INTRODUCTION

It is widely accepted that women of all ages show heightened inflammatory response to infection or sepsis and have significantly lower rates of infection-related mortality than men [1]. Epidemiological studies have also shown that women have a lower prevalence of peptic ulcer disease than men [2–5], while pregnant women or women taking estrogen-containing oral contraceptive pills exhibit a further reduced frequency of duodenal ulcer [6]. Animal studies support that female rats have reduced gastric ulcer formation as compared to male rats and have a significantly faster ulcer-healing rate [7]. Similarly, estrogen reduces the severity of dinitrobenzene sulfonic acid-induced colitis in mice [8]. However, the epidemiological studies related with the association between inflammatory bowel disease (IBD) and estrogen are conflicting. In contrast to ulcer disease, female gender, but not oral contraceptive use, was reported as a risk factor for relapse in ulcerative colitis [9]. On the other hand, hormonal replacement therapy after menopause was shown to be protective of disease activity in women with IBD, making them less likely to experience a flare of their disease [10]. Thus, in addition to numerous studies suggesting a protective role of estrogen in the development of various diseases, including cardiovascular diseases [11–13], cerebrovascular damage [14, 15], and osteoporosis [16, 17], estrogen may have protective effects in the gastrointestinal tract.

Estrogen receptor (ER), which mediates the cellular actions of estrogen, was first cloned from a human breast cancer cell line in 1986 [18]. Ten years later, a second ER was cloned from rat prostate [19], which was named as ERβ, and the first cloned traditional receptor was then named as ERα. It was shown that 17β-estradiol binds equally well to both ERs, and it is the most potent endogenous ligand for ERβ [20]. Several clinically relevant in vivo models showed that selective stimulation of ERα is sufficient to elicit many of the biological responses attributed to estrogen action [21–23]. When ERβ agonists were administered to rats with IBD, symptoms of chronic diarrhea were reversed, and intestinal inflammation was reduced [24, 25], which may be due to a greater abundance of ERβ mRNA expression in the upper gastrointestinal tract of rat [26]. On the other hand, the anti-inflammatory activity is not a unique property to ERβ, since the anti-inflammatory activity of ERα has been established in experimental autoimmune encephalomyelitis models [27, 28]. Thus, based on the aforementioned studies, the aim of the current study was to elucidate the anti-inflammatory effect of ER activation in experimental colitis and ulcer models in rats by using selective ER agonists.

MATERIALS AND METHODS

Animals

Male Wistar albino rats (200–250 g) supplied by the Marmara University (MU) Animal Center (DEHAMER) were housed in an air-conditioned room with 12-h light and dark cycles, where the temperature (22 ± 2 °C) and relative humidity (65–70 %) were kept constant. Rats were fed with standard laboratory chow, with free access to water. All experimental protocols were approved by the MU School of Medicine Animal Care and Use Committee.

Experimental Design

The animals were randomly divided as colitis (n = 40), ulcer (n = 40), and their corresponding non-ulcer (n = 40) and non-colitis (n = 40) control groups. Each group of rats was further assigned to vehicle-treated (olive oil, 1 ml/kg, intramuscularly, i.m.), selective ERα agonist PPT-treated (propylpyrazole-triol; 1 mg/kg, i.m.; Tocris, Istanbul), ERβ agonist DPN-treated (diarylpropionitrile; 1 mg/kg, i.m.; Tocris, Istanbul), non-selective ER agonist 17β estradiol-treated (E2; 1 mg/kg, i.m.; Mesigyna, Schering, Istanbul), or E2 plus non-selective ER antagonist ICI-182780-treated (fulvestrant; 1 mg/kg, i.m.; Sigma, St. Louis, MO, USA) groups consisting of eight rats each.

Induction of Colitis

Under light ether anesthesia, 1 ml of 5 % acetic acid (pH 2.6) solution was instilled through a polyethylene tubing inserted 7 cm from the anal verge for 30 s. Then, 1 ml saline was instilled to dilute the acetic acid and flush the colon. Previously, it was reported that this reproducible inflammatory model, which heals within 2–3 weeks in rats, mimics some clinical features of human IBD [29]. Control animals were handled identically except that 1 ml saline was instilled instead of 5 % acetic acid. After the application of acetic acid, the rats were allowed to recover from anesthesia. They returned to their cages to receive normal pellets and one of the five designated treatments (vehicle, PPT, DPN, E2, or E2 + ICI-182780) for the following 3 days. On the third day, body weight, daily food intake, and fecal output were recorded. Stool volume (milligram per day) and number of fecal pellets (number per day) were measured for a 24-h period. For each animal, average stool weight (milligram per number) was calculated and expressed as “fecal output.”

Then, the rats were decapitated and the distal 8 cm of the colons was opened longitudinally down to their mesenteric borders, cleansed of luminal contents, gently rinsed in saline, and dried on filter paper. The severity of colitis was assessed using macroscopic damage scoring. For macroscopic scoring of colonic damage, the 8-cm colonic segments were scored by using the criteria modified from McCafferty et al. [30]: 0, normal appearance; 1, focal hyperemia, no ulcers; 2, ulceration without hyperemia or bowel wall thickening; 3, ulceration with inflammation at one site; 4, two or more sites of ulceration and inflammation; 5, major site of damage extending >1 cm along the length of the colon; and 6–10, score is increased by 1 for each additional centimeter of damage beyond 2 cm. The scoring of colonic damage was always performed by an observer who was unaware of the treatments received by the rats. After scoring, tissue weights were recorded, corrected for body weight and expressed as tissue WI (gram per 100 g body weight).

Induction of Acetic Acid-Induced Gastric Ulcer

Ulcer was induced by a surgical method [31], which was demonstrated to develop gastric mucosal ulcers that become chronic within 2–3 days [32]. Briefly, rats were fasted overnight, with free access to water, and were anesthetized (100 mg/kg ketamine and 0.75 mg/kg chlorpromazine; intraperitoneally). Then, a midline laparotomy was made to gently exteriorize the stomach. Half a milliliter of acetic acid (80 %, vol/vol) was poured through the barrel of a 3-ml syringe placed on the serosal surface of the stomach in the corpus region (nearly an area of 60 mm2) and allowed to remain in contact for 1 min. Afterwards, the fluid was aspirated off carefully, and the area that remained in contact with acid was gently rinsed with saline. In the control rats, animals were handled identically, but instead of acetic acid, saline was placed on the serosal surface of the stomach. All rats were allowed to recover from anesthesia and returned to their cages, where they received normal pellets and one of the five designated treatments (vehicle, PPT, DPN, E2, or E2 + ICI-182780) for the following 3 days. At the end of the third day, immediately after decapitation, freshly excised stomachs were dissected out, cut along the greater curvature, and the mucosae were rinsed with normal saline for the macroscopical analysis of hemorrhagic lesions in the glandular mucosa. The length (millimeter) of each lesion was measured (three petechia were counted as 1 mm), summed per stomach, and expressed as ulcer index.

Measurement of Plasma TNF-α and IL-6 Levels

Cardiac blood samples were collected and immediately centrifuged at 3,000 × g for 10 min, stored at −80 °C for the determination of serum TNF-α and IL-6 levels. Plasma levels of TNF-α and IL-6 were quantified according to the manufacturer's instructions and guidelines using enzyme-linked immunosorbent assay kits specific for the previously mentioned rat cytokines (Invitrogen, Istanbul, Turkey). These particular assay kits were selected because of their high degree of sensitivity, specificity, inter- and intra-assay precision, and small amount of plasma sample required for conducting the assay.

Measurement of Colonic and Gastric Myeloperoxidase Activities

Myeloperoxidase (MPO) is an enzyme that is found predominantly in the azurophilic granules of polymorphonuclear leukocytes (PMN). Tissue MPO activity is frequently utilized to estimate tissue PMN accumulation in inflamed tissues and correlates significantly with the number of PMN determined histochemically in tissues [33]. MPO activity was measured in tissues in a procedure similar to that documented by Hillegass et al. [34]. Stomach or colon samples were homogenized in 50 mM potassium phosphate buffer (PB, pH 6.0) and centrifuged at 41,400 × g (10 min); pellets were suspended in 50 mM PB containing 0.5 % hexadecyltrimethylammonium bromide. After three freeze and thaw cycles, with sonication between cycles, the samples were centrifuged at 41,400 × g for 10 min. Aliquots (0.3 ml) were added to 2.3 ml of reaction mixture containing 50 mM PB, o-dianisidine, and 20 mM H2O2 solution. One unit of enzyme activity was defined as the amount of MPO present that caused a change in absorbance measured at 460 nm for 3 min. MPO activity was expressed as U/g tissue.

Histological Analysis

For light microscopic investigations, samples from the stomach or distal colon were placed in 10 % formaldehyde and processed routinely for embedding in paraffin. For each animal, four randomly taken tissue sections (5 μm) were stained with hematoxylin and eosin (H&E) and examined under an Olympus BX51 photomicroscope (Tokyo, Japan). All tissue sections were examined by an experienced histologist, who was unaware of the treatments.

For scanning electron microscopic investigations, gastric and colonic samples were fixed for 2 h in a 2.5 % phosphate-buffered glutaraldehyde solution (0.1 M, pH 7.4), postfixed in a 1 % phosphate-buffered osmium tetroxide solution, and passed through an increasing alcohol and amylacetate series. After drying the tissue samples with a Bio-Rad critical point dryer and gold coating with a Bio-Rad SC 502, tissue samples were examined under a JEOL 5200 JSM (Tokyo, Japan) scanning electron microscope (SEM).

Statistics

Statistical analysis was carried out using GraphPad Prism 3.0 (GraphPad Software, San Diego, CA, USA). All data were expressed as means±SEM. Groups of data were compared with an analysis of variance, followed by Tukey's multiple comparison tests. Values of p < 0.05 were regarded as significant.

RESULTS

Colitis induction in the vehicle-treated rats reduced food intake and increased fecal output, which then resulted in a significant reduction in the body weights (p < 0.01–0.001; Table 1). Treatment with either PPN or DPT increased food intake and stopped weight loss of rats with colitis, while diarrhea was abolished (p < 0.001). Similarly, the non-selective E2 reduced fecal output and abolished body weight loss (p < 0.01–0.001). However, the reduction in daily food intake due to colitis as compared to control rats was still observed in both E2- and E2 plus ER antagonist-treated rats (p < 0.05).

Table 1 Food Intake, Body Weight Change and Fecal Output in Colitis Groups Treated with Vehicle, DPN (ERβ Agonist), PPT (ERα Agonist), E2, or E2 + ICI 182780 (ER Antagonist)
Full size table

Induction of either colitis or ulcer increased TNF-α levels (p < 0.001; Fig. 1), which were significantly reduced by all treatments, including the ER-antagonist-treated group (p < 0.001). Similar to TNF-α results, due to colitis or ulcer induction, elevated IL-6 levels (p < 0.001; Fig. 2) were significantly depressed by DPN, PPT, or E2 treatments (p < 0.001), and the depression by E2 was not altered by the ER antagonist.

Fig. 1
figure 1

TNF-α levels in the gastric and colonic tissues of rats treated with vehicle, DPN (ERβ agonist), PPT (ERα agonist), E2, or E2 + ICI 182780 (ER antagonist). **p < 0.01; ***p < 0.001 vehicle vs. control groups. + p < 0.05; ++ p < 0.01; +++ p < 0.001 vehicle vs. ulcer or colitis groups.

Full size image
Fig. 2
figure 2

IL-6 levels in the gastric and colonic tissues of rats treated with vehicle, DPN (ERβ agonist), PPT (ERα agonist), E2, or E2 + ICI 182780 (ER antagonist). ***p < 0.001 vehicle vs. control groups. ++ p < 0.01; +++ p < 0.001 vehicle vs. ulcer or colitis groups.

Full size image

Induction of colitis or ulcer resulted in increased colonic and gastric MPO activities (p < 0.001; Fig. 3), implicating recruitment of neutrophils to these inflamed tissues. On the other hand, when the rats were treated with either selective ER agonists or the non-selective agonist 17β estradiol with or without the antagonist, the MPO activities were suppressed (p < 0.001).

Fig. 3
figure 3

MPO activity in the gastric and colonic tissues of rats treated with vehicle, DPN (ERβ agonist), PPT (ERα agonist), E2, or E2 + ICI 182780 (ER antagonist). ***p < 0.001 vehicle vs. control groups. +++ p < 0.001 vehicle vs. ulcer or colitis groups.

Full size image

Instillation of the colon by acetic acid in the vehicle and PPT-treated rats produced high macroscopic damage scores (p < 0.001), while DPN treatment and E2 treatment with or without ER antagonist significantly reduced colitis-induced damage scores (p < 0.001; Table 2). Light microscopic examination of the colonic tissue in the vehicle-treated colitis groups revealed extensive loss of surface epithelium, severe submucosal edema, vasculitis, and inflammatory cell infiltration (Fig. 4b). Scanning electron microscopic examination demonstrated serious epithelial and crypt damage, where lamina propria remained bare in many sites (Fig. 5b). In the colon of PPT-treated colitis group, regular surface epithelium and crypts were observed with inflammatory cell infiltration and submucosal edema that appeared slightly (Figs. 4c and 5c). In the DPN- (Figs. 4d and 5d) and E2-treated (Figs. 4e and 5e) colitis groups, surface epithelium with focal cell loss, regular crypts, and submucosal edema with slight to moderate inflammatory cell infiltration were observed. Similarly, in the E2 plus ER receptor antagonist-treated colitis group, the focal loss of surface epithelium, regular crypts, and moderate inflammatory cell infiltration were accompanied with submucosal edema (Figs. 4f and 5f).

Table 2 Macroscopic Damage Scores in the Colitis and Ulcer Groups Treated with Vehicle, DPN (ERβ Agonist), PPT (ERα Agonist), E2, or E2 + ICI 182780 (ER Antagonist)
Full size table
Fig. 4
figure 4

Light micrographs illustrating the histological appearances of colonic tissues in different experimental groups. Control group: a regular colon mucosa and submucosa. Vehicle-treated colitis group: b massive loss of surface epithelium (arrow), severe submucosal edema (*), inflammatory cell infiltration (arrow head), and vasculitis (v). PPT-treated colitis group: c quite regular surface epithelium (arrow), mild submucosal edema (*), and inflammatory cell infiltration (arrow head). DPN-treated colitis group: d quite regular surface epithelium (arrow), moderate submucosal edema (*), and inflammatory cell infiltration (arrow head). E2-treated colitis group: e quite regular surface epithelium (arrow) and mild inflammatory cell infiltration (arrow head). E2 + ICI 182780-treated colitis group: f quite regular surface epithelium (arrow), moderate submucosal edema (*), and inflammatory cell infiltration (arrow head). H&E staining, original magnification: ×200, inserts: ×400.

Full size image
Fig. 5
figure 5

Scanning electron micrographs illustrating the topographical appearances of the apical surface of the colon in different experimental groups. Control group: a regular topography of colon surface epithelium. Vehicle-treated colitis group: b severe damage of epithelium (arrow) with degeneration of crypts and denudation of lamina propria. PPT-treated colitis group: c regular surface epithelium (arrow). DPN-treated colitis group: d focal damage of surface epithelium (arrow). E2-treated colitis group: e quite regular surface epithelium (arrow). E2 + ICI 182780-treated colitis group: f focal damage of surface epithelium (arrow).

Full size image

Ulcer induction by acetic acid resulted in a high ulcer index (p < 0.001), which was reduced by either PPN, DPT, E2, or E2 plus ER antagonist treatments (p < 0.01–0.001; Table 2). The histological examination of the vehicle-treated ulcer group by both light and scanning electron microscopy revealed that surface mucus cells of the gastric pits were seriously damaged, glandular cells were swollen and mostly damaged, and bleeding and severe inflammatory cell infiltration were observed (Figs. 6b and 7b). In the PPT-treated ulcer group, regular gastric pits, moderate glandular damage, and inflammatory cell infiltration were observed (Figs. 6c and 7c). Similarly, in the DPN-treated ulcer group, quite regular surface mucus cells, relatively less glandular damage, and moderate inflammatory cell infiltration were observed (Figs. 6d and 7d). Quite regular surface mucous cells and glandular cells with mild inflammatory cell infiltration were observed in the E2-treated group (Figs. 6e and 7e). In the ulcer group treated with E2 and non-selective antagonist, damage to surface mucous and glandular cells, and moderate inflammatory cell infiltration were observed (Figs. 6f and 7f).

Fig. 6
figure 6

Light micrographs illustrating the histological appearances of gastric tissues in different experimental groups. Control group: a regular gastric surface epithelium and glandular cells. Vehicle-treated ulcer group: b severe damage of surface mucous cells with necrosis (arrow) and degenerated glandular cells (g), severe inflammatory cell infiltration (i), and hemorrhage in mucosa (arrow head). PPT-treated ulcer group: c quite regular surface mucous cells with gastric pits, moderate damage in glandular cells (arrow head), and inflammatory cell infiltration (*). DPN-treated ulcer group: d quite regular surface mucous cells with gastric pits, mild damage in glandular cells (arrow head), and inflammatory cell infiltration (*). E2-treated ulcer group: e quite regular surface mucous cells with gastric pits, glandular cells (arrow head), and mild inflammatory cell infiltration (*). E2 + ICI 182780-treated ulcer group: f focal damage of surface mucous cells with gastric pits, mild damage in glandular cells (arrow head), and moderate inflammatory cell infiltration (*). H&E staining, original magnifications: ×200.

Full size image
Fig. 7
figure 7

Scanning electron micrographs illustrating the topographical appearances of the apical surface of the gastric mucosa in different experimental groups. Control group: a regular topography of gastric surface epithelium and gastric pits. Vehicle-treated ulcer group: b severe damage of surface mucous cells (arrow) and denuded lamina propria (lp). PPT-treated ulcer group: c focal damage of surface mucous cells (*). DPN-treated ulcer group: d focal damage of surface mucous cells (*). E2-treated ulcer group: e focal damage of surface mucous cells (*) (right side). E2 + ICI 182780-treated ulcer group: f mild damage of surface mucous cells (*) (right side).

Full size image

DISCUSSION

The current results demonstrate that the selective ERα and ERβ agonists, as well the non-selective estrogen receptor agonist, reduced injury of the tissues induced by colitis or ulcer and abolished exaggerated neutrophil infiltration, while elevations in plasma pro-inflammatory cytokines TNF-α and IL-6 were inhibited, implicating the suppressive effect of ER activation. Diarrhea and loss in body weight following colitis were recovered in rats treated with DPN, PPT, and E2, suggesting the beneficial effects of all agents on clinical symptoms of colitis. However, colitis-reduced food intake was normalized when the animals were treated with either DPN or PPT, but not with E2, indicating a receptor-mediated effect.

Using ERα-selective ligand PPT in several models showed that PPT fully stimulates uterine wet weight increase, prevents bone resorption, ameliorates vasomotor instability, and prevents ovariectomy-induced weight gain, eliciting many of the biological responses attributed to estrogen action [23]. ERβ, co-expressed in most of these target tissues (e.g., bone, brain, and adipose tissue), plays a minor role in mediating action in classical estrogen target tissues [25]. ERβ plays a dominant role in the brain, cardiovascular system, and colon where it is expressed primarily on epithelial cells [35–38]. Even within the same organ system, ERα and ERβ may have complementary, additive, and sometimes antagonistic effects. In animals or cell lines subjected to inflammatory conditions, e.g., oxidative stress, hypoxia, trauma, and hemorrhage, ERβ mRNA expression was increased, whereas expression of ERα was decreased or not changed [39–42], suggesting an inflammation-dependent up-regulation of ERβ relative to ERα. On the other hand, controversial data exist as to whether the protective effects of E2 during myocardial ischemia–reperfusion injury are mediated via ERα, ERβ, or both [43, 44]. It was shown that E2 attenuates vascular injury by inhibiting pro-inflammatory mediator expression and neutrophil chemotaxis through an ERβ-dependent mechanism [45, 46]. Similarly, beneficial effects of estrogen and ERβ agonists in different animal models of ulcer and colitis have been reported [29, 36, 47, 48]. In the present study, both receptor agonists and E2 abolished colitis- or ulcer-induced neutrophil recruitment as well as TNF-α and IL-6 responses effectively, while a receptor-specific anti-inflammatory action was not observed.

ERβ is the predominant ER in the colonic and gastric epithelium [35, 37, 49], suggesting that effects of estrogen in the colon and stomach are mediated by ERβ. Moreover, ERβ exerts anti-inflammatory and anti-tumorigenic effects in colitis-associated colorectal cancer in mice [50]. On the other hand, it was found that overexpression of the ERα may induce apoptosis in a colon cancer line through the increase of TNF-α gene expression [51]. Current results emphasize that the gastrointestinal mucosa possesses both α and β receptors, which appear to increase their expressions in colonic and gastric inflammation. Furthermore, both receptor agonists inhibit inflammation and associated cytokine response via both receptors. Smith et al. [52] have shown that duodenal mucosal bicarbonate secretion was equally stimulated by PPT and DPN, demonstrating a reasonable explanation for the protective effect of both ligands against ulcer. Although significant alterations of ERα and ERβ protein are present in men than in women, supporting the sex-specific differences in the pathogenesis of colonic cancer, it was shown that no significant differences were present in the expressions of ERα and ERβ of the normal colonic mucosa between premenopausal women and men [53]. It was shown that women and young female rats secrete significantly less acid but more duodenal bicarbonate as compared to men and age-matched male rats [52, 54, 55]. Despite this striking sexual dimorphism, which was attributed to serum E2 levels, several studies have demonstrated that estrogens through ERs have physiological regulatory roles in both female and male rats. We have previously shown that estrogen delayed gastric emptying in male rats [56]. Using immunohistochemistry, ERα protein was shown to be extensively expressed in the squamous epithelium of the forestomach and anus in both male and female intact rats, suggesting that estrogens play a modulatory role in the proliferation and differentiation of squamous epithelium via the ERα receptors at these sites [57]. On the other hand, it was shown in ovariectomized rats that administration of physiological concentrations of 17β-estradiol decreased gastric acid output via both ERα and ERβ receptors expressed in both gastric epithelial and neuronal cells [58]. It was shown that ghrelin mRNA expression and production in the stomach is directly controlled by gastric estrogen in male and female rats, which was completely blocked by the ER antagonist ICI-182780 [59]. Based on the aforementioned studies, it appears that ER expression and the gastrointestinal response to estrogens are not gender specific, but the magnitude of the response is dependent on the availability of endogenous E2. Therefore, the current study was carried out on male rats to exclude the impact of the fluctuations in plasma estradiol levels of female rats during estrous cycles. Similar to most of the other gastrointestinal functions, the results suggest that both ERs have modulatory roles in gastric and colonic inflammation.

Previously in a model of burn-induced remote organ injury, we have shown that E2 treatment inhibited burn-induced increase in MPO activity in the liver and lung of male rats [60]. Verdu et al. [8] have reported that supraphysiological doses of 17β-estradiol decreased recruitment of PMNs at early stages of colitis in mice, probably through an ER-mediated mechanism. Physiological concentrations of E2 also significantly reduced the chemotaxis of PMNs via a receptor-mediated system [61]. In support of these findings, we showed that not only 17β-estradiol but also ER agonists blocked infiltration of neutrophils to the inflamed tissues. In accordance with the direct antioxidant properties of estrogens that are reported not to be blocked by ER antagonists [62–64], in the present study, ICI-181782 did not block E2-induced reduction in neutrophil recruitment or pro-inflammatory cytokine response, suggesting that a non-receptor-dependent, direct, and equally potent action of estrogen is evident. However, having the same antioxidant response with both ER agonists shows an additional ER-mediated effect via both subtypes. Recently, in skeletal muscles of men who had a single bout of exercise, E2-supplementation has been shown to attenuate neutrophil infiltration due to direct neutrophil/endothelial interaction, while antioxidant and membrane-stabilizing properties of E2 were not observed [65]. In inflammatory responses to experimental autoimmune diseases (e.g., arthritis, encephalitis), it was reported that treatment with ERα ligand was protective, whereas ERβ ligand treatment was not [66–69]. Furthermore, E2 and ER agonists were shown to modulate the expression of neuroinflammatory genes in the frontal cortex of female rats via both ERα and ERβ receptors [67]. Thus, it appears that distribution and participation of ER subtypes or the presence of non-receptor-dependent effect, as well as the mechanisms of action on the inflammatory response of the injured tissue, may show a great variability among different experimental models or tissues.

Neutrophils eliminate the cellular debris resulting from damage, and both their adhesion and infiltration are primarily regulated by cytokines and cell adhesion molecules [70, 71]. In parallel with the suppressed PMNs in the colonic and gastric tissues in the present study, the elevations in pro-inflammatory IL-6 and TNF-α were also abolished by the ER agonists and E2. Similarly, in the experimental autoimmune encephalitis model, ERα treatment significantly reduced TNF-α and IL-6 production, while neuroinflammation and demyelination were decreased [66]. In the experimental trauma–hemorrhage model, E2 was shown to improve macrophage and lymphocyte functions, while both ER agonists have attenuated Kupffer cell cytokine production [72]. Although there is a predominance of autoimmune disease among women [73–76], it is well recognized that increased circulatory cytokines are associated with immune cell depression and an increased susceptibility to sepsis [77], while lower IL-6 and IL-8 levels were measured in polytraumatized women associated with less multiple organ dysfunction syndrome and sepsis [78]. Taken together with the aforementioned studies, the present results suggest that E2 or ER agonists by directly and/or via receptor activation ameliorated gastric and colonic inflammation associated with the inhibition of TNF-α and IL-6.

Limitations of the current study include the use of ER agonists to delineate ER subtype-specific effects on colitis and ulcer. As it is true for other studies that have conducted their experiments using DPN and PPT, these ligands are not pure ERβ or ERα subtype agonists, but they are still the best available agonists. These agonists enabled us to demonstrate that ER activation ameliorated oxidative injury of acetic acid-induced ulcer and colitis with a concomitant improvement in diarrhea and associated weight loss in colitis. Our results support our contention that both ER-selective ligands may be therapeutically effective in the treatment of colonic or gastric inflammation, but further experimental and clinical studies are required.