Transcriptomic and epigenetic responses to short-term nutrient-exercise stress in humans

Abstract

High fat feeding impairs skeletal muscle metabolic flexibility and induces insulin resistance, whereas exercise training exerts positive effects on substrate handling and improves insulin sensitivity. To identify the genomic mechanisms by which exercise ameliorates some of the deleterious effects of high fat feeding, we investigated the transcriptional and epigenetic response of human skeletal muscle to 9 days of a high-fat diet (HFD) alone (Sed-HFD) or in combination with resistance exercise (Ex-HFD), using genome-wide profiling of gene expression and DNA methylation. HFD markedly induced expression of immune and inflammatory genes, which was not attenuated by Ex. Conversely, Ex markedly remodelled expression of genes associated with muscle growth and structure. We detected marked DNA methylation changes following HFD alone and in combination with Ex. Among the genes that showed a significant association between DNA methylation and gene expression changes were PYGM, which was epigenetically regulated in both groups, and ANGPTL4, which was regulated only following Ex. In conclusion, while short-term Ex did not prevent a HFD-induced inflammatory response, it provoked a genomic response that may protect skeletal muscle from atrophy. These epigenetic adaptations provide mechanistic insight into the gene-specific regulation of inflammatory and metabolic processes in human skeletal muscle.

Full text

1
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
www.nature.com/scientificreports
Transcriptomic and epigenetic
responses to short-term nutrient-
exercise stress in humans
R. C. Laker1, C. Garde1, D. M. Camera2, W. J. Smiles2, J. R. Zierath1,3, J. A. Hawley2,4 &
R. Barrès
1
High fat feeding impairs skeletal muscle metabolic exibility and induces insulin resistance, whereas
exercise training exerts positive eects on substrate handling and improves insulin sensitivity. To
identify the genomic mechanisms by which exercise ameliorates some of the deleterious eects of
high fat feeding, we investigated the transcriptional and epigenetic response of human skeletal muscle
to 9 days of a high-fat diet (HFD) alone (Sed-HFD) or in combination with resistance exercise (Ex-
HFD), using genome-wide proling of gene expression and DNA methylation. HFD markedly induced
expression of immune and inammatory genes, which was not attenuated by Ex. Conversely, Ex
markedly remodelled expression of genes associated with muscle growth and structure. We detected
marked DNA methylation changes following HFD alone and in combination with Ex. Among the genes
that showed a signicant association between DNA methylation and gene expression changes were
PYGM, which was epigenetically regulated in both groups, and ANGPTL4, which was regulated only
following Ex. In conclusion, while short-term Ex did not prevent a HFD-induced inammatory response,
it provoked a genomic response that may protect skeletal muscle from atrophy. These epigenetic
adaptations provide mechanistic insight into the gene-specic regulation of inammatory and
metabolic processes in human skeletal muscle.
Skeletal muscle function is critical for voluntary movement, heat production and energy homeostasis1. e role
of skeletal muscle in metabolism and the control of blood glucose is particularly important, since this organ
is responsible for up to 80% of whole body insulin-stimulated glucose uptake2. Skeletal muscle is also highly
adaptive and displays a robust molecular and morphological response to diet and habitual physical activity3,4.
High fat diets are detrimental for the function of metabolic tissues, including skeletal muscle. Indeed, increases
in circulating lipids that accompany a fat-rich diet results in lipid accumulation within metabolic tissues, dis-
ruption to normal mitochondrial function, impaired insulin signalling and loss of muscle mass5–7. However,
high-fat, low-carbohydrate diets have become popular regimes to achieve weight loss, mainly due to the satiating
properties of these fatty acids. Resistance exercise promotes muscle hypertrophy and strength through the acti-
vation of signalling pathways that ultimately increase muscle protein synthesis8. Additionally, citrate synthase,
hexokinase9 and muscle-specic lipid oxidation capacity10 is increased following resistance exercise in human
skeletal muscle. Whether resistance exercise confers protection to skeletal muscle under conditions of a high-fat,
low-carbohydrate diet has not been investigated. Early transcriptomic responses following the transition from
a normal ‘healthy’ diet to a high fat diet (HFD) may provide important information as to the initial adaptive
responses that result in loss of muscle mass and metabolic dysfunction. We therefore determined whether resist-
ance exercise, in conjunction with high-fat feeding in humans, prevents maladaptive transcriptomic responses
typically observed aer such diets.
Altered DNA methylation has previously been linked to metabolic dysfunction in skeletal muscle of people
with type 2 diabetes11 and in response to acute, intense exercise12. erefore, to further probe the regulatory
mechanisms responsible for skeletal muscle gene expression in response to diet/nutrient stimuli, we investigated
1Novo Nordisk Foundation Center for Basic Metabolic Research, University of Copenhagen, Copenhagen, Denmark.
2Mary MacKillop Institute for Health Research, Centre for Exercise and Nutrition, Australian Catholic University,
Melbourne, Australia. 3Integrative Physiology, Department of Molecular Medicine and Surgery and Department of
Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden. 4Research Institute for Sport and Exercise
Sciences, Liverpool John Moores University, Liverpool, United Kingdom. Correspondence and requests for materials
should be addressed to R.B. (email: barres@sund.ku.dk)
Received: 13 July 2017
Accepted: 27 October 2017
Published: xx xx xxxx
OPEN
Correction: Author Correction
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
2
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
the epigenetic modication of DNA methylation. We hypothesized that altered DNA methylation may be an epi-
genetic mechanism responsible for exercise and/or high-fat diet-induced adaptations in skeletal muscle. Indeed,
changes in CpG methylation within promoters or enhancers can alter DNA structure and thereby block the access
of transcriptional machinery to DNA, resulting in altered or suppressed gene expression.
We performed transcriptomic and genome-wide DNA methylation proling in human skeletal muscle before
and aer nine days of HFD with or without three bouts of resistance exercise training in middle-aged, sedentary
males. We report that diet- and exercise-induced changes in DNA methylation were associated with very specic
gene regulation in the post-intervention resting state. Our ndings also demonstrate the robust impact of resist-
ance exercise on transcriptional remodelling in skeletal muscle, which may compensate for the deleterious eects
of high-fat diets. is study provides insight into the possible initiating mechanisms of HFD-induced inamma-
tion, metabolic dysfunction and loss of tissue mass in skeletal muscle.
Research Design and Methods
Study participants and experimental design. Thirteen healthy middle-aged sedentary men were
recruited for this study. Body weight and BMI were in the normal range (Table1). Participants were provided with
oral and written information about the purpose, nature and potential risks involved with the study, and written
informed consent was obtained prior to participation. All experimental protocols and methodologies related to
the study were approved by the Australian Catholic University Human Research Ethics Committee (#2015-103 H,
clinical trial registration date 12/10/2015) and conformed with the policy statement regarding the use of human
subjects in the latest revision of the Declaration of Helsinki. e trial was registered with the Australian New
Zealand Clinical Trials Registry (ACTRN 369316). A timeline of the experimental protocol that encompassed a
parallel groups design is shown (Fig.1). Ten days prior to the start of an intervention, all participants underwent
DEXA scan (GE Lunar Prodigy Pro, GE Healthcare) to determine body composition, and preliminary exercise
testing consisting of peak aerobic power (VO2peak) and one repetition maximum leg extension and leg press
strength testing. At the commencement of the experimental period, all participants were provided with a stand-
ardized pre-packed control diet (breakfast, lunch, dinner and snacks) for three days. is diet was customized to
the subject to provide 45 kcal/kg fat-free mass (FFM) per day with 6.1 g carbohydrate/kg FFM (55% totalcaloric
intake), 1.7 g protein/kg FFM (15%) and 1.5 g fat/kg FFM (30%). Following an overnight fast, biopsies were col-
lected under local anaesthesia (2-3 mL 1% Xylocaine) from the vastus lateralis muscle using a 5-mm Bergstrom
needle, modied with suction, and denoted as the “Pre” time point. Participants then commenced a high-fat
low-carbohydrate (HFD) diet consisting of 0.8 g carbohydrate kg/FFM (8% total caloric intake), 1.7 g protein/kg
FFM (15%) and 3.9 g fat/kg FFM (77%)13 for the remaining experimental period, which was a further 9 days. is
diet has been promoted by othersto induce nutritional ketosis and to be benecial for athletic performance and
weight loss13,14. is diet contains the required minerals to maintain physiological function, protein for lean body
mass and sucient carbohydrates for brain function13. We found this diet had no impact on muscle protein turn-
over in the study participants and published the results elsewhere15. Meal plans were created using Foodworks
7.0 ® Xyris Soware (Melbourne, Australia). Compliance was monitored and participants maintained a food
checklist. Participants were divided into two groups that were pair matched for fat-free mass and strength: partic-
ipants who remained sedentary (Sed-HFD) and those that performed three bouts of resistance exercise training
(Ex-HFD), starting 1 day aer commencing HFD, which corresponds to days 4, 7 and 10 of the experimental
timeline (Fig.1). Exercise consisted of 4 × 8–10 repetitions of leg press at 80% 1-RM, 4 × 8–10 repetitions of leg
extensions at 80% 1-RM, and 4 sets of dumbbell squats. ere was a 3 min recovery period between each set.
Muscle biopsies were collected under fasted conditions on day 11 and denoted as the “Post” time point. Blood
samples were collected on the morning of day 3, 5, 8 and 11 in EDTA tubes, centrifuged at 1,000 × g at 4 °C
Figure 1. Study design. Subjects underwent body composition and exercise testing at the commencement of
the experimental period. Aer 3 days of dietary control, subjects consumed a high-fat low-carbohydrate diet
(HFD) for 9 days. Subjects either remained sedentary (Sed-HFD) or performed resistance exercise (Ex-HFD)
on days 4, 7 and 10 of the experimental period. Blood samples were collected on days 3, 5, 8 and 11 and skeletal
muscle biopsies were obtained before and aer the diet/exercise intervention.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
3
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
for 15 min and stored at −80 °C. Total RNA and DNA were simultaneously isolated from the biopsy using the
Allprep® DNA/RNA/miRNA Universal Kit (Qiagen) according to the manufacturer’s instructions.
Plasma Analysis. Plasma tumour necrosis factor α (TNF-α) and interleukin-6 (IL-6) were measured
on 96-well plates utilizing commercially available and customised Milliplex Human magnetic bead panels
(Millipore, Massachusetts, USA) following the kit-specic protocols provided by Millipore. Analytes were quan-
tied in duplicate using the Magpix system utilising xPONENT 4.2 soware. Concentrations of these analytes
were determined on the basis of the t of a standard curve for mean uorescence intensity versus pg/ml. Two
quality controls with designated ranges were run with each assay to ensure validity of data generated. Plasma
FFA concentrations were determined by an enzymatic colorimetric method (Wako Diagnostics, Tokyo, Japan).
RNA sequencing. RNA was checked for quality using the Agilent RNA 600 nano kit and Bioanalyser instru-
ment (Agilent Technologies). 1 µg of RNA per sample was subject to the Illumina TruSeq Stranded Total RNA
with Ribo-Zero Gold protocol (Illumina) and performed as described16. Briey, ribosomal RNA was removed
from the sample using 35 µl rRNA removal beads (Illumina) on a magnetic plate followed by clean-up of the
ribosomal-depleted RNA with 193 µl Agencourt RNAClean XP beads (Beckman Coulter), 70% ethanol wash and
elution into 10 µl Elution buer (Illumina). e RNA sample was fragmented for 4 min at 94 °C in Elute, Prime,
Fragment High Mix (Illumina) and then subject to rst strand cDNA synthesis with 1 µl Superscript III reverse
transcriptase (Life Technologies) per sample and thermocycler programmed to 25 °C for 10 min, 50 °C for 15 min
and 70 °C for 15 min. Second strand cDNA was synthesized by addition of Second Strand Marking Master Mix
and samples subject to 16 °C for 60 min. Samples were subject to another bead clean up prior to A-tailing and liga-
tion of adapters as per kit instructions (Illumina). Following an third bead clean-up samples were enrich for DNA
fragments by amplication using the Illumina PCR Primer Cocktail and PCR Master Mix using a pre-dened
cycle number based on each individual sample and subject to 98 °C for 30 mins then X cycles of 98 °C for 10 secs,
60 °C for 30 secs and 72 °C for 30 secs and nally 72 °C for 5 min. Samples were cleaned and validated for DNA
concentration using the Qubit dsDNA HS assay kit (Invitrogen) and for base pair size and purity using the Aglient
High Sensitivity DNA chip and Bioanalyser instrument. Libraries were subjected to 100-bp single-end sequencing
onthe HiSeq 2500 (Illumina) at the Danish National High-roughput DNA Sequencing Centre. Approximately
8.5 million reads/sample were assigned to genes with 23,373 genes surviving the expression threshold.
DNA methylation analysis. Reduced Representation Bisulfite Sequencing (RRBS) was performed
as described17. Briey, 200 ng of DNA per sample was incubated overnight at 37 °C with MpsI enzyme (NEB
#R0106L) to fragment DNA at CCGG positions to enrich for CpG regions. Samples then underwent gap lling
and A-tailing with 1 µl dNTP mix (10 mM dATP, 1 mM dCTP, 1 mM dGTP) and 1 µl Klenow fragment 3′−5′ exo
(NEB) with 30 °C for 20 min and 37 °C for 20 min. Samples underwent bead clean-up using 90 µl AMPure beads
(Beckman Coulter) on a magnetic plate, 2 × 70% ethanol wash and elution into 20 µl elution buer. Illumina
Truseq adapters (diluted 1:20) were ligated with T4 ligase (NEB) and overnight incubation at 16 °C. e enzyme
was deactived by incubation at 65 °C for 20 mins. Samples were pooled (12 samples per pool), volume adjusted
with 20% polyethylene glycol and 2.5 M NaCl prior to bead clean-up in the DynaMag magnet. Bisulte con-
version was performed using the EZ DNA methylation Kit (Zymo Research) according to the manufacturer’s
instructions with 20 hr incubation with CT conversion reagent at 50 °C. DNA was then PCR amplied using Pfu
Turbo hotstart DNA polymerase, dNTP mix (100 mM, 25 mM each) and Illumina primer cocktail. Samples were
subject to 2 min at 95 °C followed by 14 cycles of 95 °C for 30 sec, 65 °C for 30 sec and 72 °C for 45 sec and nally
72 °C for 5 min. Samples underwent nal bead clean-up and library validation for DNA concentration using the
Qubit dsDNA HS assay kit (Invitrogen) and base pair size and purity using the Aglient High Sensitivity DNA
chip and Bioanalyser instrument. Libraries were subjected to 100-bp single-end sequencing onthe HiSeq 2500
(Illumina) at the Danish National High-roughput DNA Sequencing Centre.
Accession Numbers. Sequencing data are archived for public access at the Gene Expression Omnibus
(http://www.ncbi.nlm.nih.gov/geo) under accession number GSE99965.
Bioinformatic analysis. RNA-seq reads were subjected to trimming of adapters and low quality anking
ends using Trim Galore v0.3.7 and Cutadapt v1.4.2. Pre-processed reads were mapped to hg38 using Rsubread18
and gene coverages were computed with featureCounts19 and the Gencode annotation. e gene list was ltered to
those with a read coverage larger than 0.1 rpkm in at least 5 samples. Dierential expression was computed using
Ex-HFD (n = 7) Sed-HFD(n = 6)
Age (y) 37.3 ± 5.6 38.8 ± 5.3
Body Mass (kg) 89.4 ± 12.8 84.5 ± 7.4
BMI (kg m−2) 26.9 ± 3.0 27.2 ± 2.5
Lean Mass (kg) 60.1 ± 6.0 55.7 ± 5.9
VO2peak (ml/kg/min) 38.4 ± 4.2 34.5 ± 6.0
Leg Extension 1-RM (kg) 79.6 ± 15.4 71.8 ± 18.7
Leg Press 1-RM (kg) 219.7 ± 17.9 216.0 ± 49.2
Table 1. Baseline characteristics of the participants. Values are given as mean ± SD.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
4
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
edgeR with the glmQLFit/glmQLFTest modeling framework and the following models y~Timepoint + Subject
were used for each of the HFD-Ex and HFD-Sed groups20. Genes with a false discovery rate (FDR) below 0.1 were
considered dierentially expressed.
RRBS reads were processed with the ‘rrbs’ setting of Trim Galore v0.3.7 and Cutadapt v1.4.2. Processed reads
were mapped to hg38 followed by derivation of CpG methylation using Bismark21. Dierentially methylated
regions (FDR < 0.1) were identied using BiSeq22 from the subset of CpG sites that are covered by at least half of
the samples using the following models y~Timepoint + Subject for each of the HFD-Ex and HFD-Sed groups.
Our condence in bisulte conversion eciency was assessed based on the level of non-CpG methylation, which
at CHG sites averaged1.41% and at CHH sites averaged 1.39%, which is in the expected range (Fig.S1). We also
identied CpG sites known to be highly or lowly methylated in adult human skeletal muscle as well as 50 other
tissue and cell types using RRBS data available from the Epigenome Roadmap. We found that the methylation
levels of these sites within our analysis were consistent with the expected levels (Fig.S2). Finally, we assessed the
distribution of methylation of identied CpG sites for each sample, which was bimodal with peaks in the 0–15%
and 85–100% range (Fig.S3).
Statistics. Enrichment studies were conducted using hypergeometric tests and corrected for multiple testing
using the Benjamini-Hochberg procedure. An FDR of <0.1 was used as signicance level. Two-way ANOVA fol-
lowed by Student Newman Kuel’s post-hoc tests were performed to determine dierences between Ex-HFD and
Sed-HFD groups, and time (Pre-and Post). Data are presented as mean ± SEM with P values < 0.05 indicating
statistical signicance.
Results and Discussion
Short-term resistance exercise causes major changes in the skeletal muscle transcriptome
compared with HFD alone. To identify regulatory mechanisms involved in the early adaptive response to
high-fat feeding and the interaction with resistance exercise, we proled the skeletal muscle transcriptome before
and aer 9 days of HFD, with or without 3 bouts of resistance exercise. Within the RNA-seq analysis, 23,373
genes were annotated. Principal component analysis showed a clear separation of the treatment groups between
the Pre and Post time points, with no clear separation between individuals that performed resistance exercise
and individuals that remained sedentary aer the intervention (Fig.2A). To conrm the accuracy of our RNA-
seq results we compared the expression prole of genes previously analysed by qRT-PCR in a subset of samples
from the same experiment, which were previously reported15. We found that both the dierentially expressed
and unchanged genes were consistent between the two analyses (Fig.S4). Nine days of HFD in sedentary men
(Sed-HFD) resulted in dierential expression of 412 genes in skeletal muscle (Pre vs. Post), with 264 up-regulated
and 148 down-regulated (Fig.2B and D; TableS1). Conversely, when resistance exercise was performed in com-
bination with HFD (Ex-HFD), a greater transcriptomic response was evident, with 2,617 genes changed (Pre vs.
Post) of which 1,561 were up-regulated and 1,056 were down-regulated (Fig.2C and E; TableS1). e dierent
magnitude of change in gene expression suggests that resistance exercise initiates robust transcriptional activity
in skeletal muscle, which far outweighs the impact of HFD alone.
Of the 1,561 up-regulated genes following Ex-HFD, only 240 genes were also up-regulated in the Sed-HFD
group. is implies that the remaining 1,321 genes were up-regulated as a direct response to resistance exer-
cise (Fig.3A). Similarly, of the 1,056 genes down-regulated in the Ex-HFD group, only 103 genes were also
down-regulated in the Sed-HFD group, suggesting that resistance exercise was primarily responsible for the
down-regulation of the remaining 953 genes. We performed an enrichment analysis of gene ontology (GO)
terms to identify whether the genes regulated in the Ex-HFD group were associated with specic cellular
compartments, biological processes and molecular functions (TableS2). However, due to the large number
of dierentially expressed genes, we retrieved a long list of GO terms. erefore, we used the Revigo tool to
summarize the GO terms based on semantic similarity23. We found many GO terms related to skeletal muscle
structure, myogenic activity and metabolism (Fig.3C and D; TableS2). Consistent with the biology of skeletal
muscle, we found that the cellular compartment GOs were related to neuromuscular junction, sarcomere and
mitochondrion (Fig.3C and D;TableS2), while biological process GOs included skeletal muscle satellite cell
migration, cell junction assembly and muscle cell dierentiation, among other metabolic and transcriptional
processes (Fig.3C and D; TableS2). Finally, molecular function gene ontologies were related to signal trans-
ducer activity, myogenic regulatory factor binding and structural constituent of muscle (Fig.3C and D; TableS2).
Of note, the top two GOs of the down-regulated genes were rhythmic process and circadian rhythm (Fig.3D;
TableS2). Since the muscle biopsies were taken at the same time of the day (~8am), pre- and post-intervention,
this suggests that resistance exercise may transcriptionally regulate the innate circadian oscillations of skeletal
muscle. is may be important considering the close link between circadian and metabolic gene regulation24,25
and could have widespread implications for the timing of exercise to optimize metabolic health outcomes.
Taken collectively, our analyses suggest that resistance exercise may protect skeletal muscle against the negative
impact of HFD. However, we acknowledge that functional outcomes are dicult to determine based on the
short-term exercise intervention in the current study. Of note, metabolic analyses from this cohort revealed
the Ex-HFD group showed a tendency for improved glucose tolerance15. e functional impact of resistance
exercise, when performed in conjunction with HFD, may become more apparent aer a prolonged exercise
training regimen.
Short-term HFD induces immune and inammatory genes regardless of physical activity. We
identied 343 genes that were dierentially regulated following both interventions (Sed-HFD and Ex-HFD;
Fig.4A), with 240 genes up-regulated and 103 genes down-regulated (Fig.3A and B). Only one gene, ChaC
Glutathione Specic Gamma-Glutamylcyclotransferase 1 (CHAC1), showed divergent transcriptional regulation
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
5
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
between Sed-HFD and Ex-HFD groups (Fig.4A). CHAC1 plays a role in glutathione degradation, notch sig-
nalling and activation of autophagy and apoptosis26–28. To our knowledge, the only report on CHAC1 in skel-
etal muscle suggests that CHAC1 is induced in response to re-feeding and participates in the unfolded protein
response29, which is consistent with our observation that CHAC1 is regulated by the nutritional state in humans.
Next, we performed gene ontology analysis of the 344 genes dierentially regulated by both Sed-HFD and
Ex-HFD. We identied at least 20 individual GO terms associated with immune and inammatory processes
(Fig.4B and C). Of potential interest, many of the up-regulated GO terms were associated with the extracel-
lular space, which suggests that immune and inammatory signalling is occurring outsidethe skeletal muscle,
likely through a combination of secreted and membrane-expressed proteins, as well as recruitment of mac-
rophages. Whether macrophage recruitment or muscle damage per se is driving the immune and inammatory
Figure 2. Short-term resistance exercise initiates robust transcriptional regulation compared with HFD alone.
(A) Principal component analysis (PCA) of RNA-seq for the major two principal components (PC). e 95%
condence ellipses are shown for each group. Heatmaps (B and C) and volcano plots (D and E) represent
dierentially expressed genes in skeletal muscle before (Pre) and aer (Post) 9 days of HFD (Sed-HFD; B and
D) or HFD with 3 bouts of resistance exercise training (Ex-HFD; C and E). FDR < 0.1.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
6
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
gene response aer a combined HFD and resistance exercise regimen warrants further investigation. Within the
cellular compartment category, the down-regulated GO terms were highly associated with mitochondria, and
supports the notion that HFD induces impairments in mitochondrial function (Fig.4C). Indeed, many of the
Figure 3. Short-term resistance exercise is associated with large-scale transcriptional remodelling in skeletal
muscle in the presence of HFD. Venn diagrams representing the number of genes that were up-regulated
(A) and down-regulated (B) in Sed-HFD and Ex-HFD group (excludes CHAC1 gene, which was upregulated in
Ex-HFD and downregulated in Sed-HFD). e intersection represents the genes that were regulated following
both interventions. Gene ontology analysis of genes that were up-regulated (C) and down-regulated (D)
exclusively in the Ex-HFD group. FDR < 0.1 is shown by the dotted line.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
7
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
biological function GOs terms derived from the down-regulated genes were associated with mitochondrial func-
tion and metabolic processes including cellular respiration, metabolic process, gluconeogenesis, glycolytic process
and canonical glycolysis (Fig.4C).
Figure 4. HFD induces immune and inammatory genes associated with systemic inammation, regardless of
physical activity. (A) Scatter plot of 344 dierentially expressed gene following 9 days of HFD, with or without
resistance exercise (Sed-HFD and Ex-HFD intersection). Gene ontology analysis of the 344 genes that were
either up-regulated (B) or down-regulated (C) in both the Sed-HFD and Ex-HFD groups. FDR < 0.1 is shown
by the dotted line. Plasma proles of free fatty acids (FFA; D), interleukin 6 (IL-6; E) and TNF-α (F) throughout
the intervention period. #p < 0.002 for the eect of time; $p < 0.02 for the eect of exercise.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
8
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
In support of the transcriptional profiling data, we found that circulating free fatty acids progressively
increased throughout the high-fat diet intervention in both groups (Fig.4D). Surprisingly, this occurred to a
greater extent in the exercise group compared with the sedentary group (Fig.4D). We also observed a marked
elevation of the inammatory markers IL-6 and TNF-α in both groups (Fig.4E and F). Collectively, these ndings
suggest that HFD, regardless of physical activity, elevates circulating lipids associated with systemic inammation,
and promotes local inammation/immune responses in skeletal muscle. ese observations suggest that concom-
itant resistance exercise does not fully protect skeletal muscle from deleterious HFD-induced gene regulation. We
cannot exclude the possibility that a longer exposure to an exercise stimulus (i.e. weeks, months) may reverse this
response.
Transcriptional response to short-term HFD. We report that 67 genes were uniquely altered in the
Sed-HFD group and unchanged in the Ex-HFD group. is number excludes the CHAC1 gene which was
altered in both groups but in opposite directions (see Fig.4A). is nding suggests that resistance exer-
cise preserved expression of these genes at a basal level. Although we were unable to perform gene ontol-
ogy analysis due to the small number of genes, we found that many of these 67 genes were associated with
mitochondrial function and localization, metabolic enzymes, and immune regulation (Table2). Of particu-
lar interest are the Interleukin 1 Receptor Associated Kinase 2 (IRAK2), Nuclear Protein 1, Transcriptional
Regulator (NUPR1; also known as p8) and Fibroblast Growth Factor 6 (FGF6) (Table2). IRAK2 was increased
by ~2-fold following HFD. IRAK2 is a receptor for IL1 and mediates toll like receptor (TLR) and NF-ĸB
signalling to induce transcription and mRNA stabilization for chemokine production (TNFα, IL6, IL1)30,31.
e early induction of IRAK2 by HFD feeding suggests that IRAK2 participates in the initiation of inamma-
tory signalling in skeletal muscle32–34 and may be a key mechanism underlying subsequent insulin resistance.
NUPR1 was also induced following HFD by ~1.7-fold. NUPR1 interacts directly with the critical myogenic
regulatory factor MyoD to regulate MyoD target genes35. In C2C12 myoblasts, over expression of NUPR1
represses MyoD and myogenin gene expression35. Furthermore, NUPR1 is involved in resistance to stress
induced by a change in the microenvironment36,37. us, our observation that NUPR1 is induced following
HFD may represent a stress response to the inux of fatty acids, which ultimately results in the HFD-induced
loss of muscle mass. Finally, FGF6 was decreased ~50% following nine days of HFD in skeletal muscle. FGF6
plays an role in muscle regeneration and dierentiation through stimulation of satellite cell proliferation and
early dierentiation38–40. We speculate that the decrease of FGF6 along with increased NUPR1 participates in
the loss of muscle mass with HFD.
In addition to the set of known genes that were dierentially expressed following HFD, there was a set of
eight transcripts with no identied protein product (Table3). ese transcripts exhibited an extremely robust
decrease in expression following HFD (between ~40–98% reduction; Table3). e transcripts are quite long
and could therefore, be long non-coding RNA that could confer a transcriptional response. Alternatively, some
Ensembl ID Gene Description logFC P value FDR
ENSG00000124107 SLPI secretory leukocyte peptidase inhibitor 2,82 7,17E-09 7,28E-06
ENSG00000138193 PLCE1 phospholipase C epsilon 1 0,94 2,22E-05 4,80E-03
ENSG00000014641 MDH1 malate dehydrogenase 1 −0,79 3,68E-05 7,17E-03
ENSG00000134070 IRAK2 interleukin 1 receptor associated kinase 2 1,01 5,31E-05 8,94E-03
ENSG00000176046 NUPR1 nuclear protein 1, transcriptional regulator 0,76 8,99E-05 1,29E-02
ENSG00000154518 ATP5G3 ATP synthase, H + transporting, mitochondrial Fo
complex subunit C3 (subunit 9) −0,74 1,07E-04 1,46E-02
ENSG00000112715 VEGFA vascular endothelial growth factor A −0,73 1,47E-04 1,88E-02
ENSG00000087586 AURKA aurora kinase A −1,14 1,56E-04 1,95E-02
ENSG00000110955 ATP5B ATP synthase, H + transporting, mitochondrial F1
complex, beta polypeptide −0,71 2,88E-04 3,13E-02
ENSG00000159423 ALDH4A1 aldehyde dehydrogenase 4 family member A1 −0,68 5,48E-04 4,75E-02
ENSG00000263232 ATP5A1P3 ATP synthase, H + transporting, mitochondrial F1
complex, alpha subunit 1 pseudogene 3 −0,71 6,70E-04 5,32E-02
ENSG00000132313 MRPL35 mitochondrial ribosomal protein L35 −0,70 7,28E-04 5,63E-02
ENSG00000111241 FGF6 broblast growth factor 6 −1,05 8,14E-04 6,18E-02
ENSG00000169692 AGPAT2 1-acylglycerol-3-phosphate O-acyltransferase 2 0,73 8,36E-04 6,18E-02
ENSG00000176340 COX8 A cytochrome c oxidase subunit VIIIA (ubiquitous) −0,63 9,09E-04 6,48E-02
ENSG00000166343 MSS51 MSS51 mitochondrial translational activator −0,71 1,02E-03 7,01E-02
ENSG00000184076 UQCR10 ubiquinol-cytochrome c reductase, complex III
subunit X −0,61 1,23E-03 7,98E-02
ENSG00000244482 LILRA6 leukocyte immunoglobulin-like receptor, subfamily
A (with TM domain), member 6 2,51 1,43E-03 8,74E-02
ENSG00000260318 COX6CP1 cytochrome c oxidase subunit VIc pseudogene 1 −1,05 1,72E-03 9,85E-02
ENSG00000120992 LYPLA1 lysophospholipase I −0,90 1,72E-03 9,85E-02
Table 2. Selected genes associated with mitochondrial function and localization, metabolic enzymes, and
immune regulation dierentially regulated by HFD and preserved by concomitant resistance exercise.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
9
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
of these genes appear to be subject to splicing and may be host genes to microRNAs, which are transcribed,
spliced and then degraded. In any case, the function of these transcripts remains unknown and they may
play a critical role in maintaining mass and/or metabolic homeostasis in skeletal muscle. An important future
direction will be to determine if these HFD-regulated genes also participate in HFD-induced inammation,
metabolic dysregulation or loss of muscle mass, and how resistance exercise prevents this HFD-induced genetic
regulation.
Epigenetic response to HFD and resistance exercise. We used RRBS to prole dierentially meth-
ylated DNA regions (DMRs) with the aim of investigating whether epigenetic mechanisms play a role in the
transcriptomic response to nutrient/exercise. We found that a HFD alone induced a greater degree of hypermeth-
ylation, while concomitant resistance exercise resulted in a preference towards hypomethylation of DNA (Fig.5F
and G). We found 809 DMRs following short-term HFD, while concomitant resistance exercise revealed 474
DMRs, with only 38 DMRs common between the two groups (Fig.5A–C). Gene ontology analysis of all DMRs
revealed only one term, which was regulation of transcription. Furthermore, we identied only 10 genes in the
Sed-HFD group and 54 genes in the Ex-HFD group that showed signicant association with dierential gene
expression (Fig.5A and C). is limited relationship is surprising considering that more DMRs were found in
promoters, compared with other genomic regions (Fig.5D and E), and theoretically would cause altered expres-
sion of the genes associated with these promoters. One could speculate that thespecic location of the DMR
within the promoter region will be an important factor dictating whether gene expression would be altered,
which is dependent on the recruitment of methy-CpG binding proteins specically to the transcription factor
binding sites to block their access. Another explanation is that the DMRs have no apparent impact unless there
is a stimulus to initiate transcriptional activity of the associated gene, following which the functional impact of
DNA methylation or demethylation can be realized. Indeed, in the present study, biopsies were taken 48 hours
aer the nal exercise bout, when skeletal muscle was in a resting/basal state with low levels of transcriptional
activity. erefore, the dierential methylation states identied may have regulatory functions for transcriptional
activation/repression during times of stress or stimulation. For example, DNA demethylation following resistance
exercise training may function to poise a promoter region for rapid transcriptional activation in response to a
subsequent exercise bout, as previously hypothesized41. We propose that the importance of DNA methylation for
exercise adaptation may be observed if the biopsy was sampled during a dynamic period of transcription such as
immediately aer exercise.
e glycogen phosphorylase, muscle associated (PYGM) gene was one of the genes that did show a rela-
tionship between expression and promoter methylation. e promoter of PYGM was hypermethylated in both
groups following HFD, regardless of exercise and the level of methylation was inversely correlated with gene
expression (Fig.5H). PYGM is an enzyme involved in the breakdown of glycogen to glucose-1-phosphate42,43.
e epigenetic regulation of PYGM may be an adaptive response to the high-fat low-carbohydrate diet, since
glycogen stores are expected to progressively decline in skeletal muscle and the requirement for an enzyme that
breaks down glycogen (i.e. PYGM) would be abolished in the absence of the substrate. Indeed, the decline in
muscle glycogen stores has been reported in human muscle following a similar diet for 4 weeks14. Meanwhile,
angiopoiten like 4 (ANGPTL4) was hypomethylated in both groups, with changes in gene expression only
observed following resistance exercise, but to a substantial degree (Fig.5I). ANGPTL4 is a secreted serum hor-
mone that regulates blood glucose, lipid metabolism and insulin sensitivity44–46. Hypomethylation of ANGPTL4
may be a compensatory response to HFD, yet a stimulus such as exercise may be required for transcriptional
activation. Interestingly, ANGPTL4 induces lipolysis in adipocytes44. us, increased expression of ANGPTL4 in
the Ex-HFD group may contribute to the higher levels of FFAs as compared with the sedentary group provided
with the same HFD.
In conclusion, we have reported transcriptional proles in skeletal muscle from men fed a high-fat diet with
or without a concomitant resistance exercise intervention. e extent to which resistance exercise can prevent
the deleterious impact of HFD on skeletal muscle function remain unanswered, but the dramatic number of
exercise-responsive genes identied suggest that growth and development pathways are up-regulated, along with
changes in metabolism and transcription. We found signicant changes in DNA methylation, predominantly at
promoter regions. Furthermore, genes associated with the DMRs were generally unrelated between the Sed-HFD
and Ex-HFD groups and their functional signicance has yet to be determined. Overall, our ndings suggest that
resistance exercise may be a promising intervention to maintain skeletal muscle mass and metabolic health under
conditions of a high-fat diet, currently prevalent throughout the world. Future studies are warranted to investigate
Ensembl ID Gene logFC P value FDR
ENSG00000261303 RP11–160C18.2 −5,36 6,96E-06 2,06E-03
ENSG00000262420 RP11–490O6.2 −1,44 1,19E-04 1,60E-02
ENSG00000224550 RP11–270C12.3 −0,72 2,21E-04 2,58E-02
ENSG00000272631 RP11–359E3.4 −2,27 3,88E-04 3,81E-02
ENSG00000271347 RP11–701H24.7 −1,48 1,07E-03 7,27E-02
ENSG00000277954 RP11–679B19.1 −1,18 1,22E-03 7,98E-02
ENSG00000223935 AC008074.3 −2,24 1,26E-03 8,00E-02
ENSG00000183154 RP11–863K10.7 −2,30 1,74E-03 9,91E-02
Table 3. Un-annotated genes down-regulated by HFD and protected by concomitant resistance exercise.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
10
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
the long-term functional outcomes of HFD and resistance exercise combinations, as well as dynamic exercise
studies to dissect the epigenome-transcriptome relationship in skeletal muscle.
References
1. Zierath, J. . & Hawley, J. A. Seletal muscle ber type: inuence on contractile and metabolic properties. PLoS Biol. 2, e348 (2004).
2. DeFronzo, . A., Sherwin, . S. & raemer, N. Eect of physical training on insulin action in obesity. Diabetes 36, 1379–85 (1987).
3. Hawley, J. A., Bure, L. M., Phillips, S. M. & Spriet, L. L. Nutritional modulation of training-induced seletal muscle adaptations. J.
Appl. Physiol. 110, 834–845 (2011).
4. Hawley, J. A., Hargreaves, M., Joyner, M. J. & Zierath, J. . Integrative biology of exercise. Cell 159, 738–49 (2014).
5. Martins, A. . et al. Mechanisms underlying seletal muscle insulin resistance induced by fatty acids: importance of the
mitochondrial function. Lipids Health Dis. 11, 30 (2012).
6. Itani, S. I., uderman, N. B., Schmieder, F. & B oden, G. Lipid-induced insulin resistance in human muscle is associated with changes
in diacylglycerol, protein inase C, and IappaB-alpha. Diabetes 51, 2005–11 (2002).
Figure 5. Dierential DNA methylation following short-term HFD with or without resistance exercise. Venn
diagrams representing the number of genes that were dierentially methylated and dierentially expressed
following 9 days of HFD (A) and with concomitant resistance exercise (C). Venn diagram showing the number
of dierentially methylated genes between Sed-HFD and Ex-HFD groups (B). Genomic context of DMRs
detected in the Sed-HFD (D) and Ex-HFD (E) groups. Volcano plots showing directional changes and FDR of
DMRs in the Sed-HFD (F) and Ex-HFD (G) groups. DNA methylation level relative to transcription start site
(TSS) and gene expression (insert) for the genes PYGM (H) and ANGPTL4 (I).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
11
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
7. ivas, D. A. et al. Diminished anabolic signaling response to insulin induced by intramuscular lipid accumulation is associated with
inammation in aging but not obesity. Am. J. Physiol. - egul. Integr. Comp. Physiol. 310, 561–569 (2016).
8. Coey, V. G. & Hawley, J. A. e molecular bases of training adaptation. Sports Med. 37, 737–63 (2007).
9. Tang, J. E., Hartman, J. W. & Phillips, S. M. Increased muscle oxidative potential following resistance training induced fibre
hypertrophy in young men. Appl. Physiol. Nutr. Metab. 31, 495–501 (2006).
10. Pesta, D. et al. Similar qualitative and quantitative changes of mitochondrial respiration following strength and endurance training
in normoxia and hypoxia in sedentary humans. Am. J. Physiol. egul. Integr. Comp. Physiol. 301, 1078–87 (2011).
11. Barrès, . et al. Non-CpG Methylation of the PGC-1α Promoter through DNMT3B Controls Mitochondrial Density. C ell Meta b. 10,
189–198 (2009).
12. B arrès, . et al. Acute Exercise emodels Promoter Methylation in Human Seletal Muscle. Cell Me tab. 15, 405–411 (2012).
13. Vole, J. S. & Phinney, S. D. e art and science of low carbohydrate living. (Beyond Obesity LLC, 2011).
14. Phinney, S. D., Bistrian, B. ., Evans, W. J., Gervino, E. & Blacburn, G. L. e human metabolic response to chronic etosis without
caloric restriction: preservation of submaximal exercise capability with reduced carbohydrate oxidation. Metabolism. 32, 769–76 (1983).
15. Camera, D. M., Burniston, J. G., Pogson, M. A., Smiles, W. J. & Hawley, J. A. Dynamic proteome proling of individual proteins in
human seletal muscle aer a high-fat diet and resistance exercise. FASEB J. https://doi.org/10.1096/.201700531 (2017).
16. Nylander, V. et al. Ionizing adiation Potentiates High-Fat Diet–Induced Insulin esistance and eprograms Seletal Muscle and
Adipose Progenitor Cells. Diabetes 65, 3573–3584 (2016).
17. Donin, I. et al. Obesity and Bariatric Surgery Drive Epigenetic Variation of Spermatozoa in Humans. Cell Metab. 23, 369–378
(2016).
18. Liao, Y., Smyth, G. . & Shi, W. e Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids es.
41, e108 (2013).
19. Liao, Y., Smyth, G. . & Shi, W. featureCounts: an ecient general purpose program for assigning sequence reads to genomic
features. Bioinformatics 30, 923–30 (2014).
20. obinson, M. D., McCarthy, D. J. & Smyth, G. . edge: a Bioconductor pacage for dierential expression analysis of digital gene
expression data. Bioinformatics 26, 139–40 (2010).
21. rueger, F. & Andrews, S. . Bismar: a exible aligner and methylation caller for Bisulte-Seq applications. Bioinformatics 27,
(1571–2 (2011).
22. Hebestreit, ., Dugas, M. & lein, H.-U. Detection of signicantly dierentially methylated regions in targeted bisulte sequencing
data. Bioinformatics 29, 1647–53 (2013).
23. Supe, F., Bošnja, M., Šunca, N. & Šmuc, T. EVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One 6,
e21800 (2011).
24. Hodge, B. A. et al. e endogenous molecular cloc orchestrates the temporal separation of substrate metabolism in seletal muscle.
Selet. Muscle 5, 17 (2015).
25. Masri, S., Zocchi, L., atada, S., Mora, E. & Sassone-Corsi, P. e circadian cloc transcriptional complex: metabolic feedbac
intersects with epigenetic control. Ann. N. Y. Acad. Sci. 1264, 103–109 (2012).
26. umar, A. et al. Mammalian proapoptotic factor ChaC1 and its homologues function as γ-glutamyl cyclotransferases acting
specically on glutathione. EMBO ep. 13, 1095–101 (2012).
27. Crawford, . . et al. Human CHAC1 Protein Degrades Glutathione, and mNA Induction Is egulated by the Transcription
Factors ATF4 and ATF3 and a Bipartite ATF/CE egulatory Element. J. Biol. Chem. 290, 15878–91 (2015).
28. Mungrue, I. N., Pagnon, J., ohannim, O., Gargalovic, P. S. & Lusis, A. J. CHAC1/MGC4504 is a novel proapoptotic component of
the unfolded protein response, downstream of the ATF4-ATF3-CHOP cascade. J. Immunol. 182, 466–76 (2009).
29. Bower, N. I. & Johnston, I. A. Discovery and characterization of nutritionally regulated genes associated with muscle growth in
Atlantic salmon. Physiol. Genomics 42A, 114–30 (2010).
30. Flannery, S. M., eating, S. E., Szyma, J. & Bowie, A. G. Human interleuin-1 receptor-associated inase-2 is essential for Toll-lie
receptor-mediated transcriptional and post-transcriptional regulation of tumor necrosis factor alpha. J. Biol. Chem. 286, 23688–97
(2011).
31. eating, S. E., Maloney, G. M., Moran, E. M. & Bowie, A. G. IA-2 participates in multiple toll-lie receptor signaling pathways to
NFappaB via activation of TAF6 ubiquitination. J. Biol. Chem. 282, 33435–43 (2007).
32. eyna, S. M. et al. Elevated Toll-Lie eceptor 4 Expression and Signaling in Muscle From Insulin-esistant Subjects. Diabetes 57,
2595–2602 (2008).
33. adin, M. S., Sinha, S., Bhatt, B. A., Dedousis, N. & O’Doherty, . M. Inhibition or deletion of the lipopolysaccharide receptor Toll-
lie receptor-4 confers partial protection against lipid-induced insulin resistance in rodent seletal muscle. Diabetologia 51, 336–346
(2008).
34. Shi, H. et al. TL4 lins innate immunity and fatty acid–induced insulin resistance. J. Clin. Invest. 116, 3015–3025 (2006).
35. Sambasivan, . et al. e small chromatin-binding protein p8 coordinates the association of anti-proliferative and pro-myogenic
proteins at the myogenin promoter. J. Cell Sci. 122, 3481–91 (2009).
36. Hamidi, T. et al. Nupr1-aurora inase A pathway provides protection against metabolic stress-mediated autophagic-associated cell
death. Clin. Cancer es. 18, 5234–46 (2012).
37. Jin, H.-O. et al. Nuclear protein 1 induced by ATF4 in response to various stressors acts as a positive regulator on the transcriptional
activation of ATF4. IUBMB Life 61, 1153–8 (2009).
38. Coulier, F., Pizette, S., Ollendor, V., deLapeyrière, O. & Birnbaum, D. e human and mouse broblast growth factor 6 (FGF6)
genes and their products: possible implication in muscle development. Prog. Growth Factor es. 5, 1–14 (1994).
39. Pizette, S., Coulier, F., Birnbaum, D. & DeLapeyrière, O. FGF6 modulates the expression of broblast growth factor receptors and
myogenic genes in muscle cells. Exp. Cell es. 224, 143–51 (1996).
40. Douas, J. et al. Delivery of FGF genes to wound repair cells enhances arteriogenesis and myogenesis in seletal muscle. Mol. er.
5, 517–27 (2002).
41. Barrès, . & Zierath, J. . e role of diet and exercise in the transgenerational epigenetic landscape of T2DM. Nat. ev. Endocrinol.
12, 441–51 (2016).
42. Hudson, J. W., Golding, G. B. & Crerar, M. M. Evolution of allosteric control in glycogen phosphorylase. J. Mol. Biol. 234, 700–21
(1993).
43. Newgard, C. B., Littman, D. ., van Genderen, C., Smith, M. & Fletteric, . J. Human brain glycogen phosphorylase. Cloning,
sequence analysis, chromosomal mapping, tissue expression, and comparison with the human liver and muscle isozymes. J. Biol. Chem.
263, 3850–7 (1988).
44. Sanderson, L. M. et al. Peroxisome proliferator-activated receptor beta/delta (PPAbeta/delta) but not PPA alpha serves as a plasma
free fatty acid sensor in liver. Mol. Cell. Biol. 29, 6257–67 (2009).
45. Xu, A. et al. Angiopoietin-lie protein 4 decreases blood glucose and improves glucose tolerance but induces hyperlipidemia and
hepatic steatosis in mice. Proc. Natl. Acad. Sci. USA 102, 6086–91 (2005).
46. Staiger, H. et al. Muscle-Derived Angiopoietin-Lie Protein 4 Is Induced by Fatty Acids via Peroxisome Proliferator–Activated
eceptor (PPA)-δ and Is of Metabolic elevance in Humans. Diabetes 58 (2009).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

www.nature.com/scientificreports/
12
SCIENtIFIC RePoRTS | 7:15134 | DOI:10.1038/s41598-017-15420-7
Acknowledgements
e Novo Nordisk Foundation Centre for Basic Metabolic Research is an independent research centre at the
University of Copenhagen partially funded by an unrestricted donation from the Novo Nordisk Foundation.
Author Contributions
R.C.L. performed experiments, analysed the data and wrote the manuscript; C.G. performed bioinformatics
analysis and generated gures; D.M.C. and W.J.S. performed human experiments and plasma analysis; J.R.Z.
provided expert advice and edited the manuscript; J.H. designed the study and edited the manuscript; R.B.
designed the study, analysed the data and wrote the manuscript.
Additional Information
Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-017-15420-7.
Competing Interests: e authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
© e Author(s) 2017
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com