Celiac disease in patients with type 1 diabetes: a
condition with distinct changes in intestinal immunity?
Raivo Uibo1, Marina Panarina1, Kaupo Teesalu1, Ija Talja1, Epp Sepp2, Meeme Utt1, Marika Mikelsaar2,
Kaire Heilman3, Oivi Uibo3,4and Tamara Vorobjova1
Two common chronic childhood diseases—celiac disease (CD) and type 1 diabetes (T1D)—result from complex pathological
mechanisms where genetic susceptibility, environmental exposure, alterations in intestinal permeability and immune responses play
central roles. In this study, we investigated whether these characteristics were universal for CD independently of T1D association. For
this purpose, we studied 36 children with normal small-bowel mucosa and 26 children with active CD, including 12 patients with T1D.
T1D, indicating an increase in intestinal permeability. Furthermore, these samples displayed the highest expression of forkhead box P3
(FoxP3) mRNA, a marker for regulatory T cells, as compared with other patient groups. At the same time, serum levels of IgA antibodies
specific for the CD-related antigens deamidated gliadin and tissue transglutaminase (tTG) were the highest in CD patients with T1D. In
contrast, no significant differences were found in IgA or IgG antibodies specific for bovine beta-lactoglobulin or Bifidobacterium
adolescentis DSM 20083-derived proteins. There were also no differences in the transamidating activity of serum autoantibodies
between patients and control individuals. Our results show that patients with T1D and newly detected CD exhibit severely altered
intestinal permeability, strong local immune activation and increased immunoregulatory mechanisms in the small bowel. Further study
is required to determine whether these extreme changes in this CD subgroup are due to some specific environmental factors (virus
infections), unknown genetic effects or autoimmune reactions to antigenic targets in intracellular tight junctions.
Cellular & Molecular Immunology (2011) 8, 150–156; doi:10.1038/cmi.2010.66; published online 14 February 2011
Keywords: autoimmunity; celiac disease; immunoregulation; intestinal mucosa; type 1 diabetes
With an approximate prevalence of 1.0% in Western countries, celiac
disease (CD) is one of the most common lifelong chronic disorders.1
CD and type 1 diabetes (T1D) form a significant sector of childhood
diseases with high rate of co-occurrence.2Both diseases are auto-
immune in origin and develop as a result of complex pathological
mechanisms, involving several common genetic, environmental and
immunological factors. Among these, the common susceptibility loci
of HLA and other immune system-related genes,3permeability
changes in the small-bowel mucosa,4and association with wheat con-
sumption4,5are prominent. Both diseases are increasing in most
regions of the world,6,7as are several other autoimmune diseases,
recent studies have highlighted the role of environmental and social
changes that have taken place over the last several decades.8,9
The central event in the pathogenesis of CD is damage to the small
intestinal mucosa that occurs after ingestion of gluten or related pro-
lamins in genetically susceptible individuals. During this process, sev-
eral morphological and immunological alterations have been
demonstrated in the small-bowel mucosa. These mucosal changes
are also reflected in the presence of characteristic circulating antibod-
ies10directed against tissue (type 2) transglutaminase (tTG) and dea-
midated gliadin; these antibodies are specifically associated with small
intestinal destruction in CD and are therefore widely used for serolo-
gical CD screening.11Histologically, CD is characterized by intestinal
villous atrophy with high numbers of infiltrating T cells of different
lineages.10,12Recently, much interest has been paid to the character-
ization of CD41T cells in the intestinal mucosa because the balance
between inflammatory and regulatory cells is crucial in determining
immune-mediated damage to the intestinal mucosal. Studies have
demonstrated that the number of CD41CD251FoxP31regulatory T
cells is increased in the small-bowel mucosa of CD patients; this
the elevated immune response to foreign and self-antigens.13–15
Immune responses to the self-antigen tTG may be caused by the
increased expression of tTG in the intestinal mucosa,16a conforma-
tional change in enzyme structure17or the release of intracellular tTG
that, when picked up by dendritic cells, subsequently leads to the
activation of adaptive immune responses. Antibodies against tTG
may influence either the deamidating or the transamidating activity
1Department of Immunology, University of Tartu, Tartu, Estonia;2Department of Microbiology, University of Tartu, Tartu, Estonia;3Children’s Clinic of Tartu University Hospital,
Tartu, Estonia and4Department of Pediatrics, University of Tartu, Tartu, Estonia
Correspondence: Dr R Uibo, Department of Immunology, University of Tartu, Ravila Street 19, Tartu 50411, Estonia.
Received 7 December 2010; accepted 9 December 2010
Cellular & Molecular Immunology (2011) 8, 150–156
? 2011 CSI and USTC. All rights reserved 1672-7681/11 $32.00
of the enzyme, which may lead to significant biological effects in
intestine. However, results regarding the effect of tTG autoantibodies
on enzyme function remain controversial.
Immune reactions directed against foreign antigens, such as food,
might be at least partly dependent on the integrity of the intestinal
epithelium.18,19Altered intestinal permeability in CD is well known,
particularly in connection with gluten exposure that diminishes
Importantly, removal of gluten from the diet of CD patients nor-
malizes intestinal permeability and prompts the return of normal
structure within the intestinal mucosa; this parallels an increased
expression of tight junction protein 1 (TJP1), also known as zona
occludens 1, in the small bowel. However, electron microscopic ana-
lysis has shown that damage may be irreversible in a portion of
T1D exhibit abnormalities in intestinal permeability as evaluated by
also detected before the onset of diabetes; increased intestinal per-
meability in these cases has been shown to be dependent on increased
zonulin expression, which regulates tight junctions in the intestinal
Additionally, mucosal biopsy, fecal microbiota and serological
studies have indicated that the composition of intestinal microbiota
may influence the immune mechanisms that participate in the
recent investigations have stressed the importance of Bifidobacteria
species and strains in the modulation of immune reactivity at the
intestinal mucosa level.31In particular, Bifidobacterium adolescentis
has recently received attention because this bacterium is more preval-
ent in patients with allergic disorders compared to non-allergic sub-
jects, thusindicating thatit mightbeconnected tothedevelopment of
Based on available experimental and clinical results, it has been
proposed that the pathogenesis of T1D is closely linked to events that
take place in the intestinal mucosa, where complex interplay between
the intestinal microbiota, gut permeability and mucosal immunity
determines autoimmune damage to pancreatic beta cells.34In several
associations, this has also been demonstrated in CD and other auto-
Here, we aimed to assess whether differences in intestinal permeab-
ility, characterized by TJP1 mRNA expression, and intestinal regula-
tory T cells,measured by Foxp3 mRNA expression, as wellas different
serum antibodies levels exist between CD patients with and without
accompanying T1D. Our task was also to evaluate differences in the
effect of transamidating activity of serum tTG antibodies on tTG
between patients with CD and controls.
MATERIALS AND METHODS
The study comprised three patient groups: (i) 12 patients with
active CD and T1D (five males, aged 5–14 years); (ii) 14 patients
with active CD and no T1D (aged 1–15 years, 9 males); and (iii) 36
patients with normal small-bowel mucosa revealed at biopsy for
functional dyspepsia, ulcus duodeni, erosive gastritis, neurological
disorders, allergic dermatitis, etc. (aged 1–18 years, 12 males).
Patients with CD were identified from 291 childhood T1D patients
(aged 2–18 years, 167 males) during retrospective and prospec-
tive studies assaying anti-tTG IgA and/or endomysium antibodies
between 1995 and 2007 as described elsewhere.36Small-bowel
biopsies (either with Watson capsule or gastroduodenoscope) were
performed in Tartu University Children’s Hospital for all pa-
tients. The Watson capsule biopsy samples were divided in two
portions: one half was stored for morphological examination and
the other half was immediately quick-frozen in TissueTek OCT
Compound (Sakura, Finetek, Finland) and stored at 280 uC. At
gastroduodenoscopy, one biopsy sample was used for morpho-
logical examination, while the other was stored in RNAlater
(Ambion Inc., Austin, TX, USA) at 225 uC for later analysis by
RT-PCR. The small-bowel mucosa state was morphologically eval-
uated according to the Marsh classification,37,38and the diagnosis
of CD was performed using the criteria of the European Society for
Pediatric Gastroenterology, Hepatology and Nutrition.39None of
the patients had received a diagnosis of CD before or had been on
gluten-free diets. At the time of biopsy, all patients donated blood
samples for antibody studies. This study was approved by the
Ethics Committee for Medical Investigations at the University of
Tartu. All studied children and their parents gave their written
Detection of serum antibodies
IgA and IgG type antibodies against tTG were detected using ELISA
and human recombinant tTG according to the method described by
microwells by human fibronectin (Sigma-Aldrich, St Louis, MO,
IgA and IgG type antibodies to bovine beta-lactoglobulin (Sigma-
Aldrich) were measured using an ELISA protocol suggested by
Professor E. Savilahti from Helsinki University with minor modifica-
tion. In brief, Nunc PolySorp microtiter plates (Roskilde, Denmark)
ml in phosphate-buffered saline (PBS) and incubated overnight at
4 uC. After being washed with PBS containing 0.1% Tween 20, plates
were treated for 1 h at 37 uC with a blocking solution of 2% normal
horse serum in PBS. Subsequently, serum samples diluted 1:20 in 1%
horse serum PBS were incubated on the plate (two wells with the
horseradish peroxidase conjugated rabbit anti-human IgG (DAKO,
Glostrup, Denmark) or rabbit anti-human IgA (DAKO) was applied
to the wells, followed by a 1-h incubation at 37 uC. After another
washing, tetramethyl-benzidine liquid substrate (Sigma-Aldrich) was
applied, and plates were incubated for 10 min. at room temperature.
After reaction termination with 2 M H2SO4, the optical density of the
reaction product was measured at 450 nm. The absorption value
derived from the mean value of absorption in two wells with the
antigen minus the absorption value in well without antigen was taken
as the absorbance value of the studied sera representing the relative
amount of antibodies in the studied sample.
IgA and IgG antibodies to deamidated gliadin were measured by
ELISA using a Euroimmun GAF-3X Kit (Medizinische Labordiagnostika
AG, Lu ¨beck, Germany); the reactions were performed according to man-
ufacturer’s instructions. The results are expressed in RU/ml.
proteins were detected by immunoblot using bacterial cell lysates.
Lysates were obtained by growing B. adolescentis (strain DSM
20083) in anaerobic conditions (5% CO2, 5% H2and 90% N2) on
Wilkins–Chalgren medium (Oxoid Ltd, Hampshire, UK) for 48 h in
an anaerobic glove box (Sheldon Manufacturing Inc., Cornelius, OR,
USA). Cells were collected, suspended in PBS and washed three times.
Celiac disease in type 1 diabetes
R Uibo et al
Cellular & Molecular Immunology
Subsequently, cells were disrupted with 0.1 mm glass beads (Biospec
Products, Bartlesville, OK, USA) in PBS in the presence of protease
inhibitors (Complete Tablets; Boehringer Mannheim, Mannheim,
Germany) on ice. Protein concentrations in the lysateswere estimated
by Protein Assay Solution (Bio-Rad, Hercules, CA, USA). Equal
amounts of proteins from different preparations were mixed with
300 ml SDS-PAGE sample buffer (62.5 mM Tris (pH 6.8), 2.3% SDS,
5% 2-mercaptoethanol, 10% glycerine and a few grains of bromophe-
nol blue) and heated for 15 min at 95 uC. Approximately 100 mg of
total protein was loaded for each gel.
Bacterial cell lysate proteins were loaded on 5–20% gradient gels
with a 5% concentrating gel with All Blue 10–250 kDa molecular
weight (MW) markers (Bio-Rad) as standards; proteins were sepa-
rated using the vertical electrophoresis system SE-600 (Hoefer, San
Francisco, CA, USA) connected to a thermostat. The separated pro-
teins were transferred to polyvinylidene difluoride membranes
(0.45 mm pore size) using a semidry electroblotter (Hoefer) as
described by Nilsson et al.41After transfer, the membranes were
blocked twice. First, the membranes were incubated for 15 min in a
solution containing 10 g/l polyvinyl pyrrolidone (MW: 40 kDa) and
25% methanol suspended in an ethanolamine/glycine buffer (pH 9.6)
containing 12.2 g/l ethanolamine (Merck and Co., Whitehouse
Station, NJ, USA) and 18 g/l glycine (Merck and Co.). Second, the-
membranes were incubated for another 15 min with the ethanola-
mine/glycine buffer containing Tween 20 (0.14%) and gelatin
hydrolysate (5 g/l).7,36
After membranes were cut into 5 mm wide strips, the strips were
incubated on a shaker overnight at 4 uC with sera or plasma and
diluted 1:50 for IgA and 1:100 for IgG antibody assay in an incub-
ating buffer composed of 1.25 g/l gelatin hydrolysate, 0.25 g/l Tween
20, 6.1 g/l NaCl (Merck and Co.) and 0.06 g/l Tris Base (Sigma-
Aldrich). Strips were incubated with secondary anti-human IgA or
anti-human IgG (diluted 1:500; DAKO) antibodies labeled with
horseradish peroxidase for 1 h at room temperature; subsequent sub-
in the presence of hydrogen peroxide (0.04% 3-amino-9-ethylcarba-
zole and 0.015% hydrogen peroxide) was performed for 30 min. at
room temperature. Extensive washes with PBS–Tween were applied
between incubation steps. Reactions were terminated by washing the
strips with distilled water.
(Bio-Rad). The relative MW of the bands was estimated with the Bio-
All Blue molecular markers.
Human recombinant tTG containing a C-terminal 6His-tag was
geneity by two-step chromatography as described previously.40
Briefly, 1 mg of tTG protein was coupled to 1 ml of NHS-activated
Sepharose 4 Fast Flow Beads in 2 ml centrifuge columns (Pierce
Biotechnology, Rockford, IL, USA) according to the manufacturer’s
instructions (GE Healthcare Bio-Sciences AB, Uppsala, Sweden).
Human serum (0.5 ml), diluted 1:4 in PBS, was passed through a
0.45 mm filter and applied to tTG-Sepharose column for 30min. After
column was washed four times with 3 ml PBS, bound antibodies were
eluted with 2 ml of 0.1 M glycine, pH 2.5, and pH neutralized by
adding 0.5 ml of 1 M Tris-HCl, pH 8.0. After dialysis against PBS
using Pierce Slide-A-Lyzer Dialysis Cassettes (Thermo Fisher
Scientific Inc., Rockford IL, USA), the final preparations of affinity
chloride. Protein concentrations in antibody samples were deter-
mined by the Bradford method using bovine c-globulin as a standard.
Tissue transglutaminase transamidation assays were performed
using fibronectin bound tTG as described earlier by Kira ´ly et al.42
with minor modifications. Briefly, universal binding 96-well micro-
titer plate (Thermo Fisher Scientific Oy, Vantaa, Finland) wells were
coated with 100 ml human fibronectin (Sigma-Aldrich) at a concen-
tration of 5 mg/ml in carbonate/bicarbonate buffer (pH 9.6) overnight
at 4 uC. The wells were washed five times with 300 ml TBS-T buffer
(25 mM Tris-HCl, 150 mM NaCl, 0.1% Tween 20, pH 7.4) and
human recombinant tTG (Teesalu, 2009) was added at 1 mg/ml in
TBS with 2 mM DTT for 30 min at 20 uC. After being washed with
TBS-T, the wells were incubated with 10 mg/ml of human anti-tTG,
rabbit polyclonal anti-tTG (Thermo Fisher Scientific, Fremont, CA,
USA) or CUB7402 monoclonal antibodies (5 mg/ml) in TBS, with 1%
BSA and 2 mM DTT for 30 min at 20 uC. Next, after washing with
TBS-T, 100 ml of 0.5 mM 5-(biotinamido)-pentylamine (Pierce
Biotechnology) in reaction buffer (100 mM Tris-HCl, pH 8.5, 5 mM
CaCl2, 10 mM DTT) was added and plates were incubated for 15 min
at 37 uC. The reactions were stopped by washing the wells with TBS-T
five times, followed by incubation with streptavidin-conjugated alkal-
The absorbance was read at 405 nm with 492 nm subtraction. The
binding of tTG is expressed as percentages of results without antibod-
ies in inhibition experiments. The results are expressed as percentages
of OD values obtained at the same conditions but without the addi-
tion of antibodies or inhibitors.
Detection of FoxP3 and TJP1 mRNA expression in small-bowel
Total RNA was isolated from biopsy samples using the RNeasy Mini
Kit (QIAGEN GmbH, Hilden, Germany) according to the manufac-
turer’s instructions. From successful isolations, cDNA was prepared
using the cDNA Transcription Kit (Applied Biosystems, Foster City,
CA, USA.) according to the manufacturer’s protocol. Genomic
DNA was eliminated by DNAase I treatment (Roche Diagnostics,
Mannheim, Germany) prior to cDNA synthesis. Random hexamers
(Applied Biosystems) were used to prime first-strand synthesis, and
the reaction was performed in a total volume of 20 ml with the
Multiscribe Reverse Transcriptase enzyme according to the manufac-
turer’s protocol (Applied Biosystems).
Transcription level of the FoxP3 gene (Hs00203958_m1), as a char-
acteristic of regulatory T cells, was evaluated using TaqMan Gene
Expression assay (Applied Biosystems) as described previously.14
The transcription level of the TJP1 gene (Hs.510833) was evaluated
using SYBR Green RT-PCR Gene Expression Analysis (SuperArray;
from each sample was used for each duplicate measurement for
human TJP1 (Hs.510833, SuperArray) in combination with 1 ml
cDNA to measure the housekeeping gene b-actin (ACTB, Hs.
520640). RT-PCR gene expression analysis was performed according
to the manufacturer’s instructions using SYBR Green/ROX qPCR
Master Mix (SuperArray; Bioscience Corporation). The homemade
exogenous cDNA calibrator sample of TJP1 (prepared from Caco-2
cells) was used for the interassay standard to which all other normal-
ized samples were compared. The Caco-2 cell line is known to express
high level of TJP1 mRNA.43Total RNA was isolated from Caco-2 cells
Celiac disease in type 1 diabetes
R Uibo et al
Cellular & Molecular Immunology
using the RNeasy Mini Kit (QIAGEN GmbH) and QIAshredder Spin
Columns (QIAGEN GmbH). The TJP1 cDNA was prepared from
Caco-2 cellsusing thecDNA
Biosystems) according to the manufacturer’s protocol.
The ABI Prism 7000 Sequence Detection System was programmed
for10minat95uC,followedby 40thermal cyclesof15sat95 uC and
Each measurement was performed in duplicate.
In both assays, for either FoxP3 or TJP1, the comparative threshold
(Ct) method was used to quantitate gene transcription in the samples.
The Ctof controls (18S RNA for FoxP3 and Caco-2 cells RNA for
TJP1) was subtracted from the target gene Ct. The obtained difference
was the DCtvalue. The DCtof the calibrator was then subtracted from
the DCtof the sample. The obtained difference was called the DDCt
value, and the results were expressed in relative units based on the
calculation of 2-nnCt, which yielded the relative amount of the target
gene normalized to the corresponding control and relative to the
calibrator. For the graphical FoxP3 representation, the relative num-
bers were multiplied by 1000.
Transcription levels of the FoxP3 gene in the biopsy samples posi-
tively correlate to the number of CD41CD251FoxP31regulatory T
to our experience with TJP1 immunocytochemical staining, differ-
in the intestinal mucosa are rather complicated to evaluate due to
variable staining patterns on the same sample.
The results obtained for different study groups were presented as
mean6s.e. and compared using the t-test and the analysis of variance
by ANOVA as well as by the two-tailed Spearman’s rank correlation
test. The age adjustment for mean values of antibody level was per-
formed using multiple regression analysis. Differences were consid-
ered statistically significant atP,0.05. Statistica 8.0 software was used
for all analyses.
Comparative expression of TJP1 and FoxP3 mRNA in the small-
bowel mucosa and tTG
The lowest values of TJP1 mRNA expression were found in patients
with CD and T1D (437.9652.5); this group also exhibited the highest
expression of FoxP3 mRNA (335.36102.6). In contrast, the small-
bowel mucosa of control individuals exhibited the highest TJP1
mRNA expression (713.2698.4) and the lowest FoxP3 mRNA
expression (61.1610.1) (Figure 1). In patients with CD without
T1D, the mRNA expression profiles of TJP1 and FoxP3 were between
the two extremes, CD patients with T1D and control individuals.
Overall, a significant negative correlation was found between TJP-1
were analyzed (r520.42, P50.003).
IgA and IgG type antibodies against deamidated gliadin, tTG and
Serum levels of IgA and IgG antibodies against deamidated gliadin
were significantly higher in CD patients and CD patients with T1D
compared to patients with a normal small-bowel mucosa (Figure 2).
In patients with CD and T1D, the levels of IgA antibodies against
deamidated gliadin were significantly higher compared with CD
patients without T1D (P50.003). Using multiple regression analysis,
age-adjustment showed that the values of deamidated gliadin-specific
IgA and IgG antibodies were not dependent on age (b50.15, P50.35
and b50.048, P50.76, respectively). However, there was a significant
negative correlation between the expression of TJP1 mRNA and
Figure 1 Relative expression of TJP1 mRNA and FoxP3 mRNA in small-bowel
biopsy samples in children with normal intestinal mucosa and children with
isons between the control group (normal intestinal mucosa) and children with
T1D. CD, celiac disease; FoxP3, forkhead box P3; T1D, type 1 diabetes; TJP1,
tight junction protein 1.
Figure 2 Comparative values (mean6s.e.) of IgA (a) and IgG (b) antibodies
against deamidated gliadin and tTG expressed in RU/ml and AU, respectively.
CD, celiac disease; tTG, tissue transglutaminase.
Celiac disease in type 1 diabetes
R Uibo et al
Cellular & Molecular Immunology
the level of deamidated gliadin-specific IgA antibodies (r520.24,
in CD patients with T1D (81.2613.9). Patients with CD alone had
Levels of IgG antibodies against tTG did not differ between CD
patients with and without T1D. In CD patients, the levels of anti-
tTG IgA antibodies were inversely correlated to TJP1 mRNA express-
ion in biopsy specimens (r520.84, P50.03); this correlation is also
independent of patient age (b50.06, P50.7). The level of IgA, but not
IgG antibodies against bovine beta-lactoglobulin, was significantly
elevated in CD patients compared with patients with small bowel
normal mucosa (P50.01 and P50.21, respectively; Figure 3).
Influence of tTG antibodies on the transamidating activity of tTG
did not differ between CD patients with and without T1D. In both
groups, the effect of tTG specific antibodies on the transamidating
activity of tTG was found to be not significant and within the range
of the effect demonstrated by normal sera (86–98%).
Antibodies reacting with B. adolescentis-derived proteins
Altogether, a total of 43 different proteins with molecular masses
ranging between 13 and 90 kDa from B. adolescentis lysate reacted
with serum samples in the immunoblot assay. The reactivity pattern
among different individuals within separate groups was very diverse.
B. adolescentis proteins with molecular masses between 29 and 42 kDa
were more reactive as compared with other proteins. However, no
differences in reactivity toward individual proteins between the
groups or the immunoglobulin isotypes were revealed. Similarly, the
mean number of reactive proteins between groups of subjects with
normal small-bowel mucosa (control group) and those with subtotal
villous atrophy was not statistically different. However, the mean
number of reactive bands was somewhat higher in patients with
abnormal intestinal mucosa as compared with normal individuals;
for example, IgA antibodies revealed 7.163.3 versus 5.562.4 reactive
bands, while IgG antibodies revealed 12.464.0 versus 8.766.1 bands,
As an autoimmune disorder, CD tends to be associated with other
diseasesthatareautoimmune inorigin. Amongthese,T1Dissetapart
because the prevalence of CD among T1D patients has been reported
to be about 10 times higher compared to that in the general popu-
lation.44Therefore, it is plausible to hypothesize that this frequent
association is due to common determinants that operate in both dis-
eases. Indeed, there are at least seven common genetic loci determin-
ing the susceptibility of both diseases.3However, this finding cannot
patients to develop T1D.2It is also unclear why patients with recently
developed T1D have a higher likelihood of developing CD than
patients with long-lasting T1D.36,45It is possible that a subgroup of
the development T1D and CD in the same person.
As the intestinal mucosa appears to play a central role in the
pathogenesis of both T1D and CD, we investigated individuals with
normal small-bowel mucosa and patients with intestinal villous
atrophy representing CD with or without T1D. Comparing the
expression of TJP1 mRNA as a marker of intestinal mucosa integrity,
we revealed that intestinal permeability was likely to be most severely
impaired in patients with T1D and CD. As a possible reaction to these
mucosal changes and the related increase of antigenic pressure
through impaired intestinal mucosa barrier function, the FoxP3
mRNA expression as a reflection of number of FoxP31regulatory T
lymphocytes was the highest among patients with T1D and CD. Also,
levels of IgA antibodies directed against the CD-specific antigens tTG
and deamidated gliadin were the highest in CD patients with T1D.
These findings support the hypothesis that the immune system in the
intestinal mucosa in such patients is highly activated due to the dis-
turbed integrity of intestinal barrier function. Whether this is due to
directed toward central molecules in the function of the intestinal
mucosa (possibly associated with tight junctions) is not known.
First, a portion of T1D patients have genetically or environmentally
conditioned high expression of zonulin, a molecule that has rapid and
reversible effects on intestinal tight junctions’ TJP1.4Indeed, studies on
the development of diabetes in BBreeding (BB) rat and humans have
revealed that impairment of intestinal integrity is zonulin-dependent.
Furthermore, zonulin is liberated in high amounts during infections,
which fits well the theory about the involvement of intestinal infective
agents in the development of T1D.4,26,46However, we did not have the
ability to study zonulin levels in the present study. It would be of
particular interest for further studies to determine whether CD patients
with or without T1D have different serum zonulin levels and whether
these levels are related to the presence of infectious agents (rotaviruses,
enteroviruses, etc.) in the intestinal mucosa of studied CD patients.
microflora composition between CD patients with and without T1D
despite our present finding that there were no statistical differences
between controls and patients. In our study, we examined only one
commensal strain of Bifidobacterium that does not reflect the full
scenario of changes in intestinal mucosa. However, the B. adolescentis
strain utilized has been convincingly demonstrated to be connected
with the developmentof immune dysregulation in immune-mediated
diseases.32,33However, it has also been shown that the ability of B.
adolescentis derived proteins tostimulate the immunesystem isstrain
specific and geographically variable.30The great variation in the levels
of IgA and IgG antibodies directed against the Bifidobacteria strain
between individual patients indicates a dependence of the immune
response on host-specific commensal microbiota. Furthermore, the
results of somewhat higher levels of IgA antibody reactivity among
antibodies expressed in OD units. CD, celiac disease.
Celiac disease in type 1 diabetes
R Uibo et al
Cellular & Molecular Immunology
to continue such studies in a larger context using other Bifidobacteria
biota have received attention as possible modulators of disease
development.27–30However, one could ask whether the detection of
serum antibodies against antigens of a particular commensal bacteria
reflects its presence and involvement in immune reactions at the
intestinal level. Although it is possible that we detected antibodies
induced by antigens that originated from other microorganisms, we
are still confident that a portion of the detected antibodies was direc-
ted against B. adolescentis itself or other Bifidobacterium spp. The role
of different strains of Bifidobacteria in modulating gliadin induced
inflammatory responses in intestinal epithelial cells has been recently
demonstrated in vitro.47For further studies, more information about
the specific proteins of B. adolescentis will be required.
Third, it is possible that autoantibodies against tTG have a different
roleinCDalone comparedto CDdevelopedinpatientswithT1D.Our
study results, however, did not reveal significant differences between
sera obtained from controls and CD patients with and without T1D as
evaluated by their tTG transamidating capacity in vitro. At the same
time we do not know whether there were differences between control
and patient serum samples in the deamidating capacity of tTG.
Still, one cannot exclude the presence of specific autoimmune reac-
tions in patients with CD and T1D association. T1D patients have
antibodies against different wheat proteins in the absence of CD.48,49
Although it could be taken as a secondary phenomenon related to
increased intestinal permeability, recent studies concerning immuno-
logical crossreactivity between wheat proteins and self-proteins in
T1D50suggest that autoimmune reactions to unknown antigenic pro-
teins in intestinal mucosa might also be involved in small-bowel
mucosa destruction in some T1D patients.
Although autoantibodies against the intestinal permeability modu-
lating molecule zonulin have also been detected in patients with CD
and T1D,51the actual role of these autoantibodies in the development
of these diseases remains to be investigated.
Taken together, our investigation demonstrates that in patients
with T1D, the development of CD is combined with changes in the
small-intestine mucosa characterized by increased intestinal per-
meability, activation of the humoral immune system with elevated
levels of IgA and increased mRNA expression of the regulatory cell
marker FoxP3. Further study is required to determine whether these
extreme changes in this CD subgroup are due to specific envir-
onmental factors, intestinal microbiota, unknown genetic effects or
autoimmune reactions to novel antigenic targets in intercellular tight
junctions or related structures.
This work was supported by grants 7749 and 8334 from the Estonian Science
Foundation and by the European Union through the European Regional
Development Fund and FP. The help of Mrs Tiina Ra ¨go, MD, Mrs Anu
Kaldmaa, Ms Kadri Eoma ¨e, andMrs Anu Ko ˜iveer in thepreparationof clinical
material, cell culture and the performance of antibody assays is greatly
appreciated. We also thank our colleagues from Finland, Professor Jorma
Ilonen from Turku University, Professor Outi Vaarala from the National
Institute for Health and Welfare and Professor Erkki Savilahti from the
Children’s Hospital, University of Helsinki, for their support.
Green PH, Cellier C. Celiac disease. N Engl J Med 2007; 357: 1731–1743.
diabetes mellitus in children. J Ped Endocrinol Metab 2007; 20: 1257–1264.
3 Smyth DJ, Plagnol V, Walker NM, Cooper JD, Downes K, Yang JH et al. Shared and
distinct genetic variants in type 1 diabetes and celiac disease. N Engl J Med 2008;
Visser J, Rozing J, Sapone A, Lammers K, Fasano A. Tight junctions, intestinal
permeability, and autoimmunity. Celiac disease and type 1 diabetes paradigms.
Ann NY Acad Sci 2009; 1165: 195–205.
Lefebvre DE, Powell KC, Strom A, Scott FW. Dietary proteins as environmental
modifiers of type 1 diabetes mellitus. Ann Rev Nutr 2006; 26: 175–202.
of celiacdiseaseautoimmunityina USAcohortfollowedsince1974.AnnMed2010;
Borchers AT, Uibo R, Gershwin ME. The geoepidemiology of type 1 diabetes.
Autoimmun Rev 2010; 9: A355–A365.
Lower economic status and inferior hygiene environment may protect against celiac
disease. Ann Med 2008; 40: 223–231.
Ress K, Harro M, Maaroos HI, Harro J, Uibo R, Uibo O. High prevalence of celiac disease:
need for increasing awareness among physicians. Dig Liv Dis 2007; 39: 136–139.
Jabri B, Sollid LM. Tissue-mediated control of immunopathology in coeliac disease.
Nat Rev Immunol 2009; 9: 858–870.
Lewis NR, Scott BB. Meta-analysis: deamidated gliadin peptide antibody and tissue
transglutaminase antibody compared as screening tests for coeliac disease. Aliment
Pharmacol Ther 2010; 31: 73–81.
Clin Gastroenterol 1995; 9: 273–293.
Tiittanen M, Westerholm-Ormio M, Verkasalo M, Savilahti E, Vaarala O. Infiltration of
in type 1 diabetes. Clin Exp Immunol 2008; 152: 498–507.
mellitus. Scand J Gastroenterol 2009; 44: 422–430.
Westerholm-Ormio M, Vaarala O, Tittanen M, Savilahti E. Infiltration of Foxp3- and
Toll-like receptor-4-positive cells in the intestines of children with food allergy.
J Pediatr Gastroenterol Nutr 2010; 50: 367–376.
et al. Intestinal tissue transglutaminase in coeliac disease of children and adults:
ultrastructural localization and variation in expression. Scand J Gastroenterol 2004;
Pinkas DM. Strop P, Brunger AT, Khosla C. Transglutaminase 2 undergoes a large
conformational change upon activation. PloS Biol 2007; 5: e327.
Sartor RB. Microbial influences in inflammatory bowel disease. Gastroenterology
2008; 134: 577–594.
Me ´nard S, Cerf-Bensussan N, Heyman M. Multiple facets of intestinal permeability
and epithelial handling of dietary antigens. Mucosal Immunol 2010; 3: 247–259.
Lammers KM, Lu R, Brownley J, Lu B, Gerard C, Thomas K et al. Gliadin induces an
increase in intestinal permeability and zonulin release by binding to the chemokine
receptor CXCR3. Gastroenterology 2008; 135: 194–204.
Schulzke JD, Bentzel CJ, Schulzke I, Riecken EO, Fromm M. Epithelial tight junction
structure in the jejunum of children with acute and treated celiac sprue. Pediatr Res
1998; 43: 435–441.
Secondulfo M, Iafusco D, Carratu ` R, deMagistris L, Sapone A, Generoso M et al.
Ultrastructural mucosal alterations and increased intestinal permeability in non-
coeliac, type 1 diabetic patients. Dig Liv Dis 2004; 36: 35–45.
Bosi E, Molteni L, Radaelli MG, Folini L, Fermo L, Bazzigaluppi E et al. Increased
Sapone A, de Magistris L, Pietzak M, Clemente MG, Tripathi A, Cucca F et al. Zonulin
upregulation is associated with increased gut permeability in subjects with type 1
diabetes and their relatives. Diabetes 2006; 55: 1443–1449.
Neu J, Reverte M, Mackey AD, Liboni K, Tuhacek-Tenace LM, HatchM et al. Changes
in intestinal morphology and permeability in the biobreeding rat before the onset of
type 1 diabetes. J Pediatr Gastroenterol Nutr 2005; 40: 589–595.
Watts T, Berti I, Sapone A, Gerarduzzi T, Not T, Zielke R et al. Role of the intestinal
prone rats. Proc Natl Acad Sci USA 2005; 102: 2916–2921.
an enteric bacterial pathogen accelerates insulitis in NOD mice. Diabetologia 2010;
Ou G, Hedberg M, Ho ¨rstedt P, Baranov V, Forsberg G, Drobni M et al. Proximal small
intestinal microbiota and identification of rod-shaped bacteria associated with
childhood celiac disease. Am J Gastroenterol 2009; 104: 3058–3067.
for determination of IgG antibodies against lactobacilli representing human
indigenous microflora. Pediatr Allergy Immunol 2010; 21: e772–e779.
Giongo A, Gano KA, Crabb DB, Mukherjee N, Novelo LL, Casella G et al. Toward
defining the autoimmune microbiome for type 1 diabetes. ISME J 2010; 5: 82–91.
Young SL, Simon MA, Baird MA, Tannock GW, Bibiloni R, Spencely K et al.
Bifidobacterial species differentially affect expression of cell surface markers and
cytokines of dendritic cells harvested from cord blood. Clin Diagn Lab Immunol
2004; 11: 686–690.
He F, Ouwehand AC, Isolauri E, Hashimoto H, Benno Y, Salminen S. Comparison of
mucosal adhesion and species identification of bifidobacteria isolated from healthy
and allergic infants. FEMS Immunol Med Microbiol 2001; 30: 43–47.
Celiac disease in type 1 diabetes
R Uibo et al
Cellular & Molecular Immunology
33 Stsepetova J, Sepp E, Julge K, Vaughan E, Mikelsaar M, de Vos WM. Molecularly Download full-text
assessed shifts of Bifidobacterium spp. and less diverse microbial communities are
characteristic of 5-year-old allergic children. FEMS Immunol Med Microbiol 2007;
Vaarala O, Atkinson M, Neu J. The ‘‘perfect storm’’ for type 1 diabetes. The complex
interplay between intestinal microbiota, gut permeability, and mucosal immunity.
Diabetes 2008; 10: 2555–2562.
Arrieta MC, Bistritz L, Meddings JB. Alterations in intestinal permeability. Gut 2006;
Uibo O, Heilman K, Ra ¨go T, Shor R, Paal M, Metsku ¨la K et al. Symptomless coeliac
diseasein typeI diabetes: 12-yearexperience in Estonia.PediatrInt 2010; 52: 230–
Marsh MN. Gluten, major histocompatibility complex, and the small intestine: a
molecular and immunobiologic approach to the spectrum of gluten sensitivity
(‘‘celiac sprue’’). Gastroenterology 1992; 102: 330–354.
Green PH, Rostami K, Marsh MN. Diagnosis of coeliac disease. Best Pract Res Clin
Gastroenterol 2005; 19: 389–400.
Walker-Smith J, Guandalini S, Schmitz J, Shmerling DH, Visakorpi JK. Revised
criteria for diagnosis of celiac disease. Report of working group of European Society
of Pediatric Gastroenterology and Nutrition. Arch Dis Child 1990; 65: 909–911.
Teesalu K, Agardh D, Panarina M, Utt M, Uibo O, Uibo R. A modified ELISA for
improved detection of IgA, IgG, IgM anti-tissue transglutaminase antibodies in
celiac disease. Clin Chim Acta 2009; 403: 37–41.
Nilsson I, Kornilovs’ka I, Lindgren S, Ljungh A, Wadstro ¨m T. Increased prevalence of
seropositivity for non-gastric Helicobacter species in patients with autoimmune liver
disease. J Med Microbiol 2003: 52: 949–953.
Kira ´ly R, Vecsei Z, Deme ´nyi T, Korponay-Szabo ´ IR, Fe ´su ¨s L. Coeliac autoantibodies
can enhance transamidating and inhibit GTPase activity of tissue transglutaminase:
dependence on reaction environment and enzyme fitness. J Autoimmun 2006; 26:
Caco-2 cell functional characteristics. Cell Biol Toxicol 2005; 21: 1–26.
Cronin CC, Shanafan F. Insulin-dependent diabetes mellitus and coeliac disease.
Lancet 1997; 349: 1096–1097.
of celiac disease after onset of type 1 diabetes: a 6-year prospective longitudinal
study. Pediatrics 2002; 109: 833–838.
Guttman JA, Finlay BB. Tight junctions as targets of infectious agents. Biochim
Biophys Acta 2009; 1788: 832–841.
Laparra JM, Sanz Y. Bifidobacteria inhibit the inflammatory response induced by
gliadins in intestinal epithelial cells via modifications of toxic peptide generation
during digestion. J Cell Biochem 2010; 109: 801–807.
Barnes RM, Harvey MM, Blears J, Finn R, Johnson PM. IgG subclass of human serum
antibodies reactive with dietary proteins. Int Arch Allergy Appl Immunol 1986; 81:
Savilahti E, Saukkonen TT, Virtala ET, Tuomilehto J, Kerblom HK. The Childhood
Diabetes in Finland Study Group. Increased levels of cow’s milk and b-lactoglobulin
antibodies in young children with newly diagnosed IDDM. Diabetes Care 1993; 16:
of Glo-3A antibody levels in children at increased risk for type 1 diabetes mellitus.
Pediatr Diabetes 2009; 10: 563–572.
Fasano A, Not T, Wang W, Uzzau S, Berti I, Tommasini A et al. Zonulin, a newly
discovered modulator of intestinal permeability, and its expression in coeliac
disease. Lancet 2000; 355: 1518–1519.
Celiac disease in type 1 diabetes
R Uibo et al
Cellular & Molecular Immunology