Ultrastructural injury to interstitial cells of Cajal and
communication with mast cells in Crohn’s disease
X.-Y. WANG,*1 N. ZARATE,*1 J. D. SODERHOLM,� J. M. BOURGEOIS,� L. W. C. LIU* & J. D. HUIZINGA*
*Intestinal Disease Research Program, McMaster University, Hamilton, ON, Canada
�Department of Biomedicine and Surgery, Division of Surgery, University Hospital, Linkoping, Sweden
�Department of Pathology and Molecular Medicine, McMaster University, ON, Hamilton, Canada
Abstract Crohn’s disease associated dysmotility has
been attributed to fibrosis and damage to enteric
nerves but injury to interstitial cells of Cajal (ICC)
could also be involved. We assessed ICC in specimens
obtained from patients with Crohn’s disease and
determined the relation between ICC and the
inflammatory infiltrate, particularly mast cells (MC)
using quantitative immunohistochemistry and elec-
tron microscopy. Ultrastructural injury to ICC was
patchy in all ICC subtypes but ICC-Auerbach’s plexus
(AP) showed damage more frequently, i.e. swelling of
mitochondria, decreased electron density, autophago-
somes and partial depletion of the cytoplasm. Light
microscopy confirmed a significant decrease in c-kit
immunoreactivity for ICC-AP and an increased
number of MC in the muscularis externa. Electron
microscopy showed MC exhibiting piecemeal degran-
ulation and making frequent and selective membrane-
to-membrane contact with all types of injured ICC
which suggests chronic release of granule content to
affect ICC. Extent of ICC injury was not associated
with duration of the disease. In conclusion, ultra-
structural injury and loss of ICC-AP is evident in
Crohn’s disease. Epidemiological and morphological
data suggest that ICC have the capacity to regenerate
in spite of the chronic insult. The muscularis hosts a
marked number of MC that exhibit piecemeal
degranulation associated with ICC and may facilitate
ICC maintenance.
Keywords inflammatory bowel disease, interstitial
cells of Cajal, mast cells.
INTRODUCTION
Evidence of dysmotility at different levels of the
gastrointestinal tract has been reported for both ulcer-
ative colitis and Crohn’s disease. Annese et al.1 des-
cribed a remarkable increase in the number of single
and clustered propagated contractions in patients with
inactive Crohn’s disease. In vitro mechanical studies
aimed to assess the contractility of the circular and
longitudinal muscle layers have shown abnormal
receptor-mediated muscle responses in muscle strips
obtained from Crohn’s disease patients.2 Morphologi-
cal studies have reported severe damage to enteric
neural structures3,4 and a reduced population of inter-
stitial cells of Cajal (ICC).5 Similarly, disturbed colonic
motor activity has been reported in patients with
ulcerative colitis,6 even during remission,7 particularly
in the form of low-amplitude propagating contractions
that might contribute to diarrhoea.
Motility of the gastrointestinal tract involves com-
plex processes that require the structural integrity and
functionality of each cellular element involved.
Mechanical activity of the smooth muscle cells is in
part controlled by the enteric nervous system and the
ICC. The latter are classified in different subtypes
according to their location in the gut and distribution
within the muscularis externa. In the small intestine
there are three different populations of ICC. ICC-AP,
distributed around the Auerbach’s plexus ganglia,
generate the rhythmic slow waves of depolarization
Address for Correspondence
Natalia Zarate, Centre for Academic Surgery, 3rd Floor Alex-andra Wing, The Royal London Hospital, Whitechapel,London E1 1BB, UK.Tel: +44 (0) 20 7377 7184; fax: +44 (0) 20 7377 7346;e-mail: [email protected]: 4 August 2006Accepted for publication: 23 November 2006
1These authors contributed equally to this work.
This work was presented at the 19th InternationalNeurogastroenterology and Motility Meeting, held inBarcelona, October 2003 and has been published in part, inabstract form: Neurogastroenterol. Motil. 2003;15:86.
Neurogastroenterol Motil (2007) 19, 349–364 doi: 10.1111/j.1365-2982.2006.00894.x
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 349
and repolarization that perform a pacemaker function
and subsequently determine the frequency and direc-
tion of peristaltic contractions. Within the circular and
longitudinal muscle layers ICC-IM run parallel to
smooth muscle cells and may take part in controlling
innervation.8 Finally, ICC associated with the deep
muscular plexus (ICC-DMP) are located at the inner
border of the circular muscle layer adjacent to the
submucosa where they can simultaneously establish
synapse-like contacts with nerve terminals and gap
junction contacts with smooth muscle cells.8,9 Such a
specific position and specialized contacts suggest a
principal role for ICC-DMP as modulators of neural
stimuli directed to smooth muscle cells.
Inflammation in Crohn’s disease is transmural and
chronic, and can potentially damage all ICC subtypes.
Assessment of ICC integrity in Crohn’s disease and
evaluation of both potential mechanisms influencing
their damage and the nature of their interactions
with different immune cells is warranted. Mast cells
(MC) are among the inflammatory infiltrate in
Crohn’s disease.10 We recently reported an intimate
association between MC and ICC-IM in achalasia, a
primary oesophageal motor disorder characterized by
degeneration of the oesophageal Auerbach’s Plexus
and ICC-IM injury.11 The role of MC in Crohn’s
disease is uncertain as they can participate in both
inflammatory and reparative processes.12,13 The pur-
pose of the present study was firstly to analyse the
effects of inflammation associated with Crohn’s
disease on the integrity of ICC in inflamed ileal
segments and secondly, to characterize the relations
between ICC and the inflammatory infiltrate, parti-
cularly MC.
METHODS
Surgical specimens
Surgical specimens from the terminal ileum were
obtained from 11 patients with Crohn’s disease
(Table 1). All patients fulfilled clinical, endoscopic
and histological criteria for the diagnosis of Crohn’s
disease and required surgery for disease-related com-
plications resistant to medical therapy and/or for
alleviation of persistent symptoms of mechanical
obstruction. Control surgical samples were obtained
from patients having surgery for carcinoma of the large
bowel. Informed consent was obtained from all partic-
ipants and the study was approved by the ethics
committee of McMaster University (Canada) and
University Hospital (Sweden). Human tissues were
handled according to institutional ethical guidelines.
Samples studied from patients with Crohn’s disease
were selected from areas showing macroscopic lesions,
i.e. ulcers and fissures. Whole mount preparations
exhibiting mucosal ulceration and severe fibrosis were
extremely difficult to dissect. Suitable preparations
were obtained only in two patients and they were
chosen on the basis that fibrosis did not impede
dissection. These sections without superficial ulcers
were immediately adjacent to areas exhibiting severe
ulcers or fissures.
Electron microscopy
Surgical specimens were placed in ice-cold physiologi-
cal solution and the mucosa was removed. Tissues
were cut and fixed by immersion in 2.5% glutaralde-
hyde in 0.075 mol L)1 sodium cacodylate buffer (pH
7.4, containing 4.5% sucrose and 1 mmol L)1 CaCl2)
for 6 h at room temperature. After primary fixation,
1 cm wide pieces were cut and immersed in the same
fixative for an additional fixation overnight at 4 �C;
then, all tissues were washed in 0.1 mol L)1 cacodylate
buffer, containing 6% sucrose and 1.24 mmol L)1
CaCl2 (pH 7.4) at 4 �C. They were postfixed with 1%
osmium tetroxide in 0.05 mol L)1 cacodylate buffer
(pH 7.4) for 90 min, stained with saturated uranyl
acetate for 60 min, dehydrated in graded ethanol and
propylene oxide, and embedded in Epon-Araldite (Mar-
ivac Ltd, Halifax, NS, Canada). Semithin sections were
cut and stained with 1% toluidine blue for light
microscopic examination. Following this, ultra thin
sections were cut, mounted on 100 mesh grids, and
double stained with uranyl acetate and lead citrate.
The grids were examined with a transmission electron
microscope (JEOL-1200EX Biosystem, Tokyo, Japan) at
80 kV.
Light microscopy
Immunoreactivity and quantification of kit positive
interstitial cells of Cajal Surgical samples were
immersed in modified Zamboni’s fixative containing
4% paraformaldehyde and 0.2% picric acid in
0.1 mol L)1 phosphate buffer (pH 7.4) for 2–3 h at 4 �C.
For whole mount preparations, mucosa, submucosa
and serosa were removed and the tunica muscularis
was prepared for immunostaining. For paraffin sec-
tions, tissue was embedded in wax. Five micrometre
sections were cut and mounted on coated slides,
de-waxed and re-hydrated before treated with 1% H2O2
to quench the endogenous peroxidase. Antigen
retrieval was performed on paraffin sections prior to
immunostaining by heating the slides in citrate buffer
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd350
Tab
le1
Pat
ien
tpro
file
san
dim
mu
no-q
uan
tifi
cati
on
Age
Du
rati
on
(yea
rs)
Sym
pto
ms
Tre
atm
ent
ICC
c-k
itim
mu
nore
acti
vit
y(%
tota
lar
eaon
par
affi
nse
ctio
ns)
ICC
c-k
itim
mu
nore
acti
vit
y(%
tota
lvolu
me
inw
hole
mou
nt)
Infl
amm
atio
ngr
ade
MC
qu
anti
fica
tion
(%to
tal
area
on
par
affi
nse
ctio
ns)
Pat
ien
t1
25
1.2
Ilea
lst
rict
ure
;re
sist
ance
toT
xP
redn
isolo
ne
5-A
SA
CM
0.4
9±
0.1
4A
P0.5
6±
0.4
1L
M0.0
8±
0.0
2
3C
M0.2
6±
0.0
8A
P0.0
9±
0.0
1L
M0.4
3±
0.0
6P
atie
nt
223
2A
bce
ss,
ilea
lfi
stu
la;
resi
stan
ceto
Tx
Pre
dn
isolo
ne
5-A
SA
Aza
thio
pri
ne
CM
0.3
0±
0.1
5A
P1.6
6±
0.1
0L
M0.4
7±
0.2
5
2–3
CM
0.3
0±
0.0
6A
P0.1
5±
0.0
7L
M0.4
1±
0.0
6P
atie
nt
337
21
Ilea
lst
rict
ure
and
fist
ula
;re
sist
ance
toT
xP
redn
isolo
ne
Aza
thio
pri
ne
Cip
rofl
oxac
in
CM
0.8
0±
0.4
1A
P5.3
3±
1.7
6L
M2.9
4±
1.6
3
2C
M0.4
5±
0.0
5A
P0.5
0±
0.0
9L
M0.4
7±
0.3
0P
atie
nt
431
1.8
Str
ictu
rew
ith
acti
ve
infl
amm
atio
nP
redn
isolo
ne
5-A
SA
Met
ron
idaz
ole
and
gen
tam
icin
CM
1.2
5±
0.1
8A
P1.8
2±
0.4
5L
M0.9
5±
0.8
2
2–3
CM
0.4
9±
0.1
1A
P0.5
0±
0.0
4L
M0.5
1±
0.0
3
Pat
ien
t5
26
20
Str
ictu
rew
ith
ileo
colo
nic
fist
ula
Bu
des
on
ide
5-A
SA
CM
0.8
7±
0.7
4A
P3.7
9±
1.7
4L
M0.9
2±
0.3
5
1–2
CM
0.5
1±
0.1
0A
P0.1
1±
0.0
5L
M0.5
8±
0.2
0P
atie
nt
624
0.3
Su
spic
iou
sap
pen
dic
itis
Non
eC
M0.6
5±
0.2
3A
P1.1
5±
0.0
6L
M0.5
8±
0.2
3
2C
M0.3
4±
0.0
2A
P0.2
8±
0.2
6L
M0.2
9±
0.1
2P
atie
nt
756
7P
ost
oper
ativ
efi
stu
la.
Sev
eral
pre
vio
us
rese
ctio
ns
Aza
thio
pri
ne
CM
0.5
5±
0.2
7A
P1.7
±0.5
4L
M0.5
6±
0.1
6
3C
M0.3
6±
0.1
0A
P0.1
8±
0.0
8L
M0.4
6±
0.0
5P
atie
nt
818
0.4
Su
spic
iou
sap
pen
dic
itis
Non
eC
M0.6
4±
0.5
1A
P1.0
3±
0.1
7L
M0.4
9±
0.2
9
2–3
CM
0.1
7±
0.1
7A
P0.0
2±
0.0
2L
M0.3
7±
0.0
7P
atie
nt
923
3Fis
tula
ileu
m-s
igm
oid
colo
nN
on
eC
M0.4
5±
0.1
6A
P1.3
1±
0.5
2L
M0.6
2±
0.3
6
3C
M0.2
5±
0.0
7A
P0.0
7±
0.0
0L
M0.2
5±
0.2
0P
atie
nt
10
38
4Il
eal
stri
ctu
reB
udes
on
ide
5-A
SA
14.5
5±
2.6
2
Pat
ien
t11
21
1M
ult
iple
ilea
lfi
stu
las
Pre
dn
isolo
ne
5-A
SA
12.9
9±
2.8
1
Con
trol
155
CM
0.6
5±
0.5
1A
P5.8
±1.4
LM
0.1
7±
0.0
9
CM
0.0
9±
0.0
3A
P0
LM
0.1
±0.0
3C
on
trol
260
CM
0.4
5±
0.1
7A
P3.7
2±
1.1
LM
0.3
4±
0.0
3
CM
0A
P0
LM
0
Volume 19, Number 5, May 2007 ICC injury in Crohn’s disease
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 351
(pH 6.0). Both whole mount preparations and paraffin
sections were then washed for 30 min in phosphate-
buffered saline (PBS; 0.05 mol L)1, pH 7.4, with 0.3%
Triton X-100). Non-specific antibody binding was re-
duced by incubating the tissues in 1% bovine serum
albumin for 1 h at room temperature before addition of
the primary antibodies. Tissues were then incubated in
polyclonal rabbit anti c-Kit (c-19, 1 : 500; Santa Cruz
Biotechology, Santa Cruz, CA, USA) for 48 h at 4 �C.
Secondary immunoreactions were carried out with
fluorescin isothiocynate (FITC) conjugated anti-rabbit
IgG in the whole mount preparations or with the
Vectastain ABC kits (Vector Laboratories, Burlingame,
CA, USA), in which there are biotinylated anti-rabbit
IgG, in the paraffin sections. All the secondary anti-
bodies were from Vector Laboratories (Burlingame, CA,
USA). 3,3¢-diaminobenzidine (0.05% DAB plus 0.01%
H2O2 in 0.05 mol L)1 Tris buffer saline, pH 7.6) was
used as a peroxidase substrate. All the antisera were
diluted with 1% normal goat serum and 10% human
serum in 0.3% Triton X-100 (PBS-TX, pH 7.4).
Paraffin sections were examined with a conven-
tional microscope. Pictures at ·10 were taken through
a digital camera (Sony 3CCD, Model no. DXC-930;
Tokyo, Japan) attached to the microscope using
MetaMorph Imaging System version 6.0 software
(Universal Imaging Corporation, Downingtown, PA,
USA). Pictures were taken randomly from areas
representing the circular muscle layer, the longitud-
inal muscle layer and the myenteric plexus region.
Quantification was performed using Photoshop ver-
sion 7.0 (Adobe Systems, Mountain View, San Jose,
CA, USA).14 In brief, c-kit immunoreactivity was
identified, highlighted and the area of immuno-
reactivity measured and expressed as percentage of
the total area. As c-kit immunohistochemistry recog-
nizes both ICC and MC, quantification values for MC
assessed on consecutive sections stained with tolui-
dine blue were subtracted from those obtained with
c-kit quantification for each region studied.
Whole mount preparations were examined with a
confocal microscope (Zeiss LSM 510, Jena, Germany)
with an excitation wavelength appropriate for FITC
(494 nm) and confocal micrographs were created with
Carl-Zeiss software. These are digital composites of
Z-series scans of 15–20 optical sections through a depth
of 30–60 lm. These pictures evenly covered the total
thickness of the preparation. Quantification was done
measuring the area of c-kit immunoreactivity and, as
the scanning depth of each preparation was known, it
was expressed as percentage of the total volume of the
preparation. To obtain quantification data just for ICC,
MC were erased from pictures before analysis.Tab
le1
Con
tin
ued A
geD
ura
tion
(yea
rs)
Sym
pto
ms
Tre
atm
ent
ICC
c-k
itim
mu
nore
acti
vit
y(%
tota
lar
eaon
par
affi
nse
ctio
ns)
ICC
c-k
itim
mu
nore
acti
vit
y(%
tota
lvolu
me
inw
hole
mou
nt)
Infl
amm
atio
ngr
ade
MC
qu
anti
fica
tion
(%to
tal
area
on
par
affi
nse
ctio
ns)
Con
trol
353
CM
0.7
5±
0.4
1A
P3.5
3±
0.1
5L
M0.2
6±
0.1
0
CM
0A
P0
LM
0C
on
trol
464
CM
1.0
5±
0.1
5A
P2.9
4±
0.2
7L
M0.6
6±
0.0
9
CM
0.0
5±
0.0
1A
P0
LM
0.0
3±
0.0
1C
on
trol
519.9
4±
3.6
2C
on
trol
617.8
8±
5.1
0
AP
,A
uer
bac
h’s
ple
xu
s;IC
C,
inte
rsti
tial
cell
sof
Caj
al;
CM
,ci
rcu
lar
mu
scle
layer
s;L
M,
lon
gitu
din
alm
usc
lela
yer
s,M
C,
mas
tce
ll;
Tx,
trea
tmen
t.
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd352
Identification and quantification of mast cells Mast
cell identification using c-kit staining is technically
demanding as cross sections of ICC or ICC processes
can be mistaken for MC. Therefore, specific MC
staining was performed using 0.5% toluidine blue in
combination with a light counterstain with eosin.
Inflammation scores and grades The histological
assessment of the degree of inflammation was per-
formed using a semi-quantitative method modified
from previously published work (Tables 2 and 3).15 The
histological scores were converted to inflammation
grades (Table 3), defined as follows – 0: no inflamma-
tion; 1: slight inflammation; 2: mid-level inflamma-
tion; and 3: severe inflammation. The degree of active
disease, including neutrophilic and lymphocytic infil-
trate, oedema, epithelial atrophy and ulceration, was
assessed. Atrophy was determined by looking at the
degree of crypt and enterocyte loss. The number of
mononuclear cells in the lamina propria was used to
determine lymphocyte infiltration and transmural
lymphoid hyperplasia was assessed when there was
transmural fibrosis. Each resected specimen was scored
individually.
Statistical analysis
Quantification data are expressed as mean ± SE. Com-
parisons were made using Student’s t-test. Correlation
between different variables was performed with Pear-
son-correlation test. Statistical significance required
P < 0.05.
RESULTS
Patient characteristics are shown in Table 1. The
average age was 29 years, with a range from 18 to
56 years. Duration of disease ranged from 3 months to
21 years.
Electron microscopy
Distribution of ICC subtypes was similar to previous
descriptions in control tissue.8 ICC exhibited typical
ultrastructural features16,17 such as an oval nucleus
with condensed heterochromatin distributed in the
periphery, abundant mitochondria and conspicuous
endoplasmic reticulum (ER). Smooth ER was partic-
ularly abundant within the cell processes. Caveolae
were seen along the cell membrane (Fig. 1). Type of
contacts with nearby cells varied depending on the ICC
subtype; ICC-DMP established synapse-like contacts
with nerve varicosities from the deep muscular plexus
and frequent gap junction contacts with smooth mus-
cle cells; ICC-IM exhibited no specialized contacts but
were closely associated with neighbouring nerve ter-
minals within the muscle bundles; and ICC-AP were
found in the proximity of the AP. These types of
specialized contacts contribute to our ability to dis-
criminate ICC from fibroblast-like cells.
In all the patients studied, patchy damage involving
all three subtypes of ICC was observed. Degenerative
changes were more striking and frequent in ICC-AP.
Injury was displayed in both the perinuclear cytoplasm
and the ICC processes; the main features were swelling
of mitochondria, presence of lamellar bodies and lipid
droplets, multiple autophagosomes, depletion or vac-
uolization of the cytoplasm and significant reduction
in electron density of the perinuclear cytoplasm
(Figs 2–6). Interestingly, the decrease in cytoplasmic
Table 2 Histological scoring of disease
Epithelial glands (enterocytes)Normal 0Loss of single cells 1Slight increase 2Frank ulceration 3
CryptsNormal 0Single cells 1Cryptitis 2Crypt abscesses 3
Lamina propria cells (mononuclear cells)Normal 0Loss of groups of cells 1Moderate increase 2Marked increase 3
NeutrophilsNormal 0Slight increase 1Moderate increase 2Marked increase 3
Fibrosis with depth of inflammationNormal 0Ulceration/fibrosis mucosa 1Ulceration/fibrosis submucosa 2Ulceration/fibrosis muscularis propria 3
Table 3 Conversion of histological scores to grades
Grade Total score
0 0–21 3–52 6–83 9–15
Volume 19, Number 5, May 2007 ICC injury in Crohn’s disease
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 353
electron density observed in injured ICC was accom-
panied by a remarkable decrease in the peripheral
nuclear heterochromatin condensation (Figs 2–5). Just
a few ICC showed other severe degenerative changes
such as being devoid of most organelles, in particular
cytoplasmic filaments and mitochondria (Figs 2B and
3). Injured ICC were recognizable by their sparse
perinuclear region, the caveolae lining the membrane,
intermediate filaments and abundant mitochondria.
Membrane-to-membrane contacts were frequently
seen to be preserved between injured ICC and unin-
jured ICC (Figs 2, 3 and 5) and intimate contacts
A B
DC
Figure 1 Ultrastructure of interstitial cell of Cajal (ICC) from control human ileum. (A,B) Two ICC-AP with typical ultrastructuralfeatures are neighbouring the ganglia (G) in the Auerbach’s plexus. The cytoplasm exhibits abundant mitochondria, rER (hollowarrows) and sER. The nucleus is large, oval and with condensed heterochromatin in the periphery. Fibroblast-like cells (FLC) are inthe vicinity of ICC and ganglia and are easily distinguished by the absence of caveolae. E: Endothelium cell of a capillary. (C) AnICC-IM within the septa of the circular muscle layer (CM) is close to a nerve bundle (N). (D) An ICC-DMP is intimately associated(arrow) with an outer circular muscle cell (OCM). A nerve bundle (N) of the DMP is nearby. ICM: inner circular muscle layer.
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd354
between ICC and both nerve structures (Fig. 6) and
smooth muscle cells (Fig. 7) were still present. Nerve
structures exhibiting different degrees of injury could
be seen associated indistinctly with intact or damaged
ICC (Figs 4, 6 and 8). ICC with a markedly increased
presence of rough ER (rER) and Golgi apparatus were
identified (Fig. 5). This is interpreted as a sign of ICC
regeneration.
Mast cells were abundantly distributed throughout
the musculature. They often exhibited signs of piece-
meal degranulation with numerous granules showing
complete or partial loss of their contents without
fusion of their membranes (Figs 3, 4B and 8B). A most
striking observation was the presence of frequent
intimate membrane-to-membrane contacts between
ICC exhibiting severe perinuclear damage and MC
(Figs 2–4). This association between MC and all three
different ICC subtypes proved not to be accidental
because similar close contacts with other immune
cells were far less frequent (Fig. 7). Interestingly, all
intimate contacts observed involved injured ICC,
however it cannot be ruled out that this type of contact
might also involve intact cells as previously shown in
achalasia.11 Mast cells were also seen in the proximity
of nerve structures (Figs 4B,C and 8A,B), but direct
contacts were rarely observed (Fig. 4C).
Neural structures exhibited the highest degree of
injury compared with other cell types and this was
also evident for the AP. Swollen and empty axons,
devoid of most organelles, were frequently seen
within the myenteric plexus and to a lesser extent
in the deep muscular plexus (Fig. 8); damage of nerve
A B C
Figure 2 Intimate contact between a mast cell and an injured interstitial cell of Cajal (ICC) in the Auerbach’s plexus (AP) region.(A–C) A severely injured ICC-AP (ICC*) establishes intimate membrane-to-membrane contacts (large arrows) with an intact ICC(ICC) and a mast cell (MC) in tissue from small bowel affected with Crohn’s disease (patient 5). A fibroblast like-cell (FLC) is nearbyin a. (B) Enlarged part of A: Caveolae (small arrows) are lining the cell membranes in both normal ICC and injured ICC. Damage isdemonstrated by mitochondrial vacuolization (m) and decreased cytoplasmic electron density. (C) Enlarged part of A: intimatecontact (large arrow) between the mast cell and the damaged ICC whose process contains a lipid droplet (L) and a secondarylysosome (Ly).
Volume 19, Number 5, May 2007 ICC injury in Crohn’s disease
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 355
varicosities within both muscle layers occurred but in
a scattered manner (Figs 4C and 5). Nerves displayed
degenerative changes such as autophagosomes, lipid
vacuoles and lamella bodies (Figs 8C,D). Intimate
contacts between nerve structures and inflammatory
cells were rare.
Light microscopy
Immunoreactivity and quantification of kit positive
interstitial cells of Cajal In control tissue, ICC were
distributed around the ganglia of Auerbach’s plexus
(ICC-AP) and scattered among the smooth muscle cells
in the longitudinal and circular muscle layers (ICC-IM)
but were not detected by c-kit immunohistochemistry
at the level of the deep muscular plexus (ICC-DMP),
similar to previous observations.8
Control whole mount preparations showed numer-
ous interconnections between ICC processes creating a
three-dimensional network. However, c-kit immuno-
reactivity in preparations from patients was reduced
more than 25% showing a decreased network density
(Table 1) (Fig. 9).
C-kit positivity was studied quantitatively on
sections of paraffin-embedded tissue. Interstitial cells
of Cajal distribution paralleled which was seen on
control sections and similarly, ICC-DMP could not be
Figure 3 Intimate contact between a mast cell and an injuredinterstitial cell of Cajal in the circular muscle layer. A se-verely injured ICC-IM (ICC*) establishes intimate contacts(large arrows) with a mast cell (MC) and an intact ICC (ICC)assessed in an ultra thin section from small bowel macro-scopically affected with Crohn’s disease (patient 1). The mastcell shows signs of piecemeal degranulation as numerousintra-cytoplasmic vesicles are partially empty and there is nofusion between vesicles or with the cellular membrane.Note the dramatic reduction in electron density and thepresence of mitochondrial vacuolization (m) in the injuredICC. Intact ICC exhibit morphologically normal mitochon-dria. Both ICC show caveolae along their membranes (smallarrows).
A B C
Figure 4 Intimate contacts (large arrows) between mast cells (MC) and different subtypes of injured interstitial cell of Cajal (ICC*).(A) A mast cell establishes membrane-to-membrane contact with a severely injured ICC-AP exhibiting decreased electron density,swollen mitochondria (m) and lamella bodies (*) in an section from macroscopically affected Crohn’s disease. (B) A damagedICC-IM showing swollen mitochondria (m) contacts a mast cell exhibiting signs of piecemeal degranulation. A nerve terminal isnearby and intimate contact is not established with neither the ICC nor the mast cell. (C) An injured ICC-DMP exhibiting reducedcytoplasmic electron density and vacuolization intimately contacts a mast cell that simultaneously contacts a nerve fibre (N). Allfigures are from patient 5. Large amounts of collagen and elastic fibres are evident surrounding the aforementioned cells. Smallarrows: caveolae.
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd356
identified and therefore were not quantified. Results
confirmed the electron microscopy data; c-kit immu-
noreactivity for ICC associated with AP was signifi-
cantly decreased in sections obtained from patients
(Fig. 10; Table 1). Percentage area occupied by c-kit
positivity were as follows: 0.36 ± 0.11% (control);
0.85 ± 0.26% (inflamed tissue) for the longitudinal
muscle layer (P > 0.05); 4.0 ± 0.62% (control);
2.04 ± 0.51% (inflamed tissue) for the AP region
(P < 0.05); and 0.73 ± 0.13% (control); 0.67 ± 0.09%
(inflamed tissue) for the circular muscle layer
(P > 0.05). There was no correlation between the inflam-
mation score, type of treatment and age on the one side
and the degree of ICC reduction in the AP region on the
other. Interestingly, c-kit positivity for ICC was in-
versely related with the duration of the disease (Pearson
correlation coefficient 0.9; P ¼ 0.001) as ICC was within
the normal range in the two patients with longer
duration of the disease (Table 1; Figs 11 and 12).
Mast cell distribution and quantification Mast cells
were mainly present within the mucosa, submucosa
and inner part of the circular muscle layer in control
tissue and were rare within the musculature, similar to
previous data.18 Both whole mount preparations and
paraffin sections from patients showed a remarkable
Figure 5 Interstitial cell of Cajal regeneration. An ICC (ICC)shows signs of recovery in its cytoplasm: abundant roughendoplasmic reticulum (ER) and Golgi apparatus (G). Thepresence of caveolae (small arrows) allows for distinguishingthis cell from a fibroblast. The nucleus exhibits a centralnucleolus and chromatin is condensed in the nuclear periph-ery. This ICC establishes gap junctional contact (inset figure)with a degenerating ICC (ICC*), showing decreased electrondensity and vacuolization in its cytoplasm (*) (patient 7). Amast cell (MC) is nearby.
Figure 6 Injury to interstitial cell of Cajal associated withDMP nerve varicosity. An intact ICC-DMP (ICC), locatedbetween the inner (ICM) and the outer circular muscle layer(OCM), establishes a synapse-like junction with a nerve ter-minal in tissue from small bowel macroscopically affectedwith Crohn’s disease (patient 1). Inset figure shows the inti-mate connection between both cells types (white arrows). Aninjured ICC-DMP (ICC*) exhibiting swollen endoplasmicreticulum (ER*) and a lamella body (*) is nearby. SM: sub-mucosa.
Volume 19, Number 5, May 2007 ICC injury in Crohn’s disease
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 357
infiltrate of MC within the muscularis externa (Figs 9
and 10; Table 1). Quantification on paraffin sections
demonstrated that the percentage area occupied by MC
was significantly higher in Crohn’s disease segments
compared with those of controls (0.32 ± 0.04% vs
0.02 ± 0.01%; P < 0.005).
DISCUSSION
Inflammatory bowel diseases such as Crohn’s disease
and ulcerative colitis cause patients to experience
symptoms suggestive of altered gastrointestinal motil-
ity which may be caused by changes in the neural and/
Figure 7 Contact between an ICC and a macrophage. A relatively intact ICC-DMP (ICC) with slight cytoplasmic vacuolization (*)and normally appearing mitochondria, simultaneously contacts smooth muscle cells (SM) (arrow heads) and an active macrophage(Ma) (large arrows) in the deep muscular plexus region in tissue affected with Crohn’s disease in patient 1. Lysosomes (L) and a largenumber of vacuoles can be seen in the cytoplasm of the macrophage. A moderate amount of collagen and elastic fibres is presentaround the cells. Small arrows: caveolae.
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd358
or the myogenic control systems including ICC,
resulting in abnormal propulsion of intraluminal con-
tents. The present study showed patchy ultrastructural
damage to ICC in all ileal segments obtained from
patients with Crohn’s disease. Injury involved all ICC
subtypes but was more severe for the ICC-AP, the
pacemaker cells, and frequently involved organelles
critical for cell survival such as the mitochondria.
Quantification of c-kit immunoreactivity on sections
demonstrated a reduction of ICC in the AP region. In
addition, whole mount preparations revealed a reduc-
tion in the number of cell processes and confirmed a
decreased density of the ICC network. It is important
to note that the relationship between loss of c-kit
immunoreactivity and ICC injury/depletion is incom-
pletely understood and quantification at the electron
microscopy level was not attempted. Intimate contacts
between ICC exhibiting different degrees of cellular
damage were still evident at the ultrastructural level.
This and the patchy nature of the ultrastructural
damage suggest that disruption of the network is
variable and so will be the functional implications.
Persistence of a quantitatively normal ICC population
in a few patients with long duration of the disease
suggests reparative mechanisms able to restore, com-
pletely or partially, ICC networks. Indeed, we observed
signs of ICC recovery at the ultrastructural level,
namely increased rER and Golgi apparatus. This is
consistent with the emerging concept of ICC plasti-
city.19 Differences in the density of c-kit immunore-
activity for ICC between patients could not be
attributed to factors such as age, underlying medical
A B
D
CFigure 8 Damage to enteric nervestructures. Enteric nerves, particularlyin the Auerbach’s plexus (AP), were themost severely affected cells types. (A,B)Low magnification electron microsco-py pictures showing widely spread ul-trastructural degeneration of entericganglia (G) in the myenteric plexus(patient 1). Severe vacuolization,depleted or blebbed varicosities (*) andaxons devoid of most organelles aredisplayed. ICC-AP (ICC) and mast cells(MC) exhibiting piecemeal degranula-tion are seen in the proximity of theenteric ganglia. Occasional intimatecontacts between ICC and the gangliaare observed (large arrows, A). (C,D)High magnification EM pictures des-cribing different degenerative changesof enteric structures. Many lipid drop-lets (L), lamellar bodies (*) and au-tophagosomes (**) are found in thedegenerated varicosities.
Volume 19, Number 5, May 2007 ICC injury in Crohn’s disease
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 359
treatment or clinical presentation. Specimens were
intentionally chosen from those areas exhibiting a high
degree of macroscopic inflammation with similar
inflammatory scores in all patients; thus differences
could not be attributed to variations in the degree of
inflammation. A most striking observation was the
frequent intimate contacts between the numerous MC
and the three ICC subtypes present in the small
intestine. These contacts always involved ICC exhib-
iting a certain degree of cellular injury.
A significant reduction in ICC density has been
previously reported by Porcher et al.5 in ileal
specimens macroscopically affected with Crohn’s dis-
ease. At the ultrastructural level, Rumessen20 reported
injury to the ICC located at the submuscular plexus of
the colon (ICC-SMP) in ulcerative colitis. This ICC
subtype is proposed to function as pacemakers in this
region.21,22 Whilst some of the ultrastructural injury
portrayed by ICC-SMP in ulcerative colitis was similar
to the findings in the present study, more severe cell
damage, such as nuclear injury was absent. This may
explain the integrity of slow wave parameters in
colonic muscle strips obtained from patients with
ulcerative colitis23. Direct contact between ICC-SMP
and inflammatory cells, particularly MC, was not
reported in spite of an increased MC infiltrate reported
in this condition.24 Variations in the inflammatory
mechanisms responsible for ulcerative colitis and
Crohn’s disease might account for different ICC sub-
types being affected. ICC-AP exhibited the greatest
injury in our study as did the neural structures of the
AP. This could be explained by their proximity to blood
vessels, more abundant in this region, leading to
preferential encounters with inflammatory cells migra-
ting from the blood stream.25 The Trichinella spiralis
murine model of gut inflammation is also associated
with severe ultrastructural damage to ICC-AP. In this
model, ICC-AP exhibiting severe structural abnormal-
ities were frequently found in close proximity to
macrophages and to dilated blood vessels, which were
mainly confined to the myenteric plexus region.26,27
This model also provided information regarding the
relation between ultrastructural injury to ICC net-
works on the one side and abnormal slow wave
electrical activity and mechanical contractions on the
other.26,28,29 It is currently unknown the impact of
ICC-AP damage on the small intestine pacemaker
activity in Crohn’s disease. However, evidence of
sustained activity may be deduced from the observa-
tion of normal spontaneous contractile activity in
intestinal circular muscle strips.2
ICC-IM also portrayed signs of injury at the ultra-
structural level, although not to the extent as that of
ICC-AP. Furthermore, c-kit quantification at the light
microscopy level showed no differences between
patients and controls for both the circular and longi-
tudinal muscle layers. ICC-IM in the small bowel do
not establish gap-junctions with smooth muscle cells
or synapse-like contacts with nerve terminals unlike in
the oesophagus, the fundus, or the colon.30–32 There-
fore, the proposed role as mediators of neurotransmis-
sion in the aforementioned locations may be different
in the human small intestine31,33. Ultrastructural
damage was also observed in ICC-DMP, a network of
ICC associated with the non-ganglionated deep mus-
cular plexus. We have previously shown in a morpho-
A B C
Figure 9 C-kit immunoreactivity in wholemount preparations of ileum assessed in control specimen (A) and specimen macro-scopically affected with Crohn’s disease (B,C). (A) Kit positive ICC are abundant and most exhibit a bipolar shape (arrows). Theycontact each other to form networks within the musculature. A moderate (B) and strong (C) reduction of kit positive ICC (arrows),consequently, the ICC network decreases in density. Abundant kit positive, brightly stained and round-shaped mast cells(arrowheads) are numerous in the inflamed tissues (patient 11).
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd360
logical study on human tissue that ICC-DMP may play
a role in cholinergic and nitrergic muscle innervation
as they form synapse-like contacts with both types of
nerve terminals and gap junction type contacts with
smooth muscle cells of the circular muscle layer8
consistent with data from an animal model.34 Further-
more, a role as stretch receptors has been also
proposed.35 This is supported by an observed correla-
tion between ICC-DMP injury and absence of disten-
tion-induced patterns of electrical activity, normally
associated with peristalsis, during T. spiralis infec-
tion.36
We observed a remarkable increase in MC popula-
tion within the ileal wall of Crohn’s disease specimens,
consistent with previous reports.10,37 Some cells exhib-
ited characteristic morphological changes in their
cytoplasmic vesicles defined as piecemeal-type degran-
ulation, thought to be associated with long-lasting
release of vesicle content38 and it is known that MC
mediators are released in IBD39. Mast cells can influ-
ence and be influenced by other cell types. Examples of
this are their relationship with fibroblasts, nerve
varicosities and inflammatory cells such as eosinoph-
ils.12,40,41 The present findings and our previous report
on achalasia11 indicate that ICC should be added to
this list. Mast cell granules can be found in neighbour-
ing cells, acquired by fusion of granule and cellular
membranes or by cellular capture of MC granule
A B
DC
Figure 10 C-kit immunoreactivity (A,C) and mast cell toluidine blue staining (B,D) from consecutive paraffin sections from controlileum specimen (A,B) and those macroscopically affected with Crohn’s disease (C,D). The method of quantification of c-kitimmunoreactivity used consecutive slides subtracting values obtained by toluidine blue mast-cell quantification (B,D) from c-kitstaining (A,C). (A) Distribution of Kit positive ICC was evident around the Auerbach’s plexus (AP) and scattered throughout thecircular (CM) and longitudinal muscle layers (LM). Mast cells were also identified by their rounder shape but they were rare(arrows). (B) A mast cell is observed in the same location as in (A). (C) C-kit immunoreactivity for ICC is reduced around the APregion in patient 2. A remarkable increase in mast cell infiltrate is observed. (D) Mast cell infiltrate is evident in the same locationas in (C) when assessed with toluidine blue staining.
Volume 19, Number 5, May 2007 ICC injury in Crohn’s disease
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 361
remnants (transgranulation), as seen between MC and
ICC in achalasia11 and MC and neurons in the dove
brain42, among others.43 There seems to be a specific
cellular communication between MC and ICC, rather
than random encounters. The exact nature of this
communication and the type of mediators involved are
currently unknown. Mast cells can produce membrane
bound stem cell factor (SCF) which is critical for the
survival of both MC and ICC through binding to the
c-kit receptor.44–46 Enteric nerves are also a well-
known source of SCF47,48 and their intimate associ-
ation with ICC could be important for an effective
interaction between the SCF and the ICC c-kit recep-
tor. Widespread injury to nerve structures in Crohn’s
disease could jeopardize access of ICC to this molecule
and thus influence ICC survival. Mast cell derived SCF
could be important in this context. However, various
evidences are against a crucial role for nerve-derived-
SCF on ICC viability: (i) other neighbouring cell types,
including smooth muscle cells, are able to synthesize
SCF; (ii) ICC can develop morphologically in the
absence of the enteric nervous system: GDNF)/) mice
which lack enteric neurons throughout most of the gut
develop a normal ICC population49 and explants of
chicken or mouse intestine removed from embryos
before arrival of the neural crest cells show normal
development of ICC population.50 On the other hand,
MC-derived nerve growth factor and neurotrophins can
promote cultured chicken embryonic neural crest cells
survival.51 Similarly, MC interaction with ICC could
involve cytokines potentially capable of influencing
ICC viability. This hypothesis is supported by our
recent finding that IL-9 released by MC has a prolifer-
ative effect on ICC in culture.52 Conversely, MC are a
well-known source of tryptase, released in the setting
of inflammation. The pro-inflammatory action of
tryptase, ultimately leading to increased intestinal
permeability and tissue damage53 could potentially
participate in ICC injury. Thus, the degree of protec-
tion and/or damage caused by the interaction between
MC and the various cells of the intestinal musculature,
including ICC, through piecemeal degranulation,
transgranulation or other forms of secretion remains
to be elucidated.
In conclusion, the present study shows evidence for a
decrease in the ICC-AP population, possibly a direct
effect of inflammation or indirectly as a consequence of
lack of nerve derived SCF. Ultrastructural injury to all
A
B
Figure 11 C-kit immunoreactivity in terminal ileum frompatients with long-standing Crohn’s disease show abun-dant c-kit staining for both ICC and mast cells (arrows) atthe Auerbach’s plexus region (AP) in patients 3 (A) and5 (B).
Figure 12 Positive correlation between c-kit staining for ICCand duration of Crohn’s disease. Adjusted mean of the c-kitimmunoreactivity for ICC at the muscularis externa level foreach patient was determined by subtracting the values ob-tained with the toluidine blue quantification for mast cellsfrom the c-kit quantification as the latter identifies both ICCand mast cells. Duration of the disease is expressed in years.Subjects with long-standing disease show normal c-kit stain-ing for ICC whilst this appears reduced in those with shorterduration of the condition (Pearson correlation coefficient 0.9;P ¼ 0.001).
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd362
ICC-subtypes in the ileum of Crohn’s disease patients
was evident. Mast cells were abundant, often exhibited
signs of piecemeal degranulation and established inti-
mate contacts with injured ICC. Mast cells may
promote ICC survival in Crohn’s disease because of
their capacity to secrete SCF and IL-9, which are growth
factors for ICC. The evidence of ICC recovery and the
presence of a normal ICC density in some patients with
long duration of the disease are consistent with ICC
repair taking place. Further studies are needed to provide
more evidence for or against this hypothesis.
ACKNOWLEDGMENT
This work was supported by Canadian Institutes of
Health Research (CIHR) operating grants. Dr Zarate
received a scholarship supported by Novartis, the
CIHR and the Canadian Association of Gastroenterol-
ogy and is a PhD candidate at the Autonomous
University of Barcelona. Initial ultrastructural studies
were carried out by Dr I. Berezin. The authors are
grateful to Dr J. Bienenstock for invaluable comments.
REFERENCES
1 Annese V, Bassotti G, Napolitano G et al. Gastrointestinalmotility disorders in patients with inactive Crohn’s dis-ease. Scand J Gastroenterol 1997; 32: 1107–17.
2 Vermillion DL, Huizinga JD, Riddell RH et al. Alteredsmall intestinal smooth muscle function in Crohn’s dis-ease. Gastroenterology 1993; 104: 1692–9.
3 Dvorak AM, Connell AB, Dickersin GR. Crohn’s disease: ascanning electron microscopic study. Hum Pathol 1979;10: 165–77.
4 Dvorak AM, Silen W. Differentiation between Crohn’sdisease and other inflammatory conditions by electronmicroscopy. Ann Surg 1985; 201: 53–63.
5 Porcher C, Baldo M, Henry M et al. Deficiency of inter-stitial cells of Cajal in the small intestine of patients withCrohn’s disease. Am J Gastroenterol 2002; 97: 118–25.
6 Reddy SN, Bazzocchi G, Chan S et al. Colonic motilityand transit in health and ulcerative colitis. Gastroenter-ology 1991; 101: 1289–97.
7 Bassotti G, Villanacci V, Mazzocchi A et al. Colonic pro-pulsive and postprandial motor activity in patients withulcerative colitis in remission. Eur J Gastroenterol Hepa-tol 2006; 18: 507–10.
8 Wang XY, Paterson C, Huizinga JD. Cholinergic and ni-trergic innervation of ICC-DMP and ICC-IM in the humansmall intestine. Neurogastroenterol Motil 2003; 15: 531–43.
9 Wang XY, Sanders KM, Ward SM. Intimate relationshipbetween interstitial cells of Cajal and enteric nerves in theguinea-pig small intestine. Cell Tissue Res 1999; 295: 247–56.
10 Gelbmann CM, Mestermann S, Gross V et al. Strictures inCrohn’s disease are characterised by an accumulation ofmast cells colocalised with laminin but not with fibro-nectin or vitronectin. Gut 1999; 45: 210–7.
11 Zarate N, Wang XY, Tougas G et al. Interstitial cells ofCajal, associated with mast cells, survive nitrergicnerves in achalasia. Neurogastroenterol Motil 2006; 18:556–68.
12 Levi-Schaffer F, Rubinchik E. Mast cell/fibroblast inter-actions. Clin Exp Allergy 1994; 24: 1016–21.
13 Regan MC, Flavin BM, Fitzpatrick JM et al. Stricture for-mation in Crohn’s disease: the role of intestinal fibro-blasts. Ann Surg 2000; 231: 46–50.
14 Lehr HA, Mankoff DA, Corwin D et al. Application ofphotoshop-based image analysis to quantification of hor-mone receptor expression in breast cancer. J HistochemCytochem 1997; 45: 1559–65.
15 Saverymuttu SH, Camilleri M, Rees H et al. Indium 111-granulocyte scanning in the assessment of disease extentand disease activity in inflammatory bowel disease. Acomparison with colonoscopy, histology, and fecal indium111-granulocyte excretion. Gastroenterology 1986; 90(Pt1): 1121–8.
16 Rumessen JJ, Thuneberg L. Interstitial cells of Cajal inhuman small intestine. Ultrastructural identification andorganization between the main smooth muscle layers.Gastroenterology 1991; 100 (Pt 1): 1417–31.
17 Faussone-Pellegrini MS, Thuneberg L. Guide to the iden-tification of interstitial cells of Cajal. Microsc Res Tech
1999; 47: 248–66.18 Romert P, Mikkelsen HB. C-kit immunoreactive intersti-
tial cells of Cajal in the human small and large intestine.Histochem Cell Biol 1998; 109: 195–202.
19 Faussone-Pellegrini MS, Vannucchi MG, Ledder O et al.Plasticity of interstitial cells of Cajal: a study of mousecolon. Cell Tissue Res 2006; 325: 211–7.
20 Rumessen JJ. Ultrastructure of interstitial cells of Cajal atthe colonic submuscular border in patients with ulcerativecolitis. Gastroenterology 1996; 111: 1447–55.
21 Serio R, Barajas-Lopez C, Daniel EE et al. Slow-waveactivity in colon: role of network of submucosal intersti-tial cells of Cajal. Am J Physiol 1991; 260 (Pt 1): G636–45.
22 Liu LW, Thuneberg L, Huizinga JD. Selective lesioning ofinterstitial cells of Cajal by methylene blue and light leadsto loss of slow waves. Am J Physiol 1994; 266 (Pt 1): G485–96.
23 Lu G, Vanderwinden JM, Rumessen JJ et al. Electrophys-iological, morphological and ultrastructural changes inulcerative colitis (UC), and idiopathic constipation (IC).Gastroenterology 1998; 114: A796.
24 Stoyanova II, Gulubova MV. Mast cells and inflammatorymediators in chronic ulcerative colitis. Acta Histochem
2002; 104: 185–92.25 Ozaki H, Kinoshita K, Hori M. Mechanism of the intes-
tinal dysmotility in the inflammatory bowel diseases:possible involvement of muscularis resident macrophages.Nippon Yakurigaku Zasshi 2003; 122 (Suppl.): 46P–50P.
26 Der T, Bercik P, Donnelly G et al. Interstitial cells of cajaland inflammation-induced motor dysfunction in the mousesmall intestine. Gastroenterology 2000; 119: 1590–9.
27 Wang XY, Berezin I, Mikkelsen HB et al. Pathology ofinterstitial cells of Cajal in relation to inflammation re-vealed by ultrastructure but not immunohistochemistry.Am J Pathol 2002; 160: 1529–40.
28 Huizinga JD, Thuneberg L, Kluppel M et al. W/kit generequired for interstitial cells of Cajal and for intestinalpacemaker activity. Nature 1995; 373: 347–9.
Volume 19, Number 5, May 2007 ICC injury in Crohn’s disease
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd 363
29 Ward SM, Burns AJ, Torihashi S et al. Mutation of theproto-oncogene c-kit blocks development of interstitialcells and electrical rhythmicity in murine intestine. J
Physiol 1994; 480: 91–7.30 Faussone-Pellegrini MS, Cortesini C. Ultrastructural fea-
tures and localization of the interstitial cells of Cajal inthe smooth muscle coat of human esophagus. J Submic-
rosc Cytol 1985; 17: 187–97.31 Burns AJ, Lomax AE, Torihashi S et al. Interstitial cells of
Cajal mediate inhibitory neurotransmission in the stom-ach. Proc Natl Acad Sci U S A 1996; 93: 12008–13.
32 Wang XY, Sanders KM, Ward SM. Relationship betweeninterstitial cells of Cajal and enteric motor neurons in themurine proximal colon. Cell Tissue Res 2000; 302: 331–42.
33 Ward SM, Morris G, Reese L et al. Interstitial cells of Cajalmediate enteric inhibitory neurotransmission in the loweresophageal and pyloric sphincters. Gastroenterology 1998;115: 314–29.
34 Iino S, Ward SM, Sanders KM. Interstitial cells of Cajal arefunctionally innervated by excitatory motor neurones inthe murine intestine. J Physiol 2004; 556 (Pt 2): 521–30.
35 Thuneberg L, Peters S. Toward a concept of stretch-couplingin smooth muscle. I. Anatomy of intestinal segmentationand sleeve contractions. Anat Rec 2001; 262: 110–24.
36 Wang XY, Vannucchi MG, Nieuwmeyer F et al. Changesin interstitial cells of Cajal at the deep muscular plexus areassociated with loss of distention-induced burst-typemuscle activity in mice infected by Trichinella spiralis.Am J Pathol 2005; 167: 437–53.
37 Dvorak AM, Monahan RA, Osage JE et al. Crohn’s disease:transmission electron microscopic studies. II. Immuno-logic inflammatory response. Alterations of mast cells,basophils, eosinophils, and the microvasculature. Hum
Pathol 1980; 11: 606–19.38 Dvorak AM, Tepper RI, Weller PF et al. Piecemeal
degranulation of mast cells in the inflammatory eyelidlesions of interleukin-4 transgenic mice. Evidence of mastcell histamine release in vivo by diamine oxidase-goldenzyme-affinity ultrastructural cytochemistry. Blood
1994; 83: 3600–12.39 Raithel M, Winterkamp S, Pacurar A et al. Release of mast
cell tryptase from human colorectal mucosa in inflamma-tory bowel disease. Scand J Gastroenterol 2001; 36: 174–9.
40 Mori N, Suzuki R, Furuno T et al. Nerve-mast cell (RBL)interaction: RBL membrane ruffling occurs at the contactsite with an activated neurite. Am J Physiol Cell Physiol
2002; 283: 1738–44.41 Levi-Schaffer F. Cross-talk between mast cells and eosin-
ophils. Allergy 1999; 54 (Suppl. 58): 36–8.42 Wilhelm M, Silver R, Silverman AJ. Central nervous sys-
tem neurons acquire mast cell products via transgranula-tion. Eur J Neurosci 2005; 22: 2238–48.
43 Dines KC, Powell HC. Mast cell interactions with thenervous system: relationship to mechanisms of disease. J
Neuropathol Exp Neurol 1997; 56: 627–40.44 de Paulis A, Minopoli G, Arbustini E et al. Stem cell factor
is localized in, released from, and cleaved by human mastcells. J Immunol 1999; 163: 2799–808.
45 Zhang S, Anderson DF, Bradding P et al. Human mastcells express stem cell factor. J Pathol 1998; 186: 59–66.
46 Welker P, Grabbe J, Gibbs B et al. Human mast cells pro-duce and differentially express both soluble and membrane-bound stem cell factor. Scand J Immunol 1999; 49: 495–500.
47 Young HM, Torihashi S, Ciampoli D et al. Identificationof neurons that express stem cell factor in the mouse smallintestine. Gastroenterology 1998; 115: 898–908.
48 Torihashi S, Yoshida H, Nishikawa S et al. Enteric neu-rons express steel factor-lacZ transgene in the murinegastrointestinal tract. Brain Res 1996; 738: 323–8.
49 Ward SM, Ordog T, Bayguinov JR et al. Development ofinterstitial cells of Cajal and pacemaking in mice lackingenteric nerves. Gastroenterology 1999; 117: 584–94.
50 Lecoin L, Gabella G, Le Douarin N. Origin of the c-kit-positive interstital cells in the avian bowel. Development
1996; 122: 725–33.51 Skaper SD, Pollock M, Facci L. Mast cells differentially
express and release active high molecular weight neurot-rophins. Brain Res Mol Brain Res 2001; 97: 177–85.
52 Ye J, Zhu Y, Khan WI et al. IL-9 enhances growth of ICC,maintains network structure and strengthens rhythmicityof contraction in culture. J Cell Mol Med 2006; 1016: 687–94.
53 Cenac N, Coelho AM, Nguyen C et al. Induction of intes-tinal inflammation in mouse by activation of proteinase-activated receptor-2. Am J Pathol 2002; 161: 1903–15.
X.-Y. Wang et al. Neurogastroenterology and Motility
� 2007 The AuthorsJournal compilation � 2007 Blackwell Publishing Ltd364