Impaired gastric ulcer healing in diabetic rats: role of heat shock protein, growth factors, prostaglandins and proinflammatory cytokines.

Gastric mucosa of diabetic rats is highly vulnerable to acute injury, but little is known about the influence of diabetic conditions on the healing of gastric ulcers. In this study, streptozotocin (70 mg/kg injected intraperitoneally) was used to induce diabetes mellitus in rats. Four weeks after streptozotocin injection, gastric ulcers were induced using the acetic acid method, and 10 days later, the healing rate and the gastric blood flow (GBF) were measured by planimetry and hydrogen (H(2))-gas clearance method, respectively. Six major groups of rats with gastric ulcers were used: (1) vehicle (saline); (2) streptozotocin alone; (3) insulin (4 IU/day intraperitoneally); (4) streptozotocin plus insulin; (5) pentoxifylline, an inhibitor of synthesis and release of tumor necrosis factor-alpha (TNF alpha); and (6) aspirin, a non-selective inhibitor of cyclooxygenase-1 (COX-1) and cyclooxygenase-2 (COX-2), and rofecoxib, the highly selective COX-2. In the diabetic rats, a significant delay in ulcer healing ( approximately by 300%), accompanied by a decrease in the gastric mucosal blood flow was observed. The prolongation of the healing in diabetic animals was associated with an increase in gastric mucosal expression and release of TNFalpha, interleukin-1 beta (IL-1 beta), suppression of the vascular endothelial growth factor (VEGF), platelet endothelial cell adhesion molecule-1 (PECAM-1) and the mucosal overexpression of heat shock protein 70 (HSP 70). Administration of insulin reversed the delay in ulcer healing and significantly decreased the expression of IL-1 beta and TNF-alpha, while producing the rise in the expression of VEGF and PECAM. Pentoxifylline, an inhibitor of TNF-alpha, which by itself accelerated ulcer healing in non-diabetic rats, counteracted the increase in the area of gastric ulcer induced by streptozotocin, raised significantly gastric blood flow and suppressed the plasma TNF-alpha levels. Aspirin and rofecoxib, that significantly suppressed the mucosal prostaglandin E(2) generation in ulcer area, delayed significantly the rate of ulcer healing and decreased the GBF at ulcer margin in non-diabetic rats, and these changes were significantly augmented in diabetic animals. We conclude that: (1) Experimental diabetes dramatically impairs ulcer healing, depending upon the increased release of proinflammatory cytokines and the attenuation of angiogenesis that can limit the ulcer healing effects of locally produced HSP 70 and TNF-alpha. (2) Insulin reversed this impairment of ulcer healing in diabetic rats, mainly due to the enhancement of angiogenesis and reduction in expression of cytokines in the ulcer area. (3) Classic non-steroidal anti-inflammatory drugs such as aspirin prolonged ulcer healing under diabetic conditions due to suppression of endogenous prostaglandins and the fall in the microcirculation at the ulcer margin and these effects were mimicked by selective, so called "safe" COX-2 inhibitor, rofecoxib, suggesting that both COX isoforms are important sources of prostaglandins that are essential in the ulcer healing in diabetes.