Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Insulin-dependent diabetes exerts various influences on gastric functions, such as acid secretion and gastric emptying (Feldman et al., 1979; O'Reilly and Long, 1987; Belai et al., 1991). STZ is known to possess diabetogenic properties, causing selective destruction of pancreatic  cells. Recent studies showed the aggravation of gastric mucosal ulcerogenic responses to starvation or stress in STZ-diabetic rats (Piyachatrurawat et al., 1991; Takeuchi et al., 1994a). However, the effect diabetes has on the healing of gastric lesions has not been studied.

Diabetes is characterized by multiple vascular complications including diabetic microangiopathy and is known to impair the events associated with normal wound healing (Enser and Avery, 1984; Goodson and Hunt, 1986). Diabetes results in alterations in mRNA levels of bFGF in various tissues of rats with STZ-induced diabetes, and insulin treatment partially normalized these levels (Karpen et al., 1992). These findings suggest that dysregulation of bFGF may contribute to the development of microangiopathy as well as the impaired wound healing. Indeed, results of one study showed that wound healing is markedly retarded in STZ-diabetic animals and that this retardation is antagonized by treatment with bFGF (Klingbeil et al., 1991). This angiogenic polypeptide is a potent endothelial cell mitogen that also affects the proliferation of other cell types, including fibroblasts, smooth muscle cells and epithelial cells (Folkman and Klagsbrun, 1987). Thus, many of the cell types that are required for tissue replacement at the ulcer site are modulated by bFGF. In fact, several studies have shown using an acid-stable rhbFGF, that bFGF exerts a healing promoting effect on experimentally induced gastric and duodenal ulcers in rats (Folkman et al., 1991; Satoh et al., 1991; Szabo et al., 1994), while another study demonstrated a similar effect on reepithelialization after superficial injury of isolated bullfrog fundic gastric mucosa (Paimela et al., 1993).

The purpose of our study was to examine the influence of diabetes on the healing of acute gastric lesions and the healing promoting effect of bFGF on these lesions under diabetic conditions induced in rats by STZ. We also measured the expression of naturally occurring bFGF in the damaged mucosa of normal and STZ-diabetic animals.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Male Sprague-Dawley rats (230-250 g, Charles River, Shizuoka, Japan) were used. The animals were fed standard rat food and tap water ad libitum. One week after purchase, they were given STZ (70 mg/kg) i.p. and fed normally thereafter. The control animals received an equal volume of saline.

General procedure. The animals were used in the experiments from 1 wk after i.p. injection of STZ. Blood was sampled from the tail vein and BGL were determined by Glucostar-Glucostix (Miles-Sankyo Co., Ltd., Tokyo, Japan). Animals with BGL of less than 300 mg/dl under nonfasting conditions were excluded from study.

Development and healing of gastric lesions. Two weeks after the injection of STZ, the animals were placed in individual cages with raised mesh bottoms and deprived of food but allowed free access to tap water for 18 hr before the experiment. They were given 1 ml of 0.6 N HCl orally. Some of the animals were killed 1 hr after HCl treatment; the rest were fed normally from 1 hr later and killed on various days (1, 3, 5 and 7 days) after HCl treatment. In addition, the effects of the repeated administration of insulin and basic fibroblast growth factor (rhbFGF: acid resistant mutein, CS-23) (Seno et al., 1988) on the healing of these lesions were investigated. Insulin (4 U/rat) was given s.c. once daily for 5 days to ensure better systemic action, while rhbFGF (30~1,000 ng/kg) was given orally twice daily for 5 days for expecting the local action. In all cases, the animals were killed under deep ether anesthesia; 24 hr after the final injection of insulin or 18 hr after the final administration of bFGF. To examine the effects of the above treatment on BGL, blood samples were taken immediately before HCl treatment and at the autopsy. The stomachs were removed, inflated by injecting 8 ml of 2% formalin, immersed in 2% formalin for 10 min to fix the gastric tissue wall, opened along the greater curvature and examined for lesions under a dissecting microscope with a square grid (×10). The area (mm²) of each lesion was measured, summed per stomach and used as a lesion score. The person measuring the lesions did not know the treatment given to the animals. For histological study, a strip of the stomach wall parallel to the limiting ridge across the entire upper corpus was examined under a light microscope. The tissue samples were immersed in 10% formalin, processed for routine light microscopy, sectioned at 5 µm and stained with H&E. Histological injury was assessed using the following criteria according to a previous paper (Takeuchi et al., 1994b); 1, no damage; 2, shallow damage not exceeding 25% of the mucosal depth; 3, moderate damage reaching beyond 25% of the mucosal depth but not exceeding 75% and 4, deep damage reaching beyond 75% of the mucosal depth. The length of sections with type 2, 3 or 4 damage was expressed as a percentage of the total section length and used as a histological index.

Determination of acid secretion. Acid secretion was measured in pylorus-ligated stomachs of both normal and STZ-diabetic rats on various days after HCl treatment. The animals were fasted for 18 hr before the experiment. The abdomen was incised under light ether anesthesia and the pullers ligated. Four hours after the ligation, the animals were killed under deep ether anesthesia and the gastric contents collected. After centrifugation for 15 min at 3000 rpm, each sample was measured for volume and titrated to pH 7.0 against 0.1 N NaOH using an automatic titrator (Hiranuma Comtite-8, Hiranuma, Tokyo, Japan). In some cases, the effects of insulin (4 U/rat) and rhbFGF (1000 ng/kg) on acid secretion was examined in normal and 2 wk-STZ-diabetic rats. Insulin or rhbFGF was given s.c. or i.d., respectively, immediately after the pylorus ligation. Control animals received the vehicle alone.

Determination of bFGF in gastric mucosa. The amount of bFGF in the gastric mucosa was determined in both control and STZ-diabetic rats on various days after HCl treatment. The corpus mucosa (10 mm²) was punched out using a cork borer, rinsed with cold saline and weighed. The tissue was minced in a test tube containing modified 50 mM Tris-HCl buffer (pH 7.6) in a volume 19 times the wet weight of the tissue. The Tris-HCl buffer contained 1.63 M NaCl, 10 mM ethylene-diaminetetraacetic acid (Sigma Chemical Co., St. Louis, MO), 1 mM phenylmethylsulfonyl fluoride (Wako, Osaka, Japan), 1 mM (P-amidinophenyl) methanesulfonyl fluoride HCl (Wako) and 0.05 mM N-ethyl maleimide (Sigma). The tissue was then homogenized with a polytron (T-8; Ika, Yokohama, Japan) on ice, and centrifuged for 2 min at 10,000 × g. The supernatant was stored at 80°C until the assay. The bFGF content of each sample was measured by a sandwich enzyme immunoassay according to the method of Watanabe et al. (1991), using three MAb (MAb 52, MAb 98 and MAb 3H3; Wako) for human bFGF. In brief, MAb 52 or MAb 98 was dissolved in 0.1 M carbonate buffer (pH 9.6) at a concentration of 10 µg/ml. Of a mixture of two solutions 100 µl were added to each well of a 96-well microtiter plate. After overnight incubation at 4°C, the plate was washed with PBS (0.02 M phosphate buffer, pH 7.2, containing 0.15 M NaCl) and each well incubated with 300 µl of buffer A (PBS containing 25% Blocking Ace; Snow Brand Milk Products Co., Tokyo, Japan) overnight at 4°C. After the plate was washed with PBS, 100 µl of each sample or standard rhbFGF diluted in buffer A were added to each well. After a 24-hr incubation at 4°C, the plate was washed with PBS and 100 µl of MAb 3H3 solution in buffer A was added to each well. After a 2-hr incubation at 25°C, the plate was washed and the bound peroxidase activity was measured with O-phenylenediamine. In some cases, the mucosal amount of bFGF was determined in the stomachs of STZ-diabetic rats treated with insulin. In this case, insulin (4 U/rat) was given s.c. once daily for 3 or 5 days after HCl treatment.

Preparation of drugs. Drugs used were streptozotocin (Wako), insulin (Novo, Tokyo, Japan), and rhbFGF (acid-resistant mutein: CS-23, Takeda Ltd., Osaka, Japan). rhbFGF was dissolved in 1% NaHCO₃ buffer (pH 8.5), although other drugs were dissolved in saline. Each agent was prepared immediately before use and was given p.o., i.p. or s.c. in a volume of 0.5 ml/100 g of body weight.

Statistics. Data are presented as means ± S.E. from 4 to 14 rats per group. Statistical analyses were performed using a two-tailed Dunnett's multiple comparison test (Dunnett, 1955); P < .05 were considered significant.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Body Weight and Blood Glucose Concentrations

The initial body weight of animals receiving injections with saline increased over the 21-day experimental period. In contrast, diabetic animals injected with STZ lost 11.9% of body weight within 3 wk (fig. 1). After injection of STZ, BGL under nonfasting conditions reached significantly high levels (331.3 ± 19.4 mg/dl) at 1 wk as compared with basal values (147.1 ± 7.4 mg/dl) and remained significantly elevated for 2 wk thereafter. Control rats showed stable BGL without significant changes (124.7-148.3 mg/dl) during the 3-wk test period.

View larger version (51K): [in this window] [in a new window]   Fig. 1.   Changes in blood glucose levels (BGL) and body weight in rats after i.p. injection of streptozotocin (STZ; 70 mg/kg). Data are presented as the means ± S.E. from eight rats per group. Statistically significant difference at P < .05;  from the values obtained at day 0 in normal group; *from the values obtained at day 0 in STZ group; # from normal group.

Development and Healing of HCl-Induced Gastric Lesions

Intragastric administration of 0.6 N HCl (1 ml) caused severe hemorrhagic lesions in the glandular part of the stomach, mostly in the corpus mucosa, in both control and 2-wk-STZ-diabetic rats. The area of the lesions did not significantly differ between these two groups; the lesion score obtained 1 hr after administration of HCl was 132.6 ± 19.7 mm² in control rats and 102.4 ± 8.8 mm² in STZ-diabetic rats (fig. 2). In the control group, the lesions healed rapidly within 5 days, and the lesion score on day 5 was reduced to about 18.4% (24.3 ± 4.1 mm²) of the initial score observed 1 hr after HCl treatment; the damaged portion was surrounded with granulation tissue and covered with a newly formed thin epithelium. STZ-diabetic conditions impaired the healing of the lesions; the lesion scores on day 3, 5 and 7 after HCl treatment were significantly greater in STZ-diabetic rats than in control rats. Moderate and deep histological injury was observed more frequently in the stomachs of STZ-diabetic rats than in control animals (table 1); in some diabetic rats, the damage on day 5 was still deep into the mucosa without reepithelialization.

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Healing of gastric mucosal lesions induced by 0.6 N HCl in control and STZ-diabetic rats. The animals were given 1 ml of 0.6 N HCl orally, and then fed normally 1 hr later, before being killed on various days after the administration of HCl. This treatment was given to STZ-diabetic rats 2 wk after the injection of STZ. Data are presented as the means ± S.E. from 8 to 14 rats per group. Statistically significant difference at P < .05;  from the values obtained at 1 hr in normal group; * from the values obtained at 1 hr in STZ group; # from normal group.

Acid Secretory Changes After HCl Treatment

Acid secretion in the pylorus-ligated rats was not significantly different between control and STZ-diabetic rats before HCl treatment. A significant reduction in the acid output was observed in control rats for 3 days after HCl treatment; the reduction in acid output was 63.5 and 32.7% on day 1 and day 3, respectively (fig. 3). Similar changes in acid secretion were observed after induction of gastric lesions in STZ-diabetic animals. The acid output was significantly reduced (50.4%) on day 1 but reverted to normal levels on day 3 after HCl treatment. On day 5 in STZ-diabetic rats, the acidity still showed significantly lower values than that observed before HCl treatment, but the volume of gastric juice was significantly increased, resulting in no change in acid output. Although the acidity on day 3 in STZ-diabetic rats was significantly higher than that observed on day 3 in normal rats, there was no significant difference in the acidity on day 5 between these two groups.

View larger version (50K): [in this window] [in a new window]   Fig. 3.   Gastric acid secretory activity in control and STZ-diabetic rats before and at various days after the administration of 0.6 N HCl. Acid secretion was determined in pylorus-ligated rats after 18 hr fasting. Data are presented as the means ± S.E. from eight rats per group. Statistically significant difference at P < .05;  from the values obtained before HCl treatment in normal group; * from the values obtained before HCl treatment in STZ group; # from normal group.

Effects of Insulin and rhbFGF on Healing of Gastric Lesions

Healing of lesions. Because a marked increase of BGL was observed in STZ-treated rats during the experimental period, we examined the effects of insulin on the delayed healing of HCl-induced gastric lesions in STZ-diabetic rats. Injection of saline s.c. once daily for 5 days after HCl treatment did not significantly modify the healing of gastric lesions in STZ-diabetic rats; the lesion score and BGL on day 5 were 59.8 ± 7.4 mm² and 353.4 ± 10.2 mg/dl, respectively (fig. 4). Daily injection of insulin (4 U/rat/day) to the diabetic rats for 5 days after HCl treatment decreased BGL significantly from 346.6 ± 10.2 mg/dl to within normal ranges (121.3 ± 14.1 mg/dl) and significantly enhanced the delayed healing of gastric lesions observed in STZ-diabetic rats, the lesion score being 25.4 ± 6.1 mm². Moderate and deep histological injury was observed less frequently in the stomachs of these rats than in the diabetic rats without insulin treatment (table 1). Body weight loss observed in STZ rats was not significantly affected by the 5-day-insulin treatment, the values being 241.3 ± 6.8g and 244.8 ± 4.3g, respectively, in STZ- and STZ plus insulin-treated rats.

View larger version (43K): [in this window] [in a new window]   Fig. 4.   Effect of insulin on blood glucose levels (BGL) and the healing of HCl-induced gastric mucosal lesions in STZ-diabetic rats. Insulin (4 U/rat) was given s.c. once daily for 5 days after HCl treatment. BGL was measured in each rat before and after insulin treatment. Data are presented as the means ± S.E. from eight rats per group. *Statistically significant difference from the corresponding values in STZ alone, at P < .05. Although the lesion score in STZ alone (59.8 ± 7.4 mm2) in this figure was bigger than that (72.3 ± 10.4 mm2) on day 5 in figure 2, the difference was not significant (P > .2) and may represent usual variations commonly observed in separate experiments conducted on different days.

View larger version (44K): [in this window] [in a new window]   Fig. 5.   Effect of rhbFGF on the healing of HCl-induced gastric mucosal lesions and blood glucose levels (BGL) in STZ-diabetic rats. rhbFGF (CS-23; 30-1000 ng/kg) was given orally twice daily for 5 days after HCl treatment. BGL was measured in each rat at sacrifice. Data are presented as the means ± S.E. from 8 to 10 rats per group. *Statistically significant difference from the corresponding values in STZ alone, at P < .05.

Acid secretion. Because the healing of gastric lesions is modified by luminal acidity, we examined the effects of insulin and rhbFGF on gastric acid secretion in normal and 2-wk-STZ-diabetic rats. A single injection of insulin slightly increased both the volume (17.9%) and acid output (30.3%) in normal rats when compared to controls, but neither of these changes was significant (table 2). Although STZ-diabetic rats secreted slightly less acid when compared to normal rats, similar effects were observed after administration of insulin in the diabetic animals; the volume and acid output increased 25.4 and 18.1%, respectively. However, rhbFGF given at the highest dose (1000 ng/kg) did not have any effect on gastric acid secretion in the pylorus-ligated stomachs of either normal or STZ-diabetic rats.

Mucosal Levels of bFGF During Healing of Gastric Lesions

In control rats, the amount of bFGF expressed in the stomach was about 1000 ng/g tissue. These values were significantly increased on day 3 after induction of damage by 0.6 N HCl but returned to basal levels on day 5 (fig. 6). The mucosal levels of bFGF in STZ-diabetic rats were significantly lower under basal conditions and did not significantly respond to damage by HCl; the values remained within similar ranges (600-700 ng/g tissue) before and after HCl treatment, and were significantly lower on days 3 and 5 as compared to normal groups. Daily injection of insulin restored the lowered expression of bFGF in the injured mucosa of STZ-diabetic rats and significantly increased the levels of bFGF in the gastric mucosa on day 3, the values being 1203 ± 76 ng/g tissue. The levels of bFGF (826 ± 98 ng/g tissue) on day 5 in these rats tended to increase after insulin treatment but did not reach significant levels (638 ± 58 ng/g tissue; P > .05) as compared with those observed in rats treated with STZ alone.

View larger version (38K): [in this window] [in a new window]   Fig. 6.   A, Mucosal expression of bFGF in control and STZ-diabetic rats before and at various days after the administration of 0.6 N HCl. Data are presented as the means ± S.E. from four to six rats per group. Statistically significant difference at P < .05; * from the values observed in the intact stomach (before HCl treatment) of normal group; #from the corresponding values in normal group. B, Effect of insulin on the mucosal bFGF levels in STZ-diabetic rats after HCl treatment. Insulin (4 U/rat) was given s.c. once daily for 5 days after HCl treatment. Data are presented as the means ± S.E. from four to six rats per group. Statistically significant difference at P < .05; * from the corresponding values in normal group; #from STZ alone.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Clinical conditions such as diabetes, obesity, glucocorticoid treatment and chemotherapy, are known to impair the events associated with normal wound healing. In particular, the influence of diabetes on the healing of dermal wound has been studied extensively (Klingbeil et al., 1991). Our study showed for the first time the deleterious influence of diabetes on the healing of gastric mucosal lesions in STZ-diabetic rats and further demonstrated that rhbFGF exerts healing promoting activity in the face of distinct physiological impairments observed in diabetic rats that reduce the rates of healing of acute gastric lesions.

Gastric lesions induced by 0.6 N HCl healed rapidly and were partially reepithelialized within 5 days with granulation at the injury site. We have reported that the healing of these lesions was significantly impaired by functional ablation of capsaicin-sensitive sensory nerves or the repeated administration of nitric oxide synthase inhibitors (Takeuchi et al., 1994b; 1997). As expected, diabetic conditions also had a deleterious influence on the healing of such lesions and significantly delayed the reconstruction process of the injured mucosa when examined both macroscopically and histologically. In this study, all STZ-treated animals developed a persistent hyperglycemia, which was observed 1 wk after STZ injection. The development of gastric lesions induced by HCl was not modified during the 2 wk of diabetic conditions, because no difference was observed in the severity of initial damage in response to HCl between normal and diabetic rats. However, the healing of such lesions was apparently impaired by diabetes and the mucosal defect without reepithelialization was persistently observed even 5 to 7 days after HCl treatment. Because daily injections of insulin to normalize BGL significantly antagonized the impaired healing of gastric lesions in STZ-treated rats, it is assumed that the deleterious influence of diabetes may be insulin-dependent and not caused by nonspecific action of STZ. It remains unknown, however, whether the insulin effect appears through amelioration of hyperglycemic conditions or through some other specific actions of this peptide. Because loss of body weight was similarly observed in STZ-diabetic rats with or without insulin treatment, it is unlikely that malnutrition plays an important role in the impaired healing of gastric lesions observed in such diabetic animals.

Genetically diabetic and obese mice are both hyperglycemic and obese, conditions that have been correlated with impaired healing of the dermal wound in mice and implicated with healing problems in humans (Enser and Avery, 1984; Klinbeil et al., 1991). Significant impairment of healing was observed with the full-thickness wounds in diabetic mice when compared to normal heterozygous controls. Previous studies showed using genetically diabetic mice, that the topical application of bFGF promoted the formation of granulation tissue, the healing of full-thickness incisional wounds and the epithelialization of partial wounds in pigs (McGee et al., 1988; Klingbeil et al., 1991). Similarly, we found that acid stable bFGF (rhbFGF) significantly enhanced the delayed healing of HCl-induced gastric lesions in STZ-diabetic animals. In the diabetic rats treated with rhbFGF, the histological injury on day 5 was restricted to the upper part of the mucosa and not distributed deeper in the mucosa, although in control diabetic rats most of the histological injury was still deep, reaching more than 75% of the mucosal depth. These results are in good agreement with the findings by others who showed, using an acid-stable rhbFGF, that bFGF exerts a healing promoting effect on gastric and duodenal ulcers in rats (Folkman et al., 1991; Satoh et al., 1991; Szabo et al., 1994).

The tissue reconstructing processes are influenced by several factors including luminal acid (Halter et al., 1980; Kamada et al., 1983; Takeuchi and Johnson, 1986). Because acid secretion was decreased in the damaged stomach after administration of 0.6 N HCl and because these changes in acid secretion were hardly affected by diabetes, it seems unlikely that healing was delayed by modulation of luminal acidity. In fact, neither insulin nor rhbFGF significantly affected the acid secretory ability of the stomach in diabetic rats, although both agents significantly promoted the delayed healing in STZ-diabetic rats. However, our study showed that STZ-diabetic rats displayed rapid recovery of acidity (P < .05) on day 3 after injury compared to normal rats. Because rhbFGF did not have any effect on gastric acid secretion in either normal or diabetic rats, it is unlikely that the increased expression of tissue bFGF inhibited parietal cell activity, similar to EGF and TGF- (Goldenring et al., 1993; Tsunoda et al., 1993; Wang et al., 1993). Certainly, if the expression of EGF and TGF- are increased in response to injury, similar to bFGF (Polk et al., 1992), then one may assume that the recovery rate of acidity in the injured mucosa is influenced by local production of these substances.

In general, wound healing undergoes multiple phases, such as synthesis of extracellular matrix materials, angiogenic response and epithelial migration over newly granulated tissue, ultimately resulting in tissue rebuilding (Tarnawski and Halter, 1995). bFGF is a locally acting growth factor and a potent angiogenic polypeptide that stimulates the growth of other cells such as fibroblasts and smooth muscle cells as well as epithelial cells (Folkman and Klagsbrun, 1987). The beneficial effect of rhbFGF on the impaired healing of gastric lesions may be explained by earlier observations that rhbFGF indeed enhances a cascade of cellular events that are part of the mitogenic response produced by this polypeptide (Satoh et al., 1991; Paimela et al., 1993; Szabo et al., 1994). Certainly, because the angiogenic response is closely associated with the neovascularization of the injured tissue, there is a possibility that the delay in the healing under diabetic conditions is partly due to mucosal blood flow insufficiency and the beneficial effect of rhbFGF is attributable to the amelioration of mucosal blood flow response.

It has been reported that bFGF mRNA levels are altered by diabetes in a tissue-specific fashion and that the altered production of bFGF may contribute to the pathophysiological states of various diseases associated with diabetic conditions (Karpen et al., 1992). bFGF mRNA levels increased several fold in eye, heat and brain, but decreased in skeletal muscle. In contrast, insulin-like growth factor mRNA levels were significantly decreased in all tissues except brain (Fagin et al., 1989). In our study, we measured gastric mucosal levels of bFGF before and after damage by HCl and found that bFGF levels were significantly increased on day 3 after HCl treatment. In STZ-diabetic rats, not only was there less expression of bFGF in normal stomachs, but also, bFGF levels in the injured gastric mucosa did not increase after HCl treatment. This suppression of bFGF levels in STZ-diabetic rats was significantly antagonized by daily injection of insulin, suggesting that dysregulation of bFGF production may contribute to the delayed healing of gastric lesions observed under diabetic conditions. It has been shown that bFGF is widely distributed in various cells, especially in vascular endothelial cells, fibroblasts, smooth muscle cells and macrophages (Cordon-Cardo et al., 1990), and even is generated to some extent in the epithelium (Folkman et al., 1991). Therefore, it is possible to assume that bFGF levels in the mucosa are lower in STZ-diabetic rats because of the deeper gastric injury in their stomachs (less epithelium). However, the levels of bFGF were already low in intact stomachs of STZ-diabetic rats without any injury. In addition, the mucosal bFGF levels in normal rats were higher on day 3 after injury than on day 5, although the severity of the gastric lesions was much less on the latter day. On the basis of these findings, it is unlikely that the changes in bFGF levels observed in this study are attributable to the amount of intact epithelium of the stomach. At present, it remains unclear which types of cells are responsible for the increased expression of bFGF in the injured tissue, but they may include both the injured cell itself and inflammatory cells such as macrophages.

In conclusion, our results taken together suggest that diabetic conditions have deleterious influences on the healing of acute gastric lesions in both an insulin- and bFGF-sensitive manner, and that exogenously administered rhbFGF may be effective in enhancing the delayed healing of gastric lesions in diabetic rats, through an acid-independent mechanism (table 3). Further studies with other models of gastric lesions or with other models of diabetes are required to confirm the universality of our findings in diabetic rats and to elucidate the mechanism of delayed healing of gastric lesions as well as the relationship between insulin deficiency and dysregulation of bFGF production under diabetic conditions.