Content uploaded by Steve Richard Bird
Author content
All content in this area was uploaded by Steve Richard Bird on Sep 23, 2015
Content may be subject to copyright.
Effects of sleeping with reduced carbohydrate availability on acute training
responses
Stephen C. Lane,
1
Donny M. Camera,
2
David Gray Lassiter,
3
José L. Areta,
1
Stephen R. Bird,
1
Wee Kian Yeo,
4
Nikki A. Jeacocke,
5
Anna Krook,
6
Juleen R. Zierath,
3,6
Louise M. Burke,
5
and John A. Hawley
2,7
1
Exercise and Nutrition Research Group, School of Medical Sciences, RMIT University, Bundoora, Australia;
2
Centre for
Exercise and Nutrition, Mary MacKillop Health Research Institute, Australian Catholic University, Melbourne, Australia;
3
Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden;
4
National Sports Institute of Malaysia, Kuala
Lumpur, Malaysia;
5
Sports Nutrition, Australian Institute of Sport, Belconnen, Australia;
6
Physiology and Pharmacology,
Karolinska Institutet, Stockholm, Sweden; and
7
Institute for Sport and Exercise Sciences, Liverpool John Moores University,
Liverpool, United Kingdom
Submitted 24 September 2014; accepted in final form 22 June 2015
Lane SC, Camera DM, Lassiter DG, Areta JL, Bird SR, Yeo
WK, Jeacocke NA, Krook A, Zierath JR, Burke LM, Hawley JA.
Effects of sleeping with reduced carbohydrate availability on acute
training responses. J Appl Physiol 119: 643– 655, 2015. First pub-
lished June 25, 2015; doi:10.1152/japplphysiol.00857.2014.—We de-
termined the effects of “periodized nutrition” on skeletal muscle and whole
body responses to a bout of prolonged exercise the following morning. Seven
cyclists completed two trials receiving isoenergetic diets differing in the
timing of ingestion: they consumed either 8 g/kg body mass (BM) of
carbohydrate (CHO) before undertaking an evening session of high-intensity
training (HIT) and slept without eating (FASTED), or consumed 4 g/kg BM
of CHO before HIT, then 4 g/kg BM of CHO before sleeping (FED). The
next morning subjects completed2hofcycling (120SS) while overnight
fasted. Muscle biopsies were taken on day 1 (D1) before and 2 h after HIT
and on day 2 (D2) pre-, post-, and 4 h after 120SS. Muscle [glycogen] was
higher in FED at all times post-HIT (P⬍0.001). The cycling bouts increased
PGC1␣mRNA and PDK4 mRNA (P⬍0.01) in both trials, with PDK4
mRNA being elevated to a greater extent in FASTED (P⬍0.05). Resting
phosphorylation of AMPK
Thr172
, p38MAPK
Thr180/Tyr182
, and p-ACC
Ser79
(D2) was greater in FASTED (P⬍0.05). Fat oxidation during 120SS was
higher in FASTED (P⫽0.01), coinciding with increases in ACC
Ser79
and
CPT1 as well as mRNA expression of CD36 and FABP3 (P⬍0.05).
Methylation on the gene promoter for COX4I1 and FABP3 increased 4 h
after 120SS in both trials, whereas methylation of the PPAR␦promoter
increased only in FASTED. We provide evidence for shifts in DNA meth-
ylation that correspond with inverse changes in transcription for metaboli-
cally adaptive genes, although delaying postexercise feeding failed to aug-
ment markers of mitochondrial biogenesis.
cycling; HIT; train low; sleep-low; muscle glycogen
COMMENCING ENDURANCE EXERCISE with low muscle glycogen
stores (so-called “train low”) results in a greater transcriptional
activation of enzymes involved in carbohydrate metabolism,
including the adenosine 5=-monophosphate-activated protein
kinase (AMPK), GLUT4, and the pyruvate dehydrogenase
(PDH) complex, compared with when glycogen content is
normal (6, 25, 33). Restricting carbohydrate (CHO) availability
during early (1-5 h) postexercise recovery has also been shown
to acutely upregulate various markers of substrate metabolism
and endurance training adaptation in skeletal muscle (6, 27).
Because the time course of transcriptional activation for many
exercise-induced genes occurs during the first few hours of
recovery (18), returning to basal values within 24 h (34), such
events may be linked by common signaling and/or regulatory
mechanisms, such as the restoration of muscle energy stores,
predominantly glycogen. These early adaptive responses to
acute exercise may be orchestrated through epigenetic modifi-
cations involving DNA methylation (2). Exercise-induced
changes in DNA methylation are inversely associated with
activation of some but not all genes, underpinning the adaptive
response to exercise (2, 21), and appear to be dependent on
work intensity (2) and substrate availability (1).
The original train low protocol advocated twice-a-day train-
ing sessions in which only the second exercise session was
undertaken with low glycogen availability (8). A direct conse-
quence of this strategy was that the maximal self-selected
training intensity of the second session was substantially re-
duced when it was commenced with low compared with
normal glycogen levels (14, 36). Such an outcome is counter-
intuitive for the preparation of competitive athletes for whom
high-intensity workouts are a critical component of a peri-
odized training program (12). Against this background, we
have formulated a novel approach in which we can prolong the
duration of low CHO (i.e., muscle and liver glycogen) avail-
ability, thereby potentially enhancing and extending the time
course of transcriptional activation of metabolic genes and
their target proteins, while simultaneously conserving the train-
ing intensity of the initial session and hence the “training
impulse” to the working muscles. We have termed this strategy
“train-high, sleep-low” and here for the first time have simul-
taneously measured gene, protein, and methylation responses
in skeletal muscle in response to this protocol. We hypothe-
sized that compared with the effects of rapid postexercise CHO
intake (i.e., current sport nutrition guidelines), delaying CHO
intake and thereby extending the period during which an
individual remains in a low glycogen state would enhance the
acute responses of selected genes and proteins with putative
roles in training adaptation.
METHODS
Subjects. Seven male competitive endurance-trained cyclists with a
history of ⬎3 yr endurance training and who were riding (values
means ⫾SD) an average of 406 ⫾59 km/wk (range 285– 455 km/wk)
in the 6 wk prior to commencement of the study, volunteered to
Address for reprint requests and other correspondence: J. A. Hawley, Centre
for Exercise and Nutrition, Mary MacKillop Health Research Institute, Aus-
tralian Catholic Univ., Melbourne, VIC 3065, Australia (e-mail: john.
hawley@acu.edu.au).
J Appl Physiol 119: 643–655, 2015.
First published June 25, 2015; doi:10.1152/japplphysiol.00857.2014.
8750-7587/15 Copyright ©2015 the American Physiological Societyhttp://www.jappl.org 643
participate in these trials. The subjects’ age, body mass (BM), peak
oxygen uptake (V
˙O
2peak
), and peak power output (PPO) were 29 ⫾5
yr, 76.9 ⫾9.1 kg
,
67 ⫾4.0 ml·kg
⫺1
·min
⫺1
, and 422 ⫾39 W. Prior
to giving their written consent, all subjects were informed of the
possible risks of all procedures. The study was approved by the RMIT
Human Research Ethics Committee.
Study overview. Each subject completed two experimental trials in
a randomized crossover design. In each trial they performed two
exercise bouts: the first bout, high-intensity training (HIT), was
undertaken in the evening of the first day, and the second bout,
120-min steady-state ride (120SS), on the morning of the second day.
In one trial, subjects consumed their total daily energy intake through-
out the day (i.e., 0600 –1800 h) before undertaking the HIT session in
the evening (1900 –2000 h). Following this session they consumed no
food and remained fasted overnight (FASTED). In the other trial,
subjects ate half of their energy intake prior to the evening HIT
session, consuming the remainder immediately after HIT (FED). In
both trials subjects slept in the laboratory overnight and then com-
pleted 120SS commencing at 0700 h the following morning. One hour
after the completion of 120SS, all subjects consumed a standardized
breakfast and remained in the laboratory until the completion of the
trial at ⬃1300 h. Skeletal muscle biopsies were obtained before and 2
h post-HIT, day 1 (D1), and at rest, immediately post-120SS, and after
4 h recovery, day 2 (D2). These were analyzed for selected markers of
training adaptation.
Pretesting: incremental cycle test. Approximately 2 wk prior to
commencing their first experimental trial, subjects underwent an
incremental cycling test to exhaustion on an electronically braked
cycle ergometer (Lode Excallibur Sport, Groningen, The Netherlands)
as previously described (13). During this maximal test, subjects
breathed through a Hans Rudolph two-way nonrebreathing valve and
mouth piece attached to a calibrated online gas system (TrueOne
2400, Parvomedics, Utah) interfaced to a computer that calculated the
instantaneous rates of O
2
consumption (V
˙O
2
), CO
2
production (V
˙CO
2
),
min ventilation (VE
STPD
), and the respiratory exchange ratio (RER).
Before each test, analyzers were calibrated with commercially avail-
able gasses of known O
2
and CO
2
content. V
˙O
2peak
was defined as the
highest uptake a subject attained during any 30 s of the test, while
PPO was calculated from the last completed work rate plus the
fraction of time spent in the final noncompleted work rate (13). The
maximal test and all experimental trials were conducted under stan-
dardized laboratory conditions (18 –22°C, 40 –50% relative humidity),
and subjects were fan cooled during all exercise sessions. Each
individual’s PPO recorded during the incremental test was used to
determine their prescribed cycling intensities (W) during the subse-
quent experimental trials.
Standardized diet/exercise control. Subjects consumed a prepack-
aged standardized diet for the 24-h period prior to commencing an
experimental trial (15). Dietary goals for this period were 8 g/kg BM
CHO, 1.5 g/kg BM protein, and 1.5 g/kg BM fat for a total energy
intake of ⬃220 kJ/kg BM for the 24-h period. Subjects were in-
structed to avoid any strenuous physical activity as well as alcohol and
caffeine consumption for the 24 h prior to a trial. Subjects were
provided with all foods and drinks in portion-controlled packages for
consumption during the dietary control period and were given verbal
and written instructions on how to follow the diet. Checklists were
used to record each menu item as it was consumed and to note any
deviations from the menu. Each subject’s food checklists were re-
viewed and clarified for compliance to the standardization protocols
by the primary researcher.
Experimental diet. For the experimental trials, subjects were pro-
vided with all food and fluid to be consumed prior to reporting to the
laboratory at 1700 h. Subjects received one of two isoenergetic diets
(containing 8 g/kg BM CHO, 1.5 g/kg BM protein, 1.5 g/kg BM fat,
and ⬃220 kJ/kg BM energy) that only differed in the timing of
consumption (Fig. 1). During one trial (FASTED), food was portioned
such that subjects consumed 6 g/kg BM CHO, 1.25 g/kg BM protein,
and 1.25 g/kg BM of fat throughout the day with their final days’ meal
(2 g/kg BM CHO, 0.25 g/kg BM protein, and 0.25 g/kg BM of fat)
being consumed upon arrival at the laboratory (1700 h). During the
other trial (FED), food was portioned so that subjects consumed 2
g/kg BM CHO, 0.65 g/kg BM protein, and 0.35 g/kg BM of fat before
1700 h, a further 2 g/kg BM CHO, 0.25 g/kg BM protein, and 0.25
g/kg BM of fat meal upon arrival in the lab (1700 h), with the
remainder of that day’s intake (4 g/kg BM CHO, 0.6 g/kg BM protein,
and 0.9 g/kg BM of fat) consumed after the HIT session at 2000 h. On
day 2, a breakfast containing 2 g/kg BM CHO, 0.2 g/kg BM protein,
and 0.2 g/kg BM of fat was consumed 1 h after completion of the
120SS ride in both trials.
Blood and tissue collection and analysis. Seventeen blood samples
were collected during each trial with a total 9 ml of whole blood
obtained at each sampling time point (Fig. 1). Six milliliters of blood
were collected in tubes containing EDTA. Twenty-five microliters of
blood were then immediately analyzed for glucose concentration
(YSI, Yellow Springs, Ohio), while the remaining sample was cen-
trifuged at 4°C at 4,000 rev/min for 10 min with the resulting plasma
transferred to 1.5-ml tubes and stored at ⫺80°C for subsequent
analyses of plasma insulin and catecholamine concentrations. At each
Fig. 1. Experimental design. FED (a total of 4 g CHO/kg BM prior to HIT and 4 g CHO/kg BM post-HIT); FASTED (a total of 8 g CHO/kg BM prior to HIT
and remained fasted during sleep and throughout 120SS). CHO, carbohydrate; BM, body mass; HIT, high-intensity interval training; PPO, peak power output;
120SS, 120-min steady-state ride; SS, steady state.
644 Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
time point, a further 3 ml of blood was collected in a tube containing
EGTA, which was then centrifuged, and the resulting plasma frozen
and stored (as described above) for later analyses of free fatty acids
(FFA). Catecholamine concentrations were analyzed with a commer-
cially available enzyme immunoassay (Bi-CAT EIA 17-BCTHU-
E02.1, ALPCO, Salem, NH), while plasma insulin concentrations
were determined via ELISA (80-INSHU-E01.1, E10.1, ALPCO).
Plasma FFA concentrations were determined by an enzymatic color-
imetric method (NEFAC code 279 –75401, Wako, Tokyo, Japan).
A total of five biopsies were collected during each of the experi-
mental trials from the vastus lateralis with a 5-mm Bergström needle
adapted for manual suction. Samples were immediately washed in
0.9% saline solution then snap frozen in liquid nitrogen and stored at
⫺80°C until later analysis. The sampling points were before and 2 h
after HIT (day 1) and then at rest, immediately post-120SS, and after
4 h recovery (day 2). Biopsies were subsequently analyzed for gene
expression and protein abundance of markers of training adaptation
(described subsequently).
High-intensity interval session. On the evening of the first day of an
experimental trial, subjects completed a HIT session on a Velotron
cycle ergometer (Racermate, Seattle, WA). After a standardized warm
up (10 min at 60% PPO), subjects undertook HIT, which consisted of
eight 5-min work bouts at 82.5% of individual PPO with 1 min of
active recovery (100 W) between work bouts. This protocol was
chosen as the physiological demands have previously been character-
ized, and in well-trained cyclists, the session has been demonstrated to
reduce muscle glycogen by ⬃50% of starting values (31). During
HIT, ratings of perceived exertion (RPE) were recorded at the end of
each work bout, while heart rate was averaged for each 5-min
repetition.
One hundred twenty minute steady-state ride. On the morning of
the second day of each trial, subjects completed a standardized,
steady-state cycling bout (120SS) at a fixed submaximal work rate
that was the same for both trials. During this ride, subjects cycled at
50% PPO (⬃60% of V
˙O
2peak
). RER was recorded as a 5-min average
commencing at 10, 45, 80, and 115 min, while heart rate (HR) and
RPE were recorded at the end of each 5-min collection point. Whole-
body rates of CHO and fat oxidation (g/min) were calculated from the
respiratory data collected during the 120SS ride. The calculations
were made from V
˙CO
2
and V
˙O
2
measurements, assuming a nonprotein
RER (23).
Fluid intake. During the first experimental trial (including the 24-h
standardized dietary control), subjects were allowed water ad libitum.
The volume of fluid consumed was recorded and then replicated
during the subsequent trial.
Muscle glycogen concentration. Muscle glycogen concentration
was analyzed as previously described (4). In brief, ⬃10 –15 mg of
muscle was freeze-dried and powdered, with all visible blood and
connective tissue removed under magnification. The freeze-dried
muscle sample was then extracted and glycogen concentration deter-
mined via enzymatic analyses.
DNA methylation. Bisulfite conversion is a chemical treatment of
genomic DNA utilized to convert nonmethylated cytosines into ura-
cils; the bisulfite treatment, however, does not convert methylated
cytosines. Subsequent sequencing of bisulfite-converted DNA reveals
cytosine methylation at the resolution of a single nucleotide. We
utilized the bisulfite-sequencing method to interrogate DNA methyl-
ation in our targets of interest. This method is considered the gold
standard in DNA methylation analysis, largely because of the base
pair resolution afforded by the technique (5, 32). To assay DNA
methylation in known regulatory regions for PPARGC1A, PDK4,
TFAM, PPARD, SLC2A4, COX4I1, and FABP3, we used the Qiagen
Q24 Advanced Pyromark System (No. 9002270, Qiagen, Venlo, The
Netherlands). Genomic DNA was extracted from 2–23 mg (average
18 mg) of previously flash-frozen muscle biopsies with the Qiagen
DNeasy Blood and Tissue kit (No. 69506, Qiagen). Genomic DNA
was then bisulfite converted with the Qiagen Epitect Fast kit (No,
59826, Qiagen). After bisulfite treatment, the DNA was amplified by
PCR (primer sets are listed in Table 1). Finally, the DNA was
pyrosequenced and the percentage of methylated cytosines at variable
loci were determined.
For FABP3 and GLUT4, specific regions of interest were selected
based on previously identified regions subject to changes in DNA
methylation (22, 37). The UCSC Genome Browser permits users to
add “tracks” which show the genomic addresses where dynamic
methylation has been previously observed. In the absence of specific
literature to direct our search for methylation changes in the gene
promoters of PPAR␦and COX4I1, we designed our assays around
such sites observed on the UCSC Genome Browser (Feb. 2009
GRCh37/hg19 Assembly).
RNA extraction and quantification. Approximately 20 mg of skel-
etal muscle was homogenized in TRIzol and chloroform added to
form an aqueous RNA phase. This RNA phase was then precipitated
by mixing with isopropanol alcohol, and the resulting pellet was
washed and resuspended in 50 l of RNase-free water. Extracted
RNA was quantified with a QUANT-iT analyzer kit (No. Q32852,
Invitrogen, Melbourne, Australia) and on a NanoDrop 1000 spectro-
photometer (Nanodrop Technologies, Wilmington, DE) by measuring
absorbance at 260 and 280 nm with a 260/280 ratio of ⬃1.88 recorded
for all samples.
Reverse transcription and real-time PCR. First-strand complemen-
tary DNA (cDNA) synthesis was performed with commercially avail-
able TaqMan Reverse Transcription Reagents (Invitrogen) in a final
reaction volume of 20 l. All RNA and negative control samples were
reverse transcribed to cDNA in a single run from the same reverse
transcription master mix. Serial dilutions of a template RNA (No.
AM7982, AMBION) were included to ensure efficiency of reverse
transcription and for calculation of a standard curve for real-time
quantitative polymerase chain reaction (RT-PCR). Quantification (in
duplicate) was performed with a Rotor-Gene 3000 Centrifugal Real-
Time Cycler (Corbett Research, Mortlake, Australia). Taqman-FAM-
labeled primer/probes for PGC-1␣(No. Hs01016719), TFAM
Table 1. Target genes for DNA methylation
Gene
Target Forward Primer Reverse Primer Sequencing Primer
UCSC Genome
Browser Address (Feb.
2009 GRCh37/hg19
Assembly)
PPAR␦GGAGGATGTTTTTTATTTTAGGTGAA [Biotin]AATCTAAAAAAACTCTTAACCCAATACTA GGATGTTTTTTATTTTAGGTGAAT chr6: 35, 309, 819-35,
309, 951 (TSS
⫺516 to ⫺384)
C0X4I1 GTGTTAGGATTATAGGGGTTAGT [Biotin]CCTAACTCCCCTTAATTAAATATACCT GGATTATAGGGGTTAGTT chr16: 85, 832, 015-
85, 832, 119 (TSS
⫺1158 to ⫺1054)
FABP3 TGGGTATTTGGAAGTTAGTGG [Biotin]CCCTATTCCCCAATCTTAACC ATTTGGAAGTTAGTGGATA chr1: 31, 845, 700-31,
845, 821 (TSS
⫹102 to ⫹223)
645Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
(No. Hs00273372_s1), COX4I1 (No. Hs00971639_m1), PPAR␦
(No. Hs04187066_g1), CD36 (No. Hs01567185_m1), FABP3 (No.
Hs00997360_m1), PDK4 (No. Hs01037712_m1), and GLUT-4 (No.
Hs00168966_m1) were used in a final reaction volume of 20 l. PCR
treatments were 2 min at 50°C for UNG activation, 10 min at 95°C,
then 40 cycles at 95°C for 15 s and 60°C for 60 s. Glyceraldehyde-
3-phosphate dehydrogenase (GAPDH) (No. Hs Hs99999905) was
used as a housekeeping gene, and expression was not different at any
time point or between treatments (data not shown). The relative
amounts of mRNAs were calculated by the relative quantification
(⌬⌬CT) method (17).
Western blot analysis. By using a motorized pellet pestle (Sigma-
Aldrich, St. Louis, MO) with 5-s pulses, muscle samples (⬃15 mg)
were homogenized in ice-cold buffer containing 50 mM of Tris-HCl,
pH 7.5, 1 mM of EDTA, 1 mM of EGTA, 10% glycerol, 1% Triton
X-100, 50 mM of NaF, 5 mM of sodium pyrophosphate, 1 mM of
DTT, 10 g/ml of trypsin inhibitor, 2 g/ml of aprotinin, 1 mM of
benzamidine, and 1 mM PMSF. The lysate was kept on ice at all times
and was then centrifuged at 12,000 gfor 20 min at 4°C. The
supernatant was transferred to a sterile tube and was subsequently
aliquoted for determination of protein concentration with a BCA
protein assay (Pierce, Rockford, IL). The supernatant was then resus-
pended in Laemmli sample buffer and separated by sodium dodecyl
sulfate-polyacrylamide gel electrophoresis before being transferred to
polyvinylidine fluoride membranes and incubated with primary anti-
body (1:1,000) overnight at 4°C and secondary antibody (1:2,000).
Proteins were detected via chemiluminescence (Amersham Biosci-
ences, Buckinghamshire, UK; Pierce Biotechnology) and quantified
by densitometry (Chemidoc, BioRad, Gladesville, Australia). All
sample (50 g) time points for each subject were run on the same gel.
Polyclonal antiphospho AMPK␣
Thr172
(No. 2531), -ACC
Ser79
(No.
3661P), -HSL
Ser660
(No. 4126S), monoclonal anti-phospho
p38MAPK
Thr180/Tyr182
(No. 4511) and monoclonal total adipose tri-
glyceride lipase (ATGL) (No. 2439S) and CPT1A (No. 12252S) were
purchased from Cell Signaling Technology (Danvers, MA). Data are
expressed relative to ␣-tubulin (No. 3873, Cell Signaling Technology)
in arbitrary units.
Statistical analysis. Respiratory, physiological, blood, muscle gly-
cogen, and PCR data were analyzed with SPSS software package
(version 21). Western blot data were analyzed with Sigma Stat
(version 3.1). All data were checked for sphericity by Mauchly’s test
and normality by Kolomogorov-Smirnov tests. Where mRNA or
DNA methylation data violated assumptions of sphericity or normal-
ity, data were natural log transformed before further analyses. To
compare the responses during the experimental trials, data were
analyzed by two-way analyses of variance (trial by time) with
repeated measures (RM) (␣⫽0.05). Least significant difference
and paired t-tests were used post hoc. Results and statistics
represent seven subjects unless otherwise indicated. All values are
expressed as means ⫾SD unless otherwise indicated.
RESULTS
High-Intensity Intervals
Training responses. During HIT, two subjects were unable
to complete the prescribed session during their first trial (sub-
ject 4 at interval 4 and subject 8 at interval 5). As a conse-
quence, their exercise intensity was reduced by 10 W for
subsequent intervals to allow the subjects to complete the
remaining work bouts, and an identical set of work bouts was
repeated for their second trial. Average power output sustained
for HIT sessions was 346 ⫾31 W. There were no differences
in average RPE or HR between trials (Table 2).
120-min steady-state ride. During the standardized submaxi-
mal 120SS on the morning of day 2, RER was lower in
FASTED at 10 –15, 45–50, and 115–120 min time points
compared with FED (Table 3). During both trials there was a
main effect of time (P⬍0.01), whereby there was a steady
decline in RER throughout the ride. There were no differences
in HR between trials at any time point. HR increased through-
out 120SS, although there was no main effect of time for trials
(P⫽0.056). In the FED trial only, HR was statistically higher
Table 2. Physiological and respiratory response to HIT
Int 1 Int 2 Int 3 Int 4 Int 5 Int 6 Int 7 Int 8 Mean
RPE
FED 13.9 ⫾1.2 15.4 ⫾1.6 15.5 ⫾1.0 16.4 ⫾1.0 17.0 ⫾1.3 17.1 ⫾1.3 17.0 ⫾1.7 17.0 ⫾1.7 16.2 ⫾1.4
FASTED 13.6 ⫾1.3 14.8 ⫾0.8 15.4 ⫾0.9 16.4 ⫾2.3 15.6 ⫾1.9 16.8 ⫾1.8 17.2 ⫾1.6 17.4 ⫾1.1 15.9 ⫾1.5
Heart rate
FED 166 ⫾10.4 168 ⫾11.1 171 ⫾11.7 171 ⫾9.8 171 ⫾9.7 173 ⫾9.2 173 ⫾9.3 173 ⫾10.6 171 ⫾10.2
FASTED 168 ⫾11.7 171 ⫾12.8 173 ⫾12.7 173 ⫾11.7 173 ⫾11.2 175 ⫾10.7 175 ⫾10.8 175 ⫾12.2 173 ⫾11.7
Values are means ⫾SD, P⬍0.05. Eight 5-min intervals at 82.5% peak power output. HIT, high-intensity training; RPE, ratings of perceived exertion; FED,
subjects who consumed 4 g/kg of body mass of carbohydrate before HIT, then 4 g/kg body mass of carbohydrate before sleeping; FASTED, subjects who
consumed 8 g/kg of body mass of carbohydrate before undertaking an evening session of HIT and slept without eating.
Table 3. Physiological and respiratory response to 120SS
10–15 min 45–50 min 80–85 min 115–120 min Mean
RER
FED 0.88 ⫾0.03 0.86 ⫾0.03
f
0.84 ⫾0.03
fg
0.84 ⫾0.03
fg
0.85 ⫾0.03
FASTED 0.85 ⫾0.04* 0.82 ⫾0.04*
f
0.82 ⫾0.04
fg
0.80 ⫾0.03*
fgh
0.82 ⫾0.04*
RPE
FED 10.6 ⫾1.0 10.4 ⫾1.1 10.7 ⫾1.1 10.9 ⫾0.7 10.6 ⫾0.20
FASTED 10.2 ⫾0.8 10.4 ⫾1.3 10.6 ⫾1.1 10.8 ⫾0.8 10.5 ⫾0.24
Heart rate
FED 133 ⫾7.5 136 ⫾8.1 136 ⫾8.9 138 ⫾7.6
f
136 ⫾0.65
FASTED 135 ⫾7.9 136 ⫾7.3 135 ⫾6.8 137 ⫾6.4 136 ⫾0.62
Values are means ⫾SD, P⬍0.05. One hundred twenty minutes at 50% peak power output. 120SS, 120-min steady-state ride; RER, respiratory exchange
ratio. *Different to FED;
f
, different to 10 –15 min;
g
, different to 45–50 min;
h
, different to 80 –85 min.
646 Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
at 115–120 min compared with 10 –15 min (P⫽0.01). There
were no differences in RPE between trials at any point, with all
subjects being able to complete the training bout independent
of overnight dietary status. Similarly, there was no effect of
time on RPE during trials.
Muscle glycogen concentration. Subjects commenced the
HIT session with similar muscle glycogen concentrations. Two
hours post-HIT, muscle glycogen concentration was reduced
by ⬃45% in the FASTED trial and by ⬃30% in FED compared
with REST 1 (see Fig. 8). Resting muscle glycogen concen-
trations on the morning of day 2 (REST 2) remained lower than
REST 1 (P⬍0.05), and consequently the 120SS bout was
commenced with lower glycogen concentration in the
FASTED trial compared with FED (349 ⫾141 vs. 459 ⫾159
mmol/kg dry wt; P⬍0.01). The 120SS ride further reduced
glycogen concentration in both trials by ⬃25%, with a greater
absolute reduction in the FED (121 ⫾42 mmol/kg dry wt) vs.
FASTED (83 ⫾39 mmol/kg dry wt) trial. The meal consumed
1 h after the 120SS ride elevated glycogen concentration 3 h
postfeeding (P⬍0.001) to a similar extent in both trials (⬃25
mmol/kg dry wt). Additionally, muscle glycogen concentration
in the FASTED trial remained below that of the FED trial (P
⬍0.05).
Blood glucose concentration. At REST 1, immediately be-
fore the HIT exercise bout and 2 h after the standardized meal,
blood glucose concentrations were slightly higher in the
FASTED trial compared with the FED trial (Fig. 2A,P⬍
0.05), reflecting the greater overall CHO intake in the FASTED
trial to this point in the day. Immediately post-HIT, blood
glucose concentration did not differ between the trials. The
meal consumed immediately post-HIT in the FED trial in-
creased blood glucose concentration, with levels peaking at
HIT ⫹30 (Fig. 2A). Immediately prior to 120SS ride on the
morning of day 2 (REST 2), immediately after the ride, and
then throughout recovery, blood glucose concentration in both
trials displayed similar profiles: initially increasing as a con-
sequence of the meal that was consumed 1 h after 120SS and
then declining until the end of the trials. In both trials blood
glucose concentrations peaked 40 min after the meal (120SS ⫹
100) and declined to REST 1 concentration over the subse-
quent 3 h, with a small difference between trials 150 min
postmeal (120SS ⫹210 min), at which point the blood glucose
concentration in the FASTED trial was slightly higher than that
in the FED trial.
Insulin concentration. At REST 1, plasma insulin was higher
in FASTED compared with FED (Fig. 2B), reflecting the larger
daily CHO intake in FASTED prior to this point. The meal
consumed in the FED trial immediately post-HIT significantly
elevated plasma insulin values compared with the post-HIT
concentration, and although starting to decline, these were still
elevated at HIT ⫹120 min. Insulin concentrations in FASTED
trial remained at post-HIT concentrations at all time points
measured post-HIT. The following morning at REST 2 and
immediately after the 120SS bout, the insulin concentration
was similar between trials and near resting concentrations.
Following the post-120SS meal (eaten 60-min postexercise in
both trials), plasma insulin concentration peaked 40 min
postingestion (120SS ⫹100 min) in the FED trial and 60 min
postingestion (120SS ⫹120 min) in the FASTED trial. Plasma
insulin concentrations then declined in a similar manner in both
trials returning to the REST 1 concentrations within 3 h
postingestion.
Plasma FFA concentrations. There were no differences in
plasma FFA concentrations between trials at REST 1 or im-
mediately post-HIT (Fig. 2C). In FASTED, plasma FFA con-
centration continued to increase above REST 1 and compared
with FED became greater 90 and 120 min after HIT. Resting
plasma FFA concentration after the overnight sleep had de-
clined from the previously elevated concentration in the
FASTED trial, but was still above REST 1 concentration.
There were no differences in plasma FFA level between trials
at this time. The 120-min SS ride was associated with elevated
Fig. 2. Blood glucose concentration (A), plasma insulin concentration (B), and
plasma free fatty acids (FFA) concentration (C). Values are means ⫾SD; P⬍
0.05. a, Different to REST 1; c, different to REST 2; d, different to post-120SS;
i, different to REST 1 in FED only; j, different to REST 1 in FASTED only;
k, different to 120SS ⫹60 in FED only; l, different to 120SS ⫹60; m,
different to REST 2 in FED only; Ex, exercise; M1; meal in FED, 4 g/kg BM
CHO; M2, meal in both trials, 2 g/kg BM CHO. *Difference between trials.
647Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
plasma FFAs (P⬍0.01), with no difference observed between
trials (P⫽0.179). Following breakfast, plasma FFA concen-
tration continued to decline in both trials, returning to REST 2
concentration 40 min postingestion (120SS ⫹100).
Catecholamine concentrations. Plasma concentrations of
adrenaline and noradrenaline were similar between trials at REST
1 and at all other time points throughout the trials (Fig. 3). Levels
were elevated post-HIT (P⬍0.01), but had returned to REST 1
concentration within 60 min of completing HIT. After the 120SS
ride on day 2, plasma noradrenaline was elevated in both trials,
but to a lesser magnitude than following the HIT session (P⫽
0.02). However, the plasma noradrenaline concentration was
still slightly above REST 2 concentrations 60 min postexercise
(P⫽0.02). Plasma adrenaline concentration displayed a sim-
ilar profile to the noradrenaline response and immediately
post-120SS were above REST 2 concentration, but this only
reached statistical significance in the FASTED trial (P⫽
0.005) and had returned to REST 2 concentration within 60
min postexercise in both trials.
Substrate utilization during 120SS. Total CHO oxidation
during 120SS was greater in FED (223 ⫾42 g) than FASTED
(168 ⫾28 g) (P⫽0.01), whereas total fat oxidation was
greater in FASTED (111 ⫾25 g) compared with FED (88 ⫾
17 g) (P⫽0.01, Fig. 4).
Protein Data
Signaling proteins. Baseline values for p-AMPK
Thr172
,
p-p38MAPK
Thr180/Tyr182
, and p-ACC
Ser79
were similar at rest
on day 1 (see Fig. 10). At 2 h post-HIT, p-AMPK
Thr172
tended
to be higher in FASTED compared with the FED trial (P⫽
0.058). The following morning (REST 2), phosphorylation of
AMPK
Thr172
, p38MAPK
Thr180/Tyr182
, and ACC
Ser79
was
greater in the FASTED compared with the FED trial, while
p-ACC
Ser79
was also elevated compared with REST 1 (P⬍
0.05). The 120SS bout did not increase (post-120SS) p-AMP-
K
Thr172
or p-p38MAPK
Thr180/Tyr182
above REST 2. However,
p-ACC
Ser79
further increased in FASTED at post-120SS (P⬍
0.05), but had returned to the resting value by 120SS ⫹4h.A
similar pattern was shown by p-AMPK
Thr172
.
Lipolysis and fat transport proteins. There were no differ-
ences between the two trials for CPT1, ATGL, or p-HSL
Ser 660
at rest on day 1 (Fig. 5). At HIT ⫹2 h, ATGL in the FASTED
trial was elevated compared with REST 1 (P⬍0.05). At
post-120SS, CPT1 was higher in FASTED compared with
FED (P⬍0.01) and was elevated compared with both time
points from the previous day (P⬍0.01). CPT1 remained
elevated above these levels at 120SS ⫹4 h in FASTED (P⬍
0.01). ATGL was higher at 120SS ⫹4 h in FASTED compared
with all prior time points (P⬍0.05), but despite being
substantially elevated, protein abundance was not significantly
different to FED (P⫽0.06), possibly due to a lack of statistical
power (n⫽4).
Mitochondrial genes. PGC1␣, TFAM, and COX4I1 mRNA
were not different between trials at REST 1 (Fig. 6). After HIT,
PGC1␣mRNA was increased in both trials (P⬍0.05; ⬃6-fold
change). TFAM mRNA was increased only in FED (P⬍0.01).
At REST 2, PGC1␣mRNA had declined in both trials, but
Fig. 3. Plasma Catecholamine concentrations: noradrenaline concentration (A)
and adrenaline concentration (B). Values are means ⫾SD; P⬍0.05. a,
Different to REST 1; c, different to REST 2; d, different post-120SS; o,
adrenaline different to REST 2 in FASTED only.
Fig. 4. Total carbohydrate and fat oxidation during the 120SS bout. Total
carbohydrate oxidation (A) and total fat oxidation (B). Values are means ⫾SD;
P⬍0.05. *Difference between trials.
648 Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
remained slightly elevated in FED compared with REST 1
(P⬍0.05), while TFAM was still elevated in FED and
increased in FASTED compared with REST 1. Post-120SS,
PGC1␣, and TFAM remained elevated compared with REST
1. Further increases in response to 120SS were only evident for
PGC1␣in the FED trial, which also further increased at
120SS ⫹4 h, reaching statistically higher levels compared
with FASTED (P⬍0.05). There were no differences between
trials or within trials between time points for COX4I1 mRNA
abundance. COX4I1 promoter methylation was higher in
FASTED at 120SS ⫹4 h compared with REST 1 and post-
120SS.
Lipolysis and fat transport genes. There were no differences
between trials at REST 1 for mRNA expression of PPAR␦,
FABP3, or CD36 (Fig. 7). At HIT ⫹2 h, PPAR␦mRNA was
elevated in both trials (FED ⬃2-fold, P⫽0.002 and FASTED
⬃1.5-fold, P⫽0.038), but values for both trials had returned
to pre-HIT levels by REST 2. FABP3 and CD36 mRNA
increased overnight in FASTED (P⬍0.05) with CD36 being
higher compared with FED (P⫽0.02). After 120SS, CD36
mRNA remained elevated in FASTED with FABP also becom-
ing significantly higher in FASTED compared with FED. DNA
methylation of FABP3 at 120SS ⫹4 h was greater than at
REST 2, which coincided with a trend for reduced mRNA
abundance. At 120SS ⫹4 h, PPAR␦mRNA tended to increase
in both trials, with levels only in the FED state reaching
significantly higher than previous values. Also at 120SS ⫹4h,
DNA methylation of the PPAR␦promoter was greater in
FASTED than FED (P⬍0.05).
CHO oxidation genes. There were no differences between
trials at REST 1 for mRNA of either PDK4 or GLUT4. At HIT ⫹
2, PDK4 mRNA was elevated in both trials (P⬍0.01), with
mRNA being elevated to a greater extent in FASTED (⬃65-
fold) compared with FED (⬃10-fold difference) (P⫽0.03).
Overnight, PDK4 mRNA remained elevated in FASTED com-
pared with FED (P⫽0.02) and REST 1 (P⬍0.01). The
120SS bout evoked further increases in PDK4 mRNA in both
trials, which persisted until the final time point 4 h post-120SS
(P⬍0.05). Additionally, the PDK4 mRNA was significantly
higher in the FASTED trial (P⬍0.05), compared with the
FED trial at both time points (post-120SS and 120SS ⫹4 h).
At HIT ⫹2 h, GLUT4 mRNA did not differ from REST 1,
but the following morning it was elevated in both trials com-
pared with HIT ⫹2(P⬍0.05). Post-120SS GLUT4 mRNA
was elevated above REST 1 and HIT ⫹2 in both trials (P⬍
0.05).
DISCUSSION
Skeletal muscle adaptation to exercise training is a conse-
quence of repeated contraction-induced increases in gene ex-
pression that lead to the accumulation of functional proteins
whose role is to blunt the homeostatic perturbations generated
by contraction-induced increases in energy demand and sub-
strate turnover. The development of a specific “exercise phe-
notype” is the result of new, augmented steady-state mRNA
and protein levels that stem from the training stimulus (24),
which can be modified by the prevailing energy availability
(11). The four fundamental cellular processes involved in gene
expression are transcription, mRNA degradation, translation,
and protein degradation, with each step of this cascade con-
Fig. 5. Lipolysis and fat transport proteins. Carnitine palmitoyltransferase I
(CPT1) (A), adipose triglyceride lipase (ATGL) (B), Phosphorylation of
Hormone sensitive lipase (p-HSL
Ser 660
)(C). Values are normalized to REST
1 and are expressed as means ⫾SD; P⬍0.05. At all time points CPT1 n⫽
6, ATGL n⫽4, HSL n⫽5. a, Different to REST 1; b, different to HIT ⫹2
h; c, different to REST 2; d, different to post-120SS. *Difference between
trials.
649Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
trolled by gene regulatory events (30). Recent evidence also
suggests that acute gene activation is associated with a dy-
namic change in DNA methylation in skeletal muscle and that
DNA hypomethylation is an early event in contraction-induced
gene activation (2). Here for the first time we have simultane-
ously measured gene, protein, and methylation status in skel-
etal muscle in response to a novel exercise-nutrient interven-
tion in which athletes undertook an intense bout of endurance
training late in the day and then slept with reduced CHO
availability. Specifically, we periodized the timing of nutrient
intake such that cyclists performed an evening bout of HIT
with high CHO availability, then restricted CHO intake so that
they slept with low CHO availability before undertaking a
standardized bout of submaximal exercise in the fasted state
the following morning. We found that when feeding was
withheld overnight and subjects slept with reduced energy
availability, AMPK
Thr172
, p38MAPK
Thr180/Tyr182
, and p-
ACC
Ser79
were upregulated to a greater extent the following
morning, compared with when subjects were fed a high-CHO
meal early in recovery. We also showed that when a second,
prolonged steady-state training session was commenced after
“sleeping low,” the expression of selected genes and abun-
dance of phosphorylated signaling proteins with putative roles
in lipid oxidation and transport were higher compared with
when a postexercise meal was consumed and glycogen avail-
ability was partially restored.
A major aim of this study was to circumvent the previously
observed impairment in maximal self-selected training inten-
sity when athletes perform two bouts of training within several
hours; the second session commenced with reduced muscle
glycogen content (14, 36). Under such conditions, power
output is reduced by ⬃8% (14, 36), even when caffeine is
ingested in an attempt to offset this decline (16). By undertak-
ing HIT in the evening and then withholding feeding overnight,
athletes were able to complete HIT and still train low the
following morning without compromising the total training
impulse to the working muscles. In the present study, HIT
reduced glycogen content by ⬃50%, which is consistent with
previous investigations using the same protocol and athletes of
comparable training status (31, 35). However, the glycogen
content of the well-trained cyclists in the current study was
higher than our previous work (⬃600 mmol/kg dry wt; range
⬃400 to ⬃900 mmol/kg dry wt), and its relative utilization
during HIT was similar to values we have previously reported
(⬃50%) (31, 35), hence it resulted in a substantial amount of
glycogen remaining in the muscle after the HIT session (⬃360
Fig. 6. Mitochondrial genes: mRNA expression of peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1␣)(A), mRNA expression of
mitochondrial transcription factor A (TFAM) (B), mRNA expression and DNA methylation of cytochrome c oxidase subunit IV (COX IV) (Cand D). For COX
4I1 (chr 16: 85,832019 – 20) at 120SS ⫹4h,n⫽6 and DNA methylation all time points n⫽4. Values are normalized to REST 1 and are expressed as
means ⫾SD; P⬍0.05. a, Different to REST 1; b, different to HIT ⫹2 h; c, different to REST 2; d, different to post-120SS. *Difference between trials.
650 Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
mmol/kg dry wt; Fig. 8). As such, athletes slept with reduced
but not low muscle glycogen levels and commenced the next
morning’s training session with higher than anticipated glyco-
gen availability. Of note is that the glycogen content attained in
the current study after the HIT session is higher than concen-
trations reported by others who subsequently observed signif-
icant upregulation of several training-induced signaling re-
sponses (3, 20, 25, 29, 35). Notwithstanding such differences
in exercise-induced glycogen utilization among studies, we
observed significant increases in PGC1␣mRNA expression
Fig. 7. Lipolysis and fat transport genes: mRNA expression and DNA methylation of peroxisome proliferator-activated receptor delta [PPAR␦(chr 6:
35,309,822 – 929)] (Aand B), mRNA expression and DNA methylation of fatty acid binding protein [FABP3 (chr 1: 31,845,738 – 9)] (Cand D), and
mRNA expression of cluster of differentiation 36 (CD36) (E). For mRNA 120SS ⫹4hn⫽6 and DNA methylation all time points n⫽4. Values are
normalized to REST 1 and are expressed as means ⫾SD; P⬍0.05. a, Different to REST 1; b, different to HIT ⫹2 h; c, different to REST 2; d, different
to post-120SS. *Difference between trials.
651Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
several hours after HIT (Fig. 6), consistent with a variety of
glycogen-depleting protocols (6, 7, 19, 28). PDK4 mRNA
expression was elevated compared with rest at 2 h post-HIT in
both trials (Fig. 9). However, the consumption of a high-CHO
meal immediately after HIT blunted the rise in PDK4 mRNA
such that levels in the FASTED trial were ⬃6-fold greater than
in the FED trial, with differences between trials persisting at all
subsequent time points.
The effects of sleeping with reduced muscle glycogen con-
tent can be assessed by examining markers of training adapta-
tion/substrate availability in the resting tissue samples obtained
on the morning of day 2 of the experiment. As might be
expected after withholding energy intake overnight, the abun-
dance or phosphorylated AMPK
Thr172
, p38MAPK
Thr180/Tyr182
,
and p-ACC
Ser79
protein was elevated to a greater extent in the
FASTED vs. FED trial (Fig. 10). Nevertheless, protein abun-
dance and mRNA expression of several downstream targets,
including TFAM and COX4I1 (Fig. 6) and PPAR␦(Fig. 7) did
not follow the same temporal pattern. For example, compared
with 2 h post-HIT the prior evening, PGC1␣mRNA expres-
sion declined by the morning of day 2 (some 10 h later),
irrespective of whether or not a high-CHO meal was con-
sumed. The second bout of prolonged, submaximal exercise
failed to elicit further increases in the abundance of phosphor-
ylated AMPK
Thr172
or p38MAPK
Thr180/Tyr182
protein, irrespec-
tive of whether training commenced after sleeping with low
muscle glycogen or following an evening meal that resulted in the
replenishment of some muscle glycogen (Fig. 10). However, as
might be expected from the greater oxidation of fat-based com-
pared with CHO-based fuels during this second exercise bout
(Fig. 4), sleeping with low muscle glycogen did induce a greater
increase in ACC
Ser79
phosphorylation. Fasting may promote
COX4I1 gene expression since a transient decrease in mRNA was
observed after the steady-state exercise only when participants
were fed postexercise. Furthermore, methylation of the
COX4I1promoter was increased after4hofrecovery from
steady-state exercise, which supports the notion that exercise
induces transient changes in DNA methylation (2).
In contrast to COX4I1, the responses of some other genes
with roles in mitochondrial biogenesis (PGC1␣and TFAM)
were not substantially altered by either dietary condition in
response to the second exercise bout (Fig. 6). We observed an
exercise and diet-induced elevation in PGC1␣mRNA expres-
sion in the FED, but not FASTED trial several hours after the
completion of exercise, and at a time when carbohydrate
availability was high for both conditions. This response is
difficult to explain, but suggests that withholding carbohydrate
intake immediately postexercise may influence the adaptive
responses to a subsequent training session undertaken in close
proximity (i.e., within a 12 h window or even the same day).
Conversely, when muscle glycogen was depleted by prior
exercise and a subsequent exercise bout is commenced 14 h
later with low (⬃170 mmol/kg dry wt) muscle glycogen
availability, PGC1␣mRNA expression was elevated to a
greater extent than when carbohydrate was consumed during
the recovery period (29). Differences in results between the
current and earlier (29) study are hard to reconcile. PPAR␦
mRNA expression increased in the hours after exercise, but
only to a significant extent during the FED trial. Correspond-
ingly, greater methylation of PPAR␦was observed in the
FASTED trial 4 h after the steady-state exercise, which may
underlie the comparatively reduced mRNA expression. Since
Fig. 8. Skeletal muscle glycogen concentration. Values are means ⫾SD; P⬍
0.05. a, Different to REST 1; b, different to HIT ⫹2 h; c, different to REST
2; d, different to POST 120SS. *Difference between trials.
Fig. 9. CHO oxidation genes. Pyruvate dehydrogenase kinase 4 (PDK4) (A)
and glucose transporter 4 (GLUT4) (B). Values are normalized to REST 1 and
are expressed as means ⫾SD; P⬍0.05. a, Different to REST 1; b, different
to HIT ⫹2 h; c, different to REST 2; d, different to post-120SS. *Difference
between trials.
652 Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
high- but not low-intensity exercise led to hypomethylation of
the PPAR␦gene and increased gene transcription (2), exercis-
ing in a fasted state may preclude the necessary glycolytic flux
to induce these adaptive responses at the genomic level.
While sleeping with reduced muscle CHO availability failed
to augment selected markers of training adaptation, lack of
nutrient availability resulted in marked increases in mRNA
and/or abundance of proteins involved in lipid utilization.
Compared with FED, FASTED resulted in a greater elevation
in levels of FABP3 and CD36 mRNA content the following
morning, with levels being further elevated and becoming
significantly greater than the FED trial at post-120SS. DNA
methylation of the FABP3 gene tended to be inversely related
to mRNA expression, implicating a role for transcriptional
repression by this epigenetic marker. Taken together, exercise
may induce an increase in methylation of the FABP3 gene in
the first few hours after exercise, resulting in impaired mRNA
transcription, which may be partially rescued by fasting. Im-
mediately and 4 h after the second bout of exercise undertaken
on day 2, PDK4 mRNA abundance was elevated to a greater
extent in the FASTED compared with the FED trial. Given the
differences in the contribution of fat and carbohydrate fuels to
this exercise bout (Fig. 4), this is unsurprising and highlights
the sensitivity of PDK4 to substrate availability and its role in
downregulating CHO oxidation (26). One of the main re-
sponses to withholding CHO availability postexercise, as well
as commencing subsequent exercise with reduced muscle gly-
cogen, is a marked elevation in fat transport and oxidation. Our
findings corroborate previous studies in which postexercise
feeding was withheld (27), as well as commencing exercise
with reduced muscle glycogen availability (3, 25). These ap-
proaches result in greater rates of fat oxidation measured using
tracer derived techniques (14), as well as increasing markers of
training adaptation when incorporated into a periodized train-
ing program (14, 36).
By measuring a time course of gene, protein, and methyl-
ation events in skeletal muscle, we wished to gain insight into
the temporal relationship between these cellular markers in
response to our novel diet-exercise intervention. Disappoint-
ingly, we observed only small increases in mRNA for most of
the genes under investigation that, in many cases, were discor-
dant from both protein and methylation responses. As protein
synthesis is bioenergetically costly, any gene-protein responses
are likely to be exaggerated under conditions of energy con-
straint (i.e., low CHO availability). However, given that only
⬃40% of the variance in protein levels is likely to be explained
by changes in mRNA levels (30), our results suggest that most
of the mRNAs and proteins we measured are relatively stable
and that the cellular perturbations induced by our diet-exercise
intervention did not require rapid transcriptional/translational
regulation. While we provide evidence for shifts in DNA
methylation that correspond with inverse changes in transcrip-
tion for metabolically adaptive genes, the minimal changes in
mRNA we observed would certainly not be expected to trigger
new steady-state protein levels and alter muscle phenotype
over the longer term.
While our sleeping with reduced muscle glycogen protocol
was specifically designed to prolong the time during which
subjects were exposed to low CHO availability, we also ex-
pected to observe concomitant increases in systemic factors
(i.e., elevated circulating FFA and catecholamine concentra-
Fig. 10. Signaling proteins. Phosphorylation of 5=adenosine monophosphate-
activated protein kinase (p-AMPK
Thr172
)(A), phosphorylation of p38 mitogen-
activated protein kinase (p-p38MAPK
Thr180/Tyr182
)(B), and phosphorylation
Acetyl-CoA carboxylase (p-ACC
Ser79
)(C). For all proteins at post-120SS n⫽
6; values are normalized to REST 1 and are expressed as means ⫾SD; P⬍
0.05. a, Different to REST 1; b, different to HIT ⫹2 h; c, different to REST
2; e, different to 120SS ⫹4 h. *Difference between trials.
653Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
tions) with putative roles in the adaptive processes (9, 10).
However, despite a substantially greater contribution from
lipid-based fuels to total energy expenditure during the pro-
longed steady-state ride on the morning of day 2, circulating
FFA and catecholamine concentrations were similar under both
conditions (Figs. 2 and 3). These results do not concur with
other studies in which elevated adrenergic responses have been
reported when exercise is commenced with reduced muscle
glycogen (8). The discrepancy between these studies may be
due to a combination of the higher starting muscle glycogen
levels for the second exercise bout in the present study, as well
as differences in exercise mode (i.e., single-limb kicking vs.
cycle ergometry).
In conclusion, delayed feeding after an intense evening training
session, so that cyclists sleep with lowered CHO availability, results
in a greater upregulation of several exercise responsive signaling
markers with roles in lipid oxidation the following morning com-
pared with when an evening meal was consumed (i.e., high overnight
CHO availability). Commencing prolonged, steady-state exercise
after sleeping with reduced muscle glycogen promoted greater rates
of whole body fat oxidation, compared with sleeping with at least
partially replenished muscle glycogen, but failed to elicit a greater
upregulation of cellular markers of mitochondrial biogenesis. We also
provide evidence for shifts in DNA methylation, which correspond
with inverse changes in transcription for metabolically adaptive
genes. Whether our novel, delayed postexercise feeding and sleeping
with reduced muscle glycogen protocol when incorporated into a
periodized training program undertaken over several weeks could
provide an additional stimulus to enhance the normal adaptive re-
sponses to training remains to be determined. Our results suggest that
critical absolute “thresholds” for both preexercise glycogen concen-
tration and training intensity exist if specific nutrient-exercise inter-
actions are to augment the normal training response-adaptation. Fu-
ture studies using unbiased “-omic” methodology may uncover novel
insights into different exercise and diet regimens that can be imple-
mented to optimize metabolic adaptations to training.
ACKNOWLEDGMENTS
We acknowledge the hard work and commitment of the athletes who gave
their time to participate in this research project. We also thank everyone who
assisted in running the experimental trials. We thank Dr. Bengt Saltin for input
with the design of the study protocol.
GRANTS
This project was partly funded through a collaboration with the Sports
Nutrition Department of the Australian Institute of Sport, The National Sports
Institute of Malaysia, European Research Council Ideas Program (ICEBERG,
ERC-2008-AdG23285), and The Swedish Research Council.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the author(s).
AUTHOR CONTRIBUTIONS
S.C.L., J.L.A., L.M.B., and J.A.H. conception and design of research;
S.C.L., D.M.C., D.G.L., J.L.A., N.A.J., and A.K. performed experiments;
S.C.L., D.M.C., and S.R.B. analyzed data; S.C.L., D.M.C., and S.R.B. inter-
preted results of experiments; S.C.L., D.M.C., and D.G.L. prepared figures;
S.C.L., D.G.L., and W.K.Y. drafted manuscript; S.C.L., D.G.L., S.R.B.,
W.K.Y., A.K., J.R.Z., L.M.B., and J.A.H. edited and revised manuscript;
S.C.L., W.K.Y., J.R.Z., L.M.B., and J.A.H. approved final version of manu-
script.
REFERENCES
1. Barres R, Osler ME, Yan J, Rune A, Fritz T, Caidahl K, Krook A,
Zierath JR. Non-CpG methylation of the PGC-1alpha promoter through
DNMT3B controls mitochondrial density. Cell Metab 10: 189 –198, 2009.
2. Barres R, Yan J, Egan B, Treebak JT, Rasmussen M, Fritz T,
Caidahl K, Krook A, O’Gorman DJ, Zierath JR. Acute exercise
remodels promoter methylation in human skeletal muscle. Cell Metab 15: 405–
411, 2012.
3. Bartlett JD, Louhelainen J, Iqbal Z, Cochran AJ, Gibala MJ, Gregson
W, Close GL, Drust B, Morton JP. Reduced carbohydrate availability
enhances exercise-induced p53 signaling in human skeletal muscle: im-
plications for mitochondrial biogenesis. Am J Physiol Regul Integr Comp
Physiol 304: R450 –R458, 2013.
4. Churchley EG, Coffey VG, Pedersen DJ, Shield A, Carey KA, Cam-
eron-Smith D, Hawley JA. Influence of preexercise muscle glycogen
content on transcriptional activity of metabolic and myogenic genes in
well-trained humans. J Appl Physiol 102: 1604 –1611, 2007.
5. Clark SJ, Statham A, Stirzaker C, Molloy PL, Frommer M. DNA
methylation: bisulphite modification and analysis. Nat Protoc 1: 2353–
2364, 2006.
6. Cochran AJ, Little JP, Tarnopolsky MA, Gibala MJ. Carbohydrate
feeding during recovery alters the skeletal muscle metabolic response to
repeated sessions of high-intensity interval exercise in humans. J Appl
Physiol 2010.
7. Gibala MJ, McGee SL, Garnham AP, Howlett KF, Snow RJ, Har-
greaves M. Brief intense interval exercise activates AMPK and p38
MAPK signaling and increases the expression of PGC-1alpha in human
skeletal muscle. J Appl Physiol 106: 929 –934, 2009.
8. Hansen AK, Fischer CP, Plomgaard P, Andersen JL, Saltin B, Ped-
ersen BK. Skeletal muscle adaptation: training twice every second day vs.
training once daily. J Appl Physiol 98: 93–99, 2005.
9. Hawley JA. Fat adaptation science: low-carbohydrate, high-fat diets to
alter fuel utilization and promote training adaptation. Nestle Nutr Inst
Workshop Ser 69: 59 –71, discussion 71-57, 2011.
10. Hawley JA, Burke LM. Carbohydrate availability and training adapta-
tion: effects on cell metabolism. Exerc Sport Sci Rev 38: 152–160, 2010.
11. Hawley JA, Burke LM, Phillips SM, Spriet LL. Nutritional modulation
of training-induced skeletal muscle adaptations. J Appl Physiol 110:
834 –845, 2011.
12. Hawley JA, Morton JP. Ramping up the signal: Promoting endurance
training adaptation in skeletal muscle by nutritional manipulation. Clin
Exp Pharmacol Physiol 41: 608 –613, 2014.
13. Hawley JA, Noakes TD. Peak power output predicts maximal oxygen
uptake and performance time in trained cyclists. Eur J Appl Physiol Occup
Physiol 65: 79 –83, 1992.
14. Hulston CJ, Venables MC, Mann CH, Martin C, Philp A, Baar K, Jeuken-
drup AE. Training with low muscle glycogen enhances fat metabolism in
well-trained cyclists. Med Sci Sports Exerc 42: 2046 –2055, 2010.
15. Jeacocke NA, Burke LM. Methods to standardize dietary intake before
performance testing. Int J Sport Nutr Exerc Metab 20: 87–103, 2010.
16. Lane SC, Areta JL, Bird SR, Coffey VG, Burke LM, Desbrow B,
Karagounis LG, Hawley JA. Caffeine ingestion and cycling power
output in a low or normal muscle glycogen state. Med Sci Sports Exerc 45:
1577–1584, 2013.
17. Livak KJ, Schmittgen TD. Analysis of Relative Gene Expression Data
Using Real-Time Quantitative PCR and the 2⌬⌬CT Method. Methods 25:
402–408, 2001.
18. Mahoney DJ, Parise G, Melov S, Safdar A, Tarnopolsky MA. Analysis
of global mRNA expression in human skeletal muscle during recovery
from endurance exercise. FASEB J 19: 1498 –1500, 2005.
19. Mathai AS, Bonen A, Benton CR, Robinson DL, Graham TE. Rapid
exercise-induced changes in PGC-1alpha mRNA and protein in human
skeletal muscle. J Appl Physiol 105: 1098 –1105, 2008.
20. Morton JP, Croft L, Bartlett JD, Maclaren DP, Reilly T, Evans L,
McArdle A, Drust B. Reduced carbohydrate availability does not mod-
ulate training-induced heat shock protein adaptations but does upregulate
oxidative enzyme activity in human skeletal muscle. J Appl Physiol 106:
1513–1521, 2009.
21. Nitert MD, Dayeh T, Volkov P, Elgzyri T, Hall E, Nilsson E, Yang BT,
Lang S, Parikh H, Wessman Y, Weishaupt H, Attema J, Abels M,
Wierup N, Almgren P, Jansson PA, Ronn T, Hansson O, Eriksson KF,
Groop L, Ling C. Impact of an exercise intervention on DNA methylation
654 Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
in skeletal muscle from first-degree relatives of patients with type 2
diabetes. Diabetes 61: 3322–3332, 2012.
22. Oshel KM, Knight JB, Cao KT, Thai MV, Olson AL. Identification of
a 30-base pair regulatory element and novel DNA binding protein that
regulates the human GLUT4 promoter in transgenic mice. J Biol Chem
275: 23666 –23673, 2000.
23. Peronnet F, Massicotte D. Table of nonprotein respiratory quotient: an
update. Can J Sport Sci 16: 23–29, 1991.
24. Perry CG, Lally J, Holloway GP, Heigenhauser GJ, Bonen A, Spriet
LL. Repeated transient mRNA bursts precede increases in transcriptional
and mitochondrial proteins during training in human skeletal muscle. J
Physiol 588: 4795–4810, 2010.
25. Pilegaard H, Keller C, Steensberg A, Helge JW, Pedersen BK, Saltin
B, Neufer PD. Influence of pre-exercise muscle glycogen content on
exercise-induced transcriptional regulation of metabolic genes. J Physiol
541: 261–271, 2002.
26. Pilegaard H, Neufer PD. Transcriptional regulation of pyruvate dehy-
drogenase kinase 4 in skeletal muscle during and after exercise. Proc Nutr
Soc 63: 221–226, 2004.
27. Pilegaard H, Osada T, Andersen LT, Helge JW, Saltin B, Neufer PD.
Substrate availability and transcriptional regulation of metabolic genes in
human skeletal muscle during recovery from exercise. Metabolism 54:
1048 –1055, 2005.
28. Pilegaard H, Saltin B, Neufer PD. Exercise induces transient transcrip-
tional activation of the PGC-1alpha gene in human skeletal muscle. J
Physiol 546: 851–858, 2003.
29. Psilander N, Frank P, Flockhart M, Sahlin K. Exercise with low
glycogen increases PGC-1alpha gene expression in human skeletal mus-
cle. Eur J Appl Physiol 2012.
30. Schwanhausser B, Busse D, Li N, Dittmar G, Schuchhardt J, Wolf J,
Chen W, Selbach M. Global quantification of mammalian gene expres-
sion control. Nature 473: 337–342, 2011.
31. Stepto NK, Martin DT, Fallon KE, Hawley JA. Metabolic demands of
intense aerobic interval training in competitive cyclists. Med Sci Sports
Exerc 33: 303–310, 2001.
32. Suzuki MM, Bird A. DNA methylation landscapes: provocative insights
from epigenomics. Nature reviews Genetics 9: 465–476, 2008.
33. Wojtaszewski JF, MacDonald C, Nielsen JN, Hellsten Y, Hardie DG,
Kemp BE, Kiens B, Richter EA. Regulation of 5=AMP-activated protein
kinase activity and substrate utilization in exercising human skeletal
muscle. Am J Physiol Endocrinol Metab 284: E813–E822, 2003.
34. Yang Y, Creer A, Jemiolo B, Trappe S. Time course of myogenic and
metabolic gene expression in response to acute exercise in human skeletal
muscle. J Appl Physiol 98: 1745–1752, 2005.
35. Yeo WK, McGee SL, Carey AL, Paton CD, Garnham AP, Hargreaves
M, Hawley JA. Acute signalling responses to intense endurance training
commenced with low or normal muscle glycogen. Exp Physiol 95: 351–
358, 2010.
36. Yeo WK, Paton CD, Garnham AP, Burke LM, Carey AL, Hawley JA.
Skeletal muscle adaptation and performance responses to once a day
versus twice every second day endurance training regimens. J Appl
Physiol 105: 1462–1470, 2008.
37. Zhang Y, Kent JW, 2nd Lee A, Cerjak D, Ali O, Diasio R, Olivier M,
Blangero J, Carless MA, Kissebah AH. Fatty acid binding protein 3
(fabp3) is associated with insulin, lipids and cardiovascular phenotypes of
the metabolic syndrome through epigenetic modifications in a Northern
European family population. BMC Med Genomics 6: 9, 2013.
655Sleeping with Reduced Muscle Glycogen •Lane SC et al.
J Appl Physiol •doi:10.1152/japplphysiol.00857.2014 •www.jappl.org
