PowerPoint
Depression signs and how exercise yoga helps
great graphics
Brain, Behavior, and Immunity 26 (2012) 251266
Contents lists available at SciVerse ScienceDirect
Brain, Behavior, and Immunity
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y b r b i
Review
Neuroimmunological effects of physical exercise in depression
Harris Eyre a, Bernhard T. Baune b,
a Psychiatry and Psychiatric Neuroscience Research Group, School of Medicine and Dentistry, James Cook University, 101 Angus Smith Drive, Townsville, Queensland 4811, Australia
b Discipline of Psychiatry, School of Medicine, University of Adelaide, North Terrace, Adelaide, SA 5005, Australia
a r t i c l e i n f o
Article history:
Received 15 June 2011
Received in revised form 25 September
2011
Accepted 26 September 2011
Available online 2 October 2011
Keywords:
Neuroimmunology
Neurobiology
Human
Rodent
Depression
Exercise
Physical activity
Immunology
Stress
0889-1591/$ – see front matter 2011 Elsevier Inc. A
doi:10.1016/j.bbi.2011.09.015
Corresponding author. Address: Discipline of Psy
University of Adelaide, North Terrace, Eleanor Harrald
Australia. Fax: +61 8 8222 2865.
E-mail address: [emailprotected]
a b s t r a c t
The search for an extended understanding of the causes of depression, and for the development of addi-
tional effective treatments is highly significant. Clinical and pre-clinical studies suggest stress is a key
mediator in the pathophysiology of depression. Exercise is a readily available therapeutic option, effective
as a first-line treatment in mild to moderate depression. In pre-clinical models exercise attenuates stress-
related depression-like behaviours. Cellular and humoral neuroimmune mechanisms beyond inflamma-
tion and oxidative stress are highly significant in understanding depression pathogenesis. The effects of
exercise on such mechanisms are unclear. When clinical and pre-clinical data is taken together, exercise
may reduce inflammation and oxidation stress via a multitude of cellular and humoral neuroimmune
changes. Astrocytes, microglia and T cells have an antiinflammatory and neuroprotective functions via
a variety of mechanisms. It is unknown whether exercise has effects on specific neuroimmune markers
implicated in the pathogenesis of depression such as markers of immunosenescence, B or T cell reactivity,
astrocyte populations, self-specific CD4+ T cells, T helper 17 cells or T regulatory cells.
2011 Elsevier Inc. All rights reserved.
1. Introduction
The increasing prevalence of unipolar major depressive disorder
makes the search for an extended understanding of the causes of
depression, and for the development of additional effective treat-
ments highly significant (WHO, 2008). Depression is caused by a
complex interaction of multiple factors which can be most reason-
ably understood by applying a bio-psycho-social framework. These
bio-psycho-social factors are interrelated, with chronic stress being
a major influencer (Moller-Leimkuhler, 2010). Chronic psychologi-
cal stress precedes the majority (some 80%) of episodes of clinical
depression (Kessler, 1997; Mazure, 1998; Caspi et al., 2003;
McEwen, 2003; Bartolomucci and Leopardi, 2009; Risch et al.,
2009). Similarly in animal models chronic stress is a precipitant of
depression-like behaviour (Willner, 2005; Kubera et al., 2011). The
pathophysiology of stress-associated depression is hypothesised to
be associated with various neurobiological changes which are
thought to be essential to molecular mechanisms of memory, learn-
ing, and symptoms of depression (Baune, 2009; Miller et al., 2009).
These neurobiological changes in depression occur in the mono-
amine system, hypothalamopituitaryadrenal (HPA) axis, neuro-
ll rights reserved.
chiatry, School of Medicine,
Building, Adelaide, SA 5005,
(B.T. Baune).
genesis system and the neuroimmune system. A special emphasis
has been given to neuroimmune processes since they may directly
and indirectly affect the pathophysiology of depression by effecting
other important neurobiological processes of depression (Garcia-
Bueno et al., 2008; Maes et al., 2009; Kubera et al., 2011).
Production of neuroinflammatory factors, i.e. tumour necrosis
factor alpha TNF-a, interleukin-6 IL-6, C-reactive protein
CRP, interleukin-1beta IL-1b affect the main neuroimmune
mechanisms potentially leading to symptoms of depression-like
behaviour (Garcia-Bueno et al., 2008; Anisman, 2009; Maes et al.,
2009; Kubera et al., 2011; Wager-Smith and Markou, 2011). These
findings have lead to the formulation of the cytokine model of
depression due to the capacity of pro-inflammatory cytokines to
induce sickness behaviour, which closely resembles depression-
like behaviour in humans (Dantzer et al., 2008; Capuron and Miller,
2011). Neuroinflammatory mechanisms in depression are thought
to negatively interact with various pathways and can lead to
monoamine dysfunction (e.g. low serotonin levels, creation of neu-
rotoxic tryptophan-like by-products (3-hydroxykynurenine (3-HK)
and quinolinic acid (QA)), HPA axis dysfunction (e.g. hypercortiso-
laemia and reduced glucocorticoid receptor density), neurogenesis
dysfunction (e.g. apoptosis and reduced neurotrophin creation)
and neuroimmune dysfunction (e.g. decreased T cell proliferation,
increased apoptotic rate and impaired T cell function) (Caruso
et al., 1993; Maes et al., 1995; Mellor et al., 2003; Clark et al.,
2005; Miller et al., 2009; Kubera et al., 2011). Pro-inflammatory
252 H. Eyre, B.T. Baune / Brain, Behavior, and Immunity 26 (2012) 251266
cytokines can arise from central and systemic cellular neuroim-
mune changes. Cells which are implicated in their creation include
astrocytes, microglia, macrophages and T cells (Garcia-Bueno et al.,
2008).
Few studies in depression research have directly examined the
relative expression and function of relevant T cell subsets, and
other relevant immune cells, beyond the characterisation of
CD4+, CD8+ T cells and T cell mitogen responses in depression
(Capuron and Miller, 2011). Other cellular neuroimmune mecha-
nisms have also been implicated in the pathophysiology of depres-
sion. These neuroimmune mechanisms include dysfunction of
CD4+CD25+ T regulatory (Treg) cells, T helper (TH17) cells, self-
specific CD4+ T cells, monocyte-derived macrophages, macro-
phages, astrocytes and microglia (Schwartz and Shechter,
2010a,b; Capuron and Miller, 2011). These cells are suggested to
have various roles involving regulation of inflammatory mediators,
regulating neurogenesis, regulating reactive oxygen species (ROS)
formation and also cell-to-cell interactions which may mediate
neuroimmune mechanisms of the pathogenesis of depression. In
this review we will provide a comprehensive and up-to-date re-
view of the humoral and cell-mediated neuroimmunological
mechanisms associated with depression by reviewing the most re-
cent literature. We will evaluate these mechanisms for their poten-
tial to act as novel targets for therapeutic interventions.
In recent years it has been suggested that interventions such as
antidepressants, and alternative approaches such as exercise may
exert therapeutic neuroimmune-modulating effects. In relation to
antidepressants, a recent review article by Kubera et al. (2011) sug-
gests that antidepressants may positively influence inflammatory,
oxidative, apoptotic and antineurogenic mechanisms relevant to
stress-associated depression-like behaviour. A review article by
this group (Janssen et al., 2010) presents a detailed assessment of
the cytokine response to antidepressants, and how treatment re-
sponse might be affected by genetic variants relating to cytokines.
Anti-psychotic medication and electroconvulsive therapy (ECT) are
other psychiatric interventions showing neuroimmune-modulat-
ing effects (Hestad et al., 2003; Pae et al., 2010). The efficacy of
alternative therapies in clinical depression (i.e. polyunsaturated
fatty acids (e.g. Omega-3), anti-inflammatories (Acetlysalicylic acid
and Celecoxib), exercise (resistance, aerobic and flexibility) and
mindfulness-based therapies (i.e. mindfulness-based cognitive
therapy, mindfulness meditation and mindfulness-based stress
reduction therapy)) may also be correlated with neuroimmune-
modulating abilities (Maes et al., 2000; Carlson et al., 2007; Dinan
et al., 2009; Guo et al., 2009; Song and Wang, 2011). One study has
shown that tricyclic antidepressants (TCA) cause an increase in
inflammation, as measured by CRP, however, other authors have
debated these findings (Hamer et al., 2011; Pizzi et al., 2011).
Exercise is a readily available therapeutic option, effective as a
first-line treatment in mild to moderate depression (Carek et al.,
2011). Additionally, exercise has a utility in preventing depression
and has beneficial effects on other common co-morbidities (i.e.
cardiovascular disease risk factors and glycemic control). A pro-
spective, randomised controlled trial found that exercise was as
effective as Sertraline (selective serotonin reuptake inhibitor) for
the treatment of depression the effect size of exercise was 2.0
(Blumenthal et al., 2007). Several reviews show exercise compares
favourably to antidepressants and cognitive behavioural therapy
(CBT) as a first-line treatment for mild to moderate depression
(Mead et al., 2009; Carek et al., 2011).
The efficacy of exercise in depression is classically attributed to
its impact on changing certain neurobiological mechanisms includ-
ing monoamine metabolism (e.g. increasing serotonin levels in the
CNS), HPA axis function (e.g. decreasing long-term basal levels of
cortisol), neurotrophic factors (e.g. increasing brain derived neuro-
trophic factor (BDNF) and neurogenesis) and neuroinflammation
(e.g. decreasing pro-inflammatory mediators) (Chaouloff et al.,
1985; Droste et al., 2003; Garcia et al., 2003; Greenwood et al.,
2005; Kohut et al., 2006; Nabkasorn et al., 2006; Tang et al.,
2008; Bednarczyk et al., 2009; Clark et al., 2009; Van der Borght
et al., 2009; Christiansen et al., 2010; Donges et al., 2010; Mata
et al., 2010; Rethorst et al., 2010; Sousae Silva et al., 2010). The ef-
fects of exercise on neuroimmune mechanisms other than neuroin-
flammation (e.g. cell-mediated factors such as Tregs, Th17 cells,
CNS macrophages, microglia etc.) are unclear (Beavers et al.,
2010a; Archer et al., 2011). Moreover, how these cellular changes
relate to positive effects on the monoamine system, HPA axis and
neurotrophic system also remains poorly understood (Beavers
et al., 2010a; Archer et al., 2011). Surprisingly, a comprehensive
analysis of the effects of exercise on neuroimmune mechanisms
and stress-associated depression, including both clinical and pre-
clinical research, is lacking in the literature.
In this review we provide a theoretical model whereby we show
that the beneficial effects of exercise in depression are potentially
mediated through various pathways of the neuroimmune system
(see Figs. 2 and 3). Our proposed model on the effects of exercise
will be based on evidence and empirical relationships from previ-
ously published literature. The model on various aspects of the
neuroimmune system may also be relevant for its therapeutic ef-
fects in other neuropsychiatric disorders including anxiety disor-
der, schizophrenia, Alzheimers disease, Parkinsons disease and
mild cognitive impairment (Conn, 2010a,b; Lautenschlager et al.,
2010; Petzinger et al., 2010; Tajiri et al., 2010; Carek et al., 2011;
Nation et al., 2011; Wolf et al., 2011).
The aims of this review article are to present evidence for the
involvement of the neuroimmune system in the pathogenesis of
stress-associated depression, and also to provide evidence for the
immunomodulatory effects of exercise in depression. It is proposed
that exercise will exert its action on symptoms of depression via a
variety of neuroimmunological mechanisms (Figs. 2 and 3).
2. Methods
An electronic search of reputable databases such as PubMed,
PsychoInfo, OvidSP and ScienceDirect were utilised in the creation
of this literature review. Initial searching (revealing 1500 ab-
stracts) was conducted using various combinations of the follow-
ing keywords: neurotrophin, neuroinflammation, neuroimmune,
intervention, monoamine, depression, exercise, physical activity,
cytokine, hypothesis, stress, chronic, psychological, stress-induced
depression, model, mouse, rat and human. Abstracts were selected
based on the year of publication (between 1990 and 2011), publi-
cation in the English language and of peer-reviewed type. They
were excluded if they included anecdotal evidence. In this process
1000 abstracts were excluded and the remaining 500 full text arti-
cles were sought. The resulting 500 full text articles were read
thoroughly and their utilisation in this review was based on their
journal type (i.e. peer reviewed) and salience to the aims set forth
in this review. Finally 214 articles were utilised in the making of
this literature review (Fig. 1 depicts this strategy).
3. Stress-associated depression: clinical and pre-clinical
evidence
The concept of stress-associated depression-like behaviour has
been known for many years with evidence derived from both clin-
ical and pre-clinical models. The following section will briefly out-
line most recent evidence for stress-associated depression, before
moving onto its neuroimmune correlates.
Psychological stress is a known precipitant of depressive
symptoms in the clinical setting; moreover depression is known
Fig. 1. Study inclusion flowchart.
H. Eyre, B.T. Baune / Brain, Behavior, and Immunity 26 (2012) 251266 253
to further exacerbate the stress response leading to a vicious cycle
which intensifies subsequent stressors (Kessler, 1997; Mazure,
1998; Caspi et al., 2003; McEwen, 2003; Bartolomucci and Leop-
ardi, 2009; Risch et al., 2009). Chronic stress is also associated with
precipitation and exacerbation of anxiety disorder and cognitive
impairment (e.g. mild cognitive impairment and Alzheimers dis-
ease) via similar neurobiological mechanisms which are reviewed
in: (Brady and Sinha, 2005; Miller et al., 2007; Conrad, 2010; de
Rooij et al., 2010; Nation et al., 2011).
When considering translational research between clinical and
pre-clinical models it is important to describe the stressors which
are associated with the onset of depression (Anisman et al., 2002).
Anisman et al. (2002) suggests that the stress involved in the
stress-depression continuum needs to be considered based on
severity, chronicity and predictability. Numerous investigators in
this field have found protracted, unpredictable and relatively mild
psychological stress is highly relevant to depressive symptoms in
humans (Tennant, 2002; Bartolomucci and Leopardi, 2009; Baune,
2009). Similar observations are noted in rodent studies, particu-
larly from the use of the unpredictable chronic mild stress para-
digm (Willner, 2005). Many other investigators have established
the link between unpredictable, chronic, mild stress and depres-
sion in human and rodent studies (Dura et al., 1990; Caspi et al.,
2003; McEwen, 2003; Risch et al., 2009; Frodl et al., 2010; Kubera
et al., 2011; Karg et al., 2011; Wager-Smith and Markou, 2011).
There is a large body of evidence in pre-clinical rodent models
supporting the concept of stress-associated depression-like behav-
iour. Researching depression-like behaviour in rodents includes
two main components, modelling and testing. Modelling whereby
certain variables (e.g. environment) are manipulated in order to in-
duce the required phenotype, and testing where the outcome of
the modelling is evaluated (Pollak et al., 2010).
Many models investigating rodent depression include chronic
stress paradigms (e.g. chronic mild stress or chronic foot shock
stress), adverse life events (e.g. prenatal stress) and genetic modifi-
cation (e.g. regarding depression-related genes). For the purpose of
this review, the unpredictable chronic mild stress (uCMS) paradigm
is selected as it shows strength in all descriptive validation criteria
(Willner, 1997). Additionally, the uCMS protocol is known to elicit
anxiety-like symptoms, schizophrenia-like behaviour and impair-
ments in cognition-like behaviour (Mineur et al., 2006; Conrad,
2010; Salomons et al., 2010; Wolf et al., 2011). The stressors of
uCMS are congruous in duration, intensity and predictability to
the stressors known to be associated with human depression (Ten-
nant, 2002). The uCMS paradigm consists of unpredictable and
chronic exposure to environmental changes (e.g. cage dampening,
cage tilting and food/water deprivation). The unpredictability of
the uCMS paradigm is important for the development of depres-
sion-like behaviour as predictable chronic mild stress is shown to
improve depression-like behaviour, hippocampal neurogenesis
and memory (Parihar et al., 2011). The chronicity of unpredictabil-
ity in environment is important in the development of depression
in clinical and pre-clinical models. Indeed, in clinical models, there
is a large body of literature outlining the role of uncertainty in med-
ical illnesses (i.e. exacerbations of illness in multiple sclerosis, asth-
ma, atrial fibrillation and other chronic illnesses), certain
psychological processes and traits (i.e. pessimism, hopelessness,
depressive predictive certainty, intolerance of uncertainty, neurot-
icism) and environment unpredictability in the development of
depression (Mullins et al., 2000; Kroencke et al., 2001; Lynch
et al., 2001; Miranda et al., 2008; McEvoy and Mahoney, 2011).
The specific tests examining the outcome of this model utilised
in this review will be the forced swim test (FST), tail suspension
test (TST), sucrose consumption and sucrose preference tests.
These tests show good rationale and consistently high validity.
TST and FST are based on the principle that immobility is sugges-
tive of apathy, disengagement, despair or entrapment; all of
which are well known symptoms/signs of depression (Deussing,
254 H. Eyre, B.T. Baune / Brain, Behavior, and Immunity 26 (2012) 251266
2006). Sucrose testing assesses anhedonia or loss of capacity to
experience pleasure: this is inferred by the measured consumption
or preference for a pleasureable sucrose fluid. Lower levels of con-
sumption suggest anhedonia. Together these three tests measure
depression-like behaviour.
Various molecular biological correlates are suggested to be
associated with the model of stress-associated depression-like
behaviour. These findings can be separated into four mechanisms
including (1) monoamine dysfunction, (2) HPA axis dysfunction,
(3) neurogenesis dysfunction and (4) neuroimmune system dys-
function (Fig. 2) as shown in various studies in humans with
depression and in animals investigating depression-like behaviour
(Eaton et al., 1996; Lanfumey et al., 2000; Wust et al., 2000; Pruess-
ner et al., 2003; Gronli et al., 2006; Banasr et al., 2007; Goshen
et al., 2008; Li et al., 2008; Luo et al., 2008; Pace and Miller,
2009; Elizalde et al., 2010; Frodl et al., 2010; Larsen et al., 2010;
Karg et al., 2011).
3.1. Stress-associated neuroimmunological changes in depression
For the purpose of this review, we focus on the neuroimmune
dysfunction related to the development of depression-like behav-
iour in clinical and pre-clinical studies.
3.1.1. Clinical studies
Clinical evidence suggests chronic stress induces depressive
symptoms and various neuroimmune changes. Systemic IL-6, CRP
and NF-jB are consistently elevated in association with chronic
stress-related depressive symptoms (see Table 1). Chronic stress
induces increased Natural Killer (NK) cell function, increased
Immunosenescence (i.e. lower CD4:CD8 ratio, higher proportion
of CD8+ T lymphocytes (CTL) with an effector-memory phenotype
or late differentiated (CD27CD28) and lower proportions of
CTLs in early differentiation phase (CD27+CD28+)), lower CD4+
helper T cells, higher CD8+ suppressor T cells, higher CD8+/
CD57+ activated T lymphocytes and a higher CD4+/CD8+ ratio
(Pace et al., 2006; Caserta et al., 2008; Bosch et al., 2009; Beavers
et al., 2010a). TNF-a has also demonstrated to be induced by
chronic stress (Amati et al., 2010).
Subjects with the shortshort allele of the serotonin transporter
(5-HTTLPR) polymorphism (which is correlated to lower serotonin
availability and susceptibility to stress and depression) showed a
pro-inflammatory state (increased IL-6/IL-10 ratio) when compared
to long-long counterparts. This finding may be interpreted as a pos-
sible biomarker suggesting stress susceptibility (Fredericks et al.,
2010). In depressed patients, systemic inflammatory response is
found to be exaggerated by acute stressors. Pace et al. (2006) found
patients with Major Depressive Disorder having higher levels of NF-
jB (inflammation-related nuclear transcription factor) when ex-
posed to acute stress, as opposed to controls. A pre-clinical study
Fig. 2. Neuroimmunological effects of exercise in depression. Exercise impacts
positively on neuroimmune mechanisms which in turn affect attenuation of
depression and chronic stress (dotted line). Chronic stress impacts negatively on
neuroimmune mechanisms which in turn affect initiation and perpetuation of
depression (bolded line).
by Anisman et al. reflects a similar augmentation in inflamma-
tion-related mediators after acute stress in mice with depression-
like behaviour; a significant elevation in circulating cytokine levels
(i.e. IL-6, TNF-a, IL-10 but not IL-1b, IFN-c) was found after social
isolation stress in addition to chronic cytokine-induced (IFN-a)
depression (Anisman et al., 2007). Acute stress results in a hyper-
reactivity of the pro-inflammatory response versus non-stressed
control (Maes et al., 1998; Steptoe et al., 2001; Bierhaus et al.,
2003; Brydon et al., 2004, 2008; Kop et al., 2008). Chronic stress
in the studies mentioned above was quite widely varied, such var-
iability needs to be considered when comparing neuroimmune
markers. It included parenting a child with cancer, care giving for
a family member with dementia, early life stress, maltreatment
and social isolation. There was also a wide variety of stress scales
and depressive symptom scales utilised in these articles which
diminishes the comparability of the different studies.
3.1.2. Pre-clinical studies
There is a robust literature surrounding the neuroimmune
changes associated with uCMS-related depression-like behaviour
in rodent studies (see Table 2). uCMS is associated with molecular
neuroimmune changes in the CNS including increased proinflamma-
tory cytokines (TNF-a, IL-1b, IL-6) (Sudom et al., 2004), increased
complement activity (Ayensu et al., 1995), increased TLR-4, in-
creased NK-kB, increased ROS (Lucca et al., 2009), increased COX-2
and PGE-2 (Guo et al., 2009). Increases in proinflammatory cytokines
and oxidative stress markers are seen in plasma and in the CNS in
various brain regions including then hypothalamus, pituitary, hip-
pocampus, prefrontal cortex and cortex. There are no studies assess-
ing stress-associated changes in inflammatory cytokines or
oxidative stress markers in the amygdala. uCMS is also associated
with increased systemic B cell reactivity, decreased systemic T cell
reactivity, increased systemic T cell dependent/independent humor-
al immunity markers and increased splenic mononuclear prolifera-
tion (Azpiroz et al., 1999; Edgar et al., 2002, 2003; Silberman et al.,
2004; aan het Rot et al., 2009). uCMS was also found to be associated
with a decrease in hippocampal astrocyte density (Ritchie et al.,
2004). Several studies support the evidence cited above (Kubera
et al., 1998; Silberman et al., 2002, 2005; Munhoz et al., 2006; Pal-
umbo et al., 2010; Rubinstein et al., 2010).
Stress-induced changes to neurobiological systems which were
previously thought to be unrelated to the immune system have been
discovered to influence the neuroimmune environment and induce
depression-like behaviour. These include the cannabinoid system,
IGF-1 and COX system (Duman et al., 2009; Guo et al., 2009; Beyer
et al., 2010; Park et al., 2011). The interaction of these systems with
the neuroimmune environment needs further exploration.
There are a number of putative psychoneuroimmunological fac-
tors associated with the pathogenesis of depression (Kubera et al.,
2011; Schwartz and Shechter, 2010a,b) (see Tables 1 and 2). The
most well acknowledged understanding for the development of
depression suggests that risk factors (mainly exposure to stress,
but also genetic polymorphisms) combine to trigger a cascade of
neuro-injury. Neuro-injury is thought to be mediated via activation
of pathogen associated molecular patterns (PAMPS) and danger
associated molecular pattern detectors (DAMPs) within the innate
immune system (Kubera et al., 2011; Loftis et al., 2010; Maes et al.,
2011). In addition there is modification of immune cell receptors
(e.g. toll-like receptors) resulting in the overproduction of
pro-inflammatory mediators like TNF-a, IL-1b, IL-6 and Prostaglan-
din-E2 (PGE-2) in various brain regions (i.e. hippocampus, prefron-
tal cortex and nucleus tractus solitaries) (Chang et al., 2008; Gibb
et al., 2008; Maccioni et al., 2009; Kubera et al., 2011). Pro-inflam-
matory cytokines can be released centrally (via microglia and
astrocytes) or peripherally (via monocytes, macrophages, Th17
cells and other T cells) and certain cytokine signals are able to
Table 1
Neuroimmunological changes in stress-associated depression: clinical studies.
Study Study population
(source and number)
Age,
mean
(range)
Study
design
Stressor type Psychological
measures
Biological measure Findings
Miller
et al.
(2002)
Oncology clinic, N = 50 37 Cross-
sectional
Parents of
children with
cancer
CES-D
POMS
PSS
Expose blood to cort ? measure
IL-1b, IL-6, TNF-a (i.e. measure
anti-inflamm effect of cort)
” stress = ” depressive sxs
” stress = ” IL-6
Kiecolt-
Glaser
et al.
(2003)
Nursing home, N = 225 (55
89)
Prospective
8 years
Dementia
caregiver
PSS
BDI
NLS
Plasma IL-6 Caregiving = ” stress
Caregiving = ” IL-6
Caregiving = ” depressive
symptoms
(Depressive symptoms were
not correlated with IL-6
levels)
Pace et al.
(2006)
Health volunteers,
N = 28 (MDD in 14
subjects (DSM-4))
29.9 Prospective TSST
Speech task
Examination
SCID
HDS
ZDS
CTQ (Early life
stress)
Plasma IL-6
Lymphocyte subsets
NF-jB activation within PBMCs
MDD = ” CTQ score
MDD + ” CTQ = ” IL-6
TSST = ” IL-6
MDD > non-MDD
TSST = ” NF-jB
MDD > non-MDD
TSST = ” NK cells
MDD = non-MDD
CTQ ns D with IL-6 or NF-jB
D NK = ns D NF-jB and D IL-6
Danese
et al.
(2009)
New Zealand population
sample, N = 862
32 Cross-
sectional
NZSEI (SES)
Childhood
maltreatment
measure
RCS
(childhood
social
isolation)
Psychiatric
interview for
MDD (DSM-4)
Plasma CRP MDD / definite maltreatment
RR1.69
MDD / social isolation RR1.76
” CRP / definite maltreatment
RR1.56
” CRP / social isolation
RR1.60
Fredericks
et al.
(2010)
Healthy SS or LL 5-
HTTLPR polymorphism
specific patients, N = 30
21.7 Prospective TSST
Speech task
Examination
SCID
LESS
CTQ
SSGS
RAG
SUDS
BDI
Serum IL6 and IL10
Genotyping
TSST = ” in SSGS shame and
pride subscales, SUDS and
RAG
TSST = ” IL6, ” IL10
IL-6/IL-10 ratio ” in SS vs. LL at
baseline and after TSST
NS difference in IL6 or IL10
between LL and SS genotypes
either before or after TSST
At baseline, NS difference btw
SS and LL for CTQ, BDI, RAG or
LESS
Bob et al.
(2010)
Unipolar MDD
inpatient, N = 40
42.3
(30
58)
Cross-
sectional
None BDI-2
TSC-40
DES
SDQ-20
Serum IL-6 IL-6 / BDI-2, TSC-40, SDQ-20
IL-6 not / DES
NB: Level of significance set at p < 0.05.
Legend: N = number, PSS = Perceived Stress Scale, STAI = State-Trait Anxiety Inventory, POMS Profile of Mood States, cort = cortisol, CES-D = Centre for Epidemiological
Studies Depression, sx = symptom, BDI = Beck Depression Inventory, NLS = NYU Loneliness Scale, PSSS = Perceived Social Support Scale, ROS = Role Overload Scale, BSI = Brief
Symptom Inventory, TSST Trier social stress test, PBMC = Peripheral Blood Mononuclear Cells, NF-jB = Nuclear Factor Kappa B, EMSA = Electrophoretic Mobility-Shift Assay,
AD = Adrenaline, NA = Noradrenaline, MDD = Major Depressive Disorder, HDS = Hamilton Depression Scale, ZDS = Zung Depression Scale, CTQ = Childhood Trauma Ques-
tionnaire, NK = natural killer cells, CRP = C-Reactive Protein, DSM Diagnostic and Statistical Manual, SES = Socio-Economic Status, RCS = Rutter Child Scale, NZSEI = New
Zealand Socioeconomic Index, vs. = versus CAD = Coronary Artery Disease, PCI = Percutaneous Coronary Intervention, sICAM1 = soluble Intra-cellular Adhesion Molecule-1,
D = change in, / = correlation, HJSS = Healthcare Job Satisfaction Scale, GHQ = General Health Questionnaire, MSPSS = Multidimensional Scale and Perceived Social Support,
SCID = Structured Clinical Interview for DSM-4, LESS = Life Events Scale for Students, SGSS = State Shame and Guilt Scale, RAG = Russell Affect Grid, SUDS = Subjective Units of
H. Eyre, B.T. Baune / Brain, Behavior, and Immunity 26 (2012) 251266 255
reach the brain parenchyma through humoral, neural and cellular
pathways (Ziv et al., 2006; Quan and Banks, 2007; Maes et al.,
2011; Capuron and Miller, 2011). More specifically, these path-
ways include: cytokine passage through leaky regions of the BBB
(IL-6, IL-1b, TNF-a), active transport via saturable cytokine-specific
transport molecules (IL-1, TNF), activation of brain endothelial
cells which release secondary messengers within the brain
(PGE2, nitric oxide (NO)), cytokine signal transmission via afferent
nerve fibres (IL-1) and entry into the brain of peripherally activated
monocytes via microglia production of monocyte chemoattractant
protein-1 (MCP-1) (Watkins et al., 1995; Plotkin et al., 1996;
Goehler et al., 1999; Quan and Banks, 2007; DMello et al., 2009;
Capuron and Miller, 2011).
Proinflammatory cytokines are thought to have an active role in
molecular mechanisms that influence monoamine metabolism,
neuronal genesis/survival, HPA axis sensitivity to cortisol and
certain cellular neuroimmune functions (Barrientos et al., 2003;
Miller et al., 2009; Kubera et al., 2011). The cytokines can induce
enzymatic activity increasing indoleamine-pyrrole 2,3-dioxygen-
ase and tryptophan 2,3-dioxygenase (TDO) whilst at the same time
decreasing blood tryptophan and hence serotonin levels (Goshen
et al., 2008). Reduced serotonin in turn creates more vulnerability
to stress and sets up a positive feedback loop for continued neuro-
biological damage. A byproduct of IDO/TDO is kyrunenine, a
metabolite of tryptophan, which is further metabolised by im-
mune-related cells in the brain (i.e. macrophages, microglia and
astrocytes) leading to the formation of potentially neurotoxic com-
pounds such as 3-HK and QA (Capuron and Miller, 2011). Neuronal
toxicity may cause apoptosis with lowered levels of Bcl-2 and BAG-
1 (Bcl-2 associated athanogene 1) and increased levels of caspase-3
Table 2
Neuroimmunological effects of depression-like behaviour: pre-clinical studies.
Study Animal/strain uCMS
duration
(weeks)
Test Other Biological measure Findings
Koo et al.
(2010)
Rat/SpragueDawley
(WT and NF-jB/LacZ
transgenic reporter
mice)
4 weeks
(atypical)
Or acute
stress
exposure
Sucrose
consumption
CNS infusion
of NF-jB
inhibitor or
IL-1b
BrdU injection (neurogenesis marker) SGZ,
DG
In vitro AHPs (nestin + brdU measured with
immunofluorescence and TUNEL assay)
CMS = ; sucrose consumption
CMS + NF-jB inhibitor ;
sucrose consumption
Acute or Chronic stress = ;
neurogenesis
Acute or Chronic stress + NF-jB
inhibitor ; neurogenesis
IL-1b = ; AHP
IL-1b + NF-jB inhibitor ; AHP
Silberman
et al.
(2004)
Mice/BALB/c 6 Sucrose
consumption
Serum cort
Splenic NE
T-cell dependent (SRBC and allogeneic cells)
and independent (DxB512 and LPS) humoral
response determined. Cells exposed to NE, E
and cort
Splenic lymphoid cell suspensions were
obtained
Mitogen assay PHA (lymphoid proliferation)
CMS = ; sucrose pref from 4 to
6 wks
CMS = ” cort and NE from 0 to
3 wks (ns 46 wks)
CMS = ; T cell dependent
humoral immunity markers
CMS D T cell independent
humoral immunity markers
CMS = ; T cell response to PHA
(wk 6)
Grippo et al.
(2005)
Rat/SpragueDawley 4 Sucrose pref Hypothalamus, Ant Pit, Post Pit: cytokines
Plasma: cytokines
uCMS induced anhedonia / ”
pro-inflamm cytokines (TNF-a,
IL-1b, IL-6) in the brain and in
plasma
uCMS = ” TNF-a Hypothal, Pit
uCMS = ” IL-1b Hypothal
Tannenbaum
et al.
(2002)
Mice/CD-1 54 days FST IL-1b
injection IP
PVN, ME: 5-HT
