Induction of anxiety-like behavior in mice during the initial stages
of infection with the agent of murine colonic
hyperplasia Citrobacter rodentium
Mark Lytea,b,⁎, Wang Lia, Noel Opitzb,c, Ronald P.A. Gaykemad, Lisa E. Goehlerd
aDepartment of Pharmacy Practice, School of Pharmacy, Texas Tech University Health Sciences Center, Lubbock, TX 79430, United States
bMinneapolis Medical Research Foundation, Minneapolis, MN 55404, United States
cDepartment of Veterinary and Biomedical Sciences, University of Minnesota-St. Paul, St. Paul, MN 55108, United States
dDepartment of Psychology, University of Virginia, Charlottesville, VA 22904, United States
Received 14 December 2005; received in revised form 30 March 2006; accepted 21 June 2006
Symptoms of anxiety frequently occur concomitant to the development and persistence of inflammatory bowel disease (IBD) in patients. In the
present study, we utilized an animal model of IBD, infection with Citrobacter rodentium, to determine whether the infection per se can drive
anxiety-like behavior. Nine-week-old CF-1 male mice were challenged orally with either saline or C. rodentium. Early in the infective process (7–
8 h later), mice were tested on a hole-board open field apparatus for anxiety-like behavior measurement. Immediately following behavioral testing,
plasma samples were obtained for immune cytokine analysis and colons were excised for histological analysis. In additional animals, vagal
ganglia were removed and processed for c-Fos protein detection. Challenge with C. rodentium significantly increased anxiety-like behavior as
evidenced by avoidance of the center area and increased risk assessment behavior. Plasma levels of the cytokines IFN-γ, TNF-α and IL-12 were
not different. However vagal sensory ganglia from C. rodentium-treated animals evinced significantly more c-Fos protein-positive neurons,
consistent with vagal afferent transmission of C. rodentium-related signals from gut to brain. Histological examination of the colon indicated a
lack of overt inflammation at the 8 h post-challenge time point, indicating that the differences in behavior were unlikely to follow from
inflammation-related stress. The results of the present study demonstrate that infection with C. rodentium can induce anxiety-like symptoms that
are likely mediated via vagal sensory neurons.
© 2006 Elsevier Inc. All rights reserved.
Keywords: Inflammatory bowel disease; Infection; Behavior; Citrobacter rodentium
Inflammatory bowel disease (IBD) represents a group of
disorders characterized by intestinal inflammation that affects
about 90 to 300 per 100,000 people in the general population
. Patients with IBD chronically suffer from diarrhea, ab-
dominal pain, gastrointestinal bleeding, malabsorption, and
weight loss requiring continuous medical and surgical attention.
Althoughmedicaland surgical treatmentoptions have improved
quality of life , studies have indicated poorer emotional
function in IBD patients than in the general population. In
addition, IBD patients may suffer from anxiety or depression
which further contributes to impaired quality of life [13,29].
Affective symptoms in IBD patients, as well as those with
other bowel disorders, have been attributed to psychological
factors including stress or personality (“top-down” mechan-
isms) [17,47], but other interoceptive factors related to IBD may
also contribute affective symptoms (‘bottom-up” mechanisms)
. For example, recent studies using a subclinical infection via
the per oral route with the food-borne human pathogen, Cam-
pylobacter jejuni, have demonstrated anxiety-like behavior in
mice  which may be associated with early activation of
vagal afferents and in brain regions associated with the central
Physiology & Behavior 89 (2006) 350–357
⁎Corresponding author. Department of Pharmacy Practice, Texas Tech
University Health Sciences Center, 3601 4th Street, STOP 8182, Lubbock, TX
79430-8162, United States. Tel.: +1 806 743 4200x262; fax: +1 806 743 4209.
E-mail address: firstname.lastname@example.org (M. Lyte).
0031-9384/$ - see front matter © 2006 Elsevier Inc. All rights reserved.
autonomic network . These findings suggest that intestinal
infections or inflammation, per se, can drive behavioral changes
that are likely mediated by extrinsic nerves that innervate the
gut, and thus may contribute to affective disorders associated
with IBD and other inflammatory disorders via “bottom-up”
mechanisms. This idea is in line with findings showing that
immune activation can modulate behavior and affective states,
including anxiety-like behavior [2,11,12,31,33,58].
Citrobacter rodentium belongs to a family of human and
animal enteric pathogens that include clinically significant
enterohemorrhagic and enteropathogenic Escherichia coli .
In contrast to C. jejuni, which is a serious pathogen in humans
but leads only to “subclinical” disease in mice, C. rodentium is
the causative agent of transmissible murine colonic hyperplasia,
a naturally occurring disease in mice [37,38]. The pattern of
symptomatology in infected animals closely mimics that ob-
served in IBD patients . As such, infection of mice with
C. rodentium has increasingly been used as a model for studying
IBD in humans [24,38,44]. The C. rodentium model has been
used to extensively study the molecular mechanisms governing
attachment and infection as well as the subsequent host re-
sponse [24,28,32,43,49,56], thus it may constitute an important
model for use in examining the contribution of naturalistic
infection to behavior.
The mechanismsthatmediate the behavioraleffects ofperiph-
eralimmunechallenge onthe brainhavebeenelucidatedoverthe
past decade largely employing infection-based models using
either live microorganisms [16,19] or acute models using micro-
bial-based products such as lipopolysaccharide [16,19,30,35,58].
Further, infectious organisms or their pro-inflammatory products
have been introduced into laboratory animals via routes, e.g.
intraperitoneal or intracerebroventricular, that often do not re-
present the naturalroute ofinfectionbywhichthe microorganism
gains entry into the body (e.g. [6,7,18,30,36]). Whereas these
studies have contributed significantly to the understanding of
mechanisms by which immune activation can influence brain
function, they cannot speak to the mechanisms by which chronic
infections may induce anxiety- or depressive-like behavior. Thus,
that utilize the natural route of infection of the microorganism.
In the experiments described here, we have employed a natural
route of infection with C. rodentium, a causative agent of colonic
hyperplasia, to examine possible early effects on behavioral re-
board test, with a view towards validating per oral infection with
C. rodentium as model to study affective changes in inflammatory
bowel disorders. In addition, we examined potential pathways by
which C. rodentium-induced signals reach the brain, by assessing
circulating cytokines typically associated with C. rodentium in-
fection , as well as induction of c-Fos protein (an activation
marker) in vagal sensory neurons. The use of the hole-board/open
field test has beenwell established as a behavioral tool with which
to examine anxiety-like behavior in rodents [22,26]. We have
chosen to examine an early time point in the infectious process
before the onset of inflammation, colon crypt hyperplasia and
sickness symptoms in order to reduce the possibility that pain
associated with the overt disease could account for behavioral
changes observed. We show that C. rodentium challenge is asso-
ciated with avoidance of the center regions of the open/field hole-
board, concomitant with increased risk assessment behavior, con-
sistent withanxiety-like rather than“sickness” behavior. Although
there was no evidence of increased circulating cytokines, vagal
afferents showed increased c-Fos protein expression, consistent
with a neural route of immune-to-brain signaling in this experi-
2. Materials and methods
Thirty-six, 5-week-old CF-1 male mice were purchased from
Charles River Laboratories (Wilmington, DE). Upon arrival, the
mice were housed one animal per cage (width: 16 cm, length:
22 cm, and height: 13 cm) and placed in a 12-h light/dark cycle
(dark: 2100–0900 h, light: 0900–2100 h). Food and water were
available ad libitum. All animals were maintained in the animal
facility for 4 weeks before use in experimental protocols. All
treatment procedures were approved by the Institutional Animal
Care and Use Committee.
2.2. Per oral challenge
Two days before challenge, a stock culture of C. rodentium
(#51459, American Type Culture Collection, Manassas, VA) was
use in challenge experiments, approximately 100 colony forming
units (CFU) of C. rodentium was inoculated into LB broth sup-
MO). Previous reports have shown that incubation of enteric
bacteria in norepinephrine as well as other catecholamine con-
taining medium increases both growth and expression of
virulence-related properties as well as providing a growth envi-
ronment more reflective of the in vivo milieu [40,41,54]. A
control tube containing norepinephrine-supplemented LB but
without bacteria was also setup. Both bacteria containing and
control tubes were incubated overnight at 37 °C in a 5% CO2
incubator. On the day of the experiment, both the C. rodentium-
containing and control tubes were centrifuged at 5000×g
for 10 min, the supernatant discarded, and the bacterial pellet
(C. rodentium tube) or not (control tube) resuspended in pre-
warmed phosphate-buffered saline. The tubes were centrifuged
and washed an additional 2×. Based on optical density the con-
centration of C. rodentium was adjusted to a final density of
approximately 108CFU/ml. Using a ball-tipped, stainless steel
feeding needle mice (n=8 per treatment) were per orally
challenged with 0.2 ml ofeitherC. rodentium-containing solution
approximately 2×107CFU per mouse) or control saline solution.
To confirm infection with Citrobacter, fecal pellets collected at
sacrifice were first homogenized in sterile saline and then spread
plated onto MacConkey agar and incubated overnight at 37 °C in
a 5% CO2 humidified incubator. Individual colonies were
then examined using the Api20E biochemical identification sys-
tem (bioMerieux Inc., Durham, NC) to confirm the presence of
351 M. Lyte et al. / Physiology & Behavior 89 (2006) 350–357
2.3. Experiment 1: behavior assessment with the open field
To minimize interference, the last cage and bedding changes
three acclimation periods to the test room and testing environment
animals to the testing site one cage at a time, removing them
time, animals were exposed to sounds (equipment), lighting, and
smells (alcohol cleaning solution) of the test room. The only
novelty on test day was the hole-board enclosure itself. The ac-
climation was performed a total of 3 times on 3 separate days for a
to 96 h prior to infectious challenge and subsequent behavioral
testing. A total of three separate experiments were conducted with
group sizes of 4 animals from each treatment group (8 animals
tested per day per experiment, for a total of 24 animals).
Anxiety-like behaviors were assessed in each subject using an
infrared photobeam hole-board open field apparatus manufac-
tured by Hamilton Kinder LLC (Poway, CA) approximately 7–
8 h following per oral challenge with C. rodentium-containing or
control solutions. The open field consisted of a square enclosure
(40×40 cm2) with a smooth, opaque floor. Nine holes (about
2.5 cm in diameter and 7.5 cm deep) were evenly spaced in the
apieceoffruit loopcerealasa possibleenticingfoodreward.The
center zone was defined as the 20×20 cm2area in the middle of
the enclosure. The part of the arena outside the central zone was
defined as the periphery zone adjoining the perimeter of the
enclosure (width: 10 cm and length: 40 cm). Automated data
acquisitionona real-time basis fromthe hole-boardwasachieved
through connection via a SBC-I/O converter (Hamilton–Kinder
Model MM100CC) to a computer in the test room. In addition,
each session was recorded with a video camera.
The testing room was lit with a diffuse overhead light rated at
100 lx. The testing apparatus was initially cleaned with 20%
isopropyl alcohol solution and thoroughly dried. This cleaning
onthehole-boardwas initiated byplacementofthe animals inthe
of the open field. The total time animals were monitored on the
hole-board was 5 min.
2.3.2. Behavior scoring
The breakage of individual photo beam paths on the open field
enabled detection of the animal's locomotor behavior on the open
field, which was subdivided into total distance traveled, distance
traveled in the center area, time spent in the center area, and “rest”
of the animal does not move for more than 1 s). Total numbers of
hole pokes and pokes into individual holes were also recorded to
assess pattern of exploration. Risk assessment behaviors were
scored manually from video recordings by a rater blind to experi-
defined as episodes in which the animal stretched forward, sniff-
ing, while leaving its hind feet stationary. These episodes were
followed either by exploration or by retreat, which as was also
2.3.3. Tissue preparation
Following behavioral measurement on the open field, animals
were taken to a separate procedure room for tissue preparation.
Animals were deeply anesthetized with the use of the inhalant
isoflurane and cardiac puncture was performed with heparinized
syringes to obtain blood samples for cytokine analysis. The
animals were then sacrificed by cervical dislocation and sections
of distal colon were excised and transferred into 10% neutral
buffered formalin for histological analysis. All tissue sections
were coded for blinded analysis by an independent cytopathol-
ogist. Seven separate sections per colon were analyzed.
2.3.4. Cytokine measurements
C. rodentium infection induces the cytokines tumor necrosis
factor-α (TNF-α), interferon-γ (IFN-γ), and interleukin-12 (IL-
12) . To determine whether these cytokines may contribute
to gut-to-brain signaling in early C. rodentium infection, plasma
was obtained from blood samples following centrifugation at
10,000×g and stored at −80 °C until analysis. The levels of
IFN-γ, TNF-α and IL-12 were measured using specific ELISA
kits (BD Pharmingen, San Diego, CA) as per manufacturer’s
instructions. All samples for any one cytokine were analyzed at
the same time. The detection limits of the IFN-γ and IL-12
assays were 25.38 pg/ml and for TNF-α 1.94 pg/ml.
2.3.5. Colon histology
Tissue sections were incubated overnight in plastic cassettes
in the formalin substitute Histochoice (Amresco, Solon, OH).
Tissue was cut using a Shandon AS325 microtome (Thermo-
Electron, Pittsburgh, PA) set at 4 μm. Tissue staining was
manually performed using hematoxylin and eosin stains
(Richard-Allan Scientific, Kalamazoo, MI). Colonic sections
were assessed for morphological change or active inflammation
including cryptitis or crypt abscesses, granulomata, increased
to group identification (services provided by Dr. David S. Mehr,
Central Utah Pathology LLC, American Fork, UT).
2.3.6. Statistical analysis
Results are presented as mean±S.E.M. Data were analyzed
using one-way analysis of variance (ANOVA; Statview 5).
2.4. Experiment 2: c-Fos protein induction in vagal sensory
2.4.1. Tissue preparation
In order to determine whether vagal afferents may provide a
pathway by which infection in the gut drives anxiety-like
behavior, twelve mice were inoculated with C. rodentium (n=5)
or saline (n=6), as above. Between 8 and 9 h later, they were
anesthetized with isoflurane, and transcardially perfused with
saline followed by buffered paraformaldehyde solution (4% in
352M. Lyte et al. / Physiology & Behavior 89 (2006) 350–357
0.1 M phosphate buffer, pH 7.4). The heads were stored in
fixative solution overnight at 4 °C. Vagal ganglia were dissected
bilaterally, cryoprotected, sectioned at 16 μm on a cryostat and
thaw-mounted onto charged glass slides (Super-frost Plus,
Fisher Scientific). The slides were processed immediately for
immunohistochemical detection of c-Fos protein induction.
Immunoreactivity for c-Fos protein was visualized with the
immunoperoxidase method using the avidin–biotin labeling
procedure and nickel-intensified staining with 3′3-diaminobenzi-
underwent pre-incubations in the Avidin–Biotin blocking kit
(Vector Labs). Both the avidin and biotin solutions were diluted
1:10 and applied sequentially on the slides and incubated for 1 h.
Subsequently, sections were incubated in the following order of
solutions made up of 2% normal goat serum and 0.1% sodium
azide: (1) rabbit anti-Fos (Ab5, Oncogene, Cambridge, MA;
1:5000) for overnight at 4 °C; (2) biotinylated goat anti-rabbit IgG
(Jackson ImmunoResearch; 1:200) for 2 h at 4 °C; (3) avidin–
biotin peroxidase complex (ABC Elite kit, Vector Labs, Burlin-
game, CA; 1:200) for 2 h.The ABC reaction was performed using
a glucose oxidase modification of the peroxidase staining with
stain of c-Fos-immunoreactive profiles. Specificity of the immu-
nostaining was carried out in sections that were treated identically
to the above but in which the primary antisera were omitted.
2.4.2. Microscopy and image analysis
The c-Fos immunostaining in the vagal ganglia was
evaluated using a BX51 Olympus microscope. Quantification
of c-Fos expression in vagal sensory neurons was performed
manually as previously described . Briefly, neuronal nuclei
were considered positive for c-Fos when the staining of the
nucleus was darker than the surrounding cytoplasm. Sections
counted were collected from two non-adjacent one-in-four
series, thus the section counted were 32 μm apart.
2.4.3. Statistical analysis
Statistical analyses were based on the mean number of cumu-
one-way analysis of variance (ANOVA; Statview 5).
3.1. Behavioral testing on open-field/hole-board apparatus
Treatment with C. rodentium was followed by markedly dif-
ferent behaviors in the open field/hole-board apparatus 7–8 h
later, compared to saline-treated controls (Fig. 1). C. rodentium-
Fig. 1. Reduction in center zone exploration by mice per orally infected with
C. rodentium. In comparison to the saline-treated group, infected mice showed
reduced number of entries in the center zone (A), spent less cumulative time in the
center zone (B), and traveled a distance within the center zone that was strongly
reducedby76.5%(C).D:Thetotaldistancetraveledinthe entire hole-boardarena
was also moderately reduced by 27.5%. Data are presented as mean±S.E.M.;
Fig. 2. Mice infected per orally with C. rodentium displayed a reduced number of pokes into the holes (A,⁎pb0.01), but also showed a different exploratory style as
indicatedby a stronger preference for the corner hole #1 that is close to where they were introduced into the hole-board (marked with X in the hole-board diagram), and
a reduced preference for especially the side and center holes (B). The respective holes that the diagram legends refer to are also indicated in the same grey tones inside
the hole-board diagram in the lower right.
353 M. Lyte et al. / Physiology & Behavior 89 (2006) 350–357
treated animals avoided the center region of the open field/hole-
board when compared to controls, as they exhibited fewer entries
(13.58 cm vs. 161.16 cm) [F(1,23)=7.97, pb0.01], and spent
less time in the central area (4.25 s vs. 12.22 s) [F(1,23)=5.43,
pb0.03]. The total distance traveled in the entire arena by the
C. rodentium-treated animals was reduced in comparison to the
saline-treated mice (17.19 m vs. 23.7 m), [F(1,23)=10.45,
p=0.003], which was related to the reduced tendency to
explore as illustrated below. When taking the distance traveled
in the entire arena into account, the proportion of the distance
traveled in the center zone for each mouse was still sig-
nificantly reduced in the C. rodentium-treated mice as com-
pared to the saline-treated animals (5.67% vs. 1.63%,
respectively; F(1,23)=6.43, pb0.02).
C. rodentium-challenged animals spent more time “resting”
(37.4±6.2 s) as compared to saline-treated controls (20.8±
3.1 s) [F(1,23)=5.42; p=0.029]. During the brief “rest” epi-
sodes, C. rodentium-treated animals were either backed into one
of the corners, assuming a ball-like posture while sniffing the
air, or were poking into or sniffing a hole, most frequently the
corner hole (#1) that was closest to where they were placed into
the arena (Fig. 2). Indeed, saline-treated animals poked into the
other holes more frequently than C. rodentium-treated animals
(the other 3 corner holes: 23.5±1.5 vs. 15.4±2.3 [F(1,23)=8.41,
7.74, p=0.01] as depicted in Fig. 2. In contrast, C. rodentium-
treated animals poked into corner hole #1 slightly more (10.8±
1.6)thandid the saline-treatedanimals (7.6±1.1)[F(1,23)=2.55,
p=0.12]. In addition, C. rodentium-treated animals engaged in
12.23, p=0.002], and these behaviors were more likely to
result in retreat than in exploration (retreats: 17.4±2.9 vs. 6.2±
1.1) [F(1,23)=12.45, p=0.002] (Fig. 3). At no time did any
animal exhibit signs of overt sickness such as lethargy, ruffled or
matted fur, labored breathing, hunched posture, or prolonged
3.2. Plasma cytokine levels
There were no differences in the plasma levels of TNF-α,
IFN-γ, or IL-12 between C. rodentium- and saline-challenged
animals. Levels of TNF-α were below the detection level
of 1.94 pg/ml in all animals, with the exception of one C.
rodentium-treated animal (2.74 pg/ml). Similarly, the levels of
IFN-γ were below the detection limit of 25.38 pg/ml with the
exception of one control (80.95 pg/ml) and one C. rodentium-
treated animal (304.25 pg/ml). The levels of IL-12 were below
the detection limit of the assay in all animals. Thus, neither
C. rodentium nor saline treatment was associated with induction
of these cytokines in the circulation.
3.3. Histological analysis of colon
Analysis of colonic tissue by a cytopathologist blinded to
group identification did not reveal any significant crypt hyper-
plasia (including cryptitis or crypt abscesses), granulomata, nor
architectural glandular distortion of the crypts (glands) in both
control and C. rodentium-challenged mice (Fig. 4). No evidence
Fig. 4. Histological analysis of distal colon tissue sections excised 7–8 h following challenge. Neither colonic crypt hyperplasia nor inflammatory infiltrate (e.g.
immune cells) were observed in any of the tissue sections (7 sections per colon) examined across animals from both saline (A and B) and C. rodentium (not shown)
challenged mice. Black arrow: glandular epithelium; Red arrow: lamina propria; Yellow arrow: muscularis propria. (Magnification: 4× for A; 40× for B).
Fig. 3. Mice infected per orally with C. rodentium displayed a larger number of
risk assessment behaviors, defined as stretched approach postures with their
hind feet remaining in place, and which were followed by retreat or withdrawing
their front legs (⁎pb0.005). There is no difference in the number of risk
assessment behaviors that were followed by forward movement. The insert
graph shows the total number of risk assessment behaviors (⁎pb0.005).
354 M. Lyte et al. / Physiology & Behavior 89 (2006) 350–357
of infiltrating inflammatory cells was observed for either group.
of mice utilizing the same dosage of C. rodentium employed in
the present study showed development of severe colonic hyper-
plasia in all bacterial-challenged animals by 14 days post-
challenge, as has been reported by others [25,37,38]. Further,
microbiological analysis of selected cecal contents from saline
and C. rodentium-challenged mice in the present study con-
firmed the presence of C. rodentium in C. rodentium-challenged
animals, but not in saline controls.
3.4. c-Fos induction in vagal sensory neurons
Very low basal numbers of c-Fos positive cells were seen
in the vagal ganglia from saline-treated animals (mean 2.2±
0.4 per case). However, vagal ganglia from C. rodentium-
treated animals showed an elevated number of c-Fos ex-
pressing vagal sensory neurons bilaterally (mean 23.8±3.5
per case; Fig. 5). This difference was significant [F(1,9)=
The results presented here demonstrate that during the initial
phase of infection with C. rodentium, mice developed anxiety-
like behavior, as evidenced by avoidance of the center area,
reduced exploration of the holes, and increased risk assessment
behavior on the hole-board open field apparatus. Although
induction of circulating pro-inflammatory cytokines was not ob-
served, C. rodentium, induced c-Fos protein, an activation mark-
was assessed, providing the possibility that infection-related
signals are carried by extrinsic nerves innervating the gut thatcan
drive behavioral changes typical of increased “risk aversion”
during exploration. It is important to note that there were no
indices (e.g. hunched posture, ruffled or matted fur, immobility),
associated with sickness-induced behavioral depression. Thus,
the pattern of behavior change following C. rodentium treatment
is more consistent with anxiety-like behavior.
Although the differences in behavior between C. rodentium-
infected mice and saline-treated controls are consistent with
anxiety-like behavior, the possibility exists that pain associated
with colonic inflammation may account for these behavioral
differences. However, no histological changes occurred in C.
rodentium-infected mice at the time-points assessed, nor did any
of the animals assume a posture typical of pain or sickness, such
as hunching or arching of the back. In addition to the lack of
histopathological changes in the early stage of C. rodentium-
challenged mice, measurements of immune pro-inflammatory
cytokines inthe plasma (TNF-α,IFN-γ,andIL-12)also showed
no differences betweeninfected and controlanimalsat the 7–8h
post-exposure time point, indicating that a systemic inflamma-
tory response is unlikely to account for the behavioral effects of
C. rodentium observed. Nevertheless, vagal sensory neurons
showed evidence of activation, which is consistent with a neural
pathway from the gut to the brain that is capable of influencing
behavior duringinfection. It is stressed, however, that the failure
to observe inflammatory cytokine production in the present
study does not exclude the possibility that immune stimulation
occurred within the gut and that local inflammatory mediators
IL-1 and TNF-α [15,23], as well as for prostaglandins [46,50].
These observations suggest that extrinsic nerves, notably those
of the vagus are sensitive to early signals arising from gut
infection, and that anxiety-like behavior constitutes an early
response to peripheral infection. In this way, anxiety may serve
as an early warning signal, perhaps to encourage an animal to
find safe locations in which to recuperate.
or inflammatory conditions. For example, infection of mice with
Schistosoma mansoni resulted in reduced levels of exploration
and grooming . In addition, bacterial infection or response to
Fig. 5. Colonic infection with C. rodentium induces c-Fos immunoreactivity in the sensory neurons of the vagus nerve, indicated by the dark grey-black nuclear
staining (depicted by arrows in A). Staining for c-Fos immunoreactivity was virtually absent in the vagal sensory ganglia of the saline-treated mice (B). C: Semi-
quantitative analysis revealed a significant increase in c-Fos-positive cells within the vagal ganglia of C. rodentium-treated mice in comparison to saline-treated mice
(⁎⁎pb0.005). The cumulative number of cells were counted in a total of 19.8±1.3 sections for the saline-treated and 20.8±1.2 sections for the infected mice (no
significant difference in the number of sections counted, p=0.61).
355 M. Lyte et al. / Physiology & Behavior 89 (2006) 350–357
bacterial products has also been reported to induce behavioral
changes that possibly reflect changes in emotional variables such
volunteers with bacterial lipopolysaccharide was associated with
elevatedscores on ananxietyscale early onduringexposure.
Taken together with the findings presented here, it is evident that
infection or inflammation can induce changes in behavior or
affective state beyond the “classic” sickness symptoms.
Affective symptoms, especially anxiety, are consistently
associated with IBD and other bowel disorders [5,9,14]. The
contributing factors that likely drive the occurrence of these
symptoms are complicated, and involve pain associated with
inflammation, stress from stigma and inconvenience, as well as
potential predisposition for, or previous diagnosis of, affective
disorder [17,39,48,55,57]. In general, however, findings have
indicated that most anxieties occur concomitant with inflamma-
tory episodesofboweldisease,diminishing withremission .
Assessment of “trait” vs. “state” anxiety, furthermore, has indi-
cated that as a population, IBD patients evince more “state”
anxiety than controls, but do not differ on measures of trait
anxiety . Thus, in the context of the findings presented here,
the contribution of immune activation (per se) to anxiety
associated with bowel disorders may indeed be significant,
perhaps as part of a “positive-feedback loop” in which immune-
driven anxiety contributes to stress, which then drives
exacerbation of disease, including pain and inflammation,
which increases anxiety, and so on.
The use of C. rodentium infected mice as a model for
studying human IBD [24,25,38] provides for a model system
that can enable the dissection of the relevant mechanisms
underlying affective changes during bowel disease. Although
there has been an increase in the use of C. rodentium as an
animal model of IBD [32,56,59], there has not previously been
any report of its use in behavioral testing. The present study
represents the first report demonstrating that behavioral altera-
tions can occur at very early stages during the infective process.
This in vivo intestinal infection model has particular relevance
for human IBD, and permits the investigation of pathogen–host
interaction under ecologically relevant physiological conditions
of the intestinal environment. Research on the mechanism of C.
rodentium infection-induced behavioral change may help to
explain the affective symptomatology that has been observed in
IBD patients  as well as provide insights more generally
related to mechanisms by which peripheral inflammation or
infection can influence affective states.
This work was supported by National Insititutes of Health
grant MH-50431 to M.L.
 Addolorato G, Stefanini GF, Capristo E, Caputo F, Gasbarrini A,
Gasbarrini G. Anxiety and depression in adult untreated celiac subjects
and in patients affected by inflammatory bowel disease: a personality
“trait” or a reactive illness? Hepatogastroenterology 1996;43:1513–7.
 Anisman H, Merali Z. Anhedonic and anxiogenic effects of cytokine
exposure. Adv Exp Med Biol 1999;461:199–233.
 Bauhofer A, Schwarting RK, Koster M, Schmitt A, Lorenz W, Pawlak CR.
Sickness behavior of rats with abdominal sepsis can be improved by
antibiotic and G-CSF prophylaxis in clinic modeling randomized trials.
Inflamm Res 2004;53:697–705.
 Berntson GG, Sarter M, Cacioppo JT. Ascending visceral regulation of
cortical affective information processing. Eur J Neurosci 2003;18:2103–9.
 Blanchard EB, Scharff L, Schwarz SP, Suls JM, Barlow DH. The role of
anxiety and depression in the irritable bowel syndrome. Behav Res Ther
 Bluthe RM, Bristow A, Lestage J, Imbs C, Dantzer R. Central injection of
interleukin-13 potentiates LPS-induced sickness behavior in rats. Neu-
 Bluthe RM, Dantzer R, Kelley KW. Effects of interleukin-1 receptor
antagonist on the behavioral effects of lipopolysaccharide in rat. Brain Res
 Casati J, Toner BB. Psychosocial aspects of inflammatory bowel disease.
Biomed Pharmacother 2000;54:388–93.
 Crane C, Martin M. Social learning, affective state and passive coping in
irritable bowel syndrome and inflammatory bowel disease. Gen Hosp
 Dahan S, Wiles S, La Ragione RM, Best A, Woodward MJ, Stevens MP,
et al. EspJ is a prophage-carried type III effector protein of attaching and
effacing pathogens that modulates infection dynamics. Infect Immun
 Dantzer R. Cytokine-induced sickness behavior: mechanisms and
implications. Ann N YAcad Sci 2001;933:222–34.
to activation of innate immunity. Eur J Pharmacol 2004;500:399–411.
 Dunlop SP, Jenkins D, Neal KR, Spiller RC. Relative importance of
enterochromaffin cell hyperplasia, anxiety, and depression in postinfec-
tious IBS. Gastroenterology 2003;125:1651–9.
 Ek M, Kurosawa M, Lundeberg T, Ericsson A. Activation of vagal
afferents after intravenous injection of interleukin-1beta: role of endog-
enous prostaglandins. J Neurosci 1998;18:9471–9.
 Fiore M, Moroni R, Alleva E, Aloe L. Schistosoma mansoni: influence of
infection on mouse behavior. Exp Parasitol 1996;83:46–54.
 Folks DG. The interface of psychiatry and irritable bowel syndrome. Curr
Psychiatry Rep 2004;6:210–5.
 Gibertini M. Cytokines and cognitive behavior. Neuroimmunomodulation
 Gibertini M, Newton C, Friedman H, Klein TW. Spatial learning
impairment in mice infected with Legionella pneumophila or administered
exogenous interleukin-1-beta. Brain Behav Immun 1995;9:113–28.
 Goehler LE, Gaykema RP, Hammack SE, Maier SF, Watkins LR.
Interleukin-1 induces c-Fos immunoreactivity in primary afferent neurons
of the vagus nerve. Brain Res 1998;804:306–10.
 Goehler LE, Gaykema RP, Opitz N, Reddaway R, Badr N, Lyte M.
Activation in vagal afferents and central autonomic pathways: early
responses to intestinal infection with Campylobacter jejuni. Brain Behav
 Harro J, Pold M, Vasar E. Anxiogenic-like action of caerulein, a CCK-
treatment. Naunyn Schmiedebergs Arch Pharmacol 1990;341:62–7.
and spinal sensory processing. Curr Pharm Des 2005;11:1391–409.
 Higgins LM, Frankel G, Connerton I, Goncalves NS, Dougan G,
MacDonald TT. Role of bacterial intimin in colonic hyperplasia and
inflammation. Science 1999;285:588–91.
 Higgins LM, Frankel G, Douce G, Dougan G, MacDonald TT. Citrobacter
rodentium infection in mice elicits a mucosal Th1cytokine response and
lesions similar to those in murine inflammatory bowel disease. Infect
 Hotchin J, Seegal R. Alterations in behavior resulting from persistent
lymphocytic choriomeningitis virus infection. Birth Defects Orig Artic Ser
356M. Lyte et al. / Physiology & Behavior 89 (2006) 350–357
 Husain A, Triadafilopoulos G. Communicating with patients with Download full-text
inflammatory bowel disease. Inflamm Bowel Dis 2004;10:444–50
 Iimura M, Gallo RL, Hase K, Miyamoto Y, Eckmann L, Kagnoff MF.
Cathelicidin mediates innate intestinal defense against colonization with
epithelial adherent bacterial pathogens. J Immunol 2005;174:4901–7.
 Irvine EJ. Review article: patients' fears and unmet needs in inflammatory
bowel disease. Aliment Pharmacol Ther 2004;20(Suppl 4):54–9.
 Johnson RW, Curtis SE, Dantzer R, Bahr JM, Kelley KW. Sickness
behavior in birds caused by peripheral or central injection of endotoxin.
Physiol Behav 1993;53:343–8.
 Kelley KW, Bluthe RM, Dantzer R, Zhou JH, Shen WH, Johnson RW,
et al. Cytokine-induced sickness behavior. Brain Behav Immun
 Klapproth JM, Sasaki M, Sherman M, Babbin B, Donnenberg MS,
Fernandes PJ, et al. Citrobacter rodentium lifA/efa1 is essential for colonic
colonization and crypt cell hyperplasia in vivo. Infect Immun 2005;73:
 Konsman JP, Parnet P, Dantzer R. Cytokine-induced sickness behaviour:
mechanisms and implications. Trends Neurosci 2002;25:154–9.
 Kusnecov AW, Goldfarb Y. Neural and behavioral responses to systemic
immunologic stimuli: a consideration of bacterial T cell superantigens.
Curr Pharm Des 2005;11:1039–46.
 Langhans W. Bacterial products and the control of ingestive behavior:
clinical implications. Nutrition 1996;12:303–15.
 Linthorst AC, Flachskamm C, Muller-Preuss P, Holsboer F, Reul JM.
Effect of bacterial endotoxin and interleukin-1 beta on hippocampal
serotonergic neurotransmission, behavioral activity, and free corticoste-
rone levels: an in vivo microdialysis study. J Neurosci 1995;15:2920–34.
 Luperchio SA, Newman JV, Dangler CA, Schrenzel MD, Brenner DJ,
Steigerwalt AG, et al. Citrobacter rodentium, the causative agent of
transmissible murine colonic hyperplasia, exhibits clonality: synonymy of
C. rodentium and mouse-pathogenic Escherichia coli. J Clin Microbiol
 Luperchio SA, Schauer DB. Molecular pathogenesis of Citrobacter
rodentium and transmissible murine colonic hyperplasia. Microbes Infect
 Lydiard RB. Irritable bowel syndrome, anxiety, and depression: what are
the links? J Clin Psychiatry 2001;62(Suppl 8):38–45 [discussion 46-37].
 Lyte M, Arulanandam BP, Frank CD. Production of Shiga-like toxins by
Escherichia coli O157:H7 can be influenced by the neuroendocrine
hormone norepinephrine. J Lab Clin Med 1996;128:392–8.
 Lyte M, Erickson AK, Arulanandam BP, FrankCD, Crawford MA, Francis
DH. Norepinephrine-induced expression of the K99 pilus adhesin of
enterotoxigenic Escherichia coli. Biochem Biophys Res Commun
 Lyte M, Varcoe JJ, Bailey MT. Anxiogenic effect of subclinical bacterial
infection in mice in the absence of overt immune activation. Physiol Behav
 Maaser C, Housley MP, Iimura M, Smith JR, Vallance BA, Finlay BB, et
al. Clearance of Citrobacter rodentium requires B cells but not secretory
immunoglobulin A (IgA) or IgM antibodies. Infect Immun 2004;72:
 MacDonald TT, Frankel G, Dougan G, Goncalves NS, Simmons C. Host
defences to Citrobacter rodentium. Int J Med Microbiol 2003;293:
 Mahida YR, Rolfe VE. Host–bacterial interactions in inflammatory bowel
disease. Clin Sci (Lond) 2004;107:331–41.
 Matsumoto S, Ikeda M, Yoshida S, Tanimoto T, Takeda M, Nasu M.
Prostaglandin E2-induced modification of tetrodotoxin-resistant Na+
currents involves activation of both EP2 and EP4 receptors in neonatal
rat nodose ganglion neurones. Br J Pharmacol 2005;145:503–13.
 Mayer EA, Naliboff BD, Chang L, Coutinho SV. Stress and irritable bowel
syndrome. Am J Physiol Gastrointest Liver Physiol 2001;280:G519–24.
 Miller V, Hopkins L, Whorwell PJ. Suicidal ideation in patients with
irritable bowel syndrome. Clin Gastroenterol Hepatol 2004;2:1064–8.
 Mundy R, Petrovska L, Smollett K, Simpson N, Wilson RK, Yu J, et al.
Identification of a novel Citrobacter rodentium type III secreted protein,
EspI, and roles of this and other secreted proteins in infection. Infect
 Nakamura K, Kaneko T, Yamashita Y, Hasegawa H, Katoh H, Negishi M.
Immunohistochemical localization of prostaglandin EP3 receptor in the rat
nervous system. J Comp Neurol 2000;421:543–69.
 Nataro JP, Kaper JB. Diarrheagenic Escherichia coli. Clin Microbiol Rev
 Porcelli P, Leoci C, Guerra V. A prospective study of the relationship
between disease activity and psychologic distress in patients with
inflammatory bowel disease. Scand J Gastroenterol 1996;31:792–6.
 Reichenberg A, Yirmiya R, Schuld A, Kraus T, Haack M, Morag A, et al.
Cytokine-associated emotional and cognitive disturbances in humans.
Arch Gen Psychiatry 2001;58:445–52.
 Sperandio V, Torres AG, Jarvis B, Nataro JP, Kaper JB. Bacteria–host
communication: the language of hormones. Proc Natl Acad Sci U S A
 Sykes MA, Blanchard EB, Lackner J, Keefer L, Krasner S. Psychopa-
thology in irritable bowel syndrome: support for a psychophysiological
model. J Behav Med 2003;26:361–72.
 Vallance BA, Deng W, De Grado M, Chan C, Jacobson K, Finlay BB.
Modulation of inducible nitric oxide synthase expression by the attaching
and effacing bacterial pathogen Citrobacter rodentium in infected mice.
Infect Immun 2002;70:6424–35.
 Walker EA, Gelfand MD, Gelfand AN, Creed F, Katon WJ. The
relationship of current psychiatric disorder to functional disability and
distress in patients with inflammatory bowel disease. Gen Hosp Psychiatry
 Watkins LR, Maier SF, Goehler LE. Cytokine-to-brain communication: a
review and analysis of alternative mechanisms. Life Sci 1995;57:1011–26.
 Wiles S, Clare S, Harker J, Huett A, Young D, Dougan G, et al. Organ
specificity, colonization and clearance dynamics in vivo following oral
challenges with the murine pathogen Citrobacter rodentium. Cell
357 M. Lyte et al. / Physiology & Behavior 89 (2006) 350–357