Multi-phase, multi-ethnic GWAS uncovers putative loci in predisposition to elite sprint and power performance, health and disease

Abstract

The genetic underpinnings of elite sprint and power performance remain largely elusive. This study aimed to identify genetic variants associated with this complex trait as well as to understand their functional implications in elite sprint and power performance. We conducted a multi-phase genome-wide association study (GWAS) in world-class sprint and power athletes of West African and East Asian ancestry and their geographically matched controls. We carried out genotype imputation, replications for the top GWAS signal rs10196189 in two European cohorts, and gene-based and tissue-specific functional network analyses. For the first time, we uncovered the G-allele of rs10196189 in the Polypeptide N-Acetylgalactosaminyltransferase 13 (GALNT13) being significantly associated with elite sprint and power performance (P = 2.13E-09 across the three ancestral groups). Moreover, we found that GALNT13 expression level was positively associated with the relative area occupied by fast-twitch muscle fibers in the vastus lateralis muscle. In addition, significant and borderline associations were observed for BOP1, HSF1, STXBP2, GRM7, MPRIP, ZFYVE28, CERS4, and ADAMTS18 in cross-ancestry or ancestry-specific contexts, predominantly expressed in the nervous and hematopoietic systems. From the elite athlete cohorts, we further identified thirty-six previously uncharacterized genes linked to host defence, leukocyte migration, and cellular responses to interferon-gamma, and four genes - UQCRFS1, PTPN6, RALY and ZMYM4 - associated with aging, neurological conditions, and blood disorders. Taken together, these results provide new biological insights into the genetic basis of elite sprint and power performance and, importantly, offer valuable clues to the molecular mechanisms underlying elite athletic performance, health and disease.

Keywords: Functional annotations; GWAS; Genetic diversity; Imputation; Meta-analysis; Polypeptide N-Acetylgalacto-saminyltransferase 13.

FIG. 1

Genetic associations following imputation (IMPUTE2 and Sanger Imputation Server) and meta-analysis of Jam, A-A and Jpn cohorts with athletic prowess (sprinting) across 22 autosomes. Dotted line: suggestive significance cut-off of P = 1E-05; dashed line: genome-wide significance cut-off of P = 5E-08. X-axis: chromosome number; Y-axis: -log10(P).

FIG. 2

Tissue-/cell-type specific expression of GALNT13, STXBP2, BOP1, HSF1, ZFYVE28, ADAMTS18, GRM7 and MPRIP in the nervous system (a), hematopoietic system (b), as well as other multiple systems (c), following the functional networks analysis in GIANT [37] (the data available under a CC-BY 4.0 license, see: https://humanbase.readthedocs.io/en/latest/).

FIG. 3

Cellular responses to interferon-gamma unveiled in blood-specific functional modules, linking individual genes to specific biological processes. Top three enriched GO terms are annotated for each module, and novel, previously uncharacterized genes are labelled in M1-M7 (among which, the highest ranked genes are marked in magenta based on the number of edges they associate with). The figure is adapted from the output of the modules analysis implemented in humanbase [40] (the data available under a CC-BY 4.0 license, see: https://humanbase.readthedocs.io/en/latest/)