J Appl Physiol 112: 806–815, 2012.

First published December 15, 2011; doi:10.1152/japplphysiol.00864.2011.

Exercise-associated generation of PPAR␥ ligands activates PPAR␥ signaling

events and upregulates genes related to lipid metabolism
A. W. Thomas,1* N. A. Davies,1* H. Moir,1 L. Watkeys,1 J. S. Ruffino,1 S. A. Isa,1 L. R. Butcher,1
M. G. Hughes,2 K. Morris,1 and R. Webb1
1
Cardiff School of Health Sciences and 2Cardiff School of Sport, University of Wales Institute Cardiff, Cardiff,
United Kingdom
Submitted 12 July 2011; accepted in final form 8 December 2011

Thomas AW, Davies NA, Moir H, Watkeys L, Ruffino JS, Isa THE RECENT JOINT POSITION statement released by the American
SA, Butcher LR, Hughes MG, Morris K, Webb R. Exercise- College of Sports Medicine and the American Diabetes Asso-
associated generation of PPAR␥ ligands activates PPAR␥ signaling ciation stated that “it is now well established that participation
events and upregulates genes related to lipid metabolism. J Appl in regular physical activity canѧprevent or delay Type 2 Dia-
Physiol 112: 806 – 815, 2012. First published December 15, 2011; betes (T2D), and positively affect lipids, blood pressure, car-
doi:10.1152/japplphysiol.00864.2011.—The aim of the present study
diovascular events, mortality and quality of life” (2). During
was to test the hypotheses that exercise is associated with generation
of peroxisome proliferator-activated receptor-␥ (PPAR␥) ligands in
exercise, energy demand increases dramatically in skeletal
the plasma and that this may activate PPAR␥ signaling within circu- muscle; consequently, skeletal muscle was among the first
lating monocytes, thus providing a mechanism to underpin the exer- tissues to be investigated with regard to responses to exercise
cise-induced antiatherogenic benefits observed in previous studies. A (21). However, the diversity of the effects listed above indi-
cohort of healthy individuals undertook an 8-wk exercise-training cates that exercise is a systemic phenomenon with impact that
program; samples were obtained before (Pre) and after (Post) stan- is not limited to one tissue. Indeed, more recent research has
dardized submaximal exercise bouts (45 min of cycling at 70% of shown that signaling events within a wide range of cell types
maximal O2 uptake, determined at baseline) at weeks 0, 4, and 8. are affected by exercise (52).
Addition of plasma samples to PPAR␥ response element (PPRE)- The present study focuses on the effects of exercise on
luciferase reporter gene assays showed increased PPAR␥ activity monocytes, which are important in several aspects of the
following standardized exercise bouts (Post/Pre ⫽ 1.23 ⫾ 0.10 at pathophysiology of T2D, including cardiovascular complica-
week 0, P ⬍ 0.05), suggesting that PPAR␥ ligands were generated tions such as atherosclerosis [with a link to T2D that is largely
during exercise. However, increases in PPAR␥/PPRE-luciferase ac-
due to the proatherogenic nature of glycated proteins and lipids
tivity in response to the same standardized exercise bout were blunted
during the training program (Post/Pre ⫽ 1.18 ⫾ 0.14 and 1.10 ⫾ 0.10
(48)]. Monocytes have roles in control of blood-borne lipid
at weeks 4 and 8, respectively, P ⬎ 0.05 for both), suggesting that the homeostasis via their contribution to reverse cholesterol trans-
relative intensity of the exercise may affect PPAR␥ ligand generation. port (9, 29) and in paracrine control of inflammatory signaling
In untrained individuals, specific transient increases in monocyte (4); also, adherence of monocytes to the endothelial wall and
expression of PPAR␥-regulated genes were observed within 1.5–3 h their migration into the subendothelial space and differentia-
of exercise (1.7 ⫾ 0.4, 2.6 ⫾ 0.4, and 1.4 ⫾ 0.1 fold for CD36, liver tion into macrophage foam cells are crucial to the development
X receptor-␣, and ATP-binding cassette subfamily A member 1, of atherosclerosis (48). The contribution of monocyte-medi-
respectively, P ⬍ 0.05), with expression returning to basal levels ated reverse cholesterol transport to healthy lipid homeostasis
within 24 h. In contrast, by the end of the exercise program, expres- is a major focus of the present study: after uptake by mono-
sion at the protein level of PPAR␥ target genes had undergone cytes, lipoprotein components are processed and presented to
sustained increases that were not associated with an individual exer- apolipoprotein A-1 (ApoA1), thus forming HDL-cholesterol,
cise bout (e.g., week 8 Pre/week 0 Pre ⫽ 2.79 ⫾ 0.61 for CD36, P ⬍ which acts to reduce cardiovascular risk, as cholesterol bound
0.05). Exercise is known to upregulate PPAR␥-controlled genes to to this lipoprotein is transported to the liver for safe disposal (5,
induce beneficial effects in skeletal muscle (e.g., mitochondrial bio-
9, 29). This may be of considerable clinical importance: selec-
genesis and aerobic respiration). We suggest that parallel exercise-
induced benefits may occur in monocytes, as monocyte PPAR␥ tive activation of this pathway in peripheral cells such as
activation has been linked to beneficial antidiabetic effects (e.g., monocytes has been recommended as a potential therapeutic
exercise-induced upregulation of monocytic PPAR␥-controlled genes target for atherosclerosis (5).
is associated with reverse cholesterol transport and anti-inflammatory The ligand-activated nuclear receptor peroxisome prolifera-
effects). Thus, exercise-triggered monocyte PPAR␥ activation may tor-activated receptor-␥ (PPAR␥) is a regulator of the expres-
constitute an additional rationale for prescribing exercise to type 2 sion of numerous PPAR␥ target genes [i.e., genes bearing
diabetes patients. PPAR response elements (PPREs), to which activated PPAR␥
peroxisome proliferator-activated receptor-␥; exercise training; mono-
can bind] involved in metabolism, cell differentiation, and
cytes; reverse cholesterol transport inflammation (44, 51). Thus, the actions of synthetic agonists
for PPAR␥ [e.g., thiazolidinediones such as rosiglitazone (27)]
in regulating PPAR␥ target gene expression are responsible for
beneficial effects in the context of T2D (10). However, the
transcriptional activity of PPAR␥ may also be affected by its
* A. W. Thomas and N. A. Davies contributed equally to this work.
Address for reprint requests and other correspondence: R. Webb, Cardiff
phosphorylation status, and kinases such as MAPK/ERK,
School of Health Sciences, UWIC, Cardiff CF5 2YB, UK (e-mail: rwebb AMP-activated protein kinase, and others have been associated
@uwic.ac.uk). with ligand-independent changes in PPAR␥ activity (6).
806 8750-7587/12 Copyright © 2012 the American Physiological Society http://www.jappl.org
 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM 807
With specific regard to monocytes, PPAR␥ has been re- Participant Recruitment and Exercise Procedures
ported to regulate inflammatory responses and atherogenesis The present study comprised two phases: 1) an initial analysis of
via this cell type (20, 39, 46). Importantly, because of differen- the response to an acute exercise bout (including a 24-h follow-up)
tiation of monocytes into macrophages and their subsequent and 2) an 8-wk training program comprising multiple exercise bouts
migration into a wide variety of tissues, activation of PPAR␥ (Table 1).
within monocytes/macrophages has been shown to be crucial to Recruitment. In phase 1, a cohort of healthy active untrained
many systemic thiazolidinedione antidiabetic actions throughout individuals (n ⫽ 9, 32 ⫾ 8 yr old, 179 ⫾ 11 cm height, 80 ⫾ 11 kg
the body (18, 20, 40). In particular, upregulation, within mono- body mass) was recruited. In phase 2, a cohort of healthy active
cytes/macrophages, of PPAR␥-regulated genes involved in untrained individuals (n ⫽ 8, 27.8 ⫾ 6.4 yr old, 175 ⫾ 7 cm height,
lipid uptake and clearance and reverse cholesterol transport has 67 ⫾ 6 kg body mass) was recruited. In both cases, institutional
ethical approval was granted; all participants signed confidential
been shown to promote lipid clearance and transport to the informed consent, health activity questionnaire, and blood collection
liver within HDL-cholesterol (5, 9, 29). forms prior to commencing their involvement.
However, thiazolidinediones such as rosiglitazone have been Maximal testing. For phases 1 and 2, all participants completed an
linked to negative patient outcomes; specifically, the publica- incremental cycling test to exhaustion on a Monark 824E cycle
tion of meta- and teleoanalyses of clinical rosiglitazone studies ergometer, with 30-W increases in power output every 3 min until
(38, 47, 50) has raised safety concerns regarding the increased volitional exhaustion. Heart rate (Polar s810, Polar Electro) and blood
risk of myocardial infarction and death from cardiovascular lactate and expired gases (Jaeger Oxycon Delta, Erich Jaeger, Hoe-
causes in rosiglitazone-treated patients. Therefore, because of chberg, Germany) were analyzed continuously throughout the test. All
participants achieved at least two of the criteria for attainment of
the immense clinical importance of T2D and its related con-
maximal O2 consumption (V̇O2 max) (17). Individual plots of power
ditions in Western society, alternative means of inducing vs. submaximal O2 consumption were used to allow calculation of the
beneficial PPAR␥-dependent genomic effects without also power output corresponding to 70% V̇O2 max for each participant (176 ⫾
causing such deleterious effects are urgently required. 56 and 161 ⫾ 41 W for phases 1 and 2, respectively) and, in the case
Exercise may provide such an alternative: we and others of phase 2, to establish submaximal training intensities for the
have demonstrated that participation in exercise programs can subsequent training regimen.
affect PPAR␥-mediated signaling events to bring about anti- Training procedures and submaximal testing. In phase 1, a subse-
atherogenic benefits over and above improvements in cardio- quent single 45-min bout of exercise at 70% V̇O2 max was performed.
vascular performance (7, 56, 57). For example, sedentary Blood and saliva samples were collected before and after exercise;
additional samples were collected 1.5, 3, and 24 h after completion of
participants in a low-intensity exercise program (walking the exercise bout. In phase 2, participants were assigned to an 8-wk
10,000 steps 3 times per week for 8 wk) exhibited increased training program comprising regular supervised training sessions, also
monocyte PPAR␥ DNA-binding activity and enhanced expres- with the same cycle ergometer and heart rate analysis (Table 1). The
sion within monocytes of the PPAR␥ target genes CD36, program was designed to be progressive for training volume (17). At
PPAR␥ coactivator-1␣ (PGC-1␣), and liver X receptor-␣ weeks 0, 4, and 8 of the training program, participants performed a
(LXR␣) and, consequently, upregulation of the LXR␣ target constant-load 45-min submaximal trial at a constant absolute intensity
genes ATP-binding cassette (ABC) subfamily A member 1 that was equivalent to 70% V̇O2 max, as calculated from the initial
(ABCA1) and ABC subfamily G member 1 (7). In these maximal test (161 ⫾ 41 W; sessions 1, 11, and 25 in Table 1). This
subjects, beneficial alterations were observed in serum lipid
profiles (increased HDL-cholesterol and decreased total cho- Table 1. Summary of the 8-wk training program
lesterol, LDL-cholesterol, and triglycerides), with increased
HDL-cholesterol correlating with PPAR␥ target gene expres- Week Session No. Regimen
sion, indicating that exercise had enhanced reverse cholesterol 0 1 1 ⫻ (45 min at 70% maximal O2 uptake)
transport via PPAR␥-mediated effects within monocytes (7). 1 2 2 ⫻ (25 min at 65% maximal O2 uptake; 3 min recovery)
Similarly, participation in exercise also exerted anti-inflamma- 3 3 ⫻ (10 min at 80% maximal O2 uptake; 4 min recovery)
tory effects due to upregulation of PGC-1␣ and priming of 2 4 1 ⫻ (60 min at 60% maximal O2 uptake)
5 4 ⫻ (10 min at 80% maximal O2 uptake; 3 min recovery)
monocytes for differentiation into the alternative M2 macro- 3 6 1 ⫻ (50 min at 65% maximal O2 uptake)
phage phenotype (57). 7 5 ⫻ (8 min at 85% maximal O2 uptake; 4 min recovery)
As the activity of PPAR␥ is primarily governed by the 8 2 ⫻ (20 min at 70% maximal O2 uptake; 3 min recovery)
generation and/or availability of its ligands (44, 51), it is 4 9 2 ⫻ (20 min at 75% maximal O2 uptake; 3 min recovery)
10 5 ⫻ (6 min at 85% maximal O2 uptake; 3 min recovery)
possible that the above-mentioned effects were due to gener- 11 1 ⫻ (45 min at 70% maximal O2 uptake)
ation of PPAR␥ ligands during exercise. Therefore, we aimed 5 12 1 ⫻ (60 min at 70% maximal O2 uptake)
to test the hypotheses that exercise is associated with genera- 13 4 ⫻ (10 min at 85% maximal O2 uptake; 4 min recovery)
tion of PPAR␥ ligands in the plasma and that this may activate 14 3 ⫻ (15 min at 80% maximal O2 uptake; 3 min recovery)
6 15 1 ⫻ (60 min at 65% maximal O2 uptake)
PPAR␥ signaling within circulating monocytes, thus providing 16 2 ⫻ (15 min at 85% maximal O2 uptake; 4 min recovery)
a mechanism to underpin the exercise-induced antiatherogenic 17 3 ⫻ (15 min at 80% maximal O2 uptake; 2 min recovery)
benefits observed in previous studies (7, 56, 57). 7 18 1 ⫻ (45 min at 65% maximal O2 uptake)
19 3 ⫻ (15 min at 80% maximal O2 uptake; 3 min recovery)
20 2 ⫻ (25 min at 80% maximal O2 uptake; 2 min recovery)
MATERIALS AND METHODS 21 1 ⫻ (50 min at 70% maximal O2 uptake)
8 22 1 ⫻ (60 min at 65% maximal O2 uptake)
Materials 23 4 ⫻ (10 min at 85% maximal O2 uptake; 4 min recovery)
24 2 ⫻ (25 min at 80% maximal O2 uptake; 2 min recovery)
Reagents were obtained from Sigma-Aldrich (Poole, UK) unless
25 1 ⫻ (45 min at 70% maximal O2 uptake)
stated otherwise.

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

808 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM

trial was performed under fasting conditions [i.e., tests were per- trations of 1, 3, and 5 ␮M, respectively, while plasma samples
formed at 9 AM following an overnight (ⱖ14-h) fast]. Venous blood (normalized for protein content) were added at 10% (vol/vol) to
and saliva samples were taken immediately before (Pre) and after DMEM (instead of the addition of 10% FCS). Cells were harvested
(Post) this 45-min trial, and blood lactate, heart rate, and expired gases and lysed 24 h later, and luminescence was determined using a Dual
were analyzed as previously described. On completion of the training Reporter Assay (Promega) according to the manufacturer’s instruc-
program, using the procedures described above, all participants re- tions. In each case, the ratio of luciferase to Renilla luminescence was
peated the maximal test. used to normalize luminescence values to transfection efficiency. For
Isolation of human peripheral mononuclear cells and plasma investigation of the extent to which treatment with plasma samples
samples from whole blood. Blood samples from the antecubital vein was associated with phosphorylation of PPAR␥, phospho-specific
were obtained via phlebotomy at the time points described above. Western blot experiments (see below) were carried out as previously
Heparinized blood (10 ml) was diluted 1:1 in RPMI medium, layered described (8); PPAR␥ was detected in HEK-293 total-protein extracts
over 10 ml of Histopaque-1077 Ficoll-Hypaque, and centrifuged at as a doublet of ⬃55-kDa (nonphosphorylated PPAR␥) and ⬃60-kDa
400 g for 20 min. (phosphorylated PPAR␥) bands.
Plasma samples were aspirated off as supernatants and stored at Isolation of RNA and RT-PCR assays. Total RNA was extracted
⫺80°C for subsequent analysis; an Instrumentation Laboratories IL- using a RiboPure-Blood kit (Ambion, Huntingdon, UK) according to the
300 analyzer (Diamond Diagnostics, Holliston, MA) was used to manufacturer’s instructions. Briefly, 5 ⫻ 106 mononuclear cells or
determine triglyceride, total cholesterol, and HDL-cholesterol levels monocytes were washed in ice-cold PBS and lysed, and RNA was
in the plasma samples. LDL-cholesterol was calculated using the extracted using acid phenol-chloroform extraction, ethanol precipitation,
Friedewald equation. All lipid analyses were carried out at the and resuspension in RNase-free water. RNA was quantified and checked
Diabetes Research Network Cymru laboratories (Swansea, UK). for purity using the ratio of its absorbance at 260 nm to its absorbance at
Mononuclear cell fractions were carefully removed from the Ficoll- 280 nm (only samples with a ratio ⬎1.8 were deemed suitable for use).
Hypaque interface and washed four times (500 g for 10 min) in 0.4 ml RNA samples were stored at ⫺80°C before conversion to cDNA using
of phosphatase inhibitor solution (Active Motif, Rixensart, Belgium) the High-Capacity cDNA Archive Kit (Applied Biosystems, Warrington,
and 7.6 ml of PBS. For validation purposes, direct-labeling monocyte UK) and then stored at ⫺20°C. PPAR␥, CD36, LXR␣, ABCA1, PGC-
purification experiments were carried out in some cases. Mononu- 1␣, cholesteryl ester transferase protein (CETP), lecithin-cholesteryl
clear cells were incubated with CD14-conjugated magnetic mi- acyltransferase (L-CAT), and ApoA1 mRNA expression were analyzed
crobeads [Miltenyi Biotec, Surrey, UK; 1:5 (vol/vol) at 4°C for 15 using Fast SYBR Green Master Mix (Applied Biosystems) and compared
min]. Excess microbeads were removed by centrifugation (300 g with expression of GAPDH; primers with the following sequences were
for 10 min) and removal of the resulting supernatant. Cells were designed using Primer Express software (version 2.0, Applied Biosys-
passed through a 30-␮m-pore filter and added to a separator tems): 5=-GGAAGTGATGATGAACAGCAGC-3= and 5=-GAGACT-
column (MACS LS, Miltenyi Biotec) attached to a magnetic
GTGTTGTCCTCAGCG-3= (CD36), 5=-CGCACTACATCTGCCA-
separator unit (QUADROMACS). The column was washed to
CAGT-3= and 5=-TGAGGCGGATCTGTTCTTCT-3= (LXR␣), 5=-
remove unbound cells and removed from the magnetic separator
GCACTGAGGAAGATGCTGAAA-3= and 5=-AGTTCCTGGAA-
unit, and cells were forced through the column with a syringe
GGTCTTGTTCA-3= (ABCA1), 5=-CGTGGCCGCAGATTTGAA-3=
plunger to elute labeled cells (i.e., CD14⫹ monocytes).
and 5=-CTTCCATTACGGAGAGATCCAC-3= (PPAR␥), 5=-TCAGTC-
Measurements of plasma levels of reduced glutathione and salivary
CTCACTGGGGGACA-3= and 5=-TGCTTCGTCAAAAACAG-3=
levels of IgA. Plasma reduced glutathione ([GSH]plasma) and salivary
(PGC-1␣), 5=-CGGCTCGGGCCACTTACACA-3= and 5=-GAGCAG-
IgA (sIgA) levels were measured as described previously (32, 33) via
commercially available ELISA using anti-GSH capture antibodies and GCATGGGCATTGCC-3= (CETP), 5=-TGCTCTATTTCCTGCT-
horseradish peroxidase (HRP)-conjugated secondary antibodies for GCGCCAG-3= and 5=-ATGGGGATGCCCTGGTTGTCAC-3= (L-
[GSH]plasma or anti-IgA capture antibodies and HRP-conjugated sec- CAT), 5=-CCCCCTACAGCGACGAGCTG-3= and 5=-GGTGGCCT-
ondary antibodies for sIgA [Calbiochem Biosciences (Merck), Not- TGGCGTGGTACT-3= (ApoA1), and 5=-TCCTGTGGCATCCA-
tingham, UK]. Saliva samples were collected from a salivette con- CGAA-3= and 5=-GAAGCATTTGCGGTGGAC-3= (GAPDH). Thermo-
taining a cotton pad placed under the tongue for 2 min, and the cotton cycling was carried out as follows: initial denaturation for 10 min at 95°C,
pad was centrifuged (300 g for 2 min) to yield liquid saliva. As followed by 40 cycles of 15 s at 95°C and 60 s at 60°C. Relative
reported previously (32, 33), all sIgA data were normalized for quantification of genes of interest was calculated using the 2⫺⌬⌬CT
hydration status by normalization to salivary protein content [esti- formula, in which ⌬CT is the difference between the cycle threshold (CT)
mated using a Bio-Rad protein assay (Bio-Rad Laboratories, Basing- value for the gene of interest and CT value for the housekeeping gene.
stoke, UK)]. Data were included only if plots of log(RNA) vs. ⌬CT resulted in slopes
Use of plasma samples in PPRE-luciferase reporter assays. Human between ⫺0.1 and ⫹0.1, meaning that amplicon efficiencies were ap-
embryonic kidney (HEK-293) cells were maintained under standard proximately equal, and if plots of log(template) vs. CT resulted in slopes
conditions in DMEM (Invitrogen, Glasgow, UK) supplemented with of approximately ⫺3.3 (for the housekeeping gene and the gene of
4 mM glutamine, 10% (vol/vol) FCS, and 100 IU of penicillin- interest; an ideal slope of ⫺3.32 represents 100% PCR efficiency).
streptomycin. Cells were transiently transfected in 96-well plates Western blot experiments. Western blot experiments were carried
(seeded at a density of 2 ⫻ 104 cells per well) with 100 ng each of out as described elsewhere (32, 33). Briefly, total protein extracts from
PPRE-luciferase reporter construct, PPAR␥ expression vector, and mononuclear cells were prepared by treatment with 100 ␮l of protein
Renilla reporter plasmid (pR6TK, Promega, Southampton, UK) using extraction/lysis buffer containing 1 mM protease inhibitor cocktail
Lipofectamine (Invitrogen) according to the manufacturers’ instruc- and 1 mg/ml phosphatase inhibitor (Active Motif), and protein content
tions. The PPAR␥ expression vector was made using a full-length was estimated using a protein assay (Bio-Rad Laboratories). Samples
copy of PPAR␥ cDNA cloned into pcDNA3 expression vector (In- were normalized for protein content (50 ␮g of protein in each case)
vitrogen). The plasmids-Lipofectamine cocktail in OptiMEM solution and subjected to SDS-PAGE, transferred to nitrocellulose, and probed
(Invitrogen) was incubated at room temperature for 30 min prior to with anti-CD36 or anti-PPAR␥ primary antibody (16 h, 1:1,000
addition to the cells. At 24 h after transfection, the medium was replaced dilution; Cell Signaling Technology, Danvers, MA) and then with
with fresh DMEM, and cells were exposed to rosiglitazone (GlaxoSmith- HRP-labeled anti-rabbit IgG antibody (2 h, 1:2,000 dilution; Cell
Kline, Uxbridge, UK), 15-deoxy-⌬12,14-PGJ2 (15dPGJ2; Calbiochem, Signaling Technology). Proteins of interest were detected via en-
Darmstadt, Germany), and 13-hydroxy-9,11-octadecadienoic acid (13- hanced chemiluminescence as 88-kDa (CD36) or 55- to 60-kDa
HODE), which were solubilized in DMSO and used at working concen- (PPAR␥) immunogenic bands on Western blots. Band intensities were

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM 809
determined and quantitated using AC-1 BioImaging and VisionWorks [GSH]plasma in untrained participants (week 0 Post ⫽ 68.1 ⫾
LS Software systems (UltraViolet Products, Cambridge, UK). 19.1% of week 0 Pre, P ⫽ 0.004; Fig. 1D). However, after 4 wk
Statistical Methods of training, this exercise-associated decrease in [GSH]plasma
was blunted (week 4 Post ⫽ 76.4 ⫾ 13.9% of week 4 Pre, P ⫽
Values are means ⫾ SE. Multiple comparisons were performed 0.016), and after 8 wk of training, there was no significant
using one-way ANOVA, with post hoc analysis performed using change in [GSH]plasma following an acute bout of exercise
Tukey’s method; for nonmultiple comparisons, paired t-tests were (week 8 Post ⫽ 101.4 ⫾ 13.7% of week 8 Pre, P ⫽ 0.846).
used. Significance levels were set at P ⬍ 0.05. Thus, the data shown in Fig. 1 support the hypothesis that
several systemic parameters, including physiological (heart
RESULTS
rate), metabolic (lactate production), immunological (sIgA
On completion of the 8-wk training study, the participants production), and biochemical (consumption of [GSH]plasma,
showed significantly lower heart rate (175 ⫾ 14 and 156 ⫾ 15 which reduces blood-borne reactive oxygen species to harm-
beats/min at weeks 0 and 8, respectively, P ⫽ 0.002; Fig. 1A) less forms while being converted to oxidized glutathione),
and blood lactate (5.9 ⫾ 2.8 and 3.2 ⫾ 1.4 mM at weeks 0 and were involved in training adaptations to the 8-wk exercise
8, respectively, P ⫽ 0.02; Fig. 1B) responses at the end of the program.
fixed-intensity 45-min trial, suggesting that the program was To determine whether exercise was associated with genera-
effective at promoting a training response in submaximal tion of blood-borne PPAR␥ ligands, PPRE-luciferase gene
exercise. Similarly, a borderline-significant increase in V̇O2 max reporter assays were performed. As expected, treatment of
was observed (41.7 ⫾ 8.5 and 43.4 ⫾ 6.8 ml·kg⫺1·min⫺1 at HEK-293 cells with the synthetic PPAR␥ ligand rosiglitazone
weeks 0 and 8, respectively, P ⫽ 0.08). (1 ␮M for 24 h) induced strong PPRE-luciferase activity in
sIgA significantly decreased to 53.2 ⫾ 12.7% of basal cells that had been transfected with a PPAR␥ expression vector
immediately postexercise at week 0 (P ⬍ 0.05); however, by [3.41 ⫾ 0.30 and 10.52 ⫾ 0.51 relative light units (RLU) for
week 4, the exercise-associated decrease in sIgA was nonsig- control and rosiglitazone-treated cells, respectively, P ⬍ 0.05;
nificant (77.2 ⫾ 17.4% of basal, P ⫽ 0.232), and by week 8, the Fig. 2A]. In contrast, PPRE-luciferase activity was negligible
sIgA postexercise response was further diminished (88.1 ⫾ in the presence or absence of rosiglitazone in cells not trans-
18.8% of basal, P ⫽ 0.550; Fig. 1C). Similarly, the same acute fected with the PPAR␥ expression vector (2.04 ⫾ 0.09 and
bout of exercise was associated with a significant decrease in 2.12 ⫾ 0.09 RLU for control non-PPAR␥-transfected cells and

Fig. 1. Training adaptations: responses to a 45-min bout of cycling at 70% of maximal O2 uptake (V̇O2 max), as observed during an 8-wk exercise program.
A: changes in capillary blood lactate concentration ([lactate]) following the exercise bout. B: mean heart rate [beats/min (bpm)] during the exercise bout. C: changes in
salivary IgA (sIgA) levels following the exercise bout. Open bars, plasma samples obtained before exercise (Pre); gray bars, plasma samples obtained after exercise
(Post). D: changes in plasma reduced glutathione ([GSH]plasma) following the exercise bout. Values are means ⫾ SE (n ⫽ 8 in all cases). *P ⬍ 0.05.

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

810 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM

Fig. 2. Peroxisome proliferator-activated receptor (PPAR) response element (PPRE)-luciferase activity and PPAR␥ phosphorylation, as determined in human
embryonic kidney (HEK-293T) cells treated with plasma samples from exercising participants. PPRE-luciferase activity was determined in HEK-293T cells [with
and without (⫹ and ⫺) PPAR␥ transfection] treated for 24 h with 1 ␮M rosiglitazone, 3 ␮M 15-deoxy-⌬12,14-PGJ2 (15d-PGJ2), 5 ␮M 13-hydroxy-9,11-
octadecadienoic acid (13-HODE), or plasma samples [10% (vol/vol)] obtained before (Pre) and after (Post) exercise (A) or Pre (open bars) and Post (filled bars)
plasma samples obtained at weeks 0, 4, and 8 of an exercise program (B). RLU, relative light units. C and D: representative images and densitometric data from
Western blot experiments in which total protein extracts from plasma-treated HEK-293T cells were probed for phosphorylated PPAR␥ (pPPAR␥) and
nonphosphorylated PPAR␥ (⬃60 and ⬃55 kDa, respectively). Values are means ⫾ SE (n ⫽ 8). *P ⬍ 0.05.

rosiglitazone-treated non-PPAR␥-transfected cells, respec- with Pre or Post exercise plasma samples did not induce
tively, P ⬍ 0.05 vs. respective PPAR␥-transfected samples; significant differences in the pPPAR␥-to-total PPAR␥ ratio.
Fig. 2A), indicating that this cell type does not express endog- For participants in phase 1, samples were taken before (Pre)
enous PPAR␥. Treatment with the natural PPAR␥ ligands and immediately after (Post) exercise and also at 1.5, 3, and 24
15dPGJ2 (14) and 13-HODE (37, 55) also induced increases in h after exercise. As with week 0 of the 8-wk program (see
PPRE-luciferase activity (9.90 ⫾ 1.06 and 4.80 ⫾ 0.35 RLU above) and as shown in our previous studies (32, 33), this bout
for 15d-PGJ2 and 13-HODE, respectively, P ⬍ 0.05 vs. control of exercise was associated with significant decreases in sIgA
in both cases), albeit less strongly than rosiglitazone. and [GSH]plasma (data not shown, P ⬍ 0.05 in both cases). The
Treatment of PPAR␥-expressing HEK-293 cells with week 0 extended 24-h follow-up period in this experiment revealed
plasma samples [10% (vol/vol) for 24 h] showed increased downstream signaling consequences of participation in exer-
PPAR␥/PPRE-luciferase activity; importantly, this increase cise. Significant increases (1.5- to 2.5-fold) in expression of the
was significantly greater when cells were treated with plasma monocyte PPAR␥-regulated genes CD36, LXR␣, and ABCA1
collected following the initial exercise bout (week 0 Post ⫽ (P ⬍ 0.05 in all cases) were observed in samples taken within
123.4 ⫾ 9.7% of week 0 Pre, P ⬍ 0.05; Fig. 2B). In contrast, 3 h of exercise (Fig. 3), with gene expression returning to basal
PPRE-luciferase activity was negligible in plasma-treated cells levels within 24 h of exercise. Gene expression of PPAR␥
not transfected with the PPAR␥ expression vector (1.16 ⫾ 0.09 itself, and also of PGC-1␣ [which has been reported to be
RLU, P ⬍ 0.05 vs. respective PPAR␥-transfected samples; encoded by a PPRE-bearing PPAR␥ target gene (22)], fol-
Fig. 2A). Thus, these data suggest that PPAR␥ ligands were lowed a similar trend, with nonsignificant increases at 1.5 and
generated in the plasma as a result of exercise. Interestingly, 3 h. In contrast, analysis of expression of reverse cholesterol
this effect was blunted in weeks 4 and 8 of the exercise transport genes, the expression of which is not reported to be
program (week 4 Post ⫽ 118.2 ⫾ 13.6% of week 4 Pre and increased by PPAR␥ (11, 15, 35), showed that L-CAT (Fig.
week 8 Post ⫽ 109.5 ⫾ 9.6% of week 8 Pre, P ⬎ 0.05 in both 3F) and CETP (data not shown) did not significantly increase,
cases; Fig. 2B). while ApoA1 mRNA could not be detected (suggesting that
Western blotting was used to assess PPAR␥ expression and this gene is not expressed by monocytes or, at least, that its
phosphorylation (8). While PPAR␥ could not be detected in expression levels are below the sensitivity of the assay systems
non-PPAR␥-transfected control samples (data not shown), a used in the present study). The lack of response of these
doublet of ⬃55-kDa (nonphosphorylated PPAR␥) and ⬃60- non-PPAR␥ target genes indicates that the effects in CD36,
kDa [phosphorylated PPAR␥ (pPPAR␥)] bands was detected LXR␣, and ABCA1 (and possibly also PPAR␥ and PGC-1␣)
in all other samples (Fig. 2C). The relative intensities of these were specifically induced by exercise-triggered PPAR␥ signal-
two bands indicate that PPAR␥ was in a predominantly non- ing, rather than as part of a generalized response to exercise.
phosphorylated state in all samples (pPPAR␥-to-total PPAR␥ Although the data shown in Fig. 3A are from mononuclear
ratio ⫽ 0.201, 0.102 ⫾ 0.053, and 0.162 ⫾ 0.161 for control, cells, optimization experiments using CD14⫹ selection mono-
Pre, and Post, respectively; Fig. 2D). Importantly, treatment cyte purification columns (Miltenyi Biotec) showed that

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM 811

Fig. 3. Effect of exercise on gene expression in mononuclear cells. Peripheral mononuclear cell samples were obtained from a cohort of exercising (45 min of
cycling at 70% of V̇O2 max) participants. Semiquantitative determination of mononuclear cell mRNA levels of PPAR␥ (A), CD36 (B), liver X receptor-␣ (LXR␣,
C), ATP-binding cassette subfamily A member 1 (ABCA1, D), PPAR␥ coactivator-1␣ (PGC-1␣, E), and lecithin-cholesteryl acyltransferase (L-CAT, F) was
carried out using RT-PCR. Values are means ⫾ SE (n ⫽ 9). *P ⬍ 0.05.

mRNA expression for the genes of interest in the present study expressed in monocytes, rather than in other types of mono-
was enriched by two- to sixfold in purified monocytes com- nuclear cells.
pared with mononuclear cell samples (5.71 ⫾ 1.09, 2.65 ⫾ Western blotting was used to determine expression of
0.16, 2.21 ⫾ 0.58, and 4.62 ⫾ 0.45 for PPAR␥, CD36, PPAR␥ and its target gene CD36 at the protein level in
ABCA1, and PGC-1␣, respectively). Importantly, differential mononuclear cells (week 0 Pre and week 8 Pre) in participants
expression in monocytes was comparable or greater for each of in phase 2. As shown in Fig. 4, although PPAR␥ expression
these genes than the “classical” monocyte marker gene CD14 underwent only nonsignificant increases (1.50 ⫾ 0.71 at week
(2.77 ⫾ 0.47 fold), suggesting that they are predominantly 8 compared with week 0, P ⬎ 0.05), CD36 expression was

Fig. 4. Effect of exercise on CD36 and

PPAR␥ expression in mononuclear cells.
Total protein extracts from peripheral mono-
nuclear cells were obtained from a cohort of
exercising participants on commencement
(week 0) and termination (week 8) of an 8-wk
exercise program and subjected to Western
blot experiments. A: densitometric data show-
ing overall expression levels of PPAR␥ (open
bars) and CD36 (filled bars). B: representative
image from CD36 blot. C: representative im-
age from PPAR␥ blot. D: densitometric data
showing PPAR␥ phosphorylation levels. Val-
ues are means ⫾ SE (n ⫽ 8). *P ⬍ 0.05.

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

812 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM

significantly increased at rest (2.79 ⫾ 0.61 at week 8 compared PPAR␥ activation in circulating monocytes and, hence, up-
with week 0, P ⬍ 0.05; Fig. 4B). No difference in PPAR␥ regulation of PPAR␥ target genes (Figs. 3 and 4). As men-
phosphorylation was seen between week 0 and week 8 tioned above, it is possible that the upregulation of PPAR␥-
(pPPAR␥-to-total PPAR␥ ratio ⫽ 0.231 ⫾ 0.057 and 0.242 ⫾ regulated genes in the present study may be due to ligand-
0.118, respectively; Fig. 4, C and D). Importantly, the week 8 independent alteration of PPAR␥ transcriptional activity.
samples were taken ⬎48 h after the previous bout of exercise, Several studies have demonstrated that receptor-mediated
suggesting that any increases in expression were not associated events may lead to changes in PPAR␥’s phosphorylation status
with a single exercise bout but, rather, constituted sustained and, thence, to its activation or inactivation (6, 23); for exam-
increases that occurred as a consequence of regular participa- ple, exercise-associated alterations in blood-borne levels of
tion in exercise (i.e., 8 wk of adherence to a program compris- insulin and/or in the ability of insulin to bring about its
ing regular, repeated exercise; Table 1). downstream effects (e.g., the impact of exercise-induced in-
Finally, blood lipids were determined for each participant: at flammation on insulin receptor substrate-1 tyrosine phosphor-
baseline (week 0 Pre), total cholesterol was 4.31 ⫾ 0.20 mM, ylation and, thus, the extent to which insulin receptor occu-
triglycerides were 1.53 ⫾ 0.19 mM, HDL-cholesterol was 1.33 ⫾ pancy is transduced to downstream kinase activity) have been
0.09 mM, and LDL-cholesterol was 2.29 ⫾ 0.17 mM. Thus, all reported (reviewed in Ref. 26). Also, it is possible that exercise
data fell within “healthy/low-risk” categories [American Heart may be associated with a physical change that could impact
Association reference ranges (3)] for each parameter. A single PPAR␥; for example, the rate of blood flow has been reported
submaximal standardized exercise bout did not elicit signifi- to influence PPAR␥ activation via ERK isoforms in endothelial
cant changes in blood lipids; also, when comparisons were cells (1), an effect that may be mirrored in circulating mono-
made between blood lipid levels over the duration of the cytes during exercise. Importantly, the above-mentioned ef-
exercise program, significant decreases in triglycerides (68.0 ⫾ fects could lead to monocyte PPAR␥ activation that is not
13.4%, P ⬍ 0.05) and borderline-significant decreases in total dependent on generation of PPAR␥ ligands. However, phos-
cholesterol (86.6 ⫾ 17.0%, P ⬍ 0.10), but no significant pho-specific Western blot experiments (8) suggest that treat-
changes in LDL- or HDL-cholesterol, were seen. ment of PPAR␥-expressing HEK cells with plasma from ex-
ercising participants did not increase PPAR␥ phosphorylation
DISCUSSION and, also, that there was no difference between “preexercise”
and “postexercise” samples with regard to PPAR␥ phosphor-
We present data that support the widely held view that ylation (Fig. 2, C and D).
exercise is associated with beneficial effects, and we demon- It should be stressed that HEK cells may not express all the
strate several monocyte-associated exercise-induced PPAR␥ signaling agents that might influence PPAR␥ phosphorylation;
signaling events. The aim of the present study was to elucidate therefore, this experimental system may not accurately reflect
the PPAR␥ signaling events triggered by exercise and, partic- the in vivo situation in all respects. Nevertheless, we conclude
ularly, by regular exercise training, for example, the down- that while exercise-induced PPAR␥ activation may involve a
stream effects in terms of gene expression patterns within ligand-independent component in vivo that we could not detect
monocytes and their likely accompanying functional conse- in our experimental system, the specific exercise-associated
quences with regard to antiatherogenic and/or anti-inflamma- effects in Fig. 2 are likely to be due to the generation of ligands
tory effects. in the plasma of exercising individuals and subsequent ligand-
It is well established that, in skeletal muscle, exercise up- dependent PPAR␥ activation within plasma-treated HEK cells.
regulates PPAR␥-controlled genes to promote mitochondrial Several lipid-derived agents have been identified as PPAR␥
biogenesis, aerobic respiration, and other exercise-triggered ligands [e.g., prostaglandins such as 15dPGJ2 (14) and unsat-
benefits. For example, exercise leads to increases in PPAR␥ urated fatty acids such as 13-HODE (37, 55)], and as exercise
activity within skeletal muscle (42), while comparable ⬃2.5- is associated with changes in lipid metabolism (28), it seems
fold increases in PPAR␥ expression and, also, expression of plausible that novel lipid metabolites could be generated within
the PPAR␥ coactivator PGC-1␣ were seen in two separate the blood of exercising participants [perhaps due to release of
studies 3 h after a single bout of exercise (16, 30). Moreover, the products of cyclooxygenase and/or lipoxygenase-catalyzed
a transcriptional map of the effects of exercise on skeletal reactions from circulating leukocytes (24)]. In this context,
muscle has recently been published, and many PGC-1␣/ treatment with 15dPGJ2 or 13-HODE induced significant in-
PPAR␥-regulated genes (e.g., CD36) were among those up- creases in PPRE-luciferase activity in our reporter gene assay
regulated (25). However, if exercise is a systemic phenome- system (Fig. 2A). However, precise identification of the puta-
non, its effects will not be limited to a single tissue; therefore, tive exercise-generated PPAR␥ ligands is beyond the scope of
we investigated whether parallel exercise-associated benefits the present study.
may potentially occur in monocytes via exercise-induced up- Previous reports have linked upregulation of PPRE-bearing
regulation of PPAR␥-controlled monocytic genes. monocytic genes with a variety of beneficial functional effects.
In contrast to skeletal muscle, exercise-related signaling For example, LXR␣, a PPRE-bearing transcription factor an-
events in monocytes have not been so intensively studied. To alyzed in the present study, controls expression of a number of
our knowledge, this study is novel, in that it demonstrates that LXR␣ target genes that are involved in fatty acid metabolism
exercise is associated with increased levels of PPAR␥ ligands and cholesterol homeostasis [e.g., ABCA1, which promotes the
in the plasma and that PPAR␥ signaling can be triggered by safe excretion of cholesterol and LDL by formation of Apo-I
exercise to exert beneficial effects in circulating monocytes. In and HDL particles (9, 29)]. Moreover, upregulation of the
light of the data presented in Fig. 2, we propose that generation scavenger receptor CD36, another PPAR␥-regulated gene,
of PPAR␥ ligands in the blood during exercise may lead to leads to increased import of oxidized LDL; processing of

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM 813
oxidized LDL-derived cholesterol and its LXR␣-mediated ex- profiles in studies carried out on hypercholesterolemic popu-
port within HDL particles may provide an additional link lations (31). While we acknowledge that this issue and others,
between PPAR␥ activation and improvements in lipid profiles such as the nature and small size of the cohort (and the
(7, 9). Finally, the observation that PPAR␥ activation in accompanying data variability), are limitations to the present
monocytes is linked to priming for differentiation within tis- study, ongoing studies within our research group are investi-
sues into the M2 macrophage phenotype and also to anti- gating the effects of exercise on larger cohorts and cohorts of
inflammatory paracrine effects (4, 57) suggests that activation at-risk individuals with suboptimal lipid profiles. In particular,
of PPAR␥ within monocytes/macrophages may be linked to we aim to establish whether the PPRE-luciferase system re-
systemic anti-inflammatory actions (18). sponds similarly when treated with plasma samples from T2D
However, although parallel exercise-triggered effects may and non-T2D cohorts, given that similar exercise responses
occur in muscle and monocytes, several differences exist have been shown in T2D and non-T2D cohorts (albeit with
between the respective pathways. Specifically, PPAR␥ exerts regard to skeletal muscle, rather than monocytes) in previous
different effects through upregulation of a set of target genes in exercise studies (16, 43).
monocytes that is different from that in muscle (possibly due to It is well established that regular training influences a range
differences in DNA packaging and/or availabilities of tran- of inflammatory markers (12, 49, 53) in a manner sensitive to
scriptional coactivators in the respective tissues). However, exercise intensity and duration (13, 54). The standard response
PPAR␥ is involved in the systemic regulation and integration to any effective training program is that the same absolute
of lipid metabolism throughout the body (39). Hence, PPAR␥ intensity elicits a reduced physiological perturbation compared
carries out complementary functions in diverse tissues; for with the pretraining response. By presenting diminished
example, exercise-triggered PPAR␥ signaling is involved in postexercise increases in blood-borne PPAR␥ ligands over the
stimulation of adipocyte triglyceride hydrolysis, so as to boost course of the 8-wk program, alongside the expected blunting of
release of fatty acids into the bloodstream (52) and also to postexercise heart rate, lactate, sIgA, and [GSH]plasma re-
increase expression of genes linked to oxidative metabolism of sponses over the same period, the present work has provided an
fats within skeletal muscle (30). Thus, although different genes additional parameter by which a training program may reduce
are activated in different tissues, a complementary array of the relative stimulus of a given standardized acute exercise
effects is induced, and the systemic impact is broadly benefi- bout. These observations are entirely predictable, given the
cial. Our study’s identification of monocytes as an exercise- usual principles of training planning whereby ever-greater
responsive tissue, in which PPAR␥ can be activated to upregu- absolute intensities are needed to provide sufficient disruption
late genes such as LXR␣, CD36, PGC-1␣, and ABCA1, is in of homeostasis to allow for continued training adaptations (36,
line with this concept. 51). However, they are, to our knowledge, novel in their
In the present study, upregulation of monocyte-expressed application of this principle to exercise-induced PPAR␥ sig-
PPAR␥ target genes was seen at the mRNA level ⬍3 h after a naling. Importantly, given the widely acknowledged benefits of
single bout of exercise (Fig. 3), but this effect persisted for exercise-induced PPAR␥ signaling, they provide additional
⬍24 h. In contrast, after the 8-wk program, increases in support to the concept of a need to tailor exercise programs to
PPAR␥ target gene expression were seen at the protein level in participants’ fitness levels to elicit optimally beneficial effects
samples taken ⬎48 h after the previous bout of exercise, in each case. Thus, while relatively intense exercise was
suggesting that, in this case, the increase is not due to an acute required to activate PPAR␥ signaling in the cohort of fit, active
response to an exercise bout. Thus, the following question is individuals under investigation in the present study, we and
raised: can the effects of each individual exercise bout “co- others have reported that low/moderate-intensity exercise (i.e.,
alesce,” so that, after an extended program of regular frequent training programs based on brisk walking) is sufficient to
exercise, a more sustained effect is evident? Further experi- activate PPAR␥ in cohorts of sedentary individuals (7, 56 –58).
ments are needed to investigate this possibility more fully. It has been well established for several years that exercise is
In the present study, these PPAR␥ signaling effects were associated with a multitude of benefits, including the preven-
associated with decreases in total cholesterol and triglycerides, tion and/or treatment of metabolic disorders such as T2D (2).
but not with significant changes in HDL- or LDL-cholesterol. While the present study involved healthy participants and, thus,
Because of the intensity of the exercise program to be under- detected few overt improvements in lipid profiles, individuals
taken, the participants recruited into the present study were fit, with T2D exhibit decreased PPAR␥/PGC-1␣ (34, 41) and
active individuals. It was not surprising that they exhibited dyslipidemia (2). Therefore, if individuals with T2D were to
healthy/low-risk values with regard to all four of the lipid respond in a manner similar to the participants in the present
parameters studied (3); however, one may speculate that, in study, and T2D and non-T2D cohorts have been shown to
cohorts of individuals with suboptimal blood lipid profiles, respond to a similar extent to exercise (16, 43), it is possible
exercise-induced upregulation of monocytic lipid-handling that the exercise-associated effects seen here could help nor-
genes would function to bring lipid levels to within the healthy/ malize PPAR␥ signaling and, thus, improve the blood lipid
low-risk range. In previous studies where participants did not profiles of the individuals. Thus, the effects described here
exhibit healthy/low-risk values (3) for these parameters at could potentially contribute to prevention of the development
baseline, larger statistically significant HDL- and LDL-choles- of metabolic disorders such as T2D in healthy or at-risk
terol effects were seen (7, 56, 58); in our previous study, individuals or to normalization of aberrant metabolic function
exercise-induced changes in HDL-cholesterol significantly cor- in individuals with disorders such as T2D.
related with Pre vs. Post changes in PPAR␥ target gene mRNA In conclusion, the present study presents evidence that
expression (7), while exercise (sometimes in combination with exercise-associated generation of PPAR␥ ligands activates
dietary interventions) has been shown to normalize blood lipid PPAR␥ signaling events that result in upregulation of mono-

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

814 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM

cytic PPAR␥ target genes. This may constitute a novel bene- with impaired glucose tolerance or impaired fasting glucose: a randomised
ficial effect of exercise in the context of prevention and/or controlled trial. Lancet 368: 1096 –1105, 2006.
11. Feister HA, Auerbach BJ, Cole LA, Krause BR, Karathanasis SK.
treatment of metabolic disorders such as T2D and its compli- Identification of an IL-6 response element in the human LCAT promoter.
cations. J Lipid Res 43: 960 –970, 2002.
12. Fischer CP, Plomgaard P, Hansen AK, Pilegaard H, Saltin B, Peder-
ACKNOWLEDGMENTS sen BK. Endurance training reduces the contraction-induced interleukin-6
We gratefully acknowledge the excellent technical support of Gareth mRNA expression in human skeletal muscle. Am J Physiol Endocrinol
Walters, Steve Potter, Paul Jones, and Sean Duggan. We thank Dr. Marc van Metab 287: E1189 –E1194, 2004.
Bilsen (Maastricht University, Maastricht, The Netherlands) for the kind gift of 13. Fischer CP. Interleukin-6 and acute exercise and training: what is the
PPAR␥ and PPRE-luciferase expression vectors and Dr. S. Luzio and col- biological relevance? Exerc Immunol Rev 12: 6 –33, 2006.
leagues (Diabetes Research Network Cymru) for help with blood lipid analy- 14. Forman BM, Tontonoz P, Chen J, Brun RP, Spiegelman BM, Evans
ses. RM. 15-Deoxy-⌬12,14-prostaglandin J2 is a ligand for the adipocyte
determination factor PPAR-␥. Cell 83: 803–812, 1995.
GRANTS 15. Gauthier B, Robb M, McPherson R. Cholesteryl ester transfer protein
gene expression during differentiation of human preadipocytes to adi-
H. Moir and L. R. Butcher were recipients of University of Wales Institute pocytes in primary culture. Atherosclerosis 142: 301–307, 1999.
Cardiff Ph.D. studentships. N. A. Davies and J. S. Ruffino are recipients of 16. Gibala MJ, McGee SL, Garnham AP, Howlett KF, Snow RJ, Har-
Welsh Office of Research and Development Ph.D. studentships. greaves M. Brief intense interval exercise activates AMPK and p38
MAPK signaling and increases the expression of PGC-1␣ in human
DISCLOSURES skeletal muscle. J Appl Physiol 106: 929 –934, 2009.
17. Gore CJ. Physiological Tests for Elite Athletes: Australian Sports Com-
No conflicts of interest, financial or otherwise, are declared by the authors.
mission. Champaign, IL: Human Kinetics, 2000.
18. Hevener AL, Olefsky JM, Reichart D, Nguyen MT, Bandyopadyhay
AUTHOR CONTRIBUTIONS
G, Leung HY, Watt MJ, Benner C, Febbraio MA, Nguyen AK, Folian
A.W.T., N.A.D., H.M., L.W., L.R.B., M.G.H., and R.W. are responsible for B, Subramaniam S, Gonzalez FJ, Glass CK, Ricote M. Macrophage
conception and design of the research; A.W.T., N.A.D., H.M., L.W., J.S.R., PPAR␥ is required for normal skeletal muscle and hepatic insulin sensi-
S.A.I., L.R.B., and R.W. performed the experiments; A.W.T., N.A.D., L.W., tivity and full antidiabetic effects of thiazolidinediones. J Clin Invest 117:
J.S.R., S.A.I., L.R.B., M.G.H., K.M., and R.W. analyzed the data; A.W.T., 1658 –1669, 2007.
N.A.D., L.W., S.A.I., L.R.B., M.G.H., K.M., and R.W. interpreted the results 19. Hey-Mogensen M, Højlund K, Vind BF, Wang L, Dela F, Beck-
of the experiments; A.W.T., M.G.H., and R.W. prepared the figures; A.W.T., Nielsen H, Fernström M, Sahlin K. Effect of physical training on
N.A.D., M.G.H., K.M., and R.W. edited and revised the manuscript; A.W.T., mitochondrial respiration and reactive oxygen species release in skeletal
N.A.D., H.M., L.W., J.S.R., S.A.I., L.R.B., M.G.H., K.M., and R.W. approved muscle in patients with obesity and type 2 diabetes. Diabetologia 53:
the final version of the manuscript; R.W. drafted the manuscript. 1976 –1985, 2010.
20. Hodgkinson CP, Ye S. Microarray analysis of peroxisome proliferator-
REFERENCES activated receptor-␥ induced changes in gene expression in macrophages.
Biochem Biophys Res Commun 308: 505–510, 2003.
1. Akaike M, Che W, Marmarosh NL, Ohta S, Osawa M, Ding B, Berk 21. Holloszy JO, Booth FW. Biochemical adaptations to endurance exercise
BC, Yan C, Abe J. The hinge-helix 1 region of peroxisome proliferator- in muscle. Annu Rev Physiol 38: 273–291, 1976.
activated receptor-␥1 (PPAR␥1) mediates interaction with extracellular 22. Hondares E, Mora O, Yubero P, Rodriguez de la Concepción M,
signal-regulated kinase 5 and PPAR␥1 transcriptional activation: involve- Iglesias R, Giralt M, Villarroya F. Thiazolidinediones and rexinoids
ment in flow-induced PPAR␥ activation in endothelial cells. Mol Cell Biol induce peroxisome proliferator-activated receptor-coactivator (PGC)-1␣
24: 8691–8704, 2004. gene transcription: an autoregulatory loop controls PGC-1␣ expression in
2. American College of Sports Medicine and American Diabetes Associa- adipocytes via peroxisome proliferator-activated receptor-␥ coactivation.
tion. Exercise and type 2 diabetes: joint position statement. Med Sci Sports Endocrinology 147: 2829 –2838, 2006.
Exerc 42: 2282–2303, 2010.
23. Hu E, Kim JB, Sarraf P, Spiegelman BM. Inhibition of adipogenesis
3. American Heart Association. Guidelines for Detection of High Cholesterol
through MAP kinase-mediated phosphorylation of PPAR␥. Science 274:
(Online). National Institutes of Health National Cholesterol Education
2100 –2103, 1996.
Program. http://www.nhlbi.nih.gov/guidelines/cholesterol/atp3xsum [5
24. Huang JT, Welch JS, Ricote M, Binder CJ, Willson TM, Kelly C,
July 2011].
Witztum JL, Funk CD, Conrad D, Glass CK. Interleukin-4-dependent
4. Bouhlel MA, Derudas B, Rigamonti E, Dievart R, Brozek J, Haulon S,
production of PPAR-␥ ligands in macrophages by 12/15-lipoxygenase.
Zawdiski C, Jude B, Torpier G, Marx N, Staels B, Chinetti-Gbauguidi
G. PPAR␥ activation primes human monocytes into alternative M2 Nature 400: 378 –382, 1999.
macrophages with anti-inflammatory properties. Cell Metab 6: 137–143, 25. Keller P, Vollaard NB, Gustafsson T, Gallagher IJ, Sundberg CJ,
2007. Rankinen T, Britton SL, Bouchard C, Koch LG, Timmons JA. A
5. Brewer HB Jr, Remaley AT, Neufeld EB, Basso F, Joyce C. Regulation transcriptional map of the impact of endurance exercise training on
of plasma high-density lipoprotein levels by the ABCA1 transporter and skeletal muscle phenotype. J Appl Physiol 110: 46 –59, 2011.
the emerging role of high-density lipoprotein in the treatment of cardio- 26. Kirwan JP, del Aguila LF. Insulin signalling, exercise and cellular
vascular disease. Arterioscler Thromb Vasc Biol 24: 1755–1760, 2004. integrity. Biochem Soc Trans 31: 1281–1285, 2003.
6. Burns KA, Vanden Heuvel JP. Modulation of PPAR activity via phos- 27. Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson
phorylation. Biochim Biophys Acta 1771: 952–960, 2007. TM, Kliewer SA. An antidiabetic thiazolidinedione is a high affinity
7. Butcher L, Backx K, Roberts A, Thomas A, Webb R, Morris K. ligand for peroxisome proliferator-activated receptor-␥ (PPAR␥). J Biol
Low-intensity exercise exerts beneficial effects on plasma lipids via Chem 270: 12953–12956, 1995.
PPAR␥. Med Sci Sports Exerc 40: 1–7, 2008. 28. Lehmann R, Zhao X, Weigert C, Simon P, Fehrenbach E, Fritsche J,
8. Caddy J, Singh N, Atkin L, Ahluwalia M, Roberts AW, Lang D, Machann J, Schick F, Wang J, Hoene M, Schleicher ED, Häring HU,
Thomas AW, Webb R. Rosiglitazone transiently disturbs Ca2⫹ homeo- Xu G, Niess AM. Medium chain acylcarnitines dominate the metabolite
stasis in monocytic cells. Biochem Biophys Res Commun 366: 149 –155, pattern in humans under moderate intensity exercise and support lipid
2008. oxidation. PLos One 5: e11519, 2010.
9. Chinetti G, Lestavel S, Bocher V, et al. PPAR␣ and PPAR␥ activators 29. Li AC, Glass CK. PPAR- and LXR-dependent pathways controlling lipid
induce cholesterol removal from human macrophage foam cells through metabolism and the development of atherosclerosis. J Lipid Res 45:
stimulation of the ABCA1 pathway. Nat Med 7: 53–58, 2001. 2161–2173, 2004.
10. DREAM (Diabetes REduction Assessment with ramipril and rosigli- 30. Mahoney DJ, Parise G, Melov S, Safdar A, Tarnopolsky MA. Analysis
tazone Medication) Trial Investigators, Gerstein HC, Yusuf S, Bosch of global mRNA expression in human skeletal muscle during recovery
J, et al. Effect of rosiglitazone on the frequency of diabetes in patients from endurance exercise. FASEB J 19: 1498 –1500, 2005.

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org

 PPAR␥ SIGNALING, EXERCISE, MONOCYTE, LIPID METABOLISM 815
31. Marinangeli CP, Varady KA, Jones PJ. Plant sterols combined with drial function similarly in patients with type 2 diabetes and in control
exercise for the treatment of hypercholesterolemia: overview of indepen- individuals. Diabetologia 53: 1714 –1721, 2010.
dent and synergistic mechanisms of action. J Nutr Biochem 17: 217–224, 44. Plutzky J. The potential role of PPARs on inflammation in type 2 diabetes
2006. and atherosclerosis. Am J Cardiol 92: 34 –41, 2003.
32. Moir H, Butcher L, Jones KP, Hughes MG, Neale H, Jia H, Al-Ismaily 45. Radak Z, Chung HY, Koltai E, Taylor AW, Goto S. Exercise, oxidative
Z, Webb R. AMPK inactivation in mononuclear cells: a potential intra- stress and hormesis. Ageing Res Rev 7: 34 –42, 2007.
cellular mechanism for exercise-induced immunosuppression. Appl 46. Ricote M, Huang JT, Welch JS, Glass CK. The peroxisome proliferator-
Physiol Nutr Metab 33: 75–85, 2008. activated receptor (PPAR␥) as a regulator of monocyte/macrophage func-
33. Moir H, Hughes MG, Potter S, Sims C, Butcher LR, Davies NA, tion. J Leukoc Biol 66: 733–739, 1999.
Verheggen K, Jones KP, Thomas AW, Webb R. Exercise-induced 47. Rosen CJ. Revisiting the rosiglitazone story—lessons learned. N Engl J
immuno-suppression: the roles of reactive oxygen species and AMPK Med 363: 803–806, 2010.
dephosphorylation within immune cells. J Appl Physiol 108: 1284 –1292, 48. Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med 340:
2010. 115–126, 1999.
34. Moller DE. Potential role of TNF-␣ in the pathogenesis of insulin 49. Shimizu K, Kimura F, Akimoto T, Akama T, Tanabe K, Nishijima T,
resistance and type 2 diabetes. Trends Endocrinol Metab 11: 212–217, Kuno S, Kono I. Effect of moderate exercise training on T-helper cell
2000. subpopulations in elderly people. Exerc Immunol Rev 14: 24 –37, 2008.
35. Mooradian AD, Haas MJ, Chehade J, Wong NC. Apolipoprotein A-I 50. Singh S, Loke YK, Furberg CD. Thiazolidinediones and heart failure: a
expression in rats is not altered by troglitazone. Exp Biol Med (Maywood) teleo-analysis. Diabetes Care 30: 2148 –2153, 2007.
227: 1001–1005, 2002. 51. Smith SA. Central role of the adipocyte in the insulin-sensitizing and
36. Morton JP, Kayani AC, McArdle A, Drust B. The exercise-induced cardiovascular risk modifying actions of the thiazolidinediones. Biochimie
stress response of skeletal muscle, with specific emphasis on humans. 85: 1219 –1230, 2003.
Sports Med 39: 643–662, 2009.
52. Steinberg GR. Role of the AMP-activated protein kinase in regulating
37. Nagy L, Tontonoz P, Alvarez JGA, Chen H, Evans RM. Oxidized LDL
fatty acid metabolism during exercise. Appl Physiol Nutr Metab 34:
regulates macrophage gene expression through ligand activation of
315–322, 2009.
PPAR␥. Cell 93: 229 –240, 1998.
53. Stewart LK, Flynn MG, Campbell WW, Craig BA, Robinson JP,
38. Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial
Timmerman KL, McFarlin BK, Coen PM, Talbert E. The influence of
infarction and death from cardiovascular causes. N Engl J Med 356:
2457–2471, 2007. exercise training on inflammatory cytokines and C-reactive protein. Med
39. Nunn AV, Bell J, Barter P. The integration of lipid-sensing and anti- Sci Sports Exerc 39: 1714 –1719, 2007.
inflammatory effects: how the PPARs play a role in metabolic balance. 54. Sureda A, Ferrer MD, Tauler P, Romaguera D, Drobnic F, Pujol P,
Nucl Recept 5: 1–13, 2007. Tur JA, Pons A. Effects of exercise intensity on lymphocyte H2O2
40. Odegaard JI, Ricardo-Gonzalez RR, Goforth MH, Morel CR, Subra- production and antioxidant defences in soccer players. Br J Sports Med 43:
manian V, Mukundan L, Red Eagle A, Vats D, Brombacher F, 186 –190, 2009.
Ferrante AW, Chawla A. Macrophage-specific PPAR␥ controls alterna- 55. Tontonoz P, Nagy L, Alvarez JGA, Thomazy VA, Evans RM. PPAR-␥
tive activation and improves insulin resistance. Nature 447: 1116 –1120, promotes monocyte/macrophage differentiation and uptake of oxidised
2007. LDL. Cell 93: 241–252, 1998.
41. Patti ME, Butte AJ, Crunkhorn S, Cusi K, Berria R, Kashyap S, 56. Tsekouras YE, Yanni AE, Bougatsas D, Kavouras SA, Sidossis LS. A
Miyazaki Y, Kohane I, Costello M, Saccone R, Landaker EJ, Goldfine single bout of brisk walking increases basal very low-density lipoprotein
AB, Mun E, DeFronzo R, Finlayson J, Kahn CR, Mandarino LJ. triacylglycerol clearance in young men. Metabolism 56: 1037–1043, 2007.
Coordinated reduction of genes of oxidative metabolism in humans with 57. Yakeu G, Butcher L, Isa S, Webb R, Roberts A, Thomas AW, Backx
insulin resistance and diabetes: potential role of PGC1 and NRF1. Proc K, James PE, Morris K. Low-intensity exercise triggers monocyte
Natl Acad Sci USA 100: 8466 –8471, 2003. polarisation into the M2 anti-inflammatory phenotype. Atherosclerosis
42. Petridou A, Tsalouhidou S, Tsalis G, Schulz T, Michna H, Mougios V. 212: 668 –673, 2010.
Long-term exercise increases the DNA binding activity of peroxisome 58. Yoshida H, Ishikawa T, Suto M, Kurosawa H, Hirowatari Y, Ito K,
proliferator-activated receptor-␥ in rat adipose tissue. Metabolism 56: Yanai H, Tada N, Suzuki M. Effects of supervised aerobic exercise
1029 –1036, 2007. training on serum adiponectin and parameters of lipid and glucose metab-
43. Phielix E, Meex R, Moonen-Kornips E, Hesselink MK, Schrauwen P. olism in subjects with moderate dyslipidemia. J Atheroscler Thromb 17:
Exercise training increases mitochondrial content and ex vivo mitochon- 1160 –1166, 2010.

J Appl Physiol • doi:10.1152/japplphysiol.00864.2011 • www.jappl.org