-
oni
Available online 4 August 2012
Keywords:InterferonCytokineSerotoninDepression
pathology has increasingly gained acceptance. In this regard,
several lines of evidence suggested that
sion linkage has come from reports indicating that
depressivesymptoms frequently develop in patients undergoing
immuno-therapy with cytokines, such as interferon alpha (IFN-a),
for thetreatment of some types of cancer or chronic viral diseases,
suchas hepatitis C (Capuron and Miller, 2004; Capuron et al.,
2002a).To a considerable extent, the features associated with
depression
tumor necrosis factor-a (TNF-a), treatment with IFN-a does not
af-fect plasma corticosterone levels or induces only modest effects
inthis regard (Menzies et al., 1996), and does not engender a
behav-ioral prole reminiscent of depression. Indeed, even after 3
weeksof treatment with either of two forms of pegIFN in rats, there
wasno indication of weight change, disturbed locomotor activity or
per-formance in a forced swim test (Loftis et al., 2006).
Similarly, De LaGarza et al. (2005) reported that neither acute nor
chronic IFN-atreatment affected plasma corticosterone, central
cytokine expres-sion or depressive-like behaviors (sucrose pellet
self administration,
Corresponding author. Tel.: +1 613 520 2699.E-mail addresses:
[email protected], [email protected]
Brain, Behavior, and Immunity 31 (2013) 115127
Contents lists available at
r,
.e(H. Anisman).1. Introduction
The view that activation of the inammatory immune systemmay
promote depressive illness has been supported by severallines of
research. These have included studies showing correspon-dence
between circulating cytokine levels and major depressivedisorder
(Maes, 1995), the impact of immune challenges on mood(Haroon et
al., 2012; Miller et al., 2009), and the fact that antide-pressant
treatments attenuate the symptoms of depression elicitedby
inammatory factors in animal models (Yirmiya et al., 2001)and in
humans (Musselman et al., 2001).
One of the strongest sources of support for a
cytokine-depres-
induced by immunotherapy are characterized by vegetative
symp-toms (fatigue, feelings of sickness, soporic-like effects), as
well aslow mood (Raison et al., 2010). The occurrence of depression
fol-lowing IFN-a immunotherapy tends to be more pronouncedamong
individuals with sub-syndromal symptoms or a history ofdepression
(Beratis et al., 2005; Capuron et al., 2003, 2004), as wellas low
levels of the serotonin (5-HT) precursor, tryptophan(Capuron et
al., 2003). Moreover, these symptoms diminished withtreatment
cessation (Loftis and Hauser, 2004) and can be attenu-ated by
antidepressants (Kraus et al., 2005; Musselman et al.,2001; Raison
et al., 2005).
Animal studies indicated that unlike interleukin-1b (IL-1b)
and0889-1591/$ - see front matter 2012 Elsevier Inc.
Ahttp://dx.doi.org/10.1016/j.bbi.2012.07.023interleukin-1b (IL-1b)
and tumor necrosis factor-a (TNF-a) can provoke neurochemical and
hormonalchanges akin to those associated with psychological
stressors, and that these cytokines also induce sick-ness behaviors
that resemble some of the neurovegetative features of depression.
Similarly, humandepressed patients often display marked changes of
pro-inammatory cytokine levels and immune cellactivity. Perhaps
more germane in the analysis of the cytokine-depression connection,
reports of humansundergoing interferon-a (IFN-a) treatment for
certain cancers or viral infections have indicated that
thepro-inammatory cytokine caused signs of major depression in a
substantial subset of those treated. Inthe present investigation,
we demonstrated that acute or repeated infusion of IFN-a into the
lateral ven-tricles provoked depressive-like behavior and
concomitant changes in serotonin (5-HT) and mRNAexpression of
particular 5-HT receptors and pro-inammatory cytokines. These
actions were less evidentfollowing administration directly into the
prefrontal cortex and not apparent at all when administered tothe
dorsal raphe nucleus. The data are discussed in relation to the
induction of depression elicited by IFN-a, and are presented in the
context of a mini-review that highlights potential mechanisms
through whichthe cytokine might act to promote psychomotor and
affective disturbances and interact with stressors.
2012 Elsevier Inc. All rights reserved.Article history: A role
for pro-inammatory cytokines and their neuroinammatory signaling
cascades in depressiveCentral administration of murine
interferbehavioral, brain cytokine and neurochemA mini-review and
original experiments
Shawn Hayley, Jeff Scharf, Hymie Anisman Department of
Neuroscience, Carleton University, Ottawa, Canada K1S 5B6
a r t i c l e i n f o a b s t r a c t
Brain, Behavio
journal homepage: wwwll rights reserved.-a induces
depressive-likecal alterations in mice:
SciVerse ScienceDirect
and Immunity
lsevier .com/locate /ybrbi
-
(Spalletta et al., 2006), raises the possibility that the
depressive
(22 C) and humidity (63%) kept constant, and were permitted
free
andforced swimbehavior). In fact, itwas suggested that using a
ratmod-el to assess the depressogenic action of IFN-amight not be
produc-tive (Loftis et al., 2006). However, contrary to the
negative ndings,others reported that administration of IFN-a
increased immobilityin a forced swim test,without
affectingperformance in a tail suspen-sion test or on an
elevatedplusmaze test (Makinoet al., 2000a;Orsalet al., 2008),
although the use of the forced swim test to assessdepressive-like
pathology is controversial.
The fact that IFN-a increased corticosterone in
myeloprolifera-tive disorder patients (after a single but not
repeated injections)and that direct stimulation of hypothalamic and
adrenal cultureswith IFN-a elicited corticotropin releasing hormone
(CRH) and cor-ticosterone production, indicates the potential for
the cytokine toinuence neuroendocrine processes under certain
conditions(Gisslinger et al., 1993). As well, repeated IFN-a
treatment affected5-HT turnover in the prefrontal cortex (PFC) and
amygdala (De LaGarza et al., 2005), as well as dopamine (DA) and
norepinephrine(NE) within the cortex, hypothalamus and medulla
(Kamataet al., 2000; Kumai et al., 2000; Shuto et al., 1997).
Consistent witha role for inammatory processes in promoting these
effects, pre-treatment with diclofenac, a non-steroidal
anti-inammatory drug,prevented the increased turnover of 5-HT in
prefrontal cortex andDA in the hippocampus ordinarily elicited by
intracerebroventric-ular (ICV) IFN-a administration (De La Garza et
al., 2003).
It is uncertain why the different effects of IFN-amight have
oc-curred, although it was suggested that it may be related to the
factthat the murine form of IFN-a was used in some studies,
whereashuman IFN-a was used in others, possibly reecting
differences inreceptors across species (Crnic and Segall, 1992).
Yet, it was re-ported that in both rats and mice human IFN-a
disrupted forcedswim performance, whereas murine IFN-a did not
produce a com-parable outcome. Moreover, this was the case
irrespective ofwhether IFN-a was administered intravenously or into
the brainintracisternally (Makino et al., 2000a,b,c). While not
disputing theeffects of human IFN-a, it was reported that acute
intraperitonealmurine IFN-a administration dose-dependently
increased plasmacorticosterone levels and elicited mild signs of
sickness. Interest-ingly, when the IFN-a treatment was preceded by
a psychosocialstressor the effects of IFN-a were greatly enhanced,
reected bygreater sickness, plasma corticosterone and hippocampal
norepi-nephrine utilization, as well as elevated levels of
circulating IL-6,IFN-a and IL-10 (Anisman et al., 2007, 2008a).
Based on these nd-ings, it was suggested the effects of IFN-a in
humans might simi-larly reect the conjoint effects of the cytokine
coupled with thedistress experienced by cancer or hepatitis C
patients.
Although it is thought that the effects of IFN-a, like that of
othercytokines, involves peripheral actions, cytokines and their
recep-tors are present in the brain, likely being produced within
astro-cytes and microglia, and possibly even in neurons (Huang et
al.,2011; Hulse et al., 2004; Kawanokuchi et al., 2006; Rivest,
2009).Moreover, pro- and anti-inammatory cytokine levels are
elevatedin association with a variety of neuropathological
conditions, suchas ischaemic stroke, as well as head injury (Kamm
et al., 2006; Zhuet al., 2006) and in response to systemic,
neurogenic and psycho-genic stressors (Maier et al., 1999; Miyahara
et al., 2000; Nguyenet al., 1998). It seems that these cytokines
could be acting in a ben-ecial manner (clearing debris and reducing
infection), or at highconcentrations they might act in a
neurodestructive fashion, there-by furthering psychopathology
(Rivest, 2009). In the context of thepresent investigation it may
be particularly signicant that theperiod following stroke is
frequently accompanied by depression(Pascoe et al., 2011; Stoll et
al., 1998), possibly secondary to ele-vated cytokine levels and
variations of indoleamine 2,3-dioxygen-
116 S. Hayley et al. / Brain, Behavior,ase (IDO) and and the
ensuing depletion of serotonin (Spallettaet al., 2006). In fact, it
was shown in a rodent stroke model thatmiddle cerebral artery
occlusion (MCAO) provoked a key featureaccess to food (Ralston
Purina, St. Louis, MO, USA) and water. Tolimit variability
associated with diurnal rhythms, all experimentalprocedures were
conducted between 0800 and 1200 h. The studiesmet the guidelines
set out by the Canadian Council on Animal Careand were approved by
the Carleton University Animal CareCommittee.
2.2. Surgery
Animals were anesthetised with 2.5% isouorane and
stereotax-ically implanted with a 26-gauge stainless steel guide
cannula(Plastic One, Roanoke, VA, USA) in the lateral ventricle (ML
+1.0,AP 0.22, DV 2.5 mm) relative to bregma, according to
coordi-nates from Franklin and Paxinos (1997). In Experiments 47,
thecannula was implanted either in the prefrontal cortex (L .31,
AP268, DV 2.25) or dorsal raphe nucleus (L .20, AP 4.36, DV 3.4,
low-ered at a 25 angle). The guide cannula was anchored to the
skullwith three stainless steel screws and dental cement. A
cannuladummy was inserted into the guide cannula to prevent
blockageof the guide cannula prior to intracerebroventricular (ICV)
injec-tion. Following surgery, animals were individually housed
andallowed a 7-day recovery period prior to behavioral testing.
Following testing or sacrice brains were sliced (40
micronthickness) and examined to verify cannulae placements, and
onlythe data from animals with correct placements were used for
sta-state stems directly from actions of IFNs on other processes,
ratherthan effects secondary to the peripheral actions of this
cytokine.Thus, it was of particular interest in the present
investigation to as-sess the effects of IFN-a administered directly
into the brain, there-by limiting the potential contribution of
peripheral effects ofinterferon.
2. Materials and methods
2.1. Subjects
Nave, male CD-1 strain mice obtained from Charles River Can-ada
(St. Constant, PQ) at 67 weeks of age and were accustomed tothe
vivarium for 2 weeks prior to being used as experimental sub-jects.
Mice were housed in groups of four in standard(27 21 14 cm)
polypropylene cages and maintained on a 12-hlightdark cycle (light
phase: 08002000 h), with temperatureof depression, namely that of
anhedonia (reduced consumptionof sucrose relative to water), which
could be attenuated by treat-ment with interleukin-1 antagonist
(IL-1ra) (Craft and DeVries,2006).
Given the potential links between IFN-a treatment and
depres-sion, the present investigation was conducted to
determinewhether murine IFN-a administered directly into the brain
(eitheracutely or repeatedly) would elicit behavioral effects
related toanxiety or depression, engender variations of monoamine
levelsand utilization, and affect mRNA expression 5-HT receptors
andcytokines in brain. Although cytokines, including IFN-a have
somedifculty entering the brain parenchyma, systemically
adminis-tered IFN-a may gain access to the brain (Pan, Banks &
Kastin,1997) where receptors for this cytokine are present on
microgliaand hence should be widely distributed (Wilkinson et al.,
2010;Yamada and Yamanaka, 1995). The nding that IFN-a and/orIFN-c
are increased in brain following traumatic injury (Khorooshiand
Owens, 2010), and may contribute to poststroke depression
Immunity 31 (2013) 115127tistical analyses. For ICV treatment
all animals but two had correctplacements. All animals with
placements in the prefrontal cortexwere correct. Only 60% of
animals with cannulae aimed at the
-
, anddorsal raphe were correct as this region is small and
cannulae werelowered at a 25 angle.
2.3. Blood collection and brain removal
Between 0800 and 1000 h mice were sacriced by rapid
decap-itation. Trunk blood was collected in tubes containing 10 lg
ofEDTA, centrifuged for 8 min at 3600 RPM, and the plasma storedat
80 C for subsequent corticosterone determination. Brainswere
rapidly removed and placed on a stainless steel brain matrix(2.5
3.75 2.0 cm) positioned on a block of ice. The matrix com-prised a
series of stainless steel plates that had a series of slotsspaced
500 lm apart that guided razor blades to provide coronalbrain
sections. Once the brains were sliced, tissue punches fromthe PFC
and hippocampus were collected by micropunch using hol-low 20 gauge
needles with a beveled tip following the mouse atlasof Franklin and
Paxinos (1997). The collection of these punchestook no longer than
2 min following the decapitation of the animal.Tissue punches were
stored at 80 C for subsequent determina-tion of cytokine or 5-HT
receptor mRNA expression. Samples thatwere to be used for monoamine
determinations the tissue puncheswere placed in 0.3
Monochloroacetic acid containing 10% methanoland internal
standards, and then stored at 80 C.
2.4. Plasma corticosterone determination
Plasma corticosterone levels were determined in duplicateusing a
commercially available radioimmunoassay kit (ICN Bio-medicals, CA,
USA). Assays were conducted in a single run preclud-ing inter-assay
variability, and the intra-assay variability was lessthan 8%.
2.5. High performance liquid chromatography (HPLC) assay
The levels of DA, NE and 5-HT, and their metabolites, DOPAC,MHPG
and 5-HIAA, were determined by HPLC. Tissue puncheswere sonicated
in a homogenizing solution comprising 14.17 gmonochloroacetic acid,
0.0186 g disodium ethylenediamine tetra-acetate (EDTA), 5.0 ml
methanol and 500 ml H2O. Following centri-fugation, the
supernatants were used for the HPLC analysis. Usingan Agilent
(Mississauga, Ontario) pump, guard column, radial com-pression
column (5 m, C18 reverse phase, 8 mm 10 cm), and cou-lometric
electrochemical detector (ESA Model 5100,A), 40 ll of
thesupernatant was passed through the system at a ow rate of1.5
ml/min (14001600 PSI). Each liter of mobile phase consistedof
sodium dihydrogen phosphate (90 mM), 1-octase sulfonic acid(sodium
sal) (1.7 mM), EDTA (50 mM), citric acid (50 mM), potas-sium
chloride (5 mM) and 10% acetonitrile. The mobile phase hadbeen
ltered (0.22 mm lter paper) and degassed. The area andheight of the
peaks were determined using an Agilent integrator.Protein content
of each sample was determined using bicinchoni-nic acid with a
protein analysis kit (Pierce Scientic, Brockville,Ontario) and a
Fluorostar colorimeter (BMG, Durham, NC). Thelower limit of
detection for the monoamines and metaboliteswas approximately 1.0
qg.
2.6. Reverse transcription-quantitative polymerase chain
reactionanalysis in brain
In Experiment 1 and 2, RNA from within the hippocampus andPFC
was isolated and puried by standard methodologies employ-ing Trizol
according to the manufacturers protocol (Invitrogen;Burlington, ON,
Canada). The RNA was then reverse transcribed
S. Hayley et al. / Brain, Behaviorusing Superscript II reverse
transcriptase (Invitrogen; Burlington,ON, Canada), and aliquots of
this reaction were used in simulta-neous quantitative polymerase
chain reactions (QPCR). For QPCR,SYBR green detection was used
according to the manufacturersprotocol (Stratagene Brilliant QPCR
kit). A Stratagene MX-4000 realtime thermocycler was used to
collect the data. All PCR primerpairs used generated amplicons
between 129 and 200 base pairs.Amplicon identity was checked by
restriction analysis. Primer ef-ciency was measured from the slope
relation between absolutecopy number or RNA quantity and the cycle
threshold determinedusing the MX-4000 software. All primer pairs
had a minimum of90% efciency. Primers that amplify GAPDH mRNA were
used asa control to normalize the data.
To compensate for inter-individual variability that
ordinarilyexists within the assay, the expression of each species
within thehippocampus and PFC was normalized by subtracting its Ct
fromthe GAPDH Ct, thus providing the normalized Ct values (Ctn)
foreach mRNA species. A difference between Ctns for an mRNA
spe-cies represents the fold change (i.e., as power of two) in
abundance.To simplify data presentation, the nCt values were
converted tofold changes relative to mice in the nave group,
following the pro-cedure described by Livak and Schmittgen (2001).
Primer se-quences were as follows: GAPDH, forward: GGT CGG TGT
GAACGG ATT TG, reverse: TGC CGT GAG TGG AGT CAT ACT G; MusIL-1b,
forward: TGTCTGAAGCAGCTATGGCAAC, reverse: CTGCCTGAAGCTCTTGTTGATG;
Mus IL-1R1, forward: ATGAGTTACCCGAGGTCCAGTG, reverse:
TACTCGTGTGACCGGA TATTGC; MusIL-6, forward: TCTTGGGACTGATGCTGGTG,
reverse: CAGAATTGCCATTGCACAACTC; Mus IL-6R, forward:
CTCTCCAACCACGAAGGCTG, reverse: TGCAACGCACAGTGACACTATG; Mus TNF-a,
for-ward: CTCAGCCTCTTCTCATTCCTGC reverse: CCATAGAACTG ATGAGAGGG;
Mus IL-10, forward: AATTCCCTGGGTGAGAAGCTG, re-verse:
TCATGGCCTTGTAGACACCTTG; Mus 5-HT1A (Htr1a), for-ward: TCACCTTG
AGTTTGCAGCCTC, reverse: GCAGGAGTTGGAAGCACTTAGG; Mus 5-HT1B (Htr1b),
forward: GTCAAAGTGCGAGTCTCAGACG; reverse:
ACAGATAGGCATCACCAGGGAG;Mus 5-HT2A (Htr2a), forward:
TGCCACCAACTATTTCCTG ATG, re-verse: ACATCCAGGTAAATCCAGACGG; Mus
5-HT2C (Htr2c), for-ward: GTTCAATTCGC GGACTAAGGC, reverse:
GTCAACGGGATGAAGAATGCC.
2.7. Experiment 1: acute infusion of IFN-a
Mice (N = 10/group) received infusion of murine IFN-a (100
or1000 IU) or vehicle into the lateral ventricles. Recombinant
mouseIFN-a (R&D Systems, Minneapolis, MN, USA) was dissolved in
0.5%bovine serum albumin (BSA) as a stabilizing agent and carrier
pro-tein. The control animals received equal amounts of 0.5% BSA
sal-ine solution (vehicle). IFN-a or vehicle was microinjected
intothe lateral ventricle in a 2-ll volume, infused over 5 min, via
aninjection cannula connected to an infusion pump with
polyethyl-ene tubing (Harvard Apparatus, Holliston, MA, USA).
Followingdrug infusion, the injector was left in place for an
additional2 min to ensure drug diffusion.
Immediately following drug administration, mice were re-turned
to their home cages. The cages were part of a Micromax(Accuscan)
motor activity monitoring system allowing home-cagemotor activity
to be recorded for 90 min. Measuring activity in thehome cage
permitted assessment of behavior relatively uncontam-inated by
experimental procedures (e.g., novel environment) thatcould inuence
the response to treatments.
Ninety minutes after infusion, mice were rapidly decapitated,and
plasma trunk blood was collected in tubes containing 10 lgof EDTA
for subsequent determinations of corticosterone. Bloodsamples were
centrifuged for 15 min at 3,600 rpm, and the super-natant was
stored in separate aliquots at 80 C for subsequent
Immunity 31 (2013) 115127 117analyses of corticosterone
levels.Brains were rapidly removed and placed on a stainless
steel
brain matrix (1 1.5 0.75 in.) situated on top of a block of
ice.
-
crose preference/consumption over days (and periods within
andThe brain blocker had a series of slots (spaced 500 lm apart)
thatserved as guides for razor blades to provide coronal brain
sections.Tissue from hippocampus and PFC were collected by multiple
mi-cro-punches using a hollow 20-gauge microdissection needle
fol-lowing the mouse atlas of Franklin and Paxinos (1997). In
thecase of the PFC, eight punches were used to form an inverted
trian-gle, whereas hippocampal punches comprised four
punchesextending approximately 2 mm on either side of the midline.
Tis-sue was stored at 80 C for subsequent
neurochemicaldeterminations.
2.8. Experiment 2: repeated infusion of IFN-a
Mice (N = 10/group) received intracranial infusion of IFN-a
orvehicle into the lateral ventricles (as described in Exp. 1) on
eachof 6 consecutive days. After each infusion, mice were returned
totheir individual cages. Following the fth infusion, sickness
behav-iors were recorded for 1 h at 20-min intervals following
adminis-tration of IFN-a or vehicle. The overall appearance of each
animalwas rated to assess the degree of sickness exhibited.
Sickness mea-surements were scored on a four-point scale (0 = no
symptom,1 = one symptom present, 2 = two symptoms presents, 3 =
threeor more symptoms) with respect to absent exploration and
loco-motion, curled body posture, ptosis, ragged fur, lethargy,
pilo erec-tion, drooping eyelids, and overall nonresponsiveness.
Wepreviously observed that this procedure yielded better than
90%agreement between two raters blind to the treatment mice
re-ceived. Moreover, the results obtained using this procedure
werehighly correlated with the more common procedure in which
eachsymptom was independently scored for severity on a
four-pointscale (Gibb et al., 2011). On the sixth day of treatment,
90 min afterthe nal infusion, mice were rapidly decapitated; blood
and braincollection procedures were the same as described in
Experiment 1.
2.9. Experiment 3: sucrose consumption following repeated
centralIFN-a administration
To examine whether administration of IFN-a provoked
depres-sive-like behaviors, a sucrose preference test that has been
used asa measure of anhedonia was conducted over a six-day period.
Inthis test, all mice (N = 9 or 10/group) were provided seven
daysof pre-training to establish a stable baseline of sucrose
consump-tion. To this end, mice had access to two 200 ml bottles,
one con-taining 2% sucrose and the other tap water, with the
positionaltered on a predetermined random schedule. Bottles
wereweighed and changed daily. Intake volume of each was
determinedon the basis of the bottle weights prior to vs. after the
24 h test per-iod. Following the training phase, mice received
intracranial infu-sion of IFN-a or vehicle into the lateral
ventricles (as describedin Exp. 1) on each of 6 consecutive days.
Additionally, mice wereweighed every day prior to infusion.
2.10. Experiment 4 and 5: intra-PFC infusion of IFN-a
Two additional experiments were undertaken to assess whetherthe
effects observed following ICV administration of IFN-a wouldalso be
apparent when administered directly into brain regionsthat might be
relevant to depressive behaviors. Thus, an initialstudy assessed
the effects of IFN-a administered directly into theprefrontal
cortex, a region in which 5-HT and CRH changes thatcould be
elicited by IFN-a, might contribute to depressive likebehaviors. In
this regard, we previously observed marked varia-tions of cytokine
mRNA expression and monoamine changes in
118 S. Hayley et al. / Brain, Behavior,the PFC in responses to
systemic cytokine or endotoxin challenges(e.g., Gibb et al., 2011).
Thus, in Experiment 4 and 5 we repeatedthe infusion procedure of
Experiment 1, except that IFN-a anddays) was assessed through
repeated measures ANOVA, with timeas a within-group factor. In all
instances, follow-up t tests wereconducted using Bonferroni
corrections to control for family-wiseerror.
3. Results
3.1. Experiment 1. Locomotor activity, sickness behaviors and
plasmacorticosterone levels following acute IFN-a
administration
There was no signicant effect of IFN-a treatment on
motorbehavior, nor was there any indication of sickness behaviors
beingapparent following acute IFN-a administration (Fs < 1; data
notshown). In contrast, plasma corticosterone levels, were
dependenton treatment administered, F(2,24) = 5.44, p < .05. A
single ICVtreatment of IFN-a moderately, but signicantly increased
levelsof plasma corticosterone at both 100 and 1000 IU doses(M SEM
= 12.90 2.08 and 12.11 1.14), compared to vehicletreatment (M SEM =
7.26 .26 ug/dl).
3.2. Cytokine mRNA expression
Cytokine mRNA expression in both the PFC and hippocampuswas
markedly inuenced by the treatment mice received (Fig.
1).Specically, a Treatment effect was observed in the PFC with
re-spect to mRNA expression of IL-1b, IL-6, and TNF-a, F0s(2,24) =
5.57, 9.78, and 4.52 p < .01, .01 and .05, respectively. Fol-low
up tests showed that relative to the vehicle treatment,
theadministration of both doses of IFN-a increased the mRNA
expres-sion of each of these pro-inammatory cytokines.
The cytokine variations within the hippocampus were very
sim-ilar to those seen within the PFC, such that IFN-a altered the
mRNAexpression of IL-1b, IL-6 and TNF-a, Fs(2,24) 5.57, 7.37, and
4.52,vehicle were infused directly into the medial prefrontal
region.Immediately following infusion mice were placed in their
homecages, wherein motor activity was measured as in Experiment
1.After 90 min mice were decapitated and blood and brain tissue
col-lected for determination of corticosterone as well as mRNA
expres-sion of the cytokine and 5-HT receptors as described
earlier,whereas in Experiment 5 tissue was punches were sonicated
andassayed for NE and 5-HT and their respective metabolites MHPGand
5-HIAA.
2.11. Experiment 6 and 7: intra-raphe infusion of IFN-a
Given the potential relationship between CRH and the
seroto-nergic system (Anisman et al., 2008a,b; Linthorst, 2005), we
evalu-ated whether IFN-a applied to the dorsal raphe nucleus
wouldinuence cytokine mRNA expression and monoamine variationsin
the PFC and hippocampus, thereby providing a link betweenCRH and
the dorsal raphe and anxiety/depression. The experimen-tal outcomes
and procedures were identical to those of the previ-ous two
experiments, with the exception being that the dorsalraphe was
targeted rather than the PFC.
2.12. Statistical analyses
Data for corticosterone, as well as mRNA changes for each of
thecytokines, 5-HT receptor subtypes, and the monoamines and
theirmetabolites were analyzed using one-way between-group
analy-ses of variance (ANOVA). Motor activity, sickness scoring,
and su-
Immunity 31 (2013) 115127ps < .01, .01, and .05,
respectively. The follow-up tests conrmedthat the 1000 IU dose
increased the expression of each of thesecytokines relative to
vehicle treated mice. The 100 IU dose also
-
dosage did not reach statistical signicance (.05 < p <
.10).
, and3.3. Serotonin receptor mRNA expressionincreased IL-6 mRNA
expression, whereas the rise of TNF-a at this
Fig. 1. Cytokine (IL-1b, IL-6 and TNF-a) mRNA expression (fold
changes) in theprefrontal cortex (PFC) and hippocampus following
acute i.c.v. administration ofIFN-a (100 or 1000 IU) or vehicle (M
SEM). p < 0.05 compared to vehicle.S. Hayley et al. / Brain,
BehaviorIntracerebroventricular IFN-a infusion did not affect the
expres-sion of 5-HT1A receptors within the PFC, but did promote a
reduc-tion of 5-HT2A receptor expression, F(2,24) = 5.59, p = .01,
that wasprimarily attributable to the reduced expression of this
receptor inmice treated with the 100 IU dose of IFN-a. The change
of 5-HT1Breceptor expression approached, but did not reach
statistical sig-nicance, F(2,24) = 2.48, p = .07.
Within the hippocampus, there was a treatment effect with
re-gard to 5HT1A mRNA expression, F(2,24) = 5.59, p = .05 (Fig. 2).
Spe-cically, reduced 5-HT1A expression was evident at both doses
ofIFN-a compared to vehicle (p < .05). In contrast, expression
of5-HT1B and 5-HT2A receptors was not affected by IFN-a (p = .11and
.26, respectively).
3.4. Experiment 2. Effects of repeated IFN-a administration
3.4.1. Sickness behaviorSickness scores were recorded for 60 min
following the fth day
of treatment. Analysis of sickness behaviors revealed a main
effectof Treatment, F(1,34) = 16.32, p < .001. Specically, as
depicted inFig. 3, IFN-a affected sickness at each of the
assessment times(20, 40, 60 min) relative to vehicle treated mice
(p < .05), yet theextent of the sickness was modest, being well
below that typicallyobserved following moderate doses of IL-1b or
LPS administeredsystemically.
3.4.2. Plasma corticosteroneRepeated administration of IFN-a
signicantly increased corti-
costerone levels, F(1,17) = 39.78, p < .001. Plasma
corticosteronelevels after repeated infusion of IFN-a, (M + SEM =
14.65 + 1.30)were nearly ve times that of vehicle-treated mice (M +
SEM =3.78 + 1.09) (p < .05).
3.4.3. Cytokine mRNA expressionRepeated infusion of IFN-a
increased proinammatory cytokine
mRNA expression in the PFC. As shown in Fig. 4, IL-1b
expressionwas greater in mice repeatedly treated with IFN-a than in
vehi-cle-treated mice, F(1,16) = 5.95, p = .03, as was the
expression ofTNF-a, F(1,16) = 4.53, p = .05. Furthermore, IL-6
expression withinthe PFC was also increased by IFN-a, F(1,16) =
4.61, p = .05, but thisoutcome was marginal.
Repeated infusion of IFN-a increased proinammatory cytokinemRNA
expression in the hippocampus, and as observed in Fig 4,these
effects were appreciably greater, but more variable thanthose seen
in the PFC. Specically, IFN-a increased IL-1b, IL-6,and TNF-a mRNA
expression compared to vehicle, F(1,15) = 5.74,6.64, and 4.31, ps =
.02, .03, and .05.
3.4.4. Serotonin receptor mRNA expressionFig 5 provides the fold
changes of 5-HT receptor subtypes as a
function of the IFN-a treatment mice received. As observed
follow-ing acute IFN-a administration, 5-HT1A expression was
unaffectedin the PFC. Moreover, the decline of this receptor in the
hippocam-pus of acutely treated mice was no longer apparent after
repeatedadministration of IFN-a. Likewise, the decline of 5-HT1B in
the cor-tex was entirely absent following repeated IFN-a, whereas
in thehippocampus the 5-HT1B expression was somewhat
diminished,although not signicantly so (p = .11). This same pattern
wasapparent with regard to 5-HT2C expression (p = .06), whereas
amodest rise of 5-HT2A was apparent in the hippocampus (p <
.07)just as it was following the acute IFN-a treatment.
3.5. Experiment 3. Anhedonia following repeated administration
ofIFN-a
Prior to treatment with IFN-a or vehicle, and after
establish-ment of a stable baseline, sucrose and water intake was
compara-ble (>85% preference for sucrose) in mice that had been
treatedwith either vehicle or IFN-a. Most of the consumption
(>85100%) following the experimental treatments consisted of the
su-crose solution (i.e., very little tap water was consumed), and
thegroups were comparable with respect to the amount water
con-sumed (data not shown). As within group variance associated
withpreference scores was largely attributable to changes of water
con-sumption, the absolute consumption of sucrose was used as
thedependent variable, although the ANOVA yielded comparable
re-sults irrespective of whether sucrose preference or absolute
su-crose consumption was assessed. This analysis of
variancerevealed that over the 5 test days, sucrose consumption was
signif-icantly reduced among mice that received the IFN-a
treatmentF(1,17) 13.93, p < .001. However, as seen in Fig. 6,
the reductionof sucrose consumption was particularly notable after
5 days ofIFN-a treatment as consumption had declined to about 50%
ofwhat it had been at baseline training.
3.6. Experiments 47: intra-PFC and intra-DRN infusion of
IFN-a
When IFN-a was administered directly to the PFC (Experiment4),
home cage locomotor activity was reduced relative to vehicletreated
animals, F(1,14) = 4.86, p < .05 (M + SEM = 1237 + 270 vs3264 +
878 arbitrary units, respectively). However, unlike the ef-fects
induced by ICV IFN-a administration, intra-PFC infusion didnot
inuence the plasma corticosterone levels (M + SEM =
Immunity 31 (2013) 115127 11910.89 + 2.41 and 8.63 + 1.85,
respectively). Furthermore, the proleof hippocampal cytokine
changes observed when IFN-a wasadministered to the prefrontal
cortex was markedly different from
-
and120 S. Hayley et al. / Brain, Behavior,that apparent
following ICV treatment. Specically, the IFN-atreatment did not
inuence the expression of IL-1b, TNF-a or IL-6 within the
hippocampus. Likewise, the expression of hippocam-pal 5-HT receptor
subtypes (5-HT1A, 5-HT1B, 5-HT2A and 5-HT2C)were not affected by
the cytokine treatment (data not shown).
Fig. 2. mRNA expression (fold changes) of 5HT receptors in the
prefrontal cortex (PFC) a(M SEM). p < .05, relative vehicle. p
< .0025, relative to vehicle-treated mice.
Fig. 3. Mean (SEM) sickness scores over 1 h following ICV
infusion of IFN-a orvehicle.Immunity 31 (2013) 115127As evident in
Fig. 7, in Experiment 5, IFN-a administration tothe PFC increased
the hippocampal accumulation of the NE metab-olite MHPG relative to
levels in vehicle treated mice, F(1,13) = 8.04,p = .01, and also
increased NE levels, F(1,13) = 4.34, p = .05. The5-HT metabolite,
5-HIAA, was also elevated following IFN-aadministration, F(1,13) =
5.20, p = .04, whereas the level of 5-HTwas unaffected by the
cytokine treatment.
Experiments 6 and 7 were conducted to assess the effects ofIFN-a
administered at the dorsal raphe nucleus (DRN). In contrastto the
effects observed following ICV or PFC infusion, when IFN-a(1000 IU)
was administered into the DRN no signicant changesin the expression
of cytokines, 5-HT receptors or central mono-amine levels were
apparent (data not shown).
4. Discussion
The use of IFN-a in the treatment of hepatitis-C and some
formsof cancer (hematological maligancies, leukemia and
lymphomas,melanoma) has become more prevalent owing to its ability
to as-sist in viral clearance, antiproliferative, anti-angiogenic
andapoptotic effects, as well as boosting the immune system. A
large
nd hippocampus following i.c.v. administration of IFN-a (100 or
1000 IU), or vehicle
-
Fig. 4. Expression of cytokine mRNA (fold changes) in the
prefrontal cortex (PFC) and hi(M SEM). p < 0.05, p < 0.01
compared to vehicle-treated mice.
Fig. 5. mRNA expression (fold changes) of 5HT receptors in the
PFC and hippocampus
Fig. 6. Sucrose consumption over days associated with repeated
intraventricularadministration of IFN-a (1000 IU) or vehicle (M
SEM).
Fig. 7. Mean SEM concentrations of MHPG, NE as well as 5-HIAA
and 5-HT in thehippocampus following acute administration of IFN-a
to the nedial prefrontalcortex. p < 0.05.
S. Hayley et al. / Brain, Behavior, andppocampus following
repeated i.c.v. administration of of IFN-a (1000 IU) or vehicle
following repeated i.c.v. administration of IFN-a (1000 IU) or
vehicle (M SEM).
Immunity 31 (2013) 115127 121sub-group of patients receiving
IFN-a treatment manifest depres-sive symptoms (Capuron and Miller,
2004), and although someinconsistencies are apparent in animal
studies, several reportsindicated that IFN-a can induce some
depressive-like changes inrodents (Maes et al., 2011; Myint et al.,
2009). Consistent withour earlier report concerning the effects of
systemic murine IFN-a administration (Anisman et al., 2007), in the
present investiga-tion modest sickness behavior was induced in mice
following re-peated but not acute ICV infusion of IFN-a. As well,
ICV (but notintra-PFC or DRN) administered IFN-a increased plasma
corticoste-rone and this outcome was apparent with both acute and
repeatedexposure, pointing to the independence of the
corticosterone vari-ations and the sickness behaviors.
Like the neurovegetative changes associated with IFN-a
immu-notherapy, the cytokine initially provoked only modest
reductionsof sucrose consumption, but once mice had received the
treatmentover successive days, sucrose consumption declined
appreciably,possibly reecting anhedonia, a characteristic feature
of depres-sion. Insofar as sickness behavior has been taken to
resemble thevegetative behaviors associated with major depression
(Dantzeret al., 2008, 2011), and the changes of sucrose intake
reect anhe-donia associated with depression, the present ndings are
consis-tent with the view that inammatory challenges are related
tothe emergence of a depressive-like state. Given that IFN-a
affectedanhedonia, but did not elicit sickness that might reect
somaticfeatures of depression, shouldnt necessarily be taken to
suggestthat the link between IFN-a and depressive features is
uncertainin an animal model. To be sure, depression comprises both
somaticand cognitive/affective symptoms, possibly reecting
diverseunderlying mechanisms (Anisman et al., 2008a). Thus, it may
beparticularly relevant that among hepatitis C patients responses
toquestions on the Beck Depression Inventory that targeted
cognitiveand affective symptoms appeared to provide better
validity
-
andconcerning depression relative to questions that focused on
so-matic features (Patterson et al., 2011). It might similarly be
thecase that symptoms that reect anhedonia (or cognitive
distur-bances) in an animal model might be more aligned with
depressionthan are somatic characteristics.
Admittedly, the magnitude of the anhedonic effects observed
inthe present study was modest, and it is possible that it was
second-ary to the sickness elicited by the IFN-a. As will be
discussedshortly, there are any number of reasons why repeated
treatmentwas necessary to elicit the observed behavioral changes,
but it isunlikely that the behavioral disturbances were due to
cumulativedrug effects as the half life of IFN-a, like that of
other cytokinesis brief. Signicantly, the neurovegetative effects
of IFN-a in hu-mans also appear only after a few weeks after
treatment begins,and the mood changes, including anhedonia, develop
later (Capu-ron et al., 2002a). Thus, with a more chronic regimen,
the anhe-donic effects of IFN-a might have become more notable.
Also ofsignicance is the fact that we presently found that low
doses ofIFN-a administered directly into the brain elicited
anhedonia,implicating central mechanisms as being the source of the
behav-ioral changes observed.
An important consideration regarding the IFN-a induced
behav-ioral effects is whether they actually represent a genuine
depres-sive-like syndrome. Recent data seem to support this
contentiongiven that antidepressant medications are as effective in
reducingcore depressive features in IFN-a treated patients as those
in thegeneral depressed population. Indeed, the SSRI, paroxetine,
wasshown to reduce symptoms of depression in hepatitis C
patientsthat received IFN-a with or without concomitant ribavirin
treat-ment (Capuron et al., 2002a,b; McNutt et al., 2012).
Moreover, anti-depressant treatment was equally effective in
melanoma patientsthat had IFN-a induced depressive features
relative to depressedpatients that had not been treated with this
cytokine (Navinset al., 2009). However, IFN-a induced depression
appears to beassociated with more pronounced psychomotor
retardation andweight loss, coupled with less severe feelings of
guilt, comparedto depressed patients that had not received cytokine
therapy(Capuron et al., 2009). Animal studies have paralleled these
humanndings, as IFN-a induced anhedonia in rats was attenuated by
re-peated antidepressant treatment (Sammut et al., 2002).
Interest-ingly, paroxetine was not effective as a prophylactic
treatmentwhen given to prevent the initial emergence of IFN-a
induceddepressive symptoms (McNutt et al., 2012). Similarly, a
doubleblind clinical trial failed to reveal signicant prophylactic
effectsof citalopram with respect to the onset of IFN-a induced
depres-sion (Morasco et al., 2010). Thus, SSRIs appear to help
managesymptoms after their emergence, but might not appreciably
inu-ence the pro-depressive process initially set into motion by
IFN-a.It is likely that early pro-inammatory processes engaged by
IFN-aare not substantially affected by antidepressant pre-treatment
andthat the antidepressant only becomes relevant after the
initialinammatory processes come to provoke 5-HT and other
aminer-gic changes.
4.1. Cytokine variations in brain induced by IFN-a
In contrast to its well delineated signaling pathways within
theperipheral immune system, limited data are available
concerningthe central mechanisms through which IFN-a might
provokedepressive symptoms. One obvious mechanism through
whichIFN-a could inuence emotional processes is through the
dysregu-lation of the central cytokine network, as well as the
provocationof glial-dependent pro-inammatory actions.
Immunotherapywith
122 S. Hayley et al. / Brain, Behavior,IFN-awas found to elevate
serumconcentrations of IL-6, IL-8 and IL-10 and these cytokine
changes were correlated with the manifesta-tion of depressive
feelings, although these ndings didnt speakdirectly to potential
cytokine variations in brain (Bonaccorso et al.,2001). In the
present investigation, acute or repeated administra-tion of IFN-a
directly into the brain (ICV)markedly increasedmRNAexpression of
IL-1b, IL-6 and TNF-a in the PFC and hippocampus.This is
particularly important in light of the evidence indicating
thatIL-1b and TNF-a can induce stressor-like neurochemical
anddepressive-like behavioral effects that resemble vegetative
symp-toms of depression in humans (e.g., anorexia, anxiety, reduced
mo-tor activity) (Dantzer et al., 2008; Hayley et al., 2005). By
contrast,administration of IL-6 generally does not elicit
substantial behav-ioral changes (Bluth et al., 2000), but itmay act
synergisticallywithIL-1b in promoting behavioral disturbances
(Brebner et al., 2000).
The PFC and the hippocampus have been implicated in depres-sion,
and long-termhippocampal inammation has been associatedwith
depressive behavior (Curran and OConnor, 2001; Fu et al.,2010). The
hippocampus also expresses a particularly high densityof IL-1
receptors (Parnet et al., 2002) and, as such, it might also
berelatively susceptible to the adverse consequences of
neuroinam-mation and could contribute to the cognitive disruptions
thataccompany IFN-a treatment. However, the role of specic
proin-ammatory cytokines (IL-1a, IL-6, or TNF-a) to depression
inducedby IFN-a is uncertain. Even in the present experiment,
inwhich IFN-a up-regulated each of these cytokines, it is not clear
which (if any)of these cytokines contributed to depressive
features.
A specic role for IL-6 has increasingly gained support as a
con-tributing factor in depression given that elevated circulating
IL-6levels in depressed patients normalize with successful
antidepres-sant treatment (Basterzi et al., 2005). Further evidence
has comefrom two meta-analyses, one reporting that plasma IL-1b and
IL-6 were positively associated with depression (Howren et
al.,2009), and the second that indicated signicantly higher levels
ofIL-6 in plasma of depressed patients compared to controls
(Dowlatiet al., 2010). Paralleling these conditions, immunotherapy
withIFN-a in patients was also associated with increased IL-6, and
pa-tients with elevated plasma IL-6 prior to treatment had been
foundto be especially vulnerable to subsequent depression induced
byIFN-a (Wichers et al., 2006), possibly accounting for the
variabilitythat exists concerning the depressogenic effects of
IFN-a. In thepresent investigation acute ICV IFN-a provoked a
marked elevationof IL-6 that was considerably more pronounced in
the PFC than inthe hippocampus. With repeated IFN-a treatment,
however, the IL-6 rise in the PFC was diminished, whereas that in
the hippocampuswas exaggerated. Why these region-specic variations
developedwith repeated administration of the cytokine isnt certain.
How-ever, given that the behavioral symptoms of depression also
in-creased with repeated treatment, these data point to the
IL-6hippocampal variations being more closely aligned with the
behav-ioral changes relative to the PFC variations.
4.2. Serotonergic alterations in the brain following
IFN-aadministration
Studies in humans have suggested that diminished availabilityof
5-HT, the 5-HT transporter or variations of particular
5-HTreceptors is linked to depression. For instance, in both
humanand animal studies, 5-HT1A (Albert and Franois, 2010; Nishiet
al., 2009), 5-HT1B (Sari, 2004), and 5-HT2A (Bhagwagar et al.,2006)
might contribute to affective disorders, as well as the
actionsinvolved in the effectiveness of antidepressants. Thus, it
was ofparticular interest in the present investigation to assess
5-HTreceptor mRNA expression in response to IFN-a.
The 5-HT changes induced by IFN-a in the present study were,in
several respects, consistent with experiments that implicated
5-
Immunity 31 (2013) 115127HT functioning in depression. Previous
reports indicated that a sin-gle ICV injection of IFN-a (200 or
2000 IU) reduced 5-HT and NElevels in the frontal cortex (Kamata et
al., 2000) and that its re-
-
, andpeated systemic administration also reduced cortical 5-HT
levels(Asnis et al., 2003). In the present investigation, acute ICV
infusionof murine IFN-a reduced 5-HT1B and 5-HT2A receptor
mRNAexpression within the PFC, whereas 5-HT1A was reduced in the
hip-pocampus. It is unlikely that the 5-HT receptor changes were
dueto any adverse reactions to IFN-a, as the doses used were far
belowthose thought to be associated with toxicity.
The decreased levels of 5-HT1B mRNA in the PFC are
consistentwith reports in human post-mortem studies indicating
reduced5-HT1B mRNA in individuals with major depression who had
diedby suicide (Anisman et al., 2008b). Similarly, the impact of
IFN-aon hippocampal 5HT1A mRNA expression is consistent with
studiesin humans showing that the 5-HT1A receptor expression was
notonly down-regulated in depression but was linked to SSRI
effec-tiveness (Artigas et al., 1996). Furthermore, a genetic
polymor-phism for the 5-HT1A receptor has been linked to the onset
ofIFN-a-induced depression in patients undergoing
immunotherapy(Kraus et al., 2007). Besides being of importance in
neurotransmis-sion, the 5-HT1A receptor is believed to play a
fundamental role invarious immune processes, including macrophage
phagocytosis, Tcell proliferation, as well as adhesion and
chemotaxis of mast cells(Ahern, 2011). Moreover, T cell
leukemia-derived cell lines incu-bated with IFN-a displayed reduced
5-HT1A receptor expressionand this effect was attenuated by
co-incubation with either tricy-clic or SSRI antidepressants (Cai
et al., 2005). Thus, IFN-a mighthave similar actions upon multiple
cell types that could ultimatelystimulate neuroinammatory processes
and conversely, the anti-inammatory effects, known to exist for
certain antidepressants,could facilitate benecial clinical outcomes
by antagonizing IFN-a effects at multiple nodes of the neuro-immune
axis.
The present nding concerning decreased 5-HT2A expression inPFC
and slightly elevated levels in hippocampus are in some re-spects
difcult to reconcile with ndings in humans. Specically,it has been
reported 5-HT2A receptor density was increased withinthe PFC of
individuals with a history of depression (Shelton et al.,2009) and
was also elevated among individuals that died by sui-cide (Arango
et al., 1990; Hrdina et al., 1993; Pandey et al.,2002), although
these effects could have been related to suiciderather than
depression (Pandey et al., 2002). To be sure, variationsof 5-HT2A
density within cortical brain structures are not uniformlyevident
among depressed individuals (Underwood et al., 2011). Inthis
regard, for instance, increased 5-HT2A binding was seen in
thedorsolateral prefrontal cortex among those patients with
particu-larly elevated feelings of pessimism and hopelessness
(Meyeret al., 2003). Such ndings speak to the view that individual
symp-toms ought to be given greater attention in assessing
characteris-tics of depression, but obviously characteristics such
as thesecannot be determined in an animal model of depression.
As indicated earlier, depressive symptoms in humans becomemore
pronounced with continued IFN-a treatment, with neuroveg-etative
features becoming apparent rst, followed by affectivesymptoms
(Capuron et al., 2002a; Trask et al., 2004; Wicherset al., 2007).
It likewise appeared that some of the behavioral dis-turbances
observed in the present investigation became more pro-nounced with
repeated IFN-a administration. In contrast, however,the 5-HT
receptor variations associated with acute treatment wereless
pronounced after repeated treatment. This disconnect be-tween the
effects of repeated IFN-a on behavior and 5-HT receptorchanges
appears to argue against a role for 5-HT in the behavioraloutcomes.
This said, it is certainly possible that with continuedIFN-a
treatment the initial 5-HT receptor changes could provokestill
further downstream changes that might be aligned withdepression.
For instance, dysregulation of brain derived neurotro-
S. Hayley et al. / Brain, Behaviorphic factor (BDNF) and other
trophic factors may come to disruptneurogenesis and the proper
incorporation of new neurons intohippocampal circuits, or
alternatively, existing neural and synapticconnections could become
weakened. In fact, it was reported thatsystemic IFN-a treatment to
rodents, using a schedule that wasaligned with human therapeutic
application, reduced hippocampalcell proliferation (Kaneko et al.,
2006). Likewise, humans treatedwith IFN-a displayed time-dependent
reduction of serum BDNFlevels that correlated with depressive
symptom severity, and itwas suggested that reductions in
neuroprotective capacity couldexplain such outcomes (Kenis et al.,
2011). Although antidepres-sants and cytokines are both known to
affect BDNF (Sen et al.,2008), their timing in relation to one
another as well as their exactroute of action are uncertain.
4.3. Alternative mechanisms of IFN-a induced
depressive-likepathology
It is unclear whether IFN-a could be exerting its
neurochemicaleffects through central or peripheral mechanisms,
although thepresent data certainly support a central mode of
action. Indeed,IFN-a receptors are normally located on peripheral
immune cells,such as macrophages and natural killer cells, where
they regulateantigen presentation (via MHC molecules) and
anti-viral re-sponses. Thus, it is possible that IFN-a inuenced
brain processes,secondarily by way of systemic immune cell
activity. However,IFN-a receptors have also been observed on
astrocytes and microg-lia, at least in response to immunogenic
stimuli and in neuropa-thology (Huang et al., 2011; Hulse et al.,
2004; Kawanokuchiet al., 2006; Rivest, 2009). Moreover, the closely
related receptor,IFN-c, has been localized to human microglia and
is believed tobe one of the most potent endogenous inducers of
microglial reac-tivity (Hashioka et al., 2010).Although less
convincing, there iseven evidence that IFN-a expression occurs in
cultured humanneuronal cell lines (NT2-N) (Wan et al., 2008). At
this juncture, itis unclear as to the exact cell type(s) through
which IFN-a im-parted its central effects in the current
experiments, nor is it clearwhether brain-region specic differences
in IFN-a receptor expres-sion (or easier access following infusion)
could explain the greaterneurochemical changes elicited in certain
regions.
Pro-inammatory states in the brain are known to affect
themetabolism of tryptophan and levels of 5-HT by affecting the
en-zyme, indoleamine-2,3-dioxygenase (IDO), resulting in changes
inthe kynurenine (KYN)/kynurenic (KYNA) acid ratio and degrada-tion
of tryptophan, which could favor the development of depres-sion
(Dantzer et al., 2011; Maes et al., 2011). In
particular,neuroinammation induces KYN metabolism in astrocytes
andmicroglia, and results in 5-HT being degraded by IDO into
for-myl-5-hydroxykynuramine, resulting in diminished availability
of5-HT (Myint, 2012). In the case of IFN-a therapy, the cytokinewas
shown to increase IL-6 and the KYN/KYNA ratio that was asso-ciated
with depressive pathology (Wichers et al., 2007; Maes et al.,2011).
Similarly, the immune agent, bacillus Calmette-Guerin(BCG), used to
induce depressive-like behaviors in mice, activatedthe IDO enzyme;
conversely, inhibition of IDO prevented thedepressive-like
behaviors of BCG (OConnor et al., 2009).
The link between IFN-a therapy and the induction of IDO
activ-ity might involve IFN-c, which is the major endogenous
regulatorof IDO and is strongly induced by IFN-a. Thus, it is of
signicancethat IFN-c knockout mice were resistant to the
development ofdepressive-like behaviors when challenged with BCG
(OConnoret al., 2009). Similarly, our ownwork indicated that IFN-c
null micedisplayed basal differences in open eld exploration, along
with in-creased noradrenergic and serotonergic activity within the
centralamygdala, relative to their wild-type counterparts
(Litteljohn et al.,2009, 2010). Moreover, stressor-induced
elevations of corticoste-
Immunity 31 (2013) 115127 123rone and TNF-a, as well as DA
turnover (prefrontal cortex, hypo-thalamus and central amygdala)
were blunted in IFN-c-decientmice. However, chronic stressor
exposure induced anxiety-like (re-
-
medical condition. Thus, the depressogenic actions of IFN-a
mightreect the conjoint effects of the cytokine superimposed on a
back-
andduced open eld exploration) and depressive-like
behaviors(forced swim immobility, reduced sucrose consumption)
amongboth wild-type and IFN-c knockout mice alike. Hence, the
immu-nological, hormonal and dopaminergic process affected by
endog-enous IFN-c did not appear to be translated into obvious
affectivedisturbances (Litteljohn et al., 2009, 2010).
Besides a role for IFN-c and IDO processes in depressive
symp-toms, changes in the ratio of tryptophan to large neutral
aminoacids (LNAA) (tryptophan, tyrosine, valine, leucine,
isoleucine,phenylalanine) induced by IFN-a treatment might also be
ofimportance (Capuron et al., 2002a,b). Indeed, diminished
trypto-phan levels and a reduced tryptophan/LNAA ratio was evident
fol-lowing IFN-a treatment and was associated with
anorexia,pessimistic and suicidal thoughts, as well as poor
concentration(Capuron et al., 2002b). Additionally, the fact that
pre-existing highcirculating levels of sIL-2R, IL-6 and IL-10 was
associated with agreater likelihood of depression following IFN-a
therapy indicatesthat an enhanced immune state, particularly the T
cell adaptivebranch of immunity, may render individuals vulnerable
to thedevelopment of depression (Wichers et al., 2006). Similarly,
itwas reported that pre-existing immune activation, in this case
re-ected by elevated sTNFR-1, predicted later vulnerability to
IFN-ainduced depressive symptoms (Friebe et al., 2007).
Although IFN-a treatment certainly elicits a depressive
condi-tion, it also produces a wide range of neurocognitive and
other ef-fects, many of which are unrelated to mood. Indeed,
delirium, u-like symptoms, cognitive disturbances, muscle pain and
thyroiddysfunction have all been reported to occur with fairly high
fre-quencies (Patten, 2006; Raison et al., 2005). Alterations of
5-HTand DA activity, along with psychomotor retardation and
changesin metabolic activity of the basal ganglia and anterior
cingulatecortex were also reported in non-human primates (Miller,
2009).IFN-a treatment in rodents has been associated with extreme
fati-gue, as well as cataplexy together with concomitant
disturbancesof DA accumulation and turnover (Shuto et al., 1997).
In fact, ro-dent models of chronic fatigue (with either
administration of im-mune activating agents or excessive wheel
running) implicatedIFN-a in such outcomes (Davis et al., 1998;
Katafuchi et al.,2005). Similarly, general reductions of motor
performance acrossseveral measures, including open eld, muscle
strength tests alongwith swim posture and endurance have been
reported (Dunn andCrnic, 1993). A microarray study that identied
252 up-regulatedgenes, found that 20-50-oligoadenylate synthetase
2, a gene linkedto chronic fatigue syndrome, was the only one
differentially ex-pressed in patients with IFN-a-induced depression
and fatigue(Felger et al., 2011). Thus, caution ought to be
exercised wheninterpreting IFN-a effects in relation to potential
depressive-likebehaviors, especially in animal models. In fact, it
is highly likelythat common IFN-mediated immune mechanisms
contribute toneurovegetative aspects of a range of
disorders/syndromes.
Recent studies have linked malaise, sickness and anhedonia
togeneral disturbances in cell mediated immunity, in which the
IFNsplay a critical role. In this regard,
thepossibilitywasevenentertainedthat the translocation of bacteria
normally found in the gut couldgive rise to such behaviors by
triggering oxidative and nitrosativestress pathways (Leonard and
Maes, 2012). Indeed, enhanced anti-body and oxidative stress
reactions to immune challenges (e.g.LPS) was observed in depressed
patients, which was suggested topossibly stem from leaky gut
syndrome thatmight become appar-ent with chronic depression (Haroon
et al., 2012). In such situations,individuals with pre-existing
depression coupledwith a leaky gut,would already have heightened
ongoing inammatory responses(owing tobacterial
translocation),whichwouldplace thematpartic-
124 S. Hayley et al. / Brain, Behavior,ular risk for the
deleterious consequences of IFN-a administration.We have suggested
previously (Anisman et al., 2008a,b) that pa-
tients treated with IFN-a comprise a unique population as most
aredrop of distress. Indeed, we demonstrated that although
systemicadministration of murine IFN-a in mice has modest
depressive-likebehavioral effects and limited action on brain and
plasma cyto-kines, as well as central monoamine turnover and
levels, these out-comes were markedly augmented when administered
following astressor (Anisman et al., 2007). The increased
depressive-likebehaviors observed could be due to the elevated
cytokine function-ing in brain or downstream effects of these
cytokine changes, butthe possibility cannot be dismissed that the
stress-IFN-a synergyis attributable to inammatory effects on gut
microbial processes,variations of IDO or BDNF, or other as yet
unidentied factors. Fur-ther to this same point, hepatitis C and
cancer patients are alsoundergoing substantial immunological stress
which can be ex-pected to elicit IFN-mediated inammatory reactions
resultingthe up-regulation of IFN receptor expression in brain
(Wilkinsonet al., 2010). Thus, in the context of this immunological
stress, fur-ther administration of IFN-a would be expected to have
more pro-found effects than it would otherwise.
There are certainly several limitations to the ndings of
thepresent investigation. One that bears particular mention is
thatthe effects of IFN-a were assessed only in male mice. Yet, in
hu-mans depression occurs 23 times more often in females thanmales,
and female rodents typically exhibit a greater
glucocorticoidresponse relative to males (Anisman et al., 2008b).
Furthermore, itwas reported that in humans a stressor in the form
of social exclu-sion following intravenous endotoxin treatment
elicited self-re-ported signs of depression. However, in females
(but not males)the increased levels of plasma IL-6 elicited by the
endotoxin wasassociated with increased social pain and feelings of
depressionthat appeared to be mediated by neuronal activity within
the dor-sal anterior cingulate cortex and anterior insula
(Eisenberger et al.,2009), brain regions that have frequently been
linked to majordepressive disorder. Indeed, it has been suggested
that social rejec-tion, promoted self-conscious emotions (e.g.,
humiliation, shame)and negative self-referential cognitions that
were mediated bythese brain regions, which might have inuenced HPA
functioning,inammatory processes and promotion of depressive
disorder(Slavich et al., 2010).
A second limitation of the present investigation, at least with
re-spect to its clinical relevance, is that the IFN-awas
administered di-rectly into brain as opposed to being administered
systemically asin clinical situations. Further to this, the IFN-a
treatment wasadministered acutely or involved a limited
administration protocol,unlike the lengthy treatment schedule
administered clinically totreat hepatitis C or cancer. In fact, it
would not be unrealistic to ex-pect that these different treatment
schedules and routes of admin-istration might engender
neurobiological changes that differedfrom those observed in the
present investigation, and might yieldoutcomes, including 5-HT
receptor changes, that would be morein line with those hypothesized
to be associated withmajor depres-sive disorder. However, the ICV
route of administration does havethe potential benet of ensuring
the cytokine greater access tohypothalamic neurons than a systemic
route. This may be particu-larly important given that IFN-a could
directly stimulate CRH re-lease from cultured hypothalamic neurons
(Gisslinger et al.,1993); which may help explain differences
between our resultsand some previous studies.
5. Conclusionlikely experiencing considerable distress stemming
from their
Immunity 31 (2013) 115127Treatment with IFN-a elevates brain
cytokines and plasma cor-ticosterone, and could potentially mediate
some of the behavioral
-
Ahern, G.P., 2011. 5-HT and the immune system. Curr. Opin.
Pharmacol. 11, 2933.
2008a. Serotonin receptor subtype and p11 mRNA expression in
stress-relevant
, andbrain regions of suicide and control subjects. J.
Psychiatry Neurosci. 33, 131141.
Anisman, H., Merali, Z., Hayley, S., 2008b. Neurotransmitter,
peptide and cytokineprocesses in relation to depressive disorder:
comorbidity between depressionand neurodegenerative disorders.
Prog. Neurobiol. 85, 174.
Arango, V., Ernsberger, P., Marzuk, P.M., Chen, J.S., Tierney,
H., Stanley, M., et al.,1990. Autoradiographic demonstration of
increased serotonin 5HT2 and b-adrenergic receptor binding sites in
the brain of suicide victims. Arch. Gen.Psychiatry 47,
10381046.
Artigas, F., Romero, L., De Montigny, C., Blier, P., 1996.
Acceleration of the effect ofselected antidepressant drugs in major
depression by 5-HT1A antagonists.Trends Neurosci. 19, 378383.
Asnis, G.M., De la Garza, R., Kohn, S.R., Reinus, J.F.,
Henderson, M., Shah, J., 2003. IFN-induced depression: a role for
NSAIDs. Psychopharmacol. Bull. 37, 2950.
Basterzi, A.D., Aydemir, C., Kisa, C., Aksaray, S., Tuzer, V.,
Yazici, K., Gka, E., 2005. IL-6 levels decrease with SSRI treatment
in patients with major depression. Hum.Psychopharmacol. 20,
473476.
Beratis, S., Katrivanou, A., Georgiou, S., Monastirli, A.,
Pasmatzi, E., Gourzis, P.,Tsambaos, D., 2005. Major depression and
risk of depressive symptomatologyassociated with short-term and
low-dose interferon-alpha treatment. J.Psychosom. Res. 58,
1518.
Bhagwagar, Z., Hinz, R., Taylor, M., Fancy, S., Cowen, P.,
Grasby, P., 2006. Increased 5-HT(2A) receptor binding in euthymic,
medication-free patients recovered fromdepression: a positron
emission study with [(11)C]MDL 100,907. Am. J.Psychiatry 163,
15801587.
Bluth, R.M., Michaud, B., Poli, V., Dantzer, R., 2000. Role of
IL-6 in cytokine-inducedsickness behavior: a study with IL-6
decient mice. Physiol. Behav. 70, 367373.
Bonaccorso, S., Puzella, A., Marino, V., Pasquini, M., Biondi,
M., Artini, M., Almerighi,C., Levrero, M., Egyed, B., Bosmans, E.,
Meltzer, H.Y., Maes, M., 2001.Immunotherapy with interferon-alpha
in patients affected by chronichepatitis C induces an
intercorrelated stimulation of the cytokine networkAlbert, P.R.,
Franois, B.L., 2010. Modifying 5-HT1A receptor gene expression as
anew target for antidepressant therapy. Front Neurosci. 4, 35.
Anisman, H., Poulter, M.O., Gandhi, R., Merali, Z., Hayley, S.,
2007. Interferon-aeffects are exaggerated when administered on a
psychosocial stressorbackdrop: cytokine, corticosterone and brain
monoamine variations. J.Neuroimmunol. 186, 4553.
Anisman, H., Du, L., Palkovits, M., Faludi, G., Kovacs, G.G.,
Szontagh-Kishazi, P., et al.,effects elicited by this cytokine. As
central infusion of IFN-a, at adose far below that which would
induce any effects if administeredsystemically, suggests that
central processes activated by IFN-a areresponsible for the
behavioral effects provoked. The cytokinechanges, neuroendocrine
upregulation, and serotenergic modica-tions elicited by IFN-a have
been observed in studies that involvedits systemic administration
or directly into the brain. These nd-ings are consistent with the
proposition that systemic IFN-a is ableto cross the BBB or activate
peripheral mechanisms that activatecentral processes, including
IFN-a in the CNS. Indeed, Raisonet al. (2009) reported that
pegylated IFN-a administered peripher-ally leads to increased CSF
IFN-a in patients undergoing immuno-therapy, although they were not
able to determine whether theIFN-a in the CSF was the exogenous
IFN-a or was produced endog-enously following treatment (Raison et
al., 2009). Taken together,the present data support a role for
IFN-a in the provocation of adepressive-like syndrome that is
associated with signicant alter-ations of 5-HT receptors and
central cytokine expression. It is likelythat IFN-a driven
pro-inammatory consequences are particularlyimportant for the
psychomotor and vegetative aspects of depres-sion, and those
individuals with pre-existing immune disturbancesor an extensive
previous stressor history would be especially vul-nerable to the
deleterious effects of IFN-a treatment.
Acknowledgment
The Research was supported by grants to SH and HA from
theCanadian Institutes of Health Research, and both hold
CanadaResearch Chairs in Neuroscience.
References
S. Hayley et al. / Brain, Behaviorand an increase in depressive
and anxiety symptoms. Psychiatry Res. 15, 4555.Brebner, K., Hayley,
S., Zacharko, R., Meral, i Z., Anisman, H., 2000. Synergistic
effects of interleukin-1beta, interleukin-6, and tumor necrosis
factor-alpha:central monoamine, corticosterone, and behavioral
variations. Neuropsy-chopharmacology 22, 566580.
Cai, W., Khaoustov, V.I., Xie, Q., Pan, T., Le, W., Yoffe, B.,
2005. Interferon-alpha-induced modulation of glucocorticoid and
serotonin receptors as a mechanismof depression. J. Hepatol. 42,
880887.
Capuron, L., Miller, A.H., 2004. Cytokines and psychopathology:
lessons frominterferon-alpha. Biol. Psychiatry 1, 819824.
Capuron, L., Gumnick, J.F., Musselman, D.L., Lawson, D.H.,
Reemsnyder, A., Nemeroff,C.B., et al., 2002a. Neurobehavioural
effects of interferon-alpha in cancerpatients: phenomenology and
paroxetine responsiveness of symptomdimensions.
Neuropsychopharmacology 26, 643652.
Capuron, L., Ravaud, A., Neveu, P.J., Miller, A.H., Maes, M.,
Dantzer, R., 2002b.Association between decreased serum tryptophan
concentrations anddepressive symptoms in cancer patients undergoing
cytokine therapy. Mol.Psychiatry 7, 468473.
Capuron, L., Neurauter, G., Musselman, D.L., Lawson, D.H.,
Nemeroff, C.B., Fuchs, D.,Miller, A.H., 2003.
Interferon-alpha-induced changes in tryptophan
metabolism.relationship to depression and paroxetine treatment.
Biol. Psychiatry 54,906914.
Capuron, L., Ravaud, A., Miller, A.H., Dantzer, R., 2004.
Baseline mood andpsychosocial characteristics of patients
developing depressive symptomsduring interleukin-2 and/or
interferon-alpha cancer therapy.. Brain Behav.Immun. 18,
205213.
Capuron, L., Fornwalt, F.B., Knight, B.T., Harvey, P.D., Ninan,
P.T., Miller, A.H., 2009.Does cytokine-induced depression differ
from idiopathic major depression inmedically healthy individuals?
J. Affect. Disord. 119, 181185.
Craft, T.K., DeVries, A.C., 2006. Role of IL-1 in poststroke
depressive-like behavior inmice. Biol. Psychiatry 60, 812818.
Crnic, L.S., Segall, M.A., 1992. Behavioral effects of mouse
interferons-alpha and -gamma and human interferon-alpha in mice.
Brain Res. 590, 277284.
Curran, B., OConnor, J.J., 2001. The pro-inammatory cytokine
interleukin-18impairs long-term potentiation and NMDA
receptor-mediated transmission inthe rat hippocampus in vitro.
Neuroscience 108, 8390.
Dantzer, R., OConnor, JC., Freund, G.G., Johnson, R.W., Kelley,
K.W., 2008. Frominammation to sickness and depression: when the
immune system subjugatesthe brain. Nat. Rev. Neurosci. 9, 4656.
Dantzer, R., OConnor, J.C., Lawson,M.A., Kelley, K.W., 2011.
Inammation-associateddepression: from serotonin to kynurenine.
Psychoneuroendocrinology 36, 426436.
Davis, J.M., Weaver, J.A., Kohut, M.L., Colbert, L.H., Ghaffar,
A., Mayer, E.P., 1998.Immune system activation and fatigue during
treadmill running: role ofinterferon. Med. Sci. Sports Exerc. 30,
863868.
De La Garza, R 2nd., Asnis, G.M., 2003. The non-steroidal
anti-inammatory drugdiclofenac sodium attenuates IFN-alpha induced
alterations to monoamineturnover in prefrontal cortex and
hippocampus. Brain Res. 4, 7079.
De La Garza, R 2nd., Asnis, G.M., Pedrosa, E., Stearns, C.,
Migdal, A.L., Reinus, J.F.,Paladugu, R., Vemulapalli, S., 2005.
Recombinant human interferon-alpha doesnot alter reward behavior,
or neuroimmune and neuroendocrine activation inrats. Prog.
Neuropsychopharmacol. Biol. Psychiatry 29, 781792.
Dowlati, Y., Herrmann, N., Swardfager, W., Liu, H., Sham, L.,
Reim, E.K., Lanctt, K.L.,2010. A meta-analysis of cytokines in
major depression. Biol. Psychiatry 67,446457.
Dunn, A.L., Crnic, L.S., 1993. Repeated injections of
interferon-alpha A/D in Balb/cmice. Behavioral effects. Brain
Behav. Immun. 7, 104111.
Eisenberger, N.I., Inagaki, T.K., Rameson, L.T., Mashal, N.M.,
Irwin, M.R., 2009. AnfMRI study of cytokine-induced depressed mood
and social pain: the role of sexdifferences. Neuroimage 47,
881890.
Felger, J.C., Cole, S.W., Pace, T.W., Hu, F., Woolwine, B.J.,
Doho, G.H., Raison, C.L.,Miller, A.H., 2011. Molecular signatures
of peripheral blood mononuclear cellsduring chronic interferon-a
treatment: relationship with depression andfatigue. Psychol. Med.
9, 113.
Franklin, K.B.J., Paxinos, G., 1997. A Stereotaxic Atlas of the
Mouse Brain. AcademicPress, San Diego, CA.
Friebe, A., Schwarz, M.J., Schmid-Wendtner, M., Volkenandt, M.,
Schmidt, F., Horn,M., Janssen, G., Schaefer, M., 2007. Pretreatment
levels of sTNF-R1 and sIL-6Rare associated with a higher
vulnerability for IFN-alpha-induced depressivesymptoms in patients
with malignant melanoma. J. Immunother. 30, 333337.
Fu, X., Zunich, S.M., OConnor, J.C., Kavelaars, A., Dantzer, R.,
Kelley, K.W., 2010.Central administration of lipopolysaccharide
induces depressive-like behaviorin vivo and activates brain
indoleamine 2,3 dioxygenase in murine organotypichippocampal slice
cultures. J. Neuroinammation 2, 743.
Gibb, J., Hayley, S., Poulter, M.O., Anisman, H., 2011. Effects
of stressors and immuneactivating agents on peripheral and central
cytokines in mouse strains thatdiffer in stressor responsivity.
Brain Behav. Immun. 25, 468482.
Gisslinger, H., Svoboda, T., Clodi, M., Gilly, B., Ludwig, H.,
Havelec, L., Luger, A., 1993.Interferon-alpha stimulates the
hypothalamic-pituitary-adrenal axis in vivoand in vitro.
Neuroendocrinol 57, 489495.
Haroon, E., Raison, C.L., Miller, A.H., 2012.
Psychoneuroimmunology meetsneuropsychopharmacology: translational
implications of the impact ofinammation on behavior.
Neuropsychopharmacology 37, 137162.
Hashioka, S., Klegeris, A., Schwab, C., Yu, S., McGeer, P.L.,
2010. Differentialexpression of interferon-gamma receptor on human
glial cells in vivo andin vitro. J. Neuroimmunol. 225 (12),
9199.
Immunity 31 (2013) 115127 125Hayley, S., Poulter, M.O., Meral,
Z., Anisman, H., 2005. The pathogenesis of clinicaldepression:
stressor- and cytokine-induced alterations of
neuroplasticity.Neuroscience 135, 659678.
-
andHowren, M.B., Lamkin, D.M., Suls, J., 2009. Associations of
depression with C-reactive protein, IL-1, and IL-6: a
meta-analysis. Psychosom. Med. 71, 171186.
Hrdina, P.D., Demeter, E., Vu, T.B., Sotonyi, P., Palkovits, M.,
1993. 5HT uptake sitesand 5HT2 receptors in brain of
antidepressant-free suicide victims/depressives:increase in 5HT2
sites in cortex and amygdala. Brain Res. 614, 3744.
Huang, Y., Smith, D.E., Ibez-Sandoval, O., Sims, J.E., Friedman,
W.J., 2011. Neuron-specic effects of interleukin-1b are mediated by
a novel isoform of the IL-1receptor accessory protein. Neuroscience
7, 1804818059.
Hulse, R.E., Kunkler, P.E., Fedynyshyn, J.P., Kraig, R.P., 2004.
Optimization ofmultiplexed bead-based cytokine immunoassays for rat
serum and braintissue. J. Neurosci. Methods 136, 8798.
Kamata, M., Higuchi, H., Yoshimoto, M., Yoshida, K., Shimizu,
T., 2000. Effect ofsingle intracerebroventricular injection of
alpha-interferon on monoamineconcentrations in the rat brain. Eur.
Neuropsychopharmacol. 10, 129132.
Kamm, K., Vanderkolk, W., Lawrence, C., Jonker, M., Davis, A.T.,
2006. The effect oftraumatic brain injury upon the concentration
and expression of interleukin-1beta and interleukin-10 in the rat.
J. Trauma 60, 152157.
Kaneko, N., Kudo, K., Mabuchi, T., Takemoto, K., Fujimaki, K.,
Wati, H., Iguchi, H.,Kanba, S., Tezuka, H., 2006. Suppression of
cell proliferation by interferon-alphathrough interleukin-1
production in adult rat dentate gyrus.Neuropsychopharmacology 31,
26192626.
Katafuchi, T., Kondo, T., Take, S., Yoshimura, M., 2005.
Enhanced expression of braininterferon-alpha and serotonin
transporter in immunologically induced fatiguein rats. Eur. J.
Neurosci. 22, 28172826.
Kawanokuchi, J., Mizuno, T., Takeuchi, H., Kato, H., Wang, J.,
Mitsuma, N., Suzumura,A., 2006. Production of interferon-gamma by
microglia. Multiphase Scler. 12,558564.
Kenis, G., Prickaerts, J., van Os, J., Koek, G.H., Robaeys, G.,
Steinbusch, H.W., Wichers,M., 2011. Depressive symptoms following
interferon-a therapy: mediated byimmune-induced reductions in
brain-derived neurotrophic factor? Int. J.Neuropsychopharmacol. 4,
247253.
Khorooshi, R., Owens, T., 2010. Injury-induced type I IFN
signaling regulatesinammatory responses in the central nervous
system. J Immunol. 185 (2),12581264.
Kraus, M.R., Schfer, A., Al-Taie, O., Scheurlen, M., 2005.
Prophylactic SSRI duringinterferon alpha re-therapy in patients
with chronic hepatitis C and a history ofinterferon-induced
depression. J. Viral Hepat. 12, 96100.
Kraus, M.R., Al-Taie, O., Schfer, A., Pfersdorff, M., Lesch,
K.P., Scheurlen, M., 2007.Serotonin-1A receptor gene HTR1A
variation predicts interferon-induceddepression in chronic
hepatitis C. Gastroenterology 132, 12791286.
Kumai, T., Tateishi, T., Tanaka, M., Watanabe, M., Shimizu, H.,
Kobayashi, S., 2000.Effect of interferon-alpha on tyrosine
hydroxylase and catecholamine levels inthe brain of rats. Life Sci.
30, 663669.
Leonard, B., Maes, M., 2012. Mechanistic explanations how
cell-mediated immuneactivation, inammation and oxidative and
nitrosative stress pathways andtheir sequels and concomitants play
a role in the pathophysiology of unipolardepression. Neurosci.
Biobehav. Rev. 36, 764785.
Linthorst, A.C., 2005. Interactions between
corticotropin-releasing hormone andserotonin: implications for the
aetiology and treatment of anxiety disorders.Handb. Exp. Pharmacol.
169, 181204.
Litteljohn, D., Mangano, E., Shukla, N., Hayley, S., 2009.
Interferon-gammadeciency modies the motor and co-morbid behavioral
pathology andneurochemical changes provoked by the pesticide
paraquat. Neuroscience164, 18941906.
Litteljohn, D., Cummings, A., Brennan, A., Gill, A., Chunduri,
S., Anisman, H., Hayley,S., 2010. Interferon-gamma deciency modies
the effects of a chronic stressorin mice. Implications for
psychological pathology. Brain Behav. Immun. 24,462473.
Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene
expression data usingreal-time quantitative PCR and the 2(-Delta
Delta C(T)) Method. Methods 25,402408.
Loftis, J.M., Hauser, P., 2004. The phenomenology and treatment
of interferoninduced depression. J. Affect. Disord. 82, 175190.
Loftis, J.M., Wall, J.M., Pagel, R.L., Hauser, P., 2006.
Administration of pegylatedinterferon-alpha-2a or -2b does not
induce sickness behavior in Lewis rats.Psychoneuroendocrinology 31,
12891294.
Maes, M., 1995. Evidence for an immune response in major
depression: a reviewand hypothesis. Prog. Neuropsychopharmacol.
Biol. Psychiatry 19, 1138.
Maes, M., Leonard, B.E., Myint, A.M., Kubera, M., Verkerk, R.,
2011. The new 5-HThypothesis of depression: cell-mediated immune
activation inducesindoleamine 2,3-dioxygenase, which leads to lower
plasma tryptophan andan increased synthesis of detrimental
tryptophan catabolites (TRYCATs), both ofwhich contribute to the
onset of depression. Prog. Neuropsychopharmacol. Biol.Psychiatry
35, 702721.
Maier, S.F., Nguyen, K.T., Deak, T., Milligan, E.D., Watkins,
L.R., 1999. Stress, learnedhelplessness, and brain interleukin-1
beta. Adv. Exp. Med. Biol. 461, 235249.
Makino,M., Kitano, Y., Komiyama, C., Hirohashi,M., Takasuna, K.,
2000a. Involvementof central opioid systems in human
interferon-alpha induced immobility in themouse forced swimming
test. Br. J. Pharmacol. 130, 12691274.
Makino, M., Kitano, Y., Komiyama, C., Hirohashi, M., Kohno, M.,
Moriyama, M.,Takasuna, K., 2000b. Human interferon-alpha induces
immobility in the mouseforced swimming test: involvement of the
opioid system. Brain Res. 852, 482484.
Makino, M., Kitano, Y., Komiyama, C., Takasuna, K., 2000c. Human
interferon-alpha
126 S. Hayley et al. / Brain, Behavior,increases immobility in
the forced swimming test in rats. Psychopharmacology148 (1),
106110.McNutt, M.D., Liu, S., Manatunga, A., Royster, E.B., Raison,
C.L., Woolwine, B.J.,Demetrashvili, M.F., Miller, A.H., Musselman,
D.L., 2012. Neurobehavioraleffects of interferon-a in patients with
hepatitis-C: symptom dimensions andresponsiveness to paroxetine.
Neuropsychopharmacology 37, 14441454.
Menzies, R., Phelps, C., Wiranowska, M., Oliver, J., Chen, L.,
Horvath, E., Hall, N., 1996.The effect of interferon-alpha on the
pituitary-adrenal axis. J. InterferonCytokine Res. 16, 619629.
Meyer, J.H., McMain, S., Kennedy, S.H., Korman, L., Brown, G.M.,
DaSilva, J.N., et al.,2003. Dysfunctional attitudes and 5-HT2
receptors during depression and self-harm. Am. J. Psychiatry 160,
9099.
Miller, A.H., 2009. Norman cousins lecture. Mechanisms of
cytokine-inducedbehavioral changes: psychoneuroimmunology at the
translational interface.Brain Behav. Immun. 23, 149158.
Miller, A.H., Maletic, V., Raison, C.L., 2009. Inammation and
its discontents: therole of cytokines in the pathophysiology of
major depression. Biol. Psychiatry65, 732741.
Miyahara, S., Komori, T., Fujiwara, R., Shizuya, K., Yamamoto,
M., Ohmori, M.,Okazaki, Y., 2000. Effects of repeated stress on
expression of interleukin-6 (IL-6)and IL-6 receptor mRNAs in rat
hypothalamus and midbrain. Life Sci. 66, 9398.
Morasco, B.J., Loftis, J.M., Indest, D.W., Ruimy, S., Davison,
J.W., Felker, B., Hauser, P.,2010. Prophylactic antidepressant
treatment in patients with hepatitis C onantiviral therapy: a
double-blind, placebo-controlled trial. Psychosomatics
51,401408.
Musselman, D.L., Lawson, D.H., Gumnick, J.F., Manatunga, A.K.,
Penna, S., Goodkin,R.S., Goodkin, R.S., Nemeroff, C.B., Miller,
A.H., 2001. Paroxetine for theprevention of depression induced by
high-dose interferon alfa. N. Engl. J.Med. 344, 961966.
Myint, A.M., 2012. Kynurenines: from the perspective of major
psychiatricdisorders. FEBS J. 279, 13751385.
Myint, A.M., Schwarz, M.J., Steinbusch, H.W., Leonard, B.E.,
2009. Neuropsychiatricdisorders related to interferon and
interleukins treatment. Metab. Brain Dis. 24,5568.
Navins, R., Gmez-Gil, E., Puig, S., Baeza, I., De Pablo, J.,
Martn-Santos, R., 2009.Depression in hospitalized patients with
malignant melanoma treated withinterferon-alpha-2b: primary to
induced disorders. Eur. J. Dermatol. 19, 611615.
Nguyen, K.T., Deak, T., Owens, S.M., Kohno, T., Fleshner, M.,
Watkins, L.R., Maier, S.F.,1998. Exposure to acute stress induce
brain interleukin-1b protein in the rat. J.Neurosci. 19,
27992805.
Nishi, K., Kanemaru, K., Diksic, M., 2009. A genetic rat model
of depression, Flinderssensitive line, has a lower density of
5-HT(1A) receptors, but a higher density of5-HT(1B) receptors,
compared to control rats. Neurochem. Int. 54, 299307.
OConnor, J.C., Andr, C., Wang, Y., Lawson, M.A., Szegedi, S.S.,
Lestage, J., Castanon,N., Kelley, K.W., Dantzer, R., 2009.
Interferon-gamma and tumor necrosis factor-alpha mediate the
upregulation of indoleamine 2,3-dioxygenase and theinduction of
depressive-like behavior in mice in response to
bacillusCalmette-Guerin. J. Neurosci. 29, 42004209.
Orsal, A.S., Blois, S.M., Bermpohl, D., Schaefer, M., Coquery,
N., 2008. Administrationof interferon-alpha in mice provokes
peripheral and central modulation ofimmune cells, accompanied by
behavioral effects. Neuropsychobiology 58,211222.
Pan, W., Banks, W.A., Kastin, A.J., 1997. Permeability of the
blood-brain and blood-spinal cord barriers to interferons. J
Neuroimmunol. 76 (12), 105111.
Pandey, G.N., Dwivedi, Y., Rizavi, H.S., Ren, X., Pandey, S.C.,
Pesold, C., et al., 2002.Higher expression of serotonin 5-HT(2A)
receptors in the postmortem brains ofteenage suicide victims. Am.
J. Psychiatry 159, 419429.
Parnet, P., Kelley, K.W., Bluth, R.M., Dantzer, R., 2002.
Expression and regulation ofinterleukin-1 receptors in the brain.
Role in cytokines-induced sicknessbehavior. J. Neuroimmunol. 125,
514.
Pascoe, M.C., Crewther, S.G., Carey, L.M., Crewther, D.P., 2011.
Inammation anddepression: why poststroke depression may be the norm
and not the exception.Int. J. Stroke 6, 128135.
Patten, S.B., 2006. Psychiatric side effects of interferon
treatment. Curr. Drug Saf. 1,143150.
Patterson, A.L., Morasco, B.J., Fuller, B.E., Indest, D.W.,
Loftis, J.M., Hauser, P., 2011.Screening for depression in patients
with hepatitis C using the Beck DepressionInventory-II: do somatic
symptoms compromise validity? Gen. Hosp. Psychiatry33, 354362.
Raison, C.L., Demetrashvili, M., Capuron, L., Miller, A.H.,
2005. Neuropsychiatricadverse effects of interferon-alpha:
recognition and management. CNS Drugs19, 105123.
Raison, C.L., Borisov, A.S., Majer, M., Drake, D.F., Pagnoni,
G., Woolwine, B.J., Vogt,G.J., Massung, B., Miller, A.H., 2009.
Activation of central nervous systeminammatory pathways by
interferon-alpha: relationship to monoamines anddepression. Biol.
Psychiatry 65, 296303.
Raison, C.L., Rye, D.B., Woolwine, B.J., Vogt, G.J., Bautista,
B.M., Spivey, J.R., Miller,A.H., 2010. Chronic interferon-alpha
administration disrupts sleep continuityand depth in patients with
hepatitis C: association with fatigue, motor slowing,and increased
evening cortisol. Biol. Psychiatry 68, 942949.
Rivest, S., 2009. Regulation of innate immune responses in the
brain. Nat. Rev.Immunol. 9, 429439.
Sammut, S., Bethus, I., Goodall, G., Muscat, R., 2002.
Antidepressant reversal ofinterferon-alpha-induced anhedonia.
Physiol. Behav. 75, 765772.
Immunity 31 (2013) 115127Sari, Y., 2004. Serotonin1B receptors:
from protein to physiological function andbehavior. Neurosci.
Biobehav. Rev. 28, 565582.
-
Sen, S., Duman, R., Sanacora, G., 2008. Serum brain-derived
neurotrophic factor,depression, and antidepressant medications:
meta-analyses and implications.Biol. Psychiatry 64, 527532.
Shelton, R.C., Sanders-Bush, E., Manier, D.H., Lewis, D.A.,
2009. Elevated 5-HT 2Areceptors in postmortem prefrontal cortex in
major depression is associatedwith reduced activity of protein
kinase A. Neuroscience 158, 14061415.
Shuto, H., Kataoka, Y., Horikawa, T., Fujihara, N., Oishi, R.,
1997. Repeated interferon-alpha administration inhibits
dopaminergic neural activity in the mouse brain.Brain Res. 747,
348351.
Slavich, G.M., ODonovan, A., Epel, E.S., Kemeny, M.E., 2010.
Black sheep get theblues: a psychobiological model of social
rejection and depression. Neurosci.Biobehav. Rev. 35, 3945.
Spalletta, G., Boss, P., Ciaramella, A., Bria, P., Caltagirone,
C., Robinson, R.G., 2006.The etiology of poststroke depression: a
review of the literature and a newhypothesis involving inammatory
cytokines. Mol. Psychiatry 11, 984991.
Stoll, G., Jander, S., Schroeter, M., 1998. Inammation and glial
responses inischemic brain lesions. Prog. Neurobiol. 56,
149171.
Trask, P.C., Paterson, A.G., Esper, P., Pau, J., Redman, B.,
2004. Longitudinal course ofdepression, fatigue, and quality of
life in patients with high risk melanomareceiving adjuvant
interferon. Psychooncology 13, 526536.
Underwood, M.D., Kassir, S.A., Bakalian, M.J., Galfalvy, H.,
Mann, J.J., Arango, V.,2011. Neuron density and serotonin receptor
binding in prefrontal cortex insuicide. Int. J.
Neuropsychopharmacol. 9, 113.
Wan, Q., Wang, X., Wang, Y.J., Song, L., Wang, S.H., Ho, W.Z.,
2008. Morphinesuppresses intracellular interferon-alpha expression
in neuronal cells. J.Neuroimmunol. 199, 19.
Wichers, M.C., Kenis, G., Leue, C., Koek, G., Robaeys, G., Maes,
M., 2006. Baselineimmune activation as a risk factor for the onset
of depression during interferon-alpha treatment. Biol. Psychiatry
60, 7779.
Wichers, M.C., Kenis, G., Koek, G.H., Robaeys, G., Nicolson,
N.A., Maes, M., 2007.Interferon-alpha-induced depressive symptoms
are related to changes in thecytokine network but not to cortisol.
J. Psychosom. Res. 62, 207214.
Wilkinson, J., Radkowski, M., Eschbacher, J.M., Laskus, T.,
2010. Activation of brainmacrophages/microglia cells in hepatitis C
infection. Gut 59, 13941400.
Yamada, T., Yamanaka, I., 1995. Microglial localization of
alpha-interferon receptorin human brain tissues. Neurosci. Lett.
189, 7376.
Yirmiya, R., Pollak, Y., Barak, O., Avitsur, R., Ovadia, H.,
Bette, M., Weihe, E.,Weidsenfeld, J., 2001. Effects of
antidepressant drugs on the behavioral andphysiological responses
to lipopolysaccharide (LPS) in rodents.Neuropsychopharmacology 24,
531544.
Zhu, Y., Saito, K., Murakami, Y., Asano, M., Iwakura, Y.,
Seishima, M., 2006. Earlyincrease in mRNA levels of pro-inammatory
cytokines and their interactions inthe mouse hippocampus after
transient global ischemia. Neurosci. Lett. 393,122126.
S. Hayley et al. / Brain, Behavior, and Immunity 31 (2013)
115127 127
Central administration of murine interferon- in1 Introduction2
Materials and methods2.1 Subjects2.2 Surgery2.3 Blood collection
and brain removal2.4 Plasma corticosterone determination2.5 High
performance liquid chromatography (HPLC) assay2.6 Reverse
transcription-quantitative polymerase chain reaction analysis in
brain2.7 Experiment 1: acute infusion of IFN-2.8 Experiment 2:
repeated infusion of IFN-2.9 Experiment 3: sucrose consumption
following 2.10 Experiment 4 and 5: intra-PFC infusion of I2.11
Experiment 6 and 7: intra-raphe infusion of2.12 Statistical
analyses
3 Results3.1 Experiment 1. Locomotor activity, sickness b3.2
Cytokine mRNA expression3.3 Serotonin receptor mRNA expression3.4
Experiment 2. Effects of repeated IFN- admi3.4.1 Sickness
behavior3.4.2 Plasma corticosterone3.4.3 Cytokine mRNA
expression3.4.4 Serotonin receptor mRNA expression
3.5 Experiment 3. Anhedonia following repeated a3.6 Experiments
47: intra-PFC and intra-DRN infusion of IFN-a
4 Discussion4.1 Cytokine variations in brain induced by IFN-4.2
Serotonergic alterations in the brain follow4.3 Alternative
mechanisms of IFN- induced depr
5 ConclusionAcknowledgmentReferences