EIR 14

Elvira Fehrenbach
1959 – 2008
Privatdozentin Dr. Elvira Fehrenbach has worked with me since 1994, when she
joined our Department of Transfusion Medicine at the university of Tuebingen,
Germany, to become head of our Exercise Immunology Lab. She maintained this
role with considerable success until she succumbed to cancer on Oct. 13, 2008.
Much of her work was done in close cooperation with the Department of Sports
Medicine at the university of Tuebingen: this included early work on the suppression of cytokine production by strenuous exercise, and the role of antioxidant
metabolism. She was among the first to describe the effects of exercise on the heat
shock protein system, and was also among the first to recognize and utilize the
power of microchip gene expression array analysis for our purposes.
From the beginning she helped me with the editorial work for EIR, becoming
Managing Editor for several years. Just before she was diagnosed with cancer, she
had completed her “Habilitation”, the last step in the German academic system
required to become Professor. Her final work focused on gender specific aspects
of exercise immunology, and the first results of the last study which she had carried out in collaboration with the Department of Sports Medicine are being published in this issue of EIR. By another cruel twist of fate she has been prevented
from enjoying this part of her professional success.
We will keep Elvira in our memory as a woman with a great personality who kept
pushing our common cause forward. This issue of EIR is dedicated to her.
Hinnak Northoff
EXERCISE
IMMUNOLOGY
REVIEW
VOLUME 14 • 2008
CONTENTS
From the Editors
7
Human Natural Killer Cell Subsets and Acute Exercise: A Brief Review
Brian W. Timmons and Thomas Cieslak
8
Effect of moderate exercise training on T-helper cell subpopulations
in elderly people
Kazuhiro Shimizu, Fuminori Kimura, Takayuki Akimoto, Takao Akama,
Kai Tanabe, Takahiko Nishijima, Shinya Kuno, and Ichiro Kono
24
Salmonella administration induces a reduction of wheel-running
activity via a TLR5-, but not a TLR4 dependent pathway in mice
Takashi Matsumoto, Daisuke Shiva, Noriaki Kawanishi,
Yasuko Kato, Jeffrey A. Woods and Hiromi Yano
38
Exercise-induced DNA damage: Is there a relationship with
inflammatory responses?
Oliver Neubauer, Stefanie Reichhold, Armen Nersesyan,
Daniel König, Karl-Heinz Wagner
51
Establishing a novel single-copy primer-internal intron-spanning
PCR (spiPCR) procedure for the direct detection of gene doping
Thomas Beiter, Martina Zimmermann, Annunziata Fragasso,
Sorin Armeanu, Ulrich M. Lauer, Michael Bitzer, Hua Su,
William L. Young, Andreas M. Niess and Perikles Simon
73
Gender- and menstrual phase dependent regulation of inflammatory
gene expression in response to aerobic exercise
Hinnak Northoff, Stephan Symons, Derek Zieker, Eva V. Schaible,
Katharina Schäfer, Stefanie Thoma, Markus Löffler, Asghar Abbasi,
Perikles Simon, Andreas M. Niess and Elvira Fehrenbach
86
Letter to the editor
Does prolonged exhausting exercise influence the immune system
in solid organ transplant recipients?
Ingmar Königsrainer, Derek Zieker and Alfred Königsrainer
104
Instructions for authors of EIR
106
Exercise Immunology Review
Editorial Statement
Exercise Immunology Review, an official publication of the International Society of Exercise
Immunology and of the German Society of Sports Medicine and Prevention, is committed to developing and. enriching knowledge in all aspects of immunology that relate to sport, exercise, and regular
physical ativity. In recognition of the broad range of disciplines that contribute to the understanding of
immune function, the journal has adopted an interdisciplinary focus. This allows dissemination of
research findings from such disciplines as exercise science, medicine, immunology, physiology,
behavioral science, endocrinology, pharmacology, and psychology.
Exercise Immunology Review publishes review articles that explore: (a) fundamental aspects of
immune function and regulation during exercise; (b) interactions of exercise and immunology in the
optimization of health and protection against acute infections: (c) deterioration of immune function
resulting from competitive stress and overtraining; (d) prevention or modulation of the effects of aging
or disease (including HIV infection; cancer; autoimmune, metabolic or transplantation associated disorders) through exercise. (e) instrumental use of exercise or related stress models for basic or applied
research in any field of physiology, pathophysiology or medicine with relations to immune function.
Editor: Prof. Dr. Hinnak Northoff
Managing Editor: Dr. Derek Zieker
Send editorial correspondence to:
Secretarial office EIR
Institute of clinical and experimental
Transfusion Medicine (IKET)
University of Tuebingen
Otfried-Mueller-Str. 4/1
72076 Tuebingen, Germany
ZKT.sekretariat@med.uni-tuebingen.de
Exercise Immunology Review (ISSN 1077-5552) is
published and sponsored annually by the Association for the Advancement of Sports Medicine
(Verein zur Förderung der Sportmedizin) and printed by TOM-Systemdruck GmbH, Hansanring 125.
Subscription rates are $25 in the US and €25 in
Europe and other countries. Student rates ($15 or
€15) available for up to 3 yrs. Along with payment
send name of institution and name of adviser.
Postmaster: Send address changes to Exercise
Immunology Review, TOM-Systemdruck GmbH.
Copyright © 2002 by Hinnak Northoff. Exercise Immunology Review is indexed in Sport Database,
Sport Discus, Sport Search, SciSearch, EMBASE/ Excerpta Medica, Focus on: Sports Science &
Medicine, Index Medicus, MEDLINE, Physical Education Index, Research Alert, International Bibliography of Periodical Literature, International Bibliography of Book Reviews, and CINAHL database.
Notice: authorization to photocopy items is granted for internal or personal use only. All other cases
contact Hinnak Northoff.
To subscribe or renew subscription phone +49 2571 5 78 89-0,
or write to TOM-Systemdruck, GmbH, Hansaring 125, D-48268 Greven,
e-mail: ribbers@tom-systemdruck.de
Editorial • 7
From the editors
This year´s issue of EIR holds only six full papers, but the contents are nevertheless of a heavyweight nature.
In the first article, Timmons et al. review the reaction of NK cells and their subsets to exercise, also briefly touching on the influence of gender and age and considering the possible clinical significance of these findings.
Shimizu et al., showing their own new data, demonstrate that moderate exercise is
associated with improvements in CD28 expression and the Th1 / Th2 balance in
the elderly.
Matsumoto et al., in the third article, use exercise as a suitable model to elucidate
the role of different surface molecules from salmonella in the reduction of voluntary physical activity associated with infection. They demonstrate that it is
induced by flagella antigens, mediated through TLR5.
Neubauer et al. then thoroughly review the role of DNA damage in the reaction of
the organism to strenuous exercise. They include their own novel data and conclude that DNA effects in lymphocytes are very likely not responsible for exercise
induced inflammation and vice versa.
Simon et al. also present new data. More than that, they present for the first time a
novel method capable of detecting gene doping in an unspecific way and with
high sensitivity. We are all waiting to see just how far this new investigative tool
will be used to provide data on the extent of gene doping in the future and perhaps
also in the past.
The last full article is by our group in Tübingen and shows the first data from the
last study, which Elvira Fehrenbach, longstanding managing editor of this journal,
had planned and started. It shows that female athletes show strikingly different
gene expression to exercise stress depending on the phase of their menstrual
cycle. Although Elvira could not finish the study by herself, she would have
enjoyed seeing the results.
Finally, we publish a letter by Königsrainer et al. who report on exercise in
patients with solid organ transplants. Maybe this journal will serve as a platform
to induce the international cooperation they are looking for.
Hinnak Northoff
8 • NK and exercise
Human Natural Killer Cell Subsets and
Acute Exercise: A Brief Review
Brian W. Timmons1 and Thomas Cieslak2
1 Children’s
2 Faculty
Exercise & Nutrition Centre, McMaster University,
of Physical Education and Health, University of Toronto
ABSTRACT
Natural killer (NK) cells are the most responsive immune cell to acute exercise.
This sensitivity to physiological stress combined with their role in innate immune
defences suggests that these cells may be a link between regular physical activity
and overall health status. NK cells are a heterogeneous population consisting of
at least two distinct subsets based on the expression intensity of CD56. CD56bright
and CD56dim cells exhibit different phenotypical and functional characteristics. In
this review, we examine the effects of acute exercise on NK cell subsets, with special reference to potential health implications of the findings. The available evidence suggests a differential mobilization of NK cell subsets in response to acute
exercise; CD56bright NK cells are less responsive than their CD56dim counterparts.
During the post-exercise recovery period (up to 1h), the ratio of
CD56bright:CD56dim cells favours the CD56bright subset. The potential significance
of these findings is discussed in the context of normal physiological adaptation to
exercise. We also discuss the potential role of exercise in certain clinical conditions (e.g., multiple sclerosis) as an adjunct therapy to mobilize the CD56bright
subset. Further investigation into the biology of NK cell subsets and exercise
should prove to be a fruitful area for years to come.
Keywords: CD56bright, CD56dim, acute exercise, humans, health, NK cells,
NK subsets
INTRODUCTION
Physical inactivity is considered to be an independent risk factor for various
chronic diseases of adulthood (3). While the mechanisms that translate a physically-active lifestyle into good health continue to be revealed, involvement of the
immune system has received considerable attention in recent years. Indeed, the
Correspondence to:
Brian W. Timmons, Ph.D., Children’s Exercise & Nutrition Centre
Chedoke Hospital, Evel Building, Room 469, Sanatorium Road
Hamilton, Ontario, Canada, L8N 3Z5
Telephone: (905) 521-2100 ext. 77218, Fax: (905) 385-5033
Email: timmonbw@mcmaster.ca
NK and exercise • 9
acute and chronic effects of exercise on numerous aspects of the human immune
system are the focus of substantial research, but the health significance of this
work remains to be unraveled. One aspect of the immune system that has garnished persistent interest is natural killer (NK) cells. The striking sensitivity of
NK cells to exercise stress provides strong support that these cells may be implicated as a potential link between regular physical activity and overall health status. NK cells are a heterogeneous population of lymphocytes, the biology of
which is under intense scrutiny given the clinical significance of NK cells in antiviral (2) and anti-cancer (5) defenses. New insights into the origin, development,
and interaction of these cells with other immune factors and non-immune tissue
represents an exciting and rapidly developing area of research and provides a
framework to explore the significance of exercise-induced changes in these cell
populations. Recent attention on NK cells has been driven by the presence of distinct NK cell subsets, which appear to hold diverse functions (see refs 7 and 9 for
reviews).
In this brief review, we shall consolidate the literature on NK cell subsets
and acute exercise in humans, while focusing on the heterogeneity of these cells.
This is not a comprehensive review of exercise and NK cells (these are available
elsewhere, e.g., (36)). Rather, the primary objective of this paper is to explore the
effects of acute exercise on NK cell subsets (i.e., CD56bright and CD56dim) with
special reference to potential health implications of the findings. To achieve this
objective, relevant published articles were retrieved by a PubMed database search
using the following keywords: “exercise” AND “CD56”. The reference lists of
relevant articles identified through the PubMed search were then hand-searched
for additional studies. In a few instances, data available in abstract form were
included because this was the only source of relevant information. Studies that
measured NK subsets in peripheral blood collected at a minimum before and
immediately after an acute bout of exercise were chosen.
Human NK Cell Subsets – An Overview
NK cells are large granular lymphocytes with natural cytotoxicity (9). NK cells
represent one component of innate immunity that can destroy certain virallyinfected and tumour cells, without prior sensitization (i.e., non MHC-restricted).
The widely accepted CD classification of NK cells includes the co-expression of
the Fcγ receptor III (CD16) and an isoform of the human neural cell adhesion
molecule (CD56) whose function on NK cells is unknown (9). The traditional
phenotype of human circulating NK cells therefore has been: CD3-CD16+CD56+.
More than 20 years ago the existence of two unique and functionally different NK
cell populations, based on the expression intensity of CD56 (Figure 1), was noted
(23).
CD3-CD56dim cells, which express high levels of CD16, are more cytotoxic
than CD3-CD56bright cells, which express low or no levels of CD16 (23). Mounting evidence suggests that the CD56bright subset, which comprises ~10% of circulating NK cells and possess the capacity to produce abundant cytokines (9), may
be of particular relevance in the early events of immune challenge by coordinating
“cross-talk” between innate and adaptive arms of immunity (13).
10 • NK and exercise
Further phenotypical and functional differences between CD56bright
and CD56dim cells are comprehensively contrasted in a recent review (7). A
potentially important distinction
between these subsets is the expression level of adhesion molecules. For
example, CD56bright cells feature quite
high levels of CD62L and it is
believed that enhanced expression of
these adhesion molecules on
CD56bright cells favours their trafficking to lymph nodes and sites of
inflammation, where they may initiate
or promote immune reactions (14).
Regulation of NK cell activation
Figure 1. Flow cytometry characterization of is another important issue in the conNK cell subsets. CD56bright and CD56dim cell text of NK cell subsets. For example,
CD56bright cells constitutively express
populations are derived from lymphocyte
events gated based on forward- vs sidescatter the high-affinity heterotrimeric IL-2
receptor complex, which provides this
characteristics.
subset an advantage of responding to
very low concentrations of IL-2 (7). IL-2-induced activation of CD56bright cells
results in the production of relatively large amounts of IFN-γ, which can shape the
Th1 immune response (28). Once activated, CD56bright cells are as cytotoxic as
the CD56dim subset (30). The activation of CD56dim cells is a very complex balance of activating and inhibitory signals. For example, when NK cells are
engaged with MHC class 1 molecules, the inhibitory killer immunoglobulin-like
receptors deliver a signal that prevents the NK cell from killing the target. In contrast, a number of activating receptors are present on NK cells. A full description
of activation regulation is beyond the scope of this paper so the readers are directed towards excellent recent reviews on the topic (6, 7). Needless to say, the decision for an NK cell to lyse its target must ultimately mean that activating signals
have dominated over inhibitory signals.
NK Cell Subsets – Distinct Cell Populations?
In spite of advances in our understanding of the biology of NK cell subsets, comprehension of their lineage remains an area of active research. It seems clear that
NK cells are derived from CD34+ hematopoietic progenitor cells and the known
site(s) of development and process of maturation suggest that CD56dim cells are
derived directly from the CD56bright subset (7). A recent study elegantly described
a sequential lineage whereby CD56bright, in contact with fibroblasts, can terminally differentiate into CD56dim cells (8). This study therefore suggested that the
CD56bright subset represents an immature form of NK cell that eventually reaches
the mature CD56dim phenotype with the right environment. Regardless of origin,
clear functional differences exist between these subsets.
NK and exercise • 11
There is growing evidence that NK cell subsets differ in both gene and protein expression. In a comprehensive study by Hanna and colleagues (19) gene
expression profiling of NK subsets revealed several novel functions. In the
CD56bright subset, 888 genes were found to be transcribed at significantly lower
levels (at least two fold) when compared with CD56dim cells, while 380 genes
were up-regulated. Various mRNA species for membrane proteins/receptors, signal transduction, secreted proteins, transcriptional and translational regulation,
apoptosis, cell cycle, and metabolism and structure were all found to be differentially expressed between the subsets. In some instances, 15-fold higher levels of
expression for some species (e.g., Lymphopain, HLA-DRA, and Granzyme K)
were observed in the CD56bright subset. Consistent with these findings, gene
expression of cytolytic molecules was found to be generally higher in CD56dim
than in CD56bright subsets (with the exception of Granzyme K) whereas expression of molecules involved in adhesion, migration, and cell to cell cross talk was
generally higher in the CD56bright subset (46). In total, Wendt et al.’s analysis distinguished the two NK cell subsets in the expression of 473 transcripts (46). Some
evidence also suggests that intrinsic (i.e., unstimulated) protein expression may
be greater in CD56bright than in CD56dim cells (46). For example, IL-8 expression
was greater in the former subset, although more work is required to understand
differences in intrinsic protein production between the subsets. That NK cell subsets differ in both gene and protein expression in the unstimulated state is particularly relevant from a physical activity perspective considering that both subsets
are mobilized into the peripheral circulation with acute exercise.
NK Cells and Acute Exercise
NK cells defined by the traditional phenotype (i.e., CD3−CD16+CD56+ cells)
seem to be the most sensitive and therefore responsive cell type to an acute bout
of exercise, whether that exercise is predominantly aerobic or anaerobic in nature
(36). NK cells are rapidly mobilized into the peripheral circulation most likely via
multiple mechanisms including: shear stress due to a substantial increase in
peripheral blood flow and a catecholamine-induced down-regulation of adhesion
molecule expression (29). Although NK cells present in the peripheral blood represent a very small proportion of the body’s total NK cell pool at rest (47), the
striking exercise-induced increase in the peripheral pool has been linked to an
enhanced immune surveillance (32). It is interesting to note, however, that during
very prolonged exercise (i.e., > ~3 h), circulating NK cell counts may return to
pre-exercise levels and can even drop below pre-exercise levels (16). The mechanisms for this are debatable, but clearly the exit of cells outweighs their entry into
the circulation, possibly to enter sites of muscle damage, for which there is some
evidence (26). One might argue that a blunted peripheral infiltration of NK cells,
particularly of the more cytotoxic (i.e., CD56dim) subset, might reflect a reduced
ability to defend against pathogens. Alternatively, the exit of cells from the circulation or an inhibition of their entry could mean that these cells are remaining or
trafficking to sites where they are needed to affect immune or inflammatory function. Indeed, the true health significance of exercise-induced changes in human
NK cells is open for debate. Unfortunately, mouse NK cells do not express the
12 • NK and exercise
murine homologue of CD56 and although mouse NK cells can be subdivided
based on expression intensity of CD27, which have some similarities to that of
human NK cell subsets, in vivo studies of NK cell subset function are lacking.
NK Cell Subsets and Acute Exercise
Compared with the abundance of exercise literature that has addressed the traditional CD3−CD16+CD56+ NK cell phenotype, there are only a handful of studies
that have addressed how NK cells expressing different intensities of the CD56
antigen respond to exercise. Although Horn et al. (20) reported that an acute bout
of incremental high intensity exercise mobilized NK cells with greater intensity
of CD16/CD56 expression, compared with NK cells at rest, this study simultaneously measured the expression of CD16 and CD56 antigens and, therefore, could
not truly distinguish between CD56bright and CD56dim cells. Gannon et al. (16)
determined NK cell subset counts before and after a 250-km cycling road race.
Blood samples were drawn 24 h prior to the race and 10 to 25 min following the
race, which lasted approximately 7 h. These authors found that cell counts of both
NK subsets determined following the race were actually lower than their respective pre-exercise levels (Table 1), supporting the idea of an exit of cells from the
circulation with prolonged duration of exercise. In what appears to be the first
report of the effect of acute exercise on NK cell subsets, Berk et al. (1) found that
numbers of both CD56+CD16+ (likely CD56dim cells) and CD56+CD16− (likely
CD56bright cells) lymphocytes increased in the peripheral circulation after 1 h of
treadmill running, but were below pre-exercise values after a full 3 h of running
(Table 1). The latter study should be interpreted with caution, however, since the
distinct NK cell subsets based on the expression intensity of CD56 was not
specifically examined in this study. More recently, Suzui and colleagues (38)
reported on the effects of brief, incremental exercise on NK cell subsets and
found that only the proportion of CD56dim cells increased in response to exercise.
A finding that was later confirmed by the same research team (39). The handful of
studies described above demonstrates that both NK cell subsets are responsive to
acute bouts of exercise. However, the inconsistencies in study design, blood sampling time, and flow cytometry methods make it difficult to interpret the findings.
Moreover, only one of these studies measured subsets into the post-exercise
recovery period.
In recent studies, our laboratory has addressed the issue of a differential
mobilization of NK cell subsets in response to acute exercise. Unlike the majority
of exercise immunology studies, our research is focused on the child and adolescent. In all of our studies, we used an exercise model consisting of 60 min duration at ~70% of maximal oxygen uptake (VO2max). While this type of exercise is
not consistent with most young people’s habitual physical activity patterns, our
objective was to induce significant physiological stress, thus maximizing the
mobilization and representation of NK cells in the peripheral circulation. With
this standardized approach, we confirmed a differential mobilization of CD56bright
and CD56dim subsets in response to exercise, including an elevated ratio of
CD56bright to CD56dim cells during the recovery period (43, 44). Whether studying
male (44) or female (40) children or adolescent boys and girls (43), the CD56dim
NK and exercise • 13
subset responded with greater magnitude than did the CD56bright subset after the
60 min of exercise, and this response was usually apparent after only 30 min of
exercise. After 30 and 60 min of seated resting recovery following the exercise
task, both CD56bright and CD56dim subsets had returned close to pre-exercise levels. However, while CD56bright cells remained slightly above resting levels,
CD56dim cells dipped slightly below resting levels.
Based on the literature reviewed to this point, a common theme of a differential mobilization of NK cell subsets in response to acute exercise emerges. In
almost every study, the CD56dim subset is more responsive to exercise than is the
CD56bright subset. To illustrate this conclusion, we have calculated the effect size
(ES) for each cell type’s response to exercise from four of the studies discussed
above (1, 40, 43, 44). Even though these studies implemented different testing
protocols and flow cytometry methods, the comparison of NK cell subset
responses remains valid because the comparison is made “within subjects”. Table
1 provides the cell counts (mean ± SD) and the corresponding ES, which is calculated as: ES = (post-exercise cell count − pre-exercise cell count)/average SD of
pre- and post-exercise cell counts. In these studies represented in Table 1, “postexercise cell counts” were always taken after 60 min of exercise, thus the duration
of activity is consistent across studies. In all these studies, the exercise intensity
was also similar at ~70% VO2max. The ESs were then compared using a dependent
t-test. The results of this examination clearly indicate that the ESs of CD56dim
cells were significantly (p = 0.02) greater than those of CD56bright cells, thus supporting a differential mobilization of NK cell subsets in response to acute exercise.
An alternative approach to illustrate the differential mobilization of NK cell
subsets is to calculate the ratio of CD56bright to CD56dim cells. The clinical significance of the ratio or the balance between CD56bright and CD56dim cells is disTable 1. Studies reporting the mobilization of NK cell subsets in response to acute exercise
CD56bright cells
Study
CD56dim cells
Pre-exercise
Post-exercise
Pre-exercise
Post-exercise
MEAN SD
MEAN
SD
MEAN SD
MEAN SD
ES
ES*
Timmons et al. 2007 14.7
7.1
28.1
13.3 1.0
153
58
353
178
1.7
Timmons et al. 2006 14.3
4.5
28.1
14.0 1.5
119
29
356
210
2.0
14.8
7.8
18.5
11.3
0.4
184
31
347
251
1.2
19.7
7.6
36.3
13.6 1.6
135
36
381
170
2.4
Timmons et al. 2006 19.1
9.3
45.0
18.5 1.9
173
81
537
236
2.3
17.5
6.7
34.0
13.4 1.6
132
90
336
206
1.4
14.0
9.5
22.0
15.8 0.6
300
126
470
284
Berk et al. 1990
MEAN ES
Berk et al. 1990
Gannon et al. 1997
14.0
15.0
1.2
0.8
MEAN ES
1.7
9.5
30.01
25.3 0.9
300
126
3301
190
0.2
10.0
10.02
7.0
340
200
3002
200
-0.2
-0.6
Unless otherwise indicated, values are derived from blood samples collected at rest and after 60 min of
exercise expressed as cells × 106/L. ES, effect size calculated as (post-exercise mean − pre-exercise
mean)/mean of pre- and post-exercise SD. *Indicates that ES for CD56dim cells are significantly larger
than for CD56bright cells (t=3.138, df=6; p = 0.02). The two studies included at the bottom of the table are
for comparison purposes, as values were taken from blood samples collected after either 3 h1 or 7 h2 of
exercise.
14 • NK and exercise
cussed later in this paper. However, a recent exercise-related
study (37) found that, during
sport training in healthy women,
the lowest measured whole
blood NK cell function (i.e.,
cytotoxicity) occurred concurrently with the highest blood
CD56bright:CD56dim ratio. In our
studies of children and adolescents, we observed a slight
decrease in this ratio during
exercise, but a pronounced
Figure 2. Ratio of CD56bright to CD56dim cells
increase during the recovery
before, during, and following acute exercise in
healthy children and adolescents. Values are mean ± period (Figure 2). Thus, the balSD. * significantly different from −40 min (rest).
ance of NK cell subsets during
Subjects cycled for 60 min at ~70% of maximal
recovery from physiological
oxygen uptake.
stress is in favour of the
CD56 bright subset. This is an
important observation because the recovery period from exercise is a time when
homeostatic recovery and tissue adaptation occur (25), suggesting that the
CD56bright subset may play a role in this process (see below).
Factors that Influence Mobilization of NK Cell Subsets
in Response to Acute Exercise
Notwithstanding the fact that only a few studies have addressed the impact of
acute exercise on NK cell subsets, a number of factors seem to influence their
mobilization. Consistent with the differential response of these subsets, the same
factor may have a different effect on different subsets. In the following sections, a
brief overview of some of these factors is provided.
Exercise duration and intensity
As with any marker of the immune system, the timing of a blood sample during
exercise and the intensity at which the exercise is performed are important when
interpreting the NK cell subset response. Our studies in children and adolescents
were restricted to a total of 60 min of constant-load cycling, but there was no difference in cell counts determined after 30 or 60 min of the exercise (40, 43, 44).
These findings suggest that the mobilization of NK cell subsets is relatively rapid
and complete by 30 min of exercise. These findings are supported by Berk et al.
(1), who found a relatively small increment in levels of CD56bright cells after 3
compared with 1 h of treadmill running, whereas levels of CD56dim cells had
started to return to resting levels by 3 h of running (Table 1). In contrast, a field
study found that ~7 h of road cycling resulted in NK cell subset counts that were
actually lower than pre-exercise values (Table 1). These latter results must be
interpreted with caution, however, since ~17 min passed from the end of the race
until blood collection; it is conceivable that dramatic changes in cell counts
NK and exercise • 15
occurred within this time frame. Moreover, the real-life setting for this study (i.e.,
road racing) would mean that exercise intensity would not be kept constant, as is
possible in controlled laboratory studies. Nevertheless, the limited evidence suggests that both NK cell subsets are rapidly mobilized into the circulation in
response to exercise and remain at constant levels over time when the exercise
intensity is held constant. This balance apparently reflects an equalization of entry
and exit of these cells into and out of the peripheral circulation. Although the evidence is not strong, prolonged exercise (i.e., >3 h) may result in a net exit of NK
cell subsets out of the circulation (16), as previously suggested a decade ago (17).
If true, it will be interesting to identify the fate of these cells (e.g., apoptosis or tissue infiltration) and the factors that regulate these processes.
Few studies have specifically addressed the extent to which exercise intensity alters NK cell subsets. Suzui and colleagues (38) reported on the effects of
brief, incremental exercise on NK cell subsets. Nine males exercised on a cycle
ergometer for 5 min at each of 4 increasing intensities (50, 90, 120, and 140% of
their individual ventilatory threshold), with blood samples drawn after every
workload. The authors found that only the proportion of CD56dim cells increased
in response to exercise; the proportion of CD56bright cells in the peripheral circulation did not change. However, because of an overall leukocytosis both CD56bright
and CD56dim cell counts increased with increasing exercise intensity. In a followup study (39), the same authors confirmed their earlier findings by showing that
in 6 males cycling for 30 min at 120% of their individual ventilatory threshold
(~70% VO2max) the proportion of CD56dim cells but not the proportion of
CD56bright cells increased significantly. Based on these two studies, one can conclude that the redistribution of CD56bright cells appears to be resistant to changes
in exercise intensity. However, much more work is needed to clarify how exercise
intensity influences mobilization of NK cell subsets.
Training status (fitness)
To our knowledge, only one published study has determined the extent to which
training status – or more specifically aerobic fitness – is associated with NK cell
subsets. This study by Rhind et al. (34) reported that seven endurance-trained men
exhibited a higher proportion of the CD56bright NK cell subset, as compared with
6 untrained men, although this difference was found in blood samples collected at
rest. Whether this finding was a result of the chronically trained state or of recent
training history is unclear, since Suzui and colleagues (37) demonstrated that
changes in an athlete’s training load acutely increases the proportion of CD56bright
NK cells in the circulation. To this end, we are not aware of any studies that have
determined whether training status per se influences the mobilization of NK cell
subsets in response to acute exercise. To further explore the potential relationship
between aerobic fitness and NK cell subsets, we returned to our previously published data (40, 43, 44) and performed correlation analyses on 54 boys and girls.
The results are presented in Table 2. We did not find a relationship between aerobic fitness and NK cell subsets at rest, nor could we conclude that aerobic fitness
was associated with the exercise-induced change in NK cell subsets (i.e., the magnitude of the exercise effect). The influence of training status on NK cell subsets
therefore deserves additional investigation to clarify the influence of regular physical activity.
16 • NK and exercise
Carbohydrate intake
It has been known for some time that NK cells (i.e., CD3−CD16+CD56+) are sensitive to carbohydrate (CHO) intake (usually in the form of a sport drink) during
exercise (31). To identify the effects of CHO it was suggested that exercise must
be prolonged and intense because the proposed CHO effects on NK cell redistribution in adults was due to a blunted stress hormone response mediated by mainTable 2. Pearson correlation coefficients between aerobic fitness and NK cell subsets at rest and in
response to exercise.
VO2max (ml•kg−1•BM−1)
Resting
CD56bright
cell count
Resting CD56dim cell count
VO2max (ml•kg−1•LBM−1)
−0.11 (0.43)
−0.16 (0.25)
−0.10 (0.49)
−0.04 (0.78)
Resting Ratio (CD56bright:CD56dim)
0.09 (0.52)
−0.04 (0.80)
Resting CD56bright cell proportion
−0.07 (0.60)
−0.16 (0.23)
−0.04 (0.78)
Resting CD56dim cell proportion
−0.04 (0.75)
Exercise-induced change in CD56bright cell counts
−0.01 (0.95)
−0.03 (0.83)
Exercise-induced change in CD56dim cell counts
−0.07 (0.61)
−0.02 (0.90)
Values in parentheses are p values. VO2max, maximal oxygen uptake; BM, body mass; LBM, lean body
mass. The exercise-induced change in NK cell subset is taken as the difference between the cell count at
60 min of exercise (70% VO2max) and at rest (i.e., Δ).
tained or increased blood glucose concentrations (31). While this explanation
may be adequate for the adult response, evidence supporting this theory in the
paediatric population is lacking. However, our studies did find that CHO intake
attenuated the CD56dim, but not the CD56bright, response in young boys (44) and
girls (40). In contrast, CHO intake equally attenuated both subsets in older male
and female adolescents performing the same exercise (43). In pre-pubertal and
early-pubertal boys, the attenuating effect of CHO on CD56dim cells is already
visible after only 30 min of exercise whereas in late-pubertal boys the effect
becomes apparent after 60 min of exercise (44). In spite of significant CHO
effects on NK cell subsets in the above studies, there was no evidence of an effect
on epinephrine (adrenaline), norepinephrine (noradrenaline) or cortisol – stress
hormones thought to be involved in mediating the relationship between CHO
intake and NK cell redistribution (31). This dissociation between changes in cell
counts and stress hormones with and without CHO intake suggested a direct
effect of CHO intake on NK cell subsets, possibly due to elevated blood glucose
levels and associated with normal puberty. This possibility is supported by the
observation that the one hormone affected by CHO intake in our studies was
growth hormone (GH). We found a significant correlation between exerciseinduced changes in GH and NK cell subsets in boys but this relationship was not
found in girls (unpublished observations). Due to the limited evidence to date, the
overall health significance of CHO effects on NK cell subset responses to exercise
is unclear, but these studies need to be reproduced in adults.
Sex
In one study, female sex was formally investigated as a potential mediator of the
NK cell subset response to exercise (43). Interestingly, both the CD56dim and
NK and exercise • 17
CD56bright subsets increased significantly more in female adolescents than in male
adolescents. However, the magnitude of this enhanced response was similar
between subsets as the ratio of CD56bright:CD56dim cells responded in a similar
fashion between the sexes. In two independent publications, we reported changes
in NK cell subsets in 12-yr-old boys (44) and girls (40). Since both groups were
tested under identical conditions, we were able to compare their responses to
examine whether the previously observed sex-based differences in adolescents
was present in younger children. Although some aspects of the NK cell response
were different between the boys and girls (see ref (40) for details), the actual
increase in both the CD56dim and CD56bright subsets was practically identical.
The apparent age x sex interaction in NK cell subset responses to physiological stress observed in adolescents may be relevant from a reproductive perspective. CD56bright cells found in the decidual tissue of early pregnancy could be
important in maternal-foetal tolerance (22). Whether acute exercise or regular
physical activity influences these subsets would be of considerable interest given
the interest in exercise recommendations during pregnancy (45). Moreover, studies are needed to more clearly understand the potential impact of the menstrual
cycle and sex hormones on NK cell subset responses to exercise. We have previously reported that the total lymphocyte pool is more responsive to exercise during the luteal phase in women taking oral contraceptives but not in non-users (41);
however specific effects on NK cell subsets remain to be determined.
Puberty
As in adults, NK cells (i.e., CD3-CD16+CD56+) are the most responsive cell type
to exercise in children, but the magnitude of the response to strenuous exercise is
lower in pre- and early-pubertal boys, as compared with men (42). Since work in
our laboratory is interested primarily in exercise responses during childhood, we
are particularly focused on how growth and development influence NK cell
responses to exercise. To examine this issue, we recruited boys of the same
chronological age but who varied in their pubertal development (44). We showed
that boys at the most advanced stages of puberty demonstrated the greatest
increase in the proportion of CD56dim during exercise, but that the increase in
CD56dim cell counts did not vary statistically by pubertal group due to a slightly
greater increase in total lymphocyte counts in the pre- and early-pubertal boys
(44). However, the exercise-induced increase in both the proportion and count of
the CD56bright subset was greatest in the boys at more advanced stages of puberty.
Based on these observations, responsiveness of NK cell subsets to exercise seems
to be dependent to some extent on the developmental stage of the individual.
In summary, we hope that these preliminary data will stimulate interest in
further understanding the mechanisms underlying NK cell subset mobilization
with acute exercise. Notwithstanding the few studies that have appropriately
addressed NK cell subset mobilization with exercise, the findings suggest that
several factors may be involved; in some instances the effect varies with subset
(e.g., CHO intake). An important contribution to the biology of NK cells will be
to elucidate functional responses (e.g., gene and protein expression) in subsets
sensitive to acute exercise.
18 • NK and exercise
What is the Significance of Exercise-induced Changes
in NK Cell Subsets?
Studies of exercise-induced changes in NK cell subsets have been descriptive in
nature, and the true health significance of transient changes in these cell populations remains unclear. When assessing the significance of exercise-induced
changes, we believe there are at least two possible interpretations. The first is to
consider changes in NK cells representing alterations to immune function (i.e.,
antiviral defence). In this context, one practical consequence of exercise-induced
alterations in NK cell subsets may relate to the measure of NK cell cytolytic function. A well-described phenomenon in the exercise immunology literature is a
period of relative immune function depression following high-intensity exercise,
consistently associated with depression of NK cell function and termed the “open
window”; a period of time when the host may be at increased susceptibility to
infection (32). A recent study (37) tracked changes in CD56dim and CD56bright
cells over one month of competitive sports training in healthy adult females and
found that the time during training with the lowest NK cell cytotoxicity corresponded to the time when CD56bright cell counts were highest and CD56dim cells
remained unchanged (i.e., when CD56bright:CD56dim ratio was highest). These
findings support the notion that reduced NK cell function (as measured by in vitro
cytotoxicity assays) during recovery from high-intensity acute exercise (32) and
periods of intensified training (37) may be due to disproportionate changes in NK
cell subsets; a high proportion of CD56bright cells, which have low unstimulated
cytotoxicity (9), may effectively “dampen” overall killing capacity. The findings
from our studies that show an increase in the ratio of CD56bright to CD56dim cells
during the recovery period are further evidence that observed deficits in NK cell
cytotoxic assays may be due to a disproportionate number of CD56bright cells in
the mix. However, our studies never measured NK cell function per se, and we
cannot therefore make this link conclusively.
Alternatively, the exercise-induced redistribution of NK cell subsets
observed following the end of exercise may reflect a process of homeostatic
recovery and adaptation in response to physiological stress with very little to do
with immune function per se. Based on the understanding that CD56bright cells
possess an enhanced capacity for cytokine production and express elevated levels
of adhesion molecules integral for tissue homing, it has been suggested that
CD56bright cells may be of particular relevance coordinating the early events of
immune activation in response to endogenous tissue injury (10). In support of this
hypothesis, it has been shown that CD56bright cells are enriched at the sites of
inflammation in humans (11). Given their established roles in pathological states,
it is reasonable to predict that CD56bright cells are mobilized as components of the
normal physiological adaptation to exercise. To this end, CD56bright cells express
an abundance of angiogenic growth factors (24), indicating the potential for these
cells to contribute to angiogenesis, a hallmark physiological adaptation to regular
exercise. Additional studies that measure adhesion molecule expression and
cytokine and growth factor production in NK cell subsets mobilized with exercise
are therefore required to further elucidate the potential role of NK cell subsets in
exercise adaptation.
NK and exercise • 19
NK Cell Subsets and Exercise: Clinical Implications
At this time, a reminder of the maturation of the NK cell is appropriate. A recent
review of the literature (1) strongly suggested that CD56bright NK cells give rise to
a mature CD56dim cell, which is the prevailing circulating NK cell phenotype.
Hence, if CD56dim cells are derived from CD56bright cells, then the exerciseinduced mobilization of the latter subset could reflect a mobilization of “immature” NK cells. This idea is consistent with the differential mobilization of naïve
and memory T cells by exercise reported by Gannon et al. (15). That an acute bout
of exercise can mobilize CD56bright cells to the circulation, most likely from secondary lymphoid tissue where they are abundant (7), leads one to suspect an
important clinical role for exercise.
Although the biological significance of NK cell subset responsiveness to
exercise requires further investigation, NK cells are an important first line of
defence against tumour growth, and the unique immunoregulatory properties of
the CD56dim and CD56bright subsets mark them as candidates for immunotherapy
of cancer (9). Whether the redistribution of CD56dim and CD56bright cells in
response to exercise could be of therapeutic benefit in children recovering from
cancer (12), for example, remains to be determined. In patients recovering from
bone marrow transplantation, peripheral blood is reconstituted early on by
CD56bright NK cells (21, 33) more so than CD56dim cells (33), which may be related to the maturation process of these cells. Thus, it has been suggested that stimulation of the CD56bright subset with the NK cell compartment may be a therapeutic
effect to prevent relapse of residual disease (33); exercise might be an excellent
strategy to achieve this effect. NK cell subsets are also implicated in a variety of
diseases, ageing, and female reproduction. The proportion of CD56bright cells in
the peripheral blood of individuals receiving coronary artery by-pass surgery, for
example, tend to be lower than in age-matched controls (18). Patients with multiple sclerosis treated for 12-months with IFN-β therapy demonstrate a reduction in
the proportion of CD56dim NK cells and an increase in the proportion of
CD56bright NK cells (35). (The authors suggested that CD56bright cells may have
an immunoregulatory role within the central nervous system at sites of inflammation). Finally, normal ageing is associated with a reduction in the ratio of
CD56bright to CD56dim due to an expansion of CD56dim cells (4).
Collectively, these observations create numerous exciting opportunities to
further elucidate the clinical role of NK cell subsets. Given the sensitivity of these
subsets to acute exercise and their differential mobilization, it is an intriguing idea
that exercise could be used in these clinical conditions as an adjunct therapy to
mobilize the CD56bright subset. With respect to ageing, it will be important to distinguish the impact of ageing per se from that of physical inactivity on NK cell
subsets, as this distinction is crucial for other immune markers (27).
SUMMARY AND FUTURE DIRECTIONS
The objective of this review was to explore the effects of acute exercise on NK
cell subsets (i.e., CD56bright and CD56dim) in humans with special reference to
potential clinical implications of the findings. We have argued that NK cell sub-
20 • NK and exercise
sets display a differential mobilization in response to an acute bout of exercise,
with CD56dim cells more responsive than CD56bright cells. A number of factors,
including exercise duration and intensity, CHO intake, sex, and puberty, seem to
influence the mobilization of these subsets, and much more work is needed to
identify additional moderating factors (e.g., training status) and to understand the
mechanisms of mobilization. However, we hope that the literature reviewed herein will serve as a foundation on which to pursue future studies designed to reveal
the mechanisms associated with these phenomena. We also encourage exercise
physiologists and immunologists to pool their efforts in future studies to further
expand our understanding of how (and why) exercise impacts the biology of
human NK cell subsets. Exercise, for example, may be an effective adjunct therapy to promote expansion of NK cell subsets in the development of novel
immunotherapeutic approaches. In addition to their roles in pathology (e.g.,
arthritis), NK cells may also serve a physiological role by facilitating the adaptive
processes incurred by regular physical activity. Indeed these areas, among many
others, should prove to be a fruitful area of research. It is hoped that this paper
will spark new research into the biology of NK cells and exercise.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Berk LS, Nieman DC, Youngberg WS, Arabatzis K, Simpson-Westerberg M, Lee
JW, Tan SA and Eby WC. The effect of long endurance running on natural killer
cells in marathoners. Med Sci Sports Exerc 22: 207-212, 1990.
Biron CA, Nguyen KB, Pien GC, Cousens LP and Salazar-Mather TP. Natural killer
cells in antiviral defense: Function and regulation by innate cytokines. Ann Rev
Immunol 17: 189-220, 1999.
Booth FW, Gordon SE, Carlson CJ and Hamilton MT. Waging war on modern
chronic diseases: primary prevention through exercise biology. J Appl Physiol 88:
774-787, 2000.
Borrego F, Alonso MC, Galiani MD, Carracedo J, Ramirez R, Ostos B, Pena J and
Solana R. NK phenotypic markers and IL2 response in NK cells from elderly people. Exp Gerontol 34: 253-265, 1999.
Brittenden J, Heys SD, Ross J and Eremin O. Natural killer cells and cancer. Cancer
77: 1226-1243, 1996.
Bryceson YT, March ME, Ljunggren HG and Long EO. Activation, coactivation, and
costimulation of resting human natural killer cells. Immunol Rev 214: 73-91, 2006.
Caligiuri MA. Human natural killer cells. Blood 112: 461-469, 2008.
Chan A, Hong DL, Atzberger A, Kollnberger S, Filer AD, Buckley CD, McMichael
A, Enver T and Bowness P. CD56bright human NK cells differentiate into CD56dim
cells: role of contact with peripheral fibroblasts. J Immunol 179: 89-94, 2007.
Cooper MA, Fehniger TA and Caligiuri MA. The biology of human natural killercell subsets. Trends Immunol 22: 633-640, 2001.
Cooper MA, Fehniger TA, Turner SC, Chen KS, Ghaheri BA, Ghayur T, Carson WE
and Caligiuri MA. Human natural killer cells: a unique innate immunoregulatory
role for the CD56(bright) subset. Blood 97: 3146-3151, 2001.
Dalbeth N, Gundle R, Davies RJ, Lee YC, McMichael AJ and Callan MF. CD56bright
NK cells are enriched at inflammatory sites and can engage with monocytes in a
reciprocal program of activation. J Immunol 173: 6418-6426, 2004.
NK and exercise • 21
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
Dilloo D, Laws HJ, Hanenberg H, Korholz D, Nurnberger W and Burdach SE.
Induction of two distinct natural killer-cell populations, activated T cells and antineoplastic cytokines, by interleukin-2 therapy in children with solid tumors. Exp
Hematol 22: 1081-1088, 1994.
Fehniger TA, Cooper MA, Nuovo GJ, Cella M, Facchetti F, Colonna M and Caligiuri
MA. CD56bright natural killer cells are present in human lymph nodes and are activated by T cell-derived IL-2: a potential new link between adaptive and innate immunity. Blood 101: 3052-3057, 2003.
Frey M, Packianathan NB, Fehniger TA, Ross ME, Wang WC, Stewart CC, Caligiuri
MA and Evans SS. Differential expression and function of L-selectin on CD56bright
and CD56dim natural killer cell subsets. J Immunol 161: 400-408, 1998.
Gannon GA, Rhind S, Shek PN and Shephard RJ. Naive and memory T cell subsets
are differentially mobilized during physical stress. Int J Sports Med 23: 223-229,
2002.
Gannon GA, Rhind SG, Suzui M, Shek PN and Shephard RJ. Circulationg levels of
peripheral blood leukocytes and cytokines following competitive cycling. Can J
Appl Physiol 22: 133-147, 1997.
Gannon GA, Shek P and Shepard RJ. Natural killer cells: Modulation by intensity
and duration of exercise. Exerc Immunol Rev 1: 26-48, 1995.
Hak L, Mysliwska J, Wieckiewicz J, Szyndler K, Trzonkowski P, Siebert J and Mysliwski A. NK cell compartment in patients with coronary heart disease. Immun Ageing 4: 3, 2007.
Hanna J, Bechtel P, Zhai Y, Youssef F, McLachlan K and Mandelboim O. Novel
insights on human NK cells’ immunological modalities revealed by gene expression
profiling. J Immunol 173: 6547-6563, 2004.
Horn PL, Leeman K, Pyne DB and Gore CJ. Expression of CD94 and 56(bright) on
natural killer lymphocytes - the influence of exercise. Int J Sports Med 23: 595-599,
2002.
Jacobs R, Stoll M, Stratmann G, Leo R, Link H and Schmidt RE. CD16- CD56+ natural killer cells after bone marrow transplantation. Blood 79: 3239-3244, 1992.
Koopman LA, Kopcow HD, Rybalov B, Boyson JE, Orange JS, Schatz F, Masch R,
Lockwood CJ, Schachter AD, Park PJ and Strominger JL. Human decidual natural
killer cells are a unique NK cell subset with immunomodulatory potential. J Exp
Med 198: 1201-1212, 2003.
Lanier LL, Le AM, Civin CI, Loken MR and Phillips JH. The relationship of CD16
(Leu-11) and Leu-19 (NKH-1) antigen expression on human peripheral blood NK
cells and cytotoxic T lymphocytes. J Immunol 136: 4480-4486, 1986.
Lash GE, Schiessl B, Kirkley M, Innes BA, Cooper A, Searle RF, Robson SC and
Bulmer JN. Expression of angiogenic growth factors by uterine natural killer cells
during early pregnancy. J Leukoc Biol 80: 572-580, 2006.
Mahoney DJ, Parise G, Melov S, Safdar A and Tarnopolsky MA. Analysis of global
mRNA expression in human skeletal muscle during recovery from endurance exercise. FASEB J 19: 1498-1500, 2005.
Malm C, Sjodin TL, Sjoberg B, Lenkei R, Renstrom P, Lundberg IE and Ekblom B.
Leukocytes, cytokines, growth factors and hormones in human skeletal muscle and
blood after uphill or downhill running. J Physiol 556: 983-1000, 2004.
McFarlin BK, Flynn MG, Campbell WW, Craig BA, Robinson JP, Stewart LK, Timmerman KL and Coen PM. Physical activity status, but not age, influences inflam-
22 • NK and exercise
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
matory biomarkers and toll-like receptor 4. J Gerontol A Biol Sci Med Sci 61: 388393, 2006.
Mocikat R, Braumuller H, Gumy A, Egeter O, Ziegler H, Reusch U, Bubeck A,
Louis J, Mailhammer R, Riethmuller G, Koszinowski U and Rocken M. Natural
killer cells activated by MHC class I(low) targets prime dendritic cells to induce protective CD8 T cell responses. Immunity 19: 561-569, 2003.
Nagao F, Suzui M, Takeda K, Yagita H and Okumura K. Mobilization of NK cells by
exercise: downmodulation of adhesion molecules on NK cells by catecholamines.
Am J Physiol Regul Integr Comp Physiol 279: R1251-R1256, 2000.
Nagler A, Lanier LL, Cwirla S and Phillips JH. Comparative studies of human FcRIII-positive and negative natural killer cells. J Immunol 143: 3183-3191, 1989.
Nieman DC. Influence of carbohydrate on the immune response to intensive, prolonged exercise. Exerc Immunol Rev 4: 64-76, 1998.
Pedersen BK and Ullum H. NK cell response to physical activity: possible mechanisms of action. Med Sci Sports Exerc 26: 140-146, 1994.
Raspadori D, Lauria F, Ventura MA, Tazzari PL, Ferrini S, Miggiano MC, Rondelli
D and Tura S. Low doses of rIL2 after autologous bone marrow transplantation
induce a “prolonged” immunostimulation of NK compartment in high-grade nonHodgkin’s lymphomas. Ann Hematol 71: 175-179, 1995.
Rhind SG, Shek PN, Shinkai S and Shephard RJ. Differential expression of interleukin-2 receptor alpha and beta chains in relation to natural killer cell subsets and
aerobic fitness. Int J Sports Med 15: 311-318, 1994.
Saraste M, Irjala H and Airas L. Expansion of CD56Bright natural killer cells in the
peripheral blood of multiple sclerosis patients treated with interferon-beta. Neurol
Sci 28: 121-126, 2007.
Shephard RJ and Shek PN. Effects of exercise and training on natural killer cell
counts and cytolytic activity: A meta-analysis. Sports Med 28: 177-195, 1999.
Suzui M, Kawai T, Kimura H, Takeda K, Yagita H, Okumura K, Shek PN and Shephard RJ. Natural killer cell lytic activity and CD56(dim) and CD56(bright) cell distributions during and after intensive training. J Appl Physiol 96: 2167-2173, 2004.
Suzui M, Takeda K, Yagita H, Okumura K, Shek PN and Shephard RJ. Changes in
the proportion of CD56dim and CD56bright natural killer cells during incremental
exercise. Med Sci Sports Exerc 37: S373, 2005.
Suzui M, Takeda K, Yagita H, Okumura K, Shek PN and Shephard RJ. Changes in
the proportions of CD56dim and CD56bright natural killer cells during and after acute
exercise. Med Sci Sports Exerc 38: S413, 2006.
Timmons BW and Bar-Or O. Evidence of sex-based differences in natural killer cell
responses to exercise and carbohydrate intake in children. Eur J Appl Physiol 101:
233-240, 2007.
Timmons BW, Hamadeh MJ, Devries MC and Tarnopolsky MA. Influence of gender, menstrual phase, and oral contraceptive use on immunological changes in
response to prolonged cycling. J Appl Physiol 99: 979-985, 2005.
Timmons BW, Tarnopolsky MA and Bar-Or O. Immune responses to strenuous exercise and carbohydrate intake in boys and men. Pediatr Res 56: 227-234, 2004.
Timmons BW, Tarnopolsky MA and Bar-Or O. Sex-based effects on the distribution
of NK cell subsets in response to exercise and carbohydrate intake in adolescents. J
Appl Physiol 100: 1513-1519, 2006.
NK and exercise • 23
44.
45.
46.
47.
Timmons BW, Tarnopolsky MA, Snider DP and Bar-Or O. Puberty effects on NK
cell responses to exercise and carbohydrate intake in boys. Med Sci Sports Exerc 38:
864-874, 2006.
Weissgerber TL and Wolfe LA. Physiological adaptation in early human pregnancy:
adaptation to balance maternal-fetal demands. Appl Physiol Nutr Metab 31: 1-11,
2006.
Wendt K, Wilk E, Buyny S, Buer J, Schmidt RE and Jacobs R. Gene and protein
characteristics reflect functional diversity of CD56dim and CD56bright NK cells. J
Leukoc Biol 80: 1529-1541, 2006.
Westermann J and Pabst R. Distribution of lymphocyte subsets and natural killer
cells in the human body. Clin Investig 70: 539-544, 1992.
24 • Exercise, Th cell and aging
Effect of moderate exercise training on T-helper cell
subpopulations in elderly people
Kazuhiro Shimizu 1,2, Fuminori Kimura 1, Takayuki Akimoto 3,4, Takao
Akama2, Kai Tanabe1, Takahiko Nishijima1, Shinya Kuno1, and Ichiro Kono1
1
Graduate School of Comprehensive Human Sciences, University of Tsukuba,
Ibaraki, Japan
2 Faculty of Sport Sciences, Waseda University, Saitama, Japan
3 Laboratory of Regenerative Medical Engineering, Center for Disease Biology
and Integrative Medicine, Graduate School of Medicine, University of Tokyo,
Tokyo, Japan
4 Institute for Biomedical Engineering Consolidated Research Institute for
Advanced Science and Medical Care, Waseda University, Tokyo, Japan
ABSTRACT
CD28 molecule expression on the surface of T cells plays a critical role in upregulation of various cytokines synthesis and T-helper (Th) cell proliferation and
differentiation. However, aging induces a decrease in CD28 expression and unbalance of Th1/Th2, leading to impairment of Th-cell mediated immune function.
The purpose of this study was to assess the effects of moderate exercise training
on CD28 expression and the balance of Th1/Th2 cells in elderly people. Fortyeight elderly subjects were assigned to an exercise training group (EXC: 13
males, 15 females; aged 61–76) or a non-exercise control group (CON: 7 males,
13 females; aged 62–79). Subjects in EXC participated in exercise sessions 5days a week for 6 months. Meanwhile, subjects in CON maintained their normal
physical activity levels during the study period. Blood samples were collected
before and after the training period. Samples were measured for the number of
leukocytes and lymphocytes, as well as for CD3+, CD4+, CD28+CD4+, IFNγ+CD4+, IL-4+CD4+ cells. The number of leukocytes, lymphocytes, and CD3+
cells did not change after 6 months in both EXC and CON. The number of CD4+
and CD28+CD4+ cells significantly increased after the training in EXC (P <
0.05), while CON did not show significant changes. In the EXC group, IFNγ+CD4+ cell numbers were significantly higher following the training (P < 0.05),
but the number of IL-4+CD4+ cells was not changed. In the CON group, there
Address for correspondence:
Ichiro Kono, M.D., Ph.D.
Graduate School of Comprehensive Human Sciences, University of Tsukuba
1-1-1 Tennodai, Tsukuba, Ibaraki 305-8574, Japan
E-mail: kono@taiiku.tsukuba.ac.jp
Tel: +81 29 853 2656, Fax: +81 29 853 2656
Exercise, Th cell and aging • 25
were no significant alterations in IFN-γ+CD4+ and IL-4+CD4+ cell numbers. In
conclusion, moderate exercise training in the elderly is associated with improvement of expression of CD28 on Th cells and Th1/Th2 balances. Therefore, exercise training could up-regulate Th cell-mediated immune functions and be helpful
for a decrease in the risk of infections and autoimmune diseases in elderly people.
Keywords: CD28; Th1; Th2; aging; exercise
INTRODUCTION
Human immune function undergoes distinctly adverse changes with aging that
might be explained by decreased function of, or diminished regulation of, the
immune system (34). This immune senescence potentially leads to an increased
susceptibility to infectious diseases, malignancy, and autoimmune disorders in
elderly individuals (25).
During this immune senescence, as the thymus involutes, T cells, which
have a central role in cellular immune function, show the largest age-related alterations in distribution and function (3, 11, 38). One of the most important alterations in T cell profiles with aging is declined expression of CD28 (40). CD28 is a
homodimeric immunoglobulin super-family protein expressed on the surface of T
cells (21). Ligation of CD28 with its cognate receptor on antigen-presenting cells
is both necessary and sufficient, concomitantly with T cell receptor (TCR) signaling, to induce the production of interleukin-2 (IL-2) and the expression of the IL2 receptor (IL-2R), leading to T cell proliferation (9, 15). Thus, CD28 expression
and/or function in T cells with aging can significantly affect overall immune function. In fact, T cells lacking CD28 are detected in patients with autoimmune diseases such as rheumatoid arthritis and HIV-1 infection (10). Also, CD28-deficient
mice are susceptible to infection with Pneumocystis (5). Absence of the CD28
expression may be a contributing factor to the increased incidence of infections
and autoimmune diseases in elderly people.
A certain sub-population of T cells such as the T helper (Th) cell also shows
notable alteration with aging. The features of this alteration are characterized by
the decreased absolute number of circulating Th cells (39) and by functional
changes, including decreased expression of CD28 (40), decreased production of
Th1 cytokines (IL-2 and interferon (IFN)-γ), but increased production of Th2
cytokines (IL-4), leading to a shift towards a dominance of Th2 cytokine response
(1, 13, 30, 33). CD28 plays an essential role in the commitment of Th cells toward
Th1 or Th2 cells. Signaling through CD28 stimulates production of cytokines in
Th cells (7, 21). Thus, age-related alterations in cytokine production may be influenced in part by down-regulation of CD28 expression. It has been suggested that
the dysregulation in Th1/Th2 balance may contribute to an increased rate of infections in elderly people (31). Therefore, there may be important implications for
elderly individuals in regard to improvement of CD28 expression and Th1
cytokines production that is linked to the optimization of the Th1/Th2 balance.
In recent years, the effect of exercise on human immune function has
received considerable attention. Previous evidence suggested that moderate exercise training could increase the absolute numbers of T cells and Th cells in elderly
26 • Exercise, Th cell and aging
humans (19) and the concentration of cytokines, including IL-2 and IFN-γ in
older mice (16, 17). Since co-stimulation through CD28 enhances production of
IL-2 and IFN-γ in T cells activated by antigens and/or mitogens, there is a real
possibility that exercise may have an impact on CD28 expression. To date, there
has been only one published report about the effect of exercise training on the
expression of CD28 in healthy elderly subjects. Raso et al. (29) reported that 12
months of moderate resistance training undertaken by healthy elderly people did
not alter the number of CD28 expressing Th cells or other immune parameters,
such as distribution of T cell subsets and expression of IL-2R on T and Th cells.
The expression of IL-2R on T cells in elderly subjects significantly increased following 10 months of moderate endurance training, whereas 10 months of flexibility and resistance training did not alter IL-2R expression (18). Moreover, 12
months of moderate combined (endurance and resistance) training significantly
increased the absolute number of T cells and Th cells in elderly subjects (19).
These results indicated that long term endurance exercise interventions improved
T cell responses among elderly people.
Further, there have been several studies that examined the effects of exercise
training on the level of Th1 and/or Th2 cytokines in peripheral blood (16–18),
skeletal muscle (28) and lungs (22). It has been also reported that exhaustive exercise affects Th1 and Th2 cytokine producing cells in young athletes (20). However, there is only one study showing a relation between physical activity and
peripheral Th1 and Th2 cells in elderly people. Using a cross-sectional design,
Ogawa et al. (26) showed that exercise-trained (moderate endurance training) elderly subjects had higher IFN-γ+CD4+ (Th1) cell numbers compared with their
exercise-untrained peers, but that their IL-4+CD4+ (Th2) cells showed no difference. However, there has been no longitudinal study of the effect of exercise intervention on Th1 and Th2 cells in the elderly.
The goal of the present study was to determine the effects of 6 months of
moderate combined (endurance and resistance) training on CD28 expression and
Th1/Th2 balance of Th cells in elderly subjects. We hypothesized that moderate
exercise training undertaken by elderly subjects would increase the number of
CD28 expressing Th cells and Th1 cells, but would not alter the Th2 cells.
METHODS
Subjects.
Healthy, sedentary, elderly subjects who lived independently in Japan were
recruited through municipal advertisements into two groups: exercise training
group (EXC; 13 males, 15 females; aged 61–76) and non-exercise control group
(CON; 7 males, 13 females; aged 62–79). Potential subjects were given a detailed
explanation of the risks, stress, and potential benefits of the study before they
signed an informed consent form. Based on the results of medical examinations
within 6 months prior to the study and self-reported medical histories, the following exclusion criteria were determined for all subjects: hormone replacements,
acute illness from infection within the preceding 3 months, metabolic disorders,
and major surgery during the preceding 6 months. In addition, all subjects had to
have passed a complete medical examination within the past year and received
Exercise, Th cell and aging • 27
written permission from a sports doctor to be included in the study. No subjects
had been treated with any drugs that are known to affect immune function. EXC
subjects participated in an exercise program for a period of 6 months. We asked
CON subjects not to participate in any formal exercise but just to continue their
daily activities. All participants took part in the study for 6 months. This study,
which conforms to the principles outlined in the Declaration of Helsinki, was
approved by the Ethic Committees of the Institute of Health and Sport Sciences
and the Institute of Clinical Medicine of the University of Tsukuba.
Measurement of daily physical activity.
We used an electrical pedometer (Kenz Lifecorder; Suzuken Co. Ltd., Nagoya,
Japan) in order to assess the daily physical activity in elderly subjects. With
respect to this electrical pedometer, previous study showed accuracy for the
assessment of counting steps (32). Participants were instructed to wear an electrical pedometer for 14 consecutive days during all waking hours, except during
bathing before (PRE) and after (POST) the 6-month study period. Participants
were instructed to go about their normal lives unrestricted and were asked not to
look at the electrical pedometer to see how many steps they had taken each day.
Electrical pedometer placement was standardized on the belt or waistband,
according to the manufacturer’s recommendation.
Measurement of double-product break-point.
A double-product break-point (DPBP), which is the point of accelerating double
product (DP = heart rate, HR × systolic blood pressure, SBP), has been shown to
have strong positive correlations with the lactate and ventilatory thresholds (27).
As the method to measure DPBP is non-invasive and involves no excessive strain,
it is thought to be a useful index to monitor the intensity of endurance exercise in
elderly people. In this study, the DPBP was measured at PRE and POST, according to the procedures of a previous study (27). Subjects sat and rested for at least
5 min, and then they took a cycle-ergometer (232CXL, COMBI WELLNESS,
Tokyo, Japan) ramp loading exercise test. This test consisted of 4 min of cycling
at 20 W, followed by a ramp slope at 10 W every min. The test was stopped when
the subjects reached 75% of their predicted HR max (220 – age bpm). Their DP
with HR and brachial arterial SBP were measured and recorded every 15 s via an
automated sphygmomanometer (CM-4001, Kyokko, Tokyo, Japan). The DP was
calculated from the mean HR and SBP and then plotted against the work rate. The
DPBP was determined visually as the point at which a clear and sustained
increase of the DP slope occurred.
Physical fitness tests.
Subjects took six physical fitness tests at PRE and POST, as described in “Physical Fitness Test” by Japan Ministry of Education, Culture, Sports, Science and
Technology (35). The test measures six characteristics: isometric grip strength
(based on readings from a handgrip dynamometer), muscle endurance (based on
how many sit-ups the subject could do in 30 s), balance (based on how long the
subject, with open eyes, can stand on one leg), flexibility (based on a sit-andreach exercise), agility (based on the time a subject takes to walk over a 10-m
obstacle course), and endurance (based on a 6-min walking exercise).
28 • Exercise, Th cell and aging
Exercise program.
Subjects in the EXC group participated in exercise sessions 5-days a week for 6
months. They were supervised by experienced instructors, who conducted the
tests and also were responsible for measuring their HR. The training program
involved stretching for warm-up, endurance training, resistance training, and
stretching for cool-down. The endurance training was a cycle-ergometer exercise
(30 min) at 80% work rate of the DPBP. The resistance training that requires their
muscles to work against gravity by moving their own weight up and down. This
training comprised three sets of seven exercises (squat, trunk-curl, back-extension, leg-extension, hip-extension, leg-curl, and calf-raise) without using any
weights (10 repetitions). Subjects in the CON group did not participate in the
exercise sessions; during the study they simply maintained their normal levels of
physical activity.
Blood collection.
Blood samples were obtained in the morning (between 8:30 and 9:30) both PRE
and POST. Subjects refrained from any exercise for at least 24 hours before blood
sampling. Subjects came to our experimental laboratory without taking breakfast.
Samples were collected in vacutainers containing sodium EDTA. We quantified
total leukocytes and lymphocytes from whole blood samples by using a multichannel hemocyte analysis system (SE-9000, Sysmex, Hyogo, Japan).
Determination of lymphocyte sub-populations.
We used a whole-blood staining method (19) to label the lymphocytes with fluorescent-dye. The surface antibodies used for subset identification were CD3+ for
T cells, CD4+ for Th cells, and CD28+CD4+ for CD28+Th cells. Cell surfaces
were stained with three monoclonal antibodies: CD3 (FITC, clone: UCHT1,
DakoCytomation, Glostrup, Denmark), CD4 (APC, clone: 13B8.2, Immunotech,
Marseille, France), and CD28 (FITC, clone: CD28.2, BD Biosciences, San Jose,
USA). The mouse IgG1 antibody (clone: DAK-GO1, DakoCytomation) was used
as an isotypic control.
Determination of Th1 and Th2 cells.
Cells were stimulated and the population of cytokine-producing cells was determined by flow cytometry, using the method described in a previous study (1). The
surface and intracellular cytokine antibodies used for subset identification were
IFN-γ+CD4+ for Th1 cells and IL-4+CD4+ for Th2 cells. Intracellular cytokines
were stained using monoclonal antibodies: IFN-γ (PE, clone: B27, BD Biosciences) and IL-4 (FITC, clone: MP4-25D2, BD Biosciences). The mouse IgG1
antibody (clone: DAK-GO1, DakoCytomation) was used as an isotypic control.
Whole blood samples were stimulated with phorbol 12-myristate 13-acetate (50
ng/ml) and ionomycin (500 ng/ml) for 4 h at 37°C in presence of brefeldin A (10
μg/ml). The cells were incubated with anti-human antibody: CD4 (Immunotech).
The cells were fixed with a 4% formaldehyde buffer solution. The next day, cells
were incubated with 100 l of buffer solution containing 0.5% saponin to make
the cell membranes permeable. The cytokine antibodies were then added and
incubated.
Exercise, Th cell and aging • 29
Flow cytometry analysis.
Labeled cells were analyzed by flow cytometry using a fluorescence-activated
cell sorter analyzer (FACSCalibur, BD Biosciences). The usual quantity of cells
scanned was 10,000 cells per sample. The data were analyzed using the CELLQuest software (BD Biosciences), to determine proportions of fluorescentlabeled lymphocytes. Absolute numbers of cells in specific cell subsets were calculated using the total number of cells multiplied by the percentage of positive
cells within the subset of interest.
Statistical analysis.
All data were represented as means ± SE. For all analysis, P < 0.05 was considered statistically significant. Comparison between the EXC and CON groups for
the baseline criterion measures was made by a Student t-test. ANOVA for 2
(group, EXC and CON groups) × 2 (time, PRE and POST) repeated measures was
used to determine the effect of treatment during the 6 months period between
each group. A Tukey-Kramer post-hoc test was performed whenever there were
significant effects in ANOVA. Time effect of intervention within each group was
analyzed by a Student’s t-test.
RESULTS
Physical characteristics for the EXC and the CON group are presented in Table 1.
It can be seen that the EXC group and the CON group were of similar age and
body composition before the study period. Body mass and body mass index
(BMI) did not change significantly during the study period in either EXC or
CON.
Table 1. Descriptive data for EXC and CON groups before and after 6 months.
EXC (n=28)
CON (n=20)
Characteristics
PRE
Age (yr)
Height (cm)
Body weight (kg)
BMI (kg/m2)
68.5 ± 0.7
156.2 ± 1.6
60.1 ± 1.7
24.6 ± 0.5
POST
59.8 ± 1.8
24.4 ± 0.6
PRE
69.8 ± 1.1
153.7 ± 2.1
60.2 ± 2.4
25.5 ± 0.8
POST
60.9 ± 2.1
25.8 ± 0.8
Values are means ± SE. EXC, exercise-trained group; CON, control group; BMI, body mass index; PRE,
pre-training; POST, post-training.
With regards to physical activity, the mean value ± SE of step count per day
at PRE and POST were 8161 ± 774 and 9170 ± 779 step/day in EXC, and 5827 ±
805 and 6251 ± 705 step/day in CON. Step count per day in both two groups did
not change significantly after the study period. In EXC, the mean value ± SE of
the work rate at DPBP before and after the study period were 1.00 ± 0.04 and 1.05
± 0.05 W/kg. This rate did not change significantly following exercise training.
With regard to physical fitness tests in EXC, the mean value ± SE of each fitness
tests before and after the study period were as follows: grip strength test, 29.5 ±
1.7 and 30.7 ± 1.7 kg; sit-ups test in 30 s, 11.5 ± 1.4 and 12.8 ± 1.5 times; sit-and-
30 • Exercise, Th cell and aging
Table 2. The number of leukocyte, lymphocyte, CD3+ and CD4+ cells in peripheral blood of EXC (n=28)
and CON (n=20) groups before and after 6 months.
EXC
CON
Cells
PRE
POST
PRE
POST
Leukocyte (cells/μl)
Lymphocyte ((cells/μl)
5361 ± 231
1989 ± 127
5293 ± 224
2040 ± 125
5520 ± 271
1827 ± 90
5805 ± 220
1981 ± 109
CD3+ cell (%)
CD3+ (cells/μl)
58.7 ± 3.5
1170 ± 108
62.2 ± 93
280 ± 93
62.1 ± 2.3
1141 ± 79
62.3 ± 2.4
1246 ± 92
CD4+ cell (%)
CD4+ cell (cells/μl)
42.3 ± 3.0
831 ± 70
46.9 ± 1.9
958 ± 71*
43.3 ± 2.0
796 ± 57
42.7 ± 1.9
847 ± 61
Values are means ± SE. *Significant difference from PRE, P < 0.05
reaches test, 34.6 ± 1.8 and 40.9 ± 1.4 cm; standing on one leg with open eyes
test, 60.7 ± 8.0 and 71.1 ± 8.3 s; 10-m obstacle course test, 7.81 ± 0.24 and 6.55 ±
0.21 s; 6-min walking test, 508.3 ± 12.1 and 587.1 ± 15.7 m. The EXC group did
more sit-ups, more sit-and-reaches, and showed more endurance during the 6-min
walking test at POST than at PRE (P < 0.01). Time taken for the 10-m obstacle
walk was significantly reduced following
exercise training (P < 0.01). Therefore,
muscle endurance, flexibility, agility and
endurance in EXC could be improved by 6
months of exercise training.
As shown in Table 2, the subjects in
both the EXC and the CON groups had
similar leukocyte and lymphocyte numbers in whole blood before the study period. These numbers did not change significantly after the study period in either EXC
or CON.
As shown in Table 2, the percentage
and absolute number of CD3+ and CD4+
cells at PRE did not show any inter-group
differences between EXC and CON.
There was no significant group × time
interaction in percentage and absolute
number of CD3+ cells and CD4+ cells. The
percentage and absolute number of CD3+
cells did not change in either group after
the study period. Within the EXC group,
the absolute number of CD4 + cells
Figure 1. The percentage and absolute increased after exercise training (P <
number of peripheral blood CD4+ cells 0.05). Within the CON group, CD4+ cells
expressing CD28 before and after 6 did not show significant change.
months in EXC (n = 28) and CON (n =
Figure 1 shows the changes in the
20). Values are means ± SE. *Signifi- percentage and absolute number of
cant difference from PRE, P < 0.05.
CD28+CD4+ cells in both EXC and CON.
Exercise, Th cell and aging • 31
Figure 2. Differences in CD4+ cells in activated peripheral blood between EXC (n = 28)
and CON (n = 9) before and after 6 months. (A) The percentage of IFN-γ+CD4+ cells. (B)
The absolute number of IFN-γ+CD4+ cells. (C) The percentage of IL-4+CD4+ cells. (D)
The absolute number of IL-4+CD4+ cells. Values are means ± SE. *Significant difference
from PRE, P < 0.05.
The percentage and absolute number of CD28+CD4+ cells at PRE did not show
any inter-group differences between EXC and CON. There was significant group
× time interaction in the percentage of CD28+CD4+ cells (F = 6.59, P = 0.01).
EXC showed a significant increase in the percentage of CD28+CD4+ cells after
the training (P < 0.05), whereas CON did not show any significant change. There
was no significant group × time interaction in the absolute number of
CD28+CD4+ cells. Within the EXC group, the number of CD28+CD4+ cells was
significantly increased after exercise training (P < 0.05). Within the CON group,
the number of CD28+CD4+ cells did not show any significant change.
Figure 2 shows the changes in the percentage and absolute numbers of IFNγ+CD4+ (Th1) and IL-4+CD4+ (Th2) cells in both EXC and CON. The percentages and absolute number of IFN-γ+CD4+ cells at PRE were not significantly different between EXC and CON. The group × time interaction for percentage of
IFN-γ+CD4+ cells was close to significance (F = 3.08, P = 0.09). There was no
significant group × time interaction in the absolute number of IFN-γ+CD4+ cells.
Within EXC group, the percentage and absolute number of IFN-γ+CD4+ cells
32 • Exercise, Th cell and aging
were significantly increased after the training (P < 0.05). Within the CON group,
IFN-γ+CD4+ cells did not show significant change. The percentages and absolute
number of IL-4+CD4+ cells at PRE were not significantly different between EXC
and CON. There was not significantly group × time interaction in the percentage
and absolute number of IL-4+CD4+ cells. Also, the percentage and absolute number of IL-4+CD4+ cells did not change significantly after 6 months in both EXC
and CON.
DISCUSSION
The primary finding of our investigation was that 6 months of moderate combined
(endurance and resistance) training increased spontaneously CD28 expressing Th
cells and mitogens stimulated IFN-γ producing Th1 cells in elderly subjects.
These results suggest that regular moderate exercise training can bolster Th cellmediated immune responses and have an impact on Th1 cytokines, which contribute to the alteration of the Th1/Th2 balance in elderly people.
We focused on the CD28 molecule, which plays a critical role in orchestrating immune responses, including up-regulation of various cytokines synthesis and
Th cell proliferation (21). CD28 expression on Th cells is decreased with aging
(40). Thus, decreases in the level of CD28 expression contribute to degraded Th
cell function, leading to an increased incidence of infections and autoimmune diseases in elderly people (5, 10, 40). So, improvement of expression of CD28 on Th
cells may have important implications for the immune function of elderly individuals.
In our study, Th cells and CD28 expressing Th cells were significantly
increased in elderly subjects following moderate endurance and resistance training. Raso et al. (29) reported that moderate resistance training provided no benefits to healthy elderly subjects in regard to T cell subsets, and expression of CD28
and IL-2R. However, other investigators have reported that absolute numbers of T
cells and Th cells (19) and IL-2R expression on T cells (18) increased in healthy
elderly subjects following moderate combined (endurance and resistance) or
endurance training program. Both these studies may suggest that the effects of
exercise on CD28 expression, as well as on IL-2R expression, could depend on
exercise type: endurance exercise. Specifically, moderate endurance training or
combined training, which includes endurance exercise, could up-regulate CD28
expression on Th cells in elderly people.
The molecular mechanisms underlying the up-regulation of CD28 expression through exercise training have been unclear. Possible mechanisms might be
reactive oxygen species and pro-inflammatory cytokines such as tumor necrosis
factor-α (TNF-α). Age-related increases of oxidative stress and TNF-α level
down-regulate CD28 expression (8, 23). Previous studies suggested that regular
exercise training could reduce oxidative stress and TNF-α levels (4, 37). It is
therefore possible that exercise-induced decrease of chronic oxidative stress and
inflammation could be linked to up-regulation of CD28 expression in elderly people.
CD28 signaling induces the production of IL-2 and expression of IL-2R,
leading to Th cell activation and proliferation (9, 15). The expression of IL-2R
(CD25), which has been used as a marker of T cell activation along with CD28,
Exercise, Th cell and aging • 33
on T cells significantly increased following 10 months of moderate endurance
training in healthy elderly people (18). IL-2 production was also increased following endurance training in older mice (16, 17). In our study, the number of
CD28 expressing Th cells was significantly increased in elderly subjects following moderate exercise training. Therefore, exercise-induced increase of CD28
expression could be linked to up-regulated IL-2 production and IL-2R expression.
In our study, we did not determine IL-2 and IL-2R, which reflect T cell activity.
Further studies are needed to determine these parameters, along with the expression of CD28, so that the process of exercise-induced T cell activation can be
examined closely.
Ligation of CD28 is linked to up-regulation of IFN-γ and IL-2 production
(9, 13, 15, 21). Thus, impaired expression of CD28 with aging may down-regulate
these immune competences. It has been well documented that moderate
endurance training in older mice can bolster production of IFN-γ and IL-2 in
response to mitogens and viral challenges (16, 17). Our results also indicated that
the number of IFN-γ producing Th cells in response to mitogens significantly
increased following moderate exercise training. Th1 cytokines such as IFN-γ and
IL-2 drive T cell mediated immune responses, which are essential to eliminate
many viruses. Aging is associated with deficits in Th1 cytokine productions (1,
13, 30). Also, increased susceptibility to influenza in elderly people may be related to an impairment of influenza-specific T cell responses (6). Moderate exercise
training could increase CD28 expression, leading to the bolstering of T-cell mediated antiviral immunity in elderly individuals. It could help counter the age-associated decline in the potential of Th cells to produce Th1 cytokines such as IFN-γ
and IL-2.
The impact of age on Th1/Th2 cytokines production has been examined in
an effort to elucidate the possible mechanisms that underlie age-associated alterations in human immune function. Previous investigators suggested that aging
induces a shift towards Th2 cytokine dominance (1, 27, 33). Suppressor of
cytokine signaling 3 (SOCS3) protein in Th cells acts as negative regulator of
CD28-mediated IL-2 production and IL-12 signaling which induces IFN-γ secretion (14, 24). SOCS3 protein is increased with aging (12, 36). This enhancement
of SOCS3 as well as age-related decline of CD28 expression could down-regulate Th1 cytokines activity, leading to Th2 predominance (12, 36). In our study,
moderate exercise training resulted in the following: the number of IFN-γ producing Th (Th1) cells increased in parallel with CD28 expressing Th cells,
while the number of IL-4 producing Th (Th2) cells remained constant. These
results support data from a previous cross-sectional study of elderly subjects,
which revealed that IFN-γ producing Th cells were significantly higher in
endurance-trained elderly subjects than in untrained peers and that there was no
significant difference in IL-4 producing Th cells (26). One possible mechanism
of that exercise-induced immune response that includes an increase of Th1 cells
but no change in Th2 cells may be related to SOCS3 protein. However, no relationship has been elucidated between SOCS3 in Th cells and exercise training.
Further studies need to examine this relationship. Other possible mechanism of
that may be related to catecholamines. Kohut et al. (17) suggested that the
repeated increase in circulating catecholamines that occurs with each bout of
exercise may have a great impact on Th1 cells that produce IL-2 and IFN-γ, con-
34 • Exercise, Th cell and aging
sidering that Th1 cells express β2-adrenergic receptors, whereas Th2 cells do
not.
There could be several potential mechanisms underlying the exercise training-induced enhancement of CD28 expression and Th1 cell dominance that may
be intricately intertwined with one another. Additional research is required to
fully elucidate the contribution of potential mechanisms to changed CD28 expression and Th1/Th2 balance in response to moderate exercise training undertaken
by elderly subjects. If these mechanisms were clearly understood, more effective
health-related programming could be established to enhance the immune function
in elderly people.
The present study has the following study limitations. First, elderly subjects
were not randomly assigned to groups. In this study, subjects were, in part,
recruited from elderly people who belonged to each community group, so it was
hard to assign them randomly to exercise or non-exercise control groups. Further
studies need to have subjects in the control group engage in sham-training such as
mild flexibility and calisthenics under low-intensity and low-frequency. Second,
there was a relatively small sample size that limited our power to do analysis on
immune parameters. It is related to the stringent inclusion criteria and the difficulty of finding healthy, non-frail and sedentary elderly subjects who are willing and
unable to enroll in any other formal exercise program during 6 months. Third, the
numbers of male and female subjects were not equally represented in exercise and
control groups. Although the influence of gender difference on CD28 expression
in response to exercise in elderly people is unclear, a previous study reported that
female people had higher absolute number of CD4+ cells compared with male (2).
Future studies need to examine the effects of gender and aging on immune parameters including CD28 expression and Th1/Th2 in response to exercise.
In conclusion, we demonstrated that 6 months of moderate endurance and
resistance training for healthy elderly subjects significantly enhanced CD28
expressing Th cells and Th1 cells but no change in Th2 cells. Regular moderate
exercise training may enhance CD28 expression, leading to up-regulated
cytokines activity and Th cell proliferation and differentiation. Also, moderate
exercise training could have great impact on Th1 cytokines to change the Th1/Th2
balance. These findings can help prevent infections and autoimmune diseases in
elderly people, as well as improve their immune function as they age.
ACKNOWLEDGEMENTS
We thank all of the subjects for participating in this study. We also thank Drs. T.
Otsuki and K. Koizumi (University of Tsukuba) for critical comments, as well as
Dr. R. DiGovanni (Waseda University) for comments. This study was supported
by a grant from the Tsukuba Advanced Research Alliance (TARA) Project of the
University of Tsukuba and a Grant-in-Aid for Science Research from the Ministry
of Education, Culture, Sports, Science and Technology of Japan (13558003,
18650189 to T. A. and 19300228 to I. K.).
Exercise, Th cell and aging • 35
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Alberti, S., E. Cevenini, R. Ostan, M. Capri, S. Salvioli, L. Bucci, L. Ginaldi, M. De
Martinis, C. Franceschi, and D. Monti. Age-dependent modifications of Type 1 and
Type 2 cytokines within virgin and memory CD4+ T cells in humans. Mech. Ageing
Dev. 127: 560-566, 2006.
Amadori, A., R. Zamarchi, G. De Silvestro, G. Forza, G. Cavatton, G. A. Danieli, M.
Clementi, and L. Chieco-Bianchi. Genetic control of the CD4/CD8 T-cell ratio in
humans. Nat. Med. 1: 1279-1283, 1995.
Aspinall, R., and D. Andrew. Thymic involution in aging. J. Clin. Immunol. 20: 250256, 2000.
Avula, C. P., A. R. AMuthukumar, K. Zaman, R. McCarter, and G. Fernandes.
Inhibitory effects of voluntary wheel exercise on apoptosis in splenic lymphocyte
subsets of C57BL/6 mice. J. Appl. Physiol. 91: 2546-2552, 2001.
Beck, J. M., M. B. Blackmon, C. M. Rose, SL. Kimzey, A. M. Preston, and J. M.
Green. T cell costimulatory molecule function determines susceptibility to infection
with pneumocystis carinii in mice. J. Immunol. 171: 1969-1977, 2003.
Bernstein, E., D. Kaye, E. Abrutyn, P. Gross, M. Dorfman, and D. M. Murasko.
Immune response to influenza vaccination in a large healthy elderly population. Vaccine 17: 82-94, 1999.
Bian, Y., S. Hiraoka, M. Tomura, X. Y. Zhou, Y. Yashiro-Ohhtani, Y. Mori, J.
Shimizu, S. Ono, K. Dunussi-Joannopoulos, S. Wolf, and H. Fujiwara. The capacity
of the natural ligands for CD28 to drive IL-4 expression in naïve and antigen-primed
CD4+ and CD8+ T cells. Int. Immunol. 17: 73-83, 2005.
Bryl, E., A. N. Vallejo, C. M. Weyand, and J. J. Goronzy. Down-regulation of CD28
expression by TNF-alpha. J. Immunol. 167: 3231-3238, 2001.
Cerdan, C., Y. Martin, M. Courcoul, H. Brailly, C. Mawas, F. Birg, and D. Olive.
Prolonged IL-2 receptor α/CD25 expression after T cell activation via the adhesion
molecules CD2 and CD28. Demonstration of combined transcriptional and posttranscriptional regulation. J. Immunol. 149: 2255-2261, 1992.
Effros, B. R. Costimulatory mechanisms in the elderly. Vaccine 18: 1661-1665, 2000.
Engwerda, C. R., B. S. Handwerger, and B. S. Fox. Aged T cells are hyporesponsive
to costimulation mediated by CD28. J. Immunol. 152: 3740-3747, 1994.
Han, SN., O. Adolfsson, CK. Lee, TA. Prolla, J. Ordovas, and SN. Meydani. Age and
vitamin E-induced changes on gene expression profiles of T cells. J. Immunol. 177:
6052-6061, 2006.
Haynes, L., P. J. Linton, S. M. Eaton, S. L. Tonkonogy, and S. L. Swain. Interleukin
2, but not other common gamma chain-binding cytokines, can reverse the defect in
generation of CD4 effector T cells from naive T cells of aged mice. J. Exp. Med
.190: 1013-1024, 1999.
Inoue, H., and M. Kubo. SOCS proteins in T helper cell differentiation: implications
for allergic disorders? Expert Rev. Mol. Med. 6: 1-11, 2004.
Jenkins, M. K., P. S. Taylor, S. D. Norton, and K. B. Urdahl. CD28 delivers a costimulatory signal involved in antigen-specific IL-2 production by human T cells. J.
Immunol. 147: 2461-2466, 1991.
Kohut, M. L., G. W. Boehm, and J. A. Moynihan. Moderate exercise is associated
with enhanced antigen-specific cytokine, but not IgM antibody production in aged
mice. Mech. Ageing Dev. 122: 1135-1150, 2001.
36 • Exercise, Th cell and aging
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
Kohut, M. L., J. R. Thompson, W. Lee, J. E. Cunnick. Exercise training-induced
adaptations of immune response are mediated by beta-adrenergic receptors in aged
but not young mice. J Appl Physiol 96: 1312-1322, 2004.
Kohut, M. L., and D. S. Senchina. Reversing age-associated immunosenescence via
exercise. Exerc. Immunol. Rev. 10: 6-41, 2004.
Koizumi, K., F. Kimura, T. Akimoto, T. Akama, Y. Kumai, H. Tanaka, M. Ishizu, S.
Kuno, and I. Kono. Effects of long-term exercise training on peripheral lymphocyte
subsets in elderly subjects. Jpn. J. Phys. Fitness Sports Med. 52: 193-202, 2003.
Lancester, G. I., S. L. Halson, Q. Khan, P. Drysdale, F. Wallace, A. E. Jeukendrup, M.
T. Drayson, and M. Gleeson. Effects of acute exhaustive exercise and chronic exercise
training on type 1 and type 2 T lymphocytes. Exerc. Immunol. Rev. 10: 91-106, 2004.
Lenschow, D. J., T. L. Walunas, and J. A. Bluestone. CD28/B7 system of T cell costimulation. Annu. Rev. Immunol. 14: 233-258, 1996.
Lowder, T., D. A., Padgett, and J. A. Woods. Moderate exercise early after influenza
virus infection reduces the Th1 inflammatory response in lungs of mice. Exerc.
Immunol. Rev. 12: 97-111, 2006.
Ma, S., H. Ochi, L. Cui, J. Zhang, and W. He. Hydrogen peroxide induced down-regulation of CD28 expression of jurkat cells is associated with a change of site -specific nuclear factor binding activity and the activation of caspase-3. Exp. Gerontol.
38: 1109-1118, 2003.
Matsumoto, A., Y. Seki, R. Watanabe, K. Hayashi, J. A. Johnston, Y. Harada, R. Abe,
A. Yoshimura, and M. Kubo. A role of suppressor of cytokine signaling 3
(SOCS3/CIS3/SSI3) in CD28-mediated interleukin 2 production. J. Exp. Med. 197:
425-436, 2003.
Nieman, D. C., D. A. Henson, G. Gusewitch, B. J. Warren, R. C. Dotson, D. E. Butterworth, and S. L. Nehlsen-Cannarella. Physical activity and immune function in
elderly women. Med. Sci. Sports Exerc. 25: 823-831, 1993.
Ogawa, K., J. Oka, J. Yamakawa, and M. Higuchi. Habitual exercise did not affect
the balance of type 1 and type 2 cytokines in elderly people. Mech. Ageing Dev. 124:
951-956, 2003.
Omiya, K., H. Itoh, N. Harada, T. Maeda, A. Tajima, K. Oikawa, A. Koike, T. Aizawa, T. Fu, and N. Osada. Relationship between double product break point, lactate
threshold, and ventilatory threshold in cardiac patients. Eur. J. Appl. Physiol. 91:
224-229, 2004.
Prokopchuk, O., Y. Liu, L. Wang, K. Wirth, D. Schmidtbleicher, and J. M. Steinacker. Skeletal muscle IL-4, IL-4 Ralpha, IL-13 and IL-13 Ralpha 1 expression and
response to strength training. Exerc. Immunol. Rev. 13: 67-75, 2007.
Raso, V., G. Benard, A. J. Da Silva Duarte, and V. M. Natale. Effect of resistance
training on immunological parameters of healthy elderly women. Med. Sci. Sports
Exerc. 39: 2152-2159, 2007.
Rea, I. M., M. Stewart, P. Campbell, H. D. Alexander, A. D. Crockard, and T. C.
Morris. Change in lymphocyte subsets, interleukin 2, and soluble interleukin 2
receptor in old and very old age. Gerontology 42: 69-78, 1996.
Rink, L., I. Cakman, and H. Kirchner. Altered cytokine production in the elderly.
Mech. Ageing Dev. 102: 199-209, 1998.
Schneider, P. L., S. E. Crouter, O. Lukajic, and D. R. Jr. Bassett. Accuracy and reliability of 10 pedometers for measuring steps over a 400-m walk. Med. Sci. Sports
Exec. 35: 1779-1784, 2003.
Exercise, Th cell and aging • 37
33.
34.
35.
36.
37.
38.
39.
40.
Shearer, G. M. Th1/Th2 changes in aging. Mech. Ageing Dev. 94: 1-5, 1997.
Shephard, R. J., and P. N. Shek. Exercise, aging and immune function. Int. J. Sports
Med. 16: 1-6, 1995.
The Japan Ministry of Education, Culture, Sports, Science and Technology. Physical
Fitness Test. Tokyo: Gyosei, 2000, pp. 1-135.
Tortorella, C., I. Stella, G. Piazzolla, V. Cappiello, O. Simone, A. Pisconti, and S.
Antonaci. Impaired interleukin-12-dependent T-cell functions during aging: role of
signal transducer and activator of transcription 4 (STAT4) and suppressor of
cytokine signaling 3 (SOCS3). J. Gerontol. A Biol. Sci. Med. Sci. 61: 125-135,
2006.
Tsukui, S., T. Kanda, M. Nara, M. Nishino, T. Kondo, and I. Kobayashi. Moderateintensity regular exercise decreases serum tumor necrosis factor-alpha and HbA1c
levels in healthy women. Int. J. Obes. Relat. Metab. Disord. 24: 1207-1211, 2000.
Utsuyama, M., M. Kasai, C. Kurashima, and K. Hirokawa. Age influence on the
thymic capacity to promote differentiation of T cells: induction of different composition of T cell subsets by aging thymus. Mech. Ageing Dev. 58: 267-277, 1991.
Utsuyama, M., K. Hirokawa, C. Kurashima, M. Fukayama, T. Inamatsu, K. Suzuki,
W. Hashimoto, and K. Sato. Differential age-change in the numbers of
CD4+CD45RA+ and CD4+CD29+ T cell subsets in human peripheral blood. Mech.
Ageing Dev. 63: 57-68, 1992.
Vallejo, A. N. CD28 extinction in human T cells: altered functions and the program
of T-cell senescence. Immunol. Rev. 205: 158-169, 2005.Table 2. The number of
leukocyte, lymphocyte, CD3+ and CD4+ cells in peripheral blood of EXC (n = 28)
and CON (n = 20) groups before and after 6 months.
38 • Salmonella induces reduction of wheel-running activity via TLR5
Salmonella administration induces a reduction of
wheel-running activity via a TLR5-, but not a TLR4,
dependent pathway in mice
Takashi Matsumoto1, Daisuke Shiva2, Noriaki Kawanishi3, Yasuko Kato4,
Jeffrey A. Woods5 and Hiromi Yano3*
1
Microbiology, Department of Infectious Disease, Faculty of Medicine, Oita
University, Oita 879-5593, Japan
2 Institute for Biomedical Engineering, Consolidated Research Institute for
Advanced Science and Medical Care, Waseda University, Tokorozawa, Saitama
379-1192, Japan.
3 Department of Health and Sports Science, and 4 Department of Clinical Nutrition, Kawasaki University of Medical Welfare, Kurashiki, Okayama 701-0193,
Japan
5 Department of Kinesiology and Community Health, University of Illinois at
Urbana-Champaign, Urbana, IL 61801, USA
ABSTRACT
In general, systemic bacterial infections induce sickness behavior. In mice, lipopolysaccharide (LPS), a component of gram-negative bacteria, strongly reduces
physical activity via toll-like receptor (TLR) 4. However, gram-negative bacteria,
such as Salmonella, also express flagella containing flagellin (FG) which binds to
TLR5 and induces pro-inflammatory cytokine production. It is unclear whether
FG induces sickness behavior. To determine whether Salmonella administration
regulates the reduction of voluntary physical activity in mice, male C3H/HeN
(wild type) and C3H/HeJ (tlr4 gene mutated) mice were administered living Salmonella (live) and examined for wheel-running activity. The production of TNF-α
in RAW 264 cells was measured by the ELISA assay under both live and heat-killed (HK) Salmonella conditions in vitro. Wheel-running activity in both C3H/HeJ
and C3H/HeN mice after i.p. injection of live Salmonella (1x106 CFU/kg) was
significantly lower than that in vehicle groups (p<0.01, respectively), although
wheel-running activity in C3H/HeJ mice was not reduced after i.p. injection of
HK Salmonella (1x106 CFU/kg). Furthermore, TNF-α production from RAW 264
cells with HK Salmonella treatment at the early phase was higher than that with
live Salmonella treatment. Interestingly, gentamicin-treated (GMT) Salmonella,
(which have bacterial flagella removed), did not induce reduction of wheel-running activity, although injection of the flagella-rich supernatant of GMT Salmo*Address Correspondence to:
Hiromi Yano, Ph.D., Department of Health and Sports Science,
Kawasaki University of Medical Welfare
288 Matsushima, Kurashiki, Okayama 701-0193, Japan
Tel:+81-86-462-1111 (ex.54835), Fax: +81-86-464-1109
E-mail: yanohiro@mw.kawasaki-m.ac.jp
Salmonella induces reduction of wheel-running activity via TLR5 • 39
nella significantly reduced it (p<0.01). Indeed, FG treatment also induced reduction of wheel-running activity in mice (p<0.01). Our findings suggest that the Salmonella-induced reduction of voluntary physical activity might be regulated by
FG via TLR5, but not LPS via TLR4 in mice.
Keywords: Toll-like receptors TLR4, TLR5, voluntary physical activity, flagellin, lipopolysaccharide, C3H/HeJ mouse
INTRODUCTION
Salmonella, which is a gram-negative bacterium, can invade and cause enteritis,
systemic infection and fever (3), and it possesses a range of protein [e.g. lipoprotein and flagellin (FG)] and nonprotein [e.g. lipopolysaccharide (LPS), peptidoglycan (PGN) and CpG DNA] structures that function as pathogen-associated
molecular patterns (PAMPs) (1). These PAMPs are recognized by the family of
toll-like receptor (TLRs) on host mammalian cells which signal host cells to
induce a response (1). It is well known that LPS, a component of the cell wall of
gram-negative bacteria (34), is a main mediator of pro-inflammatory cytokine
production (1) and quickly induces a range of sickness behaviors in animals
(12,17,19,45). LPS typically consist of lipid A, inner and outer cores, and O-antigen, which is the main component of the outer leaflet of the outer membrane of
Salmonella (32,35). However, the role of LPS versus other PAMPs in the induction of sickness behavior and pro-inflammatory cytokines induced by living bacterium such as Salmonella has not been systematically investigated. In fact, Royle
et al. (37) reported that administration of a lipid A antagonist prior to live Salmonella exposure had no effect on tumor necrosis factor (TNF)-α release from
macrophages. In addition, it is also known that changes in gene expression are
generally greater in cells treated with LPS than in those infected by living bacteria
(36,37). Bacteria of the Salmonella family produce a number of specialized effector proteins that can modify host cell signaling (14). Furthermore, the lipid A portion of LPS, which is responsible for the majority of immunomodulating activity
of LPS (27,30), is not expressed on outer bacterial membranes, because lipid A
locates between PGN and the outer cores/O-antigen in the Salmonella cell wall
(32,34,35). Therefore, it might be hard for host cells to recognize lipid A in the
LPS of living bacteria via TLR4. Accordingly, we hypothesized that PAMPs other
than LPS might contribute to behavioral changes following bacterial infection.
FG (the major structural protein of flagella of gram-negative bacteria) has
recently been appreciated as a major factor contributing to the host inflammatory
response to bacteria (9,13,23). FG induces an inflammatory and innate immune
response through activation of TLR5 and is known to be essential for the pathogenesis of many gastrointestinal, respiratory and renal tract bacteria (16). Signaling from FG/TLR5 as well as LPS/TLR4 activates the MyD88 dependent pathway, which sequentially activates IL-1R-associated kinase (IRAK), TNFR-associated factor 6 (TRAF6), nuclear factor kappa B (NF-κB), resulting in the induction
of genes involved in inflammatory responses (1). Although factors that regulate
sickness behavior via TLRs are poorly understood, pro-inflammatory cytokines
[such as interleukin (IL)-1β, IL-6 and TNF-α], prostaglandin and other mole-
40 • Salmonella induces reduction of wheel-running activity via TLR5
cules, which are triggered by NF-κB activation are known to play a role in the
behavioral partiality effect of infection (6,7,15,45). In fact, it was reported that
intravenous FG caused a systemic inflammatory response (10,46). Therefore, in
addition to LPS, it is possible that FG may contribute to the reduction of physical
activity following gram-negative bacterial infection. Our hypothesis was that FG,
interacting through TLR 5, contributes to the reduction in voluntary wheel-running behavior following Salmonella infection. The purpose of the present study
was to determine the components of Salmonella that regulate the reduction of voluntary physical activity in mice.
MATERIALS AND METHODS
Animals
Male 8-9-week-old C3H/HeN (wild type, n = 55) and C3H/HeJ (tlr4-gene mutation, n= 72) mice (Clea Japan, Tokyo, Japan) were used in these experiments. The
mice were housed individually in cages with a running wheel (10 x 23 x 10 cm
cage with 5.5 wide x 22 cm ø wheel, Natsume, Nagano, Japan) that was accessible 24 hours per day. The animals were under a controlled environment (20 ± 1˚C,
12:12-h light-dark cycle) and allowed unrestricted access to standard chow and
tap water. The experimental procedures followed the guiding principles for the
care and use of animals in the field of physiological sciences approved by the
Council of the Physiological Society of Japan.
Cell culture
RAW 264 cells, a mouse-derived macrophage cell line, were obtained from the
Cell Bank RIKEN Bioresource Center (Ibaraki, Japan). These cells were cultured
in DMEM containing 10% FCS supplemented 200 U/ml penicillin and 100 μg/ml
streptomycin at 37˚C in 5% CO2.
Bacteria and their treatment
Salmonella enterica (serovar Dublin) was provided by Professor Hiroko Mine of
the Department of Clinical Nutrition, Kawasaki University of Medical Welfare.
The bacteria were grown for 48 h at 35˚C in a brain heart infusion agar (Nissui,
Tokyo, Japan) and diluted in sterile physiological saline as live bacteria (Live,
1×103-5 CFU/ml) or as heat-killed (HK, 1×103-5 CFU/ml) Salmonella, which were
treated for 2 h at 62˚C. In some experiments, live Salmonella were treated for 1 hr
with gentamicin (GM, 100 μg/ml) in RPMI 1640 medium at 35°C. After GM
treatment, the medium containing Salmonella was centrifuged at 13,000 g for 5
min and then the supernatants were filtered through a 0.22 μm membrane to
remove bacteria and or any cellular debris. The purity was confirmed by SDSPAGE and Coomassie blue staining. The flagella-based motility of Salmonella
was examined by a motility test. Briefly, Salmonella were oscillating-cultured for
5 hr with brain heart infusion medium containing 0.3% agar at 35ºC (39). Western
blot analysis was performed to examine flagellin (FG) content of GM treated Salmonella and supernatant of GM treated Salmonella. Briefly, following electrophoresis, the gel was transferred to a cellulose nitrate membrane filter and
excessive proteins in the uncombined part of the membranes were saturated with
Salmonella induces reduction of wheel-running activity via TLR5 • 41
3% BSA/TBS (bovine serum albumine/tris–buffered saline, pH 7.5) overnight at
4˚C. Then the membrane was soaked in purified anti-Flagellin mouse IgG1 antibody (Biolegend, San Diego, CA) with 1%BSA/TBS solution overnight at 4˚C
(1:500). After the membranes were washed, they were soaked in POD-linked goat
IgG to mouse IgG (Nordic Immunological, Tilburg, Netherlands) with
1%BSA/TBS solution for 1 hr at room temperature (1:1,000). After the membrane were washed again, they were stained with 6 μl 30% H2O2/10 ml TBS
including 6 mg 4-chloro-1-naspthol / 2 ml ice-cold methanol with light shielding.
LPS and FG
LPS and FG (Salmonella Typhimurium) were all obtained from Sigma (St. Louis,
MO). LPS was diluted in sterile physiological saline to a final concentration of
0.5 mg/ml. The same lot and dilution of LPS were used for all experiments.
Experiment 1. Effect of Salmonella infection on wheel-running activity in
C3H/HeN and C3H/HeJ mice.
C3H/HeN (n=16) and C3H/HeJ (n=16) mice were randomly assigned to one of
two groups (n= 8 per each group): PBS (200 μl per mice as vehicle) or live Salmonella (Live, 1×106 CFU/kg) administered i.p. under light isoflurane anesthesia.
Wheel-running activity in both groups of mice was examined by observing their
running performance in a cage-adjacent wheel for 24 hr after injection. The
experimental procedure was started between 12:30 and 13:00 hours to reduce the
variability associated with diurnal rhythms (24,45).
Experiment 2. Effect of HK Salmonella on wheel-running activity.
Male C3H/HeN (n=24) and C3H/HeJ (n=24) mice were randomly assigned to one
of three groups (n= 8 per each group): sterile phosphate-buffer saline (PBS, 200
μl per mice as vehicle), LPS (0.5 mg/kg) or HK Salmonella (0.5 mg/kg). Each
mouse was lightly anesthetized with inhalant Isoflurane prior to the i.p. injection.
The experimental procedure was conducted between 12:30 and 13:00 hours to
reduce the variability associated with diurnal rhythms. Wheel-running activity in
all groups was examined by observing their running performance in a cage-adjacent wheel for 24 hr after the injection.
Experiment 3. Effect of HK and live Salmonella on TNF-α production from
macrophages in vitro.
Raw 264 cells (5x104/ well) in 96 well plates were pre-incubated for 24 hr and
then were stimulated for 0-12 hr with PBS, HK Salmonella (0-1,000 CFU/well)
or live Salmonella (0-1,000 CFU/ well). After the stimulation, the supernatants
were collected and then stored at -80˚C until analysis of TNF-α via ELISA.
Experiments 4 and 5. Effect of flagella on wheel-running activity.
Male C3H/HeJ (n=32) mice were randomly assigned to one of four groups: PBS
as vehicle (200 μl, n=9), Salmonella without gentamicin treatment (NT, 1x106
CFU/kg, n=11), Salmonella with GM treatment (GMT, 1x106 CFU/kg, n=6) and
the supernatant of GM-treated Salmonella (SG, 25 mg/kg, n=6). Each mouse was
lightly anesthetized with inhalant Isoflurane prior to the i.p. injection. The experimental procedure was conducted between 12:30 and 13:00 hours. Voluntary phys-
42 • Salmonella induces reduction of wheel-running activity via TLR5
ical activity was examined by
observing running performance in
cage-adjacent wheels for 24 hr
after i.p. injection. Moreover, we
also tested the effects of FG alone
on wheel running activity in
C3H/HeN mice. Male C3H/HeN
(n=15) mice were randomly
assigned to one of two groups: PBS
as vehicle (200 μl, n=7) or FG
(1 mg/kg, n=8). Wheel-running
activity in both groups of mice was
examined by observing their running performance in cage-adjacent
wheels for 24 hour after PBS or FG
injections. The experimental procedure was also started between
12:30 and 13:00 hours. In addition,
body weight was measured before
Fig. 1. Effect of live Salmonella injection (Live)
and 24 hr after their treatments.
6
(1 x 10 CFU/kg) on wheel-running activity (24
hr post-injection) in C3H/HeJ) and C3H/HeN
ELISA for TNF-α
mice. The values are expressed the mean ±
TNF-α was measured by an S.E.M. **p<0.01, n=8 in each group.
enzyme-linked immunosorbent
assay (ELISA) using a commercially available kit (R&D Systems,
Minneapolis,
MN).
The
absorbance was measured at 450
nm and was proportional to the
concentration of TNF-α in the
sample. The minimum detectable
dose of mouse TNF-α was typically less than 5.1 pg/ml.
Statistics
Data are expressed as the means ±
S.E.M. Statistical analyses were
performed using an analysis of
variance procedure (ANOVA) by
Stat View for Windows version 5.0.
Fisher’s protected least-significant
difference test was used for post
hoc analyses. P values of <0.05
were considered statistically signif- Fig. 2. Effect of LPS (0.5 mg/kg) or HK
(0.5 mg/kg) Salmonella injection on wheel-runicant.
ning activity (24 hr post-injection) in C3H/HeJ
(tlr 4 mutation) and C3H/HeN mice. The values
are expressed the mean ± S.E.M. **p<0.01, n=8
in each group.
Salmonella induces reduction of wheel-running activity via TLR5 • 43
RESULTS
Experiments 1 and 2. Effects of LPS and Salmonella infection on wheel-running activity in C3H/HeN and C3H/HeJ mice.
After treatment with live Salmonella, both C3H/HeN (intact tlr 4 signaling) and
C3H/HeJ (tlr 4 mutated) mice exhibited significantly reduced wheel-running
activity when compared to vehicle controls (p<0.01 and p<0.01, respectively, Fig.
1). The level of wheel-running activity was not significantly different between
C3H/HeN and C3H/HeJ mice after the injection of live Salmonella. These data
indicate that tlr 4 is not required to induce a reduction in wheel running after Salmonella infection. To verify this, and to demonstrate LPS non-responsiveness in
our C3H/HeJ mice, we administered HK Salmonella (heat-killing breaks apart
bacterial cell wall components exposing Lipid A for better binding to tlr 4) and
LPS to both strains of mice (Fig. 2). In C3H/HeN mice, wheel-running activity in
the LPS- and HK-treated groups was greatly reduced when compared to vehicle
controls (p<0.01 and p<0.01, respectively, Fig. 2). Interestingly, while LPS failed
to reduce wheel running in tlr 4 deficient C3H/HeJ mice as expected, administration of HK Salmonella did not affect wheel-running activity in C3H/HeJ mice;
indicating that heat-labile structures (not LPS acting through tlr 4) may be
responsible for reduced wheel running in response to Salmonella infection in this
strain (Fig. 2).
Experiment 3. Effect of heat-killed and live Salmonella on TNF-α production
from macrophages in vitro.
Contribution of pro-inflammatory cytokines, such as TNF-α, to various sickness
behaviors in mice has been documented (28). Therefore, to examine the extent to
which LPS and live and HK Salmonella induced macrophage TNF-α production,
Fig. 3. Effects of HK or live Salmonella (Live) challenge on in vitro production of TNF-α
in RAW264 macrophages. RAW264 cells (5×104 cell/well) were stimulated with HK (100
CFU/ml) or live Salmonella (100 CFU/ml) for 0, 1, 3, 6 or 12 hr (Fig. 3A). Cells were also
stimulated with 0, 1, 10, 100 or 1,000 CFU/ml of HK, Live, LPS (1 μg/ml) or PBS for 3 hr
(Fig. 3B) and TNF-α secretion was determined by ELISA. The values are expressed mean
± S.E.M. of two separate experiments.
44 • Salmonella induces reduction of wheel-running activity via TLR5
Fig. 4. Effect of gentamicin-treated Salmonella administration on wheel-running activity in
C3H/HeJ mice. Representative photograph of triplicate individual experiments where Salmonella was treated with the following doses of gentamycin (I = 0 mg/ml, II = 1 mg/ml, III
= 10 mg/ml, IV = 100 mg/ml and tested for motility in agar (Fig. 4A). Representative (of
triplicate gels) SDS-PAGE analysis (stained with Coomassie Blue for 10 min) of gentamicin-treated Salmonella (GMT, 20 mg) and the supernatant of gentamicin-treated Salmonella (SG, 25 mg) after centrifugation (Fig. 4B). M and FG represent the molecular weight
marker and a positive flagellin (25 mg, Sigma) control, respectively. Representative (of 3
separate blots) Western blot analysis of flagellin in GMT and SG (Fig. 4C). Voluntary
wheel-running activity in C3H/HeJ mice for 24 hr after i.p. injection with PBS vehicle
(200 μl, n=9), Salmonella without gentamicin treatment (NT, 1x106 CFU/kg, n=11), GMT
(1x106 CFU/kg, n=6) or SG (25 mg/kg n=6). The values are expressed as mean ± S.E.M.
**p<0.01 respectively (Fig. 4D).
we performed an in vitro experiment using RAW 264 macrophages. TNF-α production increased after incubation of macrophages with both live and HK Salmonella (Fig. 3A), but was more rapid in the HK condition peaking at 3 hr post vs.
12 hr post with live Salmonella. In addition, RAW 264 cells were cultured with
HK Salmonella (0-1,000 CFU/ml) and live Salmonella (0-1,000 CFU/ml) for 6 hr
(Fig. 3B). At low the low dose (1 CFU/ml), live Salmonella , TNF-α production
was not different from HK Salmonella. However, at higher doses (10-1,000
CFU/ml) HK Salmonella (10 and 100 CFU/ml) induced greater TNF-α produc-
Salmonella induces reduction of wheel-running activity via TLR5 • 45
tion when compared to live Salmonella indicating that HK treatment increases the
potency of the proinflammatory effect of Salmonella.
Experiments 4 and 5. Effect of flagellin on wheel running activity.
GM (a bacterial protein synthesis inhibitor) treatment dose-dependently reduced
Salmonella motility (Fig. 4A). SDS-PAGE analysis (Fig. 4B) and western blot
analysis (Fig. 4C)
confirmed that GM
treatment
(GMT)
reduced FG in Salmonella, but not in the
supernate (SG) of Salmonella
cultures.
Interestingly, untreated Salmonella (NT)
and SG significantly
reduced wheel running activity 24 hr
post administration
(p<0.01), whereas
GMT Salmonella did
not (Fig. 4D), indicatFig. 5. Wheel-running activity (Fig.5A) and loss of body weight ing that FG plays an
(24 hr post) (Fg.5B) in C3H/HeN mice after flagellin (FG, 1 important role in the
mg/kg, intravenous injection of 200 μl) injection. The values Salmonella-induced
are expressed as mean ± S.E.M. **p<0.01 v.s Vehicle, respec- attenuation of wheeltively. n=7-8 in each group.
running behavior. Furthermore, both wheelrunning activity and body weight were significantly reduced in mice after intravenous FG injection (p<0.01, Fig. 5A and p<0.01, Fig.5B, respectively).
DISCUSSION
Systemic bacterial infection results in a vigorous pro-inflammatory cytokine
response and various sickness-related behaviors including reduced food intake
and lethargy (20,25). Because of the widespread use of LPS as a model for gramnegative bacterially-induced physiological and behavioral changes (5), little
attention has been paid to other bacterial structures that could result in proinflammatory responses and altered sickness behavior. Accordingly, we hypothesized that Salmonella FG, which binds to TLR 5 on host cells (1), also has the
ability to promote inflammation and sickness behaviors. Our results clearly indicate that, along with LPS, FG contributes to the reduction in wheel-running activity after Salmonella infection.
We demonstrated that Salmonella infection attenuated voluntary wheel-running in C3H/HeJ mice. This is significant because this strain of mouse exhibits a
point mutation in TLR 4 (31), rendering it incapable of responding to LPS with
altered behavior (5). However, the strain still possesses the ability to recognize
46 • Salmonella induces reduction of wheel-running activity via TLR5
bacteria ligands via TLR2, TLR5, and TLR9. LPS activation of TLR4 triggers the
biosynthesis of diverse mediators of inflammation and activates the production of
costimulatory molecules required for the adaptive physical behavior (1). Indeed,
many previous studies suggest that LPS, which is made up of an outer monolayer
on the outer membranes of most gram-negative bacteria (33), induces reduction of
physical activity (12,17,19,45). Therefore, data from this C3H/HeJ experiment
suggest that bacterial components other than LPS must have been responsible for
reduced wheel-running behavior. We also subjected both strains of mice to HK
Salmonella. Interestingly, heat-killing increased the suppressive effect of Salmonella (when compared to administration of viable Salmonella) on wheel-running
activity in C3H/HeN (TLR 4 intact) mice, while having no effect on running
behavior in C3H/HeJ mice. The former result is consistent with our idea that heat
denaturation, which induces rapid and extensive killing of bacteria, induces
release of bacterial cell wall components including LPS (21). Indeed, VazquezTorres et al. (44) reported that HK Salmonella treatment increased INF-γ staining
of CD4+ T cells in C3H/HeN mice, but the adaptive cellular immune responses to
HK Salmonella were attenuated in C3H/HeJ mice. Moreover, it is also known that
macrophages respond better to nonmotile, killed bacteria than to living or motile
bacteria (36,37). Our results demonstrated that induction of a decrease in physical
activity after HK Salmonella injection in C3H/HeN mice most probably occurred
by the LPS/TLR4 signaling pathway initiating intercellular messengers and activating NF-κB (2).
Our latter result, demonstrating no wheel activity-reducing effect of HK
Salmonella when compared to live Salmonella in C3H/HeJ mice, indicated that
the structure responsible for reduced wheel-running in this strain is heat-labile.
Along these lines, it has been demonstrated that heat killing of Salmonella, while
increasing the binding efficiency of LPS, destroys other components of the bacterial wall including flagella (41,43) and the type III secretion system (4).
Our data indicating that HK treatment of Salmonella abrogated the reduction in wheel-running induced by Salmonella led us to hypothesize that the difference in wheel-running activity between live Salmonella and HK Salmonella
might be due to differences in the magnitude of the inflammatory response.
Therefore, we measured TNF-α production from the macrophage cell line RAW
264 in response to HK or live Salmonella. Contrary to our hypothesis, HK Salmonella treatment led to an earlier rise in TNF-α production than to live Salmonella.
Moreover, RAW 264 cells were more sensitive to low doses of HK Salmonella
when compared to live Salmonella. This is consistent with the effect of heatkilling on LPS binding efficiency.
The results of the experiments discussed above led us to investigate other
bacterial structures that might contribute to Salmonella-induced reductions in
wheel-running behavior. Bacteria of the Salmonella family produce a number of
specialized effecter proteins that can modify host cell signaling (14) and potentially explain the LPS-independent effects seen in our studies. Indeed, antibodies
directed against flagella prevent bacterial motility and pathogenesis in mouse
models (8,16,38). In addition, it has been reported that live Salmonella stimulates
early release of TNF-α from RAW cells without TLR4 (37). Salmonella has several PAMPs, such as FG, PGN and DNA (1). FG is the major structural protein of
flagella expressed in most gram-negative bacteria (16). Recent reports indicate
Salmonella induces reduction of wheel-running activity via TLR5 • 47
that flagella elicit host immune responses and that the responsible component is
the filament protein FG which acts by binding to TLR5 (40). Indeed, intravenous
FG causes activation of the MAPK, SAPK and IKK signaling pathways, and NFκB activation (42), inducing a systemic inflammatory response (e.g. IL-8, IL-6
and TNF-α) in mice (9,10) and in vitro (46).
In this study, we used gentamycin (GM) an antibiotic bacterial protein synthesis inhibitor, to render Salmonella immotile. GM-treated Salmonella did not
attenuate wheel-running activity in C3H/HeJ mice, whereas the supernate of
GM-treated Salmonella did result in a significant reduction in wheel-running.
SDS PAGE and western blot analysis revealed loss of FG in GM-treated Salmonella, but not in GM-treated supernates, indicating that FG may be involved in
the reduction of wheel-running induced by Salmonella treatment in C3H/HeJ
mice. Moreover, FG also reduced wheel-running activity in normal mice. These
data implicate FG, acting through TLR 5, as a moderator of reduced wheel-running behavior in mice. In contrast, we did not observe any reduction in wheelrunning activity when mice were administered either PGN or CpGDNA (data not
shown).
The mRNA expression of TNF-α in response to FG was lower than that in
response to LPS (29), and the increase in cytokines (e.g. TNF-α and IL-6) and
nitric oxide (NO) produced by FG was also lower than that of LPS (22). In this
study, however, a significant reduction in wheel-running activity was observed
after the injection of live Salmonella, the flagella-rich supernatant of Salmonella
and purified FG. Although we were unable to answer the question of what is the
direct inducer of FG-induced sickness behavior, recent studies have reported that
FG induces hypotension (9), severe liver damage (22) and upregulation of IL-8
(41), which is a chemotactic factor and activator of neutorophils, basophils and T
cells (26) and is involved in the early host response to pathogens (11,18). Clarification of what proteins dictate the FG/TLR5 signaling pathway that leads to the
reduction of physical activity will be required in the future.
In summary, our findings provide evidence that FG expressed on the surface
of the gram-negative bacterium Salmonella attenuates the reduction of voluntary
physical activity in mice.
ACKNOWLEDGEMENTS
We thank Professor Hiroko Mine of the Department of Clinical Nutrition in
Kawasaki University of Medical Welfare for the provision of bacteria. This work
was supported by a Grant-in-Aid for Scientific Research from the Ministry of
Education, Science, Sports and Culture of Japan (C-#18500634), and the Interdepartmental Research Fund of Kawasaki University of Medical Welfare (to H.
Yano).
REFERENCES
1.
Akira, S., Uematsu, S., and Takeuchi, O. Pathogen recognition and innate immunity.
Cell 124: 783-801, 2006.
48 • Salmonella induces reduction of wheel-running activity via TLR5
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Aggarwal, B.B. Nuclear factor-κB: the enemy within. Cancer Cell 6: 203-208, 2004.
Carter, P.B., and Collins, F.M. The route of enteric infection in normal mice. J. Exp.
Med. 139: 1189-1203, 1974.
Christoph, J.H. Type III protein secretion systems in bacterial pathogens of animals
and plants. Micorbiol. Mol. Biol. Rev. 62: 379-433, 1998.
Dantzer, R., Bluthé, R.M., Layé, S., Bret-Dibat, J.L., Parnet, P., and Kelley, K.W.
Cytokines and sickness behavior. Ann. N.Y. Acad. Sci. 840 : 586-590, 1998.
Dantzer, R. Cytokine-induced sickness behavior: mechanisms and implication. Ann.
NY Acad. Sci. 933: 222-234, 2001.
Dantzer, R., and Kelley, K.W. Twenty years of research on cytokine-induced sickness behavior. Brain Behav. Immun. 21: 153-160, 2007.
Drake, D., and Montie, T.C. Flagella, motility and invasive virulence of
Pseudomonas aeruginosa. J. Gen. Microbiol. 134: 43-52, 1988.
Eaves-Pyles, T., Murthy, K., Liaudet, L., Virag, L., Ross, G., Soriano, F.G., Szabo,
C., and Salzman, A.L. Flagellin, a novel mediator of Salmonella-induced epithelial
activation and systemic inflammation: IκBα degradation, induction of nitric oxide
synthase, induction of proinflammatory mediators, and cardiovascular dysfunction.
J. Immunol. 166: 1248-1260, 2001.
Eaves-Pyles, T.D., Wong, H.R., Odoms, K., and Pyles, R.B. Salmonella flagellindependent proinflammatory responses are localized to the conserved amino and carboxyl regions of the protein. J. Immunol. 167 : 7009-7016, 2001.
Eckmann, L., Kagnoff, M.F., and Fierer, J. Epithelial cells secrete the chemokine
interleukin-8 in response to bacterial entry. Infect. Immun. 61 : 4569-4574, 1993.
Engeland, C.G., Nielsen, D.V., Kavaliers, M., and Ossenkopp, K-P. Locomotor
activity changes following lipopolysaccharide treatment in mice: a multivariate
assessment of behavioral tolerance. Physiol. Behav. 72: 481-491, 2001.
Feldman, M., Bryan, R., Rajan, S., Scheffler, L., Brunnert, S., Tang, H., and Prince,
A. Role of flagella in pathogenesis of Pseudomonas aeruginosa pulmonary infection. Infect. Immun. 66: 43-51, 1998.
Galan, J.E. Salmonella interactions with host cells: type III secretion at work. Annu.
Rev. Cell Dev. Biol. 17 : 53-86, 2001.
Harden, L.M., du Plessis, I., Poole, S., and Laburn, H.P. Interleukin-6 and leptin
mediate lipopolysaccharide- induced fever and sickness behavior. Physiol. Behav.
89: 146-155, 2006.
Hayashi, F., Smith, K.D., Ozinsky, A., Hawn, T.R., Yi, E.C., Goodlett, D.R., Eng,
J.K., Akira, S., Underhill, D.M., and Aderem, A. The innate immune response to
bacterial flagellin is mediated by Toll-like receptor 5. Nature 410: 1099-1103, 2001.
Johnson, R.W., Gheusi, G., Segreti, S., Dantzer, R., and Kelley, K.W. C3H/HeJ mice
are refractory to lipopolysaccharide in the brain. Brain Res. 752: 219-226, 1997.
Jung, H.C., Eckmann, L., Yang, S.K., Panja, A., Fierer, J., Morzycka-Wroblewska,
E., and Kagnoff, M.F. A distinct array of proinflammatory cytokines is expressed in
human colon epithelial cells in response to bacterial invasion. J. Clin. Invest. 95 : 5565, 1995.
Kent, S., Kelley, K.W., and Dantzer, R. Effects of lipopolysaccharide on food-motivated behavior in the rat are not blocked by an interleukin-1 receptor antagonist.
Neurosci. Lett. 145: 83-86, 1992.
Langhans, W. Bacterial products and the control of ingestive behavior: clinical
implications. Nutrition 12 : 303-315,1996.
Salmonella induces reduction of wheel-running activity via TLR5 • 49
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Lepper, P.M., Held, T.K., Schneider, E.M., Bolke, E., Gerlach, H., and Trautmann,
M. Clinical implications of antibiotic-induced endotoxin release in septic shock.
Intensive Care Med. 28 : 824-833, 2002.
Liaudet, L., Murthy, K.G., Mabley, J.G., Pacher, P., Soriano, F.G., Salzman, A.L.,
and Szabó, C. Comparison of inflammation, organ damage, and oxidant stress
induced by Salmonella enterica serovar Muenchen flagellin and serovar Enteritidis
lipopolysaccharide. Infect. Immun. 70: 192-198, 2002.
Liaudet, L., Szabo, C., Evgenov, O.V., Murthy, K.G., Pacher, P., Virag, L., Mabley,
J.G., Marton, A., Soriano, F.G., Kirov, M.Y., Bjertnaes, L.J., and Salzman, A.L. Flagellin from gram-negative bacteria is a potent mediator of acute pulmonary inflammation in sepsis. Shock 19: 131-137, 2003.
Matsumoto, T., Takahashi, H., Kawanishi, N., Shiva, D., Kremenik, M.J., Kato, Y.,
and Yano, H. The reduction of voluntary physical activity after poly I:C injection is
independent of the effect of poly I:C-induced interferon-beta in mice. Physiol.
Behav. 93: 835-841, 2007.
McCarthy, D.O., Kluger, M.J., and Vander, A.J. Suppression of food intake during
infection: is interleukin-1 involved? Am. J. Clin. Nutr. 42: 1179-1182, 1985.
Mukaida, N., Harada, A., and Matsushima, K. Interleukin-8 (IL-8) and monocyte
chemotactic and activating factor (MCAF/MCP-1), chemokines essentially involved
in inflammatory and immune reactions. Cytokine Growth Factor Rev. 9:9-23, 1998.
Ogawa, T. Chemical structure of lipid A from Porphyromonas (Bacteroides) gingivalis lipopolysaccharide. FEBS Lett. 332: 197-201, 1993.
Palin, K., Bluthé, R.M., McCusker, R.H., Moos, F., Dantzer, R., and Kelley, K.W.
TNFα-induced sickness behavior in mice with functional 55 kD TNF receptors is
blocked by central IGF-I. J. Neuroimmunol. 187: 55-60, 2007.
Pestka,J., and Zhou, H.R. Toll-like receptor priming sensitizes macrophages to
proinflammatory cytokine gene induction by deoxynivalenol and other toxicants.
Toxicol. Sci. 92: 445-455, 2006.
Pohlman, T.H., Munford, R.S., and Harlan, J.M. Deacylated lipopolysaccharide
inhibits neutrophil adherence to endothelium induced by lipopolysaccharide in vitro.
J. Exp. Med. 165: 1393-1402, 1987.
Poltorak, , A., He, X., Smirnova, I., Liu, M., Huffel, C.V., Du, X., Birdwell, D., Alejos, E., Silva, M., Galanos, C., Freudenberg, M., Ricciardi-Castagnoli, P., Layton,
B., and Beutler, B. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice:
mutations in tlr4 gene. Science 282: 2085-2088, 1998.
Raetz, C.R.H. Biochemistry of endotoxins. Annu. Rev. Biochem. 59: 129-170, 1990.
Raetz, C.R.H., and Whitfield, C. Lipopolysaccharide endotoxins. Annu. Rev.
Biochem. 71: 635-700, 2002.
Rietschel, E.T., and Brade, H. Bacterial endotoxin. Sci. Am. 267: 54-61, 1992.
Rietschel, E.T., Brade, H., Holst, O., Brade, L., Müller-Loennies, S., Mamat, U.,
Zähringer, U., Beckmann, F., Seydel, U., Brandenburg, K., Ulmer, A.J., Mattern, T.,
Heine, H., Schletter, J., Loppnow, H., Schönbeck, U., Flad, H.D., Hauschildt, S.,
Schade, U.F., Di Padova, F., Kusumoto, S., and Schumann, R.R. Bacterial endotoxin: chemical constitution, biological recognition, host response, and immunological
detoxification. Curr. Top. Microbiol. Immunol. 216: 39-81, 1996.
Rosenberger, C.M., Scott, M.G., Gold, M.R., Hancock, R.E., and Finlay, B.B. Salmonella typhimurium infection and lipopolysaccharide stimulation induce similar
changes in macrophage gene expression. J. Immunol. 164: 5894-5904, 2000.
50 • Salmonella induces reduction of wheel-running activity via TLR5
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
Royle, M.C.J., Totemeyer, S., Alldridge, L.C., Maskell, D.J., and Bryant, C.E. Stimulation of Toll-like receptor 4 by lipopolysaccharide during cellular invasion by live
Salmonella typhimurium is a critical but not exclusive event leading to macrophage
responses. J. Immunol. 170: 5445-5454, 2003.
Stanislavsky, E. S., and Lam, J.S. Pseudomonas aeruginosa antigens as potential
vaccines. FEMS Microbiol. Rev. 21: 243-277, 1997.
Subramanian, N., and Qadri, A. Lysophospholipid sensing triggers secretion of flagellin from pathogenic salmonella. Nat. Immunol. 7 : 583-589, 2006.
Smith, K.D., Andersen-Nissen, E., Hayashi, F., Strobe, K., Bergman, M.A., Barrett,
S.L., Cookson, B.T., and Aderm, A. Toll-like receptor 5 recognizes a conserved site
on flagellin required for protofilament formation and bacterial motility. Nat.
Immunol. 4 : 1247-1253, 2003.
Teruya, H., Higa, F., Akamine, M., Ishikawa, C., Okudaira, T., Tomimori, K.,
Mukaida, N., Tateyama, M., Heuner, K., Fujita, J., and Mori, N. Mechanisms of
Legionella pneumophila-induced interleukin-8 expression in human lung epithelial
cells. BMC Microbiol. 7 : 102 (1-16), 2007.
Tallant, T., Deb, A., Kar, N., Lupica, J., de Veer, M.J., and DiDonato, J.A. Flagellin
acting via TLR5 is the major activator of key signaling pathways leading to NF-κB
and proinflammatory gene program activation in intestinal epithelial cells. BMC
Microbiol. 4: 33, 2004 [http://www.biomedcentral.com/1471-2180/4/33].
Urban, T.A., Griffith, A., Torok, A.M., Smolkin, M.E., Burns, J.L., and Goldberg,
J.B. Contribution of Burkholderia cenocepacia flagella to infectivity and inflammation. Infect. Immun. 72: 5126-5134, 2004.
Vazquez-Torres, A., Vallance, B.A., Bergman, M.A., Finlay, B.B., Cookson, B.T.,
Jones-Carson, J., and Fang, F.C. Toll-like receptor 4 dependence of innate and adaptive immunity to Salmonella: importance of the Kupffer cell network. J. Immunol.
172: 6202-6208, 2004.
Yano, H., Fujinami, Y., Matsumoto, T., and Shiva, D. Effect of prostaglandhin E2
production on LPS-induced reduction in wheel-running activity in mice. Jpn. J.
Physical Fitness Sports Med. 55: S15-S18, 2006.
Zhang, J., Xu, K., Ambati, B., and Yu, F.S. Toll-like receptor 5-mediated corneal
epithelial inflammatory responses to Pseudomonas aeruginosa flagellin. Invest.
Ophthalmol. Vis. Sci. 44: 4247-4254, 2003.
Exercise-induced DNA damage and inflammatory responses • 51
Exercise-induced DNA damage: Is there a
relationship with inflammatory responses?
Oliver Neubauer 1, Stefanie Reichhold 1, Armen Nersesyan 2, Daniel König 3,
Karl-Heinz Wagner 1
1
Department of Nutritional Sciences, Faculty of Life Sciences, University of
Vienna, Althanstraße 14, 1090 Vienna, Austria
2 Environmental Toxicology Group, Institute of Cancer Research, Medical University of Vienna, Borschkegasse 8A, 1090 Vienna, Austria
3 Centre for Internal Medicine, Division of Rehabilitation, Prevention and Sports
Medicine, Freiburg University Hospital, Hugstetterstraße 55, 79106 Freiburg,
Germany
ABSTRACT
Both a systemic inflammatory response as well as DNA damage has been observed following exhaustive endurance exercise. Hypothetically, exercise-induced
DNA damage might either be a consequence of inflammatory processes or causally involved in inflammation and immunological alterations after strenuous prolonged exercise (e.g. by inducing lymphocyte apoptosis and lymphocytopenia).
Nevertheless, up to now only few studies have addressed this issue and there is
hardly any evidence regarding a direct relationship between DNA or chromosomal damage and inflammatory responses in the context of exercise. The most conclusive picture that emerges from available data is that reactive oxygen and nitrogen species (RONS) appear to be the key effectors which link inflammation with
DNA damage. Considering the time-courses of inflammatory and oxidative stress
responses on the one hand and DNA effects on the other, the lack of correlations
between these responses might also be explained by too short observation periods. This review summarizes and discusses the recent findings on this topic. Furthermore, data from our own study are presented that aimed to verify potential
associations between several endpoints of genome stability and inflammatory,
immune-endocrine and muscle damage parameters in competitors of an Ironman
triathlon until 19 days into recovery. The current results indicate that DNA effects
in lymphocytes are not responsible for exercise-induced inflammatory responses.
Furthermore, this investigation shows that inflammatory processes, vice versa, do
not promote DNA damage, neither directly nor via an increased formation of
Address for correspondence:
Oliver Neubauer, Department of Nutritional Sciences, Faculty of Life Sciences,
University of Vienna, Althanstraße 14, A-1090 Vienna, Austria,
Phone: +43-1-4277-54932, Fax: +43-1-4277-9549
E-mail: oliver.neubauer@univie.ac.at
52 • Exercise-induced DNA damage and inflammatory responses
RONS derived from inflammatory cells. Oxidative DNA damage might have been
counteracted by training- and exercise-induced antioxidant responses. However,
further studies are needed that combine advanced –omics based techniques (transcriptomics, proteomics) with state-of-the-art biochemical biomarkers to gain
more insights into the underlying mechanisms.
Key words: DNA damage, systemic inflammatory response, lymphocytopenia,
muscle inflammatory responses, endurance exercise
INTRODUCTION
Due to extensive research in the past decades, the effects of exercise on the
immune system are well documented [20, 33, 53]. However, researchers in this
area are still puzzled by questions about the underlying molecular mechanisms of
the observed immunological alterations [32, 33]. Extremely demanding
endurance exercise has been shown to induce both a systemic inflammatory
response [15, 42, 53, 71] as well as DNA damage [21, 36, 58, 62, 80]. Exerciseinduced DNA damage in peripheral blood cells appear to be mainly a consequence of an increased production of reactive oxygen and nitrogen species
(RONS) during and after vigorous aerobic exercise [58]. Besides oxidative stress,
other factors such as metabolic, hormonal and thermal stress in addition to the
ultra-structural damage of muscle tissue are characteristic responses to prolonged
strenuous exercise, that can lead to the release of cytokines, acute phase proteins
and to the activation or inhibition of certain lines of the cellular immune system
[15, 29]. In addition to these effectors, exercise-induced modifications in DNA of
immuno-competent cells have been hypothesised to be related with immune and
inflammatory responses to prolonged intensive physical activity, either by playing
a causative role and/or by resulting from exercise-induced inflammatory processes [21, 40, 44, 53]. Nevertheless, both experimental data as well as a more mechanistic understanding regarding this relationship are still incomplete.
The aim of this review is to outline the findings and current state of knowledge on potential associations between DNA modulations and inflammatory
responses after exercise. In the first part of this article, a short description of the
most commonly applied techniques to evaluate genome stability is provided. This
is followed by a brief summary of studies that have investigated the effects of
exercise on DNA in general. The latter issue has been presented elsewhere in
detail with a focal point on methodology in an article by Poulsen et al. [58]. In the
second part of this review the focus is on studies that have investigated both, certain endpoints of DNA damage and immuno-endocrine and inflammatory parameters in the context of exercise. Since apoptosis (programmed cell death) has been
suggested to influence the regulation of leukocyte counts after exercise [53], we
also addressed studies on this topic in the present review. Furthermore, we included the few investigations that examined exercise-induced DNA modulations and
markers of muscle damage, since this issue might give some indirect evidence for
inflammatory processes following exercise. Finally, data from our own study is
presented, which aimed to get a broader and more thorough insight into oxidative
[43], myocardial [28], skeletal muscular, inflammatory and immuno-endocrine
Exercise-induced DNA damage and inflammatory responses • 53
stress responses [42] as well as genome stability [62, 63] in a large cohort of Ironman competitors. By investigating a range of divergent parameters and by quantifying the resolution of recovery up to 19 days (d) after the Ironman race, the
results specifically enabled us to verify potential interactions between several
endpoints of DNA and chromosomal damage on the one hand and inflammation
and muscle damage on the other hand.
Commonly Applied Techniques to Monitor DNA and Chromosomal Stability
in Exercise
A number of different approaches have been used to evaluate DNA stability in
exercise studies. The aim of this part of the present article is to give a brief
overview on the principles of the most frequently applied methods, since this
topic has been comprehensively reviewed in the scientific literature [8, 17, 26,
58].
Many studies in this context applied the single cell gel electrophoresis
(SCGE or COMET) assay due to its sensitivity and simplicity [8]. This technique
is based on the determination of the migration of damaged DNA out of the nucleus in an electric field, whereas the migrated DNA resembles the shape of a comet
[21, 26]. The standard version (under alkaline conditions) enables the detection of
DNA single and double strand breaks, and apurinic sites [77], while the use of the
lesion specific enzymes endonuclease III (ENDO III) and formamidopyrimidine
glycosylase (FPG) allows the detection of oxidized purines and pyrimidines,
respectively [7, 8]. Regarding the interpretation of the results that are obtained by
the SCGE assay it is important to bear in mind that endpoints are differently
reported as tail lengths of the comets, percentage DNA in tail and tail moment [8].
Contrary to the SCGE assay, the cytokinesis block micronucleus cytome
(CBMN Cyt) assay allows to assess persistent chromosomal damage [16, 21].
Endpoints of this precise method includes the formation of micronuclei (MN)
resulting from chromosomal breakage or loss, nucleoplasmic bridges (NPBs)
indicating chromosome rearrangements, and nuclear buds (Nbuds) that are
formed as a consequence of gene amplification [16, 18]. The reliability of this
MN in pathophysiological conditions has been substantiated by a recent study
which has shown an association between MN frequency and cancer incidence [3].
In several exercise studies 8-oxo-7,8-dihydro-2-deoxyguanosine (8-oxodG),
was investigated, which is formed through oxidative modification of guanine, and
mainly detected in urine or in leukocytes [26]. Measurement of urinary 8-oxodG
is thought to be the result of the repair of these lesions in DNA, excretion into the
plasma and subsequently into urine [58]. Hence, it does not necessarily reflect the
steady-state of un-repaired DNA damage [80]. Moreover, urinary 8-oxodG represents a general oxidative damage marker for the whole body, and consequently, is
not specific to DNA damage in white blood cells [60, 80]. Attention should also
be given in the interpretation of this biomarker due to methodological drawbacks
and discrepancies among divergent approaches which are currently used to
analyse 8-oxodG [26, 58].
Effects of Different Kinds of Exercise on DNA
Epidemiological as well as empirical data indicate protective effects of physical
activity on site-specific cancer risk [58, 64, 76]. However, similarly to the con-
54 • Exercise-induced DNA damage and inflammatory responses
cerns about ultra-endurance exercise and cardiovascular health [27], Poulsen et al.
hypothesised a U-shaped curve relationship between exercise and health particularly in the context of oxidative DNA modifications [58]. Data are available now
on the effects of acute bouts of very prolonged (ultra-endurance) exercise on
genome stability, which will also be presented in the following overview. According to the literature [10], ultra-endurance is defined as exercise lasting more than
4 hours (h).
Ultra-endurance Exercise (> 4 h)
Increased DNA instability as detected by the SCGE technique [36, 63] or with the
CBMN Cyt assay [62] or by analysis of urinary 8-OHdG concentrations [37, 60]
were found after an Ironman triathlon [62, 63] and ultra-marathon races [36, 37,
60]. Importantly, changes regarding the SCGE assays as well as urinary 8-OHdG
were only temporary [36, 37, 60, 62, 63] and endpoints of DNA damage measured with the CBMN Cyt assay even decreased in response to an Ironman race
and declined further 19 d post-race [62]. These responses are discussed later in
detail within the scope of our own observations.
Competitive Endurance Exercise (< 4 h)
Data regarding competitors of endurance races with a duration of less than four
hours are partly inconclusive, albeit in most studies increased DNA migration was
detected in SCGE assays after a half-marathon [44], a marathon [80] or a shortdistance triathlon race [21]. On the contrary, neither changes in the levels of
strand breaks nor in the FPG-sensitive sites, but increased ENDO III sites were
observed after a half-marathon- and a marathon [4]. However, the subjects of the
latter study were monitored only immediately post-race, while other investigations demonstrated that major DNA modulations were sustained until 5 d postrace in six short-distance triathletes [21] and for even 14 d following a marathon
[80]. Nevertheless, based on the finding of an unaltered frequency in MN, Hartmann et al. [21] concluded that intense exercise with a mean duration of 2.5 h
does not lead to chromosome damage.
Submaximal and Maximal Exercise Under Laboratory Conditions
Several studies conducted submaximal aerobic exercise protocols under laboratory conditions to investigate DNA effects. DNA damage was neither seen after
intense treadmill running in male subjects of different training status [82] nor in
well-trained endurance athletes [54]. In addition, Sato et al. showed that acute
mild exercise as well as chronic moderate training does not result in DNA damage, but rather leads to an elevation in the sanitization system of DNA damage
[66]. Interestingly, in an experiment that aimed to examine the influence of a
downhill run before and after supplementation with vitamin E, no effect was
found on the levels of leukocyte 8-OHdG in both 16 young and 16 older physically active men [65]. However, it has to be mentioned that DNA responses were not
followed until at least 1 d post-exercise in most of these studies [54, 65, 82].
Conflicting findings were reported when maximal exercise protocols, i.e. tests
until exhaustion, were conducted under laboratory conditions. Increased levels of
DNA strand breaks were observed after exhaustive treadmill running in subjects of
different training status [22, 45]. Moller et al. [38] demonstrated DNA strand breaks
Exercise-induced DNA damage and inflammatory responses • 55
and oxidative DNA damage after an maximal cycle ergometer test under highaltitude hypoxia, but not normal (normoxic) conditions. In another study, elevated
levels of MN were reported after exhaustive sprints; however, the six subjects were of
divergent training levels and gender and included one smoker [67]. On the contrary,
Pittaluga et al. [56] detected no effects of a maximal exercise test on MN in 18 young
subjects with different training status, but the authors noted chronic cellular stress
including higher MN levels at rest in the athlete group. Furthermore, there were no
differences in urinary 8-OHdG concentrations before and after supplementation with
β-carotene within the 3 d following a cycle ergometer test to exhaustion [70].
Periods of intensified training
A few studies have examined whether periods of intensified training affect
genome stability. Increased urinary 8-OHdG levels were observed in 23 healthy
males in response to a vigorous physical training programme (about 10 h of exercise for 30 d) [57] and in male long-distance runners throughout a training period
for 8 d compared to a sedentary period [47]. However, in a longitudinal study no
differences in urinary excretion of 8-OHdG between a group of long-distance runners and a sedentary control group were observed [55]. In two separate studies
that comprised a similar group of male triathletes, Palazzetti et al., reported either
no [48] or increased DNA damage [49] after 4 weeks (wk) of overload training as
detected by the SCGE assay, probably due to inter-individual differences.
In conclusion, there is growing evidence that strenuous exercise can lead to DNA
damage that with few exceptions [36] is predominantly observed not before 24 h
after the resolution of exercise [21, 44, 45, 80]. However, the diversity of methods
and endpoints used to assess DNA modifications and different study designs (i.e.
divergent exercise protocols and sampling time-points) make it difficult to determine the exact circumstances under which DNA damage occurs. Crucially, in
addition to the aforementioned factors, the heterogeneity of study cohorts (varying in gender, age and training status) most likely contributes to inconsistencies
among the studies on this topic. Nevertheless, results of the few studies that have
examined the effects of ultra-endurance exercise on genome stability indicate that
adaptations of endogenous protective antioxidant and/or repair mechanisms prevent severe and persistent DNA damage in well-trained athletes [36, 37, 45, 60,
62]. Thus, a clear dose-response relationship regarding the level of exercise that
could be detrimental cannot yet be established. Currently, there are no indications
that exhaustive endurance exercise increases the risk for cancer and other diseases
via DNA damage. However, it remains to be clarified whether perturbances of the
genomic stability of immuno-competent cells are involved in the post-exercise
temporary dysfunction of certain aspects of immunity, which may increase the
risk of subclinical and clinical infection [15, 20, 53].
Findings on Exercise-induced DNA Damage and/or Apoptosis and
Inflammatory Responses
Table 1 summarizes the small number of studies that have examined the effects of
exercise on DNA and/or apoptosis on the one side and inflammatory responses on
the other. As one of the earlier works in the context of the effects of particularly
competitive endurance exercise on DNA damage, Niess et al. [44] found that neu-
56 • Exercise-induced DNA damage and inflammatory responses
trophil counts 1 h after a half-marathon run correlated with DNA damage in
leukocytes, assessed 24 h post-race. Without examining markers of oxidative
stress, the authors could only speculate that RONS released by neutrophils might
have been responsible for the formation of DNA strand breaks. However, their
results led them to suppose that the observed DNA damage might be the key
mechanism for the modifications in the immune cell counts [44]. On the contrary,
Table 1. Studies investigating exercise-induced DNA damage and/or apoptosis and inflammatory/immune
parameters
Exercise-induced DNA damage and inflammatory responses • 57
they found no correlation between changes in DNA migration in the SCGE assay
and leukocyte counts in the 24 h after an exhaustive treadmill test [44], possibly
also because the extent of the inflammatory response was relatively low following
their exercise protocol. Although no immune and inflammatory parameters were
measured in the study by Hartmann et al. [21], their explanations have further
stimulated debate on a relationship between the activation of inflammatory cells
and the occurrence of secondary tissue and DNA lesions. Based on their observations in short-distance triathletes (no indications for oxidative DNA modifications
immediately post-race, but highest values within the standard SCGE assay 3 d
after the competition), they suggested that DNA damage might occur as a consequence of exercise-induced injury of muscle tissue rather than acute oxidative
stress during exercise [21]. The authors hypothesised that inflammatory reactions
in the course of this initial muscle damage could be responsible for the transient
DNA damage [21]. Indeed, there is evidence that activated neutrophils and
macrophages infiltrate damaged muscle [68, 78]. Although this seems to be a beneficial response in terms of muscle repair and also muscle adaptation [33, 78], it
may trigger further inflammatory processes and damage [25], in part through an
enhanced formation of RONS [29].
On the basis of these findings, researchers in this field questioned whether
damage to cellular DNA in the course of vigorous exercise could also induce
apoptosis and whether programmed cell death, in turn, might be related to the
exercise-induced regulation of leukocyte counts and, particularly, lymphocyte
trafficking and distribution [53]. A decline of the total lymphocyte concentration
is characteristic after exercise of prolonged duration and/or high intensity [33,
53]. Although the mechanisms of exercise-induced lymphocytopenia are still not
fully understood [33], it has been suggested that this effect may account, at least
partly, for the post-exercise immune dysfunction [15]. Exercise-induced changes
in corticosteroids and catecholamines are known to play a major role in characteristic post-exercise alterations of leukocyte subsets [20, 41] including leukocytosis
[42] as well as lymphocytopenia [53]. Previous studies indicated that the glucocorticoid concentrations observed after submaximal exercise are sufficient to
induce apoptosis [23]. These observations further support the assumption of a
relationship between exercise-associated induction of apoptosis and lymphocy-
58 • Exercise-induced DNA damage and inflammatory responses
topenia [53]. In response to cellular stressors that lead to DNA damage, apoptosis
is vital in preventing the propagation of severely damaged DNA and in maintaining genomic stability [30] and is regarded to be required for the regulation of the
immune response [39].
Mars et al. were the first to describe apoptosis in lymphocytes after exhaustive exercise (treadmill running) that was paralleled by DNA damage [34]. However, in the latter study, cell death was only investigated in three subjects and the
methodology (the TdT-mediated dUTP-nick end labelling or TUNEL method)
has been criticized due to its insufficient specificity [40]. Nevertheless, by the use
of flow cytometry and annexin-V to label apoptotic cells, Mooren et al. [39, 40]
confirmed that either short maximal exercise (in untrained subjects) [39] as well
as competitive endurance exercise (a marathon run) [40] has the potential to
induce lymphocyte apoptosis. This phenomenon could be explained, to a certain
extent, by an up-regulation of the expression of cell death receptors and ligands
[40] and an exercise-induced shift to a lymphocyte population with a higher density of these (CD95-)receptors [39]. Nevertheless, the authors concluded that the
changes in the proportions of apoptotic cells after exhaustive exercise were small
and, if at all, might only partially account for the concomitantly observed significant decline of lymphocytes to below baseline levels [39]. An additional finding
of Mooren et al. [40] was that apoptotic sensitivity was inversely related to the
training status of the marathon runners, since analysis of subgroups revealed that
programmed cell death occurred only in less well-trained, but not in highlytrained athletes. Recent research in this context suggests that intensive endurance
exercise does neither automatically induce apoptosis in lymphocytes nor cause
DNA damage (assessed immediately and 3 h post-exercise), provided that subjects are well-trained [54]. Since there was no correlation between the (non-significant) decrease in circulating lymphocytes and the percentage lymphocyte
apoptosis after a 2.5 h treadmill run at 75% VO2 max., Peters et al. [54] concluded
that the characteristic post-exercise lymphocytopenia is not due to apoptotic regulation by the immune system. The latter results are consistent with another study
which was conducted with a similar exercise protocol, but in untrained subjects
[69]. Steensberg et al. [69] noted that the lymphocytes which left the circulation
during the first 2 h post-exercise were characterised by not being apoptotic. Thus,
mechanisms other than apoptosis seem to play a more important role in inducing
lymphocytopenia after exercise, including a redistribution of lymphocytes and/or
a lack of mature cells that can be recruited [53]. Moreover, contrary to previous
findings [23], recent results imply that cortisol affects the cellular immune system
more by other pathways than via apoptotic regulation [54]. Furthermore, the
occurrence of DNA damage in the course of exercise does not necessarily implicate induction of apoptosis [40]. Alternative cellular outcomes to prevent the
propagation of DNA damage include cell cycle arrest or DNA repair [30].
In general, there is strong evidence which suggests that enhanced DNA stability and, most likely in turn, the absence of a change in the levels of apoptotic
lymphocytes after strenuous exercise [54] are associated with protective adaptations due to training. As mentioned above, Mastaloudis et al. [36], reported that
DNA damage in leukocytes increased temporarily mid-race of an ultra-marathon,
but returned to baseline 2 h after the competition and even decreased to below
Exercise-induced DNA damage and inflammatory responses • 59
baseline values by 6 d post-race. As probable causes for this decrease in the proportion of cells with DNA damage, the authors suggested enhanced repair mechanisms, increased clearance and/or a redistribution of damaged cells [36]. Noteworthy, plasma concentrations of inflammatory parameters, F2-isoprostanes and
antioxidant vitamins were investigated in the same subjects. Although acute oxidative and inflammatory stress responses were observed [35], the authors reported no
correlations between either of these markers with DNA damage [35, 36]. Furthermore, supplementation with vitamins E and C prevented increases in lipid peroxidation [35], but had no noticeable effects on DNA damage, on inflammation and
on muscle damage [36]. Interestingly, there were different responses regarding
oxidative stress and DNA damage in male and female runners, highlighting the
importance of studying both sexes [35, 36]. In general, these findings in ultramarathon runners indicate that the mechanism of oxidative damage is operating
independently of the inflammatory and muscle damage processes [35, 36, 79].
There are only few studies on the issue of DNA damage and immune and
inflammatory responses in the course of exercise. Briviba et al. [4] found oxidative DNA damage parallel to an increased oxidative burst ability of granulocytes
and monocytes after both a half-marathon- and a marathon race, but no correlations were detected. Again, the authors could only speculate that the exerciseinduced activation of phagocytes might have contributed to the increased RONS
production, oxidative DNA damage and the high percentage of apoptotic lymphocytes [4]. Furthermore, it is notable that the monitoring period of this study probably was too short to detect possible interactions between DNA alterations and
immune modifications.
Findings on Exercise-induced DNA Damage and Muscle Damage
As mentioned, given the scarceness of data regarding associations between DNA
modulations and inflammation in the course of exercise, we included investigations that examined exercise-induced effects on DNA together with markers of
muscle damage. These studies are summarized in Table 2. Though several major
stressors are needed and the integrity of the organism has to be challenged (e.g.
by extremely demanding endurance exercise) [29, 42, 53, 72] to induce a systemic
inflammatory response, it has been shown that leukocytes can explicitly be
mobilised in response to muscle damage [42, 51, 74], possibly due to activation of
the alternative complement pathway [51, 74]. Therefore, these studies may also
reveal whether muscle damage (induced by mechanical and/or metabolic stress
[25, 75]) and subsequent repair and inflammatory responses [78] are associated
with DNA damage. In one of the first studies on this issue, which comprised three
subjects of different gender and training history, Hartmann et al. reported a parallel increase, but no correlation between the DNA migration in the SCGE assay
and plasma creatine kinase (CK) between 6 and 24 h after intense treadmill running [22]. Likewise, applying the standard SCGE assay, Palazzetti et al. [48]
observed signs of increased oxidative stress and muscle damage induced by a
duathlon race after 4 wk of overload training, whereas no effects on leukocyte
DNA were found, probably due to efficient DNA repair. Other studies on this
topic predominantly measured 8-OHdG in urine, which reflects the average rate
of oxidative DNA damages in all cells of the body [58]. Consequently, changes in
urinary 8-OHdG excretion after muscle-damaging exercise might largely repre-
60 • Exercise-induced DNA damage and inflammatory responses
Table 2. Studies investigating exercise-induced DNA damage and muscle damage
Exercise-induced DNA damage and inflammatory responses • 61
sent DNA damage of skeletal muscles [60]. Radak et al. [60] and Miyata et al.
[37] determined urinary 8-OHdG levels and markers of muscle damage in competitors of ultra-marathon events which lasted 2 [60] and 5 d [37], respectively.
No propagation of oxidative DNA damage was observed after the first race d in
both studies [37, 60]. Interestingly, 8-OHdG significantly decreased to levels
below their peak values during the race on the second d [37], and on the fourth
race d [60], respectively. Both research groups suggested that a rapid induction of
antioxidant and repair systems occurred [37, 59]. In contrast, parameters for muscle damage continuously increased during the 2-d-race period [37] and until the
third d of the 4-d-race [60], and no correlations were reported with 8-OHdG.
Taken together, these data may show that, even if myofibrillar injury occurs, an
adaptive up-regulation of repair and nucleotide sanitization mechanisms is capable of preventing further damage of DNA. Consistently, no correlations between
biomarkers of DNA- and muscle damage were reported after a period of intensified training (despite that both 8-OHdG and muscle damage markers were found
to be increased) [47] or downhill running on a treadmill [65]. However, given that
8-OHdG levels remained unchanged, but were measured only until 1 d post-race,
the authors of the latter investigation noted that oxidative DNA damage probably
had occurred in the period between the first and the third d after exercise, when
some links amongst circulating oxidative stress markers and CK activity were
observed [65].
The prolonged monitoring period after a marathon race in an investigation
by Tsai and co-workers [80] might account for the observed significant correlations between peak levels of ENDO III-sensitive sites and urinary 8-OHdG on
the one side and plasma parameters of muscle damage and lipid peroxidation on
the other. In agreement with the conclusions of previous investigations [21, 44],
the authors suggested that inflammatory cells infiltrating into injured skeletal
muscle tissue and activated phagocytes were responsible for the increased production of RONS and consequently the delayed oxidative DNA damage during
the reparative processes after the marathon [80]. This idea is supported by a
study in rats, in which DNA damage in circulating white blood cells was closely
related to muscle damage due to exercise [81]. Nevertheless, based on these findings it is not possible to draw a clear conclusion as to whether oxidative DNA
modifications in peripheral immuno-competent cells are casually related with
immune disturbances or whether DNA damage in leukocytes, in fact, results
from oxidative stress that occurs through inflammatory processes after strenuous
exercise.
Purpose of the Current Study in Ironman Triathletes
The data presented here are part of a larger study that aimed to comprehensively
examine certain stress and recovery responses to an Ironman triathlon race. One
primary aim of the study was to test the hypothesis whether there is a relationship
between indices of muscle damage and/or inflammatory stress and endpoints of
DNA damage in lymphocytes, which were assessed by the SCGE- and the CBMN
Cyt assays for the first time in the course of competitive exercise of such duration.
Furthermore, by concomitantly exploring oxidative stress markers and antioxidant-related factors, we aimed to particularize a potential interaction of oxidative
stress between inflammatory and DNA responses.
62 • Exercise-induced DNA damage and inflammatory responses
MATERIALS AND METHODS
The study design has been described previously [28, 42]. Briefly, the study population comprised 48 non-professional, well-trained healthy male triathletes, who
participated in the 2006 Ironman Austria. Forty-two of them (age: 35.5 ± 7.0 yr,
height: 180.6 ± 5.6 cm, body mass: 75.1 ± 6.4 kg, cycling VO 2 peak: 56.6 ± 6.2 ml
kg -1 min -1, weekly net endurance exercise time: 10.7 ± 2.6 h) completed the
study and were included in the statistical analysis to investigate inflammatory and
immuno-endocrine responses as well as muscle damage [42]. The physiological
characteristics of the study participants (assessed on a cycle ergometer three
weeks before the competition), information on their training over a period of six
months prior to the race, their performance in the Ironman triathlon as well as the
only moderate (“recovery”) training thereafter have been presented in detail elsewhere [42, 43]. Of the entire study group 20 and 28 subjects were randomly
selected for the CBMN Cyt and the SCGE assays, respectively [62, 63]. Consequently, these randomized subjects were included in the data analysis for the
results that are exclusively provided within this report. All participants of the
study did not take any medication or more than 100% of RDA of antioxidant supplements (in addition to their normal dietary antioxidant intake) in the six weeks
before the Ironman race until the end of the study. The Ironman triathlon took
place in Klagenfurt, Austria on July 16th 2006 under near optimal climatic conditions and consisted of 3.8 km swimming, 180 km cycling and 42.2 km running.
Blood samples were taken 2 d pre-race, immediately (within 20 min), 1, 5 and 19
d post-race.
The samples were immediately cooled to 4°C and plasma separated at 1711
* g for 20 min at 4°C and aliquots for the measurement of biochemical parameters
were frozen at –80°C until analysis. For the analysis of DNA and chromosomal
damage in lymphocytes, blood samples were processed instantly as described previously [62, 63]. Blood samples were analysed for haematological profile, plasma
creatine kinase (CK) activity, plasma concentrations of myoglobin, interleukin
(IL)-6, IL-10, high-sensitivity C-reactive protein (hs-CRP), myeloperoxidase
(MPO), polymorphonuclear (PMN) elastase, cortisol and testosterone (see [42]).
All these values (except for the steroid hormones) were adjusted for exerciseinduced changes in plasma volume [11]. As reported previously [62, 63], the
SCGE and CBMN Cyt- assays were carried out according the methods described
by Tice et al. [77] and Fenech [17], respectively. Within the SCGE-assay, oxidative DNA base damage was assessed on the basis of the protocols of Collins et al.
[7], Collins and Dusinska [6] and Angelis et al. [1]. Analysed endpoints within the
SCGE assay included: 1.) determination of DNA migration under standard conditions to measured single and double strand breaks (determined as percentage of
DNA in the tail), and 2.) ENDO III and FPG to detect oxidized pyrimidines and
purines, respectively. Biomarkers within the CBMN Cyt block included the number of 1.) MN, 2.) NPBs, 3.) Nbuds, and 4.) necrotic and apoptotic cells.
All statistical analyses were performed using SPSS 15.0 for Windows.
Details of the data analysis has been presented previously [42, 62, 63]. For the
additional correlation analysis that is reported in this article, Pearson ´s correlation was used to examine significant relationships. In case of observed trends or
significant correlations, subjects were divided into percentile groups by the asso-
Exercise-induced DNA damage and inflammatory responses • 63
ciated variables (e.g. IL-6). One-factorial ANOVA and post hoc analyses with
Scheffé´s test were then applied to assess whether differences in endpoints of
DNA or chromosomal damage were associated with the percentile distribution.
Significance was set at a P-value <0.05 and is reported P<0.05, P<0.01 and
P<0.001.
RESULTS
Race Results
The average completion time of the whole study group was 10 h 52 min ± 1 h 1
min (mean ± SD). The estimated average antioxidant intake during the race was
393 ± 219 mg vitamin C and 113 ± 59 mg alpha-tocopherol. There were neither
significant differences in the performance nor in the consumed antioxidants
between the whole study group and the subgroups that were tested for genome
stability.
DNA and Chromosomal Damage, Apoptosis and Necrosis
As previously reported [62, 63] and briefly discussed above, the results concerning DNA and chromosomal damage were as follows: Within the CBMN Cyt
assay, the number of MN significantly (P<0.05) decreased immediately post-race,
and declined further to below pre-race levels 19 d after the Ironman competition
(P<0.01). There were no changes in the frequency of NPBs and Nbuds as an
immediate response to the triathlon, but 5 d thereafter the frequency of Nbuds was
significantly (P<0.01) higher than levels immediately post-race. However, 19 d
post-race the frequency of Nbuds returned to pre-race levels, while the number of
NPBs was significantly (P<0.05) lower than pre-race [62].
The overall number of apoptotic cells decreased significantly (P<0.01)
immediately post-race, and declined further until 19 d after the race (P<0.01).
Similarly, the overall number of necrotic cells significantly (P<0.01) declined
immediately post-race, and remained at a low level 19 d after the Ironman. Within
the SCGE assay, a decrease was observed in the level of strand breaks immediately after the race. One day post-race the levels of strand breaks increased (P<0.01),
then returned to pre-race 5 d post-race, and decreased further to below the initial
levels 19 d post-race (P<0.01). Immediately post-race there was a trend in ENDO
III and FPG-sensitive sites to decrease. The ENDO III-sensitive sites significantly
(P<0.05) increased 5 d post-race compared to 1 d post-race, but levels decreased
until 19 d (P<0.05). No significant changes were observed in the levels of FPGsensitive sites throughout the monitoring period [63].
Immune-endocrine and Inflammatory Responses, and Plasma Markers
of Muscle Damage
Briefly, as described in details elsewhere [42], there were significant (P<0.001)
increases in total leukocyte counts, MPO, PMN elastase, cortisol, CK activity,
myoglobin, IL-6, IL-10 and hs-CRP, whereas testosterone significantly (P<0.001)
decreased compared to pre-race. Except for cortisol, which decreased below prerace values (P<0.001), these alterations persisted 1 d post-race (P<0.001, P<0.01
for IL-10). Five days post-race CK activity, myoglobin, IL-6 and hs-CRP had
64 • Exercise-induced DNA damage and inflammatory responses
decreased, but were still significantly (P<0.001) elevated. Nineteen days post-race
most parameters had returned to pre-race values, with the exception of MPO and
PMN elastase, which had both significantly (P<0.001) decreased below pre-race
concentrations, and myoglobin and hs-CRP, which were slightly, but significantly
higher than pre-race [42].
Associations between Endpoints of Genome Stability and Immunoendocrine, Inflammatory and Muscle Damage Parameters
No significant correlations were found between all these markers at all timepoints with the exception of a link between IL-6 and necrosis. Immediately postrace, the plasma concentration of IL-6 correlated positively with the number of
necrotic cells (r=0.528; P<0.05). In addition, significant associations were
observed on the basis of a group distribution into percentiles by the IL-6 concentrations immediately post-race. First, the numbers of necrotic cells increased with
IL-6 across the percentiles, and the differences between all groups were P=0.012.
Second, necrosis in lymphocytes was significantly (P=0.017) higher in the subject
group with the highest IL-6 concentrations (top percentile) compared with the
lowest IL-6 values (lowest percentile).
DISCUSSION
A major finding of the present investigation is that there were no correlations
between different markers of DNA and chromosomal damage and parameters of
muscle damage and inflammation in participants of an Ironman triathlon as a prototype of ultra-endurance exercise with the exception of a link between IL-6 and
necrosis. The conclusions that can be drawn from these results are several. Overall, the current data indicate that DNA damage is neither causally involved in the
initial systemic inflammatory response nor in the low-grade inflammation that
was sustained at least until 5 d after the Ironman race [42]. Instead, based on several assessed relationships between leukocyte dynamics, cortisol, muscle damage
markers and cytokines [42], the pronounced but temporary systemic inflammatory response was most likely induced by stressors other than DNA modulations. In
fact, consistent with previous studies in this context, factors such as the initial
ultra-structural injury of skeletal muscle [51, 74], changes in concentrations of
cortisol [53] and IL-6 [71] apparently mediated leukocyte mobilization and activation [42]. Furthermore, although the temporary increased frequency of ENDO
III-sensitive sites 5 d after the Ironman competition was found simultaneously
with the moderate prolongation of inflammatory processes, correlations between
hs-CRP and markers of muscle damage suggest that the latter phenomenon was
rather related to incomplete muscle repair [42].
In addition, missing links between all these markers in the present study
indicate that exercise-induced inflammatory responses vice versa do not promote
DNA damage in lymphocytes. These results support those of Mastaloudis et al.,
who demonstrated that inflammatory and muscle damage responses, indeed, do
not directly interact with the mechanisms of oxidative DNA damage [35, 36, 79].
Nevertheless, this does not rule out the possibility that inflammatory processes
can trigger oxidative stress via oxidative burst reactions of circulating neutrophils
Exercise-induced DNA damage and inflammatory responses • 65
and an increased cytokine formation [15, 25, 29, 50, 73], which in turn might lead
to secondary (oxidative) DNA damage in immuno-competent cells [80]. In fact,
we observed correlations between markers of oxidative stress and inflammatory
parameters (unpublished results) that might point to muscular inflammatory
processes as a source of the moderate oxidative stress response 1 d after the Ironman triathlon. Nevertheless, we have recently demonstrated in the same study
participants that training- and acute exercise-induced responses in the antioxidant
defence system were able to counteract severe or persistent oxidative damage
post-race. Despite a temporary increase in protein oxidation and lipid peroxidation markers immediately and 1 d post-race (except for oxidized LDL concentrations, which actually decreased), all these markers had returned to pre-race values
5 d post-race [43]. Concomitantly, there was an increase in the plasma antioxidant
capacity following the Ironman triathlon (assessed by the trolox equivalent
antioxidant capacity- (TEAC), the ferric reducing ability of plasma- (FRAP), and
the oxygen radical absorbance capacity (ORAC)-assays) [43, 63]. These strong
antioxidant responses most likely played a significant role in counteracting sustained oxidative stress post-race in the current study, while it seems that antioxidant defences in the study group of Tsai et al. [80] were not sufficient to confer
protection against delayed oxidative damage to lipids and DNA due to reparative
processes of muscular tissue. Whatever the reasons for these discrepancies in the
oxidant/antioxidant balance are (differences in training-induced biochemical
adaptations, antioxidant status and/or antioxidant intake during the race, etc.), this
might be a major explanation for the inconsistencies between the findings of Tsai
et al. [80] and ours [43, 64, 62]. In fact, the observed negative correlations
between the ORAC and ENDO III-sensitive sites immediately and 1 d after the
Ironman race suggest that an enhanced plasma antioxidant capacity might have
prevented oxidative DNA damage [63]. These findings are in line with a recent
animal study [2], which demonstrated the protective role of an enhanced serum
antioxidant capacity in lymphocyte apoptosis. Taken together, whenever correlations between DNA damage in immuno-competent cells and inflammation [44] or
muscle damage [80] were observed, RONS derived from inflammatory cells,
appear to be the key effectors that link inflammation with DNA damage after vigorous exercise. Fig. 1 is a schematic illustration of the relationships between these
stress responses to exhaustive endurance exercise. It may be argued that results
from our study fit well into this picture insofar that antioxidant mechanisms neutralized an enhanced generation of RONS potentially resulting from inflammatory
processes due to the injury of skeletal muscle tissue, and consequently were able
to prevent lymphocyte DNA damage. It should also be noted that, similar to DNA
effects, muscle inflammatory processes and related oxidative stress responses
might be sustained for or appear days after muscle-damaging exercise [46].
Hence, potential links between these outcome measures might have been missed
in investigations with shorter monitoring periods [4, 40, 54, 65, 69]. Beyond, it is
important to note in this context that there is an additional difficulty in determining correlations between markers of oxidative DNA damage and other biomarkers
of oxidative stress, partly due to differences in the biological sites where oxidative
damage occurred [12].
The observed association between IL-6 concentration and the number of
necrotic cells immediately post-race in the present study may indicate that lym-
66 • Exercise-induced DNA damage and inflammatory responses
Fig. 1: Proposed model of exercise-induced DNA damage and inflammatory responses
phocytes partly undergo an unregulated cell death in athletes experiencing an overshooting inflammatory response. Based on recent research on the role of IL-6 in exercise [15, 19, 52], it is questionable whether IL-6, probably released by contracting
muscles [19, 52], directly modulates necrosis in lymphocytes. In this case, plasma IL6 concentrations may just serve as a marker for the pronounced initial systemic
inflammatory response. However, the (patho-)physiological relevance of this association cannot be generalised based upon the present results, since the overall number of
necrotic cells declined significantly to below pre-race values after the acute bout of
ultra-endurance exercise, and remained at these levels at all time-points investigated
[63]. Similarly, as to the decrease of necrosis, we demonstrated that levels of apoptosis also decreased immediately after the Ironman race, again remaining at these low
levels throughout the whole monitoring period [63]. Crucially, our data revealed no
link between apoptosis and post-race changes in lymphocyte counts. Mooren et al.
[40] reported an initial increase in apoptotic cells in the whole group of marathon runners, but corresponding with the findings in the current study, lymphocyte apoptosis
declined 1 d after the race. In agreement with the decrease of DNA damage after an
ultra-marathon run [36], these findings might alternatively be explained by an overshooting removal of apoptotic leukocytes by phagocytic cells in order to protect tissue
from overexposure to inflammatory and immunogenic contents of dying cells [31,
40]. Based on the concept that the phagocytic clearance of apoptotic immuno-competent cells plays a critical role in the resolution of inflammation [31, 83], this could be a
further explanation for the lack of a link between inflammatory responses on the one
hand, and DNA damage and/or apoptotic cell death on the other hand.
Finally, a reason that may also account for the lack of correlations within
most of the few studies that have addressed this issue is that the majority of these
Exercise-induced DNA damage and inflammatory responses • 67
investigations have been conducted in trained individuals [21, 36, 37, 47, 48, 54,
60, 62]. Accumulating evidence points to adaptations in protective mechanisms
due to (endurance) training - including improved endogenous antioxidant
defences and enhanced repair mechanisms [59] - that appear to be responsible for
maintaining genome integrity in immuno-competent cells in response to extremely demanding endurance exercise. While these protective mechanisms were suggested to prevent DNA damage and/or apoptosis in a number of studies [37, 40,
45, 48, 54, 60, 62], several other exercise-associated factors induce and mediate a
systemic inflammatory response [15, 53]. This indirectly further implies that
DNA damage in immuno-competent cells, if it occurs at all, might not be a major
determinant of exercise-induced inflammation.
CONCLUSION
Thus far, there is only little evidence concerning a direct relationship between DNA
damage and inflammatory responses after strenuous prolonged exercise. The most
conclusive picture that emerges from the available data is that oxidative stress seems
to be the main link between exercise-induced inflammation and DNA damage. Considering the very few studies in which markers of DNA damage were found to correlate with signs of inflammation or muscle damage, DNA damage in peripheral
immuno-competent cells, indeed, most likely resulted from an increased generation
of RONS due to initial systemic inflammatory responses or the delayed inflammatory
processes in response to muscle damage (Fig. 1). The lack of correlations between
these exercise-induced responses in most of the studies might also be explained by
the fact that the monitoring period was too short. Hence, particular attention should
be paid to the characteristic time-course of inflammatory and oxidative stress events
on the one hand and DNA effects on the other hand. Though obvious differences
exist in the manifestation and outcomes a comparable relationship is reported in
patho-physiological conditions including carcinogenesis, where (chronic) inflammation induces DNA damage and mutations via oxidative stress [13]. However, there
might be further mechanisms that link exercise-induced DNA modulations, inflammatory responses and RONS. It has been shown, that redox-sensitive signal transduction pathways including nuclear factor (NF) κB or p53 cascades are involved in
inflammation as well as “cell stress management” in response to DNA damage [24,
30]. Recent explorations of the gene expression responses to exercise have already
shed a light on hitherto unknown molecular mechanisms in exercise immunology [5,
9, 14, 61, 84, 85]. In the future, the combination of these powerful modern techniques
(transcriptomics, proteomics) with state-of-the-art biochemical biomarkers should
therefore enable researchers in this field to provide novel insights into potential further interactions between genome stability and inflammation.
REFERENCES
1.
Angelis, K. J., M. Dusinska, and A. R. Collins. Single cell gel electrophoresis: detection of DNA damage at different levels of sensitivity. Electrophoresis. 20:21332138, 1999.
68 • Exercise-induced DNA damage and inflammatory responses
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Avula, C. P., A. R. Muthukumar, K. Zaman, R. McCarter, and G. Fernandes.
Inhibitory effects of voluntary wheel exercise on apoptosis in splenic lymphocyte
subsets of C57BL/6 mice. J Appl Physiol. 91:2546-2552, 2001.
Bonassi, S., A. Znaor, M. Ceppi, C. Lando, W. P. Chang, N. Holland, M. KirschVolders, E. Zeiger, S. Ban, R. Barale, M. P. Bigatti, C. Bolognesi, A. CebulskaWasilewska, E. Fabianova, A. Fucic, L. Hagmar, G. Joksic, A. Martelli, L. Migliore,
E. Mirkova, M. R. Scarfi, A. Zijno, H. Norppa, and M. Fenech. An increased
micronucleus frequency in peripheral blood lymphocytes predicts the risk of cancer
in humans. Carcinogenesis. 28:625-631, 2007.
Briviba, K., B. Watzl, K. Nickel, S. Kulling, K. Bos, S. Haertel, G. Rechkemmer,
and A. Bub. A half-marathon and a marathon run induce oxidative DNA damage,
reduce antioxidant capacity to protect DNA against damage and modify immune
function in hobby runners. Redox Rep. 10:325-331, 2005.
Buttner, P., S. Mosig, A. Lechtermann, H. Funke, and F. C. Mooren. Exercise affects
the gene expression profiles of human white blood cells. J Appl Physiol. 102:26-36,
2007.
Collins, A. R. and M. Dusinska. Oxidation of cellular DNA measured with the comet
assay. Methods Mol Biol. 186:147-159, 2002.
Collins, A. R., S. J. Duthie, and V. L. Dobson. Direct enzymic detection of endogenous oxidative base damage in human lymphocyte DNA. Carcinogenesis. 14:17331735, 1993.
Collins, A. R., A. A. Oscoz, G. Brunborg, I. Gaivao, L. Giovannelli, M. Kruszewski,
C. C. Smith, and R. Stetina. The comet assay: topical issues. Mutagenesis. 23:143151, 2008.
Connolly, P. H., V. J. Caiozzo, F. Zaldivar, D. Nemet, J. Larson, S. P. Hung, J. D.
Heck, G. W. Hatfield, and D. M. Cooper. Effects of exercise on gene expression in
human peripheral blood mononuclear cells. J Appl Physiol. 97:1461-1469, 2004.
Coyle, E. F. Physiological determinants of endurance exercise performance. J Sci
Med Sport. 2:181-189, 1999.
Dill, D. B. and D. L. Costill. Calculation of percentage changes in volumes of blood,
plasma, and red cells in dehydration. J Appl Physiol. 37:247-248, 1974.
Dotan, Y., D. Lichtenberg, and I. Pinchuk. Lipid peroxidation cannot be used as a
universal criterion of oxidative stress. Prog Lipid Res. 43:200-227, 2004.
Federico, A., F. Morgillo, C. Tuccillo, F. Ciardiello, and C. Loguercio. Chronic
inflammation and oxidative stress in human carcinogenesis. Int J Cancer. 121:23812386, 2007.
Fehrenbach, E. Multifarious microarray-based gene expression patterns in response
to exercise. J Appl Physiol. 102:7-8, 2007.
Fehrenbach, E. and M. E. Schneider. Trauma-induced systemic inflammatory
response versus exercise-induced immunomodulatory effects. Sports Med. 36:373384, 2006.
Fenech, M. The cytokinesis-block micronucleus technique: a detailed description of
the method and its application to genotoxicity studies in human populations. Mutat
Res. 285:35-44, 1993.
Fenech, M. In vitro micronucleus technique to predict chemosensitivity. Methods
Mol Med. 111:3-32, 2005.
Fenech, M., W. P. Chang, M. Kirsch-Volders, N. Holland, S. Bonassi, and E. Zeiger.
HUMN project: detailed description of the scoring criteria for the cytokinesis-block
Exercise-induced DNA damage and inflammatory responses • 69
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
micronucleus assay using isolated human lymphocyte cultures. Mutat Res. 534:6575, 2003.
Fischer, C. P. Interleukin-6 in acute exercise and training: what is the biological relevance? Exerc Immunol Rev. 12:6-33, 2006.
Gleeson, M. Immune function in sport and exercise. J Appl Physiol. 103:693-699,
2007.
Hartmann, A., S. Pfuhler, C. Dennog, D. Germadnik, A. Pilger, and G. Speit. Exercise-induced DNA effects in human leukocytes are not accompanied by increased
formation of 8-hydroxy-2'-deoxyguanosine or induction of micronuclei. Free Radic
Biol Med. 24:245-251, 1998.
Hartmann, A., U. Plappert, K. Raddatz, M. Grunert-Fuchs, and G. Speit. Does physical activity induce DNA damage? Mutagenesis. 9:269-272, 1994.
Hoffman-Goetz, L. and S. Zajchowski. In vitro apoptosis of lymphocytes after exposure to levels of corticosterone observed following submaximal exercise. J Sports
Med Phys Fitness. 39:269-274, 1999.
Ji, L. L. Modulation of skeletal muscle antioxidant defense by exercise: Role of
redox signaling. Free Radic Biol Med. 44:142-152, 2008.
Kendall, B. and R. Eston. Exercise-induced muscle damage and the potential protective role of estrogen. Sports Med. 32:103-123, 2002.
Knasmuller, S., A. Nersesyan, M. Misik, C. Gerner, W. Mikulits, V. Ehrlich, C. Hoelzl, A. Szakmary, and K. H. Wagner. Use of conventional and -omics based methods
for health claims of dietary antioxidants: a critical overview. Br J Nutr. 99 E Suppl
1:ES3-52, 2008.
Knez, W. L., J. S. Coombes, and D. G. Jenkins. Ultra-endurance exercise and oxidative damage : implications for cardiovascular health. Sports Med. 36:429-441, 2006.
König, D., O. Neubauer, L. Nics, N. Kern, A. Berg, E. Bisse, and K. H. Wagner. Biomarkers of exercise-induced myocardial stress in relation to inflammatory and
oxidative stress. Exerc Immunol Rev. 13:15-36, 2007.
König, D., K. H. Wagner, I. Elmadfa, and A. Berg. Exercise and oxidative stress: significance of antioxidants with reference to inflammatory, muscular, and systemic
stress. Exerc Immunol Rev. 7:108-133, 2001.
Liu, B., Y. Chen, and D. K. St Clair. ROS and p53: A versatile partnership. Free
Radic Biol Med. 44:1529-1535, 2008.
Maderna, P. and C. Godson. Phagocytosis of apoptotic cells and the resolution of
inflammation. Biochim Biophys Acta. 1639:141-151, 2003.
Malm, C. Exercise immunology: a skeletal muscle perspective. Exerc Immunol Rev.
8:116-167, 2002.
Malm, C. Exercise immunology: the current state of man and mouse. Sports Med.
34:555-566, 2004.
Mars, M., S. Govender, A. Weston, V. Naicker, and A. Chuturgoon. High intensity
exercise: a cause of lymphocyte apoptosis? Biochem Biophys Res Commun.
249:366-370, 1998.
Mastaloudis, A., J. D. Morrow, D. W. Hopkins, S. Devaraj, and M. G. Traber.
Antioxidant supplementation prevents exercise-induced lipid peroxidation, but not
inflammation, in ultramarathon runners. Free Radic Biol Med. 36:1329-1341, 2004.
Mastaloudis, A., T. W. Yu, R. P. O'Donnell, B. Frei, R. H. Dashwood, and M. G. Traber. Endurance exercise results in DNA damage as detected by the comet assay. Free
Radic Biol Med. 36:966-975, 2004.
70 • Exercise-induced DNA damage and inflammatory responses
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
Miyata, M., H. Kasai, K. Kawai, N. Yamada, M. Tokudome, H. Ichikawa, C. Goto, Y.
Tokudome, K. Kuriki, H. Hoshino, K. Shibata, S. Suzuki, M. Kobayashi, H. Goto,
M. Ikeda, T. Otsuka, and S. Tokudome. Changes of urinary 8-hydroxydeoxyguanosine levels during a two-day ultramarathon race period in Japanese non-professional
runners. Int J Sports Med. 29:27-33, 2008.
Moller, P., S. Loft, C. Lundby, and N. V. Olsen. Acute hypoxia and hypoxic exercise
induce DNA strand breaks and oxidative DNA damage in humans. Faseb J. 15:11811186, 2001.
Mooren, F. C., D. Bloming, A. Lechtermann, M. M. Lerch, and K. Volker. Lymphocyte apoptosis after exhaustive and moderate exercise. J Appl Physiol. 93:147-153,
2002.
Mooren, F. C., A. Lechtermann, and K. Volker. Exercise-induced apoptosis of lymphocytes depends on training status. Med Sci Sports Exerc. 36:1476-1483, 2004.
Nagatomi, R. The implication of alterations in leukocyte subset counts on immune
function. Exerc Immunol Rev. 12:54-71, 2006.
Neubauer, O., D. König, and K. H. Wagner. Recovery after an Ironman triathlon:
sustained inflammatory responses and muscular stress. Eur J Appl Physiol. 104:417426, 2008.
Neubauer, O., D. König, N. Kern, L. Nics, and K. H. Wagner. No Indications of Persistent Oxidative Stress in Response to an Ironman Triathlon. Med Sci Sports Exerc.
40, 2008.
Niess, A. M., M. Baumann, K. Roecker, T. Horstmann, F. Mayer, and H. H. Dickhuth. Effects of intensive endurance exercise on DNA damage in leucocytes. J
Sports Med Phys Fitness. 38:111-115, 1998.
Niess, A. M., A. Hartmann, M. Grunert-Fuchs, B. Poch, and G. Speit. DNA damage
after exhaustive treadmill running in trained and untrained men. Int J Sports Med.
17:397-403, 1996.
Nikolaidis, M. G., A. Z. Jamurtas, V. Paschalis, I. G. Fatouros, Y. Koutedakis, and D.
Kouretas. The effect of muscle-damaging exercise on blood and skeletal muscle
oxidative stress: magnitude and time-course considerations. Sports Med. 38:579606, 2008.
Okamura, K., T. Doi, K. Hamada, M. Sakurai, Y. Yoshioka, R. Mitsuzono, T. Migita,
S. Sumida, and Y. Sugawa-Katayama. Effect of repeated exercise on urinary 8hydroxy-deoxyguanosine excretion in humans. Free Radic Res. 26:507-514, 1997.
Palazzetti, S., M. J. Richard, A. Favier, and I. Margaritis. Overloaded training
increases exercise-induced oxidative stress and damage. Can J Appl Physiol. 28:588604, 2003.
Palazzetti, S., A. S. Rousseau, M. J. Richard, A. Favier, and I. Margaritis. Antioxidant supplementation preserves antioxidant response in physical training and low
antioxidant intake. Br J Nutr. 91:91-100, 2004.
Peake, J. M., K. Suzuki, and J. S. Coombes. The influence of antioxidant supplementation on markers of inflammation and the relationship to oxidative stress after
exercise. J Nutr Biochem. 18:357-371, 2007.
Peake, J. M., K. Suzuki, G. Wilson, M. Hordern, K. Nosaka, L. Mackinnon, and J. S.
Coombes. Exercise-induced muscle damage, plasma cytokines, and markers of neutrophil activation. Med Sci Sports Exerc. 37:737-745, 2005.
Pedersen, B. K., T. C. Akerstrom, A. R. Nielsen, and C. P. Fischer. Role of myokines
in exercise and metabolism. J Appl Physiol, 2007.
Exercise-induced DNA damage and inflammatory responses • 71
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
Pedersen, B. K. and L. Hoffman-Goetz. Exercise and the immune system: regulation, integration, and adaptation. Physiol Rev. 80:1055-1081, 2000.
Peters, E. M., M. Van Eden, N. Tyler, A. Ramautar, and A. A. Chuturgoon. Prolonged
exercise does not cause lymphocyte DNA damage or increased apoptosis in welltrained endurance athletes. Eur J Appl Physiol. 98:124-131, 2006.
Pilger, A., D. Germadnik, D. Formanek, H. Zwick, N. Winkler, and H. W. Rudiger.
Habitual long-distance running does not enhance urinary excretion of 8-hydroxydeoxyguanosine. Eur J Appl Physiol Occup Physiol. 75:467-469, 1997.
Pittaluga, M., P. Parisi, S. Sabatini, R. Ceci, D. Caporossi, M. Valeria Catani, I. Savini, and L. Avigliano. Cellular and biochemical parameters of exercise-induced
oxidative stress: relationship with training levels. Free Radic Res. 40:607-614, 2006.
Poulsen, H. E., S. Loft, and K. Vistisen. Extreme exercise and oxidative DNA modification. J Sports Sci. 14:343-346, 1996.
Poulsen, H. E., A. Weimann, and S. Loft. Methods to detect DNA damage by free
radicals: relation to exercise. Proc Nutr Soc. 58:1007-1014, 1999.
Radak, Z., H. Y. Chung, and S. Goto. Systemic adaptation to oxidative challenge
induced by regular exercise. Free Radic Biol Med. 44:153-159, 2008.
Radak, Z., J. Pucsuk, S. Boros, L. Josfai, and A. W. Taylor. Changes in urine 8hydroxydeoxyguanosine levels of super-marathon runners during a four-day race
period. Life Sci. 66:1763-1767, 2000.
Radom-Aizik, S., F. Zaldivar, Jr., S. Y. Leu, P. Galassetti, and D. M. Cooper. Effects
of 30 min of aerobic exercise on gene expression in human neutrophils. J Appl Physiol. 104:236-243, 2008.
Reichhold, S., O. Neubauer, V. Ehrlich, S. Knasmüller, and K.-H. Wagner. No acute
and persistent DNA damage after an Ironman triathlon. Cancer Epidemiol Biomarkers Prev. 17:1913-1919, 2008.
Reichhold, S., O. Neubauer, B. Stadlmayr, J. Valentini, C. Hoelzl, F. Ferk, S. Knasmueller, and K. H. Wagner. Oxidative DNA damage in Response to an Ironman
Triathlon. Free Radic Res, [submitted].
Rogers, C. J., L. H. Colbert, J. W. Greiner, S. N. Perkins, and S. D. Hursting. Physical activity and cancer prevention : pathways and targets for intervention. Sports
Med. 38:271-296, 2008.
Sacheck, J. M., P. E. Milbury, J. G. Cannon, R. Roubenoff, and J. B. Blumberg.
Effect of vitamin E and eccentric exercise on selected biomarkers of oxidative stress
in young and elderly men. Free Radic Biol Med. 34:1575-1588, 2003.
Sato, Y., H. Nanri, M. Ohta, H. Kasai, and M. Ikeda. Increase of human MTH1 and
decrease of 8-hydroxydeoxyguanosine in leukocyte DNA by acute and chronic exercise in healthy male subjects. Biochem Biophys Res Commun. 305:333-338, 2003.
Schiffl, C., C. Zieres, and H. Zankl. Exhaustive physical exercise increases frequency of micronuclei. Mutat Res. 389:243-246, 1997.
St Pierre Schneider, B. and P. M. Tiidus. Neutrophil infiltration in exercise-injured
skeletal muscle : how do we resolve the controversy? Sports Med. 37:837-856, 2007.
Steensberg, A., J. Morrow, A. D. Toft, H. Bruunsgaard, and B. K. Pedersen. Prolonged exercise, lymphocyte apoptosis and F2-isoprostanes. Eur J Appl Physiol.
87:38-42, 2002.
Sumida, S., T. Doi, M. Sakurai, Y. Yoshioka, and K. Okamura. Effect of a single bout
of exercise and beta-carotene supplementation on the urinary excretion of 8hydroxy-deoxyguanosine in humans. Free Radic Res. 27:607-618, 1997.
72 • Exercise-induced DNA damage and inflammatory responses
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
Suzuki, K., S. Nakaji, M. Yamada, Q. Liu, S. Kurakake, N. Okamura, T. Kumae, T.
Umeda, and K. Sugawara. Impact of a competitive marathon race on systemic
cytokine and neutrophil responses. Med Sci Sports Exerc. 35:348-355, 2003.
Suzuki, K., J. Peake, K. Nosaka, M. Okutsu, C. R. Abbiss, R. Surriano, D. Bishop,
M. J. Quod, H. Lee, D. T. Martin, and P. B. Laursen. Changes in markers of muscle
damage, inflammation and HSP70 after an Ironman triathlon race. Eur J Appl Physiol. 98:525-534, 2006.
Suzuki, K., H. Sato, T. Kikuchi, T. Abe, S. Nakaji, K. Sugawara, M. Totsuka, K.
Sato, and K. Yamaya. Capacity of circulating neutrophils to produce reactive oxygen
species after exhaustive exercise. J Appl Physiol. 81:1213-1222, 1996.
Suzuki, K., M. Totsuka, S. Nakaji, M. Yamada, S. Kudoh, Q. Liu, K. Sugawara, K.
Yamaya, and K. Sato. Endurance exercise causes interaction among stress hormones, cytokines, neutrophil dynamics, and muscle damage. J Appl Physiol.
87:1360-1367, 1999.
Tee, J. C., A. N. Bosch, and M. I. Lambert. Metabolic consequences of exerciseinduced muscle damage. Sports Med. 37:827-836, 2007.
Thune, I. and A. S. Furberg. Physical activity and cancer risk: dose-response and
cancer, all sites and site-specific. Med Sci Sports Exerc. 33:S530-550; discussion
S609-510, 2001.
Tice, R. R., E. Agurell, D. Anderson, B. Burlinson, A. Hartmann, H. Kobayashi, Y.
Miyamae, E. Rojas, J. C. Ryu, and Y. F. Sasaki. Single cell gel/comet assay: guidelines for in vitro and in vivo genetic toxicology testing. Environ Mol Mutagen.
35:206-221, 2000.
Tidball, J. G. Inflammatory processes in muscle injury and repair. Am J Physiol
Regul Integr Comp Physiol. 288:R345-353, 2005.
Traber, M. G. Relationship of vitamin E metabolism and oxidation in exercising
human subjects. Br J Nutr. 96 Suppl 1:S34-37, 2006.
Tsai, K., T. G. Hsu, K. M. Hsu, H. Cheng, T. Y. Liu, C. F. Hsu, and C. W. Kong.
Oxidative DNA damage in human peripheral leukocytes induced by massive aerobic
exercise. Free Radic Biol Med. 31:1465-1472, 2001.
Umegaki, K., P. Daohua, A. Sugisawa, M. Kimura, and M. Higuchi. Influence of one
bout of vigorous exercise on ascorbic acid in plasma and oxidative damage to DNA
in blood cells and muscle in untrained rats. J Nutr Biochem. 11:401-407, 2000.
Umegaki, K., M. Higuchi, K. Inoue, and T. Esashi. Influence of one bout of intensive
running on lymphocyte micronucleus frequencies in endurance-trained and
untrained men. Int J Sports Med. 19:581-585, 1998.
Voll, R. E., M. Herrmann, E. A. Roth, C. Stach, J. R. Kalden, and I. Girkontaite.
Immunosuppressive effects of apoptotic cells. Nature. 390:350-351, 1997.
Zieker, D., E. Fehrenbach, J. Dietzsch, J. Fliegner, M. Waidmann, K. Nieselt, P.
Gebicke-Haerter, R. Spanagel, P. Simon, A. M. Niess, and H. Northoff. cDNA
microarray analysis reveals novel candidate genes expressed in human peripheral
blood following exhaustive exercise. Physiol Genomics. 23:287-294, 2005.
Zieker, D., J. Zieker, J. Dietzsch, M. Burnet, H. Northoff, and E. Fehrenbach.
CDNA-microarray analysis as a research tool for expression profiling in human
peripheral blood following exercise. Exerc Immunol Rev. 11:86-96, 2005.
Direct detection of gene doping • 73
Establishing a novel single-copy primer-internal
intron-spanning PCR (spiPCR) procedure for the
direct detection of gene doping
Thomas Beiter1, Martina Zimmermann2, Annunziata Fragasso1, Sorin
Armeanu 2, Ulrich M. Lauer 2, Michael Bitzer 2, Hua Su 3,4, William L.
Young3,5,6, Andreas M. Niess1 and Perikles Simon1
1
Department of Sports Medicine, Medical Clinic,
Department of Gastroenterology & Hepatology, Medical Clinic, University of
Tuebingen, Germany
3 Center for Cerebrovascular Research, Department of Anesthesia and Perioperative Care,
4 Cardiovascular Research Institute,
5 Department of Neurological Surgery,
6 Department of Neurology, University of California, San Francisco, California,
USA
2
ABSTRACT
So far, the abuse of gene transfer technology in sport, so-called gene doping, is
undetectable. However, recent studies in somatic gene therapy indicate that longterm presence of transgenic DNA (tDNA) following various gene transfer protocols can be found in DNA isolated from whole blood using conventional PCR
protocols. Application of these protocols for the direct detection of gene doping
would require almost complete knowledge about the sequence of the genetic
information that has been transferred. Here, we develop and describe the novel
single-copy primer-internal intron-spanning PCR (spiPCR) procedure that overcomes this difficulty. Apart from the interesting perspectives that this spiPCR procedure offers in the fight against gene doping, this technology could also be of
interest in biodistribution and biosafety studies for gene therapeutic applications.
Key Words: Gene doping, gene therapy, gene transfer, direct detection, spiPCR,
transgenic DNA
Address correspondence to
Perikles SIMON, MD, PhD, Department of Sports Medicine,
Medical Clinic, University of Tübingen
Silcherstr. 5, D-72076 Tuebingen, Germany, Phone: ++49-7071-2985163
perikles@uni-tuebingen.de
74 • Direct detection of gene doping
INTRODUCTION
The World Anti-Doping Agency (WADA) gives the following description for the
forbidden method gene doping in its upcoming 2009 Prohibited List: “The transfer of cells or genetic elements or the use of cells, genetic elements or pharmacological agents to modulat[e] expression of endogenous genes having the capacity
to enhance athletic performance, is prohibited.”
Modulation of the expression of endogenous genes is typically achieved by virtually every conventional doping substance including human or non-human peptides (1-4) and anabolic androgenic steroids (5). Accordingly, scientific articles on
gene doping usually point out or take for granted that gene transfer technology
has to be used in order to justify the term gene doping (6-16).
In the following, we will therefore use the term gene doping in its stricter definition that is generally approved within the scientific community - as the abuse of
gene transfer technology to enhance athletic performance. According to this definition an athlete who practices gene doping incorporates “an extra” amount of
genetic information (DNA or RNA) by means of gene therapeutic procedures.
The added genetic information itself can be of human origin and is not the direct
source of the performance enhancing effect. If the incorporated genetic information is DNA, it is called transgenic DNA (tDNA) and serves as a template to produce a protein within the athlete’s body that is known to improve physical performance, such as erythropoietin (EPO) (17). More recently, the development of
antisense RNA technology and advances in the delivery of such RNA molecules
have additionally opened the possibility to specifically inhibit the production of
proteins that limit or restrict physical performance on a natural basis (18).
The definition given above already implicates the three major and unique problems associated with gene doping, which are as follows:
I.
The genetically modified athlete
Depending on the stability and functionality of the introduced genetic
information an athlete could have a permanently genetically modified
physical performance. This imposes exceptional ethical concerns (14;19).
II. Undetectability
Gene doping is regarded as principally undetectable, since the introduced
gene is of human origin and the protein which mediates the performance
enhancing characteristics is even built within the athlete’s own body (810;13).
III. Safety concerns
Gene therapeutic interventions are subject to very tight safety regulations (2022). The unapproved use of gene transfer technology in athletes may not only
be of high risk to the individual, but also to others that get in contact with, or
inappropriately handle substances required for gene transfer. On top of this it
is well known that intensive exercise can severely influence the innate immune
response (23-27). Especially in elite athletes this might have unknown consequences on the first line of immune response to gene transfer related interventions that are conducted without appropriate medical supervision.
Direct detection of gene doping • 75
Currently it is only assumed that athletes may already practice gene doping (7).
Nevertheless, the subject gene doping, i.e. the consideration of candidate genes
and techniques for potential abuse and its potential detection strategies, has been
reviewed in 27 scientific articles during the past 5 years. Table 1 summarizes
solutions for detection strategies of gene doping that have been proposed and the
potential problems associated with these solutions.
Three main aspects lead to the conclusion that the abuse of gene transfer technology in sports will be very difficult to detect:
I.
Homology between gene transfer material and the normal human body
(8;18)
The material typically introduced into the body by gene transfer is frequently found in the normal human population. It is either of human origin like the tDNA itself or it contains additional non-human material and
molecules that most humans are in frequent contact with - like viral proteins, viral DNA sequences or other material that is required for the transfer or the proper function of the tDNA (Table 1; lines 1 and 2).
II. Homology of the generated protein with the natural protein (6;10;13;28)
Following gene transfer, the human body itself produces the doping relevant protein, which may therefore be principally undetectable by direct
detection methods (Table 1; line 3).
III. Limited specificity of indirect procedures
Indirect measurements on the level of the doping effect or on the level of
the bystander effects provoked on the transcriptome or proteome in
peripheral blood cells may only be of limited specificity for doping. Nevertheless, such procedures might be very important for pre-screening of
samples and may contribute an important suspicious fact (Table 1; lines 4
and 5).
Table 1
In this article we describe for the first time a procedure that principally enables
direct detection of gene doping on the level of human tDNA. As a basis for diagnostic discrimination, the gene sequences of human tDNA being employed by
76 • Direct detection of gene doping
gene transfer protocols are not 100% homologous to the human genomic DNA
(gDNA), since these do not contain the intronic sequence parts of the gDNA (see
Fig. 1, upper versus lower part). We have developed a method enabling detection
of tDNA on a single molecule level within ordinary blood samples. Detection is
based on specific amplification of tDNA even in the presence of huge amounts of
gDNA by a patent pending single-copy primer-internal intron-spanning PCR
(spiPCR) procedure (PCT/EP2007/003385;
http://www.wipo.int/pctdb/en/wo.jsp?WO=2007124861).
This spiPCR procedure can principally be used to directly detect gene doping
using any kind of gene transfer protocol that either works with DNA or leads to
the integration of tDNA into our genome. Currently, 1457 of the 1472 registrated
Clinical Gene Therapy Trials use a gene transfer procedure that leads to the generation of such tDNA sequences (http://www.wiley.co.uk/genetherapy/clinical/).
In this article we firstly describe a spiPCR protocol for the detection of gene
transfer focusing on the first line gene doping candidate genes EPO and Vascular
Endothelial Growth Factor-D (VEGF-D). Secondly, we then discuss the perspectives of spiPCR for the application as a direct detection technique for gene doping.
Fig. 1: In case primers for the PCR can be chosen in a way as illustrated for the black pair
of primers, maximum specificity is achieved by assuring that only tDNA, but not gDNA
can be primed and amplified by both primers. This principle is called primer-internal
intron-spanning (pi). In the case of the dark grey primer pair, only the reverse primer shows
this primer-internal intron-spanning, while the forward primer does not. In the case of the
light dark primers the choice of primers is either termed intron-spanning or exon-skipping
and usually is used to differentiate between gDNA and cDNA, since they can be differentiated by the size of the generated amplicons. However, in such a setting the primers can still
bind and amplify both, gDNA (upper part) and tDNA (lower part). This does not only
reduce specificity remarkably, but will also reduce sensitivity for the tDNA amplification.
Direct detection of gene doping • 77
METHODS
spiPCR-based detection of Epo and Vegf-d tDNA
The principle of spiPCR-based detection of tDNA is illustrated in Fig. 1. Whereas
coding sequences of gDNA are 100% homologous to coding sequences of any
tDNA, gDNA contains introns, whereas tDNA does not. This difference can be
used to discriminate tDNAs abused in gene doping from “parental” gDNA
sequences.
According to the above illustrated principles of tDNA detection (Fig. 1), primers
have to be chosen with respect to two main points that are important for the sensitivity and specificity of the tDNA detection:
(i) Primer-internal intron-spanning: every single primer spans an intron. For this
purpose, the first bases (5’ end) of every forward primer are located in an exon
upstream to the exon where the last bases (3’ end) are located and the first
bases of a reverse primer are located in an exon downstream to the exon
where the last bases are located (Fig. 1, black primers). Subsequent to mRNA
splicing these primers will bind only at the exon-junctions in the tDNA.
(ii) tDNA specificity: none of the primers shows a high enough homology to
hybridize anywhere else in the human genome.
Selection of primers within the coding sequence of a candidate gene has therefore
been done with respect to the gene specific exon-intron structure. We used the
Blast Like Alignment Tool (BLAT) from the UCSC Genome browser for the
alignment of reference gene coding sequences to the human genome (29). Fig. 2
shows an example for such an alignment for the locus of erythropoietin (EPO) on
chromosome 7.
Fig. 2: BLAT alignment of the reference mRNA sequence of EPO with its gene locus. Four
potential exon-intron junctions (boxes; “conserved sequence part”) can be found all of
which are suitable according to principle (i) to serve as regions where exon-intron spanning primers can be located. The “conserved sequence part” needs to be a sequence part
that is conserved among various different mRNA sequences that could be translated into a
functional protein.
For Epo a spiPCR protocol was established with an outer primer pair
”EPOs1/as3” for the amplification of a 1st round 437 bp PCR product and an inner
primer pair “EPOs2+3/EPOas3-II” flanking the 2nd round 289 bp product. Localization of the above mentioned primers to the EPO gene locus and the reference
RNA sequence of Epo is shown in Fig. 3. Sequences of primers are given in
78 • Direct detection of gene doping
Table 2. Primers were purchased from MWG (Ebersberg, Germany). All primers
are intron-spanning and are within the region that is canonical to the reference gene
mRNA sequences that is known to be protein coding (black alignment in Fig. 3).
Table 2
Primer
Sequence
EPOs1 (outer forward)
5’-ATGGGGGTGCACGAATGTC-3’
EPOas3 (outer reverse)
5’-ATGGCTTCCTTCTGGGCTC-3’
EPOs2+3 (inner forward)
5’-AGAATATCACGACGGGCTGTG-3’
EPOas3-II (inner reverse)
5’-TCCTTCTGGGCTCCCAGAG-3’
vegfD_1s (outer forward)
5’-CCTCGTACATTTCCAAACAGCTC-3’
vegfD_1as (outer reverse)
5’-TCCTGGAGATGAGAGTGGTCTTC-3’
vegfD_2s (inner forward)
5’-AAGAAGATCGCTGTTCCCATTC-3’
vegfD_2as (inner reverse)
5’-AGAGTGGTCTTCTGTTCCAGCA-3’
For Vegf-d a spiPCR protocol was established as above with primer-internal
intron-spanning primer pairs that amplify a 1st round 289 bp PCR product and a
2nd round 119 bp PCR product.
Fig. 3: Localization of EPO primers to the respective gene locus.
First and second round of the nested PCRs were prepared with Promega GoTaq®
Green Master Mix (Promega, Madison, Wisconsin, USA) containing a HotStart
Polymerase to avoid unspecific nucleotide incorporation prior to the first PCR
denaturation step. Reactions were set up under bench top UV cabinets, using
PCR-dedicated pipettes and filter tips. Preparation of PCR Master Mix, extraction
and addition of DNA samples were performed in three separate areas.
The first round (outer) PCR contained Promega GoTaq® Green Master Mix, 0.3
µM of each outer primer and ~300ng genomic DNA in a total volume of 25 or 50
µl, respectively. The positive and negative controls were pipetted by adding
defined copies of the tDNA standard or the same volume of nuclease-free water
(Promega, Madison, Wisconsin, USA), respectively.
Amplification started with a single denaturation step of 94 °C for 3 min to activate
the HotStart enzyme. To reduce unspecific amplification of by-products, a touchdown PCR protocol was used during the first six cycles of the 1st round PCR,
starting with an annealing temperature of 63 °C and decreasing the annealing
temperature by 0.5 °C/cycle to reach the optimum annealing temperature of 59 °C
which was subsequently used for additional 14 cycles. Each cycle consisted of
Direct detection of gene doping • 79
denaturation at 94 °C for 20 sec, annealing for 25 sec, and elongation at 72 °C for
35 sec. Final extension was performed at 72 °C for 7 min.
The second round (inner) PCR was performed using 2.5 (5) µl of the first PCR
product in a 25 (50) µl reaction mixture containing Promega GoTaq® Green Master Mix and 0.3 µM of each inner primer. PCR was performed as follows: initial
denaturation at 94 °C for 3 min, followed by 30 cycles of 94 °C for 20 sec, 58 °C
for 25 sec, and 72 °C for 35 sec, and a final extension for 7 min at 72 °C.
The final PCR product was analyzed on a 1.5% agarose gel and visualized by UV
illumination after staining with GelRed (Biotium Inc., Hayward, CA).
Optimum annealing temperature and the specificity of the PCR-products generated during the respective PCR amplification rounds was tested separately prior to
application of the spiPCR protocols.
The effectiveness of the spiPCR protocols was tested on different preparations of
300 ng total DNA from whole blood that were spiked with known copy numbers
ranging from 1 - 1000 of respective tDNAs as positive controls. Unspiked DNA
samples represented negative controls.
All tDNA standards were constructed by target specific PCR from cDNA
libraries. For this purpose 1 kb standards were generated that included the whole
locus of interest. Concentrations of the respective standard tDNAs were determined both photometrically and by photodensitometry from serial dilutions run
on 1.2% agarose gels using Quantity One 1-D Analysis Software (BioRad, Germany). Copy numbers were calculated and standards with defined copy numbers
were prepared by serial dilutions.
Preparation of total DNA from whole blood samples
The isolation of total DNA from 200 µl of EDTA whole blood was performed
with a silica-gel-membrane based method by applying the QIAamp DNA Blood
Mini Kit according to the manufacturer’s instruction manual (Qiagen, Hilden;
Germany) with a final elution volume of 100 µl. In some cases and for refined
analysis the yielded DNA was further concentrated by an additional isopropanol
precipitation step.
Construction of Ad-Vegf-d vector and in vitro gene transfer
A recombinant adenoviral vector encoding the Vegf-d transgene (Ad-Vegf-d) purposefully was purchased as a ready to use virus stock from Vector Biolabs
(Philadelphia, PA), thereby simulating a ´classical gene doping initiation scenario` (i.e. vector purchase via the internet).
The virus has a backbone of the human Adenovirus Type 5 with partial deletions
in the E1 and E3 domains. In this adenoviral vector, the expression of the Vegf-d
transgene was placed under the control of a CMV promoter. The adenoviral vector was amplified on 293 cells and subsequently purified by centrifugation as
described previously (30). Stocks of 1011 pfu/ml were stored at -80 °C.
Expression of the Vegf-d transgene was verified by infection of HeLa cells and
detection of VEGF-D protein in culture supernatants using a commercially available ELISA system (Quantikine-Human VEGF-D Immunoassay from R&D Systems; DVED00; data not shown).
For the spiking experiments of whole blood, the transduction efficiency of U937
cells with Ad-Vegf-d was determined by FACS analysis 7 days post-trans-
80 • Direct detection of gene doping
Results
duction using a FITC-conjugated goat anti-adenovirus antibody (Chemicon,
AB1056F, 1:100). At a multiplicity of infection (MOI) of 100 (i.e. 100 plaque
forming untis (pfu) / cell) 0,5% of the U937 cells stained positive at this time
point.
Results
RESULTS
The established spiPCR protocol for Epo tDNA
In spiPCR experiments we were able to detect down to 1 copy of Epo tDNA in the
Apresence of 300 ng of genomic DNA (Fig. 4). No by-products were detected
except primer dimers < 50 bp. All negative controls were tested negative. All
spiPCR results were verified three times.
A
Fig. 4:
Fig. 4: Outcome for the spiPCR-protocol with the primers for Epo. Lanes 1-4 represent the
controls (~300 ng gDNA), and lanes 5-16 represent ~300 ng gDNA with decreasing spike in copies of an Epo standard; copy number: 1000 (lanes 5-7), 100 (lanes 8-10),
10 (lanes 11-13), 1 (lanes 14-16).
Fig.
4:
negative
The established spiPCR protocol for Vegf-d tDNA
In spiPCR experiments we were able to detect down to 1 copy of Vegf-d tDNA
(Fig. 5). No by-products were detected except primer dimers < 50 bp. All negative
controls were tested negative and all spiPCR results were verified three times.
Fig.
Fig. 5:
5:
Outcome for spiPCRs with the above mentioned primers for Vegf-d. Lanes 1-4 represent the negative controls (~300 ng gDNA), and lanes 5-16 represent ~300 ng gDNA
with increasing spike in copies of a Vegf-d standard; copy numbers: 1 (lanes 5-7), 10 (lanes
Fig. 5:
8-10), 100 (lanes 11-13), 1000 (lanes 14-16).
Direct detection of gene doping • 81
Detection of Vegf-d from transduced cells in whole blood samples
The detectability of tDNA in body samples has been shown for extracellular tDNA.
Extracellular tDNA is expected only shortly after viral gene transfer in vivo as a
result of direct virus input into the blood circulation post-injection. A functional
gene transfer requires transduced cells. These may be target cells in a solid tissue
but also circulating blood cells that eventually got transduced. In the following
experiment transduction of blood cells with recombinant adenoviral vectors was
simulated by spiking blood with Vegf-d transduced cells at known cell numbers.
Therefore, U937 cells were infected with Ad-Vegf-d at MOI 10 and MOI 100.
Cells were washed several times and transferred to new vials to avoid free viral
particles. Transduction efficiency 7 days later was determined to be 0.1% for MOI
10 and 0.5% at MOI 100. This low transduction efficiency may be a result of a
more rapid proliferation of non-transduced cells compared to transduced cells in
the culture dish. Furthermore, a limited sensitivity of the detection of adenoviral
proteins by immunocytochemistry might underestimate the infected cell number
in our assay at least to some extent. For the further experiments U937 cells,
infected with MOI 100, were used to spike blood samples with a specified number of transduced cells and the DNA was isolated directly after spiking.
The following analysis of samples was done in a blinded fashion. Among the
samples tested in this way an additional internal negative control along with 9
samples that on a calculated basis had less than 0.5 transduced cells per sample
volume were investigated. In addition, an internal PCR negative control from previous extractions was run in parallel to ensure that no contamination did occur
during the PCR process (Fig. 6; first two lanes).
After isolation of DNA and subsequent additional precipitation with isopropanol,
all PCRs were run with the total gDNA harvested from 25 µl whole blood (~ 1 µg)
each and tested for the presence of Vegf-d tDNA. Transfected cells down to a calculated number of 2.25 cells / µl blood could be detected in blood DNA preparations (Fig. 6; lane 3).
Fig.
Fig. 6:
6:
One negative control (ø) and the probes 1-9 were precipitated with isopropanol and
run in duplicates with a spiPCR for Vegf-d. The calculated number of transduced cells / µl
blood was as follows: 112.5 cells in sample 2, 22.5 cells in sample 1, 12.5 cells in sample
4, 2.25 cells in sample 3. Samples 6-9 contained less than 0.25 cells / µl blood. Note that
the gDNA put into the sample was derived from 25 µl of whole blood sample.
DISCUSSION
To become effective as evidence of gene doping in a court of law, direct detection
Discussion
techniques for gene doping have to be developed. So far, there had been only one
report that suggested a solution for the direct detection of gene doping (28;30). At
82 • Direct detection of gene doping
the protein level it has been shown that it could be possible to discriminate
between EPO proteins derived from genomic DNA (gDNA) and proteins artificially encoded by tDNA using a conventional test for doping with recombinant
EPO. However, it is not yet elucidated why tDNA derived proteins can have differing post-translational modifications and under which circumstances such differences occur. Detection of gene doping on the level of the protein derived from
transgenes may therefore face the problem that differences in post-translational
modifications are highly variable depending on the protocol for gene transfer, the
transgene delivered, the route of vector administration and the target tissue.
Help and orientation for the development of more generalisable direct gene doping detection procedures may come from clinical research in somatic gene therapy. In somatic gene therapy the tDNA sequence transferred to a patient is known
and researchers are principally able to use tDNA specific sequence parts in order
to design a PCR that is able to detect the tDNA. PCR is therefore routinely used
for monitoring plasma and serum levels of tDNA to control for the presence of
infectious vector in the blood stream during somatic gene therapy trials and related animal studies. From these tests we know that serum and plasma probes will
only show tDNA for a very limited number of hours up to a few days following
many different kinds of administration routes and gene transfer technologies used
(8;31).
The test for presence of tDNA in whole blood is relatively rarely performed, since
it is not indicative for the presence of infectious vector, but rather for a back dated
transfection of blood cells. On top of this, it is technically much more difficult
than testing serum and plasma probes.
Importantly, in blood cells long-term presence of tDNA was found following
injection of (i) recombinant adenovirus (Ad) into the prostate of humans for 76
days (32), (ii) recombinant adeno-associated virus (AAV) into the muscle of primates for 10 months (33), (iii) recombinant adeno-associated virus into the hepatic artery to target the liver of humans for 20 weeks (34), and (iv) retroviruses into
the peripheral vein of humans for > 1 year (35). In all of these cases serum or
plasma probes were found to become predominantly negative within hours or a
few days. While up to now most publications speculate that muscle biopsies
might be necessary to detect gene doping at all, the above mentioned findings
indicate that it seems to be very promising to develop a technique that is able to
detect tDNA relevant for gene doping in whole blood samples. Additionally, it is
known that exercise increases the turnover and redistribution of leukocytes within
the human body (23;24;36), which may increase the likelihood that leukocytes,
once transfected at the site of vector application, can be found in the blood
stream.
The challenging technical difficulty for tests of whole blood in contrast to plasma
and serum is the sensitive and specific amplification of tDNA in the presence of
huge amounts of genomic DNA (gDNA). In the case of gene doping this technical
difficulty is further complicated, since the non-human part of the tDNA sequence
is completely unknown and highly divergent between different gene transfer procedures employing different sources of vectors.
Here we show for the first time that these technical difficulties can be overcome
by employing our novel spiPCR technology. This spiPCR procedure enables
direct detection of the doping relevant sequence - namely the sequence part that is
Direct detection of gene doping • 83
necessary to generate the protein that mediates the enhancement of physical performance. This sequence part has to be present in any gene doping attempt. Further studies are now under way to verify the specificity of this attempt to detect
gene doping by spiPCR.
First, we will establish more spiPCR protocols enrolling the most important
tDNA sequences that could be abused for gene doping. We will then try to develop a multiplex spiPCR that is able to detect as many as possible of these tDNA
sequences at once.
Second, we will investigate the specificity of the spiPCR in normal persons and
athletes known not to be genetically altered by gene transfer technology. For this
purpose, probes taken from athletes and persons under different conditions
including following intensive exercise will be analyzed for false positive results.
Third, the sensitivity of spiPCR will be tested in animal studies and on blood samples taken from patients that have undergone somatic gene therapy.
Apart from the interesting perspectives spiPCR offers in the fight against gene
doping, this technology may also be of interest in biodistribution and biosafety
studies for gene therapeutic applications. The crucial quality feature of spiPCR is
the high sensitivity to detect a tDNA sequence part being relevant for gene therapy or for gene doping against a high background of gDNA.
List of Abbreviations
AAV
Ad
Ad-Vegf-d
bp
FACS
FITC
gDNA
MOI
spiPCR
tDNA
WADA
adeno-associated Virus
adenovirus
adenoviral vector with Vegf-d
base pairs
fluorescence activated cell sorting
fluorescein isothiocyanate
genomic DNA
multiplicity of infection
single-copy primer-internal intron-spanning PCR
transgenic DNA
World Anti-Doping Agency
List of genes mentioned
EPO
VEGF-D
Erythropoietin
Vascular endothelial growth factor family member D
Acknowledgement
We thank Andrea Schenk and Irina Smirnow for excellent technical assistance in
virological procedures.
This project has been carried out with the support of WADA (research grant
06B7PS). PS received additional funding by the “Bundesinstitut für
84 • Direct detection of gene doping
Sportwissenschaft”, research grant IIA1-080308/08. The University of Tübingen,
Germany has a patent pending for the “Detection of transgenic DNA”
(PCT/EP2007/003385; http://www.wipo.int/pctdb/en/wo.jsp?WO=2007124861)
that describes the technique of spiPCR.
REFERENCE LIST
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
Komor M, Guller S, Baldus CD, de VS, Hoelzer D, Ottmann OG, et al. Transcriptional profiling of human hematopoiesis during in vitro lineage-specific differentiation. Stem Cells 2005 Sep;23(8):1154-69.
Kawai M, Namba N, Mushiake S, Etani Y, Nishimura R, Makishima M, et al.
Growth hormone stimulates adipogenesis of 3T3-L1 cells through activation of the
Stat5A/5B-PPARgamma pathway. J Mol Endocrinol 2007 Feb;38(1-2):19-34.
Svensson J, Tivesten A, Sjogren K, Isaksson O, Bergstrom G, Mohan S, et al. Liverderived IGF-I regulates kidney size, sodium reabsorption, and renal IGF-II expression. J Endocrinol 2007 Jun;193(3):359-66.
Simon P, Fehrenbach E, Niess AM. Regulation of immediate early gene expression
by exercise: short cuts for the adaptation of immune function. Exerc Immunol Rev
2006;12:112-31.
Labrie F, Luu-The V, Calvo E, Martel C, Cloutier J, Gauthier S, et al. Tetrahydrogestrinone induces a genomic signature typical of a potent anabolic steroid. J
Endocrinol 2005 Feb;184(2):427-33.
Sweeney HL. Gene doping. Sci Am 2004 Jul;291(1):62-9.
Adam D. Gene therapy may be up to speed for cheats at 2008 Olympics. Nature
2001 Dec 6;414(6864):569-70.
Baoutina A, Alexander IE, Rasko JE, Emslie KR. Developing strategies for detection
of gene doping. J Gene Med 2008 Jan;10(1):3-20.
Harridge SD, Velloso CP. Gene doping. Essays Biochem 2008;44:125-38.
Minunni M, Scarano S, Mascini M. Affinity-based biosensors as promising tools for
gene doping detection. Trends Biotechnol 2008 May;26(5):236-43.
Baoutina A, Alexander IE, Rasko JE, Emslie KR. Potential use of gene transfer in
athletic performance enhancement. Mol Ther 2007 Oct;15(10):1751-66.
Striegel H, Simon P. [Doping. High-tech cheating in sport]. Internist (Berl) 2007
Jul;48(7):737-42.
Haisma HJ, de HO. Gene doping. Int J Sports Med 2006 Apr;27(4):257-66.
Schneider AJ, Friedmann T. Gene doping in sports: the science and ethics of genetically modified athletes. Adv Genet 2006;51:1-110.
Unal M, Ozer UD. Gene doping in sports. Sports Med 2004;34(6):357-62.
McCrory P. Super athletes or gene cheats? Br J Sports Med 2003 Jun;37(3):192-3.
Diamanti-Kandarakis E, Konstantinopoulos PA, Papailiou J, Kandarakis SA,
Andreopoulos A, Sykiotis GP. Erythropoietin abuse and erythropoietin gene doping:
detection strategies in the genomic era. Sports Med 2005;35(10):831-40.
Wells DJ. Gene doping: the hype and the reality. Br J Pharmacol 2008
Jun;154(3):623-31.
Murray TH. Reflections on the ethics of genetic enhancement. Genet Med 2002
Nov;4(6 Suppl):27S-32S.
Direct detection of gene doping • 85
(20) Bamford KB, Wood S, Shaw RJ. Standards for gene therapy clinical trials based on
pro-active risk assessment in a London NHS Teaching Hospital Trust. QJM 2005
Feb;98(2):75-86.
(21) Manilla P, Rebello T, Afable C, Lu X, Slepushkin V, Humeau LM, et al. Regulatory
considerations for novel gene therapy products: a review of the process leading to
the first clinical lentiviral vector. Hum Gene Ther 2005 Jan;16(1):17-25.
(22) Gonin P, Buchholz CJ, Pallardy M, Mezzina M. Gene therapy bio-safety: scientific
and regulatory issues. Gene Ther 2005 Oct;12 Suppl 1:S146-S152.
(23) Horn P, Kalz A, Lim CL, Pyne D, Saunders P, Mackinnon L, et al. Exercise-recruited
NK cells display exercise-associated eHSP-70. Exerc Immunol Rev 2007;13:10011.
(24) Nagatomi R. The implication of alterations in leukocyte subset counts on immune
function. Exerc Immunol Rev 2006;12:54-71.
(25) Gleeson M, McFarlin B, Flynn M. Exercise and Toll-like receptors. Exerc Immunol
Rev 2006;12:34-53.
(26) Peake JM, Nosaka K, Muthalib M, Suzuki K. Systemic inflammatory responses to
maximal versus submaximal lengthening contractions of the elbow flexors. Exerc
Immunol Rev 2006;12:72-85.
(27) Shing CM, Ogawa K, Zhang X, Nagatomi R, Peake JM, Suzuki K, et al. Reduction
in resting plasma granulysin as a marker of increased training load. Exerc Immunol
Rev 2007;13:89-99.
(28) Lasne F, Martin L, de Ceaurriz J, Larcher T, Moullier P, Chenuaud P. "Genetic Doping" with erythropoietin cDNA in primate muscle is detectable. Mol Ther 2004
Sep;10(3):409-10.
(29) Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler AM, et al. The
human genome browser at UCSC. Genome Res 2002 Jun;12(6):996-1006.
(30) Wybranietz WA, Prinz F, Spiegel M, Schenk A, Bitzer M, Gregor M, et al. Quantification of VP22-GFP spread by direct fluorescence in 15 commonly used cell lines. J
Gene Med 1999 Jul;1(4):265-74.
(31) Gonin P, Gaillard C. Gene transfer vector biodistribution: pivotal safety studies in
clinical gene therapy development. Gene Ther 2004 Oct;11 Suppl 1:S98-S108.
(32) Freytag SO, Khil M, Stricker H, Peabody J, Menon M, Peralta-Venturina M, et al.
Phase I study of replication-competent adenovirus-mediated double suicide gene
therapy for the treatment of locally recurrent prostate cancer. Cancer Res 2002 Sep
1;62(17):4968-76.
(33) Favre D, Provost N, Blouin V, Blancho G, Cherel Y, Salvetti A, et al. Immediate and
long-term safety of recombinant adeno-associated virus injection into the nonhuman
primate muscle. Mol Ther 2001 Dec;4(6):559-66.
(34) Manno CS, Pierce GF, Arruda VR, Glader B, Ragni M, Rasko JJ, et al. Successful
transduction of liver in hemophilia by AAV-Factor IX and limitations imposed by the
host immune response. Nat Med 2006 Mar;12(3):342-7.
(35) Powell JS, Ragni MV, White GC, Lusher JM, Hillman-Wiseman C, Moon TE, et al.
Phase 1 trial of FVIII gene transfer for severe hemophilia A using a retroviral construct administered by peripheral intravenous infusion. Blood 2003 Sep
15;102(6):2038-45.
(36) Kruger K, Mooren FC. T cell homing and exercise. Exerc Immunol Rev 2007;13:3754.
86 • Gender specific gene response to exercise
Gender- and menstrual phase dependent regulation
of inflammatory gene expression in response to
aerobic exercise
Hinnak Northoff 1+, Stephan Symons2+, Derek Zieker1,4, Eva V. Schaible3,
Katharina Schäfer1, Stefanie Thoma3, Markus Löffler1,4, Asghar Abbasi1,
Perikles Simon3, Andreas M. Niess3 and Elvira Fehrenbach1
1
Institute for Clinical and Experimental Transfusion Medicine (IKET), University of Tübingen, Germany,
2 Center for Bioinformatics Tübingen (ZBIT)
3 Medical Clinic, Department of Sports Medicine; University of Tübingen, Germany
4 Department of general and transplant surgery, University of Tübingen, Germany
+ equally contributing authors
ABSTRACT
The immunological reaction to exercise has been investigated with increasing
intensity in the last 10-20 years, with most human studies performed in male subjects. Recently, gender-specific aspects have received growing attention, but studies carefully monitoring the influence of gender, including the menstrual cycle,
are rare. Here, we report gene expression patterns in response to a run at 93% of
the individual anaerobic threshold of 9 women with regular menstrual cycles and
no use of oral contraceptives who ran both at day 10 (follicular phase, F) and at
day 25 (luteal phase, L) of their cycle. 12 male subjects (M) served as controls.
The mRNA was pooled group wise and processed on a gene expression microarray encompassing 789 genes, including major genes of the inflammatory and
anti-inflammatory reaction. The differences of gene expression between time
points t0 (before run) and t1 (after run) were analyzed. Females in L showed a higher extent of regulation than females in F or men. Among those genes which were
up-regulated above 1.5 fold change (log2) pro-inflammatory genes were significantly enriched (p=0.033, after Bonferroni correction) in L, while this was not the
case in F or M. Conversely, women in L showed a strong trend towards downregulation of anti-inflammatory genes. Some prominent genes like IL6 (coding for
interleukin-6), and IL1RN (also termed IL1RA, coding for interleukin-1 receptor antagonist) were clearly regulated in opposite directions in L as opposed to F and M.
In conclusion, women in L showed a distinctly different pattern of gene regulation
in response to exercise, compared with women in F or M. The overall direction of
Address correspondence to
Prof. Dr. Hinnak Northoff, Head, Institute for Clinical and Experimental Transfusion
Medicine(IKET), University Tübingen, Med. Dir., Zentrum f. Klinische Transfusionsmedizin Tübingen gGmbH, Otfried-Mueller-Strasse 4/1, 72076 Tuebingen, Germany
Tel.: +49-(0)7071-29- 81601, Fax.: +49-(0)7071-29- 5240
mailto: hinnak.northoff@med.uni-tuebingen.de, https://www.blutspendezentrale.de
Gender specific gene response to exercise • 87
gene expression changes of women in L is clearly pro-inflammatory. This finding
accentuates a need for careful consideration of the female cyclic phase when
investigating women in exercise immunology studies. Our results may also have
implications relevant to other forms of stress in females.
Keywords: gender, inflammation, gene regulation, aerobic exercise, menstrual
cycle, stress response, IL6, IL1RN, IL1RA.
INTRODUCTION
Recent studies have documented that significant gender dimorphisms exist in certain immune responses to different types of exercise (6, 15, 27-29). Gender differences in response to exercise have clear implications for understanding genderspecific adaptations to exercise for athletic performance and overall health. However, while in general the impact of exercise on immune functions has received
considerable and increasing attention in recent years, it is still unclear to what
extent gender and fluctuations in sex hormones influence immunological responses to exercise.
Several gender-related differences in immune function under non exercise conditions have been identified, and it has been hypothesized that at least some of these
differences could be attributed to female sex hormones (7). Numerous clinical
studies have demonstrated that immune responsiveness is greater in women than
in men (7): women have lower incidence and mortality to several types of infections (7), higher serum concentrations of some immunoglobulins (IgM) (12), a
higher absolute number of T-helper lymphocytes (1), and a differential regulation
of cytokine production (12, 14). Leukocyte chemotaxis (7) is also sensitive to
gender related hormones. Mitochondria from females generate smaller amounts
of hydrogen peroxide than those of males and have higher levels of mitochondrial
reduced glutathione and antioxidant enzymes (26). Several menstrual cycle associated effects on parameters of the immune system have been described. Compared to the follicular phase (F), the luteal phase (L) of the menstrual cycle was
associated with increased concentrations of leukocytes and lymphocyte subsets
(5, 9), increased prostaglandin (PG) E2 and PGI2 release by stimulated monocytes (3, 11, 25), a greater capacity of immune cells to produce cytokines (5, 9,
13), a higher plasma cytokine activity (14), but variable effects on plasma
cytokine levels (2, 8, 13). In contrast, other studies associate the follicular phase
with greater cytokine production from immune cells (14) and higher serum IL-6
levels (2).
The fact that the majority of exercise studies has been done in males does not
really come as a surprise. However, in situations where a new hypothesis has to be
proven or disproven for the first time, it may be a forgivable or even a wise concept to start off with males only to avoid unforeseeable interferences from fluctuations of sex hormones occurring in women depending on the different phases of
their menstrual cycle. Even worse than that, we know that in competitively training female athletes the cycle is often disturbed or abolished. In addition many
females take oral contraceptives which again can have an impact on immunological functions as well (27). Thus, it can be tedious and not very easy to find well
88 • Gender specific gene response to exercise
defined and willing groups of female volunteers to do meaningful studies. Nevertheless we think that time has come to do exactly that.
A number of studies have reported no differences in cell counts and functions (4,
16-19, 31), plasma cytokine levels (16, 30), and lymphocyte apoptosis (20)
between men and women concerning the response to different kinds of exercise.
However, it appears that these studies did not control for the menstrual status of
the women at the time of testing. In contrast to studies reporting no differences,
others have reported gender differences in the immune-related responses to treadmill running (5), cycling (8, 27-29) and eccentric exercise (15, 25). In a recent
study (Fehrenbach et al. unpublished), we found out that intracellular HSP70
showed gender and menstrual cycle dependent reactions in lymphocytes and
monocytes 24 h after exercise. Timmons et al. (2005) have reported gender and
menstrual cycle dependent changes in leukocyte and cytokine responses to
cycling (27).
In the present study we used mRNA from the above mentioned HSP study to run
a microarray analysis on 789 genes, which were partly selected on the basis of
their relation to inflammatory processes. The study had a group of regularly menstruating women who ran twice, once on day 10 (follicular phase) and once on
day 25 (luteal phase) of their menstrual cycle and a group of males for comparison.
The first results of this investigation focusing on the differences in gene expression immediately after compared with before a 1 h run close to the individual
anaerobic threshold are presented here.
MATERIAL AND METHODS
Subjects
Twelve female (W) and 12 male runners (M) gave informed consent to participate
in the study. The investigation was approved by the University Ethics Committee.
All were experienced athletes with normal dietary habits. They were not on any
medication and they performed endurance training on a regular basis. The W
included in the study had regular menstrual cycles and did not use oral contraception. Determination of the cyclic phases was based on a diary, kept by the women,
beginning three months prior to the study. To confirm the cyclic phases, the hor-
Table 1: Anthropometric and physical characteristics of the subjects.
Men (n=12)
Women (n=9)
Age (yrs)
32.6 (28.7 – 36.4)
29.68 (25.4 – 33.7)
Body mass index (kg· m2)
21.6 (20.9 – 22.3)
20.9 (19.9 – 22.0)
Training sessions (1 · week-1)
5.8 (5.3 – 6.2)
4.4 (3.8 – 5.1)*
Training distance (km · week-1)
60.8 (53.9 – 67.7)
38.9 (28.6 – 49.2)*
VIAT (km · h-1)
14.0 (13.4 – 14.5)
11.8 (11.1 – 12.5)*
VIAT, running velocity at the individual anaerobic threshold. Data are presented as means (95% CI).
*p<0.01, men vs. women
Gender specific gene response to exercise • 89
monal status of W was determined by measuring oestrogen, progesterone and LH
using the ADVIA Centaur immunoassay system (Siemens Healthcare Diagnostics, Fernwald, Germany). After hormonal assessment, three women had to be
withdrawn from the study due to luteal insufficiency. The physical characteristics
of the remaining athletes are shown in Table 1.
Preliminary Testing
Before participating in the main study the athletes performed an incremental exercise test on a treadmill (Saturn, HP Cosmos, Traunstein, Germany) to determine
the running velocity (VIAT) at the individual anaerobic threshold (IAT). Capillary
blood for lactate measurement (EBIO, Eppendorf, Hamburg, Germany) was
obtained from the earlobe after every stage and heart rate was monitored continuously using a heart rate monitor (Polar Electro, Finland). VIAT was calculated by
the method of Dickhuth (23) using a PC-routine.
Continuous runs
The main investigation consisted of continuous runs (CR) on the treadmill with
duration of 60 min and a running velocity corresponding to 93% VIAT. The exercise procedure started at 09:00 a.m. The W had to perform the identical CR twice:
once in the follicular phase (F) of their cycle at day 10 and once in the luteal
phase (L) of their cycle at day 25. Capillary blood lactate was determined before
and immediately after exercise. Venous blood samples were drawn one hour
before (t0; 8:00 a.m.) and immediately after the end of the CR (t1; 10:00 a.m.).
PBMC isolation and RNA extraction
EDTA anti-coagulated venous blood samples were used for the isolation of
peripheral blood mononuclear cells (PBMC) using the Ficoll-hypaque density
gradient technique as described previously (10). After gathering the cells in RLTbuffer total RNA was extracted using an RNeasy minikit (Qiagen, Hilden Germany) in accordance with the manufacturer’s protocol. The RNA from M (n=12)
and W (L/F; n=9) was pooled using equal amounts of RNA for the corresponding
runs for t0 and t1. The integrity of extracted RNA was assessed using an Agilent
2100 Bioanalyzer (Agilent Technologies, Palo Alto, California, USA).
Microarray data generation and statistical analysis
Microarray data were generated using 65mer oligonucleotide microarrays produced at the IKET, University of Tübingen as previously described (33). We used
a 2,402 feature array including transcripts as well as buffers, controls and empty
spots. The genes on the array were selected inter alia with a focus on inflammation and regulation of inflammatory processes. Every feature was printed at least
in duplicate. The array contained 789 genes in total, while some transcripts were
contained up to 12 times in duplicate. For further details of the array used in this
study can be obtained from the National Center for Biotechnology Information
(http://www.ncbi.nlm.nih.gov/geo/) under accession number GPL5676.
An indirect reference design was used with Cy3 labeled uniRNA (Stratagene, La
Jolla, California, USA) and Cy5 labeled sample RNA. Amplification of sample
RNA was performed using Ambion´s Amino Allyl Message Amp II aRNA Amplification Kit (Ambion Inc., Austin, Texas, USA) together with Amersham CyDye
90 • Gender specific gene response to exercise
Post-labeling Reactive Pack (GE Healthcare, Buckinghamshire, UK) following
the manufacturer`s protocols, and assessing dye incorporation using a Nano Drop
ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, Delaware,
USA). After an aRNA fragmentation using Ambion´s Fragmentation reagents
(Ambion Inc., Austin, Texas, USA) hybridization was performed for 14 h at 48°C.
Subsequently, the hybridized and washed slides were scanned in a microarray
scanner (Affymetrix Inc. Santa Clara, California, USA). The photomultiplier tube
voltage was set to 100% for both green and red channels. The resulting green and
red images were overlaid using ImaGene 5.0 (Biodiscovery Inc. El Segundo, California, USA) as well as for raw data collection.
Data analysis
Further statistical and bioinformatic analysis was performed using the limma
(Linear models for microarray) package for R from the bioconductor project (24).
The data was normalized using printtip-loess intra-array normalization on the
normexp-background corrected expression values followed by inter-array quantile normalization across groups. For further analysis, normalized expression values of duplicate features were averaged. For the different pools (F, L, and M) the
fold change (fc=t1-t0) between log2 expression values of both time points was
computed. On the basis of the fold changes, up-regulated genes (fc > 1.5) and
down-regulated genes (fc < -1.5) were determined. 81 different genes from the
array with clearly pro-inflammatory impact and 43 different genes with clearly
anti-inflammatory impact were selected for a closer analysis (see addendum). For
both pro-inflammatory and anti-inflammatory gene sets and each group, the number of genes exceeding the respective fold change thresholds between t0 and t1
was calculated. For significance testing, the same number of genes contained in
the respective set was sampled 10,000 times and the fraction of genes exceeding
the threshold p value was calculated. A gene set with p < 0.05 was considered significantly enriched. Tests were omitted if no genes of the set exceeded the threshold. The result of the analysis of the above mentioned gene sets encompassing the
pro- inflammatory and anti-inflammatory genes are listed in the addendum. We
are aware that due to pooling the RNA, no classical significance testing could be
performed. To control the false-positive rate, we used rather conservative thresholds, requiring absolute fold changes of at least 1.5 (log2) for genes to be considered significantly regulated.
The raw microarray data is available in GEO (http://www.ncbi.nlm.nih.gov/geo/).
Table 2: Resting hormone levels in women and pre- and post-exercise lactate concentrations.
Men (n=12)
Women, F (n=9)
Women, L (n=9)
Estrogene (pmol · l-1)
Pre-CR
n.d.
370 (111 – 629)
491 (296 – 687)
23.8 (15.0 – 32.6)
Progesterone (nmol · l-1)
Pre-CR
n.d.
4.0 (1.9 – 6.1)*
Blood lactate (mmol· l-1)
Pre-CR
0.9 (0.7 – 1.2)
1.0 (0.9 – 1.1)
0.9 (0.7 – 1.0)
Blood lactate (mmol· l-1)
Post-CR
2.1 (1.5 – 2.6)+
2.4 (1.6 – 3.3)+
2.6 (2.1 – 3.0)+
Data are presented as means (95% CI). F, follicular phase; L, luteal phase; n.d., not detected. *p<0.01,
women, F vs. women, L; + p<0.01, post-CR vs. pre-CR. There were no significant differences between F,
L and M.
Gender specific gene response to exercise • 91
RESULTS
The treadmill runs were performed at a speed which corresponded to 93% VIAT.
At the end of exercise, blood lactate concentrations were significantly increased
in all groups, but still remained in a range typical for more intensive but still predominantly aerobic exercise. No significant differences were detected between
M, F and L (see table 2).
Statistical analysis
The enrichment analysis yielded one enriched gene set. In group L, we found the
pro-inflammatory genes enriched among the up-regulated genes (p= 0.0017, after
Bonferroni-correction for 10 tests: 0.017).
In general, L showed a high degree of regulation having 129 genes up–regulated
and 143 down-regulated, compared with F (48 / 32) and M (34 / 29). This was
especially pronounced in the gene sets specifically selected for their strong relation to inflammation. From the 81 genes judged as pro-inflammatory, 20 stood
out to be regulated above the mentioned threshold of 1.5 (log2). Of these, 13 were
up-regulated and 7 were down-regulated. 17 of the anti-inflammatory genes were
regulated above the threshold, of which 6 were up-regulated and 11 down-regulated. In M and F, much lower regulation was observed (see figure 1).
Figure 1: Major changes in expression of anti- (white) and pro (black)- inflammatory
genes (see addendum) in the three groups. Bars pointing upwards denote up-regulated
genes; bars pointing downwards denote down-regulated genes. A threshold of +/- 1.5 (log2)
was used (see addendum).
92 • Gender specific gene response to exercise
Figure 2: Box plots of log2 fold change for the selected gene lists, separately for each
group. Genes of special interest were marked at their respective positions in the respective
boxplot, + denotes outliers (below or above +- 1.5* interquartile range) not considered in
this context. For the marked pro- and anti-inflammatory genes, we observe a strong inverse
regulation. Note that the variances for both gene sets differ significantly between L and M
or F (F-test p value < 10^-9).
By arbitrarily setting another cutoff at log2 1.0 (up-regulated (fc > 1.0) or downregulated (fc < -1.0)) in either direction, 35 genes from the pro- inflammatory and
25 genes from the anti-inflammatory subset came up in L. Little changes were
detected in M (9 / 4) and F (9 / 6).
When aligning the detected genes, according to their pro-inflammatory impact on
the one hand (pro-inflammatory genes up-regulated/ anti-inflammatory genes
down-regulated) and to their anti-inflammatory impact on the other hand (proinflammatory genes down-regulated / anti-inflammatory genes up-regulated), a
strong pro-inflammatory response was revealed in L (see figure 1). Neither in F
nor in M was a comparable regulation observed.
For some genes of either set, a strong inverse regulation was detected. This was
especially pronounced for the anti-inflammatory genes IL6, the decoy receptor
interleukin 1 receptor type II (IL1R2) and IL1RN, which were up-regulated during exercise in F, while consistently down regulated in L (see figures 2 and 3).
IL1RN codes for the IL-1 receptor antagonist. In the literature the expression
IL1RA is used synonymously for the gene. Furthermore, we found several proinflammatory genes, including prostaglandin D2 receptor (PTGDR), interleukin
18 receptor accessory protein (IL18RAP) and interleukin-12 receptor beta 1
(IL12RB1) to be down-regulated in F, while strongly up-regulated in L (see figure
Gender specific gene response to exercise • 93
2). Some of the remaining genes of both sets exhibited a similar pattern of regulation. A comparable inverse regulation, into the opposite direction (pro- inflammatory impact in F, and anti-inflammatory impact in L) was exhibited by only one
anti-inflammatory gene, adrenergic receptor beta 2 (ADRB2) which was downregulated in F but up regulated in L ( for further information see addendum).
Figure 3: Profile plots for selected pro-inflammatory genes (upper row) and anti-inflammatory genes. The plots show expression values for t0 and t1 for each group. The abscissa
shows the expression value.
DISCUSSION
Among mammals, very few things are regulated with such a high species-specificity
as reproduction. Obviously there is enough flexibility built into this area of physiology to enable each species to adjust optimally to its needs. Conception susceptibility
of females decides if newborns arrive all together in spring (typical for favored victims of predators) or several times during the year (like in dogs) or every few weeks
(rodents). Human females are disposed to essentially all year long readiness for sexual activity with frequent and regular periods of conception susceptibility.
The situation as described makes animal experiments very tricky to translate to
the human situation. Nevertheless, the findings of Nickerson et al. (21), that
female rodents did not show elevated myocardial heat shock proteins (HSP) after
exercise stress, while males did, prompted us to run a study designed to explore
the reaction of HSP to exercise in controlled relation to the female menstrual
cycle. To our surprise, females showed strikingly different patterns of regulation,
depending on the phase of their menstrual cycle. While at d10 (F), they regulated
94 • Gender specific gene response to exercise
HSP upwards (like males), at d25 (L) they regulated downwards (unpublished
data). The observation, that the human females seem to take out an important cell
protective system during L in reaction to stress induced us to run a gene expression chip analysis focused on genes anyhow related to inflammation or protective
anti-inflammatory regulation.
In essence we found an impressive coordinated movement of genes in the direction of a pro-inflammatory impact. It is intriguing – and also reassuring -- that this
movement was a combined action of pro-inflammatory genes being up-regulated
and anti-inflammatory genes being down-regulated. Although only the proinflammatory up-regulation was significant, the down-regulation showed at least
a very strong trend and importantly encompassed some key markers which we
know from numerous studies as reactive to exercise. Central markers of the protective regulations following exercise like IL6, IL1RN (coding for interleukin
receptor 1 antagonist, see addendum) and IL1R2 were significantly down-regulated in L, while they were significantly or borderline significantly up-regulated in F
(see figure 2). HSPB (coding for HSP 27), a central gene in the HSP system followed essentially the same pattern. Likewise, important pro-inflammatory genes
like PTGDR, IL18RAP, arachidonate 5-lipoxygenase (ALOX5) or IL12 (see
addendum) were highly significant up-regulated in L, while they were down-regulated in F. Concerning ALOX5, a gender specific secretion pattern of
leukotrienes, governed by androgens, via regulation of extracellular signal related
kinases (ERKs) has recently been found (22).
The overall number of genes which were significantly regulated following the
exercise challenge underlines the exceptional state of the organism in the luteal
phase with females regulating 200+ genes in L while in F only about 70 genes
were regulated, similar to the number in males (60).
The question regarding what is behind these striking cycle dependent differences
is not easy to answer. It seems safe to say, that, immediately after one hour of
exercise, (t0-t1) there is a substantial change in gene expression in the direction of
an increased pro-inflammatory state in women in the luteal phase. It is also highly
likely, that this has to do with reproductive function of women. In the uterine
endometrium of adult women a steady increase in the expression of important
pro-inflammatory cytokines has already been shown starting in the mid luteal
phase and continuing up to the very late luteal phase (32). However, this situation
might be different in PBMCs. What we do not know is:
(a) Whether the observed effect is the same at other time points of the luteal phase
or whether it is specific for the last few days of the cycle;
(b) Whether the regulation on the mRNA level is accompanied by coordinated
translation into the corresponding proteins.
Concerning (a), further analysis of different time points of the cycle should show
if the observed phenomenon is characteristic throughout the luteal phase. If not,
the observed reaction could rather be understood as something that is related to
the initiation of menstruation.
Gender specific gene response to exercise • 95
Concerning (b), further studies have to be done to find out to what extent the
observed gene expression changes are accompanied by corresponding changes in
protein expression. Analysis of serum proteins will be necessary and helpful, but
not necessarily sufficient to clarify this point. Fast clearance by the kidney or
degradation is likely to occur and might blur the picture. Experiments measuring
intracellular, membrane, or ex vivo released proteins will probably be necessary.
There were some indications that a part of the pro-inflammatory genes which
were up-regulated in L had quite a low level of expression at rest. Vice versa, part
of the anti-inflammatory genes which were down-regulated in L, came from quite
high levels of expression at rest. It is therefore possible that the gene expression
changes seen in reaction to exercise in L may constitute a fast return to normal
from a highly anti-inflammatory state at rest, rather than a truly pro-inflammatory
response. Substantially more analysis, including generation of protein data will
have to be done to clarify this point. Both possibilities, may, however, make
sense.
On the one hand, the organism in L which is prepared for a pregnancy may need a
highly anti-inflammatory / immunosuppressive state in order to tolerate the fertilized egg, which, from the standpoint of immunology, is a foreign intruder. A
major external stressor like physical exercise might then induce a quick return of
this cycle specific expression pattern back to a normal pattern to be prepared for
fending off an infection. But even if the observed change of gene expression constitutes a really pro-inflammatory impulse, a second signal (e.g. danger signals)
might be necessary to provoke a prolonged inflammatory reaction.
The biological significance of the observed gene expression change can thus not
be clearly judged at present. Of course it seems possible that the inflammatory
impulse created by substantial exercise is sufficient to induce parturition of an
incumbent early pregnancy. Lynch et al. (14) showed in an elegant study that men
and women regulate the IL1/ IL1RN system in a completely different way, with
women showing differential regulation in F and L. These authors showed that ex
vivo monocytes from women secrete high amounts of IL1 and its antagonist
IL1RN in balanced amounts during F, so that no bioactivity results, while in L
there is a deficit of the antagonist, resulting in bioactivity in the supernatants.
They link this finding to the role of IL1 in parturition and during birth. In the light
of these experiments, it seems plausible that the pro-inflammatory response of
women in L may constitute a mechanism designed to end a very early pregnancy
in case of major external stress input. After all, human females get a new chance
to conceive in the next month and nature may prefer to destabilize a pregnancy
under influence of stress rather than carry it on under high risk.
In conclusion, women in their luteal phase showed a distinctly different pattern of
gene regulation in response to exercise, compared with women in their follicular
phase or men. This finding accentuates a need for careful consideration of the
female cyclic phase when investigating the stress response to exercise in women.
Our results may also have implications relevant to other forms of stress in
females.
96 • Gender specific gene response to exercise
ACKNOWLEDGMENTS
We thank the volunteers for participating in this study.
In memoriam Elvira Fehrenbach †
REFERENCES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Amadori,A., Zamarchi,R., De,S.G., Forza,G., Cavatton,G., Danieli,G.A., Clementi,M. and Chieco-Bianchi,L. Genetic control of the CD4/CD8 T-cell ratio in humans,
Nat.Med., 1: 1279-1283, 1995.
Angstwurm,M.W., Gartner,R. and Ziegler-Heitbrock,H.W. Cyclic plasma IL-6 levels
during normal menstrual cycle, Cytokine, 9: 370-374, 1997.
Arend,W.P., Smith,M.F., Jr., Janson,R.W. and Joslin,F.G. IL-1 receptor antagonist
and IL-1 beta production in human monocytes are regulated differently, J.Immunol.,
147: 1530-1536, 1991.
Barriga,C., Pedrera,M.I., Maynar,M., Maynar,J. and Ortega,E. Effect of submaximal
physical exercise performed by sedentary men and women on some parameters of
the immune system, Rev.Esp.Fisiol., 49: 79-85, 1993.
Bouman,A., Moes,H., Heineman,M.J., de Leij,L.F. and Faas,M.M. The immune
response during the luteal phase of the ovarian cycle: increasing sensitivity of human
monocytes to endotoxin, Fertil.Steril., 76: 555-559, 2001.
Brown,A.S., Davis,J.M., Murphy,E.A., Carmichael,M.D., Carson,J.A., Ghaffar,A.
and Mayer,E.P. Gender differences in macrophage antiviral function following exercise stress, Med.Sci.Sports Exerc., 38: 859-863, 2006.
Cannon,J.G. and St Pierre,B.A. Gender differences in host defense mechanisms,
J.Psychiatr.Res., 31: 99-113, 1997.
Chiu,K.M., Arnaud,C.D., Ju,J., Mayes,D., Bacchetti,P., Weitz,S. and Keller,E.T.
Correlation of estradiol, parathyroid hormone, interleukin-6, and soluble interleukin6 receptor during the normal menstrual cycle, Bone, 26: 79-85, 2000.
Faas,M., Bouman,A., Moesa,H., Heineman,M.J., de,L.L. and Schuiling,G. The
immune response during the luteal phase of the ovarian cycle: a Th2-type response?,
Fertil.Steril., 74: 1008-1013, 2000.
Fehrenbach,E., Niess,A.M., Schlotz,E., Passek,F., Dickhuth,H.H. and Northoff,H.
Transcriptional and translational regulation of heat shock proteins in leukocytes of
endurance runners, J.Appl.Physiol, 89: 704-710, 2000.
Flynn,A. Stimulation of interleukin-1 production from placental monocytes, Lymphokine Res., 3: 1-5, 1984.
Giron-Gonzalez,J.A., Moral,F.J., Elvira,J., Garcia-Gil,D., Guerrero,F., Gavilan,I.
and Escobar,L. Consistent production of a higher TH1:TH2 cytokine ratio by stimulated T cells in men compared with women, Eur.J.Endocrinol., 143: 31-36, 2000.
Konecna,L., Yan,M.S., Miller,L.E., Scholmerich,J., Falk,W. and Straub,R.H. Modulation of IL-6 production during the menstrual cycle in vivo and in vitro, Brain
Behav.Immun., 14: 49-61, 2000.
Lynch,E.A., Dinarello,C.A. and Cannon,J.G. Gender differences in IL-1 alpha, IL-1
beta, and IL-1 receptor antagonist secretion from mononuclear cells and urinary
excretion, J.Immunol., 153: 300-306, 1994.
Gender specific gene response to exercise • 97
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
MacIntyre,D.L., Reid,W.D., Lyster,D.M. and McKenzie,D.C. Different effects of
strenuous eccentric exercise on the accumulation of neutrophils in muscle in women
and men, Eur.J.Appl.Physiol, 81: 47-53, 2000.
Meksawan,K., Venkatraman,J.T., Awad,A.B. and Pendergast,D.R. Effect of dietary
fat intake and exercise on inflammatory mediators of the immune system in sedentary men and women, J.Am.Coll.Nutr., 23: 331-340, 2004.
Moyna,N.M., Acker,G.R., Fulton,J.R., Weber,K., Goss,F.L., Robertson,R.J., Tollerud,D.J.
and Rabin,B.S. Lymphocyte function and cytokine production during incremental exercise in active and sedentary males and females, Int.J.Sports Med., 17: 585-591, 1996.
Moyna,N.M., Acker,G.R., Weber,K.M., Fulton,J.R., Goss,F.L., Robertson,R.J. and
Rabin,B.S. The effects of incremental submaximal exercise on circulating leukocytes in physically active and sedentary males and females, Eur.J.Appl.Physiol
Occup.Physiol, 74: 211-218, 1996.
Moyna,N.M., Acker,G.R., Weber,K.M., Fulton,J.R., Robertson,R.J., Goss,F.L. and
Rabin,B.S. Exercise-induced alterations in natural killer cell number and function,
Eur.J.Appl.Physiol Occup.Physiol, 74: 227-233, 1996.
Navalta,J.W., Sedlock,D.A., Park,K.S. and McFarlin,B.K. Neither gender nor menstrual cycle phase influences exercise-induced lymphocyte apoptosis in untrained
subjects, Appl.Physiol Nutr.Metab, 32: 481-486, 2007.
Nickerson,M., Kennedy,S.L., Johnson,J.D. and Fleshner,M. Sexual dimorphism of
the intracellular heat shock protein 72 response, J.Appl.Physiol, 101: 566-575, 2006.
Pergola,C., Dodt,G., Rossi,A., Neunhoeffer,E., Lawrenz,B., Northoff,H., Samuelsson,B., Rådmark,O., Sautebin,L. and Werz,O. ERK-mediated regulation of
leukotriene biosynthesis by androgens: A molecular basis for gender differences in
inflammation and asthma, Proc.Natl.Acad.Sci.U.S.A, 2008. In press.
Roecker,K., Schotte,O., Niess,A.M., Horstmann,T. and Dickhuth,H.H. Predicting
competition performance in long-distance running by means of a treadmill test,
Med.Sci.Sports Exerc., 30: 1552-1557, 1998.
Smyth,GK. Bioinformatics and Computational Biology Solutions using R and Bioconductor, Springer: New York, 2005.
Stupka,N., Lowther,S., Chorneyko,K., Bourgeois,J.M., Hogben,C. and Tarnopolsky,M.A. Gender differences in muscle inflammation after eccentric exercise,
J.Appl.Physiol, 89: 2325-2332, 2000.
Sureda,A., Ferrer,M.D., Tauler,P., Tur,J.A. and Pons,A. Lymphocyte antioxidant
response and H2O2 production after a swimming session: gender differences, Free
Radic.Res., 42: 312-319, 2008.
Timmons,B.W., Hamadeh,M.J., Devries,M.C. and Tarnopolsky,M.A. Influence of
gender, menstrual phase, and oral contraceptive use on immunological changes in
response to prolonged cycling, J.Appl.Physiol, 99: 979-985, 2005.
Timmons,B.W., Tarnopolsky,M.A. and Bar-Or,O. Sex-based effects on the distribution of NK cell subsets in response to exercise and carbohydrate intake in adolescents, J.Appl.Physiol, 100: 1513-1519, 2006.
Timmons,B.W., Tarnopolsky,M.A., Snider,D.P. and Bar-Or,O. Immunological
changes in response to exercise: influence of age, puberty, and gender,
Med.Sci.Sports Exerc., 38: 293-304, 2006.
Venkatraman,J.T. and Pendergast,D. Effects of the level of dietary fat intake and
endurance exercise on plasma cytokines in runners, Med.Sci.Sports Exerc., 30:
1198-1204, 1998.
98 • Gender specific gene response to exercise
31
32
33
Venkatraman,J.T., Rowland,J.A., Denardin,E., Horvath,P.J. and Pendergast,D. Influence of the level of dietary lipid intake and maximal exercise on the immune status
in runners, Med.Sci.Sports Exerc., 29: 333-344, 1997.
von,W.M., Thaler,C.J., Strowitzki,T., Broome,J., Stolz,W. and Tabibzadeh,S. Regulated expression of cytokines in human endometrium throughout the menstrual
cycle: dysregulation in habitual abortion, Mol.Hum.Reprod., 6: 627-634, 2000.
Zieker,D., Konigsrainer,I., Traub,F., Nieselt,K., Knapp,B., Schillinger,C., Stirnkorb,C., Fend,F., Northoff,H., Kupka,S., Brucher,B.L. and Konigsrainer,A. PGK1 a
potential marker for peritoneal dissemination in gastric cancer, Cell Physiol
Biochem., 21: 429-436, 2008.
Gender specific gene response to exercise, Addendum • 99
Addendum
Anti-inflammatory genes
↑: Up regulated, fc > 1.5
↓: Down regulated, fc < -1.5
Gene
Accession Id
Description
ADRB2
NM_000024
adrenergic, beta-2-, receptor, surface
ADRB2
NM_000024
adrenergic, beta-2-, receptor, surface
ADRB2
NM_000024
adrenergic, beta-2-, receptor, surface
ADRB2
NM_000024
adrenergic, beta-2-, receptor, surface
ADRBK2
NM_005160
adrenergic, beta, receptor kinase 2
AHSA1
NM_012111
AHA1, activator of heat shock 90kDa protein
M
F
L
↑
↓
↑
ATPase homolog 1 (yeast)
CD163
NM_203416
CD163 molecule
CD19
NM_001770
CD19 molecule
↑
CD33
NM_001772
CD33 molecule
CSF3R
NM_172313
colony stimulating factor 3 receptor (granulocyte)
CSF3R
M59820.1
Human granulocyte colony-stimulating factor receptor
↑
CYC1
NM_001916
cytochrome c-1
↓
GPX1
NM_201397
glutathione peroxidase 1
GPX3
NM_002084
glutathione peroxidase 3 (plasma)
GPX4
NM_002085
glutathione peroxidase 4 (phospholipid hydroperoxidase)
GSS
NM_000178
glutathione synthetase
GSTM3
NM_000849
glutathione S-transferase M3 (brain)
↓
GSTP1
NM_000852
glutathione S-transferase pi 1
HSPB1
NM_001540
heat shock 27kDa protein 1
↓
HSPB1
NM_001540
heat shock 27kDa protein 1
↓
HSPB1
NM_001540
heat shock 27kDa protein 1
↓
IL10RB
NM_000628
interleukin 10 receptor, beta
IL13
NM_002188
interleukin 13
IL13RA2
NM_000640
interleukin 13 receptor, alpha 2
IL16
NM_172217
interleukin 16 (lymphocyte chemoattractant factor)
IL1R2
NM_173343
interleukin 1 receptor, type II
IL1RN
NM_173843
interleukin 1 receptor antagonist
IL2RB
NM_000878
interleukin 2 receptor, beta
↑
↓
↓
100 • Gender specific gene response to exercise, Addendum
Gene
Accession Id
IL4R
NM_001008699 interleukin 4 receptor
Description
IL6
NM_000600
interleukin 6 (interferon, beta 2)
IL6R
NM_181359
interleukin 6 receptor
IL6ST
NM_175767
interleukin 6 signal transducer (gp130,
LILRA2
NM_006866
leukocyte immunoglobulin-like receptor, subfamily A,
MT3
NM_005954
metallothionein 3
PPARA
BC000052.2
peroxisome proliferator-activated receptor alpha, mRNA
PPARA
NM_005036
peroxisome proliferator-activated receptor alpha
M
F
L
↑
↓
oncostatin M receptor)
member 2
↑
PPARG
BC006811.1
peroxisome proliferator-activated receptor gamma
PRDX4
NM_006406
peroxiredoxin 4
↓
PRDX5
NM_181652
peroxiredoxin 5
↓
PROC
NM_000312
protein C (inactivator of coagulation factors Va and VIIIa)
PROK2
NM_021935
prokineticin 2
PTGIS
NM_000961
prostaglandin I2 (prostacyclin) synthase
SOD1
NM_000454
superoxide dismutase 1, soluble
SOD2
AY267901
superoxide dismutase 2, nuclear gene for mitochondrial
product.
SOD3
NM_003102
superoxide dismutase 3, extracellular
STIP1
NM_006819
stress-induced-phosphoprotein 1
THBD
NM_000361
thrombomodulin
TXN
NM_003329
thioredoxin
TXN2
NM_012473
thioredoxin 2
TXNIP
NM_006472
thioredoxin interacting protein
↑
↓
Gender specific gene response to exercise, Addendum • 101
Proinflammatory genes
↑: Up regulated, fc > 1.5
↓: Down regulated, fc < -1.5
Gene
Accession Id
Description
ALOX5
NM_000698
arachidonate 5-lipoxygenase
ALOX5
NM_000698
arachidonate 5-lipoxygenase
CASP1
NM_033295
caspase 1 (interleukin 1, beta, convertase)
CASP1
NM_033292
caspase 1, transcript variant alpha
CASP1
NM_033294
caspase 1, transcript variant delta
CASP3
NM_032991
caspase 3 transcript variant beta
CASP3
NM_032991
caspase 3
CASP5
NM_004347
caspase 5
CASP5
NM_004347
caspase 5
CASP9
NM_001229
caspase 9 transcript variant alpha
CASP9
NM_032996
caspase 9, apoptosis-related cysteine peptidase
CCL4
NM_002984
chemokine (C-C motif) ligand 4
CCR1
NM_001295
chemokine (C-C motif) receptor 1
CD14
NM_000591
CD14 molecule
CD160
BC014465.1
CD160 molecule
CD1B
NM_001764
CD1b molecule
CD1B
NM_001764
CD1b molecule
CD2
NM_001767
CD2 molecule
CD44
NM_001001392 CD44 molecule (Indian blood group)
CD58
NM_001779
CD58 molecule
CD59
NM_203331
CD59 molecule, complement regulatory protein
CD69
NM_001781
CD69 molecule
CD80
NM_005191
CD80 molecule
CD83
NM_004233
CD83 molecule
COX7A2
BC100852.1
cytochrome c oxidase subunit VIIa polypeptide 2 (liver)
CSF1
NM_172212
colony stimulating factor 1 (macrophage)
CSF2
NM_000758
colony stimulating factor 2 (granulocyte-macrophage)
CX3CR1
NM_001337
chemokine (C-X3-C motif) receptor 1
M
F
L
↑
↑
↑
↑
↓
↑
↓
CXCL10
NM_001565
chemokine (C-X-C motif) ligand 10
CYSLTR1
NM_006639
cysteinyl leukotriene receptor 1
↑
DAP
NM_004394
death-associated protein
↓
102 • Gender specific gene response to exercise, Addendum
Gene
Accession Id
Description
DAPK1
NM_004938
death-associated protein kinase 1
FCGR3B
NM_000570
Fc fragment of IgG, low affinity IIIb, receptor (CD16b)
HIF1AN
NM_017902
hypoxia-inducible factor 1, alpha subunit inhibitor
HLA-DRA
NM_019111
major histocompatibility complex, class II, DR alpha
ICAM2
NM_000873
intercellular adhesion molecule 2
ICAM3
NM_002162
intercellular adhesion molecule 3
ID2
NM_002166
inhibitor of DNA binding 2, dominant negative
M
F
L
↑
↓
helix-loop-helix protein
IFNAR1
NM_000629
interferon (alpha, beta and omega) receptor 1
IFNG
NM_000619
interferon, gamma
IFNG
NM_000619
interferon, gamma
IFNG
NM_000619
interferon, gamma
IFNGR1
NM_000416
interferon gamma receptor 1
IGF2
NM_000612
insulin-like growth factor 2 (somatomedin A)
IGF2
NM_000612
insulin-like growth factor 2 (somatomedin A)
IGF2
NM_000612
insulin-like growth factor 2 (somatomedin A)
↓
IHPK3
NM_054111
inositol hexaphosphate kinase 3
↓
IL11
NM_000641
interleukin 11
IL12RB1
NM_153701
interleukin 12 receptor, beta 1
IL12RB2
NM_001559
interleukin 12 receptor, beta 2
IL15
NM_172174
interleukin 15
IL18
NM_001562
interleukin 18 (interferon-gamma-inducing factor)
IL18R1
NM_003855
interleukin 18 receptor 1
IL18RAP
BC106765.2
Homo sapiens interleukin 18 receptor accessory protein
IL1A
NM_000575
interleukin 1, alpha
IL1A
NM_000575
interleukin 1, alpha
IL1A
NM_000575
interleukin 1, alpha
IL1B
NM_000576
interleukin 1, beta
IL1R1
NM_000877
interleukin 1 receptor, type I
IL21R
NM_181079
interleukin 21 receptor
IL24
NM_181339
interleukin 24
IL5RA
NM_175728
interleukin 5 receptor, alpha
IL8RA
NM_000634
interleukin 8 receptor, alpha
IL8RA
NM_000634
interleukin 8 receptor, alpha
INDO
NM_002164
indoleamine-pyrrole 2,3 dioxygenase
IRAK1
NM_001569
interleukin-1 receptor-associated kinase 1
LBP
NM_004139
lipopolysaccharide binding protein
↓
↓
↑
↓
↓
↑
↑
↓
Gender specific gene response to exercise, Addendum • 103
Gene
Accession Id
Description
LTA
NM_000595
lymphotoxin alpha (TNF superfamily, member 1)
LTB
NM_009588
lymphotoxin beta (TNF superfamily, member 3)
MAP2K4
NM_003010
mitogen-activated protein kinase kinase 4
MAPK14
BC031574.1
Homo sapiens mitogen-activated protein kinase 14
MAPK14
NM_139014
mitogen-activated protein kinase 14
MAPK8
NM_139049
mitogen-activated protein kinase 8
MAPK8
NM_139049
M
F
L
mitogen-activated protein kinase 8
MAPKAPK2 NM_032960
mitogen-activated protein kinase-activated protein
MGST2
NM_002413
microsomal glutathione S-transferase 2
MGST3
NM_004528
microsomal glutathione S-transferase 3
NGFR
NM_002507
nerve growth factor receptor (TNFR superfamily,
NOS1
NM_000620
nitric oxide synthase 1 (neuronal)
↑
kinase 2
↑
member 16)
NOS2
NM_000625
nitric oxide synthase 2, inducible
NPY1R
NM_000909
neuropeptide Y receptor Y1
PRKCA
NM_002737
protein kinase C, alpha
PRKCB
BC036472.1
Homo sapiens protein kinase C, beta 1
PRKCQ
NM_006257
protein kinase C, theta
PRKCZ
BC014270.2
protein kinase C, zeta
PTGDR
U31099.1
Human DP prostanoid receptor (PTGDR)
PTGS1
NM_080591
prostaglandin-endoperoxide synthase 1
PTGS2
NM_000963
prostaglandin-endoperoxide synthase 2
SELE
NM_000450
selectin E
SELL
NM_000655
selectin L
SELP
NM_003005
selectin P (granule membrane protein 140kDa,
↑
↑
↓
↑
antigen CD62)
SMAD5
NM_001001419 SMAD family member 5 (SMAD5), transcript variant 2
TBXAS1
NM_030984
thromboxane A synthase 1 (platelet)
TGFB1
NM_000660
transforming growth factor, beta 1
TGFB1
NM_000660
transforming growth factor, beta 1
TIAM1
NM_003253
T-cell lymphoma invasion and metastasis 1
TIAM2
NM_012454.
T-cell lymphoma invasion and metastasis 2 transcript
variant 1
TNF
NM_000594
tumor necrosis factor (TNF superfamily, member 2)
↑
104 • Letter to the editor
Letter to the editor
Does prolonged exhausting exercise influence the immune
system in solid organ transplant recipients?
Ingmar Königsrainer, Derek Zieker and Alfred Königsrainer
Department of General, Visceral and Transplant Surgery, University of Tuebingen
Hoppe-Seylerstr 3
72076 Tübingen
ingmar koenigsrainer@med.uni-tuebingen.de
Dear sir,
Exhausting endurance exercise exhibits strong effects on the immune system (1).
Such effects have been attributed to changes in the cellular composition of peripheral blood as well as to changes in the expression of plausible candidate genes
(2,3).
The role of exhaustive exercise in transplant candidates is unclear up to now and
of great interest for the transplant society. After organ transplantation the immune
system is strongly affected by the life-long required immunosuppressive medication.
There are numerous sport events and even Olympic games for transplant recipients.
The Euregio cycling tour is a successfully performed yearly event for transplant
patients and takes place in Austria/Italy. The tour lasts 3 days, leads through the
Austrian/Italian Alps and total distance is about 330 km (day one: 140 km, day
two: 90 km, day 3: 102 km). In the last tour in June 2008 60 cyclists (doctors,
nurses, friends) and 22 transplant recipients (19 men/3 woman) who had been
successfully transplanted for liver, heart or renal failure participated successfully.
All of them were under stable immunosuppression and had normal organ function. All participants and especially patients were able to finish the race within the
3 days and suffered from no major physical problems during and after the tour.
We think that this event touches a new field of exercise related research and plan
to do microarray analysis on the next tour in 2009 to find a list of candidate genes
which might help to monitor the immunological response to exercise in this special condition as compared to healthy subjects. With this letter we like to bring
this to the attention of the international community of scientists working in the
area of exercise immunology. We would welcome any suggestion for cooperation
in this field from other countries and continents.
LITERATURE
1.
Radom-Aizik S, Leu SY, Cooper DM, Zaldivar F Jr. Serum from exercising humans
suppresses t-cell cytokine production. Cytokine. 2007;40:75-81.
Letter to the editor • 105
2.
Zieker D, Zieker J, Dietzsch J, Burnet M, Northoff H, Fehrenbach E. CDNAmicroarray analysis as a research tool for expression profiling in human peripheral
blood following exercise. Exerc Immunol Rev. 2005;11:86-96
3.
Zieker D, Fehrenbach E, Dietzsch J, Fliegner J, Waidmann M, Nieselt K, GebickeHaerter P, Spanagel R, Simon P, Niess AM, Northoff H. cDNA microarray analysis
reveals novel candidate genes expressed in human peripheral blood following
exhaustive exercise. Physiol Genomics. 2005;17;23:287-94.
106 • Instructions for authors of EIR
Instructions for authors of EIR
EIR usually solicits papers from authors with acknowledged expertise in the field
to be covered. Unsolicited papers will be considered and can also be accepted. All
papers are subject to a peer review process.
Usually the manuscripts will fit into one of two major categories: i. a review
which thoroughly covers the area indicated in the heading and includes structuring and critical discussion of existing knowledge and, if possible, the ideas of the
authors about potential practical consequences and future developments. Mere
mentioning and listing of existing literature is not considered to be a good review.
The review can be long, if necessary, or short, if the field covered by the heading
is relatively new or very focussed. ii. a paper showing original data accompanied
by an extended, review-type discussion.
The general format of the review is somewhat flexible. A review must however
have an abstract, an introduction and a conclusion around the main sections.
Reviews with three or more sections should list the headings of the sections in
form of a bullet point table at the end of the introduction. Longer sections should
also give a short interim summary at their end.
If substantial amounts of the authors‘ own new data are to be shown, a section on
methods and on results must be included. Data will only be accepted, if methods
are stated clearly and appropriate statistical evaluation of results is given.
Other types of papers, eg true meta-analyses of a circumscribed sector of literature or papers focussing on new ideas or hypotheses may also be considered.
Interested authors, please contact the editorial board.
For reference style use the one as applied by J. Appl. Physiol., with references
listed in alphabetical order. In text use ref. numbers in brackets. When giving
more than 1 reference in one bracket, use numerical order.
A short running head should appear after the title, followed by the authors and
their respective affiliations. The full address of correspondence should include an
e-mail address of the correspondent author. Up to five key words should be added
after the abstract.
Instructions for authors of EIR • 107
Send manuscript to Hinnak Northoff, Derek Zieker or one of the other editors.
Please use e-mail for all communications including manuscript submission (word
or pdf-file) if possible and paste "EIR” in the subject field of your mailing
program.
Prof. Dr. Hinnak Northoff
Editor EIR
Institute of clinical and experimental
Transfusion Medicine (IKET)
University of Tübingen
Otfried-Müller-Str. 4/1
D-72076 Tübingen
Tel.: + 49-7071-2981601
Fax: + 49-7071-295240
E-mail: hinnak.northoff@med.uni.tuebingen.de
Dr. Derek Zieker
Managing Editor, EIR
Institute of clinical and experimental
Transfusion Medicine (IKET)
University of Tübingen
Otfried-Müller-Str. 4/1
D-72076 Tübingen
Tel.: + 49-7071-2981657
Fax: + 49-7071-295240
derek.zieker@med-uni-tuebingen.de