Gut colonization by Candida albicans aggravates inflammation in the gut and extra-gut tissues in mice

Abstract

We examined whether Candida albicans gut colonization aggravates immune diseases in mice. Chronic and latent C. albicans gut colonization was established by the intragastric inoculation of C. albicans in mice fed as part of a purified diet. Allergic diarrhea was induced by repetitive intragastric administration of ovalbumin
in sensitized BALB/c mice. Contact hypersensitivity was evaluated by measuring ear swelling after topical application of 2,
4-dinitrofluorobenzene in NC/Nga mice. Arthritis was induced by intradermal injection of bovine type-II collagen emulsified
with complete Freund's adjuvant in DBA/1J mice. C. albicans gut colonization increased the incidence of allergic diarrhea, which was accompanied by gut hyperpermeability, as well as
increased infiltration of inflammatory cells in the colon. Contact hypersensitivity was also exacerbated by C. albicans gut colonization, as demonstrated by increased swelling, myeloperoxidase activity, and proinflammatory cytokines in ear auricles.
Furthermore, C. albicans gut colonization promoted limb joint inflammation in collagen-induced arthritis, in an animal model of rheumatoid arthritis.
These findings suggest that C. albicans gut colonization in mice aggravates inflammation in allergic and autoimmune diseases, not only in the gut but also in the
extra-gut tissues and underscores the necessity of investigating the pathogenic role of C. albicans gut colonization in immune diseases in humans.

Full text

immune dysfunction, coupled with multiple allergies,
contributed to this list of clinical complaints. For exam-
ple, it has been hypothesized that excessive colonization
by C. albicans in the gastrointestinal mucosa may be
aggravating factors in atopic dermatitis [6 – 8]. However,
in a recent review, Goldman and Huffnagle described that
the data are not de nitive and that more research is needed
to understand any cause-and-effect relationships between
C. albicans gut colonization and a variety of hypersensi-
tivity diseases including allergy [9].
In animal studies, Noverr et al . reported that C. albicans
gut colonization promotes allergic airway in ammation in
response to mold spore ( Aspergillus fumigatus ) or ovalbu-
min (OVA) in antibiotic-treated immunocompetent mice
[10,11]. Because these mice repeatedly received intranasal
antigen administration without systemic immunization, the
authors mentioned that C. albicans gut colonization cou-
pled with antibiotic treatment can disrupt normal airway
Received 2 April 2010; Received in  nal revised form 9 July 2010;
Accepted 23 July 2010
Correspondence: Kei Sonoyama, Laboratory of Food Biochemistry,
Research Faculty of Agriculture, Hokkaido University, Kita-9, Nishi-9,
Kita-ku, Sapporo-shi, Hokkaido 060-8589, Japan. Tel. & fax: ⫹ 81 11 706
2496; E-mail: ksnym@chem.agr.hokudai.ac.jp
Gut colonization by Candida albicans aggravates
in ammation in the gut and extra-gut tissues in mice
KEI SONOYAMA * , ATSUKO MIKI †
, RYUSUKE SUGITA
†
, HARUKA GOTO †
, MAYUMI NAKATA
†
& NATSU YAMAGUCHI *
* Research Faculty of Agriculture, and
†
Graduate School of Life Science, Hokkaido University, Sapporo, Hokkaido, Japan
We examined whether Candida albicans gut colonization aggravates immune diseases
in mice. Chronic and latent C. albicans gut colonization was established by the
intragastric inoculation of C. albicans in mice fed as part of a puri ed diet. Allergic
diarrhea was induced by repetitive intragastric administration of ovalbumin in sensi-
tized BALB/c mice. Contact hypersensitivity was evaluated by measuring ear swelling
after topical application of 2, 4-dinitro uorobenzene in NC/Nga mice. Arthritis was
induced by intradermal injection of bovine type-II collagen emulsi ed with complete
Freund ’ s adjuvant in DBA/1J mice. C. albicans gut colonization increased the inci-
dence of allergic diarrhea, which was accompanied by gut hyperpermeability, as well
as increased in ltration of in ammatory cells in the colon. Contact hypersensitivity
was also exacerbated by C. albicans gut colonization, as demonstrated by increased
swelling, myeloperoxidase activity, and proin ammatory cytokines in ear auricles.
Furthermore, C. albicans gut colonization promoted limb joint in ammation in col-
lagen-induced arthritis, in an animal model of rheumatoid arthritis. These  ndings
suggest that C. albicans gut colonization in mice aggravates in ammation in allergic
and autoimmune diseases, not only in the gut but also in the extra-gut tissues and
underscores the necessity of investigating the pathogenic role of C. albicans gut colo-
nization in immune diseases in humans.
Keywords Candida albicans , allergic diarrhea , contact hypersensitivity , collagen-
induced arthritis , mice
Introduction
Candida albicans is part of the indigenous microbial  ora
of the human gastrointestinal tract [1]. Since Truss
described in 1985 that tissue injury induced by C. albi-
cans was accompanied by mental and neurological mani-
festations [2], it was postulated that the excessive
colonization by C. albicans in the gut may be responsible
for wide ranging unspeci c chronic symptoms such as
fatigue, diarrhea, food intolerance, arthritis, and skin
problems which was called ‘ Candida hypersensitivity
syndrome ’ [2 – 5]. Mechanistically, it was believed that
© 2010 ISHAM DOI: 10.3109/13693786.2010.511284
Medical Mycology Month 2010, Early Online, 1–11
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
2 Sonoyama et al .
immune tolerance in mice [11]. In addition, we have devel-
oped a model of sustained C. albicans gut colonization by
a single intragastric inoculation of C. albicans in healthy
adult mice without administration of antibiotics or immu-
nosuppressants [12]. Although these mice appear healthy
despite lifelong C. albicans gut colonization under immu-
nocompetent conditions, disseminated infection by C. albi-
cans in visceral organs including spleen, kidneys, liver, and
lungs is induced upon treatment with immunosuppressive
agents. Thus, these mice are useful as an animal model
mimicking immunocompetent humans with chronic and
latent gut colonization by C. albicans . Using this model,
we demonstrated that serum antibody responses to repeated
oral administration of OVA were enhanced by C. albicans
gut colonization, suggesting that C. albicans gut coloniza-
tion is likely to increase the risk for food allergy [13]. In
the present study, we addressed whether C. albicans gut
colonization aggravates allergic and autoimmune diseases
in mice.
Materials and methods
Animals
The following study was approved by the Hokkaido Uni-
versity Animal Use Committee (approved no. 08-0139),
and animals were maintained in accordance with the guide-
lines of Hokkaido University for the care and use of labo-
ratory animals. Speci c pathogen-free 5-week-old female
BALB/c and NC/Nga mice and male DBA/1J mice were
purchased from Japan SLC (Hamamatsu, Japan). All mice
were housed in a temperature-controlled (23 ⫾ 2 ˚ C) room
with a dark period from 20:00 to 08:00 and allowed free
access to water and a puri ed diet prepared according to
AIN-93G [14].
Inoculation and enumeration of C. albicans
C. albicans (JCM 1542) was maintained as previously
described [12]. For inoculation, all mice were acclimatized
to the puri ed diet for 2 weeks before being deprived of
the diet for 16 h. Mice were then inoculated intragastrically
with 0.2 ml of saline containing 1 ⫻ 10 7 cells of C. albi-
cans (Candida [ ⫹ ]). Control mice were intragastrically
administered 0.2 ml of the vehicle (Candida [⫺]). Fecal
specimens were quantitatively cultured using a standard
pour plate technique as previously described [12].
Allergic diarrhea
Allergic diarrhea was induced in BALB/c mice following
the method of Kweon et al . [15]. Twelve mice were
inoculated with C. albicans as described above, and another
12 mice were uninoculated. At 1 week after the inocula-
tion, 6 mice in each group were immunized subcutaneously
with 1 mg of OVA (grade V, Sigma, St. Louis, MO) in 100
μ l of complete Freund ’ s adjuvant (CFA, Difco Laborato-
ries, Detroit, MI). Another set of 6 mice in each group was
injected subcutaneously with 100 μ l of phosphate-buffered
saline (PBS, unimmunized). Two weeks after the immuni-
zation, all mice started to receive 10 mg of OVA dissolved
in 250 μ l of saline by intragastic administration every other
day. Diarrhea was assessed by visually monitoring mice
for up to 2 h following intragastric challenge. Mice dem-
onstrating liquid stool were recorded as diarrhea-positive
animals. After the 10th challenge, mice were anesthetized
by inhalation of diethyl ether, and whole blood was drawn
from the carotid artery. Serum was separated from the
blood samples and subjected to ELISA for measurement
of OVA-speci c antibody titers as described below. Fol-
lowing a laparotomy, the entire length of the colon was
excised and, after  ushing the luminal contents with 10 ml
of ice-cold PBS, a 1-cm section of the proximal colon was
embedded in OCT compound (Sakura Finetechnical,
Tokyo, Japan) and stored at ⫺ 80 ° C for histological exam-
ination as discussed below. The remaining colon was sub-
jected to the in vitro permeation experiment as described
below.
Contact hypersensitivity
Contact hypersensitivity (CHS) to 2, 4-dinitro uoroben-
zene (DNFB) was induced in NC/Nga mice following the
method of Nagai et al . [16]. Six mice were inoculated with
C. albicans as described above, and another 6 mice were
uninoculated. At 3 weeks after the inoculation, 25 μ l of
0.15% (v/v) DNFB (Tokyo Kasei, Tokyo, Japan) in ace-
tone/olive oil (4:1, v/v) and the vehicle were applied to
each side of the right and left ear auricles of all mice,
respectively, twice at 7-day intervals. Ear thickness was
measured with a digital engineer ’ s micrometer (Mitsutoyo,
Kawasaki, Japan) before and 24, 48, and 72 h after each
application of DNFB. Ear thickness measurements were
performed by an investigator who was blinded to the mouse
treatments. DNFB-speci c ear swelling was calculated
according to the following equation:
Net swelling ⫽ (right ear thickness – left ear thickness)
at each time point – (right ear thickness – left ear thick-
ness) at 0 h.
After the last measurement of ear thickness, mice were
anesthetized by inhalation of diethyl ether, and whole
blood was drawn from the carotid artery. Serum was sepa-
rated from the blood samples and subjected to ELISA for
measurement of dinitrophenol (DNP)-speci c antibody
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
Candida albicans and in ammation 3
antibody (Zymed Laboratories, South San Francisco, CA).
Anti-OVA IgE was captured with rat anti-mouse IgE mono-
clonal antibody (LO-ME-2, Zymed Laboratories)-coated
microtiter plates and detected with digoxigenin-conjugated
OVA followed by HRP-conjugated sheep anti-digoxigenin
Fab fragments (Roche Diagnostics, Tokyo, Japan). Anti-DNP
IgG1 and IgG2a were captured with DNP-bovine serum
albumin (Calbiochem, San Diego, CA)-coated microtiter
plates and detected with HRP-conjugated rat anti-mouse
IgG1 (clone LO-MG1-2, Zymed Laboratories) or rat anti-
mouse IgG2a (clone LO-MG2a-3, Zymed Laboratories).
Anti-CII IgG was captured with CII-coated microtiter plates
and detected with HRP-conjugated goat anti-mouse IgG
polyclonal antibody (Zymed Laboratories). Plates were
developed at room temperature after the addition of o -phe-
nylenediamine (Sigma) and hydrogen peroxide. Pre-immu-
nized serum was used as a negative control. The average
extinction in negative control wells, to which three times the
standard deviation was added, provided the reference for
determination of the titer in the test sera. Antibody titers were
expressed as the reciprocal of the last dilution yielding an
extinction value higher than the reference value.
Histology
Cryostat sections (5 μ m) were prepared and stained with
hematoxylin and eosin (H&E). Toluidine blue staining was
also performed to identify mast cells in the colon tissue
sections. In the colon tissue sections, the number of eosino-
phils and mast cells was counted using a high-power  eld
in a section from each specimen. All cell counts were per-
formed by a single observer who was blinded to the mouse
treatments.
Permeation of HRP in the colon in vitro
As described by Enomoto et al . [17], translocation of HRP
for 40 min was measured in isolated segments of colon. In
brief, 5-cm segments of colon were everted,  lled with 200
μ l of Tris-HCl buffer (125 mM NaCl, 10 mM fructose, 30
mM Tris, pH 7.5), and ligated at both ends. The  lled gut
segments were incubated in Tris-HCl buffer containing 40
μ g/ml HRP (Sigma) at 37 ° C. After 40 min, gut sacs were
removed, and the contents of each sac were collected. HRP
activity in the contents of each sac was determined spec-
trophotometrically from the rate of oxidation of 3, 3⬘, 5,
5 ⬘ -tetramethylbenzidine (TMB, Sigma).
MPO activity
Tissue samples were homogenized in a 50 mM phos-
phate buffer (pH 6.0) with 0.5% hexadecyltrimethyl
ammonium bromide (Sigma). The homogenates were
titers as described below. Ear auricles were excised, and a
portion of each tissue was embedded in OCT compound
and stored at ⫺ 80 ° C for histological examination as
described below. The remaining tissue samples were sub-
jected to myeloperoxidase (MPO) activity measurement
and mRNA expression analysis as discussed below.
Collagen-induced arthritis
Bovine type-II collagen (CII, Cosmo Bio, Tokyo, Japan)
was dissolved in 0.05 M acetic acid to a concentration of
2 mg/ml by stirring overnight at 4 ° C and then emulsi ed
with an equal volume of CFA. Six DBA/1J mice were
inoculated with C. albicans as described above, and 12
mice were uninoculated. At 2 weeks after the inoculation,
150 μ l of the emulsion (150 μ g CII) was injected intrader-
mally on the back of 6 Candida [ ⫹ ] and 6 Candida [⫺]
mice. At 3 weeks after the primary immunization, mice
were boosted intradermally with 150 μ l of the emulsion
(150 μ g CII) at the base of the tail. Another set of 6 Candida
[⫺] mice was injected with PBS (i.e., untreated). After the
boosted injection, the clinical severity of arthritis (i.e.,
arthritis score) was assessed on a daily basis for visual
appearance of each paw, and was quanti ed using the fol-
lowing index: 0, normal; 1, swelling of a single  nger; 2,
swelling of multiple  ngers; 3, swelling of the entire paw.
Each paw was graded, and one mouse could thus have a
maximum score of 12. Additionally, the thickness of each
paw was measured with a digital engineer ’ s micrometer
(Mitsutoyo) on day 0, 3, 6, 10, 17, and 21 after the boosted
injection. Paw thickness measurements were performed by
an investigator who was blinded to the mouse treatments.
Blood samples were obtained from the tail vein at weekly
intervals and subjected to ELISA for measurement of CII-
speci c antibody titers as described below. On the last day
of the experiment (i.e., day 21), mice were anesthetized by
inhalation of diethyl ether, and whole blood was drawn from
the carotid artery. Serum was separated from the blood
samples and subjected to measurement of β -glucan as
described below. Each limb was excised, and the joint tis-
sues were prepared by removing the skin and then separat-
ing the paw below the ankle joint. The tissue samples were
subjected to MPO activity measurement and mRNA expres-
sion analysis as described below.
ELISA for serum antibodies
Serum antibodies speci c to OVA, DNP, and CII were mea-
sured by ELISA as previously described [12]. For anti-OVA
IgG, 96-well microtiter plates were coated with OVA,
blocked, and incubated with serially diluted serum samples.
Bound IgG was detected by incubation with horseradish per-
oxidase (HRP)-conjugated goat anti-mouse IgG polyclonal
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
4 Sonoyama et al .
then subjected to three freeze/thaw cycles. After the cen-
trifugation at 13,000 g for 30 min at 4 ° C, the superna-
tants were subjected to MPO activity measurement as
described by B á nv ö lgyi et al . [18]. In brief, the samples
were added to a 50 mM phosphate buffer (pH 6.0) sup-
plemented with hydrogen peroxide and TMB, and optical
density readings were then taken for 15 min at 620 nm.
The reaction rate ( Δ OD/time) was derived from an initial
slope of the curve. A calibration curve was then pro-
duced, with the rate of reaction plotted against the stan-
dard samples of the human MPO (Sigma). Aliquots of
tissue homogenate were subjected to mRNA expression
analysis as described below.
mRNA expression analysis
Total RNA was isolated from tissue homogenates using
Trizol reagent (Invitrogen, Carlsbad, CA) according to
the manufacturer ’ s instructions. After digestion of
genomic DNA with RQ1 RNase-free DNase (Promega,
Madison, WI), approximately 10 ng of total RNA was
annealed with Oligo (dT)
12-18 primer (Invitrogen) at
70 ° C for 10 min, and 1st strand cDNA was then synthe-
sized using M-MLV reverse transcriptase (Invitrogen),
followed by RNA digestion with DNase-free RNase H
(Invitrogen). Real-time quantitative PCR (RT-qPCR)
was performed using Thermal Cycler Dice TP800
(Takara, Ohtsu, Japan). Primer sequences for interleukin
(IL)-1 β , IL-6, tumor necrosis factor (TNF)- α , and glyc-
eraldehyde-3-phosphate dehydrogenase (GAPDH) were
identical to Giulietti et al . [19]. Ampli cation was car-
ried out in a 25- μ l reaction volume containing 12.5 μ l 1
⫻ SYBR Premix Ex Taq (Takara), 200 nM of each
primer and 1 μ l of template cDNA. The reaction condi-
tion was: 95 ° C for 10 s, followed by 40 cycles at 95 ° C
for 5 s and 60 ° C for 30 s, with dissociation curve at 95 ° C
for 15 s, 60 ° C for 30 s and 95 ° C for 15 s. Relative gene
expression levels for each sample were normalized to the
levels for GAPDH.
b -glucan measurement
Serum samples (5 μ l) were pretreated with 20 μ l of a
solution containing 0.6 M KCl and 0.125 M KOH for 10
min at 37 ° C and assayed with the Glucatell reagent kit
(Associates of Cape Cod, East Falmouth, MA) in a
kinetic, chromogenic format for 35 min at 37 ° C follow-
ing the manufacturer ’ s instruction. Optical densities at
405 nm were read by using a microplate reader (Synergy
Mx, BioTek Instruments, Winooski, VT). The concentra-
tion of β -glucan in each sample was calculated by using
a calibration curve with standard solutions of 3.125 to
50 pg/ml.
Statistics
Results are presented as mean ⫾ SEM. Unpaired t -test or
Tukey-Kramer ’ s test following one-way or two-way analy-
sis of variance was used to compare mean values. The χ 2
test was used to compare the frequencies of diarrhea. Stat-
View for Macintosh (version 5.0, SAS institute Inc., Cary,
NC) was used for the analyses.
Results
Effect of C. albicans gut colonization on allergic
diarrhea in BALB/c mice
After intragastric inoculation, a high fecal recovery of C.
albicans was observed in all Candida [ ⫹ ] mice throughout
the experimental period (Fig. 1A). Enumeration of C. albi-
cans in the gut tissues by quantitative culture after euthanasia
of animals revealed that colonization occurred in the stom-
ach, jejunum, ileum and colon in all Candida [ ⫹ ] mice (data
not shown). We detected no C. albicans in the feces and tis-
sues of Candida [⫺] mice. The present study did not examine
the colonization of other fungal species in the gut of mice.
Diarrhea was observed after two oral administrations of
OVA to the systemically immunized Candida [ ⫹ ] mice
(Fig. 1B). As described by Kweon et al . [15], diarrhea was
observed within 30 min after OVA challenge and dimin-
ished within 2 h, suggesting that acute gut allergic responses
had occurred in these mice. In the immunized Candida [⫺]
mice, diarrhea was  rst observed after six OVA challenges.
The incidence of diarrhea in both groups roughly contin-
ued to increase by the end of the experiment. However,
Candida [ ⫹ ] mice showed a higher incidence of diarrhea
throughout the induction period. Repeated OVA challenge
in mice without systemic immunization did not induce any
change in the gross appearance of feces (data not shown).
H&E staining of colon sections showed that eosinophils
in ltrated the epithelium and crypt region of the colon in
immunized mice (Fig. 1D, upper panels). Eosinophils were
rarely observed in the colon of unimmunized mice (Fig.
1E). In immunized mice, the number of in ltrated eosino-
phils was signi cantly higher in Candida [ ⫹ ]mice than in
Candida [⫺] mice, and was signi cantly higher in mice
with diarrhea than in mice without diarrhea (51 ⫾ 13 vs.
225 ⫾ 24 cells/mm
2 , P ⬍ 0.0001). Additionally, toluidine
blue staining showed the in ltration of mast cells in the
colon (Fig. 1D, lower panels). In both immunized and
unimmunized mice, the number of in ltrated mast cells
was signi cantly higher in Candida [ ⫹ ] mice than in Can-
dida [⫺] mice (Fig. 1E). In both Candida [ ⫹ ] and Candida
[⫺] mice, those with immunization had a signi cantly
higher number of mast cells in the colon as compared
to mice without immunization. In immunized mice, the
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
Candida albicans and in ammation 5
Fig. 1 Effect of Candida albicans gut colonization on OVA-induced allergic diarrhea in BALB/c mice. (A) Temporal changes in the recovery of viable organisms
from the feces in mice with (Candida [ ⫹ ], closed circle) and without (Candida [⫺], open circle) intragastric C. albicans inoculation. Values are represented
as mean ⫾ SEM of six mice per group. (B) Temporal changes in diarrhea incidence in Candida [ ⫹ ] and Candida [⫺] mice (closed and open circles, respectively).
Values with an asterisk are signi cantly different ( P ⬍ 0.05) from values of Candida [⫺] mice at each time point as estimated by χ 2 test. (C) In vitro gut permeability
in Candida [ ⫹ ] and Candida [⫺] mice (closed and open columns, respectively). Gut permeability was estimated using translocation of HRP in isolated segments
of colon. Values are represented as mean ⫾ SEM of six mice per group. Values with unlike letters are signi cantly different ( P ⬍ 0.05) as estimated by
Tukey-Kramer ’ s test. (D) Representative H&E staining and toluidine blue staining for eosinophils and mast cells, respectively, in colon sections of mice after
10th OVA challenge. Bars represent 50 μ m. (E) Numbers of in ltrated eosinophils and mast cells in colon sections of mice. Each circle represents the value
of individual mice, and closed and open circles represent mice with and without diarrhea, respectively. Horizontal bars represent mean values. Values with unlike
letters are signi cantly different ( P ⬍ 0.05) as estimated by Tukey-Kramer ’ s test. n.d., not detected. (F and G) IgG and IgE titers speci c to OVA in Candida [ ⫹ ]
and Candida [⫺] mice (closed and open columns, respectively). Values are represented as mean ⫾ SEM of six mice per group.
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
6 Sonoyama et al .
number of mast cells was signi cantly higher in mice with
diarrhea than in mice without diarrhea (24 ⫾ 9 vs. 202 ⫾
27 cells/mm
2 , P ⬍ 0.0001).
Gut permeability was estimated by measuring translo-
cation of HRP in isolated segments of colon after complet-
ing 10-times administration of OVA (Fig. 1C). In both
immunized and unimmunized mice, the translocation of
HRP was signi cantly higher in Candida [ ⫹ ] mice than in
Candida [⫺] mice. Systemic immunization had no effect
on the translocation of HRP.
Figures 1F and 1G shows serum antibody titers speci c
to OVA after the last administration. There was no signi -
cant difference between Candida [ ⫹ ] and Candida [⫺]
mice for both IgG and IgE. Mice without systemic immu-
nization showed undetectable levels of both IgG and IgE
antibodies speci c to OVA (data not shown).
Effect of C. albicans gut colonization on contact
hypersensitivity in NC/Nga mice
Intragastric inoculation of C. albicans in NC/Nga mice led
to a high fecal recovery of C. albicans in all mice through-
out the experimental period (data not shown). The  rst
application of DNFB induced no ear swelling in all mice
(Fig. 2B), but did begin to be detected 24 h after the second
challenge and tended to decrease thereafter until 72 h. Ear
swelling was signi cantly higher in Candida [ ⫹ ] mice than
in Candida [-] mice from 24 – 72 h after the second chal-
lenge. Histological examination shows moderate cell in l-
tration and severe edema in the dermis of challenged ear
auricle sections (Fig. 2A). In the challenged ear auricle
sections, thickness of dermis was signi cantly higher in
Candida [ ⫹ ] mice than in Candida [⫺] mice (268 ⫾ 32 vs.
405 ⫾ 56 μ m, P ⬍ 0.05). In addition, there was a signi -
cant correlation between the last measurement of ear swell-
ing as measured by hand-operated thickness gauge and
dermis thickness as measured in the histological section
(r ⫽ 0.891, P ⬍ 0.01).
Neutrophil in ltration in the in amed ear auricles on
day 10 after the  rst ear challenge was indirectly quan-
ti ed by MPO activity in the tissue homogenates. In the
vehicle-treated left ear auricles, no detectable levels of
MPO activity were observed (data not shown). MPO
activity in the right ear auricles was signi cantly higher
in Candida [ ⫹ ] mice than in Candida [⫺] mice (Fig.
2C). Additionally, expression of proin ammatory
cytokine genes in ear auricles at day 10 after the  rst
ear challenge was quanti ed by RT-qPCR. Figure 2D
shows the fold induction of IL-1 β , IL-6, and TNF- α
genes in DNFB-treated ear auricles, relative to the val-
ues of vehicle-treated ear auricles, which were taken as
1. These proin ammatory cytokines were upregulated
by intradermal injection of CII. The expression levels of
these genes were signi cantly higher in Candida [ ⫹ ]
mice than in Candida [⫺] mice.
There was no detectable hapten-speci c antibody in the
sera of mice without DNFB treatment (data not shown).
On day 10 after the  rst application of DNFB, all the mice
produced detectable levels of hapten-speci c IgG1 and
IgG2a antibodies (Figs. 2E and 2F, respectively). There
was no signi cant difference between Candida [ ⫹ ] and
Candida [⫺] mice relative to both IgG1 and IgG2a.
Effect of C. albicans gut colonization on collagen-induced
arthritis in DBA/1J mice
DBA/1J mice inoculated with C. albicans showed high C.
albicans fecal recovery throughout the experimental period
(data not shown). Figure 3A illustrates the gross appear-
ance of mouse limbs. As compared to untreated mice,
swelling of forepaws and hindpaws was obvious in CII-
challenged Candida [ ⫹ ] mice. However, in Candida [⫺]
mice, slight swelling was observed in hindpaws, whereas
swelling of forepaws was not evident. Quantitatively, the
arthritis score began to increase on day 5 after boosted
immunization with CII (Fig. 3B). The score was signi -
cantly higher in Candida [ ⫹ ] mice than in Candida [⫺] mice
from day 13 to day 21. Additionally, paw thickness began
to increase on day 10 after boosted immunization (Fig. 3C).
In left forepaws, the thickness was signi cantly higher in
Candida [ ⫹ ] mice than in Candida [⫺] mice from day 10
to day 21. Right forepaws also tended to be thicker in Can-
dida [ ⫹ ] mice than in Candida [⫺] mice. The thickness of
hindpaws tended to be higher in Candida [ ⫹ ] than in Can-
dida [-] mice, and the difference was statistically signi cant
in both right and left hindpaws on day 21.
Untreated mice had no detectable levels of MPO activ-
ity in joint tissue homogenates (data not shown). MPO
activity in both forepaws and hindpaws tended to be higher
in Candida [ ⫹ ] mice than in Candida [⫺] mice (Fig. 3D).
Expression of proin ammatory cytokine genes in joint tis-
sues was quanti ed for pooled samples in each group. Fig-
ure 3E shows the fold induction of IL-1 β , IL-6, and
TNF- α genes in challenged mice, relative to the values of
untreated mice, which were taken as 1. These proin am-
matory cytokines, particularly IL-6, were upregulated by
intradermal injection of CII. The increase in the expression
of these genes was more evident in Candida [ ⫹ ] mice than
in Candida [⫺] mice.
There was no detectable CII-speci c antibody in the
sera of mice without immunization (data not shown). In
immunized mice, CII-speci c IgG antibody titers contin-
ued to increase by the end of the experiment in both groups
of Candida mice (Fig. 3F). There was no signi cant differ-
ence between Candida [ ⫹ ] and Candida [⫺] mice at any
time point after boosted injection of CII.
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
Candida albicans and in ammation 7
Fig. 2 Effect of Candida albicans gut colonization on DNFB-induced CHS in NC/Nga mice. (A) Representative H&E staining of ear auricles on
day 10 after the  rst ear challenge. Bars represent 50 μ m. (B) Temporal changes in ear swelling of Candida [ ⫹ ] and Candida [⫺] mice (closed
and open circles, respectively). Values are represented as mean ⫾ SEM of six mice per group. Values with an asterisk are signi cantly different ( P ⬍
0.05) from values of Candida [⫺] mice at each time point as estimated by unpaired t -test. ( C) MPO activity in ear auricles of Candida [ ⫹ ] and
Candida [⫺] mice (closed and open columns, respectively) on day 10 after the  rst ear challenge. Values are represented as mean ⫾ SEM of six
mice per group. Values with an asterisk are signi cantly different ( P ⬍ 0.05) from values in Candida [⫺] mice as estimated by unpaired t -test.
(D) Expression of IL-1 β , IL-6, and TNF- α genes in ear auricles of Candida [ ⫹ ] and Candida [⫺] mice (closed and open columns, respectively)
on day 10 after the  rst ear challenge as estimated by RT-qPCR. Values represent the fold induction of each gene in challenged mice, relative to
the values of untreated mice, which were taken as 1. Values are represented as mean ⫾ SEM of six mice per group. Values with an asterisk are
signi cantly different ( P ⬍ 0.05) from values in Candida [⫺] mice as estimated by unpaired t -test. (E and F) IgG1 and IgG2a titers speci c to
DNP in Candida [ ⫹ ] and Candida [⫺] mice (closed and open columns, respectively) on day 10 after the  rst ear challenge. Values are represented
as mean ⫾ SEM of six mice per group.
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
8 Sonoyama et al .
Fig. 3 Effect of Candia albicans gut colonization on CII-induced arthritis in DBA/1J mice. (A) Gross appearance of forepaws and hindpaws of Candida
[ ⫹ ], Candida [⫺], and untreated mice on day 21 after boosted immunization. (B) Temporal changes in arthritis scores in Candida [ ⫹ ] and Candida [⫺]
mice (closed and open symbols, respectively). Circles and triangles represent left and right paws, respectively. Values are represented as mean ⫾ SEM
of six mice per group. Values with an asterisk are signi cantly different ( P ⬍ 0.05) from values in Candida [⫺] mice at each time point as estimated
by unpaired t -test. ( C) Temporal changes in paw thickness in Candida [ ⫹ ] and Candida [⫺] mice (closed and open circles, respectively). Values are
represented as mean ⫾ SEM of six mice per group. Values with an asterisk are signi cantly different ( P ⬍ 0.05) from values in Candida [⫺] mice at
each time point as estimated by unpaired t -test. (D) MPO activity in limb joint tissue homogenates of Candida [ ⫹ ] and Candida [⫺] mice (closed and
open columns, respectively). Values are represented as mean ⫾ SEM of six mice per group. ( E) Expression of IL-1 β , IL-6, and TNF- α genes in limb
joint tissues of Candida [ ⫹ ] and Candida [⫺] mice (closed and open columns, respectively) as estimated by RT-qPCR. Values show the measurements
of pooled samples in each group and represent the fold induction of each gene in challenged mice, relative to the values of untreated mice, which were
taken as 1. ( F) IgG titers speci c to CII in Candida [ ⫹ ] and Candida [⫺] mice (closed and open circles, respectively). Values are represented as mean
⫾ SEM of six mice per group.
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
Candida albicans and in ammation 9
Serum β -glucan concentrations were the same between
Candida [⫺] mice and Candida [ ⫹ ] mice (16.4 ⫾ 0.7 vs.
15.6 ⫾ 1.9 pg/ml, respectively).
Discussion
The present study demonstrated that chronic and latent gut
colonization by C. albicans promoted acute allergic diar-
rhea and CHS-induced ear swelling in immunocompetent
mice. The results suggest that C. albicans gut colonization
aggravates not only IgE-mediated immediate-type allergy
but also T cell-mediated delayed-type hypersensitivity. In
addition, the in ammatory responses are exacerbated not
only by C. albicans colonization within the gut tissue but
also in the extra-gut tissues. Furthermore, C. albicans gut
colonization aggravated CII-induced arthritis, an animal
model of rheumatoid arthritis, implying a contribution to
arthritic deterioration in autoimmune disease. To our
knowledge, the present study is the  rst describing the
aggravation of allergic and autoimmune diseases by C.
albicans gut colonization in immunocompetent mice.
Mice with allergic diarrhea showed the in ltration of
eosinophils and mast cells in the colon, which is consistent
with the report of Kweon et al . [15]. Exacerbation of aller-
gic diarrhea by C. albicans gut colonization was the result
of an increased in ltration of these in ammatory cells.
Additionally, hapten-induced CHS mice and CII-induced
arthritic mice showed an increase in MPO activity, a marker
of neutrophil in ltration in in amed tissues, and the activ-
ity was further increased by C. albicans gut colonization.
Furthermore, gene expression of proin ammatory cytok-
ines such as IL-1 β , IL-6, and TNF- α in in amed tissues
was upregulated by C. albicans gut colonization. Despite
these enhanced in ammatory responses, C. albicans gut
colonization exerted no changes in serum antigen-speci c
antibodies in three experimental models. Therefore, it is
likely that C. albicans gut colonization modulates local
in ammatory responses but not systemic humoral immune
responses.
Our present study showed an increased protein perme-
ation in the colon of C. albicans -colonized mice, as evi-
denced by in vitro translocation of HRP in isolated segments
of the colon. Because we previously demonstrated that C.
albicans gut colonization enhanced in vivo gut permeation
of intragastrically administered proteins (that is, HRP and
OVA) in BALB/c mice [13], it appears likely that translo-
cation of OVA in isolated segments of the colon is also
promoted by C. albicans gut colonization. In our present
investigation, increased HRP translocation in the colon was
observed not only in immunized mice but also in unim-
munized mice, even though the mice were not affected by
diarrhea. The results suggest that in mice colonized by C.
albicans , higher gut permeability is not a consequence, but
rather one of the causes of an increased incidence of aller-
gic diarrhea. Our previous study also suggested that mast
cells are responsible for increased gut permeability in C.
albicans -colonized mice. In this study, mast cell in ltration
in the colon was promoted by C. albicans colonization in
unimmunized mice. Thus, it is likely that the in ltration of
mast cells in the colon by C. albicans colonization contrib-
utes to an increased occurrence of allergic diarrhea by
increasing antigen uptake in previously sensitized mice.
According to Kweon et al ., mast cells and eosinophils are
recruited by Th2 cytokines produced by CD4
⫹
T cells in l-
trated in the colon of diarrhea-induced mice [15]. There-
fore, the increased uptake of antigen by C. albicans gut
colonization may lead to an in ltration of antigen-speci c
CD4
⫹
T cells, which in turn further recruit effector cells,
such as mast cells and eosinophils, in the colon of diarrhea-
induced mice. However, our preliminary experiment
showed that C. albicans gut colonization did not alter IL-4
production in isolated mesenteric lymph node cells restim-
ulated in vitro with OVA (1 mg/ml) in sensitized mice (0.87
⫾ 0.17 and 0.51 ⫾ 0.33 ng/ml in mice with and without
C. albicans gut colonization, respectively, Sonoyama et al.
unpublished data), thus we need to examine whether
C. albicans gut colonization promotes in ltration of anti-
gen-speci c CD4
⫹
T cells and production of Th2 cytokines
in the colonic mucosa. Mast cells act as the main effector
cells in the development of IgE-mediated allergic responses
by releasing chemical mediators such as histamines, leukot-
rienes, and cytokines, which exert clinical symptoms,
including diarrhea, and further recruit other immune and/or
in ammatory cells [20]. Taken together, it is likely that mast
cells act as the trigger and effector in promoting allergic
diarrhea in C. albicans -colonized mice.
In hapten-induced CHS and CII-induced arthritis, C.
albicans gut colonization aggravated the in ammatory
responses in tissues distal to the gut. Because no infec-
tions are evident in the visceral organs in this C. albicans
colonized mouse model [12], it is unlikely that invading
viable fungal cells are responsible for the local in amma-
tory responses. In our mouse model of C. albicans gut
colonization, serum IgG antibodies speci c to the cell
wall fraction of C. albicans become detectable after C.
albicans gut colonization [12]. Regarding rheumatoid
arthritis, gut microbiota have been widely proposed as a
potential environmental factor associated with the etiol-
ogy of this disease [21]. Indeed, the cell walls of several
enteric bacterial species are arthritogenic in animal mod-
els, and patients with early rheumatoid arthritis were
found to have a different gut microbiota from that of con-
trol patients [22]. In terms of fungi, Hida et al . reported
that injection of CII with β -glucan isolated from C. albi-
cans induces arthritis in DBA/1 mice [23], suggesting that
β -glucan, a cell wall constituent of C. albicans , acts as an
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
10 Sonoyama et al .
adjuvant for CII-induced arthritis. Taken together, these
 ndings raise the possibility that degradation products of
fungal cells and/or secreted compounds might be absorbed
by the gut, systemically circulate, and subsequently in u-
ence systemic immune responses and/or local in amma-
tory reactions. However, in CII-induced arthritis model of
our present study, serum concentrations of β -glucan in
mice with C. albicans gut colonization did not differ from
mice without colonization. In addition, serum antibody
titers speci c to CII were the same between mice with and
without C. albicans gut colonization. Our preliminary
experiments showed that C. albicans gut colonization
induced no overt symptoms of arthritis in DBA/1J mice
when CII was injected without CFA (Miki et al . unpub-
lished observation). Furthermore, we observed that serum
antibody responses to repeated oral administration of
OVA were enhanced by C. albicans gut colonization [13],
but not by daily intragastric administration of heat-killed
C. albicans (Hata et al . unpublished observation). There-
fore, it appears unlikely that tissue in ammation and/or
antibody responses are promoted by systemically circulat-
ing cell constituents such as β -glucan derived from C.
albicans colonized in the gut. We are currently investigat-
ing whether intestinal epithelial cells and/or gut-associ-
ated lymphoid tissues exposed directly to C. albicans
produce some circulating factors that promote tissue
in ammation and antibody responses.
Taken together, the present  ndings suggest that C. albi-
cans gut colonization aggravates in ammation in allergic
and autoimmune diseases in mice, not only in the gut but
also in tissues distal to the gut. However, the present  nd-
ings are derived from animal experiments and not human
studies, thus we propose the necessity of investigating the
pathogenic role of C. albicans gut colonization in allergic
and autoimmune diseases in humans. In addition, further
studies are needed to  gure out whether observed proin-
 ammatory effects are speci c to C. albicans and whether
anti-fungal treatment ameliorates enhanced in ammation.
Furthermore, we observed no signi cant relationships
between fecal C. albicans excretions and in ammation
parameters in the present study (data not shown), thus we
need to clarify the relationship between gut colonization
levels and extent of tissue in ammation. Nevertheless, the
present study provides an experimental tool for studying
the underlying molecular and cellular mechanisms for the
proin ammatory action of C. albicans colonized in the
gut.
Acknowledgments
This study was partly supported by Special Coordination
Funds for Promoting Science and Technology, a Grant-
in-Aid for Scienti c Research from The Ministry of
Education, Science, Sports and Culture of Japan (no.
19380070).
Declaration of interest: The authors report no con icts of
interest. The authors alone are responsible for the content
and writing of the paper.
References
Calderone RA. 1 Candida and Candidiasis . Washington, DC: ASM
Press, 2001.
Truss CO. Tissue injury by 2 Candida albicans . In Truss CO (ed.), The
Missing Diagnosis . Alabama: 1985: 127 – 147.
Crook WG. 3 The Yeast Connection: A Medical Breakthrough . Tennes-
see: Jackson Professional Books, 1983.
Dismukes WE, Wade JS, Lee JY, Dockery BK, Hain JD. A rand-4
omized, double-blind trial of nystatin therapy for the candidiasis
hypersensitivity syndrome. N Engl J Med 1990; 323 : 1717 – 1723.
Cater RE 2nd. Chronic intestinal candidiasis as a possible etiological 5
factor in the chronic fatigue syndrome. Med Hypotheses 1995; 44 :
507 – 515.
Faergemann J. Atopic dermatitis and fungi. 6 Clin Microbiol Rev 2002;
15 : 545 – 563.
Lacour M, Zunder T, Huber R, 7 et al . The pathogenetic signi cance
of intestinal Candida colonization – a systematic review from an
interdisciplinary and environmental medical point of view. Int J Hyg
Environ Health 2002; 205 : 257 – 268.
Nikkels AF, Pierard GE. Framing the future of antifungals in atopic 8
dermatitis. Dermatology 2003; 206 : 398 – 400.
Goldman DL, Huffnagle GB. Potential contribution of fungal infec-9
tion and colonization to the development of allergy. Med Mycol 2009;
47 : 445 – 456.
Noverr MC, Noggle RM, Toews GB, Huffnagle GB. Role of antibi-10
otics and fungal microbiota in driving pulmonary allergic responses.
Infect Immun 2004; 72 : 4996 – 5003.
Noverr MC, Falkowski NR, McDonald RA, McKenzie AN, Huff-11
nagle GB. Development of allergic airway disease in mice fol-
lowing antibiotic therapy and fungal microbiota increase: role of
host genetics, antigen, and interleukin-13. Infect Immun 2005; 73 :
30 – 38.
Yamaguchi N, Sonoyama K, Kikuchi H, 12 et al . Gastric colonization of
Candida albicans differs in mice fed commercial and puri ed diets. J
Nutr 2005; 135 : 109 – 115.
Yamaguchi N, Sugita R, Miki A, 13 et al . Gastrointestinal Candida colo-
nisation promotes sensitisation against food antigens by affecting the
mucosal barrier in mice. Gut 2006; 55 : 954 – 960.
Reeves PG, Nielsen FH, Fahey GC Jr. AIN-93 puri ed diets for labo-14
ratory rodents:  nal report of the American Institute of Nutrition ad
hoc writing committee on the reformulation of the AIN-76A rodent
diet. J Nutr 1993; 123 : 1939 – 1951.
Kweon MN, Yamamoto M, Kajiki M, Takahashi I, Kiyono H. Sys-15
temically derived large intestinal CD4
⫹
Th2 cells play a central
role in STAT6-mediated allergic diarrhea. J Clin Invest 2000; 106 :
199 – 206.
Enomoto N, Yamashina S, Schemmer P, 16 et al . Estriol sensitizes rat
Kupffer cells via gut-derived endotoxin. Am J Physiol 1999; 277 :
G671 – G677.
Nagai H, Matsuo A, Hiyama H, Inagaki N, Kawada K. Immunoglob-17
ulin E production in mice by means of contact sensitization with
a simple chemical, hapten. J Allergy Clin Immunol 1997; 100 :
S39 – S44.
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.

© 2010 ISHAM, Medical Mycology, Early Online, 1–11
Candida albicans and in ammation 11
B á nv ö lgyi A, P á link á s L, Berki T, 18 et al . Evidence for a novel protec-
tive role of the vanilloid TRPV1 receptor in a cutaneous contact aller-
gic dermatitis model. J Neuroimmunol 2005; 169 : 86 – 96.
Giulietti A, Overbergh L, Valckx D, 19 et al . An overview of real-time
quantitative PCR: applications to quantify cytokine gene expression.
Methods 2001; 25 : 386 – 401.
Gordon JR, Burd PR, Galli SJ. Mast cells as a source of multifunc-20
tional cytokines. Immunol Today 1999; 11 : 458 – 464.
Thompson-Chagoy á n OC, Maldonado J, Gil A. Colonization and im-21
pact of disease and other factors on intestinal microbiota. Dig Dis Sci
2007; 52 : 2069 – 2077.
Toivanen P. Normal intestinal microbiota in the aetiopathogenesis of 22
rheumatoid arthritis. Ann Rheum Dis 2003; 62 : 807 – 811.
Hida S, Miura NN, Adachi Y, Ohno N. Effect of 23 Candida albicans cell
wall glucan as adjuvant for induction of autoimmune arthritis in mice.
J Autoimmun 2005; 25 : 93 – 101.
This paper was  rst published online on Early Online on 02 September
2010.
Med Mycol Downloaded from informahealthcare.com by Hokkaido University on 09/05/10
For personal use only.