Anda di halaman 1dari 9

Immunology Letters 90 (2003) 311

Alteration of intestinal intraepithelial lymphocytes and increased

bacterial translocation in a murine model of cirrhosis
Toshiaki Inamura a , Soichiro Miura b, , Yoshikazu Tsuzuki b , Yuriko Hara a ,
Ryota Hokari b , Toshiko Ogawa a , Ken Teramoto a , Chikako Watanabe a ,
Hisashi Kobayashi a , Hiroshi Nagata a , Hiromasa Ishii a,1

Department of Internal Medicine, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
Second Department of Internal Medicine, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 359-8513, Japan
Received 17 May 2003; received in revised form 18 May 2003; accepted 29 May 2003

Alterations in immunological defense in the gut may lead to the bacterial infection that is frequently associated with cirrhosis of the liver.
The aim of this study was to investigate the changes in distribution and function of intestinal intraepithelial lymphocytes (IELs) in relation
to intestinal barrier dysfunction in experimental cirrhosis. Cirrhosis was induced in mice by treatment with carbon tetrachloride (CCl4 )
intraperitoneally with 5% alcohol in drinking water for 12 weeks. Bacterial translocation was assessed in mesenteric lymph nodes (MLNs)
by the transport of fluorescence-labeled latex beads and by bacteriological cultures. The lymphocyte subpopulation was compared in three
groups (cirrhosis, alcohol alone and controls). IFN- production from isolated IELs was determined by ELISA after stimulation with anti-CD3
or IL-12/IL-18. The total number of IELs significantly increased in the cirrhosis and alcohol groups. There was a preferential increase in
TCR+CD8+ population in the alcohol group, but no change in cirrhosis. Bacterial translocation was negative in the control group, and
a small number was noted in the alcohol group, whereas it was significantly noted in the cirrhosis group. Although the number of IEL was
significantly increased in the cirrhosis group, their proliferative response was decreased, and IFN- production from each IEL was markedly
diminished in either stimulation by anti-CD3 or IL-12/IL-18. These changes were more remarkable in the cirrhosis group than in the alcohol
group. In conclusion, bacterial translocation due to intestinal barrier dysfunction in cirrhosis may be closely correlated with the alteration of
the immune function in IELs.
2003 Elsevier B.V. All rights reserved.
Keywords: Intraepithelial lymphocytes; Interferon-; Liver cirrhosis; Alcohol; Bacterial translocation

1. Introduction
It is known that cirrhosis is associated with altered gastrointestinal functions such as decreased gut motility and
changes in absorptive capacity due to mucosal abnormalities and portal hypertension. These intestinal alterations may
induce failure in the gut barrier to exclude endogenous bacteria and toxins from portal and systemic circulation. These
changes may lead to complications such as spontaneous bacterial peritonitis (SBP), or the systemic inflammatory re Corresponding author. Tel.: +81-429-95-1211x2367;
fax: +81-429-96-5201.
E-mail addresses: (S. Miura), (H. Ishii).
1 Tel: +81-3-3353-1211x2260; fax: +81-3-3356-9654.

0165-2478/$ see front matter 2003 Elsevier B.V. All rights reserved.

sponse syndrome, sepsis, and multiple organ failure [14].

Although key steps in the pathogenesis of SBP have yet to
be elucidated, it is evident that the gut is a major source
of bacteria in SBP [5]. Histopathologic studies have shown
translocations of bacteria, such as Candida albicans and
Escherichia coli, by direct penetration of enterocytes associated with disruptions of the basal membrane [6]. Major
factors that could contribute to bacterial translocation might
be the suppression of immune defense, which might play a
pivotal role with other factors, such as physical disruption
of the mucosal barrier and intestinal overgrowth of bacteria.
In liver cirrhosis, there are a variety of systemic immunological abnormalities, such as polyclonal hypergammaglobulinemia, auto-antibody production, decreased cellular
immunity, and decreased natural killer activity [7,8]. In
addition to these systemic immunological changes, the

T. Inamura et al. / Immunology Letters 90 (2003) 311

alteration of local immunity in the intestinal mucosa could

be postulated [9,10], because gut-associated lymphoid tissue, as an independent immune organ, provides indispensable immunologic protection against resident microbial flora
and infectious pathogens [11,12]. In villous mucosa, lamina
propria T and B lymphocytes compose the non-aggregated
lymphoid tissue acting as effector cells, which control and
produce IgA. Intestinal intraepithelial lymphocytes (IELs)
have intimate cellular and molecular cross-talk with intestinal epithelial cells [13,14] as do a very large cell population
in the epithelial layer of the small intestine, with phenotypic and functional features distinct from those of cells in
peripheral lymphoid tissues. Because the IEL is the layer
of the gastrointestinal immune system closest to the intestinal lumen and a rich source of cytokines, we hypothesized
that the IEL changes during the development of cirrhosis.
However, there has been a paucity of data relating to the
morphological and functional changes of IELs in the intestinal mucosa of liver cirrhosis. Recently, Alpini et al.
established a murine model for cirrhosis by administering
carbon tetrachloride (CCl4 ) and ethyl alcohol [15]. As this
is a reproducible model of cirrhosis in mice, we used this
model to look at the effect of experimental cirrhosis on the
intestinal barrier function against bacteria and the intestinal
immunity focusing especially on the phenotypic composition of the IEL and the altered cytokine production within it.
We demonstrated here that the decreased immune function
of IELs could be mutually related to bacterial translocation
in cirrhosis.
2. Materials and methods
2.1. Induction of cirrhosis in mice (animal model)
Male C3H mice (2025 g) were purchased from Charles
River (Tokyo, Japan), maintained in a temperature-controlled
environment with a 12 h:12 h lightdark cycle and fed standard chow ad libitum. To induce cirrhosis, the mice were
injected intraperitoneally (IP) twice weekly with 0.1 ml of
25% CCl4 in olive oil and they consumed 5% alcohol (by
volume) in their drinking water for 12 weeks (cirrhosis
group) according to the procedure detailed by Alpini et al.
[15]. Another group of mice received only olive oil IP and
5% alcohol in drinking water (alcohol group). The control
groups were treated with material in the same manner but
without CCl4 or alcohol. Body weight and food intake were
monitored daily. These animals were used for the experiments that followed 12 weeks after being exposed to CCl4
and alcohol. All procedures carried out on the animals
were approved by the Animal Research Committee of Keio
Universitys School of Medicine.
2.2. Histologic evaluation and immunohistochemistry
Specimens for light microscopy were taken from the liver
and the small intestine. A few blocks from each liver were

placed in 10% formalin, subsequently embedded in paraffin and stained with hematoxylineosin or Azan stain. The
small intestine specimen was taken from a location 5 cm toward the anal side from the pylorus (jejunum) and from a
location 5 cm toward the oral side from the ileocecal valve
(ileum). To evaluate the morphologic changes, each sample was fixed with 10% formalin, and stained with a hematoxylin and eosin solution. For light microscopic analysis, 10
well-oriented crypt-villus units were examined per slide under a microscope. Another part of the removed intestine was
fixed for 12 h at 4 C in periodatelysineparaformaldehyde
(PLP). Subsequently, these specimens were washed and dehydrated for 12 h with PBS containing 10, 15, and 20%
sucrose. After fixation, they were embedded in Tissue-Tek
O.C.T. compound (Sakura Fineteck Inc., USA) and frozen in
liquid nitrogen. Cryostat sections of frozen tissue were cut
at 7 m. Immunohistochemistry was done with the labeled
streptavidin biotin technique (LSAB). Primary antibodies
used in the immunostaining were monoclonal antibodies that
had reacted to CD4 (RM4-5, rat IgG2a PharMingen, San
Diego, CA), and CD8 (53-6.7, rat IgG2a PharMingen, San
Diego, CA). The tissue sections were incubated with adequately diluted primary antibodies overnight at 4 C, and
treated with subclass- and host-matched biotinylated antibodies for 1 h at room temperature. They were visualized by
streptavidinFITC. Rinsing with PBS containing 1% bovine
serum albumin was performed after each step. A cover slip
was applied using glycerol jelly. These sections were observed under an fluorescence microscope (BX60 Olympus,
Tokyo). The infiltrated cells were expressed as the number of CD4 and CD8 positive cells per mm in muscularis
2.3. Measurement of translocation of inert particles
and bacteria
To investigate the translocation of inert particles
from the gut lumen, groups of mice were given 1 m
fluorescein-labeled latex beads (Polysciences, Warrington,
PA) in drinking water (1 108 /ml) for 7 days before they
were to be sacrificed. After the mice were sacrificed, mesenteric lymph nodes (MLNs) were removed, fixed in O.C.T.
compound and frozen, then sectioned at a thickness of
10 m. The latex beads in sections of MLNs were observed
through a fluorescent microscope. Six sections per MLN
were examined, and the mean number of beads per section
was calculated. The control mice were also given beads
with the same method for 7 days, and the mean number of
beads per section of MLNs was calculated.
To measure bacterial translocation, organs were harvested aseptically and cultured for bacteria as previously
described by Spaecth et al. [16]. Briefly, the MLNs, spleen
and liver were removed, weighed, and homogenized in
sterile glass tissue grinders. An equal volume of tissue
homogenate (10 l) from various experimental groups was
cultured separately on blood agar plates to culture totally

T. Inamura et al. / Immunology Letters 90 (2003) 311

aerobic bacteria and on McConkey agar plates to culture

gram-negative enteric bacilli. The agar plates were incubated for 24 and 48 h at 37 C. Bacterial colony-forming
units (CFUs) were counted. If the agar plates did not show
any bacterial growth up to 48 h, the organ was considered
negative for the presence of bacteria.
2.4. Isolation and flow cytometry of IELs
Intraepithelial lymphocytes were isolated by using the
modified procedures described previously by Ishikawa et al.
[17]. Briefly, the inverted intestine was cut into four segments and the segments were transferred to a 50 ml conical
tube containing 45 ml of 5% fetal calf serum in Ca2+ - and
Mg2+ -free Hanks balanced salt solution (HBSS; Gibco Laboratories, Grand Island, NY). The tube was shaken horizontally in an orbital shaker for 45 min at 37 C. Cell suspensions were collected and passed through a glasswool column to remove cell debris and adhering cells. The cells were
then suspended in 30% (w/v) Percoll (Pharmacia Biotech,
Uppsala, Sweden) and centrifuged for 20 min at 1800 g;
those at the bottom of the solution were subjected to Percoll
discontinuous-gradient centrifugation and IELs were recovered at the interface of 44 and 70% Percoll. The purity of
IELs was confirmed by flowcytometry using a FACSort machine (Becton-Dickinson immunocytometry systems). The
purity of CD3-positive cells was proven to be at least 96%
and IELs exhibited a strong expression of E-integrin on
their surface. The cells were washed, resuspended in RPMI
(pH 7.4) with 5% fetal calf serum, and stored on ice until
they were used.
To determine subpopulation, antibodies were obtained
from PharMingen, San Diego, CA and these consisted of:
PE-conjugated anti-mouse CD4, CD8 and FITC-conjugated
anti-mouse TCR, TCR. Isotype-matched, irrelevant
antibodies were used as negative controls. Flow cytometry was done using standard techniques. Acquisition and
analysis were undertaken on a Becton-Dickinson FACSort
(Becton-Dickinson, Mountainview, CA) using a Macintosh Power PC (Apple Computer Inc., Cupertino, CA) and
CellQuest (Becton-Dickinson) software.
2.5. IFN- production assay from IELs
IELs were stimulated in vitro with plate-coated anti-CD3
mAb, or IL-12/IL-18. IELs (4105 cells per well) were cultured in triplicate in microtiter wells coated with anti-CD3
(145-2C11, 10 g/ml, PharMingen) that were kept at 37 C
in an incubator filled with humidified air containing 5% CO2 .
The cells were cultured in 0.2 ml of RPMI medium supplemented with 5% of FCS, 2 mM of glutamine, 100 U/ml of
penicillin, 100 U/ml of streptomycin, and 20 g/ml of gentamycin sulfate. In some experiments, they were stimulated
with IL-12 (100 pg/ml, PharMingen) and IL-18 (500 pg/ml,
MBL) which were added to the cultures. Supernatants were
collected on day 3 and the IFN- in the supernatants was

measured with ELISA kits (TECHNE corporation, Minneapolis, MN).

2.6. Cell proliferation assay of IELs
To measure IEL proliferation, IELs were resuspended
in RPMI supplemented with l-glutamine (2 mM), 2mercaptoethanol (50 M), HEPES (10 mM), gentamycin
(50 mg/ml), and FCS (10%) at a density of 5 106 cells/ml.
One hundred microliters of cell suspensions were added
to the wells of a 96-well plate. The cells were cultured
at 37 C in 5% CO2 in the presence or absence of phytohemagglutinin (PHA) (15 g/ml) (DIFCO, Detroit, MI)
or plate-coated anti-CD3 mAb. After 66 h of culturing,
0.5 Ci of 3 H-thymidine (ICN Biomedicals, Costa Mesa,
CA; 6.7 Ci/mM) were added to each well containing the
cells. After an additional 6 h of incubation, cells were
harvested on glass filter papers (LKB, Gaitherburg, MD).
Radioactivity was determined using a Skatron cell plate
counter (LKB), and data was expressed as cpm.
2.7. Statistical analysis
All results were expressed as means S.E.M. The differences between groups were evaluated by one-way analysis
of variance (ANOVA) and Fishers post hoc test. Statistical
significance was set at P < 0.05.

3. Results
3.1. Histologic and immunohistochemical evaluation
There was a marked increase in collagen deposition in the
CCl4 exposed mouse livers, revealing a cirrhotic appearance
in the paraffin-embedded liver sections stained with Azan.
By 12-week exposure to CCl4 , about 72% of mice survived
and all of them developed cirrhosis (n = 86). Conversely,
in mouse livers with chronic alcohol exposure there was no
significant deposition of collagen. Fig. 1 shows morphometric analysis of intestinal morphology determined by light
microscopy in the jejunum and ileum of the small intestine.
The mucosal architecture in mice exposed to alcohol and/or
CCl4 was not significantly different from that of the controls,
and epithelial shedding was absent. However, there was a
significant decrease in the villus height of the jejunum and
ileum mucosa in both the alcohol and cirrhosis groups. Conversely, there was an increase in the crypt depth compared
with the control animals in both groups. In particular, there
was a significant increase in the crypt depth in the ileum of
the cirrhosis groups.
Fig. 2 compares the number of CD4 and CD8 positive
IELs (A) and LPLs (B) per mm in the muscularis mucosa of
the jejunum and ileum of the small intestine in the control,
alcohol, and cirrhosis groups. As we can see from Fig. 2A,
the number of CD8 positive IELs significantly increased in

T. Inamura et al. / Immunology Letters 90 (2003) 311

Fig. 1. Villus height and crypt depth in two different regions (jejunum and ileum) of the small intestine determined by light microscopy. The mucosal
architecture was compared in three different groups (control, alcohol, and cirrhosis). () P < 0.05. Values are means S.E.M. of six animals.

the alcohol and cirrhosis groups compared with the control

group in the jejunum, and the increase was more significant
in the cirrhosis group. However, in the ileum, only the cirrhosis group had a significant increase in the CD8 IEL population compared with the control animals. There was only
a small number of CD4 positive cells in the IELs, and these
were not altered after alcohol and/or CCl4 treatment. This
significant increase in CD8 cells was also observed in the
LPL populations of the alcohol and cirrhosis groups compared with the control group (Fig. 2B), although it was more

striking in the cirrhosis group than in alcohol group. When

the number of CD4 positive LPLs was determined in the
different groups, there were significant increases in the alcohol and cirrhosis groups compared with the control animals,
and the increase was more significant in the cirrhotic group.
3.2. Flow cytometric analysis of IELs
IELs were isolated from the small intestines of different groups and the yield and changes in subpopulation

T. Inamura et al. / Immunology Letters 90 (2003) 311

Fig. 2. Lymphocyte subpopulations in the intestinal mucosa determined by immunohistochemistry. Data compares the number of CD4 and CD8 positive
IELs (A) and LPLs (B) per mm muscularis mucosa in the jejunum and ileum of the small intestine in alcohol and cirrhosis groups compared with the
control group. () P < 0.05. Values are means S.E.M. of six animals.

were determined by flow cytometry. We obtained a larger

amount of IELs from the intestines of the cirrhosis groups
(2.38 0.50 105 /cm intestine) compared with the alcohol (1.25 0.21 105 /cm intestine) or control groups
(0.79 0.12 105 /cm intestine) (P < 0.05). The results of
flow cytometric analysis are summarized in Table 1. There
was a preferential increase in the TCR+CD8+ population
with a relative decrease in the other subpopulations in the
alcohol group. Despite the larger number of IELs in the
intestinal mucosa of the cirrhosis group, the subpopulation
determined by flow cytometry did not change significantly

Table 1
Flow cytometric analysis on subpopulations of small intestinal (ileal) IELs

TCR+CD4+ (n = 3)
TCR+CD8+ (n = 3)
TCR+CD8+ (n = 3)

group (%)

group (%)

group (%)

9.5 4.0
16.4 2.1
39.6 2.1

6.8 2.4
15.0 4.6
55.7 6.6

10.3 2.2
16.8 0.8
38.1 5.1

Values are expressed as Mean S.E.M.

T. Inamura et al. / Immunology Letters 90 (2003) 311

Table 2
Determination of bacterial translocation
Experimental groups

Liver (colony-forming
units (CFUs) in samples)

Spleen (colony-forming
units (CFUs) in samples)

Mesenteric lymph nodes (colonyforming units (CFUs) in samples)

Control group (n = 6)
Alcohol group (n = 6)
Cirrhosis group (n = 6)

5.4 0.8

5.2 1.0

3.0 0.6
31.0 5.0

Values are expressed as mean S.E.M. None: not detected.

compared with the control group. These suggest the possibility that there was an increase in the number of IELs in
the cirrhosis groups, while maintaining the same proportion
of subpopulations as the control groups.
3.3. Measurement of translocation of inert particles and
Fig. 3A is a representative micrograph of the translocation of fluorescein-labeled latex beads taken up in the MLNs
of the cirrhosis groups (upper panel) and controls (lower
panel). As we can see from Fig. 3B there were few particles
taken up in the MLNs of the control groups, and a small
number of latex beads were translocated to the MLNs in
the alcohol groups. However, a large number of latex beads
were found to be transferred in the MLNs in the cirrhotic
group. Table 2 compares the detection of bacterial translocation in the splanchnic organs (MLNs, spleen and liver) in
the three groups. All cultures were negative in the control
groups, and there were no colony forming units (CFUs) in
any samples. However, a small number of CFUs were noted
in the MLNs of the alcohol groups, although these numbers
were small. There was a significant number of CFUs noted
in all cultures in the cirrhotic group, with a remarkable increase in the MLNs.
3.4. Cell proliferation assay of IELs
To determine whether the increased number of IELs in the
small intestine of the cirrhosis group was due to increased
proliferation, we evaluated the mitogen (PHA)-stimulated or
TCR-mediated (anti-CD3) IEL proliferative responses of the
three groups. As we can see from Fig. 4A, the IELs obtained
from the control groups responded to PHA stimulation at
levels different from the alcohol groups. Furthermore, the
IELs obtained from the cirrhosis groups did not proliferate as
much as the controls. Similarly, the proliferative responses
to anti-CD3 were significantly suppressed in the cirrhosis
and alcohol groups (Fig. 4B).
3.5. IFN- production from IELs stimulated with anti-CD3
or with combination of IL-12 and IL-18
Fig. 5A has the IFN- concentration in the culture media
after the cells were treated with anti-CD3. IFN- could not
be detected in the culture media after 3 days when the cells
had not been stimulated, but after stimulation with anti-CD3
the culture media of IELs from the control group contained

10 1 ng/ml IFN-. There was no significant change in

anti-CD3 induced IFN- production in the alcohol groups;
in contrast, the IFN- production from IELs was markedly
diminished in the cirrhosis group.
Fig. 5B depicts the IFN- concentration in culture media
after the cells had been treated through a combination of
IL-12 and IL-18. The culture media of the IELs from the
control group contained 252 ng/ml IFN- after stimulation
with IL-12 and IL-18. IL-12 and IL-18 together increased
IFN- concentration to levels greater than those produced by
anti-CD3 treatment, consistent with the previous results [18].
The IFN- production from the IELs was slightly decreased
in the alcohol group. However, it was remarkably suppressed
in the cirrhosis group.
4. Discussion
This study demonstrated that experimental cirrhosis
induced through the administration of chronic doses of
ethanol and CCl4 resulted in a several-fold increase in
the translocation of latex particles and bacteria. We found
that the gastrointestinal tract is affected during cirrhosis
and that there are mucosal abnormalities secondary to portal hypertension [5]. The reduction in the height of villi
found in cirrhotic mice in our study is consistent with
the morphologic changes discovered in other studies [19].
The decreased number of cells in the villus region during
cirrhosis may be related to alterations in actin cytoskeletal protein, which interfere with normal cellular migration
along the crypt-villus axis [19,20]. The exact physiological
and biochemical mechanisms involved in intestinal barrier
dysfunction during cirrhosis have not been clearly understood. However, the structural changes in the epithelial cell
layer may be an important factor in facilitating intestinal
permeability, leading to bacterial translocation. A distended
interenterocyte space together with shorter microvilli has
been observed in the duodenal mucosa of cirrhotic patients
[21], and changes in brush border membrane composition,
which include decreased structural proteins and microvillus enzymes have been reported in cirrhotic rats [2224].
Recently, Ramachandran et al. focused on oxygen-free radicals and showed that cirrhosis results in oxidative stress
in the intestine, and this may affect mitochondrial function
and increased lipid peroxidation in enterocytes [19].
The other factors contributing to inducing barrier dysfunction in the gut, leading to increased bacterial translocation, include altered gut motility and bacterial overgrowth,

T. Inamura et al. / Immunology Letters 90 (2003) 311

Fig. 3. Measurement of translocation of inert particles into mesenteric

lymph nodes 7 days after administration. (A) A representative micrograph
of translocation of fluorescein-labeled latex beads taken up in MLNs of
cirrhosis groups (upper panel) and controls (lower panel). Large number
of latex beads are taken up in MLNs in cirrhotic group, while there are few
particles in control groups. (B) The number of latex beads determined in a
10 m section of MLNs through a fluorescence microscope. Six sections
per MLN were examined, and the mean number of beads per section was
calculated. () P < 0.05. Values are means S.E.M. of six animals.

which have been discovered in patients with liver cirrhosis or experimental cirrhosis [3,25,26]. However, another
intriguing and possible etiologic factor is inappropriate immune activity, especially at the level of the intestinal mucosa, since the intestinal mucosal immune system functions
as the first line of defense against enteric bacteria. The lymphoid tissues associated with the intestine are continuously
exposed to antigens in the lumen of the gut. Under healthy
conditions, a few indigenous bacteria are known to continuously translocate to the mesenteric lymph node, but because
immune defenses are intact, these bacteria do not survive.
However, it is known that injuries, caused by such factors
as alcohol and burns [27,28], as well as specific conditions,
such as obstructive jaundice [29,30] or total parenteral nutrition [31,32], frequently disrupt effective mucosal defense
to intraluminal microorganisms. When these conditions occur, there is a suppression of intestinal immune response
against injury [27] or heightened immunologic activity and
activation within intramucosal lymphoid tissue [29]. In the
present study, we found that there was a marked impairment
in IFN- production in each IEL from cirrhotic intestines
compared with the controls or alcohol groups. The exact
mechanism for reduced IFN- production from IELs is not
known in cirrhostic mice, but this may not be due to the direct
toxic effect of CCl4 because our preliminary experiments
revealed that short term injection (2 weeks) of CCl4 did not
significantly affect IFN- production. Since it has been reported that IL-12/IL-18 induction is not overridden by the
TCR pathway [33], one can speculate that by cirrhotic condition the downstream common transcriptional pathway was
inhibited. Anyhow, such decreased IFN- production during
cirrhosis may affect the phagocytic ability of macrophages
and phagocytic cells and thus allow bacteria to multiply and
transfer to extraintestinal sites. IFN- produced by intestinal T cell helps in the resolution of Yersinia enterocolitia
infection [34]. In another study [35], IFN- was shown to
confer protection against the S. typhimurium invasion of epithelial cells and fibroblasts. We also found that the proliferating response of IELs to mitogenic stimuli was significantly
suppressed, suggesting an impaired response of IELs in the
intestinal mucosa of cirrhotic condition. This finding is in
accordance with the suppression in T cell mitogenesis that
has been recorded after exposure to alcohol [28]. In our previous study using cirrhotic rats, induced by CCl4 , we also
demonstrated a remarkable suppression in the appearance of
anti-cholera-toxin containing cells in the intestinal mucosa
and mesenteric lymph nodes [10]. Such combinatorial decreases in IFN- production, mitogenic ability and the production of antigen-specific antibodies will potentially result
in disrupting the hosts complex defense network, resulting
in immune dysfunctions against microorganisms.
Despite the significant inhibition of proliferating responses in lymphocytes, we found that there was a significant increase in the number of lymphocytes both in the
intraepithelial region as well as in the lamina proprial
regions in cirrhotic animals, without a change in the


T. Inamura et al. / Immunology Letters 90 (2003) 311

Fig. 4. Proliferative response of IELs from small intestine induced by PHA (A) and anti-CD3 stimulation (B). Sixty-six hours after starting to culture
IELs in PHA (5 g/ml)-treated or anti-CD3 coated wells, the wells were pulsed with 0.5 Ci of 3 H-thymidine and 6 h later radioactivity of harvested
cells was evaluated. IELs from alcohol and cirrhosis groups are compared with control group. () P < 0.05 compared with controls. Mean S.E.M. of
four experiments.

Fig. 5. IFN- concentrations determined by ELISA in the culture media after IELs were stimulated with anti-CD3 (A) or combination of IL-12/IL-18 (B).
IELs from alcohol and cirrhosis groups are compared with control group after stimulation with plate-coated anti-CD3 (10 g/ml), or IL-12 (100 pg/ml)
and IL-18 (500 pg/ml) for 3 days. () P < 0.05. Mean S.E.M. of four experiments.

T. Inamura et al. / Immunology Letters 90 (2003) 311

subpopulations of mucosal lymphocytes. The exact reason

for the increased number of mucosal lymphocytes is not
known, but we speculate that there might be an increased
migration of these lymphocytes toward the intestinal epithelium in cirrhosis. We previously demonstrated in situ that
IELs are able to selectively adhere to the villus microvessels of the small intestine [36]. It is therefore reasonable
to assume that the accumulation of lymphocytes in the
intestinal epithelium occurs in cirrhosis to compensate for
the immunological dysfunction against enhanced intraluminal stress, although functional abilities like the cytokine
production of each IEL is significantly inhibited.
However, we also noted that there was a slight but constant
increase in the translocation of latex particles, with a significant decrease in cytokine production from IELs even in the
alcohol drinking group. It was also interesting to note that
alcohol imbibition by the mice directly induced a significant
increase in TCR+CD8+ population. These results suggest that chronic exposure to ethanol may elicit a predisposition to more significant intestinal mucosal changes by other
stressor such as CCl4 . Napolitano et al. demonstrated that
although chronic alcohol exposure itself resulted in damage
to both gut villi and the submucosal region [27], they and
other groups also found that mice receiving ethanol intake
prior to burn injuries exhibited significant impairment to intestinal reparative responses, increased bacterial translocation rates, and increased susceptibility to infection [27,28].
These results suggest chronic oral ethanol intake may have
a possible enhancer effect on significant intestinal morphology as well as contributing to immunological alterations as
we discovered in our present study.
To the best of our knowledge, this is the first direct demonstration that experimental liver cirrhosis induces a significant
reduction in IFN- production from IELs, which may be
closely correlated with the disruption of effective mucosal
defense against intraluminal microorganisms in cirrhosis.
This study was supported in part by Grants-in-Aid for
Scientific Research from the Japanese Ministry of Education, Science and Culture of Japan and by grants from Keio
Universitys School of Medicine, and from the National Defense Medical College.

[1] T.P. Almdal, P. Skinhoj, Scand. J. Gastroenterol. 22 (1987) 295300.
[2] M. Andreu, R. Sola, A. Sitges-Serra, C. Alia, M. Gallen, M.C. Vila,
S. Coll, M.I. Oliver, Gastroenterology 104 (1993) 11331138.
[3] C.S. Chang, G.H. Chen, H.C. Lien, H.Z. Yeh, Hepatology 28 (1998)
[4] E.A. Deitch, Ann. Surg. 216 (1992) 117134.
[5] A. Ramachandran, K.A. Balasubramanian, J. Gastroenterol. Hepatol.
16 (2001) 607612.


[6] J.W. Alexander, S.T. Boyce, G.F. Babcock, L. Gianotti, M.D. Peck,
D.L. Dunn, T. Pyles, C.P. Childress, S.K. Ash, Ann. Surg. 212 (1990)
[7] H. Kita, I.R. Mackay, J. Van De Water, M.E. Gershwin, Gastroenterology 120 (2001) 14851501.
[8] T. Nakamura, T. Morizane, T. Watanabe, K. Tsuchimoto, Y. Inagaki,
N. Kumagai, M. Tsuchiya, Int. J. Cancer 32 (1983) 573575.
[9] S. Miura, M. Tsuchiya, in: K. Okuda, J.-P. Benhamou (Eds.), Portal
Hypertension, SpringerVerlag, Tokyo, 1991, pp. 6384.
[10] S. Miura, H. Serizawa, Y. Hamada, S. Tanaka, M. Yoshioka, T. Hibi,
M. Tsuchiya, J. Gastroenterol. Hepatol. 1 (4 Suppl.) (1989) 7174.
[11] M. Salmi, S. Jalkanen, Gastroenterol. Clin. North Am. 20 (1991)
[12] G. Pulverer, H. Lioe Ko, J. Beuth, Scand. J. Gastroenterol.
222 (Suppl.) (1997) 107111.
[13] G.-K. Sim, Adv. Immunol. 58 (1995) 297343.
[14] H. Komano, Y. Fujiura, M. Kawaguchi, S. Matsumoto, Y. Hashimoto,
S. Oana, P. Mombaert, S. Tonegawa, H. Yamamoto, S. Itohara, M.
Nanno, H. Ishikawa, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 6147
[15] G. Alpini, I. Elias, S.S. Glaser, R.E. Rodgers, J.L. Phinizy, W.E.
Robertson, H. Francis, J. Lasater, M. Richards, G.D. LeSage, J.
Hepatol. 27 (1997) 371380.
[16] G. Spaecth, R.D. Berg, R.D. Specian, Surgery 108 (1990) 240247.
[17] H. Ishikawa, Y. Li, A. Abeliovich, S. Yamamoto, S.H.F. Kaufmann,
S. Tonegawa, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 82048208.
[18] Y. Hara, S. Miura, S. Komoto, T. Inamura, S. Koseki, C. Watanabe,
R. Hokari, Y. Tsuzuki, T. Ogino, H. Nagata, S. Hachimura, S.
Kaminogawa, H. Ishii, Immun. Lett. 86 (2003) 139148.
[19] A. Ramachandran, R. Prabhu, S. Thomas, J.B. Reddy, A. Pulimood,
K.A. Balasubramanian, Hepatology 35 (2002) 622629.
[20] T.M. Albers, I. Lomakina, R.P. Moore, Cell Biol. Int. 20 (1996)
[21] J. Such, J.V. Guardiola, J. de Juan, J.A. Casellas, S. Pascual, J.R.
Aparicio, J. Sola-Vera, M. Perez-Mateo, Eur. J. Gastroenterol. Hepatol. 14 (2002) 371376.
[22] B. Manevska, Eksp. Med. Morfol. 15 (1976) 107111 (Bulgarian
with English abstract).
[23] I. Castilla-Cortazar, J. Prieto, E. Urdaneta, M. Pascual, M. Nunez,
E. Zudaire, M. Garcia, J. Quiroga, S. Santidrian, Gastroenterology
113 (1997) 11801187.
[24] J.P. Buts, N. De Keyser, E. Collette, M. Bonsignore, L. Lambotte,
J.F. Desjeux, E.M. Sokal, Pediatr. Res. 40 (1996) 533541.
[25] A.M. Madrid, F. Cumsille, C. Defilippi, Dig. Dis. Sci. 42 (1997)
[26] C. Guarner, B.A. Runyon, S. Young, M. Heck, M.Y. Sheikh, J.
Hepatol. 26 (1997) 13721378.
[27] L.M. Napolitano, M.J. Koruda, K. Zimmerman, K. McCowan, J.
Chang, A.A. Meyer, J. Trauma 38 (1995) 198207.
[28] M.A. Choudhry, N. Fazal, M. Goto, R.L. Gamelli, M.M. Sayeed,
Am. J. Physiol. Gastrointest. Liver Physiol. 282 (2002) G937G947.
[29] F.K.S. Welsh, C.W. Ramsden, K. MacLennan, M.B. Sheridan, G.R.
Barclay, P.J. Guillou, J.V. Reynolds, Ann. Surg. 227 (1998) 205212.
[30] R.W. Parks, C.H. Stuart Cameron, C.D. Gannon, C. Pope, T. Diamond, B.J. Rowlands, J. Pathol. 192 (2000) 526532.
[31] J. Li, K.A. Kudsk, M. Hamidian, B.L. Gocinski, Arch. Surg. 130
(1995) 11641169.
[32] I. Kiristioglu, D.H. Teitelbaum, J. Surg. Res. 79 (1998) 9196.
[33] J. Yang, T.L. Murphy, W. Ouyang, K.M. Murphy, Eur. J. Immunol.
29 (1999) 548555.
[34] V.A. Kempf, E. Bohn, A. Noll, C. Bielfeldt, I.B. Autenrieth, Clin.
Exp. Immunol. 113 (1998) 429437.
[35] S. Bao, K.W. Beagley, M.P. France, J. Shen, A.J. Husband, Immunology 99 (2000) 464472.
[36] S. Koseki, S. Miura, H. Fujimori, R. Hokari, S. Komoto, Y. Hara,
T. Ogino, H. Nagata, M. Goto, S. Hachimura, S. Kaminogawa, H.
Ishii, Int. Immunol. 13 (2001) 11651174.