Athletic performance is a complex, multifactorial trait (Posthumus et al., 2016; Noble and Noble, 2023). Although success in certain sports often depends on the cultural, traditional, or economic system in which the athletes live, in some cases genetic factors allow success regardless of environmental conditions (Semenova et al., 2023; Peplonska et al., 2017). Evolutionary pressure often results in seemingly contradictory influences of genetic mutations on the phenotype, with well-known cases of phenylketonuria or sickle-cell anemia mutation carriers gaining advantages (Withrock et al., 2015). Rare mutations causally linked to monogenic diseases can sometimes be associated with exceptional athletic performance. For example, multi-medalist skier Eero Mäntyranta carried a rare mutation in the EPOR gene, which causes autosomal dominant erythrocytosis (Juvonen et al., 1991). This highlights that potentially deleterious variants may provide a strong advantage under specific conditions or at particular moments in an athlete’s life, even if they may appear harmful under other circumstances.

The objective of our study was to identify rare deleterious variants potentially influencing athletic success. We identified a putatively deleterious variant in the FANCI gene (p.Arg1285*) in 101 Polish top-elite athletes. Furthermore, we investigated whether this variant is also present in groups of 890 national-elite athletes and 1,009 sedentary controls.

The FANCI gene (also known as KIAA1794 and FLJ10719) encodes the Fanconi anemia (FA) complementation group I (FANCI) protein. FANCI is a component of the FA pathway that is essential for repairing interstrand cross-links (ICLs). The FANCI protein is a part of the FA complex and plays a crucial role in the cellular response to DNA ICLs. If left unrepaired, ICLs can lead to DNA double-strand breaks (DSBs). The FA repair pathway is a complex process involving 22 identified FA proteins (FANCA–FANCW) and several FA-associated proteins such as FAAP24 and FAAP100. FANCM recognizes DNA damage sites due to ICLs and recruits FA core complex proteins, which leads to the formation of a monoubiquitinated FANCI–FANCD2 heterodimer. Monoubiquitination of the heterodimer is the hallmark of FA pathway activation. The monoubiquitinated FANCI–FANCD2 heterodimer then recruits downstream repair factors, including SLX4 (FANCP), XPF (FANCQ), BRCA1 (FANCS), BRCA2 (FANCD1), and RAD51 (FANCR), to complete the ICL repair. Biallelic pathogenic variants in the FANCI gene cause FA. FA is an autosomal recessive or X-linked disorder (22 genes described as FA genes) characterized by congenital anomalies, defective hematopoiesis, and an increased risk of developing acute myeloid leukemia and certain solid tumors due to a deficiency in DNA repair, which results in a reduced ability to remove DNA ICLs (Kook, 2005; del Valle et al., 2020; Sondalle et al., 2019). Heterozygous mutations in FA-related genes (e.g., BRCA1, BRCA2, BRIP1, PALB2, and RAD51C) are associated with hereditary breast and/or ovarian cancer predisposition. However, results regarding cancer predisposition for the remaining FA-related genes, including FANCI, are inconclusive (del Valle et al., 2020; Daly et al., 2017).

The study included 101 top-elite Polish male athletes (age 23.5 ± 5.9 years) who were ranked among the top 100 worldwide, had won medals in national championships, or had participated in European or World Championships (Supplementary Table S1). Based on the previous methodology (Dzitkowska-Zabielska et al., 2022), athletes were divided into speed (n = 53) and endurance (n = 48) groups (Supplementary Table S1). Whole-genome sequencing (WGS) was performed as previously described (Fichna et al., 2021). The FANCI rs121918164 variant frequency was screened using the TaqMan method (Supplementary File) in 890 national-elite athletes and 1,009 sedentary controls (Fichna et al., 2021).

The presence of the variant of interest was identified in three samples via WGS and confirmed via Sanger sequencing (Supplementary File 1).

Screening the genomes of 101 Polish top-elite athletes resulted in the identification of the potentially pathogenic variant (NM_001113378.2:c.3853C>T, NP_001106849.1:p.Arg1285*, rs121918164) in the FANCI gene in three athletes (two from the speed group and one from the endurance group) (Supplementary File 1 for details). The variant was absent in a database of healthy Polish individuals (n = 1,222; The Thousand Polish Genomes‐Nasze Genomy database) (Kaja et al., 2022). The variant is present in the Genome Aggregation Database (gnomAD); date: 13.02.2024) with an MAF of 0.00004276 with the highest prevalence in the South Asian (0.00009881) and European (non-Finnish) (0.00004746) subsets (Chen et al., 2024).

We analyzed the frequency of the variant using the TaqMan method. The variant was absent in national-elite Polish athletes (n = 890) and sedentary controls (n = 1,009) from the Polish population. It should be noted that the national-elite athlete group presented a lower level of competitive success than the initial group of international top-elite athletes.

The three top-elite athletes were born in distinct regions of Poland, far from each other, and were not related. A kinship analysis was conducted using the WGS data, and the results were confirmed using the PC-relate method implemented in Hail. The analysis confirmed that the individuals were unrelated. The highest kinship value among the tested pairs was 0.0026 (on a scale of 0–0.5) (Table 1).

All other variants in the vicinity of the detected variant were analyzed, and two intergenic variants (rs974948748 and rs750419590) were found to be in perfect LD with the variant of interest. This suggests that the variant in question may be part of a specific regional haplotype (Supplementary Figure S3). These variants were not detected in a database of ∼1,000 Polish genomes and have MAFs of 0.000073 and 0.00018, respectively, in the non-Finnish European subset in gnomAD v 4.0.0.

We also performed an estimation of the false discovery rate (FDR). The probability of a random occurrence of a rare variant in the three selected samples (with MAF = 0.00005) was estimated to be 6.56*10⁻⁸. The probability of identifying such a rare variant in any trio from the group of 100 samples is approximately 1%. This 1% value can be interpreted as an estimated FDR for a coincidental finding. Fisher’s exact test comparing athletes to the non-Finnish European population yielded a p-value of 1.57 × 10⁻⁷, which is just above the genome-wide significance threshold for a full GWAS. In our study, we considered only the high-confidence LoF variants. Large-scale studies have detected 443,769 such high-confidence predicted LoF variants in humans; therefore, the Bonferroni-adjusted p-value for this variant would be 0.069 (Uhlén et al., 2015).

This estimated corrected p-value is likely higher than the actual corrected p-value as we adopted a very conservative approach. Since many of the LoF variants are very rare, a pool of 100 WGS samples is unlikely to contain as many of them as a dataset of over 100,000 WGS samples. As an example, an analysis of 60,000 human genomes revealed 179,774 high-confidence LoF variants (Karczewski et al., 2020). If we take this number as the estimate, the Bonferroni-adjusted p-value would be 0.028.

Given the extremely low frequency of the identified variant, a robust statistical assessment remains methodologically challenging. Nevertheless, we estimated that the probability of randomly identifying a triplet of such rare variants is exceptionally low (FDR < 1%). Additionally, the calculated overall p-value for the association marginally exceeds the conventional genome-wide significance threshold.

The rs121918164 FANCI variant is annotated as pathogenic in the clinical databases ClinVar (pathogenic/likely pathogenic) and LOVD (affects function/not classified) with deleterious consequences on the FANCI protein. It has been identified in a compound heterozygous state in patients with FA, in a patient with a Li–Fraumeni-like phenotype with breast cancer, and in a two-year-old boy with severe transfusion-dependent anemia and unstable Hakkari hemoglobin (Dorsman et al., 2007; Penkert et al., 2018; Mann et al., 2022).

A similar variant in the FANCI gene, c.3854G>A (p.Arg1285Gln), identified in close proximity, involves a nucleotide change at the second position within the same codon as the c.3853C>T (p.Arg1285*) variant. This underscores the potential for complex effects on protein structure and function. The c.3854G>A variant is predicted to create an additional ATM/ATR phosphorylation motif, suggesting that alterations in these pathways could significantly affect the efficacy of the DNA damage response (DDR). The DDR is essential for maintaining genomic integrity. Key regulators, including ATM, ATR, and DNA-PKcs, orchestrate repair processes and cellular processes to DNA damage, such as DSBs and replication stress (Dorsman et al., 2007).

It has also been shown that another mutation at the same codon (p.Arg1285Gly) appears to cause a milder phenotype in homozygosity (Savage et al., 2016). An in vitro study of BD0952 cells (lymphocytes from a patient with a classic FA presentation) shows that the FANCI mutation at the highly conserved Arg1285 at the C-terminus of the protein position (Arg1285Gln) causes FA in donor cells, makes the cells MMC-sensitive, and disturbs FANCD2 regulation (Smogorzewska et al., 2007). These results may indicate that the pathogenicity of the p.Arg1285* mutation depends on the genomic context and the contribution of other variants.

Despite potential harm, the presence of the p.Arg1285* mutation in three unrelated top-elite athletes may suggest both negative and, in some cases, positive effects on the phenotypes of the carrier. This phenomenon of seemingly contradictory (negative–positive) influences of genetic variants on the phenotype is well-recognized, with examples including sickle-cell anemia and malaria resistance and the CCR5 mutation and HIV resistance (Withrock et al., 2015). The case of Eero Mäntyranta also showed that carrying a specific variant in the erythropoietin receptor (EPOR) gene, which causes autosomal dominant erythrocytosis-1, is potentially beneficial for endurance sports, without causing obvious adverse clinical symptoms (Juvonen et al., 1991). The spread of pathogenic mutations in populations may be explained by natural selection, which favors traits that are useful in specific situations, rather than the overall health or longevity of carriers (Gluckman et al., 2011).

The identification of the p.Arg1285* variant in the FANCI gene in elite Polish athletes raises an interesting question about the relationship between DNA damage repair mechanisms and athletic performance. Further studies are needed to elucidate its risk–benefit mechanism. There is a risk that in vitro or in vivo model studies may not capture subtle cellular-level effects that could contribute to a larger physiological outcome, which may be observable only within a specific time window.

We are unaware of any specific biological associations that could limit the possible positive effects of the variant to a certain time window‐such as the optimal age for training‐after which any evolutionary advantage may diminish or even become detrimental. In such a case, apart from direct observation of the variant (or variants) in larger cohorts and different populations, we may not have another feasible method of assessing its impact since the biological effect may not be detectable at the cellular or model organism level.

As we did not detect the variant in national-elite athletes and sedentary controls, we cannot report the exact or the corrected p-value for the detected variant. Despite multiple efforts to exclude the coincidental detection, including calculating the probability of a completely random finding, we cannot completely exclude the possibility that the variant was identified by chance in three unrelated top-elite athletes.

We cannot distinguish whether the p.Arg1285* variant is a causative or beneficial factor for the elite athlete phenotype or whether it is only associated with a specific haplotype or a nearby causative variant.

No medical or physiological data are available; health status cannot be assessed in this study. Therefore, we cannot study the negative or positive impact of the mutation on the physiological phenotype. However, since the variant carriers are athletes, and therefore physically fit, it can be assumed that they did not experience any adverse health effects, at least until they participated in the study.