The ACTN3 R577X Polymorphism Is Associated with Cardiometabolic Fitness in Healthy Young Adults

Abstract

Homozygosity for a premature stop codon (X) in the ACTN3 “sprinter” gene is common in humans despite the fact that it reduces muscle size, strength and power. Because of the close relationship between skeletal muscle function and cardiometabolic health we examined the influence of ACTN3 R577X polymorphism over cardiovascular and metabolic characteristics of young adults (n = 98 males, n = 102 females; 23 ± 4.2 years) from our Assessing Inherent Markers for Metabolic syndrome in the Young (AIMMY) study. Both males and females with the RR vs XX genotype achieved higher mean VO₂ peak scores (47.8 ± 1.5 vs 43.2 ±1.8 ml/O₂/min, p = 0.002) and exhibited higher resting systolic (115 ± 2 vs 105 ± mmHg, p = 0.027) and diastolic (69 ± 3 vs 59 ± 3 mmHg, p = 0.005) blood pressure suggesting a role for ACTN3 in the maintenance of vascular tone. We subsequently identified the expression of alpha-actinin 3 protein in pulmonary artery smooth muscle, which may explain the genotype-specific differences in cardiovascular adaptation to acute exercise. In addition, we utilized targeted serum metabolomics to distinguish between RR and XX genotypes, suggesting an additional role for the ACTN3 R577X polymorphism in human metabolism. Taken together, these results identify significant cardiometabolic effects associated with possessing one or more functional copies of the ACTN3 gene.

About the Authors:

Chelsea L. Deschamps

Affiliation: Faculty of Kinesiology, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada

Kimberly E. Connors

Affiliation: Faculty of Kinesiology, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada

Matthias S. Klein

Affiliation: Faculty of Kinesiology, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada

Virginia L. Johnsen

Affiliation: Faculty of Kinesiology, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada

Jane Shearer

Affiliations Department of Biochemistry and Molecular Biology, Cumming School of Medicine, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada, Faculty of Kinesiology, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada

Hans J. Vogel

Affiliation: Department of Biochemistry and Molecular Biology, Cumming School of Medicine, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada

Joseph M. Devaney

Affiliation: Children’s National Medical Center, 111 Michigan Ave NW, Washington DC, United States of America

Heather Gordish-Dressman

Affiliation: Children’s National Medical Center, 111 Michigan Ave NW, Washington DC, United States of America

Gina M. Many

Affiliation: Children’s National Medical Center, 111 Michigan Ave NW, Washington DC, United States of America

Whitney Barfield

Affiliation: Children’s National Medical Center, 111 Michigan Ave NW, Washington DC, United States of America

Eric P. Hoffman

Affiliation: Children’s National Medical Center, 111 Michigan Ave NW, Washington DC, United States of America

William E. Kraus

Affiliation: Duke University, 304 Research Drive, Durham, NC, United States of America

Dustin S. Hittel

* E-mail: [email protected]

Affiliation: Department of Biochemistry and Molecular Biology, Cumming School of Medicine, University of Calgary, 2500 University Drive, Calgary, Alberta, T2N 1N4, Canada

Introduction

The sarcomeric α-actinins play an important role in generating skeletal muscle contractions by stabilizing actin thin filaments within the myofibrillar array [1, 2]. In human skeletal muscle α-actinin-2 (encoded by the ACTN2 gene) is expressed in all muscle fiber types whereas α-actinin-3 (ACTN3) is expressed in a subset of fast-twitch glycolytic muscle fibers where it contributes to the generation of rapid contractions [1]. Uniquely, α-actinin-3 expression is completely absent in 18–20% of the global population because of a common nonsense polymorphism in the ACTN3 gene wherein arginine (R) is converted into a stop codon (X) at residue 577 [3]. While the frequency of the ACTN3 577X null allele varies significantly between populations, at least one copy of the ACTN3 R577 allele is beneficial for participation in sprint and power sports [4] whereas the ACTN3 577XX genotype is overrepresented in some endurance athlete cohorts [5] [6]. While numerous studies have confirmed associations between the ACTN3 R577 allele and muscle power phenotypes in humans (i.e sprint time, fiber type) there have been far fewer studies linking the ACTN3 577X allele with endurance phenotypes [5]. Notwithstanding these observations ACTN3 knockout mice exhibit increased recovery from exercise fatigue and a shift towards slow twitch or endurance type skeletal muscle metabolism [7]. It is hypothesized that a shift towards a more efficient metabolism may be responsible for driving the (relatively) recent positive selection of the ACTN3 577X allele, which arose 40,000–60,000 years ago when anatomically modern humans migrated out of Africa [8–11].

Recently a number of studies have described a role for ACTN3 R577X in the aging process and all-cause mortality in humans [12, 13]. Notably, the R577 allele has been associated with increased survival times in patients with chronic heart failure [14] as well as increased bone mineral density/decreased fracture risk [15] suggesting an extra-sarcomeric role for α-actinin-3. While much of the research on ACTN3 has focused on skeletal muscle phenotypes related to power/strength driven sports [16] there have been relatively few studies identifying associations with specific components of cardiovascular and metabolic fitness [12, 17]. Herein we describe the identification of novel associations between the ACTN3 R577X polymorphism, cardiovascular fitness and metabolite biomarkers in healthy young adults.

Materials and Methods

Participants

This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary (Ethics ID: E23521) and is registered under the clinicaltrials.gov identifier NCT00966407. Written informed consent was obtained from all subjects before participation in any testing. All subjects (n = 98 males, n = 102 females) were a part of the Assessing Inherited Markers of Metabolic Syndrome in the Young (AIMMY) Study described previously [18]. Briefly subjects were: (1) between the ages of 18 and 35 years; (2) had completed puberty; and (3) willing and able to provide informed consent. Recruitment occurred at the University of Calgary (UCalgary) main campus using posters, information on campus wide monitors, brief classroom sessions and the university’s website. All eligible, consenting participants were considered to be healthy at the time of enrolment. Health was defined as an absence of: (1) evidence of clinically relevant systemic disease associated with disorders of glucose metabolism; (2) chronic use of glucocorticoid or appetite suppressants; (3) the use of drugs that alter glucose metabolism or other medications known to alter blood levels being tested in this study (ie statins); (4) previous diagnosis or treatment for any hematologic-oncologic disorder; (5) history or current treatment for an eating disorder; (6) current treatment for weight loss; (7) history of bariatric surgery; (8) history of neurosurgical procedure.

Blood Measures

Blood samples were collected in de-identified tubes after an 8–12 hour, overnight fast. Blood for lipoprotein assays (LDL-C, High-Density Lipoprotein Cholesterol (HDL-C), Total Cholesterol (TC), and Triglycerides (TG)) as well as insulin, glucose, C-reactive protein (CRP), and HbA1c was collected using serum stopper tubes containing a clot activator and a silicon gel separator. After collection, samples were spun at 3000 rpm for 10 minute and stored at 2–8°C at UCalgary until being transported to Calgary Lab Services (Calgary, AB) for analysis as described previously [18, 19].

Genotyping

Genomic DNA for genetic analysis was isolated from peripheral blood as described previously [18]. Blood samples were collected in tubes containing an ethylene diamine tetra-ascetic acid (EDTA) anticoagulant and were stored at 2–8°C for a maximum of one week before being sent to the Children’s National Medical Centre (CNMC) in Washington, DC without subject identification. The ACTN3 R577X SNP (rs1815739) was identified using TaqMan allele discrimination assay [20].

Cardiovascular Fitness Assessment

Family history, ethnicity, diet and physical activity levels were recorded by self-report using secure online questionnaires and an iPad as described previously [18, 19]. Hypertension was defined as BP ≥ 140/90 mmHg at two separate time points. Body mass index (BMI) was calculated by dividing the subjects height in meters by their weight in kg². Percent body fat (%BF) and bone mineral density (BMD) was measured using a dual-energy x-ray absorptiometry scan (DXA) (Hologic QDR 4500A scanner, Hologic Inc, Walthan, MA.). Resting heart rate, resting systolic (SBP) and diastolic (DBP) blood pressure, and grip strength (via an Almedic 100kg hand grip dynamometer in the UC cohort (Almedic, Montreal, QC, Canada) were performed. VO₂peak was assessed using the Bruce treadmill protocol as an indicator of cardiovascular fitness [21].

Immunoblotting

Pre-made smooth muscle cell blots (cat. # TB53) were purchased from GBiosciences (St. Louis, MO). Briefly, proteins (50 ug) isolated from smooth muscle cells were solubilized in SDS-lysis buffer and electrophoresed in a 10 well, 4–20% SDS-polyacrylamide gradient gel, followed by electroblotting onto PVDF membrane. Blots were washed with 0.05% Tween-PBS (PBST) and blocked with 5% milk in PBST for 1 hour at room temperature. For examining α-actinin 3 expression a well-characterized polyclonal antibody (a gift from A. Beggs) directed against a region within amino acids 1 and 363 of ACTN3, was used at a 1:1000 dilution and incubated overnight at 4 degrees Celsius. Membranes were washed in PBST buffer and incubated with Goat anti-rabbit secondary antibody (Cell Signaling, Beverly, MA) conjugated to horseradish peroxidase. All blots were subsequently stripped and re-probed with an beta-actin antibody (Cell Signaling) as a loading control as described previously [22].To show the specificity of the α-actinin 3 primary antibody, previously genotyped human skeletal muscle samples (50 ug each) from the STRRIDE study [23] were similarly probed with each antibody. Signals were detected using enhanced chemiluminescence substrate (ThermoScientific, Rockford, IL) and chemiluminescence was digitally captured and quantified using the Chemigenius² BioImaging System (Syngene, Frederick, MD).

Statistical Analysis

All statistical analyses were performed using SPSS Statistics, version 20 (IBM). All data are presented as mean ± SEM. All subjects included an additive genetic model (RR vs. RX vs. XX) used age, sex and ethnicity (Caucasian vs. all others) tested as possible covariates. All models were adjusted for age and sex (unless sex stratified). Analysis of covariance with the Sidak method for post-hoc multiple comparisons adjustment was used for normally distributed outcomes. Quantile regression with the Sidak method for post-hoc multiple comparisons adjustment was used for non-normally distributed outcomes (% Body Fat and CRP) as described previously [24]. All resulting adjusted means are shown as transformed values as those are the numbers used for statistical models. No adjustment for multiple testing was done on the p-values reported here. We tested a minimum of 18 outcomes (considering the total cohort, females only, and males only as separate analyses). This leaves us with a multiple testing adjusted nominal significance level of 0.003; i.e. only p-value ≤0.003 would be considered statistically significant is taking into account the numbers of statistical tests performed. In addition, Pearson correlation coefficients were calculated and the resulting p-values were corrected for multiple testing using false discovery rate (FDR) controlling at a 20% level [25].

LC-MS Metabolite Measurements and Data Analysis

As described in previous studies [26], serum samples were analyzed at Chenomx (Edmonton, AB) using LC-MS AbsoluteIDQ p150 kits (Biocrates Life Sciences, Innsbruck, Austria) employing stable-isotope labeled internal standards. Acylcarnitines, amino acids, sugars, phosphatidylcholines (PCs) and sphingomyelins (SMs) were quantified. As individual sugars were not separated by this method, the hexose values were determined as the sum of all individual sugars. Before statistical analyses, values below the lower limit of quantification (LLOQ) or above the upper limit of quantification (ULOQ) were discarded. For metabolites with only semi quantitative values available, the LLOQ was conservatively estimated as 10 times the limit of detection (LOD), as previously suggested.[Guidelines for the validation and verification of quantitative and qualitative test methods; National Association of Testing Authorities, Australia: Silverwater, NSW, Australia, 2013.] Metabolites with greater than 20% missing values (due to being below LLOQ or above ULOQ) were excluded from analysis. This rendered 87 out of originally 148 metabolites for analysis, including 30 diacyl PCs, 27 acyl-alkyl PCs, 7 lysoPCs, 8SMs, 5 hydroxySMs, 8 amino acids acetylcarnitine and hexose. Missing values were imputed with the half minimum observed value of the respective metabolite. Metabolomics statistical analyses were performed in R 3.02 (R Foundation for Statistical Computing, Vienna, Austria). Support Vector Machines with linear kernels were used to classify RR (CC) and XX (TT) samples as described previously [26]. SVM have been previously shown to be robust classifiers with superior classification performance for metabolomics data sets [27]. To create a balanced sample set for optimal SVM training, a subset of RR samples was randomly chosen to match the number of XX samples. SVM training was repeated for 10 different sets of randomly chosen RR samples, and the average accuracy was calculated. Classification performance was assessed using leave-one-out cross validation. For feature selection, metabolites were sorted according to Mann-Whitney U-test p-values, and SVMs were trained for different numbers of metabolites, and the number with maximum accuracy was chosen.

Results

Subject Characteristics

Two hundred participants (mean age = 23 ± 4.2 y; 98 males and 102 females) from the University of Calgary student population completed all three visits as part of the AIMMY study [18]. The majority (80%) were Caucasian; remaining participants were Asian (16.7%) and Egyptian (3.3%). As a means of ensuring genotyping accuracy we analyzed our AIMMY cohort and found the ACTN3 R577X locus to be in Hardy-Weinberg Equilibrium (HWE) (Table 1) with (p(R) = 0.558; p(X) = 0.442; P = 0.54) as described previously [28].

[Figure omitted. See PDF.]

Table 1. ACTN3 R577X Allele Frequencies and Hardy-Weinberg Equilibrium.

https://doi.org/10.1371/journal.pone.0130644.t001

Fitness Associations

We identified significant associations between ACTN3 genotype, systolic (SBP, p = 0.027) and diastolic blood pressure (DBP, p = 0.005) and VO₂ peak (P = 0.002) where all individuals with the RR genotype achieved greater mean values than those with the RX or XX genotypes (Tables 2, 3). When considering gender differences, the effects of ACTN3 genotype on VO₂peak were driven primarily by our female subjects whereas the effect in males approached statistical significance p = 0.052 (RR vs RX). Similarly, the association of ACTN3 genotype (RR vs RX) with peak HR was only significant in males and not in females or the combined AIMMY cohort. The association of ACTN3 genotype with SBP (RR vs XX) was only significant in the combined cohort. This phenomenon, combining two non-significant cohorts to produce a significant effect is known as Simpson’s paradox, and is the result of cohort traits (i.e. gender differences) that exhibit very different means [29] [26]. There were no significant differences in grip strength max score between ACTN3 genotypes.

[Figure omitted. See PDF.]

Table 2. Association between ACTN3 genotype and cardiometabolic traits.

https://doi.org/10.1371/journal.pone.0130644.t002

[Figure omitted. See PDF.]

Table 3. Significant associations between all ACTN3 genotypes.

https://doi.org/10.1371/journal.pone.0130644.t003

Anthropometric Associations

In female subjects only, total body weight and BMI were consistently and significantly higher (p<0.05) in RX vs XX individuals (Tables 2 & 3). A broader association was observed with DXA derived %BF values between RR and XX, as well as RX vs XX women respectively. As with VO₂peak the effect of ACTN3 genotype on body composition parameters were driven primarily by females.

ACTN3 Expression in Smooth Muscle Cells

The α-actinin 3 polyclonal antibody was first validated against previously genotyped skeletal muscle samples from the STRRIDE study (Fig 1A) which served as both positive (RR) and negative (XX) controls [30]. A ~100 kDa band corresponding to the calculated molecular weight of mature α-actinin 3 was observed in RR but not XX skeletal muscle (Fig 2A). A ~100 kDa band was also observed in pulmonary artery smooth muscle cells but not in smooth muscle cells from Bronchiole, Coronary Artery, Umbilical Artery or Uterus (Fig 1B). The normalized (to B-actin) expression level of α-actinin 3 in pulmonary artery smooth muscle was ~ 5 x lower than in skeletal muscle.

[Figure omitted. See PDF.]

Fig 1. The expression of alpha-actinin-3 in human skeletal muscle and pulmonary artery smooth muscle.

Shown are A) A rabbit polyclonal alpha-actinin 3 antibody probed against skeletal muscle samples from ACTN3 RR577 and 577XX individuals from the STRRIDE Study [30]. B) This same antibody probed against human smooth muscle (SM) cell panel. Each lane contains 50 ug of protein extract, which were stripped and reprobed with a beta-actin antibody as a loading control.

https://doi.org/10.1371/journal.pone.0130644.g001

[Figure omitted. See PDF.]

Fig 2. Heatmap and hierarchical clustering of significantly different metabolites stratified by ACTN3 RR vs XX genotype.

Yellow blocks represent high concentrations; blue blocks represent low concentrations; black blocks represent medium concentrations.

https://doi.org/10.1371/journal.pone.0130644.g002

Cardio-metabolomic Analysis

We have recently found success in identifying in novel genotype-metabolite associations using metabolomic analysis coupled to support vector machine (SVM) analysis [26]. SVMs have been consistently shown to outperform commonly used classification approaches such as partial least-squares discriminant analysis (PLS-DA) for metabolomic samples [26]. Using this approach, our metabolomic data showed classification accuracy when including only female subjects in the analysis, with 73.9±7.5% of all participants genotype (RR vs XX) predicted based on their metabolite profile. This accuracy was reached when using a set of 7 lipid metabolites for classification (Fig 2). While 4 metabolites showed reduced levels (Fig 2, Cluster A) in RR vs XX genotype females, 3 showed elevated levels (Fig 2, Cluster B), namely lyso-PC acyl C18:2, PC diacyl C38:1, and PC diacyl C40:2. Collectively, these lipid metabolites can act as a fuel source or as a cell membrane stabilizing agents. In serum, PC and lyso-PC are an important component of lipoproteins such as high-density lipoprotein (HDL) [31]. The fatty acid composition of PC is mainly determined post-synthesis by hydrolysis and re-acetylation by phospholipases and acyltransferases [32]. As these enzymes are highly active in adipose tissue, the higher % body fat observed in women might amplify the effects. The female-biased metabolomics classification is consistent with the stronger ACTN3 genotype effects observed for both VO₂peak and body composition.

Pearson Correlations and Cluster Analysis of Cardiometabolic Traits

In examining multiple Pearson correlations between the cardiometabolic fitness traits in our AIMMY cohort there is a strong relationship between VO₂peak, DBP, SBP and Peak HR, which cluster together as a block along with grip strength max score (Fig 3). This supports their statistical grouping by ACTN3 genotype (Table 2) and identifies an association between muscle strength and cardiovascular fitness. Further analysis reveals the expected clustering of body mass traits (Weight, BMI, height, BMD) and their negative correlation with HDL-Cholesterol; The negative association of cardiovascular fitness (VO₂peak) with cardiometabolic (%BF, Insulin, Triglycerides and CRP) and lipoprotein (HDL, LDL and Total Cholesterol) panels and the close relationship between glucose homeostasis/insulin resistance traits (Glucose, Insulin and HbA1c).

[Figure omitted. See PDF.]

Fig 3. Pearson correlation matrix of 18 cardiometabolic and anthropometric variables in the University of Calgary AIMMY cohort.

Positive r values are in red and negative values are in blue.

https://doi.org/10.1371/journal.pone.0130644.g003

Discussion

Whereas the ACTN3 R577X polymorphism has been associated with muscle performance phenotypes in elite and amateur power athletes, studies linking it to cardiometabolic fitness are challenged by a lack of mechanistic evidence [5]. Herein we identified significant associations between the ACTN3 ancestral R allele and cardiometabolic fitness in a population of healthy young adults. Specifically we have shown that ACTN3 RR577 individuals exhibited up to 15% higher peak oxygen consumption, VO₂ peak, compared to those of the XX genotype. Furthermore, these differences in VO₂ peak were not associated with self-reported physical activity levels among the 3 ACTN3 genotypes (S1 Table) [19]. These findings were the opposite of what we expected given the preponderance of published research linking the RR and RX genotypes to skeletal muscle power and the XX genotype to endurance sports [2, 6]. However, in reviewing the literature it became clear that alpha-actinin 3 deficiency is not consistently advantageous in endurance athlete cohorts. Indeed, Gomez-Gallego et al. [33] have shown that RR/RX professional road cyclists exhibited significantly higher peak power output and ventilatory threshold than their XX counterparts. Furthermore, the R577 allele is not only overrepresented in Russian rowers but also advantageous to competition results within this same cohort [34].

Previous attempts to identify associations between ACTN3 genotype and cardiovascular fitness tended to be ambiguous in their description of their treadmill protocol, or utilized increases in running speed to change workload [5]. As such, one interpretation of our results is that the genotype effects on VO₂ peak are unmasked by our use of the Bruce treadmill protocol, a widely used clinical test for maximal oxygen consumption that utilizes a rapidly graded incline to increase the workload [35]. It is possible that this high treadmill incline may be sufficient to differentiate between RR and RX individuals possessing a power type muscular phenotype as compared to the α-actinin-3 deficient XX individuals. This is supported by evidence from our Pearson correlation analysis (Fig 3), which identifies an association between muscle strength and cardiovascular fitness in our AIMMY cohort. Consequently, the persistence of the R allele in elite endurance athletes likely reflects the physiological demands of contemporary endurance events wherein forceful muscle contractions are increasingly essential (i.e sprint starts and finishes) [34]. To the best of our knowledge, we are the only group to use the Bruce protocol to test for maximum oxygen consumption in relation to ACTN3 genotype. An alternative interpretation of our VO₂ peak results is that we may be detecting differences in submaximal oxygen consumption in RR/RX vs XX individuals due to intrinsic skeletal muscle fiber type differences. Some studies have reported RR and RX individuals as having higher proportions of type II muscle fibers compared to XX individuals [36]. Skeletal muscle with a high proportion of fast twitch fibers consumes more oxygen than slow twitch dominant muscle at submaximal stimulation despite greater overall oxidative capacity [37]. Given that the Bruce protocol is conducted in the submaximal range it is plausible that R genotype-driven differences in fiber type could be responsible for differences in submaximal exercise oxygen consumption [35].

Notwithstanding these alterative interpretations of our results, individuals with the XX genotype also displayed a significantly lower resting systolic and diastolic blood pressure than RR individuals. Given the close relationship between arterial perfusion and VO₂ peak (Fig 3), we conducted immunoblot analysis and identified the expression of a ~100 kDa immunoreactive band corresponding to alpha-actinin 3 in human pulmonary artery smooth muscle cells (Fig 1). When it was originally identified, the expression of alpha-actinin 3 was presumed to be restricted to skeletal muscle and to a lesser extent, the brain, however recent published evidence [15] suggests a more widespread expression pattern. In addition, a cursory search of the draft (http://www.humanproteomemap.org/) identified peptide sequences unique to alpha-actinin 3 in a variety of fetal and adult human tissues and organs [38]. While our findings need to be validated in situ it is conceivable that changes vascular myogenic tone in response to alpha-actinin-3 deficiency could explain differences in resting blood pressure as well as altered vascular recruitment in response to an acute exercise stress test [39, 40]. Furthermore, it has not escaped our attention that the expression (or lack thereof) of alpha-actinin 3 in pulmonary artery could explain the gene-stratified differences in survival of patients with chronic heart failure [14]. Regardless of the underlying mechanism, these novel associations have clinical relevance given the wide spread use of the Bruce protocol to assess cardiac function and aerobic capacity [35].

Previous studies speculated that differences in androgen hormones could explain the sexual dimorphic effects of the ACTN3 R577X polymorphism on muscle strength characteristics [28, 41]. As with VO₂ peak, we identified female-driven associations between ACTN3 genotype and body fat where females with XX genotype exhibited significantly higher % BF than individuals with the RX or RR genotypes. In fact, according to American College of Sports Medicine (ACSM) guidelines, the %BF values for the RR and RX groups were classified as “low” while the %BF for the XX group were classified as “moderate”[19]. Indeed, within our AIMMY cohort a higher percent body fat negatively correlates with VO₂ peak and grip strength and positively with fasting insulin levels (Fig 3). As such, we believe that the interaction of ACTN3 genotype with cardiometabolic parameters has added clinical significance.

Initially, the absence of significant genotype associations with a clinical biochemical panel (i.e Cholesterol, Triglyceride, Glucose) suggested that the metabolic differences described of ACTN3 deficient mice may not be present in XX humans [7]. However, previous expertise with targeted metabolomics [26, 42] in our laboratory allowed us to identify a panel of 7 lipid metabolites, all phosphatidylcholines, that accurately predicted RR vs XX genotype in female subjects. Phosphatidylcholines are important components of biological membranes that have been previously shown to vary in the serum of obese adults and adolescents [43–45]. Furthermore, increased levels of lyso-PC in RR individuals may indicate elevated breakdown of phosphatidylcholines. Lyso-PC has been shown to mediate numerous physiological and cellular processes such as inflammation [46] and G-protein cell signaling [47]. It is plausible that alterations in the ratios of phosphatidylcholines may be associated with higher % body fat in female subjects or more intriguingly, the latitudinal gradient of the ACTN3 577X allele by contributing to metabolic efficiency and/or cellular membrane stability in colder climates [8]. While the relationship between metabolome and ACTN3 genotype needs to be explored in a larger cohort including clinical populations (i.e heart failure, COPD) ours is the first to report metabolite differences in humans related the ACTN3 genotype.

In summary, we have identified novel associations between the ACTN3 R577X polymorphism peak VO₂ and blood pressure in a population of healthy young adults. We also identified a plausible mechanism by which alpha-actinin 3 deficiency in pulmonary artery smooth muscle may influence both intrinsic aerobic capacity and the pathophysiology of heart failure. Finally, we used serum metabolomics to identify for the first time, a role for the ACTN3 R577X polymorphism in human lipid metabolism. Taken together, these associations suggest that prospective and retrospective ACTN3 genotyping [48] may provide novel insights into human athletic performance and cardiovascular disease risk stratification.

Supporting Information

[Figure omitted. See PDF.]

S1 Table. Comparison of Paffenbarger Survey Scores between ACTN3 genotypes.

Self-reported yearly physical activity scores between ACTN3 genotypes. All analyses were performed by ANCOVA as described previously. No statistically significant differences were identified between ACTN3 genotypes RR, RX or XX (p<0.05).

https://doi.org/10.1371/journal.pone.0130644.s001

(DOCX)

Author Contributions

Conceived and designed the experiments: DH EPH JS. Performed the experiments: CLD KEC MK VLJ GMM WB. Analyzed the data: HG-D MK JMD. Contributed reagents/materials/analysis tools: EPH WEK HJV. Wrote the paper: CLD DSH.

Citation: Deschamps CL, Connors KE, Klein MS, Johnsen VL, Shearer J, Vogel HJ, et al. (2015) The ACTN3 R577X Polymorphism Is Associated with Cardiometabolic Fitness in Healthy Young Adults. PLoS ONE 10(6): e0130644. https://doi.org/10.1371/journal.pone.0130644

References

1. Mills M, Yang N, Weinberger R, Vander Woude DL, Beggs AH, Easteal S, et al. Differential expression of the actin-binding proteins, alpha-actinin-2 and -3, in different species: implications for the evolution of functional redundancy. Human molecular genetics. 2001;10(13):1335–46. Epub 2001/07/07. pmid:11440986.

2. Ma F, Yang Y, Li X, Zhou F, Gao C, Li M, et al. The association of sport performance with ACE and ACTN3 genetic polymorphisms: a systematic review and meta-analysis. PloS one. 2013;8(1):e54685. Epub 2013/01/30. pmid:23358679; PubMed Central PMCID: PMC3554644.

3. North KN, Yang N, Wattanasirichaigoon D, Mills M, Easteal S, Beggs AH. A common nonsense mutation results in alpha-actinin-3 deficiency in the general population. Nature genetics. 1999;21(4):353–4. Epub 1999/04/07. pmid:10192379.

4. Bouchard C, Hoffman EP, Commission IM. Genetic and molecular aspects of sports performance: Wiley Online Library; 2011.

5. Silva MS, Bolani W, Alves CR, Biagi DG, Lemos JR, da Silva JL, et al. Influences of ACTN3 R577X Variant in Oxygen Uptake are Eliminated by Endurance Training in Healthy Individuals. International journal of sports physiology and performance. 2015. Epub 2015/01/09. https://doi.org/10.1123/ijspp.2014-0205 pmid:25569611.

6. Alfred T, Ben-Shlomo Y, Cooper R, Hardy R, Cooper C, Deary IJ, et al. ACTN3 genotype, athletic status, and life course physical capability: meta-analysis of the published literature and findings from nine studies. Human mutation. 2011;32(9):1008–18. Epub 2011/05/05. pmid:21542061; PubMed Central PMCID: PMC3174315.

7. MacArthur DG, Seto JT, Chan S, Quinlan KG, Raftery JM, Turner N, et al. An Actn3 knockout mouse provides mechanistic insights into the association between alpha-actinin-3 deficiency and human athletic performance. Human molecular genetics. 2008;17(8):1076–86. Epub 2008/01/08. pmid:18178581.

8. Friedlander SM, Herrmann AL, Lowry DP, Mepham ER, Lek M, North KN, et al. ACTN3 allele frequency in humans covaries with global latitudinal gradient. PloS one. 2013;8(1):e52282. Epub 2013/01/30. pmid:23359641; PubMed Central PMCID: PMC3554748.

9. Sankararaman S, Mallick S, Dannemann M, Prufer K, Kelso J, Paabo S, et al. The genomic landscape of Neanderthal ancestry in present-day humans. Nature. 2014;507(7492):354–7. Epub 2014/01/31. pmid:24476815; PubMed Central PMCID: PMC4072735.

10. Head SI, Chan S, Houweling PJ, Quinlan KG, Murphy R, Wagner S, et al. Altered Ca2+ Kinetics Associated with alpha-Actinin-3 Deficiency May Explain Positive Selection for ACTN3 Null Allele in Human Evolution. PLoS genetics. 2015;11(2):e1004862. Epub 2015/01/16. pmid:25590636; PubMed Central PMCID: PMC4295894.

11. Amorim CE, Acuna-Alonzo V, Salzano FM, Bortolini MC, Hunemeier T. Differing Evolutionary Histories of the ACTN3*R577X Polymorphism among the Major Human Geographic Groups. PloS one. 2015;10(2):e0115449. Epub 2015/02/24. pmid:25706920.

12. Fiuza-Luces C, Ruiz JR, Rodriguez-Romo G, Santiago C, Gomez-Gallego F, Yvert T, et al. Are 'endurance' alleles 'survival' alleles? Insights from the ACTN3 R577X polymorphism. PloS one. 2011;6(3):e17558. Epub 2011/03/17. pmid:21407828; PubMed Central PMCID: PMC3048287.

13. Seto JT, Chan S, Turner N, MacArthur DG, Raftery JM, Berman YD, et al. The effect of alpha-actinin-3 deficiency on muscle aging. Experimental gerontology. 2011;46(4):292–302. Epub 2010/11/30. pmid:21112313.

14. Bernardez-Pereira S, Santos PC, Krieger JE, Mansur AJ, Pereira AC. ACTN3 R577X polymorphism and long-term survival in patients with chronic heart failure. BMC cardiovascular disorders. 2014;14:90. Epub 2014/07/26. pmid:25059829; PubMed Central PMCID: PMC4113663.

15. Yang N, Schindeler A, McDonald MM, Seto JT, Houweling PJ, Lek M, et al. alpha-Actinin-3 deficiency is associated with reduced bone mass in human and mouse. Bone. 2011;49(4):790–8. Epub 2011/07/26. pmid:21784188.

16. Eynon N, Hanson ED, Lucia A, Houweling PJ, Garton F, North KN, et al. Genes for elite power and sprint performance: ACTN3 leads the way. Sports medicine. 2013;43(9):803–17. Epub 2013/05/18. pmid:23681449.

17. Grealy R, Smith CL, Chen T, Hiller D, Haseler LJ, Griffiths LR. The genetics of endurance: frequency of the ACTN3 R577X variant in Ironman World Championship athletes. Journal of science and medicine in sport / Sports Medicine Australia. 2013;16(4):365–71. Epub 2012/10/25. pmid:23092649.

18. Karlos A, Shearer J, Gnatiuk E, Onyewu C, Many G, Hoffman EP, et al. Effect of the SORT1 low-density lipoprotein cholesterol locus is sex-specific in a fit, Canadian young-adult population. Appl Physiol Nutr Metab. 2013;38(2):188–93. Epub 2013/02/27. pmid:23438231.

19. Many GM, Lutsch A, Connors K, Shearer J, Brown HC, Ash G, et al. Examination of Lifestyle Behaviors and Cardiometabolic Risk Factors in University Students Enrolled in Kinesiology Degree Programs. Journal of strength and conditioning research / National Strength & Conditioning Association. 2015. Epub 2015/02/04. https://doi.org/10.1519/JSC.0000000000000871 pmid:25647655.

20. Livak KJ. Allelic discrimination using fluorogenic probes and the 5'nuclease assay. Genetic Analysis: Biomolecular Engineering. 1999;14(5–6):143–9.

21. Bruce R.A., Kusumi F, Hosmer D. Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. American heart journal. 1973;85:545–62.

22. Watts R, Ghozlan M, Hughey CC, Johnsen VL, Shearer J, Hittel DS. Myostatin inhibits proliferation and insulin-stimulated glucose uptake in mouse liver cells. Biochemistry and cell biology = Biochimie et biologie cellulaire. 2014;92(3):226–34. Epub 2014/06/03. pmid:24882465.

23. Hittel DS, Kraus WE, Tanner CJ, Houmard JA, Hoffman EP. Exercise training increases electron and substrate shuttling proteins in muscle of overweight men and women with the metabolic syndrome. J Appl Physiol (1985). 2005;98(1):168–79. Epub 2004/09/07. pmid:15347626.

24. Sprouse C, Gordish-Dressman H, Orkunoglu-Suer EF, Lipof JS, Moeckel-Cole S, Patel RR, et al. SLC30A8 nonsynonymous variant is associated with recovery following exercise and skeletal muscle size and strength. Diabetes. 2014;63(1):363–8. Epub 2013/10/09. pmid:24101675.

25. Reiner A, Yekutieli D, Benjamini Y. Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics. 2003;19(3):368–75. Epub 2003/02/14. pmid:12584122.

26. Klein MS, Connors KE, Shearer J, Vogel HJ, Hittel DS. Metabolomics reveals the sex specific effects of the SORT1 low-density lipoprotein cholesterol locus in healthy young adults. Journal of proteome research. 2014. Epub 2014/09/04. https://doi.org/10.1021/pr500659r pmid:25182463.

27. Hochrein J, Klein MS, Zacharias HU, Li J, Wijffels G, Schirra HJ, et al. Performance evaluation of algorithms for the classification of metabolic 1H NMR fingerprints. Journal of proteome research. 2012;11(12):6242–51. Epub 2012/11/03. pmid:23116257.

28. Clarkson PM, Devaney JM, Gordish-Dressman H, Thompson PD, Hubal MJ, Urso M, et al. ACTN3 genotype is associated with increases in muscle strength in response to resistance training in women. J Appl Physiol (1985). 2005;99(1):154–63. Epub 2005/02/19. pmid:15718405.

29. Hernan MA, Clayton D, Keiding N. The Simpson's paradox unraveled. International journal of epidemiology. 2011;40(3):780–5. Epub 2011/04/02. pmid:21454324; PubMed Central PMCID: PMC3147074.

30. Slentz CA, Duscha BD, Johnson JL, Ketchum K, Aiken LB, Samsa GP, et al. Effects of the amount of exercise on body weight, body composition, and measures of central obesity: STRRIDE—a randomized controlled study. Archives of internal medicine. 2004;164(1):31–9. Epub 2004/01/14. pmid:14718319.

31. Cole LK, Vance JE, Vance DE. Phosphatidylcholine biosynthesis and lipoprotein metabolism. Biochimica et biophysica acta. 2012;1821(5):754–61. Epub 2011/10/08. pmid:21979151.

32. Lands WE. Stories about acyl chains. Biochimica et biophysica acta. 2000;1483(1):1–14. Epub 1999/12/22. pmid:10601692.

33. Gomez-Gallego F, Santiago C, Gonzalez-Freire M, Muniesa CA, Fernandez Del Valle M, Perez M, et al. Endurance performance: genes or gene combinations? International journal of sports medicine. 2009;30(1):66–72. Epub 2008/07/25. pmid:18651373.

34. Ahmetov II, Druzhevskaya AM, Astratenkova IV, Popov DV, Vinogradova OL, Rogozkin VA. The ACTN3 R577X polymorphism in Russian endurance athletes. British journal of sports medicine. 2010;44(9):649–52. Epub 2008/08/23. pmid:18718976.

35. Shah BN. On the 50th anniversary of the first description of a multistage exercise treadmill test: re-visiting the birth of the 'Bruce protocol'. Heart. 2013;99(24):1793–4. Epub 2013/05/18. pmid:23680884.

36. Pereira A, Costa AM, Leitao JC, Monteiro AM, Izquierdo M, Silva AJ, et al. The influence of ACE ID and ACTN3 R577X polymorphisms on lower-extremity function in older women in response to high-speed power training. BMC geriatrics. 2013;13:131. Epub 2013/12/10. pmid:24313907; PubMed Central PMCID: PMC4029788.

37. Kushmerick MJ, Meyer RA, Brown TR. Regulation of oxygen consumption in fast- and slow-twitch muscle. The American journal of physiology. 1992;263(3 Pt 1):C598–606. Epub 1992/09/01. pmid:1415510.

38. Kim MS, Pinto SM, Getnet D, Nirujogi RS, Manda SS, Chaerkady R, et al. A draft map of the human proteome. Nature. 2014;509(7502):575–81. Epub 2014/05/30. pmid:24870542.

39. Heinonen I. Muscle-specific functional sympatholysis in humans. Experimental physiology. 2014;99(2):344–5. Epub 2014/02/04. pmid:24487247.

40. Thomas GD. Functional sympatholysis in hypertension. Autonomic neuroscience: basic & clinical. 2015;188:64–8. Epub 2014/12/03. pmid:25458424; PubMed Central PMCID: PMC4336594.

41. Zempo H, Tanabe K, Murakami H, Iemitsu M, Maeda S, Kuno S. Age differences in the relation between ACTN3 R577X polymorphism and thigh-muscle cross-sectional area in women. Genetic testing and molecular biomarkers. 2011;15(9):639–43. Epub 2011/04/16. pmid:21491997.

42. Connors KE, Karlos AE, Gnatiuk EA, Shearer J, Reimer RA, Hittel DS. SORT1 Protective Allele is Associated with Attenuated Postprandial Lipaemia in Young Adults. Circulation Cardiovascular genetics. 2014. Epub 2014/07/22. https://doi.org/10.1161/circgenetics.114.000534 pmid:25042869.

43. Wishart DS, Tzur D, Knox C, Eisner R, Guo AC, Young N, et al. HMDB: the Human Metabolome Database. Nucleic acids research. 2007;35(Database issue):D521–6. Epub 2007/01/05. pmid:17202168; PubMed Central PMCID: PMC1899095.

44. Reinehr T, Wolters B, Knop C, Lass N, Hellmuth C, Harder U, et al. Changes in the serum metabolite profile in obese children with weight loss. European journal of nutrition. 2015;54(2):173–81. Epub 2014/04/18. pmid:24740590.

45. Wahl S, Yu Z, Kleber M, Singmann P, Holzapfel C, He Y, et al. Childhood obesity is associated with changes in the serum metabolite profile. Obesity facts. 2012;5(5):660–70. Epub 2012/10/31. pmid:23108202.

46. Olofsson KE, Andersson L, Nilsson J, Bjorkbacka H. Nanomolar concentrations of lysophosphatidylcholine recruit monocytes and induce pro-inflammatory cytokine production in macrophages. Biochemical and biophysical research communications. 2008;370(2):348–52. Epub 2008/03/29. pmid:18371300.

47. Xu Y. Sphingosylphosphorylcholine and lysophosphatidylcholine: G protein-coupled receptors and receptor-mediated signal transduction. Biochimica et biophysica acta. 2002;1582(1–3):81–8. Epub 2002/06/19. pmid:12069813.

48. Schadock I, Schneider A, Silva ED, Buchweitz MR, Correa MN, Pesquero JB, et al. Simple Method to Genotype the ACTN3 r577x Polymorphism. Genetic testing and molecular biomarkers. 2015. Epub 2015/04/02. https://doi.org/10.1089/gtmb.2014.0299 pmid:25831089.