Mechanisms of curcumin-induced gastroprotection against ethanol-induced gastric mucosal lesions

Abstract

Background: Curcumin, a pleiotropic substance used for centuries in traditional medicine, exhibits antioxidant, anti-inflammatory and antiproliferative efficacy against various tumours, but the role of curcumin in gastroprotection is little studied. We determined the effect of curcumin against gastric haemorrhagic lesions induced by 75% ethanol and alterations in gastric blood flow (GBF) in rats with cyclooxygenase-1 (COX-1) and COX-2 activity inhibited by indomethacin, SC-560 or rofecoxib, inhibited NO-synthase activity, capsaicin denervation and blockade of TRPV1 receptors by capsazepine.

Methods: One hour after ethanol administration, the gastric mucosal lesions were assessed by planimetry, the GBF was examined by H₂ gas clearance, plasma gastrin was determined by radioimmunoassay, and the gastric mucosal mRNA expression of Cdx-2, HIF-1α, HO-1 and SOD 2 was analysed by RT-PCR.

Results: Curcumin, in a dose-dependent manner, reduced ethanol-induced gastric lesions and significantly increased GBF and plasma gastrin levels. Curcumin-induced protection was completely reversed by indomethacin and SC-560, and significantly attenuated by rofecoxib, L-NNA, capsaicin denervation and capsazepine. Curcumin downregulated Cdx-2 and Hif-1α mRNA expression and upregulated HO-1 and SOD 2, and these effects were reversed by L-NNA and further restored by co-treatment of L-NNA with L-arginine.

Conclusions: Curcumin-induced protection against ethanol damage involves endogenous PG, NO, gastrin and CGRP released from sensory nerves due to activation of the vanilloid TRPV1 receptor. This protective effect can be attributed to the inhibition of HIF-1α and Cdx-2 expression and the activation of HO-1 and SOD 2 expression.

Keywords: Afferent sensory fibres; Calcitonin gene related peptide; Curcumin; Gastric blood flow; Gastrin; Nitric oxide; Prostaglandin; Transient potential vanilloid receptor.

Fig. 1

Mean area of ethanol-induced gastric lesions, alterations in gastric blood flow (GBF) and plasma gastrin levels in rats pretreated intragastrically (i.g.) with vehicle (saline) or curcumin applied i.g. in graduated doses from 2.5 mg/kg to 100 mg/kg and, for comparison, with the proton pump inhibitor omeprazole (20 mg/kg i.g). The results are mean ± SEM obtained from seven rats per group. An asterisk indicates a significant change (p < 0.05) relative to the vehicle-control values

Fig. 2

The mean area of ethanol-induced gastric lesions and the alterations in the gastric blood flow (GBF) in rats pretreated intragastrically (i.g.) with curcumin in a dose of 50 mg/kg with or without the concurrent treatment with indomethacin (INDO; 5 mg/kg i.p.), SC-560 (SC; 5 mg/kg i.g.) or rofecoxib (ROFE; 10 mg/kg i.g.) with or without combination with 16,16 dimethyl PGE₂ (5 μg/kg i.g). Results are mean ± SEM from 6 to 8 rats per group. An asterisk indicates a statistically significant (p < 0.05) change relative to the vehicle control values. The asterisk and cross indicate a statistically significant change (p < 0.05) relative to curcumin administered alone. The cross indicates a significant change (p < 0.05) relative to each group treated with COX-1 and COX-2 inhibitor alone

Fig. 3

Representative gross macroscopic and microscopic appearance (using H&E staining) of the gastric mucosa pretreated with (A) vehicle (saline) or (B) curcumin (50 mg/kg i.g.) 30 min before per os instillation of 75% ethanol (1 ml/rat) and curcumin-pretreated gastric mucosa of rats treated with L-NNA (20 mg/kg i.p.) applied alone (C) or combined with l-arginine (200 mg/kg i.g.) (D) and exposed to 75% ethanol. The severe gastric haemorrhagic lesions are clearly visible in the gastric oxyntic mucosa pretreated with vehicle (saline) and exposed to 75% ethanol (A1). The gastric mucosal lesions were markedly reduced in gastric mucosa pretreated with curcumin (50 mg/kg i.g.) and exposed to ethanol (B1). In rats pretreated with L-NNA (20 mg/kg i.p.) and administered curcumin (50 mg/kg i.g.), the protective effect of curcumin against ethanol injury was lost (C1). When l-arginine (200 mg/kg i.g.) was administered with L-NNA, a curcumin-induced reduction in the formation of haemorrhagic gastric lesions was observed (D1). In the vehicle-pretreated rats compromised by ethanol, a severe disruption of gastric mucosa, accompanied by the loss of glandular architecture and leucocyte infiltration, is observed (A2). Of note, in curcumin-pretreated gastric mucosa, relatively mild disruption of gastric mucosa along with mild oedema and partial preservation of glandular structure can be observed in comparison with vehicle-control (B2). The loss of glandular structure and the presence of deep mucosal lesions were observed in gastric mucosa of rats treated with a combination of L-NNA and curcumin (C2). In contrast, the mild disruption of gastric mucosa and a partial restoration of glandular structure along with mild signs of inflammation were observed in gastric mucosa of l-arginine-treated rats administered a combination of L-NNA and curcumin and compromised 30 min later by 75% ethanol (D2)

Fig. 4

The mean area of ethanol lesions and accompanying changes in the GBF in rats pretreated with curcumin (50 mg/kg i.g.), l-arginine (200 mg/kg i.g.) alone or in those pretreated with a combination of L-NNA administered in a dose of 20 mg/kg i.p. and curcumin (50 mg/kg i.g.) with or without l-arginine (200 mg/kg i.g.). The results are mean ± SEM recorded in seven rats per group. The asterisk indicates a significant change (p < 0.05) relative to the respective values recorded in the vehicle-control group. The asterisk and cross indicate a statistically significant change (p < 0.05) compared with the values obtained in animals pretreated with l-arg alone. The cross indicates a statistically significant change (p < 0.05) relative to the values obtained in animals pretreated with curcumin. Double crosses indicate a significant change (p < 0.05) relative to animals treated with a combination of L-NNA and curcumin

Fig. 5

The mean area of ethanol-induced gastric lesions and the changes in gastric blood flow (GBF) in rats with intact sensory nerves or in those with capsaicin denervation pretreated with CGRP (10 μg/kg s.c.) or curcumin (50 mg/kg i.g.) applied alone or administered in combination with CGRP (10 μg/kg s.c.). The results are mean ± SEM from 6 to 8 rats per group. The asterisk indicates a significant change (p < 0.05) relative to the vehicle-control values. The double asterisk indicate a significant change (p < 0.05) relative to the values obtained in animals pretreated with CGRP alone. The asterisk and cross indicate a statistically significant change (p < 0.05) as compared to treatment with curcumin alone. The cross indicates a significant change (p < 0.05) as compared with values obtained in vehicle-control- or curcumin-treated rats with intact sensory nerves. The double crosses indicate a significant change (p < 0.05) compared with capsaicin-denervated rats treated with curcumin

Fig. 6

The area of ethanol-induced gastric damage and the changes in gastric blood flow (GBF) in rats pretreated with curcumin (50 mg/kg i.g.), with or without combination with capsazepine (5 mg/kg i.g.). The results are mean ± SEM recorded in seven rats. The asterisk indicates a significant change (p < 0.05) relative to the vehicle-control values. The cross indicates a significant change (p < 0.05) relative to animals pretreated with curcumin alone

Fig. 7

Densitometry analysis of mRNA expression for Cdx-2, HIF-1α, HO-1 and SOD 2 compared to the expression of mRNA for β-actin in the intact gastric mucosa (line 1), that with vehicle (saline) and curcumin (50 mg/kg i.g.) pretreatment (lines 2 and 3, respectively), and in those administered L-NNA (20 mg/kg i.p.), with or without l-arginine (200 mg/kg i.g.) followed by curcumin (50 mg/kg i.g.) (lines 5 and 6, respectively) and exposed to 75% ethanol. Results are mean ± SEM of three experiments. The asterisk indicates a significant change (p < 0.05) relative to the intact values (graphs 1 and 2) or vehicle-control (graphs 3 and 4). The cross indicates a significant change (p < 0.05) relative to animals pretreated with the vehicle-control (graphs 1 and 2) or with curcumin (50 mg/kg i.g.) (graphs 3 and 4). The asterisk and cross indicate a significant change (p < 0.05) relative to the values obtained in animals pretreated with curcumin (50 mg/kg i.g.) alone (graphs 1 and 2) or in combination with L-NNA (20 mg/kg i.p.; graphs 3 and 4). Double crosses indicate a significant change (p < 0.05) relative to those obtained in animals treated with a combination of L-NNA and curcumin (50 mg/kg i.g.)