- Email: [email protected]
HeJ mice
GASTROENTEROLOGY 1994$07:1726-1735
Spontaneous, Heritable Colitis in a New Substrain of C3H/HeJ Mice JOHN
P. SUNDBERG,*
and EDWARD
CHARLES
0.
ELSON,’
HENDRICK
BEDIGIAN,*
H. BIRKENMEIER*
*The Jackson Laboratory, Bar Harbor, Maine; and tDivision of Gastroenterology,
University of Alabama, Birmingham, Alabama
Background/Aims: C3H/HeJ mice at the Jackson Labo-
ing, lack of immunologic
ratory have periodically been culled because of the occurrence of soft feces, perianal ulceration, and rightsided colitis. No pathogens have been isolated. The goal of the current study was to establish a substrain with a high Incidence of this disease. Methods: Affected male and female C3H/HeJ mice were bred. The clinical, pathological, microbiological, and genetic features of 216 mice of the resulting pedigree were characterized. Results: A severely affected female crossed with a normal male resulted in a new substrain, de noted C3H/HeJBir, with a high incidence of right-sided colitis. Histologically, lesions occurred primarily in the cecum and proximal colon, characterized by acute and chronic inflammation, crypt abscesses, ulcerations, regenerative hyperplasia, and submucosal scarring. Such colitis peaked at 3-6 weeks; however, similar disease was found sporadically in animals more than 1 year of age. Small lesions at the anorectal junction were common throughout life. An extensive search for patho gens was negative. Genetic analysis of CBH/HeJBir mice suggested that the disease was inherited as a quantitative trait. Conclusions: CSH/HeJBir mice develop a spontaneous, heritable form of idiopathic inflammatory bowel disease and will be a valuable resource for genetic and immunologic studies of this disease.
have limited
hronic
inflammatory
plex disorders
bowel diseases (IBD) are com-
whose pathogeneses
are poorly under-
stood. Colitis can be induced in experimental animals to study various aspects of IBD.’ These models have increased our understanding of nonspecific inflammatory mechanisms. However, most of these induced models do not allow study of the complex interactions between genetic, environmental, and immune components important in disease initiation and progression. The model with a high degree of homology to many forms of human IBD is spontaneous colitis that develops in cotton-topped tamarins.‘-3 Unfortunately, the endangered species status of these primates, difficulty in handling, expense of hous-
and other problems
Models of disease in genetically bred laboratory
defined
mice circumvent
investigation
of chronic
diseases, including particularly
the identification Thus,
of genes important
a mouse model
useful model system to investigate mechanisms
secondary
perianal
of IBD,
would
the primary
be a
immuno-
that can lead to IBD.
Because of the occurrence and/or
or autoimmune
one that occurs spontaneously,
logic and genetic
of in-
animals. Mice and genetic
inflammatory
in disease susceptibility.*
strains
many of the problems
inherent in models using larger, outbred are the premier species for immunologic
of soft, light-colored
ulceration,
feces
mice of the C3H/
HeJ strain in the research and production colonies of the Jackson Laboratory have been culled in disproportionate numbers
compared
with
fected mice had histological pathogens
were never
Because it seemed netic predisposition
other strains.
At necropsy,
evidence
of colitis, although
identified
with
routine
likely that C3H/HeJ to develop
af-
testing.
mice had a ge-
a form of IBD, we selec-
tively bred affected C3H/HeJ mice with perianal ulcerations to generate a pedigreed “high-susceptibility” substrain, C3H/HeJBir, with a high incidence of spontaneous colitis. The purpose of this study was to describe the natural history, clinical and pathological features, and probable
C
reagents,
their usefulness.
mode of inheritance
in C3H/HeJBir
of spontaneous
colitis
mice.
Materlab and Methods Materials
and Animals
The studies were approved by the Animal Care and Use Committee of the Jackson Laboratory. C3H/HeJ mice with P&anal ulcerations were initially obtained from the Jackson Laboratory’s production colonies in Morrel Park (Bar Harbor, ME). These mice and their offspring were selectively bred to Abbreviations used in this paper: IL, interleukln; charide. 0 1994 by the American Gastroenterological 00165085/94/$3.00
LPS, lipopolysacAssociation
IBD IN CM/HeJBir
December 1994
derive a new substrain, of perianal
To determine present
C3H/HeJBir,
ulceration,
as described
the onset, severity,
in the gastrointestinal
longitudinal between
and type of inflammation
tract
of C3HIHeJBir
study of disease progression
mice, a
was instituted.
the ages of 10 days and 90 weeks underwent
at weekly intervals.
The mice were later divided
groups
The first group
for analysis.
to 3 months second group months
consisted
of mice from 0
of mice from 4 to 12 months
88 total); the
of age (37 males
65 total); and the third group
of age (21 males and 42 females,
Mice
necropsy
into three age
of age (56 males and 32 females,
and 28 females,
earlier findings.
with a high frequency in Results.
of mice >12
63 total). A total of
Serum was collected, immediately
large intestine, all animals.
and mice were killed by CO* asphyxia-
before cecum,
necropsy.
The stomach,
and perianal
Intestinal
small
skin were collected
rolls were prepared
by placing
and from
them on
an index card, twisting
them
in concentric
centrifugal
around
pick
in the plane
of the card.’ The
a central
unopened using
tooth
intestinal
rolls were gently
a 27-gauge
further
needle
solution stained
processed
circles
with
fixative
were tabulated Redmond, erpoint;
to 70% ethanol.
routinely,
sectioned
Microsoft
a spreadsheet
subacute,
(Excel; Microsoft
using a graphics
or chronic)
moderate, using
program
Corp., (Pow-
based on loca-
or severe), and quality commonly
accepted
(acute,
histological
Briefly, mild lesions were small, focal, or widely sepa-
rated multifocal
areas of inflammation
to the lamina propria. or represented
Moderate
or fibrosis extending
and/or fibrosis limited
areas of inflammation
and/
into the submucosa.
Severe lesions were
ulcers that covered large areas (a millimeter
or more of mucosa)
of cecum or colon. The ulcers were covered neutrophils
fibrosis and loss of crypts.
crypt necrosis and/or ulceration phils
infiltrating
had a mixture cells with
by an exudate
of
and necrotic debris, or the areas were epithelialized
with underlying
and/or
minimal
Acute lesions had
with a predominance
covering
of neutrophils
the ulcers.
and mononuclear
or no fibrosis.
defined beds of granulation
were stained
Chronic
were removed
examination.
All sections
the proximal
for gram-negative
and Campylobacter spp. (Campy Systems,
Microbiology
GDVII,
colons and
aerobes (GN broth)
Pouch; Becton Dickinson
Cockeyville,
Systems).
immunosorbent pneumonia
of males and fe-
The cecums
stains. In addition,
were cultured
crobiology
Ten C3Hl
with H&E, Brown and Brenn (tissue gram), and
Warthin-Starry cecums
Laboratory. substrain,
Equal numbers
for histopathologic
MD).’ Specimens
Serum
Mi-
were also
(CDT; Becton Dickinson
was tested
by enzyme-linked
following
mouse
assay for the
pathogens:
virus of mice, Sendai virus, mouse hepatitis
minute
virus of mice, ectromelia,
reo 3, epizootic
meningitis,
cytomegalovirus,
mouse
Mycoplasma pulnzonis,
diarrhea
adenovirus
virus,
lymphocytic
of infant (MAD),
chorio-
mice,
mouse
polyoma
virus,
and Mycoplasnzu arthvitidis. a
Genetic Quality Control and Genetic Studies
of neutro-
Subacute
At the time affected mice were initially the production
colony,
for 204-208
been inbred type integrity
used to compare
areas
inflammatory
lesions had well-
C3HIHeJ
generations.
To compare
markers
methods
Defective This mutation, immune
to bacterial C3H/HeJ
designated
genetic
and sub-
quality control
assay for H-2.“,“’ is a fea-
mice from other C3H strains.
Lpsd, has a pronounced
effect on the
for their lipopolysaccha-
by an in vitro splenocyte
of IBD in C3H/HeJBir
affected C3H/HeJBir
C3Hi
control were
lipopolysaccharide
Mice were tested
ride (LPS) responsiveness crossing
standard
and a microcytotoxic
response
system.”
Inheritance
quality
had
the geno-
with the progenitor
for genetic
from
strain
female mice from both the primary
strain colonies. Tests included isoenzyme
obtained
the progenitor
of the new substrain
HeJ colony, standard
ture that distinguishes
lesions were either multifocal
by locally extensive
were also evaluated.
and processed
and
and the results
Corp.). Lesions were graded
(mild,
HeJBir,
evidence
were obtained
of the Jackson
mice from the newly established
males were used in both groups.
Tissues
at 6 pm,
by light microscopy,
WA) and graphed
tion, severity criteria.
using
colonies
additional
mice without
examination
1727
The rolls were
with H&E.
Samples were evaluated
from the production
in Fekete’s acid-alcohol-formalin
for 12 hours, then transferred
were trimmed,
inflated
and 5-mL syringe.
fixed by immersion
C3H/HeJ
of physical
tested for Clostridiwn di#icile antigens
114 male and 102 female mice were studied. tion
Ten normal
of IBD at the time
MICE
assay.”
mice was evaluated
by
mice with inbred Mus castanets
(CAST/Ei) and examining offspring in the F1 and F2 generations for evidence of disease.”
ReSUttlB
Colony Establishment
tissue under ulcers or reepithelial-
ized regions of cecum. Inflammatory
cells were predominantly
lymphocytes. For purposes of comparison to human anatomy and therefore human disease, the term right colon hereafter includes the cecum and colon up to the hepatic
flexure in the mouse.6
Microbiology Although no mouse pathogens were identified during prior diagnostic tests of affected C3H/HeJ mice, the newly established substrain was also tested for pathogens to verify
Over a period of 1 year, a pedigreed
substrain
was
established starting initially with one severely affected female and a clinically normal male. Because we did not know the etiology of bowel disease in mice originating from production colonies, each affected mouse was treated as an individual case, and sick mice were therefore not housed together. Once a pedigreed colony was established, affected individual mice were bred to each other. We initially used several sporadic severe cases to start multiple pedigrees. The pedigree with the highest inci-
1728
SUNDBERG ET AL.
GASTROENTEROLOGY Vol. 107, No. 6
developed
rua * AA AuAAUAA
clinical
of perianal
disease as determined
by the presence
ulcers and soft, yellow feces.
ClinicopathologicCorrelationWith Natural History of Disease A longitudinal
study determined
the age of onset,
duration, and pathological progression of right-sided colitis in more than 200 C3H/HeJBir mice between 2 and
90 weeks of age. The microscopic compared
**
*** **
with the clinical
of death. Clinically,
(*I.
in young mice between
were also positive,
including ulceration
was confined
to a relatively
short period
was selected for further generations
study. Offspring
were evaluated
ence of perianal
clinically
ulcerations,
CA; Table
for IBD by the pres-
soft feces, and positive
occult test results (SmithKline HeJBir
in all successive
Diagnostics
1). This new substrain
and is now maintained
Inc., San Jose,
was designated by strict
HemC3H/
brother-sister
matings (currently Fi3) with at least one parent positive for disease. The initial pedigree of the substrain is shown in Figure 1.
or histological
with age. Results
the number
evidence
are summarized
in Figure
2.
indicator of disease
mice. There was a correlation
positive Hemoccult inflammation
of mice with
of disease decreased
Occult blood was the most predictive in C3H/HeJBir
after weaning
weeks old. Once healed, these
ulcers did not recur. Overall, 1)
primarily
some at 1 year of age. The presence of perianal
out clinical (Figure
was then at the time
4 and 17 weeks of age. However,
males
when the mice were 4-7
dence of disease in the F2 and Fs generations
present
occult blood was detected
a few of the older
figure 1. Pedigree of CBH/HeJBir mice. Incidence of IBD is indicated by perianal ulceration (W, male; ?? , female) and soft, yellow feces
pathology
findings
between
test result and histological
a
evidence of
in the colon. Mice with trace amounts of occult
blood in the feces usually had small ulcers or erosions evident at the mucocutaneous amination
anorectal junction
on histological
ex-
of the tissues (see below).
Gross evidence ence of moist
of severely affected mice was the pres-
areas of perianal
The ulcers were consistently
ulceration
(Figure
found in animals
3A).
with soft,
Clinical Features at 3-5 Weeks of Age
yellow feces. Microscopically,
During
small to extensive areas of ulceration of the squamous epithelium of the haired skin adjacent to or extending
the 2 years of this study, the QH/HeJBir
colony generated 678 mice. Each mouse was examined once between 3 and 5 weeks of age for clinical evidence of IBD. The clinical
data on the 326 female
male mice are summarized
in Table
and 352
1. Approximately
these lesions
from the anus. The underlying with
granulation
collected
tissue
for evaluation
dermis
was usually
filled
the tissues
were
(Figure
nation
to be positive
blood when tested during
this
developed perianal ulcerations. In 50% of the animals, the only clinical sign of disease was trace amounts of occult blood in the feces. Other animals had various combinations of symptoms including occult blood with perianal ulceration, occult blood with soft stools, and occult blood with both soft stools and perianal ulceration. Within the substrain colony, 21% of the females and 12 % of the males appeared normal and never developed clinical signs of IBD. Both serological testing and routine fecal cultures did not identify any pathogens in the C3H/ HeJBir mice. Less than 1% of C3H/HeJ mice in the production colony from which this substrain was derived
Table
of the gastrointestinal
3B).Removal and exami-
the anus showed that changes were limited
for occult
1.
Frequency
tract from the stomach
of Clinical
C3H/HeJBir
Disease
Blood only Perianal ulceration only Soft feces only Blood and ulceration Blood and soft feces Ulceration and soft feces Ulceration, soft feces, and blood None
to
to the cecum
in the Colony
of
Mice Female
Clinical signs
of
by the time
79% of the female and 88% of the male mice were found time. Approximately 25% of the mice of both sexes has soft stools, and 8% of the females and 27% of the males
consisted
Male
Total of mice (56)
No.
%
No.
%
50 0 0 9 16 0
171 0 0 14 61 0
52 0 0 4 19 0
170 0 0 44 47 0
48 0 0 13 13 0
9 16
12 68
4 21
50 41
14 12
IBD IN CBH/HeJBir
December1994
lOS%
MICE
1729
A
80%
8@= $ ke 40%
20%
0%
1 o-3
o-3
13 - 18
4-12
4.12
13-18
Age in month intmwls
Age in month intavals
Figure 2. Summary of clinical and pathological data. (A) Percentages of CBH/HeJBir mice in the longitudinal study with clinically (M, positive occult blood, perianal ulcer, soft stool) or histologically confirmed (0, ulceration, inflammation, edema, or fibrosis) disease. (6) Percentages of CBH/HeJBir mice with various histological grades of severity of cecal disease in the three age groups.
and colon. Normally, has a prominent moist material.
the lower intestinal
cecum that usually
tract of a mouse
contains
dark green
As the feces enter the colon, they become
dark green to black, oval, and hard (Figure
4A). In con-
trast, severely affected C3H/HeJBir mice had small, shrunken, yellow cecums that often contained little or no ingesta (Figure 4B). The colons were slightly distended with mustard yellow soft feces that formed soft oval pellets. This soft fecal material became matted in the perianal
ure 5C). In focal areas, where crypts the lamina propria was highly cellular, necrosis
5D). Adjacent mitotic
crypts activity.
were hyperchroAnimals
4 and 5 weeks of age had small focal ulcers. maining
lamina
mixed,
propria
predominantly
was thickened mononuclear
necrosis lamina
between The re-
and contained cell infiltrate
ure 6A and B). In these areas, the epithelium going the
hair.
(Figure
matic with increased
were short and there was crypt
a
(Fig-
was under-
of individual
cells with separation from 6B). Once the mucosa (Figure
propria
Microscopically, lesions were primarily limited to the cecum and right colon. Lesions in the cecum progressed
sloughed,
through a series of stages. The earliest age at which mice underwent necropsy was 10 days. All suckling
uded into the lumen (Figure 6C). Vessels in the muscular walls were markedly dilated (Figure 60). By 6-7
pups examined
weeks of age, the ulcers
at this age had normal
gastrointestinal
tracts. However, during the suckling to weaning transition period beginning at 14 days of age, all mice examined had marked edema of the submucosa and, to a lesser extent, the lamina propria (Figure 5A). Although the lining epithelium of the cecum was intact, there were large numbers of neutrophils in the lamina propria and scattered neutrophils in the edematous submucosa (Figure 5B). The cecum was not uniformly involved, with the most severe lesions observed at the entry of the ileum and directly across from it. A patchy, multifocal pattern of inflammation involved other parts of the cecum. Between 3 and 4 weeks of age, the inflammation became subacute, consisting of a mixture macrophages, lymphocytes, and neutrophils in the lamina propria. The submucosa remained edematous with a loose, diffuse infiltrate similar to that of the lamina propria (Fig-
numbers
the submucosa of neutrophils
inflammatory
cell
became
edematous
in the lamina
and large
propria
were reepithelialized
infiltrate
had
become
were ex-
and the
subacute
to
chronic. Margination of lymphocytes within small veins in the submucosa was prominent (Figure 6D). By 10 weeks of age, the ulcers had healed, leaving areas without crypts surrounding small islands of mucosa with regenerative erative
crypt
crypts (Figure epithelium
6E). The surrounding apparently
replaced
regenthe de-
pressed areas lacking crypts, because after 10 weeks of age, no evidence of ulceration was found. Ulcers or chronic scarring with small ulcerations were occasionally found in the cecums of older mice. The latter were located above gut-associated lymphoid tissue. Microscopic lesions of the anorectal mucosa were common and consisted of small erosions or ulcerations at the mucocutaneous junction, with a mild to moderate exudation of neutrophils into the lumen. Severe cases of
1730
SUNDBERG
ET AL.
GASTROENTEROLOGY
Vol. 107,
No. 6
Flgure 3. Perianal ulceration. (A) Perianal ulceration extending over the scrotum and ventral tail head in a 5-week-old male C3H/HeJ mouse with severe cecal lesions. (B) Desiccated proteinaceous material and necrotic debris cover a large perianal ulcer and a bed of granulation tissue. Adjacent epidermis is thickened (H&E; original magnification 25x).
Figure 4. Gross pathology of mouse cecum and colon. (A) Normal distended cecum (so/id arrow), entrance of the ileum (open arrow), and colon with three dark fecal pellets in an unaffected C3H/HeJ mouse. (B) The cecum is small, pale, and difficult to identify (so/id arrow) next to the entrance of the ileum (open arrow) in an affected C3H/HeJBir mouse. The colon contains poorly formed, lightcolored fecal pellets.
December 1994
IBD IN C3H/HeJBir
MICE
1731
figure S. Histopathology of cecum and colon of CBH/HeJBir mice at 2-3 weeks of age. (A) Cecum of a 2-week-old male mouse has marked edema of the submucosa. (B) Higher magnification of boxed area in A showing the locally extensive infiltration of neutrophils into the lamina propria. (C) Cecum of a 3week-old male mouse that has a large area of early crypt loss, mixed inflammatory cell infiltrate in the lamina propria, and mild edema of the submucosa. (/I) Higher magnification of boxed area in C. Early crypt loss is associated with necrotic debris in crypts (arrow) adjacent to crypts undergoing necrosis (H&E; original magnifications: A, 25x; B, 100x; C, 100x; D, 400x).
1732
SUNDBERG
ET AL.
GASTROENTEROLOGY
Vol. 107,
No. 6
Flgure 6. Histopathology of cecum and colon of CBH/HeJBir mice at 4-10 weeks of age. (A) Cecum of a a-week-old male mouse in which there is a localized area of total ctypt loss associated with separation of the lining epithelium. (6) Higher magnification of boxed area in A. Necrosis of lining epithelial cells (arrowheads) is associated with the separation of the epithelium. (C) Cecum from a 4-week-old male mouse in which the epithelium has been denuded to form a large ulcer. Exudation of neutrophils covers the defect. (D) Cecum from a B-weekaId male mouse showing margination of lymphocytes along the wall of a small vein in the submucosa (H&E; original magnifications: A, 100x; 6, 400x; C, 100x; D, 400x). (Continued on next page.)
IBD IN CSH/HeJBir
December 1994
ined
in this
study
inflammatory When
and was probably
collectively,
moderate
and histological
it is apparent
to severe cecal and mild, but frequent,
anorectal
appear together
in affected mice and constitute
features of the disease. Compilation
the criteria
indicated
that young
affected
period,
during
through
adult
puberty,
would occasionally
extend
relatively
compared
with
those of the cecum.
In
markers
cosa and lamina propria
strains
tration of neutrophils in local areas of the mid colon. One of the 3-week-old mice had similar mild colonic involvement
limited
to a small area approximately
0.5
genetic
existed
between
No differences tested
allele,
represented
an extension
of the ulcerative
standard
genetic
and
C3H/HeJBir
between
Chr 1 (I&
dif-
the two
1, Akp 1, Pep
Gr I), Chr 9 (Apoa 1, TrJ Mod 1, Mpi I), Chr 11 (Es
toxin
ulcerations,
used
3), Chr 3 (Amy 1, Cur 2), Chr 4 (Pgm 2, Gpd l), Chr 5 (Pgm I), Chr 7 (Gpi 1, Hbb), Chr 8 (Es 1, Es 11, Got 2,
indicating
inflammation
assays
any easily identifiable
C3H/HeJ
for 24 markers:
two substrains,
perianal
of
were observed
consisted primarily cells and neutrophils mice with prominent
from the presence
all microbiological
analyses whether
cm from the anus. Mice 4 weeks of age and older had a mild, inflammatory cell infiltrate in the distal colon that
Because the distal colonic
of
mouse pathogens.
3), Chr 14 (Es lo),
of lymphocytes with some plasma with mild edema of the submucosa.
incidence
we tested 20 affected animals
colonies,
for known
to determine
ferences mice.
with a very mild infil-
the increased
Genetic Analyses
mice 2 weeks of age, there was mild edema of the submuassociated
in the oldest
mild.
Bir mice resulted
C3H/HeJ
were negative
Initial subtle
whether
or viral pathogens,
production
to include
the rest of the colon were mild and
affected with
by culture and serology as described in Materials and Methods. As previously found for sporadic cases in the
this region. Lesions involving
in
Analysis
To determine bacterial
was pres-
uncommon
Lesions evident
3) mice were generally
IBD in C3H/HeJ
to weaning
cecal inflammation
mice. As mice aged, the percentage
Microbial
ulceration
the suckling
it was relatively
any of the lesions decreased. (group
of all of
male and female mice
a time when they first start to eat an
adult chow diet. Although ent
are
bleeding,
were initially
perianal
features
the primary
transition
to the
that occult
pathology
figure 6 (cent’d.). (E) Cecum from a l@week-old male mouse with an island of spared mucosa with hyperplastic crypts surrounded by areas lacking crypts. (H&E; original magnification 100x).
unrelated
1733
process in the cecum.
all of the clinical
viewed
MICE
Netl 1, H2-K). resistance
Chr 15 (Gpt l), or Chr 17 (G/o 1,
LPS responsiveness
was also similar
they both
carried
Lpsd. We concluded
in the
the endothat
C3Hl
it most likely
HeJBir mice were genetically indistinguishable from C3H/HeJ mice in standard tests used to assure the ge-
process.
netic integrity
was observed
in
Small intestinal lesions were essentially nonexistent. In one 7 1-week-old female mouse, there was a solitary ulcer in the duodenum adjacent to the pylorus. The villi had undergone necrosis that left a short linear area of necrotic debris containing large numbers of neutrophils. The tip on one adjacent villus was necrotic and filled with neutrophils. No pathogenic organisms were observed in or around the ulcer. No other intestinal lesions were observed in this animal. This was the only small intestinal lesion observed in any of the C3H/HeJBir mice exam-
of C3H/HeJ
mice.
The negative microbiological results raised the interesting possibility that genetic background alone may be a primary cause of spontaneous IBD in C3H/HeJBir mice. To further explore this possibility, a cross was made between affected C3H/HeJBir mice and inbred Mus castaneozls (CASTlEi) mice. We evaluated 19 female and 19 male mice in the F, generation, ranging in age at the time of necropsy from 8 to 31 weeks. Although these mice were not assayed for the presence of infectious agents, none of these offspring showed clinical or histo-
1734
SUNDBERG
ET AL.
GASTROENTEROLOGY
logical signs of disease. These results
suggest
that more
carries a known
than one gene is involved in determining the high incidence of IBD in our C3H/HeJBir colony. They also pro-
certain
vide evidence
mice,
agent
that disease is not caused by a microbial
that is a strict pathogen
Mice
should
in all production
colonies
Over the past decade,
gested
for this with
that
reason
other
taneous
These
breeding
One of approximately
mice have
high
frequency
to establish
nal study was performed of the clinical This curred
mainly
in the second the transition mother’s
a pro-
resulted
in
study
to fourth
history
that
the colitis
includes
the mice are changed Overt
oc-
colon and began
week of life, which diet.
clinical
from
signs
of
disease, if present, usually become evident later at 4-6 weeks of age. The clinical features of most mice were mild
with
symptom
>lO%
of the animals
complex characterized
very soft feces, and extensive time the mice were young
developing
by positive
perianal adults
the full
occult blood,
ulceration.
By the
at 8 weeks of age or
older, virtually all appeared to be clinically normal. We can only speculate on the reasons for the high frequency of lesions at the time of weaning. in the mice at this critical
transition
models,
Many changes time;
occur
two that are
possibly significant relative to the development of colitis are worthy of mention. First, this is the time when bacteria colonize the gut, and the largest numbers and types of bacteria would occur mainly in the cecum and right colon where the lesions are most frequent.‘* Second, there is a retrovirus present in the mother’s milk of this strain, known as mouse mammary tumor virus,15 which encodes a superantigen that interacts with certain T-cell receptor subsets with an initial stimulation and then deletion. Both of these possible explanations for the colitis are under active study. The role that the immune system plays in the colitis of C3H/HeJBir mice is unknown. The C3HlHeJ strain
that
of antibodies
by homologous
C3Hl
in mice.
system
In at least one of these
(IL-2) knockout,
develops milder
recombi-
in the immune
the mice do not
do when kept under conditions.” The IL-
IBD, mainly
limited
colon, when the colony is maintained
pathogen-free
of
with the
Several recent reports indi-
when germ-free but specific pathogen-free
10 knockout
a variety
data indicate
frequency
deletion
the interleukin
LPS
LPS is
in the normal enteric bacterial
strains.”
in colitis
get disease conventional
However,
and interact
of other genes important
can result
obvious
to bacterial
are not seen in the normal
cate that the selective nation
macrophage
that contain
unpublished
present
HeJ or other inbred
right
features.
showed
when
to a solid
a longitudi-
the natural
in the cecum and right period
milk
this substrain,
and pathological
longitudinal
Recent
to various proteins
mice was begun.
to colitis.
mice have a high
flora; such antibodies
pedigrees
to determine
system.
sug-
this possibility,
of these
also has this re-
responsiveness
of bacteria
mice
a new substrain, CSH/HeJBir, with a high incidence of perianal ulceration and diarrhea in the offspring. During the time required
immune
in
system
in certain
also able to stimulate
of C3H/HeJ
of affected
10 C3H/HeJ
but one component
C3H/HeJBir
to have resulted
substrain
be more susceptible
constituents
to
to LPS. It is not immediately
observations
than most strains to a spon-
form of IBD. To examine
gram of selective
C3H/HeJ
background
made them more susceptible
for evalu-
feces or perianal
at a fairly
strains.
the genetic
of the Jackson
laboratory
if they have soft, light-colored
been culled
reactivity
why mice with reduced
are sent to the diagnostic
compared
defects
The C3H/HeJBir
No. 6
their response
of the immune
most prominently
duced
ulceration.
that reduces
LPS.16 This defect is known abnormalities
functions.”
Discussion Laboratory ation
for mice.
bacterial
mutation
Vol. 107,
environment.”
to the
in a specific
Interestingly,
the
IL-10
knockout
develops
colonic
inflammation
at the time of
weaning,
a timing
similar
to that seen in C3H/HeJBir
mice. A reasonable
unifying
hypothesis
is that a dysregu-
lated immune response to the normal bacterial flora may represent a common pathogenesis in all of these mouse models,
including
the C3H/HeJBir
mice.
The high frequency of IBD in Jews and in certain families along with the IO-fold increased risk of ulcerative
colitis
strongly
and
suggests
Crohn’s that
cause.*’ The increased
disease
these
in
family
disorders
incidence
of IBD in offspring
two affected parents raises the possibility be only a limited number of genes leading tibility.*’
members
have a genetic
In light of these findings,
of
that there may to IBD suscep-
an important
ques-
tion to consider is whether any of the susceptibility genes in the human population are homologous to susceptibility genes in C3H/HeJBir mice.** If so, genetic mapping studies using C3H/HeJBir mice may identify homologous regions in human chromosomes where IBD susceptibility genes are located. High-resolution mapping in mice is now possible and could identify candidate genes to investigate
in human
patients.
In summary, a new substrain of mice has been established with a heritable form of colitis. The disease has a natural history and pathology that are different from other described mouse models.’ Because of the abundance of reagents for immunologic and genetic studies in mice,
IBD IN CBH/HeJBir
December 1994
this model provides new avenues of investigation
into
IBD, particularly investigation of the complex interplay between genetic, immune, and environmental components that are involved in colitis induction and progression.
References I. Kim H-S, Berstad A. Experimental colitis in animal models. Stand J Gastroenterol 1992; 27:529-537. 2. Chalifoux LV, Bronson RT. Colonic adenocarcinoma associated with chronic colitis in cotton top marmosets, Saguinus oedipus. Gastroenterology 1981; 80:942-946. 3. Madara JL, Podolsky DK, King NW, Sehgal PK, Moore R, Winter HS. Characterization of spontaneous colitis in cotton-top tama rins (Saguinus Oedipus) and its response to sulfasalazine. Gastroenterology 1985;88:13-19. 4. Sundberg JP. Conceptual evaluation of inbred laboratory mouse models as tools for the study of diseases in other species. Lab Anim 1992;22:48-51. 5. Moolenbeek C, Ruitenberg El. The “Swiss roll”: a simple technique for histologic studies of the rodent intestine. Lab Anim 1981; l5:57-59. 6. Popesko P, Rajtova V, Horak J. A colour atlas of the anatomy of small laboratory animals. 2. Rat, mouse, golden hamster. London: Wolfe, 1992:146. 7. Sundberg JP, Cordy WR, King LE Jr. Alopecia areata in aging C3H/HeJ mice. J Invest Dermatol 1994;102:847-856. 8. Bedigian HG. Animal health and genetic quality control report. Bar Harbor, ME: Jackson Laboratory, 1993:1-2. 9. Green MC. Catalog of mutant genes and polymorphic loci. In: Lyons MC, Searle AG, eds. Genetic variants and strains of the laboratory mouse. 2nd ed. Oxford, England: Oxford University Press, 1989:12-403. 10. Shiroshi T, Sagai T, Moriwaki K. A simplified micro-method for cyto-toxicity testing using a flat-type titration plate for the detection of H-2 antigens. Microbial lmmunol 1981;25:1327-1334. 11. Vetvicka V. C3H/HeJ mice. In: Rihova BR, Vetvicka V, eds. Immunological disorders in mice. Boca Raton, FL: CRC, 1991:153171. 12. Leiter EH, Beamer WG, Shultz LD. The effect of immunosuppression on streptozotocin-induced diabetes in C57BL/KsJ mice. Diabetes 1983; 32:148-155.
MICE
1735
13. Dietrich W, Katz H, Lincoln SE, Shin H-S, Friedman J, Dracopoli NC, Lander ES. A genetic map of the mouse suitable for typing intraspecific crosses. Genetics 1992; 131:423-447. 14. Schaedler RW, Dubos R, Costello R. The development of the bacterial flora in the gastrointestinal tract of mice. J Exp Med 1965; 122:59-66. 15. Gross L. Oncogenic viruses. 2nd ed. Oxford, England: Pergamon Press, 1970:238-280. 16. Glode LM, Rosenstreich DL. Genetic control of B cell activation by bacterial lipopolysaccharide is mediated by multiple distinct genes or alleles. J lmmunol 1976;117:2061-2066. 17. Brandwein S, McCabe RP, Dadrat A, Ridwan BU, Birkenmeier EH, Sundberg JP, Elson CO. Immunologic reactivity of colitic C3H/ HeJBir mice to enteric bacteria (abstr). Gastroenterology 1994; 106:A656. 18. Sadlack B, Merz H, Schorle H, Schimpl A, Feller AC, Horak I. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 1993;75:253-261. 19. Kuhn R, Lohler J, Rennick D, Rajewsky K, Muller W. InterleukinIO-deficient mice develop chronic enterocolitis. Cell 1993;75:263-274. 20. Orholm M, Mukholm P, Langholz E, Nielsen OH, Sorensen TIA, Binder V. Familial occurrence of inflammatory bowel disease. N Engl J Med 1991;324:84-88. 21. Yang H, Shohat T, Rotter JI. The genetics of inflammatory bowel disease. In: MacDermott RP, Stenson WF, eds. Curr Top in Gastroenterol Elsevier, 1992:17-51. 22. Rotter JI, Yang H. Resolving the genetics of IBD: the challenge for the ’90s. Prog Infl Bowel Dis 1993; 14:1-7.
Received April 4,1994. Accepted August 4,1994. Address requests for reprints to: John P. Sundberg, D.V.M., Ph.D., The Jackson Laboratory, 600 Main Street, Bar Harbor, Malne 04609 1500. Fax: (207) 2882119 or -5079. Supported by a grant from the Crohn’s and Colitis Foundation of America (to E.H.B., J.P.S.); the National Alopecia Areata Foundation (to J.P.S.); and by grant DK44240 from the National Instltutes of Health (to C.O.E., E.H.B., J.P.S.). The authors thank Aletha Torrey for outstanding technical assistance and her dedication to the development, maintenance, and characterization of the C3H/HeJBlr colony: Colleen M. Vallee for technical assistance with the necropsies; and Beth A. Sundberg for computer graphics.
Spontaneous, Heritable Colitis in a New Substrain of C3H/HeJ Mice JOHN
P. SUNDBERG,*
and EDWARD
CHARLES
0.
ELSON,’
HENDRICK
BEDIGIAN,*
H. BIRKENMEIER*
*The Jackson Laboratory, Bar Harbor, Maine; and tDivision of Gastroenterology,
University of Alabama, Birmingham, Alabama
Background/Aims: C3H/HeJ mice at the Jackson Labo-
ing, lack of immunologic
ratory have periodically been culled because of the occurrence of soft feces, perianal ulceration, and rightsided colitis. No pathogens have been isolated. The goal of the current study was to establish a substrain with a high Incidence of this disease. Methods: Affected male and female C3H/HeJ mice were bred. The clinical, pathological, microbiological, and genetic features of 216 mice of the resulting pedigree were characterized. Results: A severely affected female crossed with a normal male resulted in a new substrain, de noted C3H/HeJBir, with a high incidence of right-sided colitis. Histologically, lesions occurred primarily in the cecum and proximal colon, characterized by acute and chronic inflammation, crypt abscesses, ulcerations, regenerative hyperplasia, and submucosal scarring. Such colitis peaked at 3-6 weeks; however, similar disease was found sporadically in animals more than 1 year of age. Small lesions at the anorectal junction were common throughout life. An extensive search for patho gens was negative. Genetic analysis of CBH/HeJBir mice suggested that the disease was inherited as a quantitative trait. Conclusions: CSH/HeJBir mice develop a spontaneous, heritable form of idiopathic inflammatory bowel disease and will be a valuable resource for genetic and immunologic studies of this disease.
have limited
hronic
inflammatory
plex disorders
bowel diseases (IBD) are com-
whose pathogeneses
are poorly under-
stood. Colitis can be induced in experimental animals to study various aspects of IBD.’ These models have increased our understanding of nonspecific inflammatory mechanisms. However, most of these induced models do not allow study of the complex interactions between genetic, environmental, and immune components important in disease initiation and progression. The model with a high degree of homology to many forms of human IBD is spontaneous colitis that develops in cotton-topped tamarins.‘-3 Unfortunately, the endangered species status of these primates, difficulty in handling, expense of hous-
and other problems
Models of disease in genetically bred laboratory
defined
mice circumvent
investigation
of chronic
diseases, including particularly
the identification Thus,
of genes important
a mouse model
useful model system to investigate mechanisms
secondary
perianal
of IBD,
would
the primary
be a
immuno-
that can lead to IBD.
Because of the occurrence and/or
or autoimmune
one that occurs spontaneously,
logic and genetic
of in-
animals. Mice and genetic
inflammatory
in disease susceptibility.*
strains
many of the problems
inherent in models using larger, outbred are the premier species for immunologic
of soft, light-colored
ulceration,
feces
mice of the C3H/
HeJ strain in the research and production colonies of the Jackson Laboratory have been culled in disproportionate numbers
compared
with
fected mice had histological pathogens
were never
Because it seemed netic predisposition
other strains.
At necropsy,
evidence
of colitis, although
identified
with
routine
likely that C3H/HeJ to develop
af-
testing.
mice had a ge-
a form of IBD, we selec-
tively bred affected C3H/HeJ mice with perianal ulcerations to generate a pedigreed “high-susceptibility” substrain, C3H/HeJBir, with a high incidence of spontaneous colitis. The purpose of this study was to describe the natural history, clinical and pathological features, and probable
C
reagents,
their usefulness.
mode of inheritance
in C3H/HeJBir
of spontaneous
colitis
mice.
Materlab and Methods Materials
and Animals
The studies were approved by the Animal Care and Use Committee of the Jackson Laboratory. C3H/HeJ mice with P&anal ulcerations were initially obtained from the Jackson Laboratory’s production colonies in Morrel Park (Bar Harbor, ME). These mice and their offspring were selectively bred to Abbreviations used in this paper: IL, interleukln; charide. 0 1994 by the American Gastroenterological 00165085/94/$3.00
LPS, lipopolysacAssociation
IBD IN CM/HeJBir
December 1994
derive a new substrain, of perianal
To determine present
C3H/HeJBir,
ulceration,
as described
the onset, severity,
in the gastrointestinal
longitudinal between
and type of inflammation
tract
of C3HIHeJBir
study of disease progression
mice, a
was instituted.
the ages of 10 days and 90 weeks underwent
at weekly intervals.
The mice were later divided
groups
The first group
for analysis.
to 3 months second group months
consisted
of mice from 0
of mice from 4 to 12 months
88 total); the
of age (37 males
65 total); and the third group
of age (21 males and 42 females,
Mice
necropsy
into three age
of age (56 males and 32 females,
and 28 females,
earlier findings.
with a high frequency in Results.
of mice >12
63 total). A total of
Serum was collected, immediately
large intestine, all animals.
and mice were killed by CO* asphyxia-
before cecum,
necropsy.
The stomach,
and perianal
Intestinal
small
skin were collected
rolls were prepared
by placing
and from
them on
an index card, twisting
them
in concentric
centrifugal
around
pick
in the plane
of the card.’ The
a central
unopened using
tooth
intestinal
rolls were gently
a 27-gauge
further
needle
solution stained
processed
circles
with
fixative
were tabulated Redmond, erpoint;
to 70% ethanol.
routinely,
sectioned
Microsoft
a spreadsheet
subacute,
(Excel; Microsoft
using a graphics
or chronic)
moderate, using
program
Corp., (Pow-
based on loca-
or severe), and quality commonly
accepted
(acute,
histological
Briefly, mild lesions were small, focal, or widely sepa-
rated multifocal
areas of inflammation
to the lamina propria. or represented
Moderate
or fibrosis extending
and/or fibrosis limited
areas of inflammation
and/
into the submucosa.
Severe lesions were
ulcers that covered large areas (a millimeter
or more of mucosa)
of cecum or colon. The ulcers were covered neutrophils
fibrosis and loss of crypts.
crypt necrosis and/or ulceration phils
infiltrating
had a mixture cells with
by an exudate
of
and necrotic debris, or the areas were epithelialized
with underlying
and/or
minimal
Acute lesions had
with a predominance
covering
of neutrophils
the ulcers.
and mononuclear
or no fibrosis.
defined beds of granulation
were stained
Chronic
were removed
examination.
All sections
the proximal
for gram-negative
and Campylobacter spp. (Campy Systems,
Microbiology
GDVII,
colons and
aerobes (GN broth)
Pouch; Becton Dickinson
Cockeyville,
Systems).
immunosorbent pneumonia
of males and fe-
The cecums
stains. In addition,
were cultured
crobiology
Ten C3Hl
with H&E, Brown and Brenn (tissue gram), and
Warthin-Starry cecums
Laboratory. substrain,
Equal numbers
for histopathologic
MD).’ Specimens
Serum
Mi-
were also
(CDT; Becton Dickinson
was tested
by enzyme-linked
following
mouse
assay for the
pathogens:
virus of mice, Sendai virus, mouse hepatitis
minute
virus of mice, ectromelia,
reo 3, epizootic
meningitis,
cytomegalovirus,
mouse
Mycoplasma pulnzonis,
diarrhea
adenovirus
virus,
lymphocytic
of infant (MAD),
chorio-
mice,
mouse
polyoma
virus,
and Mycoplasnzu arthvitidis. a
Genetic Quality Control and Genetic Studies
of neutro-
Subacute
At the time affected mice were initially the production
colony,
for 204-208
been inbred type integrity
used to compare
areas
inflammatory
lesions had well-
C3HIHeJ
generations.
To compare
markers
methods
Defective This mutation, immune
to bacterial C3H/HeJ
designated
genetic
and sub-
quality control
assay for H-2.“,“’ is a fea-
mice from other C3H strains.
Lpsd, has a pronounced
effect on the
for their lipopolysaccha-
by an in vitro splenocyte
of IBD in C3H/HeJBir
affected C3H/HeJBir
C3Hi
control were
lipopolysaccharide
Mice were tested
ride (LPS) responsiveness crossing
standard
and a microcytotoxic
response
system.”
Inheritance
quality
had
the geno-
with the progenitor
for genetic
from
strain
female mice from both the primary
strain colonies. Tests included isoenzyme
obtained
the progenitor
of the new substrain
HeJ colony, standard
ture that distinguishes
lesions were either multifocal
by locally extensive
were also evaluated.
and processed
and
and the results
Corp.). Lesions were graded
(mild,
HeJBir,
evidence
were obtained
of the Jackson
mice from the newly established
males were used in both groups.
Tissues
at 6 pm,
by light microscopy,
WA) and graphed
tion, severity criteria.
using
colonies
additional
mice without
examination
1727
The rolls were
with H&E.
Samples were evaluated
from the production
in Fekete’s acid-alcohol-formalin
for 12 hours, then transferred
were trimmed,
inflated
and 5-mL syringe.
fixed by immersion
C3H/HeJ
of physical
tested for Clostridiwn di#icile antigens
114 male and 102 female mice were studied. tion
Ten normal
of IBD at the time
MICE
assay.”
mice was evaluated
by
mice with inbred Mus castanets
(CAST/Ei) and examining offspring in the F1 and F2 generations for evidence of disease.”
ReSUttlB
Colony Establishment
tissue under ulcers or reepithelial-
ized regions of cecum. Inflammatory
cells were predominantly
lymphocytes. For purposes of comparison to human anatomy and therefore human disease, the term right colon hereafter includes the cecum and colon up to the hepatic
flexure in the mouse.6
Microbiology Although no mouse pathogens were identified during prior diagnostic tests of affected C3H/HeJ mice, the newly established substrain was also tested for pathogens to verify
Over a period of 1 year, a pedigreed
substrain
was
established starting initially with one severely affected female and a clinically normal male. Because we did not know the etiology of bowel disease in mice originating from production colonies, each affected mouse was treated as an individual case, and sick mice were therefore not housed together. Once a pedigreed colony was established, affected individual mice were bred to each other. We initially used several sporadic severe cases to start multiple pedigrees. The pedigree with the highest inci-
1728
SUNDBERG ET AL.
GASTROENTEROLOGY Vol. 107, No. 6
developed
rua * AA AuAAUAA
clinical
of perianal
disease as determined
by the presence
ulcers and soft, yellow feces.
ClinicopathologicCorrelationWith Natural History of Disease A longitudinal
study determined
the age of onset,
duration, and pathological progression of right-sided colitis in more than 200 C3H/HeJBir mice between 2 and
90 weeks of age. The microscopic compared
**
*** **
with the clinical
of death. Clinically,
(*I.
in young mice between
were also positive,
including ulceration
was confined
to a relatively
short period
was selected for further generations
study. Offspring
were evaluated
ence of perianal
clinically
ulcerations,
CA; Table
for IBD by the pres-
soft feces, and positive
occult test results (SmithKline HeJBir
in all successive
Diagnostics
1). This new substrain
and is now maintained
Inc., San Jose,
was designated by strict
HemC3H/
brother-sister
matings (currently Fi3) with at least one parent positive for disease. The initial pedigree of the substrain is shown in Figure 1.
or histological
with age. Results
the number
evidence
are summarized
in Figure
2.
indicator of disease
mice. There was a correlation
positive Hemoccult inflammation
of mice with
of disease decreased
Occult blood was the most predictive in C3H/HeJBir
after weaning
weeks old. Once healed, these
ulcers did not recur. Overall, 1)
primarily
some at 1 year of age. The presence of perianal
out clinical (Figure
was then at the time
4 and 17 weeks of age. However,
males
when the mice were 4-7
dence of disease in the F2 and Fs generations
present
occult blood was detected
a few of the older
figure 1. Pedigree of CBH/HeJBir mice. Incidence of IBD is indicated by perianal ulceration (W, male; ?? , female) and soft, yellow feces
pathology
findings
between
test result and histological
a
evidence of
in the colon. Mice with trace amounts of occult
blood in the feces usually had small ulcers or erosions evident at the mucocutaneous amination
anorectal junction
on histological
ex-
of the tissues (see below).
Gross evidence ence of moist
of severely affected mice was the pres-
areas of perianal
The ulcers were consistently
ulceration
(Figure
found in animals
3A).
with soft,
Clinical Features at 3-5 Weeks of Age
yellow feces. Microscopically,
During
small to extensive areas of ulceration of the squamous epithelium of the haired skin adjacent to or extending
the 2 years of this study, the QH/HeJBir
colony generated 678 mice. Each mouse was examined once between 3 and 5 weeks of age for clinical evidence of IBD. The clinical
data on the 326 female
male mice are summarized
in Table
and 352
1. Approximately
these lesions
from the anus. The underlying with
granulation
collected
tissue
for evaluation
dermis
was usually
filled
the tissues
were
(Figure
nation
to be positive
blood when tested during
this
developed perianal ulcerations. In 50% of the animals, the only clinical sign of disease was trace amounts of occult blood in the feces. Other animals had various combinations of symptoms including occult blood with perianal ulceration, occult blood with soft stools, and occult blood with both soft stools and perianal ulceration. Within the substrain colony, 21% of the females and 12 % of the males appeared normal and never developed clinical signs of IBD. Both serological testing and routine fecal cultures did not identify any pathogens in the C3H/ HeJBir mice. Less than 1% of C3H/HeJ mice in the production colony from which this substrain was derived
Table
of the gastrointestinal
3B).Removal and exami-
the anus showed that changes were limited
for occult
1.
Frequency
tract from the stomach
of Clinical
C3H/HeJBir
Disease
Blood only Perianal ulceration only Soft feces only Blood and ulceration Blood and soft feces Ulceration and soft feces Ulceration, soft feces, and blood None
to
to the cecum
in the Colony
of
Mice Female
Clinical signs
of
by the time
79% of the female and 88% of the male mice were found time. Approximately 25% of the mice of both sexes has soft stools, and 8% of the females and 27% of the males
consisted
Male
Total of mice (56)
No.
%
No.
%
50 0 0 9 16 0
171 0 0 14 61 0
52 0 0 4 19 0
170 0 0 44 47 0
48 0 0 13 13 0
9 16
12 68
4 21
50 41
14 12
IBD IN CBH/HeJBir
December1994
lOS%
MICE
1729
A
80%
8@= $ ke 40%
20%
0%
1 o-3
o-3
13 - 18
4-12
4.12
13-18
Age in month intmwls
Age in month intavals
Figure 2. Summary of clinical and pathological data. (A) Percentages of CBH/HeJBir mice in the longitudinal study with clinically (M, positive occult blood, perianal ulcer, soft stool) or histologically confirmed (0, ulceration, inflammation, edema, or fibrosis) disease. (6) Percentages of CBH/HeJBir mice with various histological grades of severity of cecal disease in the three age groups.
and colon. Normally, has a prominent moist material.
the lower intestinal
cecum that usually
tract of a mouse
contains
dark green
As the feces enter the colon, they become
dark green to black, oval, and hard (Figure
4A). In con-
trast, severely affected C3H/HeJBir mice had small, shrunken, yellow cecums that often contained little or no ingesta (Figure 4B). The colons were slightly distended with mustard yellow soft feces that formed soft oval pellets. This soft fecal material became matted in the perianal
ure 5C). In focal areas, where crypts the lamina propria was highly cellular, necrosis
5D). Adjacent mitotic
crypts activity.
were hyperchroAnimals
4 and 5 weeks of age had small focal ulcers. maining
lamina
mixed,
propria
predominantly
was thickened mononuclear
necrosis lamina
between The re-
and contained cell infiltrate
ure 6A and B). In these areas, the epithelium going the
hair.
(Figure
matic with increased
were short and there was crypt
a
(Fig-
was under-
of individual
cells with separation from 6B). Once the mucosa (Figure
propria
Microscopically, lesions were primarily limited to the cecum and right colon. Lesions in the cecum progressed
sloughed,
through a series of stages. The earliest age at which mice underwent necropsy was 10 days. All suckling
uded into the lumen (Figure 6C). Vessels in the muscular walls were markedly dilated (Figure 60). By 6-7
pups examined
weeks of age, the ulcers
at this age had normal
gastrointestinal
tracts. However, during the suckling to weaning transition period beginning at 14 days of age, all mice examined had marked edema of the submucosa and, to a lesser extent, the lamina propria (Figure 5A). Although the lining epithelium of the cecum was intact, there were large numbers of neutrophils in the lamina propria and scattered neutrophils in the edematous submucosa (Figure 5B). The cecum was not uniformly involved, with the most severe lesions observed at the entry of the ileum and directly across from it. A patchy, multifocal pattern of inflammation involved other parts of the cecum. Between 3 and 4 weeks of age, the inflammation became subacute, consisting of a mixture macrophages, lymphocytes, and neutrophils in the lamina propria. The submucosa remained edematous with a loose, diffuse infiltrate similar to that of the lamina propria (Fig-
numbers
the submucosa of neutrophils
inflammatory
cell
became
edematous
in the lamina
and large
propria
were reepithelialized
infiltrate
had
become
were ex-
and the
subacute
to
chronic. Margination of lymphocytes within small veins in the submucosa was prominent (Figure 6D). By 10 weeks of age, the ulcers had healed, leaving areas without crypts surrounding small islands of mucosa with regenerative erative
crypt
crypts (Figure epithelium
6E). The surrounding apparently
replaced
regenthe de-
pressed areas lacking crypts, because after 10 weeks of age, no evidence of ulceration was found. Ulcers or chronic scarring with small ulcerations were occasionally found in the cecums of older mice. The latter were located above gut-associated lymphoid tissue. Microscopic lesions of the anorectal mucosa were common and consisted of small erosions or ulcerations at the mucocutaneous junction, with a mild to moderate exudation of neutrophils into the lumen. Severe cases of
1730
SUNDBERG
ET AL.
GASTROENTEROLOGY
Vol. 107,
No. 6
Flgure 3. Perianal ulceration. (A) Perianal ulceration extending over the scrotum and ventral tail head in a 5-week-old male C3H/HeJ mouse with severe cecal lesions. (B) Desiccated proteinaceous material and necrotic debris cover a large perianal ulcer and a bed of granulation tissue. Adjacent epidermis is thickened (H&E; original magnification 25x).
Figure 4. Gross pathology of mouse cecum and colon. (A) Normal distended cecum (so/id arrow), entrance of the ileum (open arrow), and colon with three dark fecal pellets in an unaffected C3H/HeJ mouse. (B) The cecum is small, pale, and difficult to identify (so/id arrow) next to the entrance of the ileum (open arrow) in an affected C3H/HeJBir mouse. The colon contains poorly formed, lightcolored fecal pellets.
December 1994
IBD IN C3H/HeJBir
MICE
1731
figure S. Histopathology of cecum and colon of CBH/HeJBir mice at 2-3 weeks of age. (A) Cecum of a 2-week-old male mouse has marked edema of the submucosa. (B) Higher magnification of boxed area in A showing the locally extensive infiltration of neutrophils into the lamina propria. (C) Cecum of a 3week-old male mouse that has a large area of early crypt loss, mixed inflammatory cell infiltrate in the lamina propria, and mild edema of the submucosa. (/I) Higher magnification of boxed area in C. Early crypt loss is associated with necrotic debris in crypts (arrow) adjacent to crypts undergoing necrosis (H&E; original magnifications: A, 25x; B, 100x; C, 100x; D, 400x).
1732
SUNDBERG
ET AL.
GASTROENTEROLOGY
Vol. 107,
No. 6
Flgure 6. Histopathology of cecum and colon of CBH/HeJBir mice at 4-10 weeks of age. (A) Cecum of a a-week-old male mouse in which there is a localized area of total ctypt loss associated with separation of the lining epithelium. (6) Higher magnification of boxed area in A. Necrosis of lining epithelial cells (arrowheads) is associated with the separation of the epithelium. (C) Cecum from a 4-week-old male mouse in which the epithelium has been denuded to form a large ulcer. Exudation of neutrophils covers the defect. (D) Cecum from a B-weekaId male mouse showing margination of lymphocytes along the wall of a small vein in the submucosa (H&E; original magnifications: A, 100x; 6, 400x; C, 100x; D, 400x). (Continued on next page.)
IBD IN CSH/HeJBir
December 1994
ined
in this
study
inflammatory When
and was probably
collectively,
moderate
and histological
it is apparent
to severe cecal and mild, but frequent,
anorectal
appear together
in affected mice and constitute
features of the disease. Compilation
the criteria
indicated
that young
affected
period,
during
through
adult
puberty,
would occasionally
extend
relatively
compared
with
those of the cecum.
In
markers
cosa and lamina propria
strains
tration of neutrophils in local areas of the mid colon. One of the 3-week-old mice had similar mild colonic involvement
limited
to a small area approximately
0.5
genetic
existed
between
No differences tested
allele,
represented
an extension
of the ulcerative
standard
genetic
and
C3H/HeJBir
between
Chr 1 (I&
dif-
the two
1, Akp 1, Pep
Gr I), Chr 9 (Apoa 1, TrJ Mod 1, Mpi I), Chr 11 (Es
toxin
ulcerations,
used
3), Chr 3 (Amy 1, Cur 2), Chr 4 (Pgm 2, Gpd l), Chr 5 (Pgm I), Chr 7 (Gpi 1, Hbb), Chr 8 (Es 1, Es 11, Got 2,
indicating
inflammation
assays
any easily identifiable
C3H/HeJ
for 24 markers:
two substrains,
perianal
of
were observed
consisted primarily cells and neutrophils mice with prominent
from the presence
all microbiological
analyses whether
cm from the anus. Mice 4 weeks of age and older had a mild, inflammatory cell infiltrate in the distal colon that
Because the distal colonic
of
mouse pathogens.
3), Chr 14 (Es lo),
of lymphocytes with some plasma with mild edema of the submucosa.
incidence
we tested 20 affected animals
colonies,
for known
to determine
ferences mice.
with a very mild infil-
the increased
Genetic Analyses
mice 2 weeks of age, there was mild edema of the submuassociated
in the oldest
mild.
Bir mice resulted
C3H/HeJ
were negative
Initial subtle
whether
or viral pathogens,
production
to include
the rest of the colon were mild and
affected with
by culture and serology as described in Materials and Methods. As previously found for sporadic cases in the
this region. Lesions involving
in
Analysis
To determine bacterial
was pres-
uncommon
Lesions evident
3) mice were generally
IBD in C3H/HeJ
to weaning
cecal inflammation
mice. As mice aged, the percentage
Microbial
ulceration
the suckling
it was relatively
any of the lesions decreased. (group
of all of
male and female mice
a time when they first start to eat an
adult chow diet. Although ent
are
bleeding,
were initially
perianal
features
the primary
transition
to the
that occult
pathology
figure 6 (cent’d.). (E) Cecum from a l@week-old male mouse with an island of spared mucosa with hyperplastic crypts surrounded by areas lacking crypts. (H&E; original magnification 100x).
unrelated
1733
process in the cecum.
all of the clinical
viewed
MICE
Netl 1, H2-K). resistance
Chr 15 (Gpt l), or Chr 17 (G/o 1,
LPS responsiveness
was also similar
they both
carried
Lpsd. We concluded
in the
the endothat
C3Hl
it most likely
HeJBir mice were genetically indistinguishable from C3H/HeJ mice in standard tests used to assure the ge-
process.
netic integrity
was observed
in
Small intestinal lesions were essentially nonexistent. In one 7 1-week-old female mouse, there was a solitary ulcer in the duodenum adjacent to the pylorus. The villi had undergone necrosis that left a short linear area of necrotic debris containing large numbers of neutrophils. The tip on one adjacent villus was necrotic and filled with neutrophils. No pathogenic organisms were observed in or around the ulcer. No other intestinal lesions were observed in this animal. This was the only small intestinal lesion observed in any of the C3H/HeJBir mice exam-
of C3H/HeJ
mice.
The negative microbiological results raised the interesting possibility that genetic background alone may be a primary cause of spontaneous IBD in C3H/HeJBir mice. To further explore this possibility, a cross was made between affected C3H/HeJBir mice and inbred Mus castaneozls (CASTlEi) mice. We evaluated 19 female and 19 male mice in the F, generation, ranging in age at the time of necropsy from 8 to 31 weeks. Although these mice were not assayed for the presence of infectious agents, none of these offspring showed clinical or histo-
1734
SUNDBERG
ET AL.
GASTROENTEROLOGY
logical signs of disease. These results
suggest
that more
carries a known
than one gene is involved in determining the high incidence of IBD in our C3H/HeJBir colony. They also pro-
certain
vide evidence
mice,
agent
that disease is not caused by a microbial
that is a strict pathogen
Mice
should
in all production
colonies
Over the past decade,
gested
for this with
that
reason
other
taneous
These
breeding
One of approximately
mice have
high
frequency
to establish
nal study was performed of the clinical This curred
mainly
in the second the transition mother’s
a pro-
resulted
in
study
to fourth
history
that
the colitis
includes
the mice are changed Overt
oc-
colon and began
week of life, which diet.
clinical
from
signs
of
disease, if present, usually become evident later at 4-6 weeks of age. The clinical features of most mice were mild
with
symptom
>lO%
of the animals
complex characterized
very soft feces, and extensive time the mice were young
developing
by positive
perianal adults
the full
occult blood,
ulceration.
By the
at 8 weeks of age or
older, virtually all appeared to be clinically normal. We can only speculate on the reasons for the high frequency of lesions at the time of weaning. in the mice at this critical
transition
models,
Many changes time;
occur
two that are
possibly significant relative to the development of colitis are worthy of mention. First, this is the time when bacteria colonize the gut, and the largest numbers and types of bacteria would occur mainly in the cecum and right colon where the lesions are most frequent.‘* Second, there is a retrovirus present in the mother’s milk of this strain, known as mouse mammary tumor virus,15 which encodes a superantigen that interacts with certain T-cell receptor subsets with an initial stimulation and then deletion. Both of these possible explanations for the colitis are under active study. The role that the immune system plays in the colitis of C3H/HeJBir mice is unknown. The C3HlHeJ strain
that
of antibodies
by homologous
C3Hl
in mice.
system
In at least one of these
(IL-2) knockout,
develops milder
recombi-
in the immune
the mice do not
do when kept under conditions.” The IL-
IBD, mainly
limited
colon, when the colony is maintained
pathogen-free
of
with the
Several recent reports indi-
when germ-free but specific pathogen-free
10 knockout
a variety
data indicate
frequency
deletion
the interleukin
LPS
LPS is
in the normal enteric bacterial
strains.”
in colitis
get disease conventional
However,
and interact
of other genes important
can result
obvious
to bacterial
are not seen in the normal
cate that the selective nation
macrophage
that contain
unpublished
present
HeJ or other inbred
right
features.
showed
when
to a solid
a longitudi-
the natural
in the cecum and right period
milk
this substrain,
and pathological
longitudinal
Recent
to various proteins
mice was begun.
to colitis.
mice have a high
flora; such antibodies
pedigrees
to determine
system.
sug-
this possibility,
of these
also has this re-
responsiveness
of bacteria
mice
a new substrain, CSH/HeJBir, with a high incidence of perianal ulceration and diarrhea in the offspring. During the time required
immune
in
system
in certain
also able to stimulate
of C3H/HeJ
of affected
10 C3H/HeJ
but one component
C3H/HeJBir
to have resulted
substrain
be more susceptible
constituents
to
to LPS. It is not immediately
observations
than most strains to a spon-
form of IBD. To examine
gram of selective
C3H/HeJ
background
made them more susceptible
for evalu-
feces or perianal
at a fairly
strains.
the genetic
of the Jackson
laboratory
if they have soft, light-colored
been culled
reactivity
why mice with reduced
are sent to the diagnostic
compared
defects
The C3H/HeJBir
No. 6
their response
of the immune
most prominently
duced
ulceration.
that reduces
LPS.16 This defect is known abnormalities
functions.”
Discussion Laboratory ation
for mice.
bacterial
mutation
Vol. 107,
environment.”
to the
in a specific
Interestingly,
the
IL-10
knockout
develops
colonic
inflammation
at the time of
weaning,
a timing
similar
to that seen in C3H/HeJBir
mice. A reasonable
unifying
hypothesis
is that a dysregu-
lated immune response to the normal bacterial flora may represent a common pathogenesis in all of these mouse models,
including
the C3H/HeJBir
mice.
The high frequency of IBD in Jews and in certain families along with the IO-fold increased risk of ulcerative
colitis
strongly
and
suggests
Crohn’s that
cause.*’ The increased
disease
these
in
family
disorders
incidence
of IBD in offspring
two affected parents raises the possibility be only a limited number of genes leading tibility.*’
members
have a genetic
In light of these findings,
of
that there may to IBD suscep-
an important
ques-
tion to consider is whether any of the susceptibility genes in the human population are homologous to susceptibility genes in C3H/HeJBir mice.** If so, genetic mapping studies using C3H/HeJBir mice may identify homologous regions in human chromosomes where IBD susceptibility genes are located. High-resolution mapping in mice is now possible and could identify candidate genes to investigate
in human
patients.
In summary, a new substrain of mice has been established with a heritable form of colitis. The disease has a natural history and pathology that are different from other described mouse models.’ Because of the abundance of reagents for immunologic and genetic studies in mice,
IBD IN CBH/HeJBir
December 1994
this model provides new avenues of investigation
into
IBD, particularly investigation of the complex interplay between genetic, immune, and environmental components that are involved in colitis induction and progression.
References I. Kim H-S, Berstad A. Experimental colitis in animal models. Stand J Gastroenterol 1992; 27:529-537. 2. Chalifoux LV, Bronson RT. Colonic adenocarcinoma associated with chronic colitis in cotton top marmosets, Saguinus oedipus. Gastroenterology 1981; 80:942-946. 3. Madara JL, Podolsky DK, King NW, Sehgal PK, Moore R, Winter HS. Characterization of spontaneous colitis in cotton-top tama rins (Saguinus Oedipus) and its response to sulfasalazine. Gastroenterology 1985;88:13-19. 4. Sundberg JP. Conceptual evaluation of inbred laboratory mouse models as tools for the study of diseases in other species. Lab Anim 1992;22:48-51. 5. Moolenbeek C, Ruitenberg El. The “Swiss roll”: a simple technique for histologic studies of the rodent intestine. Lab Anim 1981; l5:57-59. 6. Popesko P, Rajtova V, Horak J. A colour atlas of the anatomy of small laboratory animals. 2. Rat, mouse, golden hamster. London: Wolfe, 1992:146. 7. Sundberg JP, Cordy WR, King LE Jr. Alopecia areata in aging C3H/HeJ mice. J Invest Dermatol 1994;102:847-856. 8. Bedigian HG. Animal health and genetic quality control report. Bar Harbor, ME: Jackson Laboratory, 1993:1-2. 9. Green MC. Catalog of mutant genes and polymorphic loci. In: Lyons MC, Searle AG, eds. Genetic variants and strains of the laboratory mouse. 2nd ed. Oxford, England: Oxford University Press, 1989:12-403. 10. Shiroshi T, Sagai T, Moriwaki K. A simplified micro-method for cyto-toxicity testing using a flat-type titration plate for the detection of H-2 antigens. Microbial lmmunol 1981;25:1327-1334. 11. Vetvicka V. C3H/HeJ mice. In: Rihova BR, Vetvicka V, eds. Immunological disorders in mice. Boca Raton, FL: CRC, 1991:153171. 12. Leiter EH, Beamer WG, Shultz LD. The effect of immunosuppression on streptozotocin-induced diabetes in C57BL/KsJ mice. Diabetes 1983; 32:148-155.
MICE
1735
13. Dietrich W, Katz H, Lincoln SE, Shin H-S, Friedman J, Dracopoli NC, Lander ES. A genetic map of the mouse suitable for typing intraspecific crosses. Genetics 1992; 131:423-447. 14. Schaedler RW, Dubos R, Costello R. The development of the bacterial flora in the gastrointestinal tract of mice. J Exp Med 1965; 122:59-66. 15. Gross L. Oncogenic viruses. 2nd ed. Oxford, England: Pergamon Press, 1970:238-280. 16. Glode LM, Rosenstreich DL. Genetic control of B cell activation by bacterial lipopolysaccharide is mediated by multiple distinct genes or alleles. J lmmunol 1976;117:2061-2066. 17. Brandwein S, McCabe RP, Dadrat A, Ridwan BU, Birkenmeier EH, Sundberg JP, Elson CO. Immunologic reactivity of colitic C3H/ HeJBir mice to enteric bacteria (abstr). Gastroenterology 1994; 106:A656. 18. Sadlack B, Merz H, Schorle H, Schimpl A, Feller AC, Horak I. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 1993;75:253-261. 19. Kuhn R, Lohler J, Rennick D, Rajewsky K, Muller W. InterleukinIO-deficient mice develop chronic enterocolitis. Cell 1993;75:263-274. 20. Orholm M, Mukholm P, Langholz E, Nielsen OH, Sorensen TIA, Binder V. Familial occurrence of inflammatory bowel disease. N Engl J Med 1991;324:84-88. 21. Yang H, Shohat T, Rotter JI. The genetics of inflammatory bowel disease. In: MacDermott RP, Stenson WF, eds. Curr Top in Gastroenterol Elsevier, 1992:17-51. 22. Rotter JI, Yang H. Resolving the genetics of IBD: the challenge for the ’90s. Prog Infl Bowel Dis 1993; 14:1-7.
Received April 4,1994. Accepted August 4,1994. Address requests for reprints to: John P. Sundberg, D.V.M., Ph.D., The Jackson Laboratory, 600 Main Street, Bar Harbor, Malne 04609 1500. Fax: (207) 2882119 or -5079. Supported by a grant from the Crohn’s and Colitis Foundation of America (to E.H.B., J.P.S.); the National Alopecia Areata Foundation (to J.P.S.); and by grant DK44240 from the National Instltutes of Health (to C.O.E., E.H.B., J.P.S.). The authors thank Aletha Torrey for outstanding technical assistance and her dedication to the development, maintenance, and characterization of the C3H/HeJBlr colony: Colleen M. Vallee for technical assistance with the necropsies; and Beth A. Sundberg for computer graphics.