Int J Sports Med
DOI: 10.1055/a-1198-5496
Review

Mapping Robust Genetic Variants Associated with Exercise Responses

Javier Alvarez-Romero
1  Institute for Health and Sport, Victoria University, Melbourne, Australia
,
Sarah Voisin
1  Institute for Health and Sport, Victoria University, Melbourne, Australia
,
Nir Eynon
1  Institute for Health and Sport, Victoria University, Melbourne, Australia
2  MCRI, Murdoch Childrens Research Institute, Parkville, Australia
,
1  Institute for Health and Sport, Victoria University, Melbourne, Australia
› Author Affiliations
 

Abstract

This review summarised robust and consistent genetic variants associated with aerobic-related and resistance-related phenotypes. In total we highlight 12 SNPs and 7 SNPs that are robustly associated with variance in aerobic-related and resistance-related phenotypes respectively. To date, there is very little literature ascribed to understanding the interplay between genes and environmental factors and the development of physiological traits. We discuss future directions, including large-scale exercise studies to elucidate the functional relevance of the discovered genomic markers. This approach will allow more rigour and reproducible research in the field of exercise genomics.


#

Introduction

Both aerobic and strength exercise training lower the incidence of many chronic diseases via a number of mechanisms, including increased skeletal muscle mitochondrial function [1], modulation of the sympathetic nervous and immune systems, and optimization of the neuroendocrine system [2]. These mechanisms act as buffers against chronic diseases, minimizing inflammatory state, and enhancing neuroplasticity and growth factor expression [3]. However, large inter-individual differences exist in the physiological responses to any given exercise training (also called “trainability”) [4] [5], and recently new statistical methods have been developed to properly isolate individual responses from random error [6]. Large trainability has been observed in many physical fitness parameters [7], including maximal oxygen uptake (VO2max) [8] [9], resting heart rate [9], exercise heart rate [9], aerobic threshold [10], anaerobic threshold [9], resting muscle glycogen content, muscle enzyme activity [11], as well as muscle mass and strength [12] [13].

The heritable component of trainability is large, with genetics explaining 47% of the variance in VO2 peak trainability, and around 52% in resistance variability [14]. The contribution of familial factors (genetics and environment) to trainability was demonstrated in the seminal HERITAGE family study [15]. This study indicated that VO2max was more variable between families than within families at baseline [16], and in response to exercise training [17], thus suggesting that DNA sequence variations could modulate exercise responses [4] [18]. Pinpointing the responsible gene variants could illuminate the fundamental mechanisms driving this heterogeneity in response to exercise training [18].

The genetic contribution to trainability has been investigated by two different approaches: candidate genes and genome-wide association (GWAS) study. The GWAS approach involves scanning several hundred thousand (currently up to 5 million) DNA markers across the human genome to find genetic variations associated with a particular trait. One of the advantages of the GWAS approach is that it is unbiased and hypothesis-free. In contrast, candidate gene studies require knowledge of the trait of interest and is particularly useful to validate the functional impact of gene loci such as those identified by GWAS [19]. GWAS have demonstrated that trainability is polygenic (i. e., influenced by many genetic variants), and that people harbouring the same genotypes in specific gene variants respond more similarly to exercise training than people harbouring different genotypes [20] [21] [22] [23]. These variants may modulate gene expression that is essential to the molecular adaptation to exercise training, since molecular processes mediate metabolism, angiogenesis, cardiac and skeletal myofibre hypertrophy, and other processes that lead to better fitness [24].

While many SNPs have been associated with exercise response and trainability. The vast majority of the genes previously identified have not been replicated [25]. Replication in an independent cohort is important as it increases the likelihood that results are true and reduces the number of false positives [26] [27]. In this review we summarised SNPs associated with both resistance and aerobic trainability and have been replicated in two independent cohorts. In addition, we have screened these SNPs with the goal of identifying SNPs at trainability-associated loci that may have functional relevance. Further, we discussed future directions of performing large-scale exercise studies to elucidate the functional relevance of the discovered genomic markers. This approach will allow more rigour and reproducible research in the field of exercise genomics.


#

Materials and Methods

To provide a robust and comprehensive narrative review, a semi-structured search was performed (July 2019) to identify all studies relating to genetic variants and exercise trainability. Three electronic databases (PUBMED, MEDLINE and SCOPUS) were used to identify relevant articles using the following keywords “genes”, “genome”, “exercise”, “physical activity”, “aerobic capacity”, “resistance”, “strength”, “power”. We excluded studies where the sole focus was on populations with a diagnosed medical condition such type 2 diabetes mellitus, any inflammatory conditions, and cardiovascular disease. Articles were separated in two categories: genetic variants associated with either aerobic or resistance trainability ([Tables 1] and [2]). This review was conducted in accordance with the IJSM’s ethical standards of the journal [28]

Table 1 Gene variants associated with aerobic trainability.

Author, Date

Sample Size

Sex (% Males)

Age

Ancestry/Country /ethnicity

Chromosome

Annotated gene

Variant

Genotype and training response (+/−/0)

Intervention (if any)/exercise

Duration

Type of study

Alves (2009) [92]

N=83

Males only

20–35 yrs

Brazil

17
1

ACE
ATG

rs4340
rs699

ACE (0) VO2max
TT (+) LVM

Moderate intensity endurance training

3 days/week
16 weeks

Candidate Gene

Bouchard 2011 [23]

N=742

Males (N/A) and Females

17–65 yrs

HERITAGE study
Caucasian and African-American
U.S.A

4
6
9
3
9
3
1
1
20
11
14
15
11
14
2
4
11
3
22
11
6

ACSL1
PRDM1
GRIN3A
KCNH8
C9orf27
ZIC4
CAMTA1
RGS18
BIRC7
DBX1
DAAM1
NDN
CXCR5
TTC6
LOC400950
LOC100289626
LOC100130460
NLGN1
MN1
CD44
ENPP3

rs6552828
rs10499043
rs1535628
rs4973706
rs12115454
rs11715829
rs884736
rs10921078
rs6090314
rs10500872
rs1956197
rs824205
rs7933007
rs12896790
rs4952535
rs2053896
rs2198009
rs2030398
rs738353
rs353625
rs10452621

(+) VO2max

Endurance training Moderate: at 55% HR first two weeks and intense: last 6 weeks 75% HR

20 weeks

GWAS

Dionne (1991) [93]

Males
N=46

Males only

17–29 yrs

Canada, USA

Mitochondria

MTND2
MTND5

MTN2 (-) VO2max
MTND5 (+) VO2max

Endurance training at 85% of HRR

3–5 days/week
20 weeks

Candidate gene

Hautala et al. 2007 [35]

N=478

Males (48.3%) and
Females

17–65 yrs

HERITAGE study
Caucasian and African-American
Canada, U.S.A

22

PPARD

rs2016520
rs2076167

African American only
rs2016520
CC (-) VO2max, PPO
rs2076167 (0)

Endurance training moderate 55% of VO2 and absolute 75% of VO2 intensity

20 weeks

Candidate gene

He et al.
2008 [94]

N=181

Males only

19±1

Han Chinese

7
15

NRF-1
NRF-1
NRF-2

rs2402970
rs6949152
rs6949152

rs2402970
CC (+) VT, RE
rs6949152
AA (+) VT, RE
rs6949152
AA (+) VO2max

Endurance training
95% to 105% ventilatory threshold

18 weeks

Candidate gene

He et al.
2006 [95]

N=181

Males only

19±1

Han Chinese

11

HBB

rs10768683

C (+) RE

Endurance training
95% to 105%
ventilatory threshold

18 weeks

Candidate gene

He et al. 2007 [96]

N=181

Males only

19±1

Han Chinese

15

NRF-2
NRF-2
NRF-2

rs12594956
rs8031031
rs7181866

ATG haplotype
(+) RE

Endurance training
95% to 105% ventilatory threshold

18 weeks

Candidate gene

He et al. 2008 [43]

N=181

Males only

19±1

Han Chinese

4
4
4

PPARGC1A
PPARGC1A
PPARGC1A

rs17847357
rs8192678
rs6821591

rs17847357,
rs8192678
(0) VO2max
rs6821591 G (+) VO2max

Endurance training
High intensity
95% to 105% HR

18 weeks

Candidate gene

He et al. 2010 [97]

N=181

Males only

19±1

Han Chinese

4
4
4
2
9

PPP3CA
PPP3CA
PPP3CA
PPP3R1
PPP3R2

rs2850965
rs3804423
rs3804358
rs4671887
rs3739723

G (+) VO2max
G (+) VO2max
G (+) VO2max
A (+) VO2max
A (+) RE

Aerobic endurance 95% to 105% of ventilatory threshold

18 weeks

Candidate gene

He et al. 2010 [98]

N=181

Males only

19±1

Han Chinese

8
8
8
8
8

PPP3CC
PPP3CC
PPP3CC
PPP3CC
PPP3CC

rs1879793
rs1075534
rs7430
rs2461483
rs10108011

CC (+) SV
AA (+) SV, CO
GG (+) SV
CC (+) SV
GG (+) SV

Aerobic endurance 95% to 105% of ventilatory threshold

18 weeks

Candidate gene

Leon et al. 2004 [99]

N=766

Males (43%) and Females

17–65 yrs

HERITAGE study
Caucasian and African-American
U.S.A

19

APOE

E2, E3, E4

(0)VO2max

Endurance training Moderate: at 55% HR first two weeks and intense: last 6 weeks 75% HR

20 weeks

Candidate Gene

McKenzie 2011 [22]

N=109

Males (46.7%) and Females

50–75 yrs

Caucasian
U.S.A

14

AKT1

rs1130214

Men:
GG (+) VO2max
Females:
(0)

Aerobic training moderate 50–70%

24 weeks

Candidate gene

McPhee et al 2011 [100]

N=58

Females only

Age 18–37 yrs

Caucasian UK

14

HIF1A

rs11549465

T (+) VO2max

Aerobic 75–90% of HRmax

6 weeks

Candidate gene

Pickering et al. 2018 [42]

N=42

Males only

16–19 yrs

European (UK)

4

PPARGC1A
VEGF
ADBR2
ADBR2
CRP

rs8192678
rs2010963
rs1042713
rs1042714
1205

Endurance genotype (+) Yo-Yo Test

Aerobic training moderate to intense

8 weeks

Candidate gene

Prior et al. 2003 [101]

N=233

Males (39.3%) and Females

50–75 yrs

Caucasian and African-American
U.S.A

14

HIF1A

rs28708675
rs11549465

African American cohort:
rs28708675
AA (+) VO2max
Caucasian cohort:
rs11549465
CC (+) VO2max

Aerobic training
moderate 50–70%

24 weeks

Candidate gene

Prior et al. 2006 [102]

N=146

Males (42%) and Females

50–75 yrs

Caucasian and African-American
U.S.A

6

VEGF

rs699947
rs1570360
rs2010963

AAG & CGC haplotypes (+) VO2max

Aerobic training
moderate 50–70%

24 weeks

Candidate gene

Rankinen et al. 2000 [103]

N=472

Males (49%) and Females

Age 17–65 yrs

HERITAGE study
Caucasian
U.S.A

1

ATP1A2

Polymorphisms at exon 1 and 21–22

2α haplotype (+) VO2max and PP

Endurance training Moderate: at 55% HR first two weeks and intense: last 6 weeks 75% HR

20 weeks

Candidate Gene

Rankinen et al. 2000 [104]

N=472

Males (48.7%) and females

Age 17–65 yrs

HERITAGE study
Caucasian
U.S.A

17
1

ACE
ATG

rs4340
rs699

Males:
ACE I/D (0)
ATG M (+) reduced diastolic blood pressure.
Females:
ACE I/D (0)
ATG M/T (0)

Endurance training Moderate: at 55% HR first two weeks and intense: last 6 weeks 75% HR

20 weeks

Candidate Gene

Rico-Sanz et al. 2003 [105]

N=779

Males (N/A) and Females

Age 17–65 yrs

HERITAGE study
Caucasian and African-American
U.S.A

1

AMPD1

rs17602729

TT (-) VO2max

Endurance training Moderate: at 55% HR first two weeks and intense: last 6 weeks 75% HR

20 weeks

Candidate Gene

Ring-Dimiriou et al. 2014 [40]

N=24

Males only

45–65 yrs

Austria

4

PPARGC1A

rs8192678

GG (+) VO2peak

70–90% of Vo2peakk

3 days/week 10 weeks

Candidate Gene

Rivera et al. 1997 [106]

N=240

Males (47.5%) and Females

17–65 yrs

HERITAGE study
Caucasian and African-American
U.S.A

19

CKMM

rs8111989

CC (-) VO2max

Endurance training Moderate: at 55% HR first two weeks and intense: last 6 weeks 75% HR

20 weeks

Candidate Gene

Sonna et al. 2001 [107]

N=147

Males (42.2%) and Female

Age 21.7±3.6 yrs

USA: 57% Caucasians,
25% African-Americans, 14% Hispanics, 3% Asians, and 1% Native American

17

ACE

rs1799752

ACE I/D (0) VO2max

2 aerobic days and 2 strength training days
per week

8 weeks

Candidate Gene

Stefan et al. (2007) [38]

N=136

Males (46%) and Females

Age 19–67 yrs

Germany

22
22
22
22
4

PPARD
PPARD
PPARD
PPARD
PPARGC1A

rs2267668
rs6902123
rs2076167
rs1053049
rs8192678

rs2267668 G (-) AT, VO2peak
rs6902123 (0)
rs2076167 (0)
rs1053049 (0)
rs8192678 A (-) AT

Unsupervised:
3 h of moderate sports per week

9 months

Candidate Gene

Steinbacher et al. 2015 [41]

N=28

Males Only

50–69 yrs

Austria

4

PPARGC1A

rs8192678

AA (-) decreased fibre type 1 transformation

70–90% of Vo2peakk

3 days/week 10 weeks

Candidate Gene

Yoo et al. 2016 [108]

N=79

Males (64.6%) and Females

Age 30–60 yrs

Korea

12
18
2
3
6
2
2

AMN1
CDH2
ASB3
SRGAP3
UST
PUM2
KCNH7

rs11051548
rs2542729
rs1451462
rs13060995
rs6570913
rs11096663
rs12613181

(+) VO2 max
(+) VO2 max
(+) VO2 max
(+) VO2 max
(+) VO2 max
(+) VO2 max
(+) VO2 max

HIIT
60%–84% of VO2max

9 weeks

GWAS

Yu et al. 2014 [109]

N=360

Males (50%) and Females

Age 18–40 yrs

China

19

APOE

E2, E3, E4

E2/E3 (+) VO2max
E3/E4 (+) VO2max

Aerobic 60%–85%

6 months

Candidate gene

Zarebska et al. 2014 [110]

N=66

Females only

Age 19–24 yrs

Caucasian
Poland

11

GSTP1

rs1695

G (+) VO2max and VEmax

Aerobic training
50% to 70% of HRmax

12 weeks

Candidate gene

Zhou et al. 2006 [111]

N=102

Males Only

18.8±0.9 yrs

China

19

CKMM

rs1803285

AG (-) RE

Distance running program 95–105% of VT

18 weeks

Candidate Gene

Table 2 Gene variants associated with resistance trainability.

Author, Date

Sample Size

Sex (% Males)

Age

Ancestry/County of origin/ethnicity

Chromosome

Gene

Variant

Genotype and training response (+/−/0)

Intervention

Duration

Type of study

Ash (2016) [65]

N=602

Males (38.5%) and Females

Age 18–40 yrs

FAMuSS study:
Predominately European-American Ancestry

5

NR3C1

rs10482614
rs10482616
rs4634384

Females: rs4634384 T (+) Hypertrophy
Males: rs10482616 GG (+) MVC
rs10482614 AA (+) MVC

Upper arm, Unilateral resistance program

12 weeks

Candidate Gene

Charbonneau (2008) [55]

N=243

Males (35.3%)
And Females

Age 50–85 yrs

U.SA. Caucasian

17

ACE

rs1799752

Females: ACE (0)
Males: ACE (0)

Knee Extension unilateral resistance program

10 weeks
3days/weeks

Candidate Gene

Clarkson (2005) [66]

N=602

Males (41%)
and Females

Age 18–40 yrs

FAMuSS study:
Predominately European-American Ancestry

11

ACTN3

rs1815739

Females: ACTN3 XX (+) Maximal dynamic strength (1RM).
Males: ACTN3 (0)

Upper arm, Unilateral resistance program

12 weeks

Candidate Gene

Delmonico (2007) [112]

N=157

Males (45.2%) and Females

Age=50–85 yrs

Caucasian
USA

11

ACTN3

rs1815739

Females: ACTN3 RR (+) PP
Males: ACTN3 (0)

Knee Extension unilateral resistance program

3days/week
10 weeks

Candidate Gene

Erskine (2012) [113]

N=51

Males only

Age 20.3±3.1 yrs

Caucasian
UK

8

PTK2

rs7843014
rs7460

rs7843014 AA (+) Strength (MVC)
rs7460 TT (+) Strength (MVC)

Knee Extension unilateral resistance program

3days/week
9 weeks

Candidate Gene

Erskine (2013) [61]

N=51

Males only

Age 20.3±3.1 yrs

Caucasian
UK

17
11

ACE
ACTN3

rs1799752
rs1815739

ACE (0)
ACTN3 (0)

Knee Extension unilateral resistance program

3days/week
9 weeks

Candidate Gene

Folland (2000) [56]

N=33

Males only

Age 18–30 yrs

UK

17

ACE

rs4646994

Isometric training: ACE DD/ID (+) Isometric strength (MVC)
Dynamic training: ACE DD/ID (0)

Isometric Training
Dynamic training

3days/week
9 weeks

Candidate Gene

Giaccaglia (2006) [57]

N=213

Males (N/A) and Females

Age > 60 yrs

Predominately Males and Females of European-American Ancestry

17

ACE

rs4646994

ACE DD (+) strength (MVC)

Light resistance training

3days/week
18 months

Candidate Gene

Harmon (2010) [67]

N=874

Male (41.1%) and Females

Age 18–40 yrs

FAMuSS study:
Predominately European-American Ancestry

17
3

CCL2
CCR2

CCL2
(rs17652343), (rs1860189), (rs3917878), (rs2857654), (rs1024611), (rs1024610), (rs3760396),
(rs2857656), (rs2857657),
(rs4586),
(rs13900)
CCR2
(rs17141010), (rs768539), (rs3918358), (rs1799864), (rs1799865).

Females:
CCL2 (0) and CCR2 (0)
Males:
CCL2 T (rs1024610) (+) Maximal Isometric strength (MVC)
Males and Females
CCR2 (AA) rs3918358 and (TT) rs1799865 (+) Isometric strength (MVC)

Upper arm, Unilateral resistance program

2 days/week
12 weeks

Candidate Gene

He (2019) [59]

N=40

Females only

Age 53–66 yrs

Chinese, Beijing

17

ACE

rs4646994

ACE DD (+) Maximal Isometric strength (MVC), muscle hypertrophy and grip strength

Whole body resistance training

3 days/week
8 weeks

Candidate Gene

Hong (2014) [74]

N=83

Males only

Age 22.6±1.4 yrs

South Korean

11

CNTF

rs1800169

CNTF G/A (0)

Resistance training of
the upper extremities

3 days/week
8 weeks

Candidate Gene

Jamshidi et al. (2002) [114]

N=144

Males only

19.6 (2.4) yrs

UK

6

PPARA

rs425778

C (+) LV mass

Upper and lower body training program

10 weeks

Candidate Gene

Jones (2006) [13]

Study 1, N=28.
Study 2 N =39

Males only

18–20 yrs

Caucasian UK

17
11

(Power-related polygenic risk score)

ACE D (rs1799752)
ACTN3 (rs1815739)
ADRB2 C (rs1042714)
AGT C (rs699)
IL-6 G/C (rs1800795)
PPARA C (rs4253778)
TRHR G (rs8192676)
VDR A (rs1544410)

Power genotype (+) Power (CMJ) after high intensity resistance training but not low intensity resistance training.

Low intensity
(~30% of 1 RM and high repetitions) and high-intensity
(~70% of 1 RM and low repetitions) resistance training

8 weeks of high or low resistance training
1 to 2 days per week

Polygenic Score

Keogh (2015) [115]

N=58

Males (31%) and Females

Age 69.8±5.3

New Zealand (European ancestry)

17

ACE
UCP2

rs4646994
rs7109266

ACE ID (0)
UCP2 GG (+) Lower body strength (8ft Up and Go time)

Resistance training light to moderate intensity

2days/week, 12 weeks

Candidate Gene

Kostek (2005) [116]

N=67

Males (47.7%) and females

50–85 yrs

U.S.A Caucasian

12

IGF1

IGF1 192

IGF1 192/192+192/- (+) dynamic (1RM) muscle strength

Unilateral resistance program

10 weeks
3days/wk

Candidate Gene

Li (2014) [117]

N=94

Males only

Age 18–22 years

Han Chinese

2

MTSN

rs1805086
rs1805065

MTSN KR (+) Hypertrophy in Biceps and Quadriceps
MTSN AT + TT (+) Hypertrophy in Biceps

Arm and Leg resistance training

3–4 days/ wk
8 weeks

Candidate Gene

Pereira (2013) [58]

N=139

Females only

Age 65.5 (8.2)

Portugal, Caucasian

17
11

ACE
ACTN3

rs1799752
rs1815739

ACE D/D (+) maximal dynamic strength 1RM, power (CMJ), functional capacity (STS)
ACTN3 RR (+) maximal dynamic strength (1RM), power (CMJ), functional capacity (STS)

High-speed power training

12 weeks 3 days/week

Candidate Gene

Pescatello (2006) [60]

N=631

Males (42%) and females

Age 18–40 yrs

FAMuSS study:
Predominately European-American Ancestry

17

ACE

rs4646994

Trained Arm
Post Intervention: ACE II/ID (+) Maximal Isometric strength (MVC)
Untrained Arm
Post Intervention: ACE DD/ID (+) maximal dynamic strength (1RM), muscle size (CSA of Type II fibres).

Upper arm, Unilateral resistance program

12 weeks, 2 days/week

Candidate Gene

Pistilli (2008) [70]

N=748

Males (40.2%) and
Females

18–40 yrs

Caucasian

10

IL15RA

rs2296135

rs2296135 CC (+) MVC

RT program

12 weeks 2 days/week

Candidate gene

Reichman (2004) [71]

N=153

Males (49.6%) and
Females

Aged 18–31 years

Predominantly European-American Ancestry

10

IL15RA

rs3136617
rs3136618
rs2296135

rs3136617 C (+) muscle hypertrophy
rs2296135 C (+) muscle hypertrophy

Whole body resistance training @75% of 1RM

10 weeks, 3 days/week

Candidate Gene

Sprouse (2014) [68]

N=874

Males (50%) and females

Age: 18–40 years

FAMuSS study:
Predominately European-American Ancestry

8

SLC30A8

rs13266634

Females: SCL30A8 (0)
Males: SCL30A8(0)

Upper arm, Unilateral resistance program

Acute and 12-week Intervention

Candidate Gene

Thomis (2004) [63]

N=57

Males only

22.4 (3.7) yrs

Flemish Brabant, Belgium

17
2

ACE
MTSN

rs4646994
rs1805086
rs1805065

ACE (I/D) (0) strength, isometric and concentric torque
or arm muscle cross-sectional area
MTSN: Unable to be determined

High resistance training program

10 weeks, 3days/week

Candidate Gene

Walsh (2009) [73]

N=745

Males (40%) and Females

Age 18–40 yrs

FAMuSS study:
Predominately European-American Ancestry

11

CNTF

rs1800169

Females: CNTF GG (+) isometric (MVC) and dynamic (1RM) muscle strength
Males: CNTF (0)

Upper arm, Unilateral resistance program

12 weeks,
2 days/wk

Candidate Gene

Walsh (2012) [69]

N=560

Males (N/A) and Females

Age 18–40 yrs

FAMuSS study:
Predominately European-American Ancestry

1

LEP
LEPR

rs2167270
rs1137100
rs1137101
rs1805096
rs8179183

LEP (GG/GA) rs2167270 (+) Muscle hypertrophy
LEPR (0)

Upper arm, Unilateral resistance program

12 weeks,
2 days/wk

Candidate Gene

Finally, we selected SNPs that were classified as robust and separated them according to whether they were related to the aerobic trainability or resistance trainability. We chose this criteria as it reflects the reliability of the findings and increases the likelihood that there is true association of the SNP with trainability [27]. It also allows us to identify and summarise SNPs with biological relevance which is useful for researchers to ‘select’ candidate SNPs to identify causality and purpose of gene [29].

SNPs were considered robust if:

  1. Consistent association with a given phenotype in at least two independent cohorts.

  2. SNPs were shown to have functional relevance in an animal model or cell culture, with gene expression/DNA methylation Quantitative Trait Loci (QTLs) analysis or network, and enrichment analysis.

Aerobic Trainability

Twin and family studies indicate that ~22–57% of aerobic fitness variability between individuals can be explained by genetics and therefore plays an important role in the range of aerobic phenotypes observed in a population [30]. Here, we briefly describe some of the robust SNPs that have been associated with aerobic trainability, which means they were replicated in at least 2 independent cohorts and were shown to have functional relevance.

A bioinformatic analysis study conducted by Ghosh et al. found that the greatest number of SNPs were annotated to the PPAR signalling pathway suggesting its importance in VO2max trainability [31]. As such the most widely studied genes within this pathway are the peroxisome proliferator-activated receptors (PPARA, PPARG, and PPARD) and their transcriptional coactivators (PPARGC1A and PPARGC1B). These genes have been linked to multiple aerobic phenotypes, including muscle morphology, aerobic capacity and endurance performance [32] [33] [33] . PPARD is expressed predominantly in adipocytes and skeletal muscle where it promotes fatty acid oxidation [34]. In the HERITAGE family study, the rs2016520 SNP (C allele) located in PPARD was associated with reduced VO2max and maximal power output after a 20 week endurance training intervention in African-Americans but not in Caucasians [35]. In vitro and animal studies show that the minor allele (C allele) in this SNP (rs2016520) results in higher PPARD transcriptional activity, which in turn promotes lipid accumulation and the alters normal regulation of lipid uptake and storage [34] [36] [37]. In a European cohort it was shown that the PPARD rs2267668 SNP was associated with VO2peak and anaerobic threshold after a 9-month lifestyle intervention [38]. They then confirmed that in human primary cell lines that those carrying the minor allele at rs2267668 (G allele) were associated with lower mitochondrial activity, demonstrating a potential functional effect [38]. Taken together, PPARD locus may play a role in aerobic trainability, but larger cohorts of different ancestries and, more in depth functional studies to determine causal SNP are needed to confirm this.

The transcriptional co-activator PPARGC1A interacts with PPARD and regulates mitochondrial biogenesis, angiogenesis, lipolysis and adipogenesis [39]. Four candidate gene studies, predominantly in men, found consistent associations of rs8192678 within PPARGC1A and aerobic capacity in Europeans [38] [40] [41] [42]. While in the Han Chinese cohort another nearby SNP (rs6821591) was associated with VO2max specifically, the G allele was associated with increased VO2max compared to those carrying the A allele [43]. Work conducted in a Han Chinese cohort found that the PPARGC1A rs6821591 SNP had functional significance as gene expression was altered and this was dependent on genotype (A v G allele) with the G allele displaying increased PGC-1α gene expression [44]. Overexpression of PGC-1α in an animal model showed increased Type 1 fibres in muscles that are normally Type II fibre type dense and this induced increases in resistance to fatigue, inferring increased aerobic capacity [45]. These population-specific results indicate that it is the PPARGC1A locus itself, rather than individual SNPs located within that locus, may be important for trainability [43] [46].

Currently 26 SNPs associated with VO2max trainability were identified in a GWAS and were validated in 2 separate cohorts (detailed in [Table 2]) [23]. They accounted for 49% of VO2max trainability and were able to classify responders and non-responders [23] [47]. Whether these SNPs are directly involved in gene function or regulation of genes is the next step to validate these findings. The most robust is the SNP rs6552828 located near the ACSL1 gene which was the strongest predictor (~6%) of aerobic trainability (VO2max) [23]. It has subsequently been validated in a bioinformatics pathway analysis and found to be strongly correlated to the aerobic electron transport chain phenotype and the PPAR signalling pathway providing a robust candidate gene in VO2max trainability [31]. ACSL1 regulates lipid metabolism by facilitating the transport of long chain fatty acids into the mitochondria and is an essential step in fatty acid oxidation [48]. Timmons et al. integrated RNA profiles with genetic variants and found the following genes CD44, and DAAM1, also discovered in the Bouchard et al. GWAS, were associated with gene expression changes [49]. Gene expression of CD44 was up-regulated in response to endurance training [49] and was strongly associated with phenotypic terms associated with aerobic exercise such as: cardiovascular physiological processes, muscle contraction, physical fitness and aerobic electron transport chain [31] indicating that this gene and any alterations to its function (i. e. via SNPs) may play in important role in aerobic trainability. While these genes certainly provide robust genes, there are still limitations in determining the causality of these particular SNPs in the molecular mechanisms affecting aerobic trainability.

Many candidate gene and GWAS studies have been conducted and this review highlights the large collection of candidate genes that have been associated with aerobic trainability. Only 12 SNPs have been robustly associated with aerobic trainability ([Table 3]) meaning that have been validated in at least 2 independent cohorts and were shown to have some functional relevance. Subsequent studies should focus on understanding the functional role of the SNPs that have been replicated as this review highlights the lack of understanding of the molecular mechanism and limits our understanding of aerobic trainability.

Table 3 Robust SNPs associated with aerobic or resistance trainability.

Aerobic trainability

Resistance trainability

SNP

Nearest gene

Beneficial allele

SNP

Nearest gene

Beneficial allele

rs6552828

ACSL1

G

rs4646994*

ACE

D

rs699

AGT

T

rs1799752*

ACE

D

rs6090314

BIRC

A

rs4340*

ACE

D

rs12580476

C12orf36

TBC

rs13447447*

ACE

D

rs884736

CAMTA1

G

rs1815739

ACTN3

R

rs353625

CD44

TBC

rs2296135

IL15 RA

C

rs1956197

DAAM1

G

rs4253778

PPARA

C

rs17117533

NDN

A

rs8192678

PPARGC1A

G

rs10921078

RGS18

A

rs7531957

RYR2

TBC

rs11715829

ZIC4

G


#

Resistance Trainability

Muscular strength and power show a heritability estimated around 52% [14]. Skeletal muscle strength is defined as the force produced by muscle contraction. A variety of measures have been investigated, including muscle strength, maximal voluntary contraction (MVC), 1 repetition maximum (1RM) and handgrip strength. While the production of skeletal muscle power is defined as how much force can be produced and the velocity at which it is produced. The production of power can be measured at the by undertaking tests such as Wingate’s, counter movement jumps (CMJ) and vertical jumps (VJ).

The ACE I/D and ACTN3 R/X SNPs are two of the most extensively studied gene loci. We have chosen not to discuss ACTN3 here as it has recently been reviewed in detail by Del Coso et al. [50] and instead focus on the ACE I/D SNP. The ACE gene encodes the angiotensin-converting enzyme that is a central component of the renin-angiotensin-system [51]. The ACE I/D results in either an insertion (I) or deletion (D) of a 287-basepair region in intron 16 of the gene [52] and can alter the levels of ACE in the blood [52]. It has recently been shown that the polymorphism can manipulate the activity of the C- and N-terminal domain in the enzyme [53]. Further, exercise can decrease the enzyme activity in the C-terminal domain and increase the activity in the N- terminal domain which results in improved blood flow and proliferation of red blood cells [53]. It is thought that the I allele confers enhanced endurance performance while the D allele is thought to confer increased muscle power and strength [54]. The D allele was consistently shown across 6 separate candidate gene studies to be associated with greater gains in strength after resistance training and this was consistent across sex and age [55] [56] [57] [58] [59] [60]. While the literature is consistent regarding muscular strength, the association with muscular power is less convincing [55] [61] [62] [63]. The D allele in ACE was associated with CMJ in older females after a 12-week power training program [58] and in young males after a high intensity training program [13]. However, it was the I allele in ACE that was associated with a higher baseline VJ at baseline in males and females [62]. Another two studies did not find any association between the ACE I/D and skeletal muscle power at baseline or in response to resistance training [61] [63]. ACE provides a robust candidate gene for explaining variation in muscular strength but not muscular power suggesting that this gene loci may only explain some of the inter-individual resistance variability dependent on type of resistance exercise.

Many of the candidate genes in resistance trainability came from a large multi-centre trial (FAMuSS) which aimed to identify nonsynonymous SNPs with functional effects on muscle power and strength [64]. These include: Glucocorticoid receptor (NR3C1) [65] , alpha-actinin 3 (ACTN3) [66], Chemokine (C-C motif) ligand 2 (CCL2) [67] , Chemokine (C-C motif) ligand 2 Receptor (CCR2) [67], ACE [60], Solute carrier family 30 (zinc transporter), member eight gene (SLC30A8) [68], Leptin (LEP) and Leptin receptor (LEPR) [69]. The FAMuSS study was conducted in young (18–40 years old) males (N=247) and females (N=355) of predominantly European-American ancestry. Participants underwent a 12-week unilateral resistance program consisting of upper arm exercises in the non-dominant arm [60]. Only IL-15RA, ACTN3 and ACE from this series of studies were replicated in separate cohorts and have functional relevance. In the IL-15RA locus the rs2296135 SNP was associated gains in muscular strength and replicated in two different studies in cohort of European ancestry [70] [71]. When the gene IL-15RA is knocked down in an animal model it altered the contractile properties and fatigability in skeletal muscle fibres [72]. While the locus is important it not yet clear which SNPs is responsible for altering the function of IL-15RA protein. Although SNPs within CCL2, CCR2 and CNTF have not been replicated they interestingly showed sex-specific associations with muscle strength. CTNF polymorphisms were associated with strength gains only in females [73], which was subsequently confirmed in a South Korean cohort [74]. SNPs in CCL2 and CCR2 were associated strength gains in males only [67]. This indicates potential sex-specific differences in the genetic architecture of complex traits and should be incorporated into study design [75] [76]. In addition PTK2, CNTF, IL-6, PPARA and VDR candidate genes have been replicated with functional relevance [13] [73].

In total 7 SNPs ([Table 3] ) were robustly associated with resistance variability. While there are plethora of candidate gene studies no GWAS have been conducted that specifically focuses on resistance trainability.


#

Functional Validation

We have identified 12 SNPs and 7 SNPs that are robustly associated with variance in aerobic and resistance trainability respectively. The next steps are to a) identify the causal SNP, b) annotate the casual SNP to the correct gene and then c) to establish the functional relevance of the gene [47]. The overall evidence from literature connecting causal genes to trainability is relatively low [31]. If we hope to identify the casual variants or genes it is vital that we begin to integrate “omic” technologies from the genome and epigenome to transcriptome to proteome and metabolome which can capture a complete picture of complex human traits such as aerobic and resistance trainability [77] [78].

There have been attempts to associate molecular pathways or “molecular phenotypes” with physiological phenotypes of aerobic and resistance trainability [79] [80] [81]. Sarzynski et al. applied this systems biology approach by combining the 21 SNP identified in a GWAS from the HERITAGE study cohort ([Table 2]) [15] [23] and examined the joint contributions of these SNPs to exercise response [47]. This approach identified potential pathways in calcium signalling, energy sensing and partitioning, mitochondrial biogenesis, angiogenesis, immune functions, and regulation of autophagy and apoptosis, providing important pathways that can be investigated more closely [47]. Another integrative approach is expression quantitative trait loci (eQTLs) analysis that leverages gene loci identified from GWAS and integrate these with gene expression data to identify differential gene expression levels to try and uncover the ‘molecular phenotype’ that lead to these variations in exercise response [82] [83]. Willems et al. identified the rs6565586 SNP in ACTG1 as a strong candidate gene in inter-individual variability in the resistance-related phenotype (hand grip strength) and correlated this with a lower expression of mRNA in skeletal muscle. ACTG1 encodes Actin Gamma 1 and is involved in the structure and function of skeletal muscle fibres. Interestingly, in a knock out mouse model, animals displayed overt muscle weakness [84]. This type of analysis presented an ideal candidate gene to begin understanding the molecular mechanisms in human skeletal muscle.

To establish causality of genetic variants in aerobic and resistance trainability the field needs to move forward beyond association analysis. The type of follow-up experiment will depend on the location of SNP within the gene. For SNPs within coding regions ideally experiments are performed to study the effect of the SNP has on protein structure and function. For SNPs within in non-coding regions it more difficult to determine as they may not directly affect a gene but alter/regulate transcription factors and mediate alterations in genes this way [77]. However, with the introduction of the large epigenetic database ENCODE (Encyclopaedia of DNA elements) we can now identify the transcription factor association, chromatin structure and histone modification of target genes [85] and more recently enhancers providing candidate gene targets for follow up analysis [86]. With the discovery of CRISPR Cas-9 genome-editing tool in 2012 [87], this has paved the way for establishing causality of SNPs and the functional effects of them. This has been used to great effect for establishing causal genes implicated in insulin resistance whereby they were able to determine the casual effect of 12 candidate genes that had previously been identified in a GWAS [88]. To date no experiments have been conducted using this gene-editing tool to establish the function and causality of candidate genes of trainability beyond association analysis.

There is still much work to do before personalised exercise prescription (both in a clinical and elite athlete setting) can be based on an individual’s genetics. However, there are concerted efforts taking place to make this possible such as the Athlome Project Consortium and the Gene SMART (Skeletal Muscle Response to Training), recently launched with the aim of uncovering the genetic variation underlying athletic performance, adaptation to exercise training, and exercise-related musculoskeletal injuries [89] [90]. These, and other initiatives will allow for population-based approach to understand the role of genes and environmental factors contributing to the complex exercise response phenotype [91].

This review summarised robust genetic variants that have been associated with aerobic and resistance trainability. To date, there is very little literature ascribed to understanding the interplay between genes and environmental factors and the development of physiological traits. Therefore, much work remains to identify causal variants and functional relevance of genes associated with aerobic and resistance trainability.


#
#
#

Conflict of interest

The authors declare that they have no conflict of interest.


Correspondence

Dr. Nir Eynon
Institue of Sports Exercise and Active Living, Victoria University
8001 Melbourne,
Australia   
Phone: +61 399195615   
Fax: +61 399195615   

Publication History

Received: 00 00 2020

Accepted: 01 June 2020

Publication Date:
21 July 2020 (online)

© Georg Thieme Verlag KG
Stuttgart · New York