Review

Cytokines as mediators of depression: What can we learn

from animal studies?

Adrian J. Dunna,*, Artur H. Swiergiela,b, Renaud de Beaurepairec

aDepartment of Pharmacology, Louisiana State University Health Sciences Center, P.O. Box 33932, Shreveport, LA 71130-3932, USAbInstitute of Genetics and Animal Breeding, Polish Academy of Sciences, Jastrzebiec, Poland

cHopital Paul Guiraud, Villejuif, and INSERM 513 Creteil, France

Abstract

It has recently been postulated that cytokines may cause depressive illness in man. This hypothesis is based on the following

observations: 1. Treatment of patients with cytokines can produce symptoms of depression; 2. Activation of the immune system is

observed in many depressed patients; 3. Depression occurs more frequently in those with medical disorders associated with immune

dysfunction; 4. Activation of the immune system, and administration of endotoxin (LPS) or interleukin-1 (IL-1) to animals induces

sickness behavior, which resembles depression, and chronic treatment with antidepressants has been shown to inhibit sickness behavior

induced by LPS; 5. Several cytokines can activate the hypothalamo–pituitary–adrenocortical axis (HPAA), which is commonly activated

in depressed patients; 6. Some cytokines activate cerebral noradrenergic systems, also commonly observed in depressed patients; 7.

Some cytokines activate brain serotonergic systems, which have been implicated in major depressive illness and its treatment. The

evidence for each of these tenets is reviewed and evaluated along with the effects of cytokines in classical animal tests of depression.

Although certain sickness behaviors resemble the symptoms of depression, they are not identical and each has distinct features. Thus the

value of sickness behavior as an animal model of major depressive disorder is limited, so that care should be taken in extrapolating

results from the model to the human disorder. Nevertheless, the model may provide insight into the etiology and the mechanisms

underlying some symptoms of major depressive disorder. It is concluded that immune activation and cytokines may be involved in

depressive symptoms in some patients. However, cytokines do not appear to be essential mediators of depressive illness.

q 2005 Elsevier Ltd. All rights reserved.

Keywords: Cytokines; Depression; Interleukin-1; Interferon; Norepinephrine; Serotonin; HPA Axis; Sickness behavior; Animal models

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892

1.1. Origins of the cytokine hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892

2. Experimental evidence for the cytokine hypothesis of depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892

2.1. Responses to cytokine treatment in man . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892

2.2. Immune activation in depressed patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893

2.3. Depression in immune-related medical conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894

2.4. Sickness behavior and depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894

2.4.1. Effects of antidepressant treatments on sickness behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898

2.5. HPAA activation by cytokines: the relationship to depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898

2.6. Brain catecholaminergic activation induced by cytokines: relationship to depression . . . . . . . . . . . . . . . . . . . . . . . . . . 899

2.7. Cytokines, depression and serotonin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900

3. Effects of immune activation and cytokines in classical tests of depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901

Neuroscience and Biobehavioral Reviews 29 (2005) 891–909

www.elsevier.com/locate/neubiorev

0149-7634/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.neubiorev.2005.03.023

* Corresponding author. Tel.: C1 318 675 7850; fax: C1 318 675 7857.

E-mail address: adunn@lsuhsc.edu (A.J. Dunn).

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909892

3.1. The Porsolt or forced swim test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901

3.2. Tail suspension test (TST) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902

3.3. Involvement of cytokines in other models of depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902

3.4. Usefulness of the cytokine induced sickness behavior model for the study of depression . . . . . . . . . . . . . . . . . . . . . . . 902

4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904

1. Introduction

In recent years, it has been postulated that depression may

be caused by cytokine secretion associated with activation of

the immune system. This hypothesis was proposed initially

because the incidence of immune abnormalities is higher in

depressed patients than in the general population, and

because depression is a common side effect of cytokine

therapy. The hypothesis gained momentum when it was

shown that immune activation and administration to animals

of endotoxin (lipopolysaccharide, LPS) or the cytokine,

interleukin-1 (IL-1), could induce a behavioral pattern

resembling that commonly observed in sick animals,

including man. This behavioral pattern has become known

as sickness behavior. There are many similarities between

sickness behavior and the symptoms of depression. Thus it

was suggested that cytokines could induce depression, or

indeed were the cause of depression. The hypothesis has

come to be called the cytokine hypothesis of depression,

although proponents differ on whether cytokines are a cause

of major depressive disorder, or the cause.

1.1. Origins of the cytokine hypothesis

The cytokine hypothesis of depression posits that

depression is caused by the actions of cytokines. Cytokines

are proteins and glycoproteins secreted by immune cells that

function as signals among and between immune cells. They

are the hormones of the immune system. It is now known

that cytokines have effects on cells outside the immune

system, and that non-immune cells can synthesize and

secrete cytokines. Thus cytokines can be regarded as

classical hormones that can function locally or systemically

to orchestrate immune responses, and can also coordinate

immune responses with those of other physiological systems

in the body, including the nervous system.

The cytokine hypothesis of depression derives from both

clinical and experimental observations. The clinical obser-

vations were made on patients treated with interferons

(IFN’s) and interleukin-2 (IL-2). Such patients often

displayed influenza-like symptoms and nonspecific neuro-

psychiatric symptoms, some of which are characteristics of

depression. However, the initial interest was not focused on

depression, although depression is a relatively common

side-effect. Immune abnormalities in depressed patients

have been widely reported for more than 30 years. The

earliest studies suggested deficient immune function in

depressed patients, but the results were inconsistent and

highly variable (see Weisse, 1992). Subsequent studies

indicated that depressed patients often exhibited immune

activation, and that depression was more frequent in

patients with diseases having an immunological component

(Kronfol, 2002). However, the results were quite varied;

immune function in depressed patients was decreased in

some aspects, but activated in others.

The interest in an immune system involvement in

depression intensified around 15 years ago with the

juxtaposition of the above observations on immune

abnormalities in depressed patients with the effects of

immune challenges in animals. In the animal studies, it was

shown that administration of LPS or IL-1 induced a

behavioral syndrome that became known as ‘sickness

behavior’ (Kent et al., 1992). Smith first suggested that

depression was associated with increased secretion of

cytokines (especially IL-1) by macrophages (Smith,

1991). He noted that administration of IL-1 mimicked not

only some of the behavioral characteristics of depression,

but also activated the hypothalamo–pituitary–adrenocorti-

cal axis (HPAA). This macrophage theory of depression

attracted immediate attention, because it was compatible

with a number of earlier findings in depressed patients.

2. Experimental evidence for the cytokine hypothesis

of depression

The experimental evidence for the cytokine hypothesis of

depression is summarized in Table 1. Each of these points

will be reviewed in turn, focusing on the evidence from

animal models.

2.1. Responses to cytokine treatment in man

Cytokines have been shown to be effective in the

treatment of medical conditions, such as hepatitis C,

multiple sclerosis, some infections, leukemia, Kaposi’s

sarcoma, melanoma, myeloma, renal carcinoma and other

forms of cancer. The cytokines most commonly used are

IFNa, IFNb, IFNg and IL-2. Each of these cytokines has

been reported to produce side effects such as asthenia,

myalgia, confusion and influenza-like symptoms.

Depression is most commonly associated with treatment

with IFNa and IL-2, and occasionally with IFNb but

not with IFNg (Valentine et al., 1998; Gohier et al., 2003).

Table 1

Experimental evidence cited for a cytokine hypothesis of depression

Evidence Principal cytokines implicated

Treatment of patients with cytokines induces symptoms of major depressive disorder IL-2, IFNa

Activation of the immune system is observed in many depressed patients. This may be reflected in elevated

circulating concentrations of cytokines

IL-6

Depressive disorders occur more often in medical disorders involving immune dysfunction than those that do not

Activation of the immune system, and administration of LPS or cytokines to animals induces sickness behavior,

which resembles major depressive disorder. Chronic treatment with antidepressants inhibits LPS-induced sickness

behavior

IL-1

Cytokines can activate the HPA axis. Elevated plasma cortisol is commonly observed in depressed patients IL-1, IL-6, TNFa, IFNa

Cytokines activate brain noradrenergic systems. Elevated NE and its catabolites are commonly observed in the

CSF of depressed patients

IL-1, TNFa

Cytokines activate brain serotonergic systems. Abnormalities of brain serotonin have been implicated in major

depressive disorder and its treatment

IL-1, IL-6, TNFa

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 893

The evidence has been reviewed recently by de Beaurepaire

et al. (2005) and will only be summarized here. The

incidence of depression associated with cytokine therapy is

highly variable, ranging from 0 to 45% in different studies.

The reasons for this are multifold. They are in part related to

the disease under treatment and the cytokines and doses

used, but are also related to the assessment measures, the

patient’s psychiatric history, and the choice of comparison

(control) subjects (de Beaurepaire et al., 2005).

In most cases, the depressive symptoms can be treated

effectively with antidepressants. Interestingly, however,

antidepressants are sometimes more effective on the mood

symptoms, and not on the neurovegetative symptoms, like

anorexia, psychomotor retardation and sleep disorders

(Miyaoka et al., 1999; Musselman et al., 2001a; Capuron

et al., 2002), whereas it is the latter that are the focus of most

of the animal studies.

2.2. Immune activation in depressed patients

Immune function in depressed patients has been the

subject of study for around half a century. Interestingly, the

early studies indicated a tendency to decreased immune

function in depressed patients (Weisse, 1992), whereas the

more recent studies have emphasized immune activation

(see review de Beaurepaire et al., 2005). According to the

reviews of Maes (1995) and Kronfol (2002) major

depression appears to be associated with increased plasma

concentrations of positive acute-phase proteins (hapto-

globin, ceruloplasmin, C-reactive protein, hemopexine,

a1-antitrypsin, and a1-acid glycoprotein), and with lower

plasma concentrations of negative acute-phase proteins

(transferrin, albumin, and retinol-binding protein). Increases

of acute-phase proteins associated with decreases in

negative acute-phase proteins are considered to be indica-

tive of an inflammatory state. Thus, it has been postulated

that chronic depression is associated with chronic inflam-

mation (Maes, 1995). Consistent with this are the reports

that depressed patients display elevated concentrations of

inflammatory markers, such as prostaglandins (Lieb et al.,

1983) and complement (Kronfol and House, 1989). Maes

suggested that the chronic inflammatory state he believes

underlies depression can be explained by increased

production of cytokines by circulating monocytes and

macrophages (Maes, 1995). In his review, Kronfol cites

reports of increased circulating leukocytes (Kronfol and

House, 1989) (see below), activated T-cells (Maes et al.,

1992), and elevated plasma concentrations of cytokines

(Kronfol, 2002). Other studies have focused on the ability of

immune cells obtained from patients to produce cytokines in

vitro (Maes, 1999; Anisman et al., 1999). Such assays

have been argued to provide functional measures of immune

activity, but the biological significance of such ex vivo

measures is not at all clear. Also, neopterin, considered to

be a marker of activation of cell-mediated immunity,

was increased in the plasma and urine of depressed

patients (Duch et al., 1984; Dunbar et al., 1992). However,

other authors have failed to find evidence for immune

activation in depressed patients (Landmann et al., 1997;

Natelson et al., 1999).

Several studies have focused on the possible elevation of

cytokines in depressed patients. IL-1b, interleukin-6 (IL-6)

and the IFN’s have been reported to be increased in the

plasma of depressed patients (Maes et al., 1993a,b, 1994).

Increases of plasma IL-1b reported in some studies

(Griffiths et al., 1997; Owen et al., 2001) were not replicated

in others (Brambilla and Maggioni, 1998). Some have

reported higher IL-6 concentrations in cancer patients with

depression (Musselman et al., 2001b), but others found only

normal concentrations of circulating cytokines in psychia-

tric patients, including those diagnosed with depression

(Haack et al., 1999). One recent study indicated increases in

IL-1b and decreases in IL-6 in the cerebrospinal fluid (CSF)

of depressed patients (Levine et al., 1999). In this study, the

increase in IL-1b correlated with the severity of the

depression.

A recent meta-analysis of the clinical data on immune

abnormalities in depressed patients by Zorrilla et al. (2001)

found that patients with major depression exhibited: an

overall leukocytosis, manifested as a relative neutrophilia

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909894

and lymphopenia; increased CD4/CD8 ratios; increased

circulating haptoglobin, prostaglandin E2,(PGE2) and IL-6

concentrations; reduced natural killer (NK)-cell cytotox-

icity; and reduced lymphocyte proliferative responses to

mitogens. Interestingly, plasma concentrations of IL-1, the

only cytokine that clearly induces depression-like symp-

toms in animals, were not consistently altered. The authors

commented that ‘the degree of heterogeneity of the studies’

results raises questions about their robustness’. They also

noted that among these biological markers, only three were

associated with chronic stress: increased CD4/CD8 ratios,

reduced mitogen-induced proliferative responses, and

reduced NK cell cytotoxicity (Zorrilla et al., 2001).

Pollmacher et al. noted that certain cytokines (most notably

IL-1b) are undetectable in human plasma under normal

physiological conditions as well as during experimental

endotoxemia (Pollmacher et al., 2002). They also pointed

out that the physiological fluctuations of the detectable

cytokines (IL-6 and tumor necrosis factor a, TNFa) are very

poorly characterized in normal and pathological states, and

the alterations of circulating cytokines observed in the

depressed (as well as in immune-related medical conditions)

are extremely modest compared with the concentrations of

circulating cytokines that occur during cytokine treatments.

These observations pose important issues for proponents of

a simple cytokine hypothesis of depression. It is also clear

that the immune activations observed in depressed patients

are not of the magnitude observed when sickness behavior is

induced by IL-1, LPS or infections with pathogen (see

below). Another important caveat is that depression may

occur in response to an undetected medical condition that

itself results in immune activation.

Overall the evidence for an activation of immune

responses in patients with major depressive disorder is not

very consistent (de Beaurepaire et al., 2005; Zorrilla et al.,

2001), although there are trends to leukocytosis, increased

CD4/CD8 ratios, elevated plasma concentrations of hapto-

globulin, IL-6 and PGE2 and decreased NK cell activity and

lymphocyte proliferative responses. It is clear that immune

activation does not occur in all depressed patients.

2.3. Depression in immune-related medical conditions

As indicated above, there is some evidence for a higher

incidence of immune activation in depressed patients. If

immune activation were a direct cause of depression, it

should be possible to obtain evidence for this by assessing

symptoms of depression and immune states in different

medical conditions. Depression should be more prevalent in

patients that suffer from medical conditions associated with

immune activation, and it should be possible to correlate the

symptoms of depression with evidence of immune acti-

vation (such as circulating cytokines) in those conditions.

Depression has been reported to be more common in

non-infectious diseases associated with a chronic activation

of the immune system, such as multiple sclerosis (Minden

and Schiffer, 1990), allergy (Marshall, 1993), rheumatoid

arthritis (Dickens et al., 2002), and stroke (Schwartz et al.,

1993). In these diseases, the immune abnormalities precede

the development of depression (Foley et al., 1992). Never-

theless, data on the correlations between depression

and immune activation or plasma concentrations of

cytokines in patients with autoimmune diseases or chronic

infections or inflammation are few and conflicting (see

Pollmacher et al., 2002).

A complicating factor is that depression may also have a

higher prevalence in diseases characterized by a loss of

immune function, although the loss of some immune

mechanisms in these diseases is often accompanied by

increases in others, for example in asthma or in stroke. If

depression were accompanied by the loss of immune

mechanisms of defense, diseases that develop because of a

loss of these mechanisms should have a higher incidence in

depressed patients. Thus, for example, a higher prevalence

of infections and cancers should occur in depressed

patients. Some epidemiological evidence indicates that

this may be true, but there are major methodological

problems, and the biological mechanisms are very uncertain

(see Kronfol, 2002).

2.4. Sickness behavior and depression

The concept of sickness behavior was defined by Ben Hart

in a classic review (Hart, 1988). Sick animals, including

humans, exhibit decreases in a number of activities, most

notably feeding, exploration and sexual activity. They also

exhibit increased body temperature (i.e. fever) and increased

sleep. Contrary to the prevailing opinion that the behavior of

sick animals was a nonspecific malaise, Hart argued that it

was ‘not a maladaptive and undesirable effect of illness but

rather a highly organized strategy that is at times critical to

the survival of the individual if it were living in the wild

state’. The behavioral responses were important in facilitat-

ing recovery from the illness, and defending the animal

against pathogens and predators. The fever was thought to

increase the efficacy of the host’s immune system, and in

some cases to decrease the proliferation of pathogens. The

behavioral responses were considered to drive the animal to

hide, and remain safe, while the immune system cleared

invading pathogens and aided the repair of damaged tissue.

This revolutionary hypothesis immediately gained wide-

spread support. Hart also recognized that endotoxin and IL-1

induced behavioral patterns that resembled very closely

those present in sick animals. Most authors have supported

and applauded his insight.

The term ‘sickness behavior’ was coined in a subsequent

review (Kent et al., 1992), in which it was argued that the

relationships between sickness behavior and LPS and IL-1

suggested that the symptoms of sickness might be treated

with antagonists of the immune activation, such as cytokine

antagonists. The link to depression was made in Smith’s

‘macrophage theory of depression’ (Smith, 1991) in

Table 2

Sickness behaviors

Hypomotility (Lethargy)

Hyperthermia (Fever)

Hypophagia (Anorexia)

Hyperalgesia

Decreased interest in exploration

Decreased sexual activity in females

Increased time spent sleeping

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 895

which he noted the similarities between sickness behavior

and depression, and proposed specifically that IL-1 secreted

by macrophages was responsible for the depression. This

exciting series of proposals caught the imagination of many

researchers, and has achieved folklore status. However, the

simplest form of this hypothesis requires careful

assessment.

Sickness behaviors induced by IL-1 and LPS

include: hyperthermia, hypomotility, hypophagia, hyperal-

gesia, decreased interest in exploring the environment,

decreased libido, and increased time spent sleeping

(Kent et al., 1992; Dantzer et al., 2001; Larson and Dunn,

2001) (Table 2). There are also indications that motivation

and some cognitive functions may be affected (see Dantzer

et al., 2001; Larson and Dunn, 2001).

The administration of interferons in animals does not

induce classical sickness behavior. In humans, depression is

often observed following treatment with IFNa, but many

other neurobehavioral symptoms are observed, the most

common of which is fatigue (Valentine et al., 1998; Meyers,

1999; Cleeland et al., 2003). IFNa administration to rodents

does produce some behavioral symptoms (reviewed below),

but they are generally less marked than those observed in

humans. Most notably, IFNa does not typically induce fever

or HPA axis activation in mice as it does in humans

(Valentine et al., 1998, see below).

The analogies between sickness behavior and depression

are far from perfect. From a general biological perspective,

Table 3

Psychobehavioral symptoms or concomitants of major depressive disorder and an

Humans An

DSM IV symptoms of depression Cytokine

therapy

LP

Depressed mood C In

Anhedonia C CSignificant loss of weight or appetite (Weight Gain in

Atypical)

Anorexia Hy

Insomnia (Hypersomnia in atypical) C Sl

Psychomotor agitation or retardation C Lo

Fatigue or loss of energy C Lo

,

Feelings of worthlessness 0 In

Diminished ability to think or concentrate, or indeci-

siveness

C In

Recurrent thoughts of death or suicidal ideation G In

A diagnosis of depressive disorder requires symptoms 1 or 2, plus four of 3–9 for

not cited in the text.

sickness behavior is considered to be a coping strategy that

bestows benefits on the individual, and has evolutionary

advantages (Kent et al., 1992; Hart, 1988; Larson and Dunn, in

press). Although it has been argued that psychiatric states have

adaptive value (Nesse, 2000), this view has been disputed,

and the issue is very controversial (Dubrovsky, 2002;

McLoughlin, 2002). Whether or not there are any benefits of

depression for the suffering patient is questionable.

With regard to the specific symptoms associated with

sickness behavior and with depression, fever is almost

universally observed with pathogen infection and other

pathologies, but hyperthermia is not characteristic of

depression (although there is a trend to a nocturnal

hyperthermia in depressed patients (Dietzel et al., 1986).

Hypomotility is characteristic of sickness behavior in

animals, which is consistent with the psychomotor retar-

dation observed in depressed patients, although depressed

patients are sometimes agitated.

LPS and IL-1 induce a hypersensitivity to pain (Watkins

et al., 1994), but depressed patients are not hypersensitive to

pain and are more likely to suffer hypo-algesia (Adler et al.,

1993).

IL-1, LPS and infections decrease feeding (frequently

referred to as anorexia or aphagia, but truly it is

hypophagia), and consequent loss of body weight (Swiergiel

et al., 1997a). Depressed patients may exhibit either hypo-

or hyperphagia; the hypophagia is characteristic of typical

depression, whereas the hyperphagia is a defining charac-

teristic of atypical depression.

IL-1 and LPS administration to rodents increases sleep

time, specifically an increase in the duration of slow-wave

sleep (Krueger et al., 2003). However, hypersomnia is not

typical of depression; depressed patients are most often

insomniacs, although increased sleep occurs in atypical

depression.

Depressed mood is difficult, if not impossible, to assess in

animals, but may be common to sickness behavior and

imal models of depression

imals

S or IL-1 administration Chronic mild stress (CMS)

determinate Indeterminate

C (Willner, 1997)

pophagia Hypophagia (Forbes et al., 1996)

ow-wave sleep [ Disturbed sleep (Cheeta et al., 1997)

comotor activity Y Locomotor activity ? (Gorka et al., 1996)

comotor activity Y,,,,,,,,,Y

Locomotor activity ? (Gorka et al., 1996)

determinate Indeterminate

determinate Indeterminate

determinate Indeterminate

a period of two weeks or more. References are included only when they are

Table 4

Comparison of physiological parameters in human depression, sickness behavior, and animal models

Physiological parameter Humans Animals

Depression Cytokine therapy LPS or IL-1 injection Chronic mild stress

HPAA (Plasma corticosteroids) C C C Ca

Plasma catecholamines C C C (Dobrakovova et al., 1982)b

Plasma tryptophan K Kc

Brain NE activity C ?d C 0 (Azpiroz et al., 1999)

Brain 5-HT activity C

Brain tryptophan CBrain CRF G G CBNST (Stout et al., 2000)

Plasma IL-1b 0 0 C K (Mormede et al., 2003)

Plasma IL-6 C C C C (Mormede et al., 2003)

DST C C (Froger et al., 2004)

CRF test C

DST: dexamethasone suppression test; BNST: bed nucleus of stria terminalis. References are included only when they are not cited in the text.a Bielajew et al. (2002) and Ayensu et al. (1995) observed increases; Azpiroz et al. (1999) did not.b Chronic immobilization.c Decreased by IL-1, but not IFNa.d Depending on the brain region.

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909896

depression. Table 3 lists the Diagnostic and Statistical

Manual (DSM-IV) criteria for a major depressive episode.

One or more such episodes in the absence of manic symptoms

are necessary for a diagnosis of major depressive disorder.

These symptoms are compared with those observed in

patients treated with IFNa or IL-2, and with sickness

behavior, and chronic stress in rodents. Table 4 compares

the various physiological parameters in the same way.

A key symptom of depression is anhedonia, the inability

to experience pleasure. Needless to say, one cannot

directly assess anhedonia in animals. It has been argued

that the consumption of palatable foods, such as solutions of

sucrose or saccharin, or sweetened milk may reflect hedonia,

so that reductions in these reflects anhedonia, although it is

difficult experimentally to distinguish the hedonic from the

appetitive effects (see Merali et al., 2003). This is confounded

even more when studies are performed in food-deprived

animals. The most popular models use saccharin, which it is

argued appeals to the taste buds, but because it lacks caloric

value, is not considered to reflect feeding (Yirmiya, 1996;

Yirmiya et al., 1999; De La Garza, 2005).

Hedonia has also been assessed using intracranial self-

stimulation (ICSS) paradigms. It is well established that rats

will press a bar to stimulate themselves via electrodes

placed in various parts of the brain, especially in the medial

forebrain bundle (MFB). In a series of studies, Anisman’s

group has shown that peripheral (ip) injection of IL-2

altered self-stimulation via electrodes in the MFB (Anisman

et al., 1996, 1998). Curiously, there was relatively little

effect immediately after the IL-2, only a modest increase in

the threshold at intermediate currents. But over the next

several days, more substantial increases in the ICSS

threshold were observed. These effects were interpreted to

indicate that IL-2 induced a long-lasting anhedonia. No such

effects were observed following administration of IL-1b or

IL-6 (Anisman et al., 1998). Given the widely accepted role

of DA in hedonia, the effect of IL-2 would be consistent

with the reduction in apparent DA release, although the

latter has only been reported after acute injection of IL-2

(Anisman et al., 1996). Similar effects of IL-2 were

observed in mice stimulated in the ventral tegmental area.

In this case icv application of IL-2 (5 ng) elevated

the frequency required to elicit self-stimulation from the

dorsal, but not the ventral A10 region (Hebb et al., 1998).

Taken together these results suggest that IL-2, but not IL-1

and IL-6, can have long-lasting anhedonic effects in rats

and mice.

There has been less research on cognitive behavior.

Certainly, sick animals may show cognitive deficits, for

example, in learning and memory tasks (see examples in

Dantzer et al., 2001; Larson and Dunn, 2001), but several of

the above characteristics of sickness behavior, could explain

the apparent cognitive deficits. Major cognitive deficits are

not often observed in depressed patients, and when they are,

they are generally mild, normally restricted to some

cognitive bias (McKenna et al., 2000). Clinical research

has shown that cytokine administration (IFNa and IL-2) in

cancer patients can affect cognitive functions, with effects

reminiscent of the dysfunctions observed in neurodegen-

erative diseases (Meyers, 1999).

Several observations are not compatible with a simple

model of cytokine-induced depression. For example, IL-1

antagonists, such as the IL-1-receptor antagonist (IL-1ra)

can effectively prevent the effects of IL-1, but the effects of

LPS are only partially prevented, and those of influenza

virus are only slightly attenuated (Swiergiel et al., 1997b).

Also, LPS induces hypophagia in IL-1-knockout mice

(Fantuzzi and Dinarello, 1996) and in IL-6-knockout mice

(Swiergiel and Dunn, 2003). It must be concluded that,

whereas IL-1 is sufficient to induce sickness behaviors, it is

not necessary, and that other factors must contribute to the

behavioral responses. Therefore, the mechanisms involved

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 897

in the expression of sickness behavior are complex and must

involve multiple mechanisms. It is certainly possible that

these mechanisms involve cytokines other than IL-1, but no

cytokine that induces sickness behavior as effectively as

IL-1 has yet been identified.

A popular version of the cytokine hypothesis of

depression proposes a role for cytokines within the brain.

According to this model, immune activation in the periphery

can induce the synthesis or appearance of cytokines and

their receptors in the brain parenchyma. This could be

regarded as a local inflammation within the brain. The major

route by which this is believed to occur involves stimulation

of afferents of the vagus nerve (Watkins et al., 1995; Bluthe

et al., 1996) affecting cells in the brain stem, which in turn

project to the forebrain. The hypothesis states that it is the

cytokines induced within the brain that are responsible for

the depressive symptoms (Licinio and Wong, 1999). The

major cytokine implicated has been IL-1.

This is a novel and an intriguing idea, but more

substantial evidence will be necessary to establish such a

mechanism. The experimental data cited to support this

hypothesis are the induction of the appearance of IL-1 in

various regions of the brain, and the ability of intracere-

brally administered IL-1 antagonists to prevent the effect of

the peripheral treatments. Most of the studies indicating

induction of IL-1 in the brain have involved peripheral

administration of very high doses of LPS. Many of them

have reported only the presence in the brain of mRNA for

IL-1 and/or the IL-1 Type I receptor (Laye et al., 1994;

Wong and Licinio, 1994; Wong et al., 1997), and many of

those have used non-quantitative techniques. Demonstration

of the presence of the mRNA only indicates the propensity

for the synthesis of the proteins, and in the case of IL-1, only

the synthesis of its precursor, pro-IL-1b which must be

cleaved by interleukin-1 converting enzyme (caspase 1) to

produce active IL-1b. Quan et al. (1998) showed that

administration of LPS (2.5 mg/kg ip) to rats induced the

mRNA for IL-1 in the pituitary and the circumventricular

organs (CVO’s), i.e. outside the blood-brain barrier. Small

amounts of mRNA for IL-1b were found in the brain, but

most was present in microglia, thought to be derived from

macrophages that penetrated the endothelia and/or accessed

the brain via the CVO’s (Quan et al., 1998).

Whether or not active IL-1 occurs in the normal brain is

controversial. As indicated above, most of the evidence

indicates that IL-1b appears in the brain after high doses of

LPS and is largely localized in glia. An early immunohis-

tochemical study indicated the presence of IL-1 in post

mortem human brain (Breder et al., 1988), and immuno-

histochemical findings have subsequently been reported

from the rat (Lechan et al., 1990) and the pig (Molenaar

et al., 1993). The anatomical distributions of the IL-1

expression differed somewhat among the individual human

brains in the human study, and there was little concordance

among the anatomical distributions observed in human, rat

and pig. Van Dam et al. (1995) observed immunoreactivity

for IL-1 in microglia in rat brain and sparsely in astroglia

after peripheral administration of a high dose of LPS

(2.5 mg/kg ip or intravenously (iv)). This result is consistent

with those of Quan et al. (1994) using an immunoassay for

IL-1 after a similar dose of LPS.

A more recent study on the brains of multiple sclerosis

patients and controls also found IL-1b-like immunoreactiv-

ity primarily im microglia, especially in the sclerotic plaques

(Huitinga et al., 2000). However, IL-1b-like immunoreac-

tivity was also found in neuronal cell bodies in the

hypothalamus particularly in the paraventricular nucleus,

but also in several other hypothalamic nuclei and in certain

neuronal fibers. The neuronal IL-1b immunoreactivity was

dramatically reduced in the brains of multiple sclerosis

patients. Curiously, all of the IL-1b-positive neurons also

contained oxytocin, although only about one-third of the

oxytocin neurons contained IL-1b. None of the cortico-

tropin-releasing factor (CRF)- or vasopressin-positive

neurons contained IL-1b immunoreactivity. Huitinga et al.

(2000) considered whether the association of IL-1b with

oxytocin neurons could be artifactual refecting another

protein with similar antigenic properties, although it

occurred with three different antibodies to IL-1b. Any

association of IL-1b in oxytocin-containing neurons with

sickness behaviour requires explanation.

Several studies from Maier’s group have used ELISA’s

to demonstrate the presence of IL-1 in the brain, especially

in the hippocampus, following a variety of treatments

including LPS and various stressors (see Watkins et al.,

1995). However, in none of these studies was the purported

IL-1 purified, even partially, prior to the assay, so we cannot

be certain that the apparent immunoreactivity truly reflected

IL-1, especially because the amounts of IL-1 reported were

around the limits of the sensitivity of the assays. However,

in one study, Quan et al., 1996) employed a bioassay to

demonstrate the activity of partially purified IL-1 in the

brains of untreated rats. Classical criteria for the identifi-

cation of hormones and neurotransmitters require more

rigorous techniques, involving extensive purification of the

candidate molecule, and demonstration of its identity using

both immunological and bioassays. However to date, not

even a Western blot has been published to demonstrate the

existence of IL-1 in the brain. Moreover, it cannot be

excluded that the presence of the limited amounts of IL-1 in

these reports reflects local pathologies. The IL-1 may well

have been present in microglia involved in the phagocytosis

of dead cells or cellular components.

There are similar problems with the studies purporting to

demonstrate the presence of IL-1 receptors in the brain. An

early study claimed widespread binding of [125I]IL-1

(mouse) to slices of rat brain (Farrar et al., 1987). Careful

studies by Haour et al. (1992) and Takao et al. (1992) found

that the majority of IL-1 binding in the rat and the mouse

brain was associated with the cells in the endothelia, which

may explain the earlier results of Farrar et al. (1987).

Neither group found any evidence for binding of IL-1 in

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909898

the brain parenchyma of the rat, although both observed

limited binding in the hippocampus of the mouse. It is

possible that there are so few IL-1 receptors in the rat and

mouse brain that they could not be detected by these binding

techniques. It is thought that the presence of a very small

number of IL-1 receptors on a cell may be sufficient to

induce biologic responses (Krakauer and Oppenheimer,

1999). Nevertheless, it is clear that the existence of IL-1

receptors in the brain parenchyma is limited at best. Other

studies have identified mRNA for the IL-1 Type 1 receptor

in the brain (Wong and Licinio, 1994; Yabuuchi et al., 1994;

Ericsson et al., 1995). However, the presence of mRNA is

insufficient to prove the existence of functional receptors,

for which other proteins, such as the IL-1 receptor accessory

protein, are required.

The major evidence cited to indicate a role for cytokines

in the brain is based on the use of intracerebral injections of

IL-1ra (Dantzer et al., 1999; Pugh et al., 1999; Borsody and

Weiss, 2004). Such data are useful and suggestive of a role

for IL-1, but much more evidence will be needed to define

and establish a physiological role for IL-1 in the brain, not

the least in depression.

2.4.1. Effects of antidepressant treatments on sickness

behavior

Depression in humans responds reasonably well to

antidepressant drugs. Antidepressants also have useful

anti-stress effects (Strohle and Holsboer, 2003), but they

are not particularly effective in alleviating the feeling of

being sick. In a creative series of studies, Yirmiya used

ingestion of a saccharin solution by rats as a measure of

hedonia (Yirmiya, 1996). He showed that treatment of rats

with LPS, which is a potent stimulator of the production and

secretion of the pro-inflammatory cytokines, IL-1, IL-6,

TNFa and IFNg, decreased the frequency with which rats

pressed a bar to obtain the saccharin solution. This response

was considered to reflect anhedonia, a cardinal symptom of

depression. This hypothesis was tested by treating the rats

chronically (for 3–5 weeks) with the classic tricyclic

antidepressant, imipramine, which inhibits the reuptake of

both norepinephrine (NE) and serotonin (5-hydroxytrypta-

mine, 5-HT). This treatment prevented the induction by LPS

of the ‘anhedonia’ in the rats. Similar results were obtained

by Shen et al. using the NE-selective reuptake inhibitor,

desmethylimipramine (Shen et al., 1999), and by Yirmiya

using the 5-HT-selective reuptake inhibitor (SSRI), fluox-

etine, although in the latter case, the prevention was less

complete (Yirmiya et al., 2001). However, Shen et al.

observed no such effect of the SSRI, paroxetine, nor of the

5-HT and NE reuptake inhibitor, venlafaxine (Shen et al.,

1999). Thus the effect of antidepressants appears to be less

evident with the SSRI’s than with the tricyclic antidepress-

ants (Yirmiya et al., 1999; Shen et al., 1999). However, the

atypical antidepressant, tianeptine was effective chronically

but not acutely (Castanon et al., 2003). We failed to observe

any effect of chronic imipramine or venlafaxine, on the

inhibitory effect on LPS on the drinking of sweetened milk

in mice (Dunn and Swiergiel, 2001). Yirmiya has also

indicated a failure to observe an effect of antidepressants on

LPS-induced saccharin-drinking in mice (Yirmiya et al.,

1999). From the above cited studies, it appears that the

effect of chronic antidepressant treatment is observed most

often with LPS, is less evident with IL-1, and the effects on

LPS are evident in rats and not in mice. Other evidence

suggests that antidepressants may affect the induction of

IL-1 production by LPS (Yirmiya et al., 1999; Castanon

et al., 2003; Connor et al., 2000), and reflect a peripheral

rather than a central mechanism. A peripheral mechanism of

action for antidepressants is highly unlikely, so that

Yirmiya’s exciting initial observation has failed to provide

strong support for an IL-1 hypothesis of depression.

Another important aspect of sickness behavior is that it

does not appear to be stereotyped but is adaptive. This

means that the specific expression of sickness behaviors can

vary according to the needs of the organism (see Larson and

Dunn, in press) For example, although hypophagia is a

typical symptom of sickness behavior, it may not be

expressed if the food supply is limited, or the animals are

food-deprived (McCarthy et al., 1986; Kent et al., 1994;

Larson et al., 2002). It may also be diminished if there is

only one source of food (Aubert et al., 1997a). Priorities for

food and warmth may also be altered (Aubert et al., 1997b).

Another example is sexual activity, which is inhibited by

IL-1 and LPS in females, but not in males (Avitsur et al.,

1998). This presumably reflects the obvious danger for the

sick gravid female and her offspring, but is not a threat for

the male.

In sum, although there are many similarities between

sickness behavior and the behavior of patients suffering

from major depressive disorder, there are many aspects in

which the sickness behavior model does not resemble the

human syndrome. Furthermore, it has been shown that in

experimental endotoxemia, the fever and corticosterone

responses occur at a time when there are no detectable

increases in circulating endotoxin or cytokines (Campisi

et al., 2003). These discrepancies suggest that cytokines

may not be necessary for the activation of the endocrine and

sickness responses, nor for major depressive disorders.

2.5. HPAA activation by cytokines: the relationship

to depression

The potent activation of the HPAA by IL-1 (Besedovsky

et al., 1986) was a seminal discovery because it extended the

concept of stress to immune activation. In this sense, the

immune system acts a sensor of invading pathogens,

signalling the brain of a threatening environmental event

(Blalock, 1984). It is highly significant that the major

mechanism by which immune stimuli activate the HPAA is

the same as that involved in the responses to psychological

and physical stressors, namely activation of CRF-containing

neurons in the paraventricular nucleus of the hypothalamus,

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 899

and subsequent secretion of anterior pituitary ACTH

(although there is evidence that IL-1 may activate the

HPAA by multiple mechanisms (see Silverman et al., 2003;

Dunn, 2005). Interestingly, there is little evidence for

habituation (desensitization) of the HPAA response to IL-1,

although there is rapid tolerance to the responses to LPS

(Fan and Cook, 2004).

The HPAA-activating effect of IL-1 is shared by other

cytokines, most notably, IL-6 and TNFa, but the latter are

markedly less potent than IL-1, and may have limited

physiological significance, except perhaps in the absence of

IL-1 (Silverman et al., 2003; Dunn, 2005). Treatment

with antibodies to IL-6 and studies in IL-6-knockout mice

indicate that IL-6 contributes only modestly to the elevation

of plasma ACTH and corticosterone by LPS in mice (Wang

and Dunn, 1999; Swiergiel and Dunn, 2003) However,

because plasma concentrations of IL-6 can be dramatically

elevated by infections and other pathologies in which the

HPAA is activated, IL-6 may contribute to this activation

(Silverman et al., 2004) Depending on the physiological

conditions, IL-6 apparently has the ability to activate the

HPAA via the hypothalamus, the pituitary and the adrenal

cortex (Silverman et al., 2003; Silverman et al., 2004; Dunn,

2004; Dunn, 2005).

Administration of IFNa and IFNg, but not IFNb causes

marked activation of the HPAA in man (e.g. Shimizu et al.,

1995; Capuron et al., 2003). However, curiously, IFNgelevates cortisol, but has little effect on ACTH (Holsboer

et al., 1988). There appear to be marked species differences,

because administration of relatively high doses of IFNa(human or mouse) to mice did not alter plasma corticoster-

one (Dunn, 1992, 2005). In rats, both peripheral (ip) and

intracerebroventricular (icv) administration of hIFNainduced decreases in plasma ACTH and corticosterone

(Saphier, 1989; Saphier et al., 1993), although one study

showed modest increases in ACTH and corticosterone

following iv rat IFNa (Menzies et al., 1996).

In depressed patients, activation of the HPAA is perhaps the

most consistent biological marker, yet it occurs in only

50–70% of patients. A similar proportion of depressed patients

are abnormal in the dexamethasone suppression test and the

CRF challenge tests (Strohle and Holsboer, 2003). Smith cited

the ability of IL-1 to activate the HPAA as support for a

cytokine hypothesis of depression (Smith, 1991). It is relevant

that the HPAA activation by IL-1 can be dissociated from its

behavioral effects. Inhibitors of the enzyme cyclooxygenase

(COX) more or less prevent several behavioral responses to

IL-1 (Swiergiel et al., 1997a; Dunn and Swiergiel, 2000), but

the HPAA activation is affected only after iv injection of IL-1,

and in the early phase after ip injections (Dunn and Chuluyan,

1992). The role of CRF also differs in the HPAA and the

behavioral responses. It appears to be involved in the HPAA

responses to IL-1, because antibodies to CRF prevent IL-1-

induced HPAA activation in rats and mice (Dunn, 1993;

Turnbull and Rivier, 1999). Also, HPAA activation is

minuscule in CRF-knockout mice (Muglia et al., 2000;

Dunn and Swiergiel, 1999), while the hypophagic responses

to IL-1 and LPS are normal (Dunn and Swiergiel, 1999).

The principal factors mitigating against a hypothesis of

IL-1-induced HPAA activation in depression is the failure to

observe consistent elevations of plasma concentrations of

IL-1 in depressed patients, and the fact that one third or

more of depressed patients do not exhibit HPAA activation,

or other abnormalities of HPAA function.

2.6. Brain catecholaminergic activation induced

by cytokines: relationship to depression

In mice and rats, central noradrenergic systems are

markedly activated by IL-1 as indicated by classical

neurochemical studies showing increases in the NE

catabolite, 3-methoxy,4-hydroxyphenylethyleneglycol

(MHPG) (Dunn, 1988; Dunn, 2001; Kabiersch et al.,

1988). The noradrenergic activation is not uniform, the

ventral system (ventral noradrenergic ascending bundle,

VNAB) being activated substantially more than the dorsal

system. Consistent with this, reductions in the hypothalamic

content of NE have also been observed in rats (Fleshner

et al., 1995). Microdialysis studies in rats have indicated a

similar activation (Shintani et al., 1993; Smagin et al., 1996;

Wieczorek and Dunn, 2003). There is no evidence for

noradrenergic activation associated with IL-6 (Wang and

Dunn, 1998), although TNFa has such an effect in mice at

high doses (Ando and Dunn, 1999). The latter could be

attributed to TNFa-induced IL-1 secretion.

In contrast to these robust effects of IL-1 on NE, this

cytokine does not consistently affect dopamine (DA)

metabolism, although small increases are occasionally

observed (Dunn, 2001). IL-6 and TNFa have not been

reported to affect DA. However, IL-2 (1 mg ip) induced a

marked reduction in microdialysate concentrations of DA

from the nucleus accumbens of the rat (Anisman et al.,

1996). The latter result is consistent with the increase in

thresholds for ICSS observed in rats following similar

treatments with IL-2, indicative of anhedonia (see above).

Reports of the effects of interferons on brain catechol-

amines have been quite varied (see also Schaefer et al.,

2003). Shuto et al. found that chronic (but not acute)

administration of hIFNa (15!106 Units ip) induced small

decreases in whole brain (minus cerebellum) DA and its

catabolite, 3,4-dihydroxyphenylacetic acid (DOPAC), but

there were no changes in the DOPAC:DA ratio nor in NE

(Shuto et al., 1997). They also reported a decrease in the

apparent turnover of DA, assessed by blocking its synthesis

with a-methyl-p-tyrosine. Kumai et al. reported that

repeated subcutaneous treatment (7 days) with hIFNaincreased the DA and NE contents of the cortex,

hypothalamus and medulla, but not of the hippocampus or

thalamus (Kumai et al., 2000). We observed no effects of a

single ip injection of mouse IFNa (1000 or 10,000 U/

mouse) on DA, NE or any of their metabolites (Dunn, 2001).

On the other hand, a single icv injection of IFNa (200 or

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909900

2000 U of hIFNaA/D) was reported to decrease frontal

cortical NE (Kamata et al., 2000). However, in yet another

study, 1000 U hIFNa injected icv increased apparent DA

turnover (DOPAC:DA ratio) in the hippocampus (primarily

caused by a substantial decrease in DA), although no such

effect was observed in the prefrontal cortex or striatum

(De La Garza and Asnis, 2003).

Many studies have indicated hypersecretion of brain NE

in depression, principally assessed by increased CSF or

urinary concentrations of the NE catabolite, MHPG

(3-methoxy,4-hydroxyphenylethyleneglycol), but also by

direct measurement of CSF NE (Wong et al., 2000). Thus an

IL-1-mediated elevation of cerebral NE activity could be

taken as evidence for a role of IL-1 in depression, although

as in the case of the HPAA activation, the failure to observe

consistent elevations of plasma IL-1 in depressed patients

tempers this evidence.

2.7. Cytokines, depression and serotonin

Currently, the prevailing hypothesis is that depression is

caused by some disorganization of brain serotonergic

systems, perhaps a deficiency in serotonergic neurotrans-

mission. This hypothesis is based largely on the observation

that many effective antidepressants affect serotonergic

transmission. Also, serotonin receptor abnormalities are

found in the brains of depressed patients (for reviews, see

Bell et al., 2001; Coppen and Wood, 1978; Wichers and

Maes, 2004). Antidepressants are presumed to work by

normalizing a defect in brain serotonergic transmission

through plastic changes or neurotrophic effects (Manji et al.,

2001).

An interesting mechanism by which infections might

induce depression, relates to the metabolism of tryptophan

and 5-HT). Infections, especially viral ones, induce IFNg,

which is a potent inducer of indoleamine 2,3-dioxygenase

(IDO) in macrophages and certain other cells. IL-2 and

IFNa have similar effects on IDO, but to a lesser extent.

This enzyme degrades tryptophan, so that infections are

typically associated with decreases in plasma tryptophan.

IDO enables the conversion of tryptophan to kynurenine,

and the production of neopterin, both of which are elevated

in the plasma of infected subjects. The resulting catabolism

of tryptophan decreases its circulating concentrations.

Lower plasma concentrations of tryptophan have been

associated with depression (Coppen and Wood, 1978), but

low plasma tryptophan is not a particularly reliable biologic

marker for depressive illness (Bell et al., 2001). It has been

postulated that such a decrease may limit the availability of

tryptophan to the brain, thus limiting serotonin synthesis,

and precipitating depression (Wichers and Maes, 2004).

Such a mechanism for the induction of depression has not

been adequately demonstrated, but lowering plasma trypto-

phan can precipitate a relapse in depressed patients treated

with SSRI’s (Delgado et al., 1994, 1999), but not in those

treated with desmethylimipramine (whose primary

mechanism of action is thought to be inhibition of

NE uptake) (Delgado et al., 1999) or cognitive therapy

(O’Reardon et al., 2004). It appears that medicated patients

are more likely to experience relapse following tryptophan

depletion than unmedicated patients (Moore et al., 2000).

Tryptophan depletion may also induce depression in

susceptible individuals (Bell et al., 2001). There is also

evidence that individuals that exhibit mood changes in

response to rapid tryptophan depletion may be at risk for

depression (Moreno et al., 2000).

There is some limited support for this mechanism from

animal studies. Repeated tryptophan depletion of rats

increased anxiety-like behavior in the open field and

immobility in the forced swim test (thought to reflect

depression-like behavior, see below), but did not alter

behavior in the Morris water-maze (Blokland et al., 2002).

However, acute tryptophan depletion did not alter the

behavior of rats in the open-field test, the home cage

emergence test, and the forced swim test, but decrements

were observed in an object recognition test (Lieben et al.,

2004).

IL-1, IL-6 and TNFa have been shown to activate brain

serotonergic systems, increasing brain tryptophan concen-

trations and the metabolism of 5-HT as indicated by

increases in the brain concentrations of its catabolite,

5-hydroxyindoleacetic acid (5-HIAA) (Dunn, 2001;

Zalcman et al., 1994; Clement et al., 1997). Apparent

5-HT release by peripherally administered IL-6 has also

been indicated by microdialysis and in vivo chronoampero-

metry (Zhang et al., 2001). The acute effects of IL-1 and

IL-6 appear to increase the available tryptophan in the brain,

perhaps enhancing serotonergic activity, which could be

considered to have an antidepressant effect. The problem for

a hypothesis that cytokines induce depression by activating

serotonin is that such an effect is apparently in the wrong

direction, because SSRI’s also induce increased extracellular

5-HT. This could perhaps be explained if the chronic effects

of the cytokines differed from the acute ones. The effects of

chronic administration of the cytokines on serotonin

metabolism have not been studied extensively, but one

report indicated that repeated TNFa administration increased

its effects on 5-HT (Hayley et al., 2002). IL-1b has also been

reported to activate the serotonin transporter, which could

decrease extracellular 5-HT (Morikawa et al., 1998).

There are few data on the effects of IFN’s on brain 5-HT.

A single icv injection of 200 or 2000 U of hIFNa was

reported to decrease the 5-HT content of the frontal cortex,

and both 5-HT and 5-HIAA were decreased in the midbrain

and the striatum of rats (Kamata et al., 2000). However, we

observed no effects of ip human or mouse IFNa on 5-HT or

5-HIAA in mice at doses (400–16,000 U/mouse) that

induced behavioral changes (Dunn, 2001). However, in

another study a single icv injection of IFNa (1000 U)

increased 5-HIAA:5-HT ratios in the prefrontal cortex, but

not in the striatum or the hippocampus of rats (De La Garza

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 901

and Asnis, 2003). This latter effect appeared to be prevented

by diclofenac pretreatment.

Therefore, there are several hypothetical mechanisms by

which cytokines could act on brain serotonergic systems and

thus contribute to the pathophysiology of depression.

However, relationships between these potential mechanisms

and depression remain to be demonstrated in depressed

patients.

3. Effects of immune activation and cytokines in classical

tests of depression

The cytokine hypothesis of depression would be strength-

ened if immune stimulation or cytokine administration

proved positive in classical models of depression. Generally,

these models are based on the assumption that a particular

behavioral pattern resembles a human depressive symptom,

and/or it is selectively sensitive to clinically effective

antidepressant treatments. The most recent reviews indicate

that there are some half dozen rodent models of depression

(Cryan et al., 2002; Cryan and Mombereau, 2004; Porsolt,

2000). The two tests most commonly used are the forced

swim test (FST), and the tail suspension test (TST).

3.1. The Porsolt or forced swim test

The forced swim test introduced by Porsolt et al. (1977a)

is by far the most frequently used. The test involves placing

a rat or a mouse in a tall cylinder of water from which it

cannot escape, so that it must swim or float to survive. In the

original form of the test, rats were placed in the cylinder of

water on the first day for 15 min. It has been suggested that

this session may be necessary for the rats to learn that escape

is impossible (Borsini and Meli, 1988). The rats are tested

again 24 h after the first trial after potential antidepressant

treatments. In its original form, drugs or other treatments

were administered immediately after the session on the first

day, and then again five hours and one hour before the test

on the second day (Porsolt et al., 1977a). However, Porsolt

subsequently showed that many antidepressant treatments

tested positively with only one or two of the three

treatments, although in many cases the three-treatment

regimen was more effective (Porsolt et al., 1978). The key

parameters affected by antidepressants are the time spent

immobile (floating) and the latency to stop active swimming

(struggling) or attempting to escape. Antidepressants

increase the latency to float and decrease the floating

time; whereas chronic stress and certain other treatments

considered depressogenic increase the time spent floating.

This test has been shown to work for most antidepressant

treatments, including atypical antidepressants, such as the

monoamine oxidase inhibitors (MAOI’s) and electrocon-

vulsive therapy (ECT) (Cryan and Mombereau, 2004;

Porsolt, 2000), although the SSRI’s are less effective and

some are ineffective. Thus the test has been validated

pharmacologically, even though the drug treatments work

after one, two or three treatments, rather than the chronic

treatments required in depressed patients. Porsolt also

developed a test for use in mice, but in this test the mice

are forced to swim only once, usually with the antidepress-

ant treatments applied shortly before the test. Immobiliz-

ation is scored during the last 4 min of the 6-minute test

(Porsolt et al., 1977b). However some researchers have

employed a 2-day test in mice like that in rats.

Surprisingly, there are few data in the literature on the

effects of cytokines or immune stimulation using the FST. In

an early report, del Cerro and Borrell found that IL-1 (5 U)

administered icv reduced floating in rats by more than two-

fold (del Cerro and Borrell, 1990), an apparent antidepress-

ant effect. In direct contradiction to these results, Huang and

Minor reported (in abstract form only) that icv injection of

2 ng of IL-1b significantly increased floating in rats in the

FST, and that this response was prevented by icv

administration of 6 mg of IL-1ra or by ip caffeine, the latter

implicating adenosine receptors (Huang and Minor, 2000).

As discussed above, patients medicated with IFNa may

display depressive symptoms that can be successfully

treated with antidepressants (Musselman et al., 2001a).

Makino et al. reported that recombinant human IFNa(60,000 U/kg iv) increased floating in the FST in rats,

without altering locomotor activity (Makino et al., 2000a).

Human IFNa-2, and IFNa-2b also increased floating in the

FST in mice (Makino et al., 1998). Curiously, this effect was

observed only with human IFNa; hIFNb and hIFNg, and

mouse IFNa, IFNb and IFNg were all ineffective. Pretreat-

ment with naloxone, but not indomethacin prevented the

FST response to hIFNa (Makino et al., 2000b). We have

observed consistent increases in floating in the FST in rats

injected icv with 150 or 1500 U rat IFNa (Dunn and

Swiergiel, unpublished observations). In preliminary results

in mice, we have also observed increased floating times

following icv administration of mIFNa (500–5000 U)

(Dunn and Swiergiel, unpublished observations).

Several other reports pertain indirectly to a role of

cytokines in behavior in the FST. Tannenbaum et al. showed

that chronic stress increased sickness scores and the

duration of floating in the FST, and although they reported

that chronic stress augmented the sickness behavior induced

by ip IL-1b, they did not report the effects of IL-1b in the

FST (Tannenbaum et al., 2002). Deak et al. showed that

exposure to the FST did not alter the concentrations of IL-1

measured in the brain or in peripheral tissues, and concluded

that the production of IL-1 in the brain was unlikely to play

a role in the behavioral consequences of the forced swim

(Deak et al., 2003).

A potential role for IL-6 was demonstrated by Sakic et al.

(2001) who observed that a spontaneous increase in serum

interleukin-6 (IL-6) concentrations in lupus-prone

(MRL-lpr) mice coincided in time with increased immo-

bility in the FST. Also, sucrose intake was decreased when

IL-6 was over-expressed systemically in healthy mice,

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909902

suggesting that sustained elevation of plasma IL-6 might

affect other behaviors. High circulating concentrations of

IL-6, were achieved in mice by infecting them with

Ad5mIL6 adenovirus. This treatment reduced exploration

of a novel object, as well as food, water, and sucrose intake,

that is, resembling changes observed in sickness. Further-

more, in lupus-prone mice the presence of CD45-positive

cells (leukocytes) in the choroid plexus and in the brain

parenchyma correlated positively with floating in the FST

(Farrell et al., 1997). These results suggest that immune

system alterations may affect the behavior in the FST.

However, elevations of IL-6 are commonly associated with

sickness, but whether or not the IL-6 is responsible for the

symptoms has not been established. It is significant that IL-6

administration has not been shown to induce any sickness

behaviors (Bluthe et al., 1998; Swiergiel and Dunn, 1999).

Comparing results derived from sickness behavior studies

and the FST requires careful interpretation. If sickness

behavior is considered adaptive, its functional adaptive

behavioral equivalent in the FST may be an increased floating

(usually interpreted as an increase in behavioral despair), as

opposed to active swimming that could be considered

maladaptive (Nishimura et al., 1986). An appropriate response

to LPS or IL-1 would be to increase flotation, not because it

produces depression, but because it is an appropriate strategy

for survival. Thus the biological significance of the effects of

immune challenges, LPS or cytokines in behavioral patterns

observed in the FST remains unclear.

3.2. Tail suspension test (TST)

The tail suspension test, the second most widely used test

for depression, is conceptually similar to the FST. This test

is most commonly used in mice, because it is very difficult

to implement in rats. The animal is suspended by its tail and

the latency to cease struggling and the duration of passive

immobility are scored. The duration of immobility dis-

played by the suspended animals is reduced by antidepress-

ant treatments (Cryan et al., 2002; Cryan and Mombereau,

2004). The TST has the advantage of eliminating con-

founding effects of thermal effects of the swimming (Cryan

and Mombereau, 2004). Nevertheless, the FST and TST

may differ in the biological substrates underlying the

observed behaviors (Cryan and Mombereau, 2004). Lines

of helpless mice responding to antidepressants, have been

developed by selective breeding for responses in the TST

(Vaugeois et al., 1996).

Relatively few data are available on the effects of immune

activation or cytokines on responses in the TST. Yamano et al.

reported that YM643 (2!108–2!109 U iv), a consensus

IFNa, and sumiferon (2–20!106 U), a natural IFNa,

administered iv, increased immobility time in the TST in

mice (Yamano et al., 2000). Icv and repeated subcutaneous

injections were also effective. Peripheral pretreatment with

imipramine or CP-154,526 (a CRFR1 receptor antagonist), but

not with naloxone or indomethacin, reversed the IFN-induced

immobility and it was concluded that the IFNa induced

depression-like behavior by acting centrally (Yamano et al.,

2000). Yamano et al. suggested that IFNa induced depression

via IL-1b release, which in turn was responsible for the CRF

secretion. In preliminary studies, we have also

observed increases in the time spent immobile in the TST

after icv mIFNa (500–5000 U/mouse) (Dunn and Swiergiel,

unpublished observations).

A summary of the biological effects observed in

depression and in animal models is provided in Table 4.

Even though the data are incomplete, the table clearly

indicates the existence of common features between

depression and sickness behavior, as well as some aspects

of chronic stress in animals.

3.3. Involvement of cytokines in other models

of depression

There are very limited data on the involvement of

cytokines in other models of depression. In the chronic mild

stress (CMS) model for depression, Kubera et al. showed

that 3 weeks of CMS increased the ability of splenocytes to

produce IL-1 and IL-2, an effect that was partially prevented

by chronic treatment with imipramine (Kubera et al., 1996;

Kubera et al., 1998). A decrease in NK cell activity was also

observed (Kubera et al., 1998). Mormede et al. (2003)

observed decreases in body weight after 21 days of CMS,

but these were not accompanied by changes in the

hematocrit or plasma concentrations of corticosterone.

However, the consumption of sucrose was decreased,

consistent with an anhedonic effect. Spleen cells did not

display any changes in the production of IFNg or IL-6.

However, the mRNA’s for IL-1b and IL-6 were decreased

in the liver, while they were increased in the hypothalamus

(Mormede et al., 2003).

A number of immune abnormalities have been observed

in the rat olfactory bulbectomy model (see Leonard and

Song, 2002). In the only published study involving

cytokines, the serum concentrations of IL-1b and TNFawere increased by LPS administration, but this response was

decreased by bulbectomy (Connor et al., 2000). Treatment

with desmethylimipramine decreased the responses in both

bulbectomized and control rats. These changes paralleled

the increases in plasma corticosterone.

3.4. Usefulness of the cytokine-induced sickness behavior

model for the study of depression

The value of animal models of human diseases is to

provide insight into the mechanisms involved in the

diseases, which may suggest potential therapies that can

then be tested in the models. Because such models will

normally involve artificial interventions, it is important to

use them intelligently. Thus if cytokine administration is

used to induce a model for depression, then the introduc-

tion of appropriate cytokine antagonists should prevent

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 903

the symptoms in the model. However, unless cytokines are

the cause (or a cause) of major depressive disorder,

cytokine antagonists are unlikely to be effective clinically.

Moreover, provided they are safe and non-toxic, the

cytokine antagonists could be tested directly in patients,

negating the need for the model. The model could be used

to assess the toxicity of the treatments or the efficacy of

chronic treatment, but such tests would normally employ

standard tests for toxicity and not use the disease model. If

cytokine antagonists are not effective in the clinical

situation, there would be no need to test them in the

animal model, and the model would have little value.

Models are most useful if they are based on the etiology of

the human disease, or to test a proposed underlying

mechanism.

It is unlikely that treatments effective in the model will

work if the method used to induce the model is unrelated to

the basic mechanism(s) of the disease. A example of the

successful use of an animal model is the development of

treatments for attention deficit hyperactivity disorder

(ADHD). When it was observed that methylphenidate (an

amphetamine-like drug that stimulates DA release) para-

doxically reversed the hyperactivity observed in 6-hydroxy-

dopamine-treated rats (Shaywitz et al., 1976), it was

reasoned that it might counter the hyperactivity observed

in children with ADHD. Such treatment has proven very

effective in humans (Shaywitz et al., 2001). In this case, the

model was useful most probably because ADHD in children

involves an abnormality in brain catecholamine systems.

But not all models have proven so useful. For example, the

olfactory bulbectomy model mimics depression in several

tests, and the effects respond appropriately to very many

treatments effective in depressed patients (see Cryan et al.,

2002; Cryan and Mombereau, 2004; Kelly et al., 1997).

Nevertheless, although the model is of considerable

scientific interest, it has failed to provide insight into the

mechanism of depression and has so far failed to inspire

useful clinical therapies.

Needless to say, we do not yet know the value of the

cytokine-induced sickness behavior model of depression,

but we do have some insight into what will probably not

work. For example, there is general agreement that many

behavioral responses to IL-1 are largely prevented by

inhibition of COX, a key enzyme for the production of

prostaglandins and other eicosanoids (Swiergiel et al.,

1997a; Uehara et al., 1989; Bluthe et al., 1992). This

observation led Charlton (2000) (and subsequently others)

to propose that COX inhibitors (the non-steroidal anti-

inflammatory drugs (NSAID’s), such as aspirin, indometha-

cin and ibuprofen) should have antidepressant activity.

However, depressed patients have frequently taken

NSAID’s, but there is no good evidence for any anti-

depressant activity. Reference to the animal data indicates

something of the problem. Whereas COX inhibitors more or

less prevent several sickness behaviors induced by IL-1,

they are much less effective against LPS, and have virtually

no effect against more complex immune challenges, such as

influenza virus infection or tumors (Swiergiel et al., 1997a;

McCarthy and Daun, 1993; Jain et al., 2001).

4. Conclusions

The above review indicates some associations between

the appearance of cytokines and major depressive disorder.

However, no clear causal relationship has been established.

Administration of certain cytokines to humans (especially

IFNa and IL-2) induces symptoms of depression in some

patients, but such responses occur in only a minority of

patients, and many other neuropsychiatric symptoms may

also be induced. Immune activation appears more frequently

in depressed patients than in the general population, but is

not observed in all depressed patients. It is possible that the

immune activation may reflect other medical conditions,

which may themselves induce depression. Alternatively,

depression may ensue when a patient hears a diagnosis with

a poor prognosis. Nevertheless, it is possible that excessive

production of cytokines may induce symptoms of

depression in some patients.

Although there is some evidence that depressed patients

exhibit elevated plasma concentrations of cytokines, such

effects are not observed consistently. Elevations are not

observed in all depressed patients, and when they do occur

they are frequently quite small. IL-1 is the major cytokine

known to induce depression-like symptoms, but the

evidence for elevations of this cytokine in depressed

patients is sparse. The only cytokine consistently elevated

in the plasma of depressed patients is IL-6. Plasma

concentrations of IL-6 are frequently elevated during

infections and other pathological conditions (and following

some forms of stress), Thus there is unlikely to be any

specific relationship between plasma IL-6 and depression.

Moreover, in most studies in which IL-6 was administered

to rats and mice, it failed to induce sickness behavior.

Sickness behavior induced by immune stimulation or IL-1

has many similarities with the behavior of depressed patients.

However, there are important differences, for example, in the

effects on body temperature, in the sensitivity to pain, and in

the specific changes in sleep patterns. Therefore, it may be

premature to consider experimentally induced sickness

behavior as a true model for depressive illness. Nevertheless,

certain aspects of depression may be studied in this way, if

the results are interpreted conservatively. Nevertheless,

experiments in which animals were treated chronically

with antidepressant drugs have failed to provide strong

support for the idea that such treatments work by antagoniz-

ing the actions of cytokines.

The cytokine hypothesis is not in conflict with earlier

hypotheses of depression, such as those involving hyper-

activity of the HPAA, of CRF, or of noradrenergic systems,

because IL-1 activates the HPAA and brain noradrenergic

systems. Also, CRF appears to be involved in the HPAA

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909904

responses to IL-1, because antibodies to CRF prevent IL-1-

induced HPAA activation in rats and mice, and HPA

responses to IL-1 are minuscule in CRF knockout mice. A

cytokine hypothesis could also be consistent with a

serotonin hypothesis of depression, because IL-1, IL-6 and

TNFa have each been shown to affect brain serotonergic

transmission. However, IL-1, IL-6 and TNFa administered

acutely appear to increase 5-HT release, which would

achieve an effect similar to that induced by the many

commonly used antidepressants that inhibit 5-HT re-uptake,

so that these cytokines ought to have antidepressant effects.

Although drugs that inhibit serotonin reuptake are the

most useful currently available for the treatment of

depression, there is no strong direct evidence that

abnormalities in 5-HT cause depression. More research on

the effects of antidepressants on the immune system would

be useful. In particular, the associations of abnormalities in

tryptophan and serotonin with pathogen infection, with

immune activation, and with depression, are potentially

very significant, and require further investigation.

These observations do not exclude a role for cytokines in

inducing depression. It is certainly possible that increased

cytokine production may induce depression in some patients,

and certain cytokines may contribute to a variety of

neuropsychiatric symptoms in patients with a variety of

diseases. Nevertheless, one can conclude with some

confidence that the actions of cytokines cannot account for

all cases of depression. It is also possible that by activating

the HPAA axis, and central CRF, noradrenergic and

serotonergic mechanisms, cytokines may complement (or

even synergize with) other factors that induce depression.

Acknowledgements

The authors’ research included or referred to in this

article was supported by the US National Institutes of

Health (MH46261 and NS35370), and by the EU Frame-

work 6 NEWMOOD Integrated Project and the Polish

Committee for Scientific Research, (KBN) Grant 3 PO

4C05825. We thank Christina A. Meyers, PhD (University

of Texas MD Anderson Cancer Center) for suggestions for

improving the manuscript.

References

Adler, G., Gattaz, W.F., 1993. Pain perception threshold in major

depression. Biol. Psychiat. 34, 687–689.

Ando, T., Dunn, A.J., 1999. Mouse tumor necrosis factor-a increases brain

tryptophan concentrations and norepinephrine metabolism while

activating the HPA axis in mice. Neuroimmunomodulation 6, 319–329.

Anisman, H., Kokkinidis, L., Merali, Z., 1996. Interleukin-2 decreases

accumbal dopamine efflux and responding for rewarding lateral

hypothalamic stimulation. Brain Res. 731, 1–11.

Anisman, H., Kokkinidis, L., Borowski, T., Merali, Z., 1998. Differential

effects of interleukin (IL)-1b, IL-2 and IL-6 on responding for

rewarding lateral hypothalamic stimulation. Brain Res. 779, 177–187.

Anisman, H., Ravindran, A.V., Griffiths, J., Merali, Z., 1999. Endocrine and

cytokine correlates of major depression and dysthymia with typical or

atypical features. Mol. Psychiatr. 4, 182–188.

Aubert, A., Kelley, K.W., Dantzer, R., 1997a. Differential effects of

lipopolysaccharide on food hoarding behavior and food consumption in

rats. Brain Behav. Immun. 11, 229–238.

Aubert, A., Goodall, G., Dantzer, R., Gheusi, G., 1997b. Differential effects

of lipopolysaccharide on pup retrieving and nest building in lactating

mice. Brain Behav. Immun. 11, 107–118.

Avitsur, R., Pollak, Y., Weidenfeld, J., Yirmiya, R., 1998. Gender

differences in the behavioral response to interleukin-1: neurochemical

mechanisms. Neuroimmunomodulation 5, 63.

Ayensu, W.K., Pucilowski, O., Mason, G.A., Overstreet, D.H.,

Rezvani, A.H., Janowsky, D.S., 1995. Effects of chronic mild stress

on serum complement activity, saccharin preference, and corticosterone

levels in Flinders lines of rats. Physiol. Behav. 57, 165–169.

Azpiroz, A., Fano, E., Garmendia, L., Arregi, A., Cacho, R., Beitia, G.,

Brain, P.F., 1999. Effects of chronic mild stress (CMS) and imipramine

administration, on spleen mononuclear cell proliferative response,

serum corticosterone level and brain norepinephrine content in male

mice. Psychoneuroendocrinology 24, 345–361.

Bell, C., Abrams, J., Nutt, D., 2001. Tryptophan depletion and its

implications for psychiatry. Br. J. Psychiatr. 178, 399–405.

Besedovsky, H.O., del Rey, A., Sorkin, E., Dinarello, C.A., 1986.

Immunoregulatory feedback between interleukin-1 and glucocorticoid

hormones. Science 233, 652–654.

Bielajew, C., Konkle, A.T., Merali, Z., 2002. The effects of chronic mild

stress on male Sprague–Dawley and Long Evans rats: I. Biochemical

and physiological analyses. Behav. Brain Res. 136, 583–592.

Blalock, J.E., 1984. The immune system as a sensory organ. J. Immunol.

132, 1067–1070.

Blokland, A., Lieben, C.K.J., Deutz, N.E.P., 2002. Anxiogenic and

depressive-like effects, but no cognitive deficits, after repeated

moderate tryptophan depletion in the rat. J. Psychopharmacol. 16,

39–49.

Bluthe, R.M., Crestani, F., Kelley, K.W., Dantzer, R., 1992. Mechanisms of

the behavioral effects of interleukin 1. Role of prostaglandins and CRF.

Ann. NY Acad. Sci. 650, 268–275.

Bluthe, R-M., Michaud, B., Kelley, K.W., Dantzer, R., 1996. Vagotomy

attenuates behavioural effects of interleukin-1 injected peripherally but

not centrally. NeuroReport 7, 1485–1488.

Bluthe, R.M., Michaud, B., Poli, V., Bernay, F., Parnet, P., Dantzer, R.,

1998. Interleukin-6 is active only in presence of other proinflammatory

cytokines to induce sickness behavior. Neuroimmunomodulation 5, 67.

Borsini, F., Meli, A., 1988. Is the forced swimming test a suitable model for

revealing antidepressant activity? Psychopharmacology 94, 147–160.

Borsody, M.K., Weiss, J.M., 2004. The effects of endogenous interleukin-1

bioactivity on locus coeruleus neurons in response to bacterial and viral

substances. Brain Res. 1007, 39–56.

Brambilla, F., Maggioni, M., 1998. Blood levels of cytokines in elderly

patients with major depressive disorder. Acta Psychiatr. Scand. 97,

309–313.

Breder, C.D., Dinarello, C.A., Saper, C.B., 1988. Interleukin-1 immuno-

reactive innervation of the human hypothalamus. Science 240, 321–

324.

Campisi, J., Hansen, M.K., O’Conner, K.A., Biedenkapp, J.C.,

Watkins, L.R., Maier, S.F., Fleshner, M., 2003. Circulating cytokines

and endotoxin are not necessary for the activation of the sickness or

corticosterone response produced by peripheral E. coli challenge.

J. Appl. Physiol. 95, 1873–1882.

Capuron, L., Gumnick, J.F., Musselman, D.L., Lawson, D.H.,

Reemsnyder, A., Nemeroff, C.B., Miller, A.H., 2002. Neurobehavioral

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 905

effects of interferon-a in cancer patients: phenomenology and

paroxetine responsiveness of symptom dimensions. Neuropsychophar-

macology 26, 643–652.

Capuron, L., Raison, C.L., Musselman, D.L., Lawson, D.H.,

Nemeroff, C.B., Miller, A.H., 2003. Association of exaggerated HPA

axis response to the initial injection of interferon-alpha with develop-

ment of depression during interferon-alpha therapy. Am. J. Psychiatr.

160, 1342–1345.

Castanon, N., Konsman, J.-P., Medina, C., Chauvet, N., Dantzer, R., 2003.

Chronic treatment with the antidepressant tianeptine attenuates

lipopolysaccharide-induced Fos expression in the rat paraventricular

nucleus and HPA axis activation. Psychoneuroendocrinology 28, 19–

34.

Charlton, B.G., 2000. The malaise theory of depression: major depressive

disorder is sickness behavior and antidepressants are analgesic. Med.

Hypotheses 54, 126–130.

Cheeta, S., Ruigt, G., van Proosdij, J., Willner, P., 1997. Changes in sleep

architecture following chronic mild stress. Biol. Psychiatr. 41, 419–427.

Cleeland, C.S., Bennett, G.J., Dantzer, R., Dougherty, P.M., Dunn, A.J.,

Meyers, C.A., Miller, A.H., Payne, R., Reuben, J.M., Wang, X.S.,

Lee, B.-N., 2003. Are the symptoms of cancer and cancer treatment due

to a shared biologic mechanism? A cytokine-immunologic model of

cancer symptoms. Cancer 97, 2919–2925.

Clement, H.W., Buschmann, J., Rex, S., Grote, C., Opper, C., Gemsa, D.,

Wesemann, W., 1997. Effects of interferon-g, interleukin-1b, and

tumor necrosis factor-a on the serotonin metabolism in the nucleus

raphe dorsalis of the rat. J. Neural. Transm. 104, 981–991.

Connor, T.J., Harkin, A., Kelly, J.P., Leonard, B.E., 2000. Olfactory

bulbectomy provokes a suppression of interleukin-1b and tumour

necrosis factor-a production in response to an in vivo challenge with

lipopolysaccharide: effect of chronic desipramine treatment. Neuroim-

munomodulation 7, 27–35.

Coppen, A., Wood, K., 1978. Tryptophan and depressive illness. Psychol.

Med. 8, 49–57.

Cryan, J.F., Mombereau, C., 2004. In search of a depressed mouse: utility of

models for studying depression-related behavior in genetically modified

mice. Mol. Psychiatr. 9, 1050–1062.

Cryan, J.F., Markou, A., Lucki, I., 2002. Assessing antidepressant activity

in rodents: recent developments and future needs. Trends Pharmacol.

Sci. 23, 238–245.

Dantzer, R., Aubert, A., Bluthe, R.-M., Gheusi, G., Cremona, S., Laye, S.,

Konsman, J.-P., Parnet, P., Kelley, K.W., 1999. Mechanisms

of the behavioural effects of cytokines. Adv. Exp. Med. Biol. 461,

83–105.

Dantzer, R., Bluthe, R-M., Castanon, N., Chauvet, N., Capuron, L.,

Goodall, G., Kelley, K.W., Konsman, J.-P., Laye, S., Parnet, P.,

Pousset, F., 2001. Cytokine effects on behavior. In: Ader, R., Felten, D.,

Cohen, N. (Eds.), Psychoneuroimmunology. Academic Press, San

Diego, CA, pp. 703–727.

Deak, T., Bellamy, C., D’Agostino, L.G., 2003. Exposure to forced swim

stress does not alter central production of IL-1. Brain Res. 972, 53–63.

de Beaurepaire R, Swiergiel AH, Dunn AJ., 2005. Neuroimmune

mediators: are cytokines mediators of depression? In: Licinio J,

Wong M-L, editors. Biology of Depression. Vol. 2, Weinheim:

Wiley, pp. 557–581.

De La Garza R., 2005. Endotoxin or pro-inflammatory cytokine-induced

sickness behavior as an animal model of depression: Focus on anhedonia.

Neurosci. Behav. Rev. 29, 761-770

De La Garza, R., Asnis, G.M., 2003. The non-steroidal anti-inflammatory

drug diclofenac sodium attenuates IFN-a induced alterations to

monoamine turnover in prefrontal cortex and hippocampus. Brain

Res. 977, 70–79.

del Cerro, S., Borrell, J., 1990. Interleukin-1 affects the behavioral despair

response in rats by an indirect mechanism which requires endogenous

CRF. Brain Res. 528, 162–164.

Delgado, P.L., Price, L.H., Miller, H.L., Salomon, R.M., Aghajanian, G.K.,

Heninger, G.R., Charney, D.S., 1994. Serotonin and the neurobiology

of depression- effects of tryptophan depletion in drug-free depressed

patients. Arch. Gen. Psychiatr. 51, 865–874.

Delgado, P.L., Miller, H.L., Salomon, R.M., Licinio, J., Krystal, J.H.,

Moreno, F.A., Heninger, G.R., Charney, D.S., 1999. Tryptophan-

depletion challenge in depressed patients treated with desipramine or

fluoxetine: implications for the role of serotonin in the mechanism of

antidepressant action. Biol. Psychiatr. 46, 212–220.

Dickens, C., McGowan, L., Clark-Carter, D., Creed, F., 2002. Depression in

rheumatoid arthritis, a systematic review of the literature with meta-

analysis. Psychosom. Med. 64, 52–60.

Dietzel, M., Saletu, B., Lesch, O.M., Sieghart, W., Schjerve, M., 1986.

Light treatment in depressive illness: polysomnographic,

psychometric, and polyendocrinal findings. Eur. Neurol. 25 (Suppl 2),

93–103.

Dobrakovova, M., Kvetnansky, R., Oprsalova, Z., Macho, L., 1982. The

effect of chronic stress on the activity of the sympathetic-adrenome-

dullary system. Bratisl. Lek. Listy 91, 587–592.

Dubrovsky, B., 2002. Evolutionary psychiatry. Adaptationist and non-

adaptationist conceptualizations. Prog. Neuropsychopharmacol. Biol.

Psychiatry 26, 1–19.

Duch, D.S., Woolf, J.H., Nichol, C.A., Davidson, J.R., Garbutt, J.C., 1984.

Urinary excretion of biopterin and neopterin in psychiatric disorders.

Psychiatr. Res. 11, 83–89.

Dunbar, P.R., Hill, J., Neale, T.J., Mellsop, G.W., 1992. Neopterin

measurement provides evidence of altered cell-mediated immunity in

patients with depression, but not with schizophrenia. Psychol. Med. 22,

1051–1057.

Dunn, A.J., 1988. Systemic interleukin-1 administration stimulates

hypothalamic norepinephrine metabolism parallelling the increased

plasma corticosterone. Life Sci. 43, 429–435.

Dunn, A.J., 1992. The role of interleukin-1 and tumor necrosis factor a in

the neurochemical and neuroendocrine responses to endotoxin. Brain

Res. Bull. 29, 807–812.

Dunn, A.J., 1993. Role of cytokines in infection-induced stress. Ann. NY

Acad. Sci. 697, 189–202.

Dunn, A.J., 2001. Effects of cytokines and infections on brain neurochem-

istry. In: Ader, R., Felten, D.L., Cohen, N. (Eds.), Psychoneuroimmu-

nology. Academic Press, New York, pp. 649–666.

Dunn AJ., 2005. Cytokine activation of the hypothalamo–pituitary–

adrenal axis. In: Steckler T, Kalin N, editors. Handbook on Stress,

Immunology and Behaviour. New York: Elsevier. Amsterdam,

pp. 157–174.

Dunn A.J., 2004. IL-6 access to the axis. Endocrinology. 145, 3578–3579.

Dunn, A.J., Chuluyan, H., 1992. The role of cyclo-oxygenase and

lipoxygenase in the interleukin-1-induced activation of the HPA axis:

dependence on the route of injection. Life Sci. 51, 219–225.

Dunn, A.J., Swiergiel, A.H., 1999. Behavioral responses to stress are intact

in CRF-deficient mice. Brain Res. 845, 14–20.

Dunn, A.J., Swiergiel, A.H., 2000. The role of cyclooxygenases in

endotoxin- and interleukin-1-induced hypophagia. Brain Behav.

Immun. 14, 141–152.

Dunn, A.J., Swiergiel, A.H., 2001. The reductions in sweetened milk

intake induced by interleukin-1 and endotoxin are not prevented

by chronic antidepressant treatment. Neuroimmunomodulation 9,

163–169.

Dunn AJ, Swiergiel AH. Unpublished observations.

Ericsson, A., Liu, C., Hart, R.P., Sawchenko, P.E., 1995. Type 1

interleukin-1 receptor in the rat brain: distribution, regulation, and

relationship to sites of IL-1-induced cellular activation. J. Comp.

Neurol. 361, 681–698.

Fan, H., Cook, J.A., 2004. Molecular mechanisms of endotoxin tolerance.

J. Endotoxin Res. 10, 71–84.

Fantuzzi, G., Dinarello, C.A., 1996. The inflammatory response in

interleukin-1b-deficient mice: comparison with other cytokine-related

knock-out mice. J. Leukocyte Biol. 59, 489–493.

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909906

Farrar, W.L., Kilian, P.L., Ruff, M.R., Hill, J.M., Pert, C.B., 1987.

Visualization and characterization of interleukin 1 receptor in brain.

J. Immunol. 139, 459–463.

Farrell, M., Sakic, B., Szechtman, H., Denburg, J.A., 1997. Effect of

cyclophosphamide on leukocytic infiltration in the brain of MRL/lpr

mice. Lupus 6, 268–274.

Fleshner, M., Goehler, L.E., Hermann, J., Relton, J.K., Maier, S.F.,

Watkins, L.R., 1995. Interleukin-1b induced corticosterone elevation

and hypothalamic NE depletion is vagally mediated. Brain Res. Bull.

37, 605–610.

Foley, F.W., Traugott, U., LaRocca, N.G., Smith, C.R., Perlman, K.R.,

Caruso, L.S., Scheinberg, L.C., 1992. A prospective study of depression

and immune dysregulation in multiple sclerosis. Arch. Neurol. 49,

238–244.

Forbes, N.F., Stewart, C.A., Matthews, K., Reid, I.C., 1996. Chronic mild

stress and sucrose consumption: validity as a model of depression.

Physiol. Behav. 60, 1481–1484.

Froger, N., Palazzo, E., Boni, C., Hanoun, N., Saurini, F., Joubert, C.,

Dutriez-Casteloot, I., Enache, M., Maccari, S., Barden, N., Cohen-

Salmon, C., Hamon, M., Lanfumey, L., 2004. Neurochemical and

behavioral alterations in glucocorticoid receptor-impaired transgenic

mice after chronic mild stress. J. Neurosci. 24, 2787–2796.

Gohier, B., Goeb, J.L., Rannou-Dubas, K., Fouchard, I., Cales, P.,

Garre, J.B., 2003. Hepatitis C, alpha interferon, anxiety and depression

disorders: a prospective study of 71 patients. World J. Biol. Psychiatr. 4,

115–118.

Gorka, Z., Moryl, E., Papp, M., 1996. Effect of chronic mild stress on

circadian rhythms in the locomotor activity in rats. Pharmacol.

Biochem. Behav. 54, 229–234.

Griffiths, J., Ravindran, A.V., Merali, Z., Anisman, H., 1997. Neuroendo-

crine measures and lymphocyte subsets in depressive illness: influence

of a clinical interview concerning life experiences. Psychoneuroendo-

crinology 22, 225–236.

Haack, M., Hinze-Selch, D., Fenzel, T., Kraus, T., Kuhn, M., Schuld, A.,

Pollmacher, T., 1999. Plasma levels of cytokines and soluble cytokine

receptors in psychiatric patients upon hospital admission: effects of

confounding factors and diagnosis. J. Psychiatr. Res. 33, 407–418.

Haour, F., Ban, E., Marquette, C., Milon, G., Fillion, G., 1992. Brain

Interleukin-1 receptors: mapping, characterization and modulation. In:

Rothwell, N.J., Dantzer, R.D. (Eds.), Interleukin-1 in The Brain.

Pergamon Press, Oxford, pp. 13–25.

Hart, B.L., 1988. Biological basis of the behavior of sick animals. Neurosci.

Biobehav. Rev. 12, 123–137.

Hayley, S., Wall, P., Anisman, H., 2002. Sensitization to the neuroendo-

crine, central monoamine and behavioural effects of murine tumor

necrosis factor-a: peripheral and central mechanisms. Eur. J. Neurosci.

15, 1061–1076.

Hebb, A.L.O., Zacharko, R.M., Anisman, H., 1998. Self-stimulation from

the mesencephalon following intraventricular interleukin-2 adminis-

tration. Brain Res. Bull. 45, 549–556.

Holsboer, F., Stalla, G.K., von Bardeleben, U., Hammann, K., Muller, H.,

Muller, O.A., 1988. Acute adrenocortical stimulation by recombinant

gamma interferon in human controls. Life Sci. 42, 1–5.

Huang, Q., Minor, T.R., 2000. Adenosine mediates interleukin-1beta

induced behavioral depression in rats. Brain Behav. Immun. 14, 101.

Huitinga, I., van der Cammen, M., Salm, L., Erkut, Z., van Dam, A.-M.,

Tilders, F., Swaab, D., 2000. IL-1b immunoreactive neurons in the

human hypothalamus: Reduced numbers in multiple sclerosis.

J. Neuroimmunol. 107, 8–20.

Jain, N.K., Kulkarni, S.K., Singh, A., 2001. Lipopolysaccharide-mediated

immobility in mice: reversal by cyclooxygenase enzyme inhibitors.

Methods Find. Exp. Clin. Pharmacol. 23, 441–444.

Kabiersch, A., del Rey, A., Honegger, C.G., Besedovsky, H.O., 1988.

Interleukin-1 induces changes in norepinephrine metabolism in the rat

brain. Brain Behav. Immun. 2, 267–274.

Kamata, M., Higuchi, H., Yoshimoto, M., Yoshida, K., Shimizu, T., 2000.

Effect of single intracerebroventricular injection of a-interferon on

monoamine concentrations in the rat brain. Eur. Neuropsychopharma-

col. 10, 129–132.

Kelly, J.P., Wrynn, A.S., Leonard, B.E., 1997. The olfactory bulbectomized

rat as a model of depression: an update. Pharmacol. Therap. 74,

299–316.

Kent, S., Bluthe, R-M., Kelley, K.W., Dantzer, R., 1992. Sickness behavior

as a new target for drug development. Trends Pharmacol. Sci. 13,

24–28.

Kent, S., Rodriguez, F., Kelley, K.W., Dantzer, R., 1994. Reduction in food

and water intake induced by microinjection of interleukin-1b in the

ventromedial hypothalamus of the rat. Physiol. Behav. 56, 1031–1036.

Krakauer, J.V., Oppenheimer, J.J., 1999. Pro-inflammatory cytokines. TNF

and IL-1 families, chemokines, TGF-beta and others. In: Paul, W.E.

(Ed.), Fundamental Immunology. Lippincott-Raven, Philadelphia, PA,

pp. 775–811.

Kronfol, Z., 2002. Immune dysregulation in major depression: a critical

review of existing evidence. Int. J. Neuropsychopharmacol. 5, 333–343.

Kronfol, Z., House, J.D., 1989. Lymphocyte mitogenesis, immunoglobulin

and complement levels in depressed patients and normal controls. Acta

Psychiatr. Scand. 80, 142–147.

Krueger, J.M., Majde, J.A., Obal, F., 2003. Sleep in host defense. Brain

Behav. Immun. 17, S41–S47.

Kubera, M., Symbirtsev, A., Basta-Kaim, A., Borycz, J., Roman, A.,

Papp, M., Claesson, M., 1996. Effect of chronic treatment with

imipramine on interleukin-1 and interleukin-2 production by spleno-

cytes obtained from rats subjected to a chronic mild stress model of

depression. Pol. J. Pharmacol. 48, 503–506.

Kubera, M., Basta-Kaim, A., Holan, V., Simbirtsev, A., Roman, A.,

Pigareva, N., Prokopieva, E., Sham, J., 1998. The effect of mild chronic

stress, as a model of depression, on the immunoreactivity of C57BL/6J

mice. Int. J. Immunopharmacol. 20, 781–789.

Kumai, T.T.T., Tanaka, M., Watanabe, M., Shimizu, H., Kobayashi, S.,

2000. Effect of interferon-alpha on tyrosine hydroxylase and catechol-

amine levels in the brain of rats. Life Sci. 67, 663–669.

Landmann, R., Schaub, B., Link, S., Wacker, H.R., 1997. Unaltered

monocyte function in patients with major depression before and after

three months of antidepressive therapy. Biol. Psychiatr. 15, 675–681.

Larson SJ, Dunn, AJ., in press. Behavioral mechanisms for defense against

pathogens.In: Berczi I, Sventivanyi A, editors. Natural immunity. New

York: Elsevier.

Larson, S.J., Dunn, A.J., 2001. Behavioral effects of cytokines. Brain

Behav. Immun. 15, 371–387.

Larson, S.J., Romanoff, R.L., Dunn, A.J., Glowa, J.R., 2002. Effects of

interleukin-1b on food-maintained behavior in the mouse. Brain Behav.

Immun. 16, 398–410.

Laye, S., Parnet, P., Goujon, E., Dantzer, R., 1994. Peripheral adminis-

tration of lipopolysaccharide induces the expression of cytokine

transcripts in the brain and pituitary of mice. Mol. Brain Res. 27,

157–162.

Lechan, R.M., Toni, R., Clark, B.D., Cannon, J.G., Shaw, A.R.,

Dinarello, C.A., Reichlin, S., 1990. Immunoreactive interleukin-1B

localization in the rat forebrain. Brain Res. 514, 135–140.

Leonard, B.E., Song, C., 2002. Changes in the immune system in rodent

models of depression. Int. J. Neuropsychopharmacol. 5, 345–356.

Levine, J., Barak, Y., Chengappa, K.N., Rapoport, A., Rebey, M.,

Barak, V., 1999. Cerebrospinal cytokine levels in patients with acute

depression. Neuropsychobiology 40, 171–176.

Licinio, J., Wong, M-L., 1999. The role of inflammatory mediators in the

biology of major depression: central nervous system cytokines

modulate the biological substrate of depressive symptoms, regulate

stress-responsive systems, and contribute to neurotoxicity and neuro-

protection. Mol. Psychiatr. 4, 317–327.

Lieb, J., Karmali, R., Horrobin, D., 1983. Elevated levels of prostaglandin

E2 and thromboxane B2 in depression. Prostaglandins Leukotriene

Med. 10, 361–367.

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 907

Lieben, C.K.J., van Oorsouw, K., Deutz, N.E.P., Blokland, A., 2004. Acute

tryptophan depletion induced by a gelatin-based mixture impairs object

memory but not affective behavior and spatial learning in the rat.

Behav. Brain Res. 151, 53–64.

Maes, M., 1995. Evidence for an immune response in major depression: a

review and hypothesis. Prog. Neuropsychopharmacol. Biol. Psychiatr.

19, 11–38.

Maes, M., 1999. Major depression and activation of the inflammatory

response system. In: Dantzer, R., Wollman, E.E., Yirmiya, R. (Eds.),

Cytokines, Stress, and Depression. Kluwer Academic/Plenum Publish-

ers, New York, pp. 25–46.

Maes, M., Lambrechts, J., Bosmans, E., Jacobs, J., Suy, E., Vandervorst, C.,

de Jonckheere, C., Minner, B., Raus, J., 1992. Evidence for a systemic

immune activation during depression, results of leukocyte enumeration

by flow cytometry in conjunction with monoclonal antibody staining.

Psychol. Med. 22, 45–53.

Maes, M., Bosmans, E., Meltzer, H.Y., Scharpe, S., Suy, E., 1993a.

Interleukin-1b: a putative mediator of HPA axis hyperactivity in major

depression? Am. J. Psychiatr. 150, 1189–1193.

Maes, M., Scharpe, S., Meltzer, H., Bosmans, E., Suy, E., Calabrese, J.,

Cosyns, P., 1993b. Relationships between interleukin-6 activity, acute

phase proteins, and function of the hypothalamic–pituitary–adrenal axis

in severe depression. Psychiatr. Res. 49, 11–27.

Maes, M., Scharpe, S., Meltzer, H.Y., Okayli, G., Bosmans, E.,

D’Hondt, P., Vanden Bossche, B.V., Cosyns, P., 1994. Increased

neopterin and interferon-gamma secretion and lower availability of

L-tryptophan in major depression, further evidence for an immune

response. Psychiatr. Res. 54, 143–160.

Makino, M., Kitano, Y., Hirohashi, M., Takasuna, K., 1998. Enhancement

of immobility in mouse forced swimming test by treatment with human

interferon. Eur. J. Pharmacol. 356, 1–7.

Makino, M., Kitano, Y., Komiyama, C., Takasuna, K., 2000a. Human

interferon-a increases immobility in the forced swimming test in rats.

Psychopharmacology 148, 106–110.

Makino, M., Kitano, Y., Komiyama, C., Hirohashi, M., Kohno, M.,

Moriyama, M., Takasuna, K., 2000b. Human interferon-a induces

immobility in the mouse forced swimming test: involvement of the

opioid system. Brain Res. 852, 482–484.

Manji, H.K., Drevets, W.C., Charney, D.S., 2001. The cellular neurobiol-

ogy of depression. Nat. Med. 7, 541–547.

Marshall, P.S., 1993. Allergy and depression, a neurochemical threshold

model of the relation between the illnesses. Psychol. Bull. 113, 23–43.

McCarthy, D.O., Daun, J.M., 1993. The effects of cyclooxygenase

inhibitors on tumor-induced anorexia in rats. Cancer 71, 486–492.

McCarthy, D.O., Kluger, M.J., Vander, A.J., 1986. Effect of centrally

administered interleukin-1 and endotoxin on food intake of fasted rats.

Physiol. Behav. 36, 745–749.

McKenna, P.J., McKay, A.P., Laws, K., 2000. Memory in functional

psychosis. In: Berrios, G.E., Hodges, J.R. (Eds.), Memory Disorders in

Psychiatric Practice. Cambridge University Press, Cambridge.

McLoughlin, G., 2002. Is depression normal in human beings? A critique of

the evolutionary perspective. Int. J. Ment. Health Nurs. 26, 170–173.

Menzies, R.A., Phelps, C.P., Wiranowska, M., Oliver, J., Chen, L.T.,

Horvath, E., Hall, N.R.S., 1996. The effect of interferon-alpha on the

pituitary-adrenal axis. J. Interferon Cytokine Res. 16, 619–629.

Merali, Z., Brennan, K., Brau, P., Anisman, H., 2003. Dissociating anorexia

and anhedonia elicited by interleukin-1b: antidepressant and gender

effects on responding for ‘free chow’ and ‘earned’ sucrose intake.

Psychopharmacology 165, 413–418.

Meyers, C.A., 1999. Mood and cognitive disorders in cancer patients

receiving cytokine therapy. Adv. Exp. Med. Biol. 461, 75–81.

Minden, S.L., Schiffer, R.B., 1990. Affective disorders in multiple

sclerosis. Review and recommendations for clinical research. Arch.

Neurol. 47, 98–104.

Miyaoka, H., Otsubo, T., Kamijima, K., Ishii, M., Onuki, M., Mitamura, K.,

1999. Depression from interferon therapy in patients with hepatitis C.

Am. J. Psychiatr. 156, 1120.

Molenaar, G.J., Berkenbosch, F., van Dam, A.-M., Lugard, C.M.J.E., 1993.

Distribution of interleukin 1b immunoreactivity within the porcine

hypothalamus. Brain Res. 608, 169–174.

Moore, P., Landolt, H.-P., Seifritz, E., Clark, C., Bhatti, T., Kelsoe, J.,

Rapaport, M., Gillin, J.C., 2000. Clinical and physiological con-

sequences of rapid tryptophan depletion. Neuropsychopharmacology

23, 601–622.

Moreno, F.A., Heninger, G.R., McGahuey, C.A., Delgado, P.L., 2000.

Tryptophan depletion and risk of depression relapse: a prospective

study of tryptophan depletion as a potential predictor of depressive

episodes. Biol. Psychiatr. 48, 327–329.

Morikawa, O., Sakai, N., Obara, H., Saito, N., 1998. Effects of interferon-a,

interferon-g and cAMP on the transcriptional regulation of the

serotonin transporter. Eur. J. Pharmacol. 349, 317–324.

Mormede, C., Castanon, N., Medina, C., Moze, E., Lestage, J., Neveu, P.J.,

Dantzer, R., 2003. Chronic mild stress in mice decreases peripheral

cytokine and increases central cytokine expression independently of IL-

10 regulation of the cytokine network. Neuroimmunomodulation 10,

359–366.

Muglia, L.J., Bethin, K.E., Jacobson, L., Vogt, S.K., Majzoub, J.A., 2000.

Pituitary-adrenal axis regulation in CRH-deficient mice. Endocr. Res.

26, 1057–1066.

Musselman, D.L., Lawson, D.H., Gumnick, J.F., Manatunga, A.K.,

Penna, S., Goodkin, R.S., Greiner, K., Nemeroff, C.B., Miller, A.H.,

2001a. Paroxetine for the prevention of depression induced by high-

dose interferon alfa. New Engl. J. Med. 344, 961–966.

Musselman, D.L., Miller, A.H., Porter, M.R., Manatunga, A., Gao, F.,

Penna, S., 2001b. Higher than normal plasma interleukin-6 concen-

trations in cancer patients with depression: preliminary findings. Am.

J. Psychiatr. 158, 1252–1257.

Natelson, B.H., Denny, T., Zhou, X.D., LaManca, J.J., Ottenweller, J.E.,

Tiersky, L., DeLuca, J., Gause, W.C., 1999. Is depression associated

with immune activation? J. Affect. Disorders 53, 179–184.

Nesse, R.M., 2000. Is depression an adaptation? Arch. Gen. Psychiatry 57,

14–20.

Nishimura, H., Tsuda, A., Oguchi, M., Ida, Y., Tanaka, M., 1986. Is

immobility of rats in the forced swim test ‘behavioral despair’? Physiol.

Behav. 42, 93–95.

O’Reardon, J.P., Chopra, M.P., Bergan, A., Gallop, R., DeRubeis, R.J.,

Crits-Christoph, P., 2004. Response to tryptophan depletion in major

depression treated with either cognitive therapy or selective serotonin

reuptake inhibitor antidepressants. Biol. Psychiatr. 55, 957–959.

Owen, B.M., Eccleston, D., Ferrier, I.N., Young, A.H., 2001. Raised levels

of plasma interleukin-1b in major and postviral depression. Acta

Psychiatr. Scand. 103, 226–228.

Pollmacher, T., Haack, M., Schuld, A., Reichenberg, A., Yirmiya, R., 2002.

Low levels of circulating inflammatory cytokines—do they affect

human brain functions. Brain Behav. Immun. 16, 525–532.

Porsolt, R.D., 2000. Animal models of depression: utility for transgenic

research. Rev. Neurosci. 11, 53–58.

Porsolt, R.D., Le Pichon, M., Jalfre, M., 1977a. Depression: a new animal

model sensitive to antidepressant treatments. Nature 266, 730–732.

Porsolt, R.D., Bertin, A., Jalfre, M., 1977b. Behavioural despair in mice: a

primary screening test for antidepressants. Arch. Int. Pharmacodyn.

229, 327–336.

Porsolt, R.D., Anton, G., Blavet, N., Jalfre, M., 1978. Behavioural despair

in rats: a new model sensitive to antidepressant treatments. Eur.

J. Pharmacol. 47, 379–391.

Pugh, C.R., Nguyen, K.T., Gonyea, J.L., Fleshner, M., Watkins, L.R.,

Maier, S.F., Rudy, J.W., 1999. Role of interleukin-1 beta in impairment

of contextual fear conditioning caused by social isolation. Behav. Brain

Res. 106, 109–118.

Quan, N., Sundar, Q.N., Weiss, J.M., 1994. Induction of interleukin-1 in

various brain regions after peripheral and central injections of

lipopolysaccharide. J. Neuroimmunol. 49, 125–134.

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909908

Quan, N., Zhang, Z., Emery, M., Bonsall, R., Weiss, J.M., 1996. Detection

of interleukin-1 bioactivity in various brain regions of normal healthy

rats. Neuroimmunomodulation 3, 47–55.

Quan, N., Whiteside, M., Herkenham, M., 1998. Time course and

localization patterns of interleukin-1 beta messenger RNA expression

in brain and pituitary after peripheral administration of lipopolysac-

charide. Neuroscience 83, 281–293.

Sakic, B., Gauldie, J., Denburg, J.A., Szechtman, H., 2001. Behavioral effects

of infection with IL-6 adenovector. Brain Behav. Immun. 15, 25–42.

Saphier, D., 1989. Neurophysiological and endocrine consequences of

immune activity. Psychoneuroendocrinology 14, 63–87.

Saphier, D., Welch, J.E., Chuluyan, H.E., 1993. a-Interferon inhibits

adrenocortical secretion via m1-opioid receptors in the rat. Eur.

J. Pharmacol. 236, 183–191.

Schaefer, M., Schwaiger, M., Pich, M., Lieb, K., Heinz, A., 2003.

Neurotransmitter changes by interferon-alpha and therapeutic impli-

cations. Pharmacopsychiatry 36 (Suppl 3), S203–S206.

Schwartz, J.A., Speed, N.M., Brunberg, J.A., Brewer, T.L., Brown, M.,

Greden, J.F., 1993. Depression in stroke rehabilitation. Biol. Psychiatr.

33, 694–699.

Shaywitz, B.A., Klopper, J.H., Yager, R.D., Gordon, J.W., 1976.

Paradoxical response to amphetamine in developing rats treated with

6-hydroxydopamine. Nature 261, 153–155.

Shaywitz, B.A., Fletcher, J.M., Shaywitz, S.E., 2001. Attention

deficit hyperactivity disorder. Curr. Treat Options Neurol. 3, 229–236.

Shen, Y., Connor, T.J., Nolan, Y., Kelly, J.P., Leonard, B.E., 1999.

Differential effect of chronic antidepressant treatments on lipopoly-

saccharide-induced depressive-like behavioural symptoms in the rat.

Life Sci. 65, 1773–1786.

Shimizu, H., Ohtani, K.-I., Sato, N., Nagamine, T., Mori, M., 1995.

Increase in serum interleukin-6, plasma ACTH and serum cortisol

levels after systemic interferon-a administration. Endocr. J. 42, 551–

556.

Shintani, F., Kanba, S., Nakaki, T., Nibuya, M., Kinoshita, N., Suzuki, E.,

Yagi, G., Kato, R., Asai, M., 1993. Interleukin-1b augments release of

norepinephrine, dopamine, and serotonin in the rat anterior hypothala-

mus. J. Neurosci. 13, 3574–3581.

Shuto, H., Kataoka, Y., Horikawa, T., Fujihara, N., Oishi, R., 1997.

Repeated interferon-a administration inhibits dopaminergic neural

activity in the mouse brain. Brain Res. 747, 348–351.

Silverman, M.N., Pearce, B.D., Miller, A.H., 2003. Cytokines and HPA

axis regulation. In: Kronfol, Z. (Ed.), Cytokines and Mental Health.

Kluwer, Norwell, Mass, pp. 85–122.

Silverman M.N, Miller A.H, Biron C.A, Pearce B.D., 2004.

Characterization of an IL-6 and ACTH dependent, immune

to adrenal pathway during viral infection. Endocrinology 145, 3580–

3589.

Smagin, G.N., Swiergiel, A.H., Dunn, A.J., 1996. Peripheral administration

of interleukin-1 increases extracellular concentrations of norepi-

nephrine in rat hypothalamus: comparison with plasma corticosterone.

Psychoneuroendocrinology 21, 83–93.

Smith, R.S., 1991. The macrophage theory of depression. Med. Hypotheses

35, 298–306.

Stout, S.C., Mortas, P., Owens, M.J., Nemeroff, C.B., Moreau, J.-L., 2000.

Increased corticotropin-releasing factor concentrations in the bed

nucleus of the stria terminalis of anhedonic rats. Eur. J. Pharmacol.

401, 39–46.

Strohle, A., Holsboer, F., 2003. Stress responsive neurohormones in

depression and anxiety. Pharmacopsychiatry 36 (Suppl 3), S207–S214.

Swiergiel, A.H., Dunn, A.J., 1999. The roles of IL-1, IL-6 and TNFa in the

feeding responses to endotoxin and influenza virus infection in mice.

Brain Behav. Immun. 13, 252–265.

Swiergiel, A.H., Dunn, A.J., 2003. Feeding and anxiety-related

behaviors are intact in IL-6 deficient mice. Brain Behav. Immun. 17,

214.

Swiergiel, A.H., Smagin, G.N., Dunn, A.J., 1997a. Influenza virus infection

of mice induces anorexia: comparison with endotoxin and interleukin-1

and the effects of indomethacin. Pharmacol. Biochem. Behav. 57,

389–396.

Swiergiel, A.H., Smagin, G.N., Johnson, L.J., Dunn, A.J., 1997b. The

role of cytokines in the behavioral responses to endotoxin and

influenza virus infection in mice: effects of acute and chronic

administration of the interleukin-1-receptor antagonist (IL-1ra). Brain

Res. 776, 96–104.

Takao, T., Culp, S.G., Newton, R.C., De Souza, E.B., 1992. Type I

interleukin-1 receptors in the mouse brain-endocrine-immune axis

labelled with [125I]recombinant human interleukin-1 receptor antagon-

ist. J. Neuroimmunol. 41, 51–60.

Tannenbaum, B., Tannenbaum, G.S., Sudom, K., Anisman, H., 2002.

Neurochemical and behavioral alterations elicited by a chronic

intermittent stressor regimen: implications for allostatic load. Brain

Res. 953, 82–92.

Turnbull, A.V., Rivier, C., 1999. Regulation of the hypothalamic–pituitary–

adrenal axis by cytokines: actions and mechanisms of action. Physiol.

Rev. 79, 1–71.

Uehara, A., Ishikawa, Y., Okumura, T., Okamura, K., Sekiya, C.,

Takasugi, Y., Namiki, M., 1989. Indomethacin blocks the anorexic

action of interleukin-1. Eur. J. Pharmacol. 170, 257–260.

Valentine, A.D., Meyers, C.A., Kling, M.A., Richelson, E., Hauser, P.,

1998. Mood and cognitive side-effects of interferon-a therapy. Sem.

Oncol. 25, S39–S47.

Van Dam, A.M., Bauer, J., Tilders, F.J.H., Berkenbosch, F., 1995.

Endotoxin-induced appearance of immunoreactive interleukin-1b in

ramified microglia in rat brain: a light and electron microscopic study.

Neuroscience 65, 815–826.

Vaugeois, J.M., Odievre, C., Loisel, L., Costentin, J., 1996. A genetic

mouse model of helplessness sensitive to imipramine. Eur.

J. Pharmacol. 316, R1–R2.

Wang, J.P., Dunn, A.J., 1998. Mouse interleukin-6 stimulates the HPA axis

and increases brain tryptophan and serotonin metabolism. Neurochem.

Int. 33, 143–154.

Wang, J.P., Dunn, A.J., 1999. The role of interleukin-6 in the activation of

the hypothalamo–pituitary–adrenocortical axis induced by endotoxin

and interleukin-1b. Brain Res. 815, 337–348.

Watkins, L.R., Wiertelak, E.P., Goehler, L.E., Smith, K.P., Martin, D.,

Maier, S.F., 1994. Characterization of cytokine-induced hyperalgesia.

Brain Res. 654, 15–26.

Watkins, L.R., Maier, S.F., Goehler, L.E., 1995. Cytokine-to-brain

communication: a review & analysis of alternative mechanisms. Life

Sci. 57, 1011–1026.

Weisse, C.S., 1992. Depression and immunocompetence, a review of the

literature. Psych. Bull. 111, 475–489.

Wichers, M.C., Maes, M., 2004. The role of indoleamine 2,3-dioxygenase

(IDO) in the pathophysiology of interferon-alpha-induced depression.

J. Psychiatry Neurosci. 29, 11–17.

Wieczorek, M., Dunn, A.J., 2003. Dissociation of noradrenergic and HPA

responses to interleukin-1 using indomethacin: a microdialysis study.

Brain Behav. Immun. 17, 209.

Willner, P., 1997. Validity, reliability and utility of the chronic mild stress

model of depression: a 10-year review and evaluation. Psychopharma-

cology 134, 319–329.

Wong, M-L., Licinio, J., 1994. Localization of interleukin-1 type I receptor

mRNA in rat brain. Neuroimmunomodulation 1, 110–115.

Wong, M.L., Bongiorno, P.B., Rettori, V., McCann, S.M., Licinio, J., 1997.

Interleukin (IL) 1b, IL-1 receptor antagonist, IL-10, and IL-13 gene

expression in the central nervous system and anterior pituitary during

systemic inflammation: pathophysiological implications. Proc. Natl

Acad. Sci. USA 94, 227–232.

Wong, M.L., Kling, M.A., Munson, P.J., Listwak, S., Licinio, J., Prolo, P.,

Karp, B., McCutcheon, I.E., Geracioti, T.D., DeBellis, M.D.,

Rice, K.C., Goldstein, D.S., Veldhuis, J.D., Chrousos, G.P.,

Oldfield, E.H., McCann, S.M., Gold, P.W., 2000. Pronounced

and sustained central hypernoradrenergic function in major

A.J. Dunn et al. / Neuroscience and Biobehavioral Reviews 29 (2005) 891–909 909

depression with melancholic features: relation to hypercortisolism and

corticotropin-releasing hormone. Proc. Natl Acad. Sci. USA 97, 325–

330.

Yabuuchi, K., Minami, M., Katsumata, S., Satoh, M., 1994. Localization of

type 1 interleukin-1 receptor mRNA in the rat brain. Mol. Brain Res. 27,

27–36.

Yamano, M., Yuki, H., Yasuda, S., Miyata, K., 2000. Corticotropin-

releasing hormone1 receptors mediate consensus interferon-a YM643-

induced depression-like behavior in mice. J. Pharmacol. Exp. Therap.

292, 181–187.

Yirmiya, R., 1996. Endotoxin produces a depressive-like episode in rats.

Brain Res. 711, 163–174.

Yirmiya, R., Weidenfeld, J., Pollak, Y., Morag, M., Morag, A., Avitsur, R.,

Barak, O., Reichenberg, A., Cohen, E., Shavit, Y., Ovadia, H., 1999.

Cytokines, ‘Depression due to a general medical condition’, and

antidepressant drugs. Adv. Exp. Med. Biol. 461, 283–316.

Yirmiya, R., Pollak, Y., Barak, O., Avitsur, R., Ovadia, H., Bette, M.,

Weihe, E., Weidenfeld, J., 2001. Effects of antidepressant drugs on the

behavioral and physiological responses to lipopolysaccharide (LPS) in

rodents. Neuropsychopharmacology 24, 531–544.

Zalcman, S., Green-Johnson, J.M., Murray, L., Nance, D.M., Dyck, D.,

Anisman, H., Greenberg, A.H., 1994. Cytokine-specific central

monoamine alterations induced by interleukin-1, -2 and -6. Brain

Res. 643, 40–49.

Zhang, J-J., Terreni, L., De Simoni, M-G., Dunn, A.J., 2001. Peripheral

interleukin-6 administration increases extracellular concentrations of

serotonin and the evoked release of serotonin in the rat striatum.

Neurochem. Int. 38, 303–308.

Zorrilla, E.P., Luborsky, L., McKay, J.R., Rosenthal, R., Houldin, A.,

Tax, A., McCorkle, R., Seligman, D.A., Schmidt, K., 2001.

The relationship of depression and stressors to immunological

assays: a meta-analytic review. Brain Behav. Immun. 15, 199–226.