A novel and effective approach of developing aggressive experimental autoimmune gastritis in neonatal thymectomized BALB/c mouse by polyinosinic:polycytidylic acid

Abstract

Neonatal thymectomy induces autoimmune gastritis (AIG) in 40–70% of BALB/c mice. We presumed that induction of autoimmunity by polyinosinic:polycytidylic acid (poly I:C) might allow development of a more aggressive model of AIG. A group of BALB/c mice were thymectomized on day 3 after birth. Neonatal thymectomized mice were either injected with poly I:C or phosphate-buffered saline (PBS). Non thymectomized neonatal BALB/c mice were injected with only poly I:C. All neonatal thymectomized mice injected with poly I:C developed 3 cardinal features of AIG: (1) moderate to severe degree gastritis (2) presence of autoantibody to H⁺/K⁺ ATPase and (3) loss of parietal cells. However, only 70% of the PBS-treated neonatal thymectomized BALB/c mice developed some, but not, all features of AIG. A mild degree of AIG was seen in 12 of 31 nonthymectomized BALB/c mice administered with only poly I:C. Administration of poly I:C in neonatal thymectomized BALB/c mice in the first and second week appeared to be the most effective for induction of aggressive AIG. The levels of interleukin (IL)-6, IL-12p70, interferon-γ and tumour necrosis factor-α were significantly higher in poly I:C-injected thymectomized mice compared to PBS-injected neonatal thymectomized mice (P < 0·05). The frequencies of CD4⁺CD25⁺ regulatory T cells in the spleen were significantly decreased in neonatal thymectomized mice administered with poly I:C compared to PBS-treated neonatal thymectomized mice (P < 0·01). Taken together, these results suggest that induction of inflammatory cytokines and reduction of regulatory T cells by poly I:C might contribute to the development of an aggressive model of AIG in neonatal thymectomized BALB/c mice.

INTRODUCTION

Autoimmune gastritis (AIG), or chronic gastritis (Type A), is an organ-specific autoimmune disease that is the underlying pathologic cause of a pernicious anaemia [1]. AIG is characterized by lymphocytic infiltrates within the mucosa and submucosa of the body and fundus of the stomach, loss of parietal cells and prevalence of circulating autoantibodies reactive with the gastric parietal cell-associated hydrogen/potassium adenosine triphosphatase (H⁺/K⁺ ATPase) [2,3].

To study the cellular and molecular events underlying the pathogenesis of AIG, several approaches have been taken to construct animal models of AIG. The most characterized AIG model is that induced by thymectomy of BALB/c mice between days 2 and 4 after birth [4–6]. AIG has also been induced in neonatal mouse without thymectomy by immunosuppressive drugs such as cyclosporin [7]. It is also possible to develop AIG in young thymectomized mice by irradiation [8], in adult mouse after adoptive transfer of lymphocytes from mouse with AIG [9], and by immunization with gastric H⁺/K⁺ ATPase [10] or whole stomach lysates [11]. Recently, Toh et al. [12] and Alderuccio et al.[13] summarized the utility and limitation of animal models of AIG. Their data show that although it is possible to induce AIG in neonatal or young or adult BALB/c mice, few studies could show the induction of AIG in all mice. In addition, few animal models of AIG show all the characteristics features of AIG. For example, Toh et al. [12] have shown that only 40% BALB/c mice develop AIG due to neonatal thymectomy, however, the prevalence of autoantibodies and loss of parietal cells in animal model of AIG have not been properly explored in some studies.

The aim of this study is to develop a more aggressive model of AIG so that AIG develops in all mice with all three characteristic features of AIG. Previous studies have shown that AIG could be developed in BALB/c mice by inducing two autoimmune processes like thymectomy and irradiation at the same time [14]. Studies on the mechanism of action of AIG development in neonatal thymectomized mice indicate that like most other organ-specific autoimmune diseases such as experimental autoimmune encephalitis and diabetes in nonobese diabetic mouse, the development of AIG in thymectomized BALB/c mice is mediated by inflammatory cytokines. Martinelli et al. [15] have detected various proinflammatory cytokines like interleukin (IL)-2, IL-3, IL-5, IL-6, IL-10, interferon (IFN)-γ, tumour necrosis factor (TNF)-α and granulocyte-macrophage colony stimulating factor (GM-CSF) in the gastric mucosa from thymectomized AIG mice. Based on these studies, we postulated that administration of various cytokines or inducer of cytokines might facilitate the development of AIG in neonatal thymectomized mice.

Polyinosinic:polycytidylic acid (poly I:C) is a synthetic double-stranded polyribonucleotide. Because of its structural resemblance to double-stranded viral RNA, poly I:C induces a variety of proinflammatory cytokines such as IFN-α, IFN-γ and IL-12 [16]. Indeed, it has been shown that administration of poly I:C accelerates autoimmune pancreatitis in MRL mice, a genetically susceptible model of pancreatitis containing a lymphoproliferative gene [17].

In this study, we injected poly I:C into neonatal thymectomized BALB/c mice at different duration after thymectomy. Administration of poly I:C in thymectomized BALB/c mice led to the development of gastritis and antibody to H⁺/K⁺ ATPase and loss of parietal cells in all thymectomized BALB/c mice within 4 weeks after thymectomy. The mechanism of action underlying the synergistic effect of poly I:C during the development of AIG in neonatal thymectomized BALB/c mice was studied from the production of proinflammatory cytokines and the frequencies of regulatory T cells in these mice.

MATERIALS AND METHODS

Animals

Adult 8–10 week-old BALB/c (H-2^d) mice were purchased from The Charles River Inc (Nagoya, Japan). They were maintained separately at the animal centre of Ehime University School of Medicine, Ehime, Japan, under controlled conditions (22°C, 55% humidity, and 12-h day/night rhythm), and were fed a standard laboratory chow. All animals received adequate care according to good laboratory practice guidelines. The Committee of Animal Experimentation, Ehime University School of Medicine, approved the study.

Development of AIG by neonatal thymectomy and administration of poly I:C

The experimental protocol is shown in Table 1. Neonatal thymectomy was performed on 3-day-old BALB/c mice under cold anaesthesia, as previously described [18]. After neonatal thymectomy, these mice received either poly I:C (Sigma Chemical Co., St Louis, MO, USA) at a dose of 5 mg/kg, intraperitoneal, for 4 weeks (5th and 6th days in the first week, 12th and 13th days in second week, 19th and 20th days in third week and 26th and 27th days in fourth week) or phosphate-buffered saline (PBS). Some nonthymectomized neonatal BALB/c mice were injected with only poly I:C.

Table 1.

Protocol of administration of poly I:C in neonatal thymectomized BALB/c mice. Neonatal thymectomy was performed in all BALB/c mice on day 3 after birth. Neonatal thymectomized mice were administered with poly I:C (5 mg/kg) or PBS, intraperitoneally at different times after thymectomy.

	Administration of poly I:C
	1st week		2nd week		3rd week		4th week
(days after thymectomy)	5th	6th	12th	13th	19th	20th	26th	27th
1st week	•	•	○	○	○	○	○	○
2nd week	○	○	•	•	○	○	○	○
3rd week	○	○	○	○	•	•	○	○
1st and 2nd week	•	•	•	•	○	○	○	○
3rd and 4th week	○	○	○	○	•	•	•	•
1st–4th week	•	•	•	•	•	•	•	•

•Poly I:C;

○PBS

Histopathology

Stomachs were fixed in 10% formalin in PBS and embedded in paraffin. Sections (5 µm) were cut and stained with haematoxilin and eosin. Histologically, the development of AIG was classified into four stages according to the infiltration of mononuclear cells, loss of parietal cells and hypertrophy of the gastric mucosa, as described by Nishio et al. [19] (Fig. 1). In the early stage (stage 1), mononuclear cells expanded around the subglandular region (Fig. 1b). These mononuclear cells expanded upward and there was the loss of parietal cells (stage 2, Fig. 1c). In stage 3, the glands became atrophic and the gastric mucosa became hypertrophic due to expansion of immature cells (Fig. 1d). The histological finding of normal gastric mucosa was regarded as stage 0 (Fig. 1a).

Fig. 1.

Histologically, the severity of gastritis was classified according to the infiltration of mononuclear cells, loss of parietal cells and hypertrophy of the gastric mucosa, as described in Methods. (b) In the early stage (stage 1), mononuclear cells expanded around the subglandular region. (c) These mononuclear cells expanded upward and there was the loss of parietal cells (stage 2). (d) In stage 3, the glands became atrophic and the gastric mucosa became hypertrophic due to expansion of immature cells. (a) The histological finding of normal gastric mucosa was regarded as stage 0.

Assessment of loss of parietal cells in mice with AIG

Frozen sections of gastric tissue specimens were incubated with an optimum dilution of anti-H⁺/K⁺ ATPase monoclonal antibody against the α-(clone: 1H9) and β-(clone: 2B6) subunit (Research Diagnostics Inc., Flanders, NJ, USA) for 30 min at room temperature. After three washes in PBS, bound antibody was detected by incubation with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse-IgG F(ab′)₂ (dilution 1 : 200, ICN Pharmaceuticals, Inc. Aurora, OH, USA) diluted in PBS for 30 min at room temperature. The positive signals were examined by epifluorescence microscope (Axioskop 2 plus, Carl Zeiss Fluorescence Microscope, Gottingen, Germany) with an excitation wavelength of 450 nm. Parietal cells in the gastric mucosa were stained greenish.

Specimens stained without anti-H⁺/K⁺ ATPase monoclonal antibody or stained with an unrelated antibody were used as negative controls.

Detection of autoantibodies against the parietal cells

Indirect immunofluorescence method was employed to detect the prevalence of autoantibodies against H⁺/K⁺ ATPase in the murine sera. Frozen sections of the normal stomachs were incubated with different dilutions of sera (1 : 10, 1 : 100, 1 : 500, and 1 : 1000) in PBS for 30 min at room temperature, followed by three washes with PBS. Bound antibody was detected by incubation with FITC-conjugated goat anti-mouse IgG F(ab′)₂ (1 : 200, ICN Pharmaceutical Inc) for 30 min at room temperature. After further three washes with PBS, sections were examined by epifluorescence microscopy (Axioskop 2 plus, Carl Zeiss Fluorescence Microscope) with an excitation wavelength of 450 nm.

Measurements of cytokines

The levels of IL-6, IL-10, monocyte chemoattractant protein (MCP)-1, IL-12p70, IFN-γ, and TNF-α in the sera from different mouse were determined using Cytometric Bead Array (CBA) kits (BD Biosciences Pharmingen, San Diego, CA, USA), exactly according to the instruction of the manufacturer. Cytokine-bound cytometric profiles were analysed using BD CBA software (BD Biosciences Pharmingen) and the levels of cytokines were expressed as pg/ml.

Flow cytometric analysis

To enumerate the frequencies of CD4⁺CD25⁺ regulatory T cells, single-cell suspensions of splenocytes were washed and resuspended with PBS containing 1% fetal bovine serum and 0·02% sodium azide. After blocking the Fc receptors, spleen cells were incubated with FITC-conjugated anti-CD3 monoclonal antibody (mAb) (clone: 17A2), allophycocyanin (APC)-conjugated anti-CD25 mAb (clone: PC61) and peridinin chlorophyll protein (PerCP)-conjugated anti-CD4 mAb (clone: RM4-5) for 30 min at 4°C in a dark place. Three-colour staining profiles were analysed on FACS Calibur (Becton Dickinson, Franklin Lakes, NJ, USA) after the removal of dead cells and debris by gating.

Isolation of cells

CD4⁺ T cell populations were isolated from splenocytes using a CD4⁺ T cell isolation kit (Miltenyi Biotech Gmbh, Bergisch Gladbach, Germany). Briefly, CD8⁺ T cells, B cells and dendritic cells were depleted from splenocyte suspensions using mAb against CD8a (clone 53–6·7), CD11b (clone M1/70·15·11·5), CD45R (clone RA3–6B2), CD49b (clone DX5) and Ter-119 (clone Ter-119). For isolation of CD4⁺CD25^– T cells, CD25⁺ cells were then depleted from the CD4⁺ T cell enriched fraction using a mAb against CD25 (clone 7D4). To isolate CD4⁺CD25⁺ regulatory T cells, CD4⁺ T cells were stained with FITC-conjugated anti-CD3 mAb, Phycoerythrin-conjugated anti-CD25 mAb and PerCP-conjugated anti-CD4 mAb. CD4⁺CD25⁺ T cells were purified using an EPICS® ALTRA™ HyPerSort (Beckman Coulter, Inc., Fullerton, CA, USA).

In vitro T cell proliferation assay

Proliferation of T cells was assessed as described [20]. Briefly, 5 × 10⁴ CD4⁺CD25^– T cells were incubated with 5 × 10⁴ irradiated syngenic splenocytes in the presence or absence of 1 µg/ml of anti-CD3ɛ (clone 145–2C11) and the presence or absence of 5 × 10⁴ CD4⁺CD25⁺ T cells in 96-well round-bottom tissue culture plate containing 200 µl of RPMI 1640 plus 10% fetal calf serum at 37°C in 5% CO₂. Following 48-h incubation, ³H-thymidine (1 µCi/ml) was added to each well and the plates cultured for further 16 h. Cells were harvested and cell-associated ³H-thymidine was determined and expressed as counts per minute (CPM). Assays were set up in triplicate.

Statistical analysis

The histological stages of gastritis, the numbers of CD4⁺CD25⁺ regulatory T cells and the serum levels of various cytokines were shown as mean ± standard deviation (mean ± SD). Means were compared with the unpaired t-test. In cases of differences as assessed by an F-test, t-test was adjusted for unequal variances (Man-Whitney's U-test). P < 0·05 was considered to be statistically significant. Statistical calculation was performed using Stat view version 5·0 statistical program.

RESULTS

Poly I:C accelerated the development and severity of gastritis in thymectomized BALB/c mice

Seven of 12 PBS-treated neonatal thymectomized BALB/c mice developed gastritis at 4 weeks after thymectomy (Fig. 2a). Administration of only poly I:C in nonthymectomized BALB/c mice induced gastritis in only few mice (Fig. 2b). However, when poly I:C was administered for 8 times within 4 weeks in neonatal thymectomized mice, all mice developed gastritis by 4 weeks (Fig. 2c). The stages of gastritis were also significantly higher in these mice compared to only neonatal thymectomized mice at 4 weeks (P < 0·05).

Fig. 2.

Histological scores of gastritis in (a) PBS-treated neonatal thymectomized BALB/c mice, (b) in neonatal mice administered with only poly I:C and (c) poly I:C-treated neonatal thymectomized mice. The stage of gastritis of individual mouse is shown by circle. Poly I:C was administered twice in a week for 4 consecutive weeks as described in Table 1. The mice were killed at 4, 6 and 8 weeks after thymectomy. At the age of 4 weeks, histological staging of gastritis was significantly higher in poly I:C-treated neonatal thymectomized mice (c) compared to that of PBS-treated neonatal thymectomized mice (a). *P < 0·05 compared with PBS-treated neonatal thymectomized mice.

Development of autoantibody to parietal cells in all thymectomized mice administered with poly I:C

Six of 12 PBS-treated neonatal thymectomized BALB/c mice developed autoantibody to parietal cells in the sera at 4 weeks after thymectomy (Fig. 3a). However, administration of poly I:C for 8 times did not induce this autoantibody in any nonthymectomized BALB/c mice (Fig. 3b). Interestingly, all neonatal thymectomized BALB/c mice (n = 21) administered with poly I:C for 8 times developed antibody to parietal cells at 4 weeks after thymectomy (Fig. 3c). The titres of autoantibody to parietal cell were significantly higher in these mice compared to neonatal thymectomized mice at 4 weeks (P < 0·05).

Fig. 3.

Autoantibodies to parietal cells in BALB/c mice receiving different types of treatment. (a) Halves of the PBS-treated neonatal thymectomized mice dveloped autoantibody to parietal cells in the sera, 4 weeks after thymectomy. (b) Administration of only poly I:C did not induce autoantibody to parietal cells in BALB/c mice. However, when poly I:C was administered, twice in week for 4 consecutive weeks in neonatal thymectomized mice, all mice developed autoantibody to parietal cells in the sera with 4 weeks of thymectomy (c). *P < 0·05 compared with PBS-treated neonatal thymectomized mice.

Loss of parietal cells in mice with experimental AIG

The frequencies of parietal cells in the gastric mucosa were studied by indirect immunofluoresence. Representing staining pattern of parietal cells in gastric mucosa is shown in Fig. 4. Many parietal cells were detected in the gastric mucosa from normal untreated BALB/c mice (Fig. 4a,b). The frequencies of parietal cells reduced slightly at gastric mucosa at 4 weeks in PBS-treated neonatal thymectomized BLAB/c mice (Fig. 4c,d). However, the numbers of parietal cells reduced drastically at the gastric mucosa in poly I:C-treated neonatal thymectomized BALB/c mice (Figs 3, 4e,f).

Fig. 4.

Parietal cells were detected in the gastric mucosa by indirect immunofluorescence using parietal cell-specific monoclonal antibody, as described in the methods. Parietal cells were stained greenish. Consecutive sections of gastric mucosa were also stained by haematoxylin-eosin. (a, b) Many parietal cells were detected in the gastric mucosa of PBS-treated nonthymectomized BALB/c mouse. (c, d) In PBS-treated neonatal thymectomized mice, the frequencies of parietal cells reduced compared to normal BALB/c mice. (e, f) In neonatal thymectomized mice treated with poly I:C, there was a drastic reduction of parietal cells. Magnification, × 10.

The most important period was from 3 days after birth to 2 weeks of age for development of AIG

Although gastritis, development of autoantibodies and loss of parietal cells were seen in thymectomized mice administered with poly I:C at 4 weeks, it was important to evaluate the optimum timing of administration of poly I:C to get the maximum effect. To have insight about this, a group of neonatal thymectomized BALB/c mice were administered with poly I:C for 2 or 4 or 8 times as shown in the protocol in Table 1. As shown in Table 2, the administration of poly I:C in the first or second week induced higher frequencies of parietal cell autoantibody in thymectomized BALB/c mice. Indeed, administration of poly I:C for 8 times in 4 weeks induced autoantibody to parietal cells in 100% neonatal thymectomized BALB/c mice.

Table 2.

Prevalence of autoantibody to parietal cells in neonatal thymectomized BALB/c mice administered with poly I:C

		Prevalence of parietal cell autoantibodies
Neonatal thymectomy	Administration of poly I:C	Number	Parietal cell antibodies + ve	%
+	None	12	6	50·0
+	1st week	12	10	83·3
+	2nd week	14	12	85·7
+	1st and 2nd weeks	11	9	81·8
+	3rd week	10	5	50·0
+	3rd and 4th weeks	10	5	50·0
+	1st- 4th weeks	21	21	100·0
–	1st- 4th weeks	10	0	0

BALB/c mice were thymectomized on day 3 after birth. These thymectomized mice were either administered with poly I:C or PBS. Poly I:C was administered twice in a week, as described in Table. 1. The presence of autoantibody to parietal cell was estimated by indirect immunofluorescence.

Impact of thymectomy and poly I:C administration on the relative proportion and function of CD4⁺CD25⁺ regulatory T cells in the spleen

The numbers of CD4⁺CD25⁺ regulatory T cells among the spleen cells were counted by three-colour flow cytometry at 4 weeks after starting administration of poly I:C or PBS or thymectomy. As shown in Fig. 5, the ratio of CD4⁺CD25⁺ regulatory T cells among total spleen cells decreased significantly in PBS-treated neonatal thymectomized mice and poly I:C administered nonthymectomized mice compared to controls mice (P < 0·01). Administration of poly I:C for 8 times (shown in Table 1) in neonatal thymectomized mice resulted in further decrease of regulatory T cells among whole spleen cells (Fig. 5). To have insights about the ratio of CD4⁺CD25⁺ regulatory T cells among CD4⁺ T cells, we estimated the ratio of CD4⁺ T cells among whole spleen cells in different groups of mice. The ratio of CD4⁺ T cells among whole spleen cells were 16·8 ± 0·3% (n = 6) in PBS-treated nonthymectomized mice. However, it was decreased to 9·3 ± 0·1% (n = 6) in poly I:C-treated nonthymectomized mice. The ratio of CD4⁺ T cells to whole spleen cells were 3·2 ± 0·3% (n = 6) and 2·2 ± 0·3% (n = 6) in PBS-treated neonatal thymectomized mice and poly I:C-treated neonatal thymectomized mice, respectively. The ratio of CD4⁺CD25⁺ regulatory T cells among CD4⁺ T cells was significantly lower in poly I:C-treated thymectomized mice (17·1 ± 4·9%, n = 6) compared to PBS-treated thymectomized mice (28·9 ± 2·9, n = 6) [P < 0·01]. Functional analysis revealed that CD4⁺CD25⁺ regulatory T cells from neonatal thymectomized mice and treated with poly I:C suppressed the proliferation of CD4⁺CD25^– T cells due to stimulation with anti-CD3 (Table 3, b). CD4⁺CD25⁺ regulatory T cells from poly I:C-injected nonthymectomized BALB/c induced massive suppression of proliferation of CD4⁺CD25^– T cells due to stimulation with anti-CD3 (Table 3, c)

Fig. 5.

Low prevalence of CD4⁺CD25⁺ regulatory T cells in the spleen from neonatal thymectomized mice treated with poly I:C. A total of 24 mice were randomly assigned to receive either PBS (n = 6) or only poly I:C (n = 6, twice in a week for 4 weeks) or neonatal thymectomy (n = 6) or poly I:C, twice in a week for 4 weeks after neonatal thymectomy (n = 6). All mice were killed 4 weeks after thymectomy and spleen cells were isolated. The frequencies of regulatory T cells among whole spleen cells were estimated by flow cytometry. Data are shown as mean ± SD. *P < 0·05, **P < 0·01.

Table 3.

Immunosuppressive effect of CD4⁺CD25⁺ regulatory T cells from poly I:C-administered neonatal thymectomized BALB/c mice

Types of cells in culture	Numbers	Incorporation of 3H-thymidine (CPM)
(a) CD4+CD25– T cells plus anti-CD3 plus syngenic splenocytes	3	52890 ± 3316
(b) CD4+CD25– T cells plus anti-CD3 plus syngenic splenocytes plus CD4+CD25+ regulatory T cells from poly I:C-treated neonatal thymectomized mice	3	32941 ± 3901*
(c) CD4+CD25– T cells plus anti-CD3 plus syngenic splenocytes plus CD4+CD25+ regulatory T cells from poly I:C-treated nonthymectomized mice	3	3886 ± 624*

CD4⁺CD25^– T cells were isolated from normal BALB/c mice and CD4⁺CD25⁺ regulatory T cells were isolated from poly I:C-treated neonatal thymectomized BALB/c mice or poly I:C-treated nonthymectomized BALB/c mice. CD4⁺CD25^– T cells (5 × 10⁴ cells) were cultured with irradiated syngenic splenocytes (5 × 10⁴) and anti-CD3 (1 µg/ml) in presence or absence of CD4⁺CD25⁺ regulatory T cells (5 × 10⁴). Data are shown as mean ± SD of three separate experiments.

^*P < 0·05 compared with cultures containing no CD4⁺CD25⁺ regulatory T cells. Culture of only CD4⁺CD25^– T cells exhibited very low levels of proliferation (48·5 ± 7·9 CPM, n = 3).

Up-regulation of various cytokines in the sera due to administration of ploy I:C in thymectomized mice

To evaluate the effect of poly I:C on cytokine secretion, we measured serum levels of various cytokines by CBA method using flow cytometry. The levels of IL-6, MCP-1, TNF-α, IFN-γ and IL-12p70 in the sera in thymectomized mice administered with poly I:C (twice in a week for 4 consecutive weeks) were significantly increased compared to that in PBS-treated neonatal thymectomized mice (Table 4). However, the levels of IL-10 remain almost unchanged in these two groups of mice.

Table 4.

Increased levels of proinflammatory cytokines in neonatal thymectomized mice treated with poly I:C

	n	IL-10 (pg/ml)	IL-6 (pg/ml)	MCP-1 (pg/ml)	IFN-γ (pg/ml)	TNF-α (pg/ml)	IL-12p70 (pg/ml)
(a)Poly I:C-treated thymectomized	3	36·5 ± 12·6	368·3 ± 115·4*	303·5 ± 61·9*	25·8 ± 9·0*	72·5 ± 14·5*	268·6 ± 121·9*
(b)Poly I:C-treated nonthymectomized	3	33·7 ± 9·2	250·1 ± 115·2*	200·6 ± 49·0*	19·8 ± 2·0*	59·1 ± 16·0*	165·5 ± 92·6*
(c)PBS-treated thymectomized	3	17·0 ± 10·2	143·1 ± 86·0	96·7 ± 33·8	10·0 ± 6·2	33·5 ± 14·1	56·4 ± 25·2

BALB/c mice were thymectomized on day 3 after birth (a, c) or remained untreated (b). Neonatal thymectomized mice received either poly I:C (a) or PBS (c), twice in a week for 4 consecutive weeks. Non-thymectomized mice received poly I:C (b). All mice were killed 4 weeks after thymectomy or poly I:C administration. The levels of cytokines in the sera were estimated by CBA method.

^*P < 0·05 compared to PBS-treated thymectomized mice.

DISCUSSION

There are three characteristic features of experimental AIG in BALB/c mice: progressive gastritis, presence of autoantibodies to H⁺/K⁺ ATPase and loss of parietal cells. Administration of poly I:C in neonatal thymectomized mice caused the development of AIG with all three features of experimental AIG in all neonatal thymectomized BALB/c mice. However, only 70% of neonatal thymectomized BALB/c mice treated with PBS developed some features of experimental AIG. Experimental AIG also developed at all 4-week-old BALB/c mice. This newly developed model of experimental AIG would allow further study about the cellular and molecular events related to pathogenesis of AIG and this model might be use for trial of newer drugs.

Various factors became evident regarding the mechanism of AIG development in thymectomized BALB/c mice from this study. First, it became evident that poly I:C by itself is not an inducer of AIG in BALB/c mice, because injection of poly I:C in non thymectomized BALB/c mice did not induce AIG. This indicates that there is a need to break central tolerance by thymectomy in murine model of AIG for optimum functioning of poly I:C. The timing of poly I:C administration is also important. Administration of ploy I:C for only two times in the first week after thymectomy was enough to have experimental AIG in most of the neonatal thymectomized BALB/c mice. However, if poly I:C was administered at third or fourth week only, that did not potentiate the lesions of AIG in neonatal thymectomized mice. Finally, it became evident that the synergistic effect of poly I:C to promote AIG was possibly mediated by various inflammatory cytokines in neonatal thymectomized BALB/c mice. Indeed, it is known that proinflammatory cytokines play a dominant role in the induction of AIG in neonatal thymectomized BALB/c mice. Martinelli et al. have shown the production of a mixture of Th1- and Th2-associated cytokines, including IL-2, IFN-γ, IL-3, IL-5, IL-6, IL-10, and GM-CSF, with the notable absence of IL-4 in the gastric mucosa in thymectomized BALB/c mice [15]. Barrett et al. [21] have shown the abrogation of development of gastritis in neonatal thymectomized mice by neutralization of IFN-γ with anti-IFN-γ antibodies, indicating a major role of IFN-γ in experimental AIG. A role of IL-12 has already been shown in the induction of several autoimmune diseases including chronic experimental colitis [22]. In this study, serum levels of IL-12p70, TNF-α and IFN-γ were increased significantly in nonthymectomized as well as thymectomized mice treated with poly I:C compared to PBS-treated thymectomized mice. However, the levels of IL-10 were not increased due to poly I:C administration in thymectomized or nonthymectomized BALB/c mice. In combination, it appears that administration of poly I:C induces a variety of proinflammatory cytokines, which might be responsible for the development of aggressive AIG in neonatal thymectomized BALB/c mice.

Recently, it has been reported that CD4⁺CD25⁺ regulatory T cells play a key role for induction of AIG [20]. CD4⁺ CD25⁺ regulatory T cells are thymus derived and comprises approximately 5–10% of the mature CD4⁺ T cell population in the thymus and periphery. They appear in the periphery around 3 days after birth, reaching near-adult levels by 2 weeks of age [23]. Day 3 neonatal thymectomy-induced AIG is thought to be due to the lack of regulatory effect of CD4⁺CD25⁺ T cells because administration of these regulatory T cells into neonatal thymectomized mice prevents the development of AIG [24]. The administration of poly I:C in thymectomized mice resulted in reduction of functional CD4⁺CD25⁺ regulatory T cells in the spleen. However, administration of poly I:C in nonthymectomized although resulted in decreased frequencies of CD4⁺CD25⁺ regulatory T cells, these mice did not develop gastritis. This might be due to the functional differences of CD4⁺CD25⁺ regulatory T cells in these two groups of mice (Table 3). Indeed, factors other than regulatory T cells such as inflammatory mucosal milieu might play an important role during the induction of AIG in thymectomized BALB/c mice.

In conclusion, we demonstrated that administration of poly I:C accelerated the onset and intensity of experimental AIG in thymectomized BALB/c mice. Increased production of proinflammatory cytokines and decreased proportion of regulatory T cells might have contributed to the synergistic effect of poly I:C during the induction and propagation of AIG in neonatal thymectomized BALB/c mice. Administration of poly I:C during the initial phase of thymectomy seems to be critical for development of aggressive model of AIG in neonatal thymectomized BALB/c mice.