Muscle health and performance in monozygotic twins with 30 years

Vol.:(0123456789)

1 3

European Journal of Applied Physiology

https://doi.org/10.1007/s00421-018-3943-7

ORIGINAL ARTICLE

Muscle health andperformance inmonozygotic twins with30 years

ofdiscordant exercise habits

KatherineE.Bathgate

· JamesR.Bagley

· EdwardJo

· RobertJ.Talmadge

· IreneS.Tobias

·

LeeE.Brown

· JaredW.Coburn

· JoseA.Arevalo

· NancyL.Segal

· AndrewJ.Galpin

Received: 30 April 2018 / Accepted: 10 July 2018

Abstract

Introduction Physical health and function depend upon both genetic inheritance and environmental factors (e.g., exercise

training).

Purpose To enhance the understanding of heritability/adaptability, we explored the skeletal muscle health and physiological

performance of monozygotic (MZ) twins with > 30 years of chronic endurance training vs. no speciﬁc/consistent exercise.

Methods One pair of male MZ twins (age = 52 years; Trained Twin, TT; Untrained Twin, UT) underwent analyses of: (1)

anthropometric characteristics and blood proﬁles, (2) markers of cardiovascular and pulmonary health, and (3) skeletal

muscle size, strength, and power and molecular markers of muscle health.

Results This case study represents the most comprehensive physiological comparison of MZ twins with this length and

magnitude of diﬀering exercise history. TT exhibited: (1) lower body mass, body fat%, resting heart rate, blood pressure,

cholesterol, triglycerides, and plasma glucose, (2) greater relative cycling power, anaerobic endurance, and aerobic capac-

ity (VO

max), but lower muscle size/strength and poorer muscle quality, (3) more MHC I (slow-twitch) and fewer MHC

IIa (fast-twitch) ﬁbers, (4) greater AMPK protein expression, and (5) greater PAX7, IGF1Ec, IGF1Ea, and FN14 mRNA

expression than UT.

Conclusions Several measured diﬀerences are the largest reported between MZ twins (TT expressed 55% more MHC I ﬁbers,

12.4ml/kg/min greater VO

max, and 8.6% lower body fat% vs. UT). These data collectively (a) support utilizing chronic

endurance training to improve body composition and cardiovascular health and (b) suggest the cardiovascular and skeletal

muscle systems exhibit greater plasticity than previously thought, further highlighting the importance of studying MZ twins

with large (long-term) diﬀerences in exposomes.

Keywords Fiber type· Myosin heavy chain· Maximal oxygen consumption· Endurance training· FN14· PAX7· Body

composition· AMPK· Aerobic exercise· Aging

Abbreviations

%BF Body fat percentage

AMPK 5′ AMP-activated protein kinase

B2M β-2-microglobulin

Communicated by William J. Kraemer.

Katherine E. Bathgate and James R. Bagley contributed equally to

this work.

* Andrew J. Galpin

[email protected]

Biochemistry andMolecular Exercise Physiology

Laboratory, Center forSport Performance, Department

ofKinesiology, California State University, Fullerton, 800

North State College Blvd., KHS-121, Fullerton, CA92834,

USA

Muscle Physiology Laboratory, Department ofKinesiology,

San Francisco State University, 1600 Holloway Ave. GYM

101, SanFrancisco, CA94132, USA

Human Performance Research Laboratory, California State

Polytechnic University, 3801 W. Temple Ave. 66-213,

Pomona, CA91768, USA

Department ofBiological Sciences, California State

Polytechnic University, 3801 W. Temple Ave., Pomona,

CA91768, USA

Department ofPsychology, California State University,

Fullerton, 800 North State College Blvd., Fullerton,

CA92834, USA

European Journal of Applied Physiology

1 3

BMC Bone mineral content

BMD Bone mineral density

CHOL Total cholesterol

CS Citrate synthase

CSA Muscular cross-sectional area

DBP Diastolic blood pressure

DXA Dual-energy X-ray absorptiometry

ECL Enhanced chemiluminescent

EI Echo intensity

FEV

Forced expiatory volume in the ﬁrst second

FM Fat mass

FVC Forced vital capacity

HbA1c Glycosylated hemoglobin

HDL High-density lipoprotein

IGF1Ea Insulin-like growth factor a

IGF1Ec Mechano-growth factor

kcal Kilocalorie

LDL Low-density lipoprotein

LM Lean mass

MHC Myosin heavy chain

MSTN Myostatin

MT Muscle thickness

MVIC Maximal voluntary isometric contraction

MyHC Myosin heavy chain gene

MZ Monozygotic

NOS3 Endothelial nitric oxide synthase

PP Peak power

PPIA Cyclophilin

QRT-PCR Quantitative reverse transcriptase polymerase

chain reaction

RER Respiratory exchange ratio

RHR Resting heart rate

RPE Rating of perceived exertion

RPM Rotations per minute

SBP Systolic blood pressure

TFAM Transcription factor A of the mitochondria

TNF Tumor necrosis factor-α

TRIG Triglycerides

TT Trained twin

UT Untrained twin

VAT Visceral adipose tissue

VEGFA Vascular endothelial growth factor

VL Vastus lateralis

max Maximal aerobic capacity

WAnT Wingate anaerobic test

WEEE Weekly estimated energy expenditure

Introduction

Chronic physical exercise reduces the risk of all-cause

mortality and increases longevity. The eﬃcacy of exercise

training is widespread, well documented, and evident across

all physiological systems. Human research studies have dif-

ﬁculty controlling for genetic inheritance, making it chal-

lenging to discern which speciﬁc physiological character-

istics adapt to exercise, and to what extent (Boomsma etal.

2002). Fortunately, monozygotic (MZ) human twins (who

share all their genes) with diﬀering physical activity patterns

alleviate this issue and are therefore preeminent research

targets (Hannukainen etal. 2011; Leskinen etal. 2011; Rot-

tensteiner etal. 2015; Boomsma etal. 2002; Segal 2017).

Previous investigations have analyzed MZ twins with

either long-term (several decades) and moderate diﬀer-

ences in physical activity (e.g., walkers vs. sedentary) or

short-term (months-years) and large diﬀerences in physical

activity (e.g., high-intensity endurance training programs

vs. sedentary) (Adams etal. 1985; Fagard etal. 1991, 1987;

Klissouras and Pigozzi 2009). This heterogeneity in study

designs may partially explain why heritability estimates of

measures, such as maximal oxygen consumption (VO

max),

range broadly from ~ 25 to 96% (Adams etal. 1985; Fagard

etal. 1991, 1987; Klissouras and Pigozzi 2009). The plastic-

ity of other markers of sport performance, physical ﬁtness,

muscle function, body composition, metabolic health, and

the epigenome are equally diﬃcult to interpret (Leskinen

etal. 2010, 2011; Rottensteiner etal. 2015; Hannukainen

etal. 2011; Segal 2017; Danis etal. 2003; Thomis etal.

1998). To date, no investigation has comprehensively

assessed the physiological proﬁles of human MZ twins with

both long-term and large diﬀerences in exercise habits.

We had the unique opportunity to study a pair of 52-year-

old male MZ twins with similar exposomes (i.e., totality

of human environmental exposures) during development

(to approximately age 20 years), but signiﬁcantly divergent

exercise patterns for the past three decades. One brother

[the exercise-trained twin (TT)] regularly engaged in rig-

orous endurance training, competing in triathlons and dis-

tance running events for the past 30 years; conversely, the

untrained twin (UT) did not engage in regular exercise other

than normal activities of daily living. The purpose of this

study was to explore the eﬀects of 30years of endurance

training on skeletal muscle health and physiological perfor-

mance in these twins.

Methods

Participants

Two male monozygotic twins (age = 52 years) volunteered

for this study. TT regularly engaged in various modes of

endurance exercise and competed in multiple marathons and

triathlons [logged ~ 63,458 running km (39,431 miles) from

July 1993 to June 2015; All Word Bronze Level Ironman

qualiﬁer in 2005]. Conversely, UT did not engage in regular

European Journal of Applied Physiology

1 3

exercise other than normal activities of daily living. The

twins self-identiﬁed as Caucasian and originated from the

Midwestern United States; TT worked as a high school track

coach and UT was a truck driver. Written informed consent

was obtained from subjects prior to testing; this study was

approved by the University’s Institutional Review Board for

Human Subjects.

Experimental design andcontrols

The twins underwent a battery of tests (conducted over

three consecutive days) to analyze (1) anthropometric char-

acteristic and blood proﬁles, (2) markers of cardiovascular

and pulmonary health, and (3) whole muscle size, strength,

and power and molecular markers of skeletal muscle health

(see detailed methods in subsequent sections). Prior to vis-

iting the laboratory, participants completed informed con-

sent forms, submitted DNA samples to conﬁrm zygosity,

underwent fasted blood panels from personal physicians,

and tracked normal physical activity patterns and dietary

intake using a cell phone application (MyFitnessPal; San

Francisco, CA, USA) for seven consecutive days before the

ﬁrst visit. Additionally, the twins completed detailed written

questionnaires and follow-up oral interviews to determine

exercise histories; TT also provided proof of participation

in competitive events (e.g., pictures, medals, certiﬁcates,

race numbers, etc.) along with hand-written training logs

from 1980-current. To determine current physical activity

and dietary habits, researchers extracted exercise frequency,

intensity, duration, weekly energy expenditure due to exer-

cise (WEEE), kilocalorie (kcal) consumption, and macronu-

trient composition from the twin’s MyFitnessPal accounts.

The twins’ monozygosity was conﬁrmed by Aﬃliated

Genetics (now Taueret Laboratories; Salt Lake City, UT,

USA) via analyses of ﬁfteen short tandem repeat markers

(STRs). The probability of the twins being MZ was greater

than 99%.

During Visit 1, researchers collected medical and exer-

cise history questionnaires and performed anthropometric,

pulmonary, whole muscle size (ultrasound), strength, and

power measures; additionally, the twins were familiarized

with VO

max procedures and equipment. During Visit 2,

the twins completed body composition testing via dual-

energy X-ray absorptiometry (DXA) and VO

max tests

on a cycle ergometer. Finally, during Visit 3 they under-

went resting muscle biopsies (vastus lateralis, VL) and

completed Wingate anaerobic test (WAnT). All testings

were conducted at the same location/time of day, and by

the same technicians; however, TT and UT were tested on

separate days to avoid competition or other unnecessary

interferences, and to minimize halo eﬀects. The twins were

also cautioned not to discuss any of the tests until both

had completed them and they complied fully. Furthermore,

neither twin was given access to any ﬁndings until both

had completed all visits.

The twins were housed in a hotel adjacent to the labora-

tory the day before Visit 1 through the completion of Visit

3. Participants were instructed to refrain from extraordi-

nary physical activity, drugs, and alcohol at least 48h

prior to Visit 1, sleep ≥ 8h before each visit, and eat their

typical diets during the testing period. They were allowed

to continue with normal caﬀeine habits (although neither

twin regularly consumed coﬀee). Participants were asked

to consume 500ml of water the night before and 1L of

water the morning of each visit to standardize hydration.

Anthropometric measures andblood proﬁles

Height and body mass were measured with a stadiometer

and digital Health-o-meter scale (792KL, Bridgeview, IL)

and body composition [lean mass (LM), fat mass (FM),

total body fat percentage (%BF), visceral adipose tissue

(VAT), bone mineral content (BMC), and bone mineral

density (BMD)] was assessed via DXA (Hologic Discov-

ery-QDR Series Densitometer, Bedford, MA). A three-

compartment model of body composition was applied

through which FM and non-bone LM (i.e., lean mass—

bone mineral content) was analyzed for the whole body.

Prior to the scan, participants were made free from metal-

lic clothing and accessories, required to empty their blad-

der, and refrained from consuming liquids or mineral sup-

plements. The DXA machine was calibrated before each

scan using a manufacturer-provided phantom. All DXA

measurements and analyses were conducted by a single

certiﬁed technologist.

Each twin also obtained fasting blood proﬁles prior to

Visit 1 [TT was conducted by LabCorp (Dublin, OH) and

UT by Kaiser Permanente (Panorama City, CA, USA)].

Analyses included total cholesterol (CHOL), low-density

lipoproteins (LDL), high-density lipoproteins (HDL), tri-

glycerides (TRIG), plasma glucose, and glycosylated hemo-

globin (HbA1c).

Cardiorespiratory health

Resting heart rate (RHR), blood pressure, VO

max, and pul-

monary function were measured to assess cardiorespiratory

health of the TT and UT. For resting measures, participants

laid supine for 20min before analysis of RHR via a polar

heart rate monitor (Polar, Olulu, Finland) and systolic (SBP)

and diastolic (DBP) blood pressures via e-sphyg 2 sphyg-

momanometer (American Diagnostic Corporation, Haup-

pauge, NY, USA).

European Journal of Applied Physiology

1 3

Maximal oxygen consumption

max was assessed using maximal graded cycling test

protocols with an open-circuit indirect calorimeter (Par-

voMedics TrueOne

2400, Salt Lake City, Utah, USA)

and a cycle ergometer (Cosmed LC7, Monark Exercise

AB, Varburg, Sweden) administered in a thermo-neutral

(~ 24°C) laboratory. Participants reported to the labora-

tory following at least 8h of no strenuous activity and

rested quietly seated for 10min before testing. Subjects

were then ﬁtted with a rubber-ventilated mask with a one-

way valve interfaced to the indirect calorimeter via corru-

gated plastic tubing as well as a wireless heart rate moni-

tor (Polar, Oulu, Finland). Participants were instructed to

cycle at a constant rate of 70 revolutions per minute (rpm)

during the graded exercise test. Due to the large diﬀer-

ence in anticipated ﬁtness level, TT initiated the test at a

constant workload of 125W with a 25W increase every

minute until exhaustion. UT started at a constant work-

load of 110W with 15W increases each minute for the

ﬁrst 5min, and a 25W increase every minute thereafter

until exhaustion. After each stage, participants were asked

to report a rating a perceived exertion (RPE) using the

Borg Scale. A valid VO

max measurement was deﬁned

as the highest VO

reached within the last minute of the

test while satisfying the following criteria: RPE > 18,

respiratory exchange ratio (RER) > 1.1, heart rate > 90%

of the predicted maximum heart rate (220-age), and VO

plateaued within the last 30s. Prior to each test, calibra-

tion procedures were conducted for the ﬂow meter using a

3.0-L syringe and for the gas analyzer with veriﬁed gases

of known concentrations.

Pulmonary function

Pulmonary function was assessed using a spirometer

(Spirolab II, SDI Diagnostics, Easton, MA, USA). Partici-

pants wore a nose clip and sealed their lips around the dis-

posable turbine ﬁlter. They were instructed to breathe in as

hard as possible (maximal inspiration), then immediately

exhale as fast, hard, and long as possible (maximal expira-

tion). The best forced vital capacity (FVC) and forced expia-

tory volume in the ﬁrst second (FEV

) of ﬁve attempts was

recorded.

Skeletal muscle health

Whole muscle size, strength, and power were assessed.

Additionally, protein and gene expression were measured for

various markers of ﬁber type, metabolism, growth, repair,

and inﬂammation.

Whole muscle size andquality

The right VL of each twin was assessed for muscular cross-

sectional area (CSA), muscle thickness (MT), and echo

intensity (EI; a proxy for muscle quality) using a portable

brightness mode (B-mode) ultrasound machine (GE/Logic

e, Wauwatosa, WI) with a 10-MHz linear array probe by the

same researcher for both twins. The researchers marked a

line at the midpoint of the lateral knee joint surface and the

anterior superior iliac spine. Two more lines were drawn 2

and 4cm distal to the midpoint. Constant minimal pressure

was applied with the probe over the skin to avoid compres-

sion of the musculature and transmission gel was used to

improve acoustic collection. Images were acquired as the

transducer was manually moved mediolaterally along the

marked lines on the skin. Three images from each of the

three marked locations were analyzed using publicly avail-

able software. CSA, MT, and EI of each location were deter-

mined by the mean of the three measurements for each loca-

tion. CSA, MT, and EI were taken as the mean values of the

three locations.

Whole muscle strength

Quadriceps strength of the right limb was assessed by maxi-

mum voluntary isometric contraction (MVIC) at 90° of knee

ﬂexion. The participant was seated and fastened securely

to a HUMAC NORM isokinetic dynamometer (CSMI Inc.,

Stoughton, MA). Three extensions at 50, 75, and 90% of

maximal perceived eﬀort were performed as a warm-up.

Testing consisted of maximally extending the leg for ~ 4–6s.

One min of rest was given before repeating the trial (par-

ticipants remained seated in the dynamometer). Four total

attempts were given, and the highest torque was recorded

as the MVIC. Additionally, handgrip strength was tested in

both hands using a Jamar Plus handgrip dynamometer (Pat-

terson Medical, Warrenville, IL). The participants performed

three warm-up repetitions at 50, 75, and 90% of their per-

ceived maximal eﬀort, respectively. They then performed

ﬁve maximal eﬀort trials by squeezing the device as hard as

possible for ~ 2–4s. Volitional rest was taken between trials

and the top score was recorded.

Whole muscle power

Muscle power was measured by a WAnT. After a self-

selected warm-up (~ 5min of light stretching and 5min of

light cycling; Monark ergometer 893E, Monark Exercise

AB, Varburg, Sweden), participants performed ﬁve sprints

(5s each) at ~ 95% of perceived maximal eﬀort (1min rest

between each sprint) prior to a maximal eﬀort WAnT. Partic-

ipants cycled between 30 and 50 rotations per minute (RPM)

for 10s at one-third of the prescribed resistance. After 10s,

European Journal of Applied Physiology

1 3

participants increased their RPM to a near-maximal rate as

the technician loaded the basket to the full resistance (7.5%

body mass). The basket was immediately dropped and the

participants cycled for 30s at maximum eﬀort. The highest

power produced over a 5-s period was considered the peak

power (PP). Anaerobic capacity was deﬁned by the aver-

age power throughout the 30s. Absolute power decline was

considered as PP subtracted by the lowest 5s power output.

Muscle biopsy

During Visit 3, after an overnight fast (> 12h) and refraining

from exercise (≥ 16h since VO

max test), participants rested

supine for 30min. A mark was made mid-muscle belly (half-

way between the greater trochanter and patella), cleaned

with iodine and anesthetized with 1% lidocaine/xylocaine

(without epinephrine). A small incision was made, then tis-

sue (~ 100mg) was obtained using the Bergström technique

with applied suction as described previously (Murach etal.

2016; Bagley etal. 2017; Tobias etal. 2018). The muscle

samples were immediately cleansed of excess blood, fat and

connective tissue, cut into multiple ~ 15mg bundles, and

either (1) stored in cold skinning solution [(in mm): 125K

propionate, 2.0 EGTA, 4.0 ATP, 1.0 MgCl

, 20.0 imidazole

(pH 7.0), and 50% (v/v) glycerol] at 4°C or (2) rapidly fro-

zen in liquid nitrogen, and stored at − 80°C. The incision

site was cleaned and covered with sterile gauze and cohesive

bandage tape.

Muscle ﬁber type

Muscle ﬁber composition was classiﬁed by myosin heavy

chain protein (MHC) isoform, as well as myosin heavy chain

gene (MyHC) expression. MHC isoforms were identiﬁed

using both single ﬁbers (to determine ﬁber type % distribu-

tion) and homogenized samples (to determine % area that

each ﬁber type comprises).

Detailed methods for single muscle ﬁber typing via

SDS-PAGE were published previously (Murach etal.

2016; Bagley etal. 2017; Tobias etal. 2018). Brieﬂy, indi-

vidual muscle ﬁbers were randomly selected and extracted

longitudinally in a physiological buﬀer using ﬁne-tipped

tweezers under a light microscope at room temperature and

placed into 80µl SDS buﬀer [10% SDS, 6mg ml

−1

EDTA,

0.06m Tris (pH 6.8), 2mg ml

−1

bromophenol blue, 15%

glycerol, and 5% b-mercaptoethanol]. For homogenate

analysis, entire bundles were mechanically homogenized,

and boiled for 90s then placed in 1ml of sodium dodecyl

sulfate (SDS) buﬀer. Aliquots were loaded into individual

wells in a 3.5% loading gel and 5% separating gel and run

at 5°C (SE 600 Series; Hoefer, San Francisco, CA, USA)

and silver stained as described previously (Murach etal.

2016; Bagley etal. 2017; Tobias etal. 2018). Fiber types

were identiﬁed as MHC I, I/IIa, IIa, IIa/IIx, IIx, or I/IIa/IIx

for single ﬁbers. For homogenate samples, band densitom-

etry was analyzed via ImageJ software (National Institutes

of Health, Bethesda, MD) to calculate the relative amount

of MHC I, MHC IIa, and MHC IIx isoform in each sample.

The TT and UT muscle samples were also analyzed for

relative levels of MyHC 1 (MYH7), MyHC 2a (MYH2),

and MyHC 2x (MYH1) mRNAs (see Muscle gene expres-

sion section for detailed methods).

AMPK protein expression

5′AMP-activated protein kinase (AMPK) protein subu-

nits were analyzed as markers of cellular metabolism.

Primary antibodies for AMPK α1 (ab32047) and AMPK

γ1 (ab32508) were purchased from AbCam (Cambridge,

MA). The primary antibody for tubulin (2125) and the sec-

ondary HRP-conjugated anti-rabbit IgG antibody (7074)

were from Cell Signaling Technology. Detailed methods

for AMPK protein analysis were published previously

(Tobias etal. 2018). Brieﬂy, ﬁber bundles were homog-

enized using a pellet pestle in Buﬀer A (1% SDS, 23mM

EDTA, 10% glycerol, 400mM β-mercaptoethanol, pH

6.8), incubated at 95°C for 5min, solubilized and stored

at − 20°C until analysis. Total protein concentration was

measured using a Bradford assay and adjusted via dilu-

tion in Buﬀer A to normalize total protein loading. Fiber

homogenates stored in Buﬀer A were combined with 20%

Buﬀer B (1% SDS, 23mM EDTA, 10% glycerol, 400mM

β-mercaptoethanol, pH 6.8) for western blot analysis of

AMPK isoforms. Samples were run on an 8% separating

polyacrylamide gel at 200V for 90min using Buﬀer C

(β-mercaptoethanol, pH 6.8) and wet transferred to PVDF

membrane for 3h at 80V using a transfer buﬀer contain-

ing 192mM glycine, 25mM Tris-base and 20% methanol.

Immediately following transfer, membrane was blocked

for 60min in 5% non-fat milk powder (BioRad, Hercu-

les, CA, USA) dissolved in Buﬀer D (20mM Tris-base,

150mM NaCl, 0.05% Tween-20). Membrane was then

rinsed (3×, 5min each) in Buﬀer D and rocked overnight

at 4°C in primary antibody diluted in 5% BSA in Buﬀer D

(1:2000 for AMPK α1 and 1:1000 for tubulin) or 5% milk

in Buﬀer D (1:1000 for AMPK γ1). Following another

round of rinsing three times in Buﬀer D, membrane was

then rocked at room temperature for 90min in secondary

antibody diluted 1:2000 in 5% non-fat milk in Buﬀer D.

Following a ﬁnal round of rinsing three times in Buﬀer D,

membrane was developed with SuperSignal

™

West Femto

enhanced chemiluminescent (ECL) substrate diluted 1:10

in SuperSignal

™

West Pico ECL substrate and imaged

using an Omega Lum™ C imaging system (Aplegen Gel

Company, San Francisco, CA, USA).

European Journal of Applied Physiology

1 3

Muscle gene expression

Quantitative reverse transcriptase polymerase chain reac-

tion (QRT-PCR) methods were used to quantify selected

mRNAs associated with: (1) skeletal muscle fiber type

[MyHC 1 (MYH7), MyHC 2a (MYH2), MyHC 2× (MYH1)]

(Welle etal. 1999), (2) oxidative metabolism [transcription

factor A of the mitochondria (TFAM), citrate synthase (CS)]

(Mancini etal. 2017; Scribbans etal. 2017), (3) angiogenesis

[endothelial nitric oxide synthase (NOS3), vascular endothe-

lial growth factor (VEGFA)] (Mancini etal. 2017), (4) mus-

cular growth and repair [myostatin (MSTN), Pax7 (PAX7),

mechano-growth factor (IGF1Ec), insulin-like growth factor

a (IGF1Ea), MyoD (MYOD1)] (Hameed etal. 2003; Pessina

etal. 2010; Silvennoinen etal. 2015), and (5) inﬂammatory

responses [TWEAK (TNFSF12), FN14 TWEAK receptor

(TNFRSF12A), tumor necrosis factor-α (TNF)].

Brieﬂy, a portion of each sample (stored at − 80°C) was

used for total RNA isolation using Trizol, puriﬁed with

RNAqueous micro kit reagents and minicolumns (Invitrogen,

ThermoFisher, Waltham, MA, USA) and stored at − 80°C.

Reverse transcription was performed using Vilo Super-

script IV reverse transcriptase (Invitrogen, ThermoFisher,

Waltham, MA, USA) and random primers. The relative

levels of each mRNA were quantiﬁed using the speciﬁc

primer pairs (see Table1) in a real-time Sybr Green-based

RT-PCR procedure using the Fast SYBR

™

Green Master

Mix (Applied Biosystems, ThermoFisher, Waltham, MA,

USA). For QRT-PCR quantiﬁcation, the cycle thresholds

were obtained for β-2-microglobulin (B2M) and cyclophilin

(PPIA) as normalizing genes (Li etal. 2015) and the ∆∆CT

method was used for calculation of relative expression levels

(Livak and Schmittgen 2001). Cycle threshold levels were

set at the mid-exponential phase for each gene from all sam-

ples (in triplicate) in a single run.

Data analysis

Percent diﬀerences between TT and UT were analyzed for

select variables using MS Oﬃce Excel 2013 (Microsoft,

Redmond, WA, USA). Relative gene expression data were

calculated as described above (presented in arbitrary units).

Results

Exercise histories

The twins exhibited divergent physical patterns since the

mid-1980s. Both TT and UT participated in recreational

sports (e.g., baseball and basketball) from ages 10–20

(1973–1983). UT remained recreationally active in bas-

ketball and light cycling (~ 4× per week) from ages 20–39

(1983–1999), but has not been involved in recreational

sports and been otherwise relatively inactive since ~ 1999

(around this time, UT sustained a moderate ankle injury with

pain that has persisted to date).

TT began running cross-country/track competitively in

high school, competed in collegiate cross-country/track, and

was selected to All-Conference Team in 1985. After college,

Table 1 Primers for selected genes used for polymerase chain reaction (PCR)

Primers for TWEAK, FN14 and TNF were designed for this study using published sequences in the NCBI database

Human gene name (common name) Forward primer (5′–3′) Reverse primer (5′–3′)

MYH1 (MyHC 2×) (38) TTT ATC TAA CTG CTG AAA GGT GAC TCT CCA AAA GTC ATA AGT ACA AAA TG

MYH2 (MyHC 2a) (38) ATG TCC TGA TGC CATGG CAA ACT ACC CTA TGC TTT ATT TCC

MYH7 (MyHC 1) (38) CTT TGC CAC ATC TTG ATC TG TGC TTT ATT CTG CTT CCT CC

TFAM (T

Fam

) (24) CGC TCC CCC TTC AGT TTT GT CAC TCC GCC CTA TAA GCA TC

CS (citrate synthase) (31) ACT GTG GAC ATG ATG TAT GGTG GTA GCA GTT TCT GGC ATT CAG

NOS3 (eNOS) (24) CGG GGA TTC TGG CAG GAG C CGT AGG TCT TGG GGT TGT CA

VEGFA (VEGF) (24) GAG GAA AGG GAA AGG GGC A CTC GGC TTG TCA CAT CTG C

MSTN (myostatin) (33) CTA CAA CGG AAA CAA TCA TTA CCA GTT TCA GAG ATC GGA TTC CAG TAT

PAX7 (Pax7) (27) CAC TGT GAC CGA AGC ACT GT GTC AGG TTC CGA CTC CAC AT

TNFSF12 (TWEAK) CGA TCG CAG CCC ATT ATG AAG TGT TGA TTC TGG CTT CCT CCC

TNFRSF12A (FN14) CTC TGA GCC TGA CCT TCG TG GTC TCC TCT ATG GGG GTG GT

TNF (TNA-α) GCT GCA CTT TGG AGT GAT CG CTT GTC ACT CGG GGT TCG AG

IGF1Ec (MGF) (12) CGA AGT CTC AGA GAA GGA AAGG ACA GGT AAC TCG TGC AGA GC

IGF1Ea (IGF-1) (12) GCC TGC TCA CCT TCA CCA GC TCA AAT GTA CTT CCT TCT GGG TCT TG

MYOD1 (MyoD) (12) GCA GGT GTA ACC GTA ACC ACG TAC AAA TTC CCT GTA GC

PPIA (cyclophilin) (22) TCC TGG CAT CTT GTC CAT TGC TGG TCT TGC CAT TCC T

B2M (β-2-microglobulin) (22) CTA TCC AGC GTA CTC CAA AG GAA AGA CCA GTC CTT GCT GA

European Journal of Applied Physiology

1 3

TT continued endurance training and began competing in

running road races (best marathon time: 3:01:07 in 1993).

His best marathon times were relatively consistent over 20

years, with only a 6.9% increase in race time from 1985

(3:03:10) to 2005 (3:15:05). From ages 30–40 (1993–2003)

TT continued endurance training and began resistance train-

ing (2–3× per week). From ages 40–50 (2003–2013), TT

recorded ~ 34,195 running kilometers (~ 21,248 miles). He

also engaged in cycling, resistance training, and swimming

2–3× per week. He competed in one Ironman triathlon (time:

12:33:59), two half Ironman triathlons, two Olympic triath-

lons, two marathons, and 45 other running races.

Pre‑study physical activity anddietary habits

Self-recorded physical activity patterns diﬀered between

TT and UT during the week before testing. TT reported

averaging 69min of endurance exercise per day [running

at 5.3min/km (8.5min/mile) pace] and three days of upper

body resistance training per week. UT recorded an aver-

age of 22min of walking each day. This resulted in an

estimated WEEE of 7594kcal for TT and 204kcal for UT

(189.5% diﬀerence). TT reported consuming an average

of 3517 ± 515kcal per day (51% carbohydrate, 33% fat,

and 15% protein), while UT consumed 1,741 ± 787kcal

per day (31% carbohydrate, 42% fat, and 27% protein). TT

consumed ~ 173kcal over his estimated total daily energy

expenditure (TDEE), while UT consumed ~ 208kcal over

his estimated TDEE (TDEE includes estimated resting meta-

bolic rate and energy expenditure from physical activity).

Anthropometric measures

TT was 186cm tall and weighed 94.0kg (BMI: 27.2kg/

, overweight), while UT was 183cm tall and weighed

104.5kg (BMI: 31.2kg/m

, obese class I).

Body composition

Total LM was the same between twins (74.72 vs. 74.64kg),

but UT had 47.3% greater total body fat mass. In addition,

UT demonstrated 43.3% greater estimated VAT mass and

area than TT (551 vs. 355g). Both total BMC and BMD

were similar between twins (see Table2).

Resting heart rate, blood pressure, andblood

proﬁles

RHR was 30.3% higher in UT (42 vs. 57bpm). Resting BP

for the TT was 122/57mmHg (Category: Elevated) and

132/77mmHg for UT (Category: Stage 1 hypertension)

(Whelton etal. 2017). TT displayed lower total cholesterol,

LDL, triglycerides, plasma glucose, and cholesterol/HDL

(2.7 vs. 3.5) (see Table2).

Maximal oxygen consumption

TT had a 30.1 and 23.8% higher relative (47.5 vs. 35.1ml/

kg/min) and absolute (4.66 vs. 3.67L/min) VO

max, respec-

tively. The ﬁnal completed stages were 350W for TT and

300W for UT. Maximum HR during the test was 157bpm

for TT and 178bpm for UT (11.8% diﬀerence). Peak RER

was 1.17 for TT and 1.25 for UT (6.4% diﬀerence) and maxi-

mum ventilation (VE) was of 124.2L/min for TT and 118.7

for UT (4.4% diﬀerence). Final RPE was 18 for TT and 19

for UT (on the 6–21 Borg Scale).

Pulmonary function

FVC was similar (6.01 vs. 5.88l) for TT and UT, respec-

tively. However, UT slightly exceeded TT in FEV

(4.47 vs.

4.76l, 6.3% diﬀerence) and FEV

% (74.4 vs. 81.0%, 8.5%

diﬀerence).

Muscle size andperformance

Average CSA and MT of the vastus lateralis diﬀered by 4.5%

(larger in UT; 1,050 vs. 1098cm

and 15.4 vs. 16.1cm,

respectively). EI was also higher in the UT (64.1 vs. 77.1

AU), but to a much larger magnitude (18.4%).

Muscle strength

The TT knee extensors produced 60% less torque during the

IMVC (137 vs. 254Nm). When normalized to LM, TT still

produced less torque (1.83 vs. 3.40Nm/kg, 60% diﬀerence).

Table 2 Body composition and blood proﬁles of male monozygotic

(MZ) twins with 30 years of discordant exercise training histories

NF% body fat percentage, BMC bone mineral content, BMD bone

mineral density, LDL low-density lipoprotein, HDL high-density lipo-

protein, HbA1c glycosylated hemoglobin

Body composition Trained

twin (TT)

Untrained

twin (UT)

% Diﬀerence

BF% 19.2 27.8 − 36.6

BMC (g) 3.65 3.58 1.9

BMD (g/cm

) 1.39 1.35 2.9

Fasted blood proﬁle

Total cholesterol (mg/dl) 172 202 − 16.0

LDL (mg/dl) 92 124 − 29.6

HDL (mg/dl) 64 57 11.6

Triglycerides (mg/dL) 81 107 − 27.7

Plasma glucose (mg/dL) 86 102 − 17.0

HbA1c (%) 5.8 5.7 1.7

European Journal of Applied Physiology

1 3

Similarly, TT produced 24 and 17% less force in his right

(44.3 vs. 56.5kg) and left hand (43.7 vs. 51.7kg) than UT,

respectively. When normalized to LM TT still displayed less

force in both hands (right = 59 vs. 76% of LBM, 25.2% dif-

ference; left = 58 vs. 69% of LM, 17.3% diﬀerence).

TT and UT displayed similar absolute peak power dur-

ing the WAnT (788 vs. 784W). However, TT had a greater

relative peak power (8.38 vs. 7.53W/kg, 10.7% diﬀerence)

and 43.7% less absolute power decline (237 vs. 370W) and

power drop (34.2% vs. 48.9%, 5.33% diﬀerence). TT dis-

played higher average absolute power (665 vs. 585W, 12.9%

difference) and higher average relative power (7.08 vs.

5.62W/kg, 23.0% diﬀerence). Despite this, TT took 46.7%

more time to reach peak power than UT (3.20 vs. 1.99s).

Muscle ﬁber type

Single ﬁber MHC

A total of 213 single ﬁbers were analyzed for MHC isoform.

Fiber-type distributions for TT and UT are shown in Fig.1a.

TT had 2.4× more MHC I ﬁbers, 13.3× fewer MHC II ﬁbers,

and 10× fewer hybrid ﬁbers (i.e., MHC I/IIa, IIa/IIx, and I/

IIa/IIx).

Homogenate MHC

The composition of MHC content (i.e., from homogenate

muscle samples) displays the total myoﬁbrillar protein area

that each MHC-type encompasses. These data mirrored the

single ﬁber MHC ﬁndings (see Fig.1b). TT muscle had 1.5

times more MHC I, 5 times less MHC IIa and 14 times less

MHC IIx protein vs. UT.

MHC gene expression

As shown in Fig.1-C, MyHC 2x was expressed at a lower

level in TT relative to UT. No diﬀerences were observed in

MyHC I or IIa expression between subjects.

AMPK protein expression

Relative isoform expression was measured for AMPK

α1 (catalytic subunit) and AMPK γ1 (regulatory subu-

nit) with tubulin as a standard loading control in the ﬁber

Fig. 1 Vastus lateralis ﬁber

type in monozygotic (MZ)

twins [trained twin (TT) vs.

untrained twin (UT)]. a Single

muscle ﬁber myosin heavy

chain (MHC) ﬁber-type percent

(%) distribution (total hybrids:

MHC I/IIa, IIa/IIx, and I/IIa/

IIx ﬁbers). b Homogenate

MHC ﬁber-type % composition

(unable to identify hybrids with

homogenate method). c Relative

myosin heavy chain (MyHC)

mRNA expression at rest

(arbitrary units, AU). MyHC I

(MYH7), MyHC IIa (MYH2),

MyHC IIx (MYH1)

European Journal of Applied Physiology

1 3

homogenates (Fig.2). TT exhibited higher expression of

both isoforms.

Muscle gene expression

No diﬀerences were observed between TT and UT in mark-

ers of oxidative metabolism or angiogenesis gene products

(Fig.3a, b). Some diﬀerences were observed in the expres-

sion of genes related to muscle growth and repair, including

elevations in PAX7, IGF1Ec and IGF1Ea, and lower MYOD1

in the TT (Fig.3c). The only diﬀerence in the expression of

inﬂammatory response markers was an elevation in FN14

in TT (Fig.3d).

Discussion

This study represents the most comprehensive physiologi-

cal comparison of MZ twins with this length and magni-

tude of diﬀering exercise histories to date. Major ﬁndings

were that TT exhibited: (1) lower total body mass and %BF,

(2) lower RHR, BP, total cholesterol, LDL, triglycerides,

and plasma glucose levels, (3) a higher aerobic capacity

(VO

max), (4) lower quadriceps muscle size and strength,

poorer muscle quality, but greater relative cycling power

and anaerobic endurance, (5) more MHC I (slow-twitch) and

fewer MHC IIa (fast-twitch) and hybrid muscle ﬁbers, (6)

greater muscle AMPK α1 and γ1 protein expression, and (7)

elevated expression of muscle growth, repair, and inﬂamma-

tory mRNA markers (i.e., PAX7, IGF1Ec, IGF1Ea, FN14)

compared to the UT. The magnitude of diﬀerence in several

of the health, cardiovascular, and muscular strength meas-

ures are larger than research on MZ twins with 3–30years

of discordant physical activity patterns (Hannukainen etal.

2011; Rottensteiner etal. 2015; Leskinen etal. 2010) (see

Table3). The variables with the largest diﬀerences between

twins (i.e., > 24%) were TRIG, LDL, WEEE, body composi-

tion, VO

max, grip strength, and leg strength; all of which

are independent and signiﬁcant predictors of mortality (Vol-

aklis etal. 2015).

Body composition diﬀered to a greater extent than similar

twin research (Leskinen etal. 2010) (see Table3). Total

muscle mass was similar between TT and UT, meaning the

diﬀerences in total body mass (10.5kg) were almost entirely

explained by fat mass (17.8 vs. 28.8kg). When compared

to age- and sex-speciﬁc normative data for %BF, UT was

placed in the 38th percentile, while TT was in the second

percentile (Kelly etal. 2009). These body composition dif-

ferences may not be explained solely by the divergent exer-

cise habits as other lifestyle factors likely contributed as well.

For example, our short-term dietary tracking data indicate

UT consumed 35 more kcals over his estimated TDEE than

TT. This small diﬀerence may account for ~ 7kg of addi-

tional body mass [(35 kcal*52 weeks*30years)/7700kcal/

kg] when extrapolated over 30years. These might explain

why UT demonstrated ~ 43% greater VAT mass than TT.

As anticipated, our data show that chronic endurance

training improves most markers of cardiovascular health.

Previous studies have reported an average decrease of

3–5bpm in RHR after 10–20 weeks of endurance training

(Cornelissen etal. 2010; Wilmore etal. 1996), which is far

less than the 15bpm diﬀerence found in the current study.

In concert, a meta-analysis concluded that on average, exer-

cise reduces SBP by 3.84mmHg and DBP by 2.58mm Hg

(Whelton etal. 2002). Our data suggest the physiological

potential for a much larger diﬀerence as TT was 10mmHg

(SBP) and 20mm Hg (DBP) lower than UT. The length of

the discordance in activity (30years) could likely explain

the dissimilarity as most of the studies included in the

meta-analysis only ranged from 3 weeks to 2years. TT also

demonstrated healthier CHOL, TRIG, LDL, HDL, and fast-

ing blood glucose concentrations, which agrees with other

research (Leskinen etal. 2010; Rottensteiner etal. 2015).

However, the magnitude of diﬀerence between our twins

was similar (HDL), less than (TRIG), and greater (fasting

glucose and CHOL) than comparable research (Leskinen

etal. 2010; Rottensteiner etal. 2015) (see Table3).

VL size (CSA and MT) was similar between TT and

UT. Yet, VL quality (EI) and overall strength were better

in UT. This has important health implications as quadricep

Fig. 2 Relative expression of

5′AMP-activated protein kinase

(AMPK) protein isoforms,

AMPK α1 (catalytic subunit)

and AMPK γ1 (regulatory

subunit) with tubulin as a stand-

ard loading control in vastus

lateralis homogenate samples

from the trained twin (TT) vs.

untrained twin (UT). AMPK γ1

was below the limit of detection

in UT

European Journal of Applied Physiology

1 3

EI correlates with functional strength in middle-aged adults

(Fukumoto etal. 2012). The lower muscle quality in TT

diﬀers from previous ﬁndings (Leskinen etal. 2009), where

inactive twins exhibited more intramuscular thigh fat com-

pared to their trained counterparts. This diﬀerence could be

explained by our use of ultrasound vs. magnetic resonance

imaging (MRI) used by Leskinen etal. 2009. While MRI is

considered a “gold standard” to measure intramuscular fat

and muscle size, there is no general consensus on standard

assessment tools for muscle quality. Diagnostic ultrasound

is gaining popularity as an imaging modality due to its port-

ability and relative aﬀordability (Correa-de-Araujo etal.

2017).

The stronger grip of UT is possibly a result of his delivery

truck occupation, yet this remains speculative. Explanation

for the diﬀerences in VL quality and strength is also not

clear, and counter to previous research which suggested the

more active twin should be favored (Leskinen etal. 2010).

Fig. 3 Resting vastus later-

alis mRNA expression of

select genes related to skeletal

muscle: a oxidative metabo-

lism [transcription factor A

of the mitochondria (TFAM),

citrate synthase (CS)], b

angiogenesis [endothelial

nitric oxide synthase (eNOS),

vascular endothelial growth

factor (VEGF)], c growth and

repair [myostatin (MSTN),

Pax7 (PAX7), mechano-growth

factor (MGF), insulin-like

growth factor a (IGF-1), MyoD

(MYOD1)], and d inﬂammatory

responses [TWEAK (TNFSF12),

FN14 (TWEAK receptor;

TNFRSF12A), tumor necrosis

factor-α (TNF-α)] in the trained

twin (TT) vs. untrained twin

(UT). Relative gene expression

represented in arbitrary units

European Journal of Applied Physiology

1 3

One difference between our study and similar research

(Leskinen etal. 2010) is our TT utilized much higher train-

ing volume and a diﬀerent exercise mode. In regard to mus-

cle power, PP during the WAnT did not diﬀer between TT

and UT, but it took TT almost twice (47%) as long to reach

PP. A previous study by Jacob and colleagues also found

endurance-trained individuals produced signiﬁcantly less

power during a WAnT than both untrained and sprint-trained

individuals (Jacob etal. 2002). Thus, it is unclear if TT’s

decades of endurance-running training, or other lifestyle fac-

tors, compromised the quadriceps quality, torque production,

and time to PP.

As expected, TT displayed markedly higher anaerobic and

aerobic endurances. However, the magnitude of diﬀerence

between the twins (> 30%) was larger than most previous

reports for VO

max (Table3), albeit still within the range

proposed by previous heritability research (Schutte etal.

2016). In our study, TT’s relative VO

max placed him in

the 90th percentile for males ages 50–59, while UT’s value

placed him in the 60th percentile (Riebe etal. 2018). The

divergence in maximal cardiovascular ﬁtness has important

health implications for the twins as each metabolic equiva-

lent of task (1 MET) decrease all-cause mortality by 13–15%

(Kodama etal. 2009). The minimal diﬀerences in maximal

spirometry found here is consistent with the prior conclusion

that these measures change little with exercise training in

healthy individuals. However, previous studies report elite

marathon runners possess ~ 7% higher FVC and FEV1% than

age-matched controls (Eastwood etal. 2001), indicating a

greater reliance on genetic inheritance than exercise habits.

MHC ﬁber-type distribution is highly variable among

individuals, and it has been suggested that ~ 45% of one’s

ﬁber-type variance is solely determined by genetic factors

(Simoneau and Bouchard 1995). To provide a comprehen-

sive view of muscle ﬁber type, we analyzed both MHC

ﬁber-type distribution (i.e., ﬁber type of single muscle ﬁb-

ers) and homogenate MHC ﬁber-type composition (i.e.,

ﬁber type of homogenized muscle samples). As shown in

Fig.1 (A, B), compared to UT, TT’s VL had a greater per-

cent composition of MHC I muscle ﬁbers, as well as more

total MHC I ﬁbers and fewer hybrid ﬁbers. The UT ﬁber-

type distribution (~ 40% MHC I, ~ 40% MHC IIa, ~ 30%

hybrids) was similar to previous research in healthy/non-

exercise-trained males (~ 36% MHC I, ~ 32% MHC IIa,

~ 32% hybrids) (Dickinson etal. 2010). The predominance

of MHC I ﬁbers (and fewer hybrid and MHC IIx ﬁbers)

in TT was expected, but 94% MHC I ﬁbers is amongst the

highest documented in the literature. Succinctness among

the single ﬁber, homogenate, and gene data support the

pronounced bias away from MHC IIx expressing ﬁbers in

TT. Interestingly, basal MyHC I and IIa gene expression

levels were similar between twins, suggesting that down-

stream mechanisms may dictate MHC protein isoform

synthesis. This further highlights the sometimes limited

syncing of single acute measures of molecular mecha-

nisms with long-term eventual protein expression. Another

striking ﬁnding was the magnitude of diﬀerence between

ﬁber-type compositions in the twins. Previous estimates

suggest ~ 30% of diﬀerences in MHC I ﬁber type between

individuals may be explained exclusively by environment

factors (Simoneau and Bouchard 1995). Here we report

a much larger diﬀerence in MHC I ﬁbers between twins,

highlighting the magnitude of muscle adaptability with

exercise training may be more reliant on exposomal factors

than previously thought.

AMPK regulates cellular metabolism and is activated

during acute exercise (Kjobsted etal. 2018). The cata-

lytic subunit isoform (AMPK α1) and regulatory subunit

isoform (AMPK γ1) are known to form a complex with

AMPK β2 that is only activated during long duration,

low-intensity exercise (Kjobsted etal. 2018). We found

higher expression of both AMPK α1 and γ1 isoforms in

TT, which parallels previous research showing elevated

α1 and γ1 protein concentrations after chronic endurance

Table 3 Percent diﬀerences (%)

in body composition, muscular

strength, and cardiorespiratory

ﬁtness variables between the

trained twin (TT) and untrained

twin (UT) with 30 years of

discordant exercise training

histories (current study)

compared to the literature (13,

20, 29)

↑ TT greater than UT, ↓ TT less that UT, 3 years 3years of discordance between twins, 30 years 30years

of discordance between twins, BMI body mass index, FM fat mass, FFM fat free mass, BF% body fat per-

centage, IMVC isometric voluntary contraction (quadriceps), Rel. VO

max relative maximal oxygen con-

sumption (ml/kg/min)

Variable Current study 3years (Hannukainen

etal. 2011)

3years (Rottensteiner

etal. 2015)

30years (Leski-

nen etal. 2010)

Body mass ↓ 10.6 03.8 02.6 08.7

BMI ↓ 13.7 04.1 03.6 07.4

FM ↓ 47.3 16.4 18.2 32.3

FFM ↓ 00.1 00.8 02.5 00.9

BF% ↓ 36.6 10.8 14.8 22.7

IMVC ↓ 59.9 – 04.6 17.6

Grip strength ↓ 16.8 00.6 – 01.3

Rel. VO2max ↑ 30.1 15.9 15.6 20.7

European Journal of Applied Physiology

1 3

training (Frosig etal. 2004). Thus, these two isoforms may

be particularly associated with long-term endurance exer-

cise adaptations.

Diﬀerences in key mRNA markers of muscle growth,

repair, and inﬂammation were also observed (Fig.3). Curi-

ously, TT did not show higher expression of genes related to

oxidative capacity (TFAM or CS) or angiogenesis (NOS3 or

VEGFA). These results were unexpected, considering eleva-

tions in oxidative capacity and muscle capillarization are

hallmark adaptations with endurance training. Three possi-

ble reasons could explain these ﬁndings. One, TT may have

reached a new steady state for mitochondrial and capillary

density and no longer required elevated expression of the

related genes. Second, it is possible, yet unlikely, that TT

was in a detraining state and mRNAs associated with oxida-

tive capacity and angiogenesis returned to baseline. Third,

since the muscle biopsies were obtained approximately

48h after the muscle performance testing, it is possible that

UT would have a greater response to the muscle activity

required for the performance analyses resulting in elevations

in TFAM, CS, NOS3 (eNOS), and VEGFA expression that

approached and superseded those of TT.

A clear elevation existed in the expression of two splice

variants of the IGF-1 gene in TT. Both mechano-growth fac-

tor (IGF-1Ec) and IGF-1Ea were elevated more than 50% in

TT. As both of these isoforms are secreted by skeletal mus-

cle and thought to activate muscle satellite cells (Hameed

etal. 2003), we evaluated the expression of Pax7 (a marker

for satellite cell number). A modest elevation in Pax7 was

observed for TT, suggesting elevated satellite cell content,

which would be expected with endurance training and more

MHC I ﬁbers (slow muscle ﬁbers have more myonuclei than

fast ﬁbers) (Allen etal. 1995). Thus, as ﬁbers were trans-

formed from fast to slow a greater number of myonuclei

would likely be required to maintain the elevated expression

for nuclear DNA-encoded mitochondrial proteins. Addition-

ally, MyoD (a marker for muscle precursor cell diﬀeren-

tiation) was decreased in TT, likely reﬂecting the smaller

proportion of MHC IIa muscle ﬁbers as relative levels of

MyoD are highly correlated with fast muscle ﬁbers (Hughes

etal. 1993).

TT showed an elevated FN14 gene expression, a recep-

tor for TWEAK (Tajrishi etal. 2014). The proinﬂammatory

cytokine system involving TWEAK, TNFa, and FN14 have

been implicated in inducing catabolic changes in muscle

with sarcopenia (Tajrishi etal. 2014). While the precise

roles of TWEAK-FN14 signaling in human skeletal muscle

require more research, early investigations suggest moder-

ate intensity endurance exercise elevates FN14 (Raue etal.

2015). The modest elevation in FN14 mRNA in TT in our

study likely reﬂects his primarily MHC I ﬁber-type composi-

tion, as resting FN14 levels have been shown to be higher in

MHC I vs. IIa ﬁbers (Trappe etal. 2015).

This is the ﬁrst study to perform a comprehensive analysis

of skeletal muscle health and physiological performance in

MZ twins with both long-term and large divergences in exer-

cise habits. Magnitudes of diﬀerence in several measures are

the largest reported between MZ twins (i.e., TT expressed

55% more MHC I ﬁbers, 12.4ml/kg/min greater VO

max,

and 8.6% lower %BF vs. UT). Our ﬁndings support utilizing

chronic endurance exercise training to improve body com-

position and cardiovascular health and suggest these physi-

ological systems exhibit greater plasticity than previously

thought. These results highlight the need to further study

the exposome’s role on physiological adaptations across the

lifespan.

Acknowledgements The authors would like to thank Kathryn McLe-

land, Cassio Ruas, Nathan Serrano, Kara Lazauskas, and Colleen

Gulick for their assistance with this project. This research was funded

by a California State University Development of Research and Creativ-

ity (CSU-DRC) Grant to J.R. Bagley.

Author contributions JRB and AJG conceived and designed this work.

KEB, JRB, EJ, RJT, IST, JAA, and AJG performed the experiments.

All authors collected and analyzed the data. KEB, JRB, LEB, JWC,

NLS, and AJG interpreted the results of experiments. KEB, AJG,

and JRB drafted the manuscript. All authors read and approved the

manuscript.

Compliance with ethical standards

Conflict of interest The authors declared no conﬂicts of interest.

Ethical standards All procedures performed in this study were in

accordance with the ethical standards of the University’s Institutional

Review Board for Human Subjects and with the 1964 Declaration of

Helsinki and its later amendments.

Informed consent Informed consent was obtained from all individual

participants included in the study.

References

Adams TD, Yanowitz FG, Fisher AG, Ridges JD, Nelson AG, Hagan

AD, Williams RR, Hunt SC (1985) Heritability of cardiac size: an

echocardiographic and electrocardiographic study of monozygotic

and dizygotic twins. Circulation 71(1):39–44

Allen DL, Monke SR, Talmadge RJ, Roy RR, Edgerton VR (1995)

Plasticity of myonuclear number in hypertrophied and atro-

phied mammalian skeletal muscle ﬁbers. J Appl Physiol (1985)

78(5):1969–1976

Bagley JR, McLeland KA, Arevalo JA, Brown LE, Coburn JW, Gal-

pin AJ (2017) Skeletal muscle fatigability and myosin heavy

chain ﬁber type in resistance trained Men. J Strength Cond Res

31(3):602–607

Boomsma D, Busjahn A, Peltonen L (2002) Classical twin studies and

beyond. Nat Rev Genet 3(11):872–882

Cornelissen VA, Verheyden B, Aubert AE, Fagard RH (2010) Eﬀects

of aerobic training intensity on resting, exercise and post-exercise

blood pressure, heart rate and heart-rate variability. J Hum Hyper-

tens 24(3):175–182

European Journal of Applied Physiology

1 3

Correa-de-Araujo R, Harris-Love MO, Miljkovic I, Fragala MS,

Anthony BW, Manini TM (2017) The need for standardized

assessment of muscle quality in skeletal muscle function deﬁcit

and other aging-related muscle dysfunctions: a symposium report.

Front Physiol 8:87. https ://doi.org/10.3389/fphys .2017.00087

Danis A, Kyriazis Y, Klissouras V (2003) The eﬀect of training in male

prepubertal and pubertal monozygotic twins. Eur J Appl Physiol

89(3–4):309–318

Dickinson JM, Lee JD, Sullivan BE, Harber MP, Trappe SW, Trappe

TA (2010) A new method to study invivo protein synthesis in

slow- and fast-twitch muscle ﬁbers and initial measurements in

humans. J Appl Physiol (1985) 108(5):1410–1416

Eastwood PR, Hillman DR, Finucane KE (2001) Inspiratory mus-

cle performance in endurance athletes and sedentary subjects.

Respirology 6(2):95–104

Fagard R, Van Den Broeke C, Bielen E, Amery A (1987) Maximum

oxygen uptake and cardiac size and function in twins. Am J Car-

diol 60(16):1362–1367

Fagard R, Bielen E, Amery A (1991) Heritability of aerobic power

and anaerobic energy generation during exercise. J Appl Physiol

(1985) 70(1):357–362

Frosig C, Jorgensen SB, Hardie DG, Richter EA, Wojtaszewski JF

(2004) 5′-AMP-activated protein kinase activity and protein

expression are regulated by endurance training in human skeletal

muscle. Am J Physiol Endocrinol Metab 286(3):E411–E417

Fukumoto Y, Ikezoe T, Yamada Y, Tsukagoshi R, Nakamura M, Mori

N, Kimura M, Ichihashi N (2012) Skeletal muscle quality assessed

from echo intensity is associated with muscle strength of middle-

aged and elderly persons. Eur J Appl Physiol 112(4):1519–1525

Hameed M, Orrell RW, Cobbold M, Goldspink G, Harridge SD (2003)

Expression of IGF-I splice variants in young and old human

skeletal muscle after high resistance exercise. J Physiol 547(Pt

1):247–254

Hannukainen JC, Borra R, Linderborg K, Kallio H, Kiss J, Lepomaki

V, Kalliokoski KK, Kujala UM, Kaprio J, Heinonen OJ, Komu

M, Parkkola R, Ahotupa M, Lehtimaki T, Huupponen R, Iozzo P,

Nuutila P (2011) Liver and pancreatic fat content and metabolism

in healthy monozygotic twins with discordant physical activity. J

Hepatol 54(3):545–552

Hughes SM, Taylor JM, Tapscott SJ, Gurley CM, Carter WJ, Peter-

son CA (1993) Selective accumulation of MyoD and myogenin

mRNAs in fast and slow adult skeletal muscle is controlled by

innervation and hormones. Development 118(4):1137–1147

Jacob C, Zouhal H, Vincent S, Gratas-Delamarche A, Berthon PM,

Bentue-Ferrer D, Delamarche P (2002) Training status (endur-

ance or sprint) and catecholamine response to the Wingate-test in

women. Int J Sports Med 23(5):342–347

Kelly TL, Wilson KE, Heymsﬁeld SB (2009) Dual energy X-Ray

absorptiometry body composition reference values from

NHANES. PLoS One 4(9):e7038

Kjobsted R, Hingst JR, Fentz J, Foretz M, Sanz MN, Pehmoller C,

Shum M, Marette A, Mounier R, Treebak JT, Wojtaszewski JFP,

Viollet B, Lantier L (2018) AMPK in skeletal muscle function and

metabolism. FASEB J 201700442R

Klissouras V, Pigozzi F (2009) Genetic limits of sport performance:

quo vadis? J Sports Med Phys Fitness 49(1):1–5

Kodama S, Saito K, Tanaka S, Maki M, Yachi Y, Asumi M, Sugawara

A, Totsuka K, Shimano H, Ohashi Y, Yamada N, Sone H (2009)

Cardiorespiratory ﬁtness as a quantitative predictor of all-cause

mortality and cardiovascular events in healthy men and women:

a meta-analysis. JAMA 301(19):2024–2035

Leskinen T, Sipila S, Alen M, Cheng S, Pietilainen KH, Usenius

JP, Suominen H, Kovanen V, Kainulainen H, Kaprio J, Kujala

UM (2009) Leisure-time physical activity and high-risk fat:

a longitudinal population-based twin study. Int J Obes (Lond)

33(11):1211–1218. https ://doi.org/10.1038/ijo.2009.170

Leskinen T, Rinnankoski-Tuikka R, Rintala M, Seppanen-Laakso T,

Pollanen E, Alen M, Sipila S, Kaprio J, Kovanen V, Rahkila P,

Oresic M, Kainulainen H, Kujala UM (2010) Diﬀerences in mus-

cle and adipose tissue gene expression and cardio-metabolic risk

factors in the members of physical activity discordant twin pairs.

PLoS One 5:(9)

Leskinen T, Usenius JP, Alen M, Kainulainen H, Kaprio J, Kujala UM

(2011) Leisure-time physical activity and artery lumen diameters:

a monozygotic co-twin control study. Scand J Med Sci Sports

21(6):e208–e214

Li X, Yang Q, Bai J, Xuan Y, Wang Y (2015) Identiﬁcation of appropri-

ate reference genes for human mesenchymal stem cell analysis by

quantitative real-time PCR. Biotechnol Lett 37(1):67–73

Livak KJ, Schmittgen TD (2001) Analysis of relative gene expres-

sion data using real-time quantitative PCR and the 2(-Delta Delta

C(T)) Method. Methods 25(4):402–408

Mancini A, Vitucci D, Labruna G, Imperlini E, Randers MB, Schmidt

JF, Hagman M, Andersen TR, Russo R, Orru S, Krustrup P, Sal-

vatore F, Buono P (2017) Eﬀect of lifelong football training on

the expression of muscle molecular markers involved in healthy

longevity. Eur J Appl Physiol 117(4):721–730

Murach KA, Bagley JR, McLeland KA, Arevalo JA, Ciccone AB,

Malyszek KK, Wen Y, Galpin AJ (2016) Improving human skel-

etal muscle myosin heavy chain ﬁber typing eﬃciency. J Muscle

Res Cell Motil

Pessina P, Conti V, Pacelli F, Rosa F, Doglietto GB, Brunelli S, Bossola

M (2010) Skeletal muscle of gastric cancer patients expresses

genes involved in muscle regeneration. Oncol Rep 24(3):741–745

Raue U, Jemiolo B, Yang Y, Trappe S (2015) TWEAK-Fn14 path-

way activation after exercise in human skeletal muscle: insights

from two exercise modes and a time course investigation. J Appl

Physiol 118(5):569–578

Riebe D, Ehrman JK, Liguori G, Magal M (2018) ACSM’s guidelines

for exercise testing and prescription, 10thedn. Wolters Kluwer,

Philadelphia

Rottensteiner M, Leskinen T, Niskanen E, Aaltonen S, Mutikainen

S, Wikgren J, Heikkila K, Kovanen V, Kainulainen H, Kaprio J,

Tarkka IM, Kujala UM (2015) Physical activity, ﬁtness, glucose

homeostasis, and brain morphology in twins. Med Sci Sports

Exerc 47(3):509–518

Schutte NM, Nederend I, Hudziak JJ, Bartels M, de Geus EJ (2016)

Twin-sibling study and meta-analysis on the heritability of maxi-

mal oxygen consumption. Physiol Genomics 48(3):210–219

Scribbans TD, Edgett BA, Bonaﬁglia JT, Baechler BL, Quadrilatero J,

Gurd BJ (2017) A systematic upregulation of nuclear and mito-

chondrial genes is not present in the initial postexercise recov-

ery period in human skeletal muscle. Appl Physiol Nutr Metab

42(6):571–578

Segal NL (2017) Twin mythconceptions: false beliefs, fables, and facts

about twins. Twin mythconceptions. Academic Press, San Diego

Silvennoinen M, Ahtiainen JP, Hulmi JJ, Pekkala S, Taipale RS, Nindl

BC, Laine T, Hakkinen K, Selanne H, Kyrolainen H, Kainulainen

H (2015) PGC-1 isoforms and their target genes are expressed dif-

ferently in human skeletal muscle following resistance and endur-

ance exercise. Physiol Rep 3 (10)

Simoneau JA, Bouchard C (1995) Genetic determinism of ﬁber type

proportion in human skeletal muscle. FASEB J 9(11):1091–1095

Tajrishi MM, Zheng TS, Burkly LC, Kumar A (2014) The TWEAK-

Fn14 pathway: a potent regulator of skeletal muscle biology in

health and disease. Cytokine Growth Factor Rev 25(2):215–225

Thomis MA, Beunen GP, Maes HH, Blimkie CJ, Van Leemputte

M, Claessens AL, Marchal G, Willems E, Vlietinck RF (1998)

Strength training: importance of genetic factors. Med Sci Sports

Exerc 30(5):724–731

Tobias IS, Lazauskas KK, Arevalo JA, Bagley JR, Brown LE, Gal-

pin AJ (2018) Fiber type-speciﬁc analysis of AMPK isoforms in

European Journal of Applied Physiology

1 3

human skeletal muscle: advancement in methods via capillary

nanoimmunoassay. J Appl Physiol (1985) 124(4):840–849

Trappe S, Luden N, Minchev K, Raue U, Jemiolo B, Trappe TA (2015)

Skeletal muscle signature of a champion sprint runner. J Appl

Physiol 118(12):1460–1466

Volaklis KA, Halle M, Meisinger C (2015) Muscular strength as a

strong predictor of mortality: A narrative review. Eur J Intern

Med 26(5):303–310

Welle S, Bhatt K, Thornton CA (1999) Stimulation of myoﬁbrillar

synthesis by exercise is mediated by more eﬃcient translation of

mRNA. J Appl Physiol 86(4):1220–1225

Whelton SP, Chin A, Xin X, He J (2002) Eﬀect of aerobic exercise on

blood pressure: a meta-analysis of randomized, controlled trials.

Ann Intern Med 136(7):493–503

Whelton PK, Carey RM, Aronow WS, Casey DE Jr, Collins KJ, Den-

nison Himmelfarb C, DePalma SM, Gidding S, Jamerson KA,

Jones DW, MacLaughlin EJ, Muntner P, Ovbiagele B, Smith SC

Jr, Spencer CC, Staﬀord RS, Taler SJ, Thomas RJ, Williams KA,

Sr., Williamson JD, Wright JT Jr. (2017) 2017 ACC/AHA/AAPA/

ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for

the Prevention, Detection, Evaluation, and Management of High

Blood Pressure in Adults: A Report of the American College of

Cardiology/American Heart Association Task Force on Clinical

Practice Guidelines. J Am Coll Cardiol

Wilmore JH, Stanforth PR, Gagnon J, Leon AS, Rao DC, Skinner JS,

Bouchard C (1996) Endurance exercise training has a minimal

eﬀect on resting heart rate: the HERITAGE Study. Med Sci Sports

Exerc 28(7):829–835