and demonstrated that horses with two copies of the elite-performing haplotype had lower blood pressure during exercise and better racing performance results, compared to horses with two copies of the sub-elite performing haplotype. In addition, horses with the elite-performing haplotype also had higher levels of Endothelin3 in plasma. The results reported here are important for understanding the biological mechanisms behind blood pressure regulation in relation to racing performance in both horses and humans.

Funding: The Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS) (to GL) and The Swedish Research Council (VR) (to GL). This project has also received funding from the CNRS, University Paul Sabatier (AnimalFarm IRP) (to LO), and the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement 681605-PEGASUS) (to LO). K.F. and A.R. were supported by grants from FORMAS (2020-01135) (to AR). K.F. received salary from FORMAS (2020-01135). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability: The racing performance data that support the findings of this study are available from the Swedish Trotting Association (Stockholm, Sweden), but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. However, data are available from The Swedish Trotting Association (Svensk Travsport, Box 201 51, 161 02 Bromma, Sweden). The contact person is Christina Olsson, head of the breeding department at The Swedish Trotting Association, kundtjanst@travsport.se . The WGS data has been deposited in the BioProject database (NCBI repository SRA) with ID PRJNA1045044 according to their guidelines. According to the ethical permit related to the study (application number 2006/784-31/1 and 2012/1633-31/4) approved by the Human Research Ethics Committee at Karolinska Stockholm, Sweden, the sequencing data from living patients cannot be shared due to patient privacy regulations. We therefore only share the summary or processed form of the sequencing data in S5 – S7 Tables. More information can be found at Etikprövningsmyndigheten, Box 2110, 750 02, Uppsala, or registrator@etikprovning.se .

Genomic regulatory regions, such as putative enhancers, are difficult to study simply because they are hard to characterize [ 17 ]. They can be cell type and condition specific, acting both upstream and downstream of target genes, or over long genomic distances, where genome looping brings them into proximity to their target genes [ 18 ]. Targeted chromosome conformation capture (HiCap) is an experimental method that can detect genomic loops mediating regulatory interactions between such regions and the promoters of their target genes [ 19 , 20 ]. In this study, we fine-mapped the harness racing selective sweep, and used comparative data to interrogate its potential function. In addition to equine epigenetic data, we have used HiCap datasets produced using twelve different human cell types, including vascular endothelial cells, to indicate a role of our potential regulatory region and the target genes involved. Further, since the sweep region flanks the GNAS-EDN3 region identified in multiple human blood pressure GWAS [ 21 , 22 ], we evaluated the effect of the minimal sweep horse haplotypes on blood pressure, as well as EDN1 and EDN3 plasma levels, before and during exercise. Finally, key genetic variants within the sweep region were characterized for spatial and temporal distribution within past and present horse populations. Taken together, the results presented here indicate the identification of a regulatory unit, likely important for vascular traits and performance in horses and humans.

The harness racing selective sweep study revealed a 19.6 kb genomic region on chromosome 22, located approximately 50 kb downstream of the Endothelin 3 (EDN3) gene, under selection in CBT [ 15 ]. Five SNVs in high linkage disequilibrium (LD, r 2 = 0.92–0.94) were significantly associated with racing performance, including the number of victories, earnings, and racing times. Variant rs69244086 C>T (EquCab3.0, ECA22:46717860) showed the strongest association within the sweep region, and was further genotyped in 18 additional horse breeds. The favorable T allele was found in high frequency in breeds used for racing, while it generally remained at low frequency in ponies and draught horses. We hypothesized that the identified region might contain a regulatory element influencing either the expression of EDN3 or other genes nearby [ 15 ].

We have previously used a unique, three breed admixture Nordic horse model, to study the genetics of athletic performance traits [ 15 , 16 ]. The basis for this model is that racing Coldblooded trotters (CBTs) predominantly originate from the North Swedish Draught horses (NSDs), a sturdy breed used in farming and forestry. It is well established that some crossbreeding occurred between Standardbreds (SBs) (the most commonly used breed for harness racing) and CBTs before obligatory paternity testing was introduced in Sweden in the 1960s. However, a remarkable improvement in the racing performance of the CBT has occurred during the last fifty years. In part, it is likely that this improvement could be explained by crossbreeding, with a marked increase of favorable genetic variants originating from SBs and introduced to the CBTs. This process should leave “genetic footprints” in the genome of CBT in the form of chromosome segments originating from SBs. We have identified such a footprint using pair-wise allele frequency distribution and selective sweep mapping [ 15 ]. Notably, our model, which is solely based on the hybrid origin of the CBT, provides a unique opportunity to study genes influencing body constitution and complex morphological traits of importance for racing success.

Despite the apparent potential of domestic animal models to provide valuable insights into the natural biological mechanisms driving human traits and disease, their use to date has been limited. In the past, the use of domestic animals as models for genomic research has provided basic knowledge concerning gene function and biological mechanisms and a complementary view on genotype–phenotype relationships [ 1 ]. Recent advances in the availability and quality of human and mammalian reference genomes, plus the technological advances required for their alignment, are revealing both coding and non-coding conserved bases, key to unraveling shared gene function and regulation [ 2 , 3 ]. The horse is one of the most popular species for studying athletic performance. They have been intensively selected for centuries, to encompass the optimal physical capacity for strength, speed, and endurance [ 4 , 5 ]. Additionally, their recent population history, involving closed populations selected for similar phenotypic traits within breeds and large variations across breeds, has created a favorable genome structure for genetic mapping. These factors combined make the horse an optimal model for studying the molecular genetics underlying athletic performance and the complex biological processes activated by exercise [ 1 , 4 , 6 , 7 ]. Previous research in horses has begun to unravel the genetics of complex traits, such as muscle mass and locomotion patterns, on athletic performance, with many potential candidate genetic variants identified [ 8 – 14 ]. However, it is much more challenging to understand the mechanisms by which the identified variants exert their functions.

ELISA tests were used to measure the plasma concentration of EDN1 and EDN3 at rest and during exercise for each haplotype group. Samples with a coefficient of variation (CV) value above 20% were excluded from the analysis ( S10 and S11 Tables). Horses homozygous for SPH (n = 8) had a significantly higher plasma concentration of EDN1 at rest and during exercise, compared to the other groups (Tukey´s HSD test) ( Fig 5 ). In addition, SPH horses had lower plasma concentrations of EDN3 at rest (P = 0.06) and during training (P = 0.003) compared to horses homozygous for the EPH ( Fig 5 ). Individual plasma values are presented in S10 and S11 Tables. There was a significant correlation between the plasma EDN3 and DPB (R = -0.66, P = 0.03) and a borderline significant correlation between EDN3 and MAP (R = -0.57, P = 0.07) during exercise. Also, there was a significant correlation between EDN1 and DBP (R = 0.73, P = 0.01), SBP (R = 0.7, P = 0.02) as well as MAP (R = 0.81, P = 0.003) during exercise.

At rest, blood pressure measurements were also taken from eight heterozygous horses (HET). For all traits, a significant interaction was calculated between haplotype and time point. On average, horses homozygous (HOM) for SPH had significantly higher systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) during exercise, measured directly after the last uphill interval ( Fig 4A–4C ). In addition, five minutes after the uphill interval, the SPH group showed significantly higher SBP, MAP, and pulse pressure (PP) than the EPH group ( Fig 4D and 4E ). There were no statistically significant differences in blood pressure measurements at rest before the exercise. Although there was a higher proportion of males in the SPH group, there was no significant impact of sex on any of the blood pressure measurements. All blood pressure values are presented in S9 Table .

For all 14 SNVs within the minimum shared haplotype in the trotters, we investigated the allele frequency over time using the mapDATAge package (35) and 431 previously characterized ancient genomes [ 4 , 36 – 40 ]. For all SNVs except two, rs69244085 and rs69244086, there has been an increase of the alternate allele observed from at least 7,500 to 5,500 years ago ( Fig 3 ). For all 14 SNVs, the two alleles were shown to segregate in specimens pre-dating the rise and spread of the DOM2 genetic lineage of modern domestic horses [ 35 ].

In both SBs and CBTs the rs69244089 genotypes distributed according to HWE (P = 0.16 and P = 0.35, respectively). For breeds other than CBT and SB, each individual’s performance status is unknown. However, a trend for an increased frequency of the C allele in traditional performance breeds was observed, e.g., Arabian horses, Thoroughbreds, and Warmbloods. In contrast, this allele was low or absent in draft horses and ponies ( Table 5 ).

Given that all 14 associated variants within the minimum shared 5.5 kb locus were in perfect LD in CBTs and SBs, we used the genotypes at rs69244089 as a proxy for EPH (C allele) and SPH (T allele) haplotypic pairs and re-assessed the association between haplotypes and racing performance. For CBTs (n = 516), EPH homo- or heterozygotes outperformed the SPH homozygotes ( Table 3 ). In SBs, there were no statistically significant differences between the three genotypes ( Table 4 ). However, when analyzing horses carrying at least one T allele versus CC horses, the results were significant for several performance traits ( Table 4 ), including the number of wins, placings and earnings. While the number of wins and placings was higher in horses with at least one T allele, the CC horses earned more money than the TT/TC horses ( Table 4 ).

Moving back to the horse genome, we investigated the in silico regulatory potential of each allele from the 14 SNVs associated with racing performance and included in the minimum shared 5.5 kb region, to alter transcription factor binding. Using the EquCab3 genome reference and the Homo sapiens Comprehensive Model Collection (HOCOMOCO) transcription factor binding model database [ 33 ] in motifbreakR [ 34 ], we found that all alleles had the potential to cause alterations of high effect. SNVs rs69244086 C>T and rs69244089 T>C are illustrated as examples in S1 Fig . In S8 Table , the transcription factor binding scores for all analyzed variable sites are listed.

In step four, we assessed the potential functional impact of the minimum shared 5.5 kb region using comparative data from both horse and human resources ( Fig 1C and 1D ). Epigenetic data drawn from nine tissues was generated by the Equine section of Functional Annotation of Animal Genomes (FAANG) [ 24 ] for two Thoroughbreds, both homozygous for the EPH haplotype. No clear signals were observed in the minimal region for four marks (H3K27ac, H3K27me3, H3K4me1, H3K4me3) measured across eight different tissues (adipose, brain, heart, liver, lung, muscle, ovary, skin). But in lamina we observed an insulator and an active enhancer in our 5.5 kb minimum shared region ( Fig 1 ). To access additional functional data points, we lifted the minimum shared 5.5 kb region and contained SNVs to the human reference genome, hg38. In comparative analyses, the liftover indicated that, although not well conserved, the region had regulatory potential, including Open Regulatory Annotation database (ORegAnno) elements [ 25 ], ENCODE cCREs [ 18 ] and GTEx cis-eQTL variants regulating EDN3 in esophagus mucosa tissue [ 26 ] ( Fig 1D ). We further investigated the regulatory potential of the minimum shared 5.5 kb region using comparative chromatin interaction profiles in 12 different human cell types, including iPS cells and a neural cell line (see Methods and S5 Table ). SNVs lifted to the minimum 5.5 kb region showed interactions with the promoters of multiple genes, but only in the vascular endothelial cell datasets ( Fig 1C and 1D and S5 Table ). With a relaxed threshold, requiring support evidence from at least two samples, direct interaction between human to horse lifted SNVs and seven genes were observed ( Fig 1C ). These interactions included the proximal cell membrane-cytoskeleton dynamics gene, PHACTR3 [ 27 ], through to the megabase distal cell reprogramming transcription factor gene, TFAP2C [ 28 ]. While specific SNVs in humans are unlikely to have the exact same role in horses, we never-the-less explored the human direct, and indirect, interaction network to gain a wider perspective of genes which may be regulated by the minimum shared 5.5 kb region ( Fig 2 ). Here, support was required from five samples, and all human loci are considered, even if these are not currently supported with horse annotation evidence. We see that the 5.5 kb region (putative regulatory locus, PutRegLocus) interacts with the EDN3 promoter indirectly, via GNAS and two other promoters ( Fig 2 ). GWAS variant rs16982520, located within ZNF831, is associated with hypertension [ 29 , 30 ], systolic blood pressure [ 31 ] and mean arterial pressure [ 32 ] and interacts directly with EDN3 and four direct interactors of the putative regulatory region (GNAS, RP4-614C15.2, SPO11 and ZNF831), suggesting a complex regulatory pattern. In total, there were 15 protein-coding genes that directly or indirectly interacted with the putative regulatory region ( S6 Table ). We performed a gene enrichment analysis for these 15 genes and found that several human phenotype terms related to thyroid hormone regulation were enriched ( S7 Table ).

In step three, we prioritized variants for further analysis. The 400 bp deletion (ECA22:46,714,602–46,715,003) was absent in horses with low earnings per start and either heterozygous (SB) or homozygous (CBT) in elite performing horses. However, this variant was not significantly associated with racing performance traits following genotyping and analyses in a further 497 CBTs (linear models, S4 Table ). Due to technology constraints, 25 of the available 82 single base pair variants were selected for genotyping in a larger horse material. Variant selection was based on available pooled allele frequency data matching admixture expectation in CBTs, SBs and NSD [ 16 ], variant spacing across the region, and success in MassArray design (see Methods ). The 24 SNVs and one single base pair deletion ( Fig 1B ) were genotyped in 391 horses sampled from 12 different breeds (including 210 CBTs), as well as three Przewalski horses. SNV rs69244086 was not included in this set, but was either directly genotyped or imputed across the dataset. After quality control, 24 SNVs and 394 horses (including the Pzrewalskis) were available for further analysis. Pairwise LD calculations with SNV rs69244086 revealed 15 SNVs in strong LD (r2 ≥ 0.53, 14 SNVs with r2 > 0.98) (ECA22:46,708,983–46,719,042). LD decayed outside the region (r 2 < 0.07), leaving 14 SNVs for haplotype analysis ( Fig 1B ). Phasing revealed in total 14 different haplotypes, with seven haplotypes found at a frequency > 2% in the total sample set, or within each breed represented by five or more individuals ( Table 1 ). Generalized linear model (GLM) regression in the R environment (23) was used to test the association between the two haplotypes harbored by CBTs (n = 165) and SBs (n = 38), and harness racing performance traits. For both CBTs and SBs, the H1 haplotype, carrying the rs69244086-T high-performance associated allele, was the most common ( Table 2 ). In both breeds, the H2 haplotype demonstrated a significantly negative effect on all performance traits tested, except for the number of starts in SBs (P = 0.59) and the number of wins in SBs (P = 0.06) ( Table 2 ). The above analysis showed that the minimum 5,564 bp, 14 SNV shared haplotype (ECA22: 46,713,478–46,719,042), significantly associated with racing performance in both CBTs and SBs. This allowed for the definition of H1 as an elite-performing haplotype (EPH: AAGCGTTTCTCAAA) and H2 as a sub-elite-performing haplotype (SPH: GGTTACCCTCTGGG). Of note, the reference EquCab3.0 haploid reference genome, generated from a Thoroughbred, represents the SPH haplotype.

In step two, we performed whole genome sequencing (WGS) of two CBTs and two SBs to increase variant density across the 19.6 kb sweep region ( Fig 1B ). In each breed, horses were selected to be homozygous for different alleles at SNV rs69244086 (CC and TT, respectively) and had either high (TT) or low (CC) earnings per start ( S2 Table ). From the 19.6 kb sweep region, 78 SNVs and six indels (one 400 bp deletion, one 23 bp deletion and four single bp INDELs) were identified ( Fig 1B and S3 Table ).

A) Location of the 2018 selective sweep. B) Fine-mapped region including discovery variants, fine-mapped SNVs associated with racing performance, the refined minimum sweep region as well as active elements in lamina from FAANG data. Discovery SNVs are indicated in black, deletions in red. C) HiCap interacting SNV and genes, lifted from humans. D) The span of the minimum sweep region covers an extended 11 kb space in the human reference genome, hg38. HiCap interacting SNVs are indicated in relation to known regulatory elements and SNVs reported from GWAS studies. Zoonomia phyloP scores and cactus alignments show the conservation of this region across mammals. The yellow boxes represent distal enhancer-like signatures and the blue box represent a CTCF-only.

With a four-step process, we reduced the 19.6 kb sweep region from the previous study [ 15 ] to a minimum shared 5.5 kb region ( Fig 1A and 1B ). In step one, we repeated the racing performance association analysis using 251 more horse samples (total n = 629) with SNV genotypes extracted from the 670K Axiom Equine Genotyping Array. However, supplementing the original 378 CBTs with additional samples did not narrow the 19.6 kb region [ 15 ]. Seven SNVs within the sweep (five significantly associated with racing performance and two flanking SNVs) [ 15 ] were extracted from the array, including the most significant one: rs69244086 [ 15 ]. Pairwise LD centered on the previously most associated SNV, rs69244086 [ 15 ], resulted in the same core five SNV region (ECA22:46,717,451–46,718,964, r 2 > 0.6), and analysis of linear models showed that these SNVs remained significantly associated with harness racing performance traits (i.e., number of victories, earnings, and race times, P ≤ 0.05) ( S1 Table ).

Discussion

Naturally occurring animal models, such as livestock and companion animal species, can provide complementing views to de novo generated animal models, where genome editing is required and the genomic context of the mutation may be lost. Natural model populations can carry specific genetic variants that have been under artificial selection to obtain desired phenotypes. In this study, we have used such a model to study the regulatory genomics of blood pressure modulation. More fundamental knowledge on exercise related blood pressure modulation and blood pressure tuning in general is warranted, since dysregulation of these systems are linked to both cardiovascular disease and metabolic syndrome [41,42]. By performing multiple experiments, to first refine a selective sweep region associated with racing performance, followed by comparative transcription factor binding analyses and 3D genome interaction mapping, as well as horse plasma ELISA and blood pressure measurements, we identified a regulatory element comparatively active in human endothelial cells, potentially affecting horse blood pressure regulation and athletic performance. Here, we demonstrate a regulatory role of a potentially admixed genomic region, advantageous for harness racing performance in horses, and show that this region interacts with several well-known blood pressure linked genes, as identified in human GWAS studies [21,30].

The intergenic selective sweep identified in 2018 [15] is located 50 kb downstream of the EDN3 gene. Given both the known role of EDN3 in blood pressure regulation and its proximal location, we hypothesized that 1) the identified non-coding region harbored a regulatory element that acted on the EDN3 gene and 2) a key phenotype for elite athletic performance is blood pressure regulation. Our results demonstrated significant associations between the two minimal haplotypes identified in CBTs and SBs and multiple racing performance traits. In addition, in comparison with SPH homozygotes, horses homozygous for the EPH had significantly higher levels of EDN3 in their plasma and lower blood pressure readings during exercise.

The role of endothelins in blood pressure regulation is well studied. Endothelins are involved in one of the most potent vasoregulatory systems, where EDN1 and EDN3 have opposite effects and act synergistically to regulate blood pressure [43]. EDN1 increases blood pressure by vasoconstriction, while EDN3 is involved in nitric oxide release, in turn resulting in vasodilation and decreased blood pressure [43]. Further, EDN3 also stimulates the secretion of vasopressin, which increases blood volume by retaining water in the kidneys, while EDN1 has the opposite effect in the same organ [44–47]. Although EDN3 clearly contributes to blood pressure regulation, and was a strong candidate for our regulatory region, in the absence of appropriate horse samples, we used already produced HiCap datasets from 12 different human cell types, to investigate the comparative chromatin interaction profile of our identified region. Spanning 12 cell types covering multiple organs and life stages, including iPS cells and a neural cell line (S5 Table), the results showed that interactions were only observed between our putative regulatory region and different genes in human primary vascular endothelial cells. Endothelial cells make up the inner layer of blood vessels, and regulate the exchange between the blood and the surrounding tissues, by adjusting the vessel diameters according to the tissue demand [48]. Endothelial cells are critical regulators of the vascular tone and blood pressure regulation, as they influence the contractile ability of vascular smooth muscle cells within small arteries [49].

From the human HiCap data analysis, genes with direct interactions with our putative regulatory region included GNAS, SPO11 and ZNF831, all with orthologues and conserved gene order in horses (Fig 1C). Also included was the human specific lincRNA RP4-614C15.2 (Fig 2). Each of these genes encode proteins with known functional roles related to endothelial dysfunction but are certainly not limited to this function. GNAS is a complex locus which is imprinted in humans and mice [50,51]. It gives rise to multiple gene products through the use of both alternative promoters and splicing mechanisms [52]. This inclusion of varied first exons together provides instructions for the alpha subunit of the protein complex, called a guanine nucleotide-binding protein (G protein) [53]. This transmembrane protein complex is involved in calcium and potassium ion concentration changes within cells [54,55]. These changes are crucial for regulating heritable traits such as cardiac output and peripheral vascular resistance. In fact, blood pressure regulation in humans is a classical complex genetic trait with heritability estimates of 30–50% [56]. In line with this, GWAS of blood pressure measures has identified hundreds of genetic variants associated with systolic- and diastolic blood pressure as well as hypertension in humans [30]. A genomic region frequently identified in human GWAS for blood pressure measures is the wider GNAS-EDN3 region [21,22]. The SPO11 (Initiator Of Meiotic Double Stranded Breaks) gene is involved in DNA damage repair [57], and the stressors leading to DNA damage can be found during exercise.

In both humans and horses, intense exercise induces a physical stress response, in order to supply optimal energy and oxygen for the working muscles. Endothelial cells in the vascular wall have a significant role in maintaining blood flow by regulating the diameter of the blood vessels and by preventing the blood from clotting [58,59]. The blood flows more efficiently, and with less turbulence if the blood vessels are wide and if the blood is less viscous [60]. As such, blood vessels must be dilated during intense exercise, in order to supply working muscles with sufficient energy and oxygen. An increasing body of evidence suggests that oxidative stress, which results in an excessive generation of reactive oxygen species (ROS), is vital in the pathogenesis of hypertension [61]. Oxidative stress and inflammation significantly contribute to vascular remodeling by promoting exaggerated contractility and proliferation of vascular cells [62]. These factors also favor DNA damage, thereby linking SPO11 to endothelial dysfunction. It should be noted that several genes with known function in DNA damage and response pathways previously have been associated with pulmonary arterial hypertension [63]. ZNF831 encodes a transcription factor that has been associated, via GWAS, with a wide range of human phenotypes, including body height [64], platelet component distribution width [65], hair color [66] and diastolic blood pressure [21,67]. The ZNF831 mRNA is highly expressed in whole blood and spleen, and the protein is identified in the heart and T-lymphocytes [68], making it a likely candidate transcription factor for endothelial cells. While not orthologous in horse, the human lincRNA RP4-614C15.2 is a gene with a known role in angiogenesis [69]. LincRNAs are found explicitly in the nucleus, functioning in cell differentiation and identity. Imminent evidence suggests that long non-coding RNAs (lncRNAs) play critical roles in pulmonary vascular remodeling and pulmonary arterial hypertension (PAH) [70]. LncRNAs are implicated in regulating chromatin structure, thereby leading to pulmonary arterial endothelial dysfunction by modulating endothelial cell proliferation, angiogenesis, endothelial mesenchymal transition, and metabolism [71]. Targeting epigenetic regulators may lead to new, potential therapeutic possibilities in treating PAH [71]. It is clear that the search for potential lncRNAs regulating this wider genomic region in horses are required, but even so, these four directly interactive genes comprise promising candidates that warrant further investigation. The results observed from the human HiCap data were supported by the analysis of ChIP-seq data provided by the equine section of FAANG [72,73]. We demonstrated that there is an active enhancer in lamina, located in our 5.5 kb region (Fig 1). Laminae can be described as finger-like protrusions of tissue in the hoof and is affected in equine laminitis. There is plentiful evidence that a compromised blood flow through the lamina develops in horses with laminitis e.g., [74], but little is known about the cause. One piece to this puzzle has shown that the concentration of EDN1 in laminar connective tissues obtained from laminitic horses was higher than in non-laminitic horses [75,76]. Horses with laminitis are hypertensive [77], consistent with digital hypoperfusion. All in all, these findings may be linked to the active enhancer in our 5.5 kb region in lamina.

Each SNV within the minimum shared 5.5 kb span was found to segregate in the genome of horses that lived before the rise and spread of the modern domestic lineage, around ~4,200 years ago (DOM2) [40]. Temporal allelic trajectories portray an increase in the alternative allele frequency, from 7,500–5,500 years ago, prior to the earliest archaeological evidence of horse husbandry [39,78]. This concerted rise in alternate allele frequency following the spread of the DOM2 lineage is compatible with a haplotype undergoing positive selection, possibly following artificial selection for horses showing improved athletic performance. Interestingly, the allele trajectories are in striking contrast with those previously described at the myostatin locus, which is driving performance in short-distance racing [8]. In that case selection was recent, indicated to start only within the last 1,000 years. It may be that performance traits represent recurrent selection targets over the history of horse domestication.

Although the current study demonstrated statistically significant differences in horse EDN1 and EDN3 levels between the different haplotype groups, there were no observed differences in blood pressure at rest. In adult horses, normal resting blood pressure is approximately 130/95 mmHg (systolic/diastolic) [79], in line with the values observed in the current study. In humans, it has been demonstrated that elevated blood pressure at rest is negatively correlated with athletic performance [80]. Individuals with elevated blood pressure had significantly lower maximal oxygen consumption, ventilatory anaerobic thresholds, and heart rate (HR) reserves (difference between an individual’s resting HR and maximum HR) [80]. However, unlike humans, high blood pressure in horses is very uncommon, and most blood pressure measurements on horses are performed to evaluate and monitor hypotension. An increase in blood pressure at rest is most commonly seen as a result of diseases such as laminitis, chronic renal failure, or equine metabolic syndrome [81–83].

Perhaps the differences in observed blood pressure measurements between humans and horses is driven by species differences in the red blood cell reservoir, the spleen [60]. A previous study on horses demonstrated a significant correlation between blood pressure and spleen volume, with the splenic volume decreasing when hypertension was induced [84]. The equine spleen reservoir is larger than in any other domestic animal, and when there is increased demand, erythrocytes stored in the spleen can be released into the system [60]. Given the relationship between spleen size and blood pressure, further studies examining the relationship between the identified regulatory region, spleen size, and blood pressure represent promising avenues for enhancing the understanding of the cardiovascular system in horses in relation to exercise. While it may be viewed as a limitation, in the current study, blood pressure was measured using an external cuff. In general, direct blood pressure measurements via a fluid-filled cannula inserted into an artery are more accurate. However, the indirect blood pressure measurements are less invasive and horse cuff protocols have previously been successfully evaluated [85].

The regulatory potential of the region identified in this study is likely not limited to blood pressure regulation. For example, the EDN3 ligand has a broad expression across a range of tissues and has been implicated in diseases such as Hirschsprung’s disease, Multiple Sclerosis, and Waardenburg syndrome in humans, as well as melanocyte development in mice [86–90]. Also, the GNAS gene encodes a G protein with multifaceted functionality. Much of vertebrate physiology is based on the versatile functionality of G protein complexes associated with their G protein coupled receptors (GPCR) [91]. Notably, the G protein encoded by the GNAS gene helps stimulate the activity of an enzyme called adenylyl cyclase [92]. Adenylyl cyclase is the key enzyme that synthesizes cAMP and it is involved in cellular processes that help regulate the activity of endocrine glands such as the thyroid, pituitary gland, gonads, and adrenal glands [93,94]. In addition, adenylyl cyclase is thought to influence bone development signaling pathways, limiting the spatial production of bone with tissue [95]. Further, we see a trend associated with the haplotype and conformational type differences across horse breeds, warranting further studies. Another intriguing avenue is via thyroid hormone regulation, a network with profound effects on the cardiovascular system [96,97]. Therefore, it is unsurprising that the 15 protein-coding genes that directly, or indirectly, interacted with the putative regulatory region, showed gene ontology terms related to thyroid hormone regulation in an enrichment analysis (S7 Table). Thyroid hormone likely mediates the new demands and adaptations on the cardiovascular system needed for elite performance.

Taken together, this study underscores the pivotal role of a specific genomic region in regulating endothelial tissue integrity across species boundaries. By shedding light on the intricate mechanisms governing blood pressure modulation, these findings hold promise for advancing our understanding of cardiovascular health and disease. Moreover, the implications extend beyond blood pressure regulation, potentially encompassing a myriad of physiological processes influenced by hormone signaling. This elucidation of molecular interactions underscores the interconnectedness of mammalian physiology and highlights avenues for future research into therapeutic interventions and preventive measures against cardiovascular disorders.

[END]
---
[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1011285

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/