Article Text

Download PDFPDF
Polymorphisms in the IGF1 signalling pathway including the myostatin gene are associated with left ventricular mass in male athletes
  1. Ruth-Jessica Karlowatz1,
  2. Jürgen Scharhag2,3,
  3. Jörg Rahnenführer4,
  4. Ulrich Schneider5,
  5. Ernst Jakob5,
  6. Wilfried Kindermann2,
  7. Klaus Dieter Zang1
  1. 1Institute of Human Genetics, University of Saarland/IGD Saar GmbH, Homburg/Saar, Germany
  2. 2Institute of Sports and Preventive Medicine, University of Saarland, Saarbrücken, Germany
  3. 3University Outpatient Clinic, Centre for Sports Medicine, University Potsdam, Germany
  4. 4Fakultät Statistik, Technische Universität Dortmund, Dortmund, Germany
  5. 5Department of Sports Medicine, Sportsclinic Hellersen, Lüdenscheid, Germany
  1. Correspondence to PD Dr.med. Jürgen Scharhag, University Outpatient Clinic, Centre for Sports Medicine, University Potsdam, Am Neuen Palais 10, Haus 12, 14469 Potsdam, Germany; scharhag{at}uni-potsdam.de

Abstract

Background Athlete's heart as an adaptation to long-time and intensive endurance training can vary considerably between individuals. Genetic polymorphisms in the cardiological relevant insulin-like growth factor 1 (IGF1) signalling pathway seem to have an essential influence on the extent of physiological hypertrophy.

Objective Analysis of polymorphisms in the genes of IGF1, IGF1 receptor (IGF1R) and the negative regulator of the cardiac IGF1 signalling pathway, myostatin (MSTN), and their relation to left ventricular mass (LVM) of endurance athletes.

Methods In 110 elite endurance athletes or athletes with a high amount of endurance training (75 males and 35 females) and 27 male controls, which were examined by echocardiographic imaging methods and ergometric exercise-testing, the genotypes of a cytosine–adenine repeat polymorphism in the promoter region of the IGF1 gene and a G/A substitution at position 3174 in the IGF1R gene were determined. Additionally, a mutation screen of the MSTN gene was performed.

Results The polymorphisms in the IGF1 and the IGF1R gene showed a significant relation to the LVM for male (IGF1: p=0.003; IGF1R: p=0.01), but not for female athletes. The same applies to a previously unnoticed polymorphism in the 1 intron of the MSTN gene, whose deletion allele (AAA→AA) appears to increase the myostatic effect (p=0.015). Moreover, combinations of the polymorphisms showed significant synergistic effects on the LVM of the male athletes.

Conclusions The authors' results argue for the importance of polymorphisms in the IGF1 signalling pathway in combination with MSTN on the variant degree of physiological hypertrophy of male athletes.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Highly trained endurance athletes may show several morphological cardiac changes in response to the enhanced haemodynamic load, including an increase in heart volume by the balanced increase in left and right myocardial masses and end-diastolic volumes, commonly described as athlete's heart.1 This physiological cardiac hypertrophy is characterised by a normal or enhanced cardiac function and with maintenance of normal architecture and organisation of cardiac structure.2 In mammals, at birth or soon after, apparently most cardiac myocytes lose their ability to proliferate, and physiological growth of the adult heart occurs primarily as a result of an increase in cardiomyocyte size.3

What is already known on this topic

  • Athlete's heart as an adaptation to long-time intensive endurance training shows considerable individual differences, and genetic disposition seems to play an important role.

What this study adds

  • Our results argue for the importance of polymorphisms in the insulin-like growth factor 1 signalling pathway in combination with myostatin on the variant degree of physiological hypertrophy of male athletes.

One of the most important growth factors for the postnatal growth of the heart is the insulin-like growth factor 1 (IGF1). IGF1 is predominantly bound to specific binding proteins in plasma and tissues, which apparently regulate the effective concentration as well as the biological activity of IGF1.4 5 The IGF1 signalling pathway, which is initiated upon binding of IGF1 to the IGF1 receptor (IGF1R), stimulates glucose metabolism and cell growth, and prevents apoptosis.6 Transgenic mice with enhanced cardiac IGF1 signalling developed a cardiac hypertrophy with normal life span and normal or enhanced cardiac function.7,,9 Furthermore, cardiac IGF1 expression was reported to be higher in athletes than in control subjects.10 These data suggest that the IGF1 signalling pathway is important for physiological growth of the heart.

On the other hand, Shyu et al11 identified myostatin (MSTN) as a negative regulator of the cardiac IGF1 signalling pathway. Primarily, MSTN was identified as a potent inhibitor of skeletal muscle growth,12 but MSTN mRNA and protein are also expressed in fetal and adult cardiomyocytes.13 Furthermore, there is evidence that MSTN could play an important role in cardiomyocyte growth.11 13,,16 Therefore, IGF1 and IGF1R as well as MSTN seem to have essential relevance in both the induction and inhibition of cardiac hypertrophy.

However, not all endurance athletes develop an athlete's heart. It has been reported that the amount of training should account for only 11% of the variable increase in left ventricular mass (LVM),17 suggesting a high importance of individual genetic disposition. Polymorphisms of the DNA sequence of the genes of IGF1, IGF1R and MSTN, which may modify the expression or function of their proteins, could be of decisive influence on the individual hypertrophic response of the heart. Interestingly, two polymorphisms seem to influence the serum IGF1 levels: One of them is a short but highly polymorphic microsatellite sequence composed of a variable number of cytosine-adenosine (CA) repeats in the promoter region of the IGF1 gene. In the literature, its relation to the IGF1 level was discussed with conflicting results: The 19 allele was found by several groups to be associated with the lowest serum IGF1 level.18 19 In contrast, in a large Dutch cohort study, non-carriers of the 19 allele (≠19/≠19) had lower serum IGF1 concentrations than carriers,20 and Allen et al21 even found no association between this polymorphism and serum IGF1 concentration. In Caucasian populations, the number of (CA) repeats ranges from 10 to 23. The most common (‘wild type’) allele has 19 repeats. The homozygous absence of the 19 allele was reported to be associated with an increased risk of type 2 diabetes mellitus, myocardial infarction and the development of a left ventricular hypertrophy.20 22

The other polymorphism which seems to affect the IGF1 plasma level is located in the gene of the IGF1R: a G◊A transition in the exon16 at position 3174, codon 1013. Bonafè et al23 showed that the A allele was associated with lower plasma IGF1 levels than the G allele. In the MSTN gene, Schuelke et al found a rare loss-of-function mutation within intron 1 (a missplicing change: IVS1+5 G>A), which has been associated with muscle hypertrophy in a child.24 Regarding cardiac hypertrophy, so far there have been no reports on functional polymorphisms in the MSTN gene.

The aim of our study was a combined analysis of the two polymorphisms in the IGF1 and the IGF1R gene, and of the complete coding sequence of the MSTN gene, regarding possible implications on heart size in elite endurance athletes.

Materials and methods

Study population

A group of 110 elite athletes (35 females (f) and 75 males (m); competitive athletes) and 27 male untrained controls (c) were studied. Each athlete had undergone training programmes for periods of 8±4 years, with averaged 14±5 h endurance training per week. Athletes participated in long-distance running (n=10), badminton (n=2), biathlon (n=2), cross-country skiing (n=25), cycling (n=9), rowing (n=21), swimming (n=15), triathlon (n=25) and soccer (n=1). The mean ages were as follows: f: 20±3 years; m: 23±5 years; c: 26±4 years; mean weight: f: 66±6 kg; m: 73±8 kg; c: 72±8 kg; and mean height: f: 175±6 cm; m: 181±8 cm; c: 178±6 cm. Subjects were tested by stepwise incremental cycle ergometry or by treadmill ergometry until exhaustion. None of the athletes and controls had a history of hypertension or coronary artery disease, and ECG at rest and during exercise was without pathologies. No medications with an influence on the cardiovascular system or anabolic steroids were taken by any of the subjects. All athletes gave their written informed consent to take part in the study, which was approved by the ethics committee of the Ärztekammer des Saarlandes, Germany.

Echocardiographic examinations

Left ventricular internal diameters and wall thickness were measured in the M-mode according to Sahn et al25 and Lang et al.26 LVM was calculated according to the corrected equation of Teichholz,27 and indexed to body surface area to yield the left ventricular mass index (LVMI) (table 1).

Table 1

Echocardiographic parameters

Determination of genotypes

Venous blood (≥5 ml; EDTA) was collected and stored at −80°C until genomic DNA was extracted.

Genotyping of the CA repeat polymorphism of the IGF1 gene was performed as follows: PCR according to Vaessen et al,20 but using single wells instead of well plates and a final volume of 25 µl with 50–100 ng of genomic DNA. Forward primer was labelled with JOE. The size of PCR products was determined by autosequencer (ABI 310) in comparison with internal lane standard 600 (ILS600; Promega, Madison, Wisconsin). The results were divided into three genotypes: homozygous for the 19 allele (19/19), heterozygous for the 19 allele and any other allele (19/≠19) and homozygous or heterozygous for any non-19 allele (≠19/≠19).

The IGF1R polymorphism was determined by restriction fragment-length polymorphism analysis: PCR products were amplified with primers as described by Balogh et al.28 The PCR was carried out for 35 cycles, each consisting of 60 s for denaturing at 95°C, 20 s for annealing at 56°C and 60 s for extension at 72°C. After digestion by 3U MnlI restriction enzyme, allele A yielded two restriction products of 123 and 84 bp; allele G yielded three restriction products of 103, 84 and 20 bp.

In order to obtain the whole coding sequence of MSTN, we split the three exons into nine fragments. Additionally we analysed the region around the IVS1+5 G>A mutation described before.24 First, we screened only 20 subjects for all 10 fragments. The PCRs followed the cycling profile were as follows: 5 min denaturation at 95°C and 35 cycles, each consisting of 60 s for denaturing at 95°C, 45 s for annealing at 55°C (fragments Ex1a, 1b, 2a, 2b, 3a, 3b, 3c, 3d, 3e) or 30 s for annealing at 60°C (fragment IVS1) and 60 s for elongation at 72°C. The screen was performed using denaturing high-performance liquid chromatography (DHPLC analysis on a WAVE MD Mutation Detection System (Transgenomic, Glasgow, UK). Prior to DHPLC, aliquots of untreated PCR products were denatured at 95°C for 10 min followed by gradual reannealing to 64°C over 30 min, allowing the formation of heteroduplexes in the PCR products. The temperatures of the oven for optimal heteroduplex separation (table 2) were obtained using the Navigator software 1.4.2 (Transgenomic). Amplicons with abnormal DHPLC elution profiles were analysed with forward and reverse sequencing using the Big Dye Terminator Cycle sequencing kit (Applied Biosystems, Foster City, California) on an ABI 310 Genetic Analyser, following the manufacturer's directions. The only polymorphic fragment, IVS1, was analysed for the whole study group with DHPLC.

Table 2

PCR primers for the amplification of the three exons and flanking intron regions of the myostatin gene as well as the analysis temperatures for the denaturing high-performance liquid chromatography

Statistical analysis

Data were represented as mean±SD. Statistical analyses were carried out using the statistical programming language R.29 Mean changes in the LVMI values between different genotype groups were modelled using analysis of variance (ANOVA). In these models, measures for LVM were dependent variables, and genotypes, age of the athletes, hours of training per week and years of active sports training were independent variables. No interaction terms were included in the models. Reported mean values for echocardiographic variables dependent on genotype values are thus adjusted for the effect of age, training hours and years of sports training. χ2 Tests were used to assess the fits of observed genotype frequencies with the Hardy–Weinberg equilibrium. p Values below p=0.05 were regarded as statistically significant. p Values are reported with at least two significant digits and corresponding effect sizes. The ‘coefficient of determination’ (r2) was used to assess correlation in ANOVA models and to study the correlation between variables.

Results

‘New’ MSTN polymorphism

We frequently found a point mutation (rs11333758), a deletion of one of three adenines at position 88–90 bp in intron1 (IVS1+88_90delA, denomination according to den Dunnen and Antonarakis30) of the MSTN gene. To our knowledge, there have been no reports so far on the functional relevance of this polymorphism. In the following, the genotypes will be referred to as homozygous for the wild type (A/A), heterozygous for the deletion (A/–) and homozygous for the deletion (–/–).

Subjects molecular characteristics

Genotyping of all three polymorphisms was successful in all participants of the study. The genotype frequencies of the IGF1, IGF1R and MSTN polymorphisms (table 3) are in Hardy–Weinberg equilibrium in the whole population with the two exceptions of the group of female athletes: related to the IGF1 polymorphism, carriers heterozygous for the 19 allele (19/≠19) were under-represented (p<0.01); with regard to the MSTN polymorphism, carriers homozygous for the deletion allele (–/–) were over-represented (p<0.01).

Table 3

Genotype and allele frequencies (n (%)) in our sample

Echocardiographic findings

LVM in the study group was significantly associated with the amount of endurance training (mean: +0.9±2 g/m2/h/week; p=0.0005), gender (mean: −10.2±17 g/m2 for female gender; p=0.0015) and age (mean: +1.3±3 g/m2/year; p=0.0001). No relation was found to years participating in competitive sport (p=0.2).

The values of LVMI for the IGF1, IGF1R and MSTN genotypes are shown in table 4.

Table 4

Left ventricular-mass index genotypes

In the group of male endurance athletes (m), there were significant differences between genotypes of all three polymorphisms, but not in the two other groups (f, c).

Homozygous carriers of the 19 allele (19/19) had significantly lower LVM than non-carriers of the IGF1 19 allele (≠19/≠19) (effect: −16.4±16 g/m2; p=0.003; r2=0.32). Controls and female athletes showed no statistically significant differences (p>0.2) (figure 1).

Figure 1

Correlation between insulin growth factor 1 cytosine-adenosine n repeat polymorphism and left ventricular mass index in the whole study group (c, male controls; f, female athletes; m, male athletes). The data are adjusted for age and training. Balks are mean.

Furthermore, we found a relation between the IGF1R polymorphism and LVM in our population (figure 2)

Figure 2

Correlation between insulin growth factor receptor polymorphism 3174G/A and left ventricular mass index in the whole study group (c, male controls; f, female athletes; m, male athletes). The data are adjusted for age and training. Balks are mean.

Male athletes who carried the G/G genotype showed a significantly lower LVM than those with A/A (effect: −12.8±19 g/m2; p=0.01; r2=0.30). Again, we could not find any significant differences in the groups of controls and female athletes (p>0.1).

The results of the analysis of the previously unnoticed MSTN polymorphism (IVS1+88_90delA) with regard to LVM are shown in figure 3.

Figure 3

Correlation between myostatin polymorphism IVS1+88_90delA and left ventricular mass index in the whole study group (c, male controls; f, female athletes; m, male athletes). The data are adjusted for age and training. Balks are mean. MSTN, myostatin.

In the group of male athletes, carriers heterozygous for deletion (A/–) had a significantly lower LVM than carriers homozygous for wild type (A/A) (effect: −10.3±16 g/m2; p=0.015; r2=0.29). Indeed, carriers homozygous for deletion (–/–) also showed a lower LVM than the A/A carriers but this was not significant. In turn, controls and female athletes showed no statistically significant differences (p>0.3).

Synergistic effects

We additionally looked for potential synergistic effects between the three investigated polymorphisms on LVM. Because of the insufficient number of cases for controls and for female athletes, we analysed only the group of male athletes. Table 5 contains the results of the statistical analysis.

Table 5

Synergistic effects of the insulin-like growth factor 1 (IGF1), IGF1 receptor and myostatin (MSTN) polymorphisms on left ventricular mass index (LVMI) in our population of male athletes

We found a significant association of distinct genotype combinations of IGF1 and MSTN gene polymorphisms as well as of IGF1R and MSTN gene polymorphisms, but not between IGF1 and IGF1R gene polymorphisms with LVM. Male athletes with IGF1 19/19 and MSTN A/– genotypes had a significant lower LVM than non-carriers of the combination (effect: −16.6±19 g/m2; p=0.003). This effect was stronger than those of the single genotypes. Carriers of the combination IGF1R G/G and MSTN A/– also had a lower LVM than non-carriers (effect: −15.2±19 g/m2; p=0.01), and the values were even lower than those of the single genotypes.

Discussion

The novel result of this cross-sectional study was the strong association of athlete's heart with genetic polymorphisms in the IGF1 signalling pathway. Our findings support previous studies which demonstrated that the signalling pathway of IGF1 plays an important role in regulating a physiological cardiac hypertrophy.31 Admittedly, our study groups were rather small and consisted of about 70% of male athletes. However, it is difficult to recruit large numbers of elite endurance athletes with standardised, controlled training conditions. Nevertheless, we could demonstrate for our male elite endurance athletes a significant association between the alternative molecular forms of all three polymorphisms and the extent of left ventricular hypertrophy assessed by LVMI.

In this study, male athletes homozygous for the 19 allele in the promoter region of the IGF1 gene (19/19) showed a significantly lower LVM than non-carriers of the 19 allele (≠19/≠19). A comparable result was observed for homozygous carriers of the G allele (G/G) of the IGF1R polymorphism in comparison with homozygous carriers of the A allele (A/A). Surprisingly, also in the MSTN gene, the previously unnoticed IVS1+88_90delA polymorphism amidst the intron 1 was significantly associated with the LVM: carriers heterozygous for the deletion (A/–) had a smaller LVM than A/A carriers. Homozygous (–/–) athletes were too few for a statistical evaluation.

These data are based on DNA findings and not on expression data. Serum levels do not reflect correctly the tissue concentrations of the components of the IGF1 signalling pathway, which is largely autonomously regulated in different tissues and also in the heart.5 Since cardiac biopsies from our subjects were not feasible, we do not know to which extent the polymorphisms influence these cardiac protein levels. Therefore, at the moment, our results can only be interpreted as follows: if the wild type allele (19) is associated indeed with a lower IGF1 level,18 19 this should reduce training induced growth response of the cardiomyocytes in homozygous carriers (19/19). If, on the contrary, the wild type allele should result in a higher IGF1 level,20 homozygous (19/19) carriers should reach increased threshold levels for IGF1 causing an activation of MSTN which would attenuate cardiac hypertrophy.

For the ‘silent’ IGF1R 3174G/A mutation, there is less controversial evidence about its influence on the IGF1 level: the A allele seems to be associated with a lower serum IGF1 level.23 Nevertheless, in our study, the A/A genotype was correlated with a higher LVM. This could be connected with the known ability of the IGF1R to reduce the IGF1 expression in a negative-feedback mechanism.28 Therefore, the A allele may be associated with a higher activity of the IGF1R, maybe via a linkage disequilibrium, indeed resulting in an increased negative feedback suppressing the IGF1 expression but at the same time causing an enhanced intracellular signalling, which may lead to an increased growth response of the cardiomyocytes in homozygous carriers (A/A).

The functional relevance of the IVS1+88_90delA polymorphism in the MSTN gene has not yet been published. Our results indicate an increased MSTN effect for the deletion allele. This means that carriers of at least one deletion allele may show an attenuated training induced growth response of the heart resulting in a lesser LVM increase. It is unclear whether the MSTN polymorphism—although untranscribed—is a functional variant or a marker for other functional variants influencing LVM. In any case, it should be noticed that this MSTN polymorphism is located in an evolutionary highly conserved sequence.

Interestingly, there was an unexpected gender difference for the effect of the three analysed polymorphisms on LVM, with almost no effect in female athletes. Admittedly, our female group is too small for any detailed statistical evaluation. However, our results and also the published studies indicate a sex hormonal influence on the IGF1 signalling pathway,32,,34 which could explain at least in part the sex differences. In the male athletes group, it seems evident that interactions of genotypes of the IGF1 signalling pathway increase the influence on cardiac hypertrophy. A positive (additive) association of IGF1 and IGF1R genotypes was not detected in this study. This may result from the negative feedback between IGF1 and its receptor. However, IGF1 and MSTN genotypes as well as IGF1R and MSTN genotypes seemed to exert synergistic effects on LVM, confirming the assumption of an interaction between the three proteins in a network. In the future, this novel and important finding deserves further validation on larger groups of athletes, whereat the determination of cardiac protein levels might reveal additional important information.

Information box

  • Athlete's heart as an adaptation to long-time intensive endurance training shows considerable individual differences. Genetic disposition seems to play an important role. Our results argue for the importance of polymorphisms in the insulin-like growth factor 1 signalling pathway in combination with myostatin on the variant degree of physiological hypertrophy of male athletes.

Conclusion

The genetic disposition with respect to the IGF1 signalling pathway seems to play a crucial role in the development of an athlete's heart in males. We consider our study as the basis for a larger and, if possible, prospective study with a more detailed analysis of patterns of molecular variation in genes involved in the physiological and maybe also in the pathological hypertrophy of the heart. A future goal is to identify genetic patterns with predictive value for the extent of left ventricular hypertrophy, both for athletes and for patients.

References

Footnotes

  • Funding The study was supported by a grant from the Landesforschungsförderungsprogramm of the Saarland (Grant D4-14.2.1.1 (LFFP 0430)) and the Landessportverband für das Saarland (LSVS).

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by the ethic committee of the Ärztekammer des Saarlandes, Germany (see Materials and Methods).