Professor C. H. Cho, Department of Pharmacology, Faculty of Medicine, The University of Hong Kong, 5 Sassoon Road, Hong Kong, China. E-mail: firstname.lastname@example.org
Previous studies have shown that intragastric administration of unfractionated heparin enhances gastric ulcer healing in rats. As the large molecule of heparin may be partially degraded in the upper gastrointestinal tract, it is likely that fragments of heparin, derived from the unfractionated parent compound, are involved in the anti-ulcer action in the stomach. Therefore, it is possible that low molecular weight heparin may have a similar ulcer healing effect.
Male Sprague–Dawley rats with acetic acid-induced gastric ulcers were given a 3.0-kDa low molecular weight heparin (0.6–6.0 mg/kg) intravenously or intragastrically once daily for 4 days. Ulcer healing, mucosal histological changes, angiogenesis and gastric mucus production both in vivo and in vitro were determined. The bleeding time was measured to indicate the anticoagulation activity.
Both intravenous and intragastric low molecular weight heparin dose dependently accelerated gastric ulcer healing, which was accompanied by a significant increase in mucosal regeneration and proliferation, angiogenesis and mucus content in the stomach. The drug also stimulated the mucus production in MKN-28 cells. Drug administration by either route did not alter the bleeding time in rats.
A 3.0-kDa low molecular weight heparin possesses an ulcer healing effect similar to that of unfractionated heparin in the stomach of the rat. This smaller molecular drug is superior to the unfractionated form, does not affect the coagulation activity and may show better absorption in the gastrointestinal tract.
Experimental evidence has shown that intravenous (i.v.) or intragastric (i.g.) administration of unfractionated heparin enhances gastric ulcer healing in rats.1,2 However, anti-ulcer doses of heparin also produce an anticoagulant effect when the drug is given by i.v. injection. Interestingly, oral administration of unfractionated heparin increases ulcer healing without affecting the coagulation function of the animal.2 It is traditionally known that oral administration of the large molecule of heparin is ineffective as an anticoagulant because of poor absorption in the gastrointestinal tract. It is possible that oral unfractionated heparin may be partially degraded in the gastrointestinal tract, such that fragments of unfractionated heparin may contribute to ulcer healing in the stomach when the drug is taken orally.
In order to examine this hypothesis, a 3.0-kDa low molecular weight heparin was used in the present study to investigate its ulcer healing properties and its possible mechanisms of action in the stomach of the rat. Histological changes, such as gastric mucosal regeneration, proliferation and mucosal angiogenesis, as well as gastric mucus production, after low molecular weight heparin treatment were also studied because all of these parameters are important in gastric ulcer healing.3–5
MATERIALS AND METHODS
The use of animals in the present study was approved by the Committee on the Use of Live Animals in Teaching and Research of the University of Hong Kong. Male Sprague–Dawley rats (180–200 g) were reared on a standard laboratory diet (Ralston Purina Co., USA) and given tap water. They were kept in a room in which the temperature (22 ± 1 °C), humidity (65–70%) and day/night cycle (12 h:12 h light:dark) were controlled. The rats were fasted for 24 h, but with free access to water, before being subjected to acetic acid to produce gastric ulcers.
Induction of gastric ulcers
Gastric ulcers were produced by luminal application of an acetic acid (E. Merck, Darmstadt, Germany) solution to rats as described previously,6 with a few modifications including the ulcer size and volume of acetic acid.3 One day after the injection of acetic acid solution, the ulcer size was about 125 mm2, which was derived from the sum of the ulcers on both sides of the gastric wall.
Drug treatment and measurement of ulcer size
A 3.0-kDa low molecular weight heparin, at doses of 0, 0.6, 3.0 or 6.0 mg/kg, was given either intravenously or intragastrically once daily for 4 days, starting 1 day after ulcer induction, to observe the ulcer healing effect. The doses of low molecular weight heparin were equal to 100, 500 and 1000 U/kg of unfractionated heparin, respectively, used in previous studies.1,2 After drug treatment, the rats were killed and their stomachs were removed, opened along the greater curvature and spread on a glass board. The ulcers (mm2) in the anterior and posterior walls were determined and summed in each stomach by a person unaware of the treatments. After measuring the ulcer size, gastric tissues were excised for histological and immunohistochemical analysis.
Gastric tissues were processed and finally stained with haematoxylin and eosin for histological studies. The length of newly regenerated gastric mucosa (mm) from the ulcer edge and the ruptured muscularis mucosae (mm) were determined under a light microscope (Nikon, × 40) as described by Ogihara and Okabe.7
Assessment of epithelial cell proliferation at the ulcer margin
To determine cell proliferation, a single intraperitoneal injection of 100 mg/kg 5-bromo-2′-deoxyuridine (Sigma, St Louis, MO, USA) was administered 1 h before the stomach was removed. Cell proliferation was assessed by immunohistochemical staining with anti- 5-bromo-2′-deoxyuridine antibody (Sigma) as described previously.8,9 The percentage of cells labelled with 5-bromo-2′-deoxyuridine relative to the total number of mucosal cells was counted in both margins of the gastric ulcer crater, using a Leica image processing and analysis system (Q500IW, Leica Image Systems, Cambridge, UK) for each rat, and finally expressed as the labelling index by taking the mean of labelling cells at both margins.
Determination of angiogenesis at the ulcer margin and base
The microvessels at the ulcer margin and base in the granulation tissue of the submucosa were identified by immunohistochemical staining with von Willebrand factor antibody (DAKO, A/S, Denmark).10 The number of microvessels was quantified at the two sides of the ulcer margin and the ulcer base with a Leica image processing and analysis system. The number of microvessels at the ulcer margin was expressed by taking the average of both sides of the ulcer margin.
Assessment of gastric mucus content
Sections were stained with the periodic acid–Schiff technique. They were first immersed in 1% (v/v) periodic acid for 5 min, followed by successive washing with distilled water. They were transferred to Schiff’s reagent for 3 min and then flushed with running tap water for 3 min. Finally, they were counter-stained with Mayer’s haematoxylin. The amount of mucus within the mucosa was assessed by a single-blind method by measuring the relative thickness of the adherent mucus layer11 with a Leica image analyser in five consecutive fields of each side of the ulcer crater. The results were averaged from two sides and expressed as the ratio of the thickness of the mucus layer to the thickness of the total mucosa.
Determination of mucus synthesis in the MNK-28 cell line
MKN-28 is a secretory type of human gastric carcinoma cell line, and was provided by the Japanese Cancer Research Resource Bank (Tokyo, Japan). The cells were routinely cultured in a 25-cm2 flask (Coster Corporation, Cambridge, USA) in RPMI-1640 medium (Sigma), supplemented with 10% heat-inactivated fetal calf serum (Life Technologies, USA), 0.2 g/L streptomycin and 0.1 g/L penicillin G, as well as 2 mM NaHCO3, until the cells reached confluence. When mucus synthesis was determined, the confluent MKN-28 cells were digested with 0.25% 1 mM trypsin–ethylenediaminetetra-acetic acid (Life Technologies) and washed twice with phosphate-buffered saline. The cells (3 × 105 cells/1 mL of medium) were then incubated overnight at 37 °C in a 24-well culture plate (Coster Corporation, USA) under 5% CO2 in air for confluence.
The rate of mucus synthesis was determined by measuring the incorporation of D-[6-3H]-glucosamine (Amersham, USA) into gastric mucus glycoprotein according to the method of Terano et al.12 with modifications. When the cells reached confluence, they were incubated with 1 mL of medium containing [3H]-glucosamine-HCl in the presence of 50–800 μg/mL of low molecular weight heparin or vehicle at 37 °C under 5% CO2 in air for 6 h. Subsequently, medium was aspirated and disposed. The remaining cells were washed twice with Ca2+, Mg2+-free phosphate-buffered saline and solubilized with 0.35 mL of 0.3 M NaOH followed by neutralization with 0.35 mL of 0.3 M HCl. Thereafter, 0.6 mL of 50% trichloroacetic acid was added. The mixture was then centrifuged at 2500 g for 10 min at room temperature. The pellets were washed with 1 mL of a mixture of chloroform and methanol (1:1) solution, followed by centrifugation at 2500 g for 3 min at room temperature. The radioactivity in the cell lysates was measured by mixing 9 mL of scintillation fluid (BCS, Amersham, UK) into the aliquot (9:1) and counted in a liquid scintillation counter (Beckman, USA). Mucus synthesis was expressed as the ratio of [3H]-glucosamine incorporation relative to that of the control group.
Assessment of anticoagulant activity
The bleeding time was used as a determinant of coagulation function.13 One hour after low molecular weight heparin treatment, rats were anaesthetized with an intraperitoneal injection of pentobarbitone sodium and the abdomen was opened with an incision to expose the liver. A piece of liver was excised from the edge of a lobe; pieces of filter paper were dipped at 10-s intervals into the blood oozing from the cut surface until the end-point was reached, indicated by a piece of blood clot clinging to the filter paper. The bleeding time was taken as the time elapsing between cutting the liver edge and the end-point of bleeding. Three separate consecutive readings were taken by a person unaware of the type of treatment from three lobes in each rat and the values were averaged.
The 3.0-kDa low molecular weight heparin was purchased from Sigma Chemicals Co. (St Louis, MO, USA). It was dissolved in normal saline (0.9% NaCl, w/v) before administration.
All data were presented as the mean ± S.E.M. Statistical analysis was performed with an analysis of variance (ANOVA) followed by a Dunnett t-test or Student’s two-tailed unpaired t-test. P values less than 0.05 were considered to be statistically significant.
Effects of i.v. or i.g. low molecular weight heparin on gastric ulcer healing
The three doses (0.6, 3.0 and 6.0 mg/kg) of low molecular weight heparin, when given for 4 days either by the i.v. or i.g. route, produced a similar effect in hastening the reduction of gastric ulcer size. Both routes of low molecular weight heparin decreased the ulcer size in a dose-dependent fashion and a significant effect was found at the two higher doses (Figure 1A,B).
Effects of i.v. or i.g. low molecular weight heparin on histological changes after ulcer induction
The length of the regenerated gastric mucosa at the ulcer margin was increased by i.v. low molecular weight heparin after 4 days of drug treatment. This effect was more remarkable at the two higher doses, and was significantly different from that of the control. However, i.v. low molecular weight heparin did not affect the length of the ruptured muscularis mucosae (Figure 2A,B). A similar effect on the regeneration of the gastric mucosa was found with i.g. low molecular weight heparin. This effect occurred in a dose-related manner. The increases in the regenerated gastric mucosa at the two higher doses were statistically significantly greater than that of the control. Intragastric low molecular weight heparin also did not shorten the length of the ruptured muscularis mucosae (Figure 3A,B), indicating that the drug had no contractile action at the ulcer base.7
Effects of i.v. or i.g. low molecular weight heparin on epithelial cell proliferation at the ulcer margin
Both i.v. and i.g. low molecular weight heparin given for 4 days dose-dependently increased the incorporation of 5-bromo-2′-deoxyuridine into the epithelial cells of the gastric mucosa at the ulcer margin. The 5-bromo-2′-deoxyuridine labelling indices at the two higher doses of low molecular weight heparin in both treatment routes were significantly higher than that of the corresponding control (Figure 4A,B).
Effects of i.v. or i.g. low molecular weight heparin on angiogenesis at the ulcer margin and base
There was a dose-related increase in the microvessel number both at the ulcer margin and base after 4 days of treatment with i.v. low molecular weight heparin. Significant effects were observed at the two higher doses at the ulcer margin and at the highest dose at the ulcer base (Figure 5A,B). Intragastric low molecular weight heparin produced a similar result to i.v. low molecular weight heparin on angiogenesis after the same treatment duration. The microvessel number was profoundly elevated by the two higher doses of low molecular weight heparin both at the ulcer margin and base (Figure 6A,B).
Effects of i.v. or i.g. low molecular weight heparin on the thickness of the gastric mucus layer
Histologically, in five consecutive examined fields starting from the ulcer edge, both i.v. and i.g. low molecular weight heparin increased the thickness of the gastric mucus layer in a dose-dependent manner. A significant effect for i.v. low molecular weight heparin was found at the highest dose. However, the increase in the mucus layer thickness was found to be more marked in the i.g. group than in the i.v. group. The three doses of i.g. low molecular weight heparin significantly elevated the thickness of the mucus layer (Figure 7A,B).
Effect of low molecular weight heparin on mucus synthesis in MKN-28 cells
Low molecular weight heparin produced a stimulatory effect on mucus synthesis in MKN-28 cells. Low molecular weight heparin, at concentrations of 50, 100, 200, 400 or 800 μg/mL, elevated the incorporation of [3H]-glucosamine in MKN-28 cells after 6 h of incubation in a concentration-dependent manner. The effect produced by 200–800 μg/mL of low molecular weight heparin was significantly higher than that of the control group (Figure 8).
Effects of i.v. or i.g. low molecular weight heparin on blood coagulation
The three doses of i.v. low molecular weight heparin and the highest dose of i.g. low molecular weight heparin used in this experiment did not affect the bleeding time 1 h after administration. The bleeding time was similar among the control and low molecular weight heparin groups (Figure 9).
The present study indicated that low molecular weight heparin was able to increase ulcer healing by either the i.v. or i.g. route, and this effect occurred via the stimulation of gastric mucosal cell proliferation and regeneration, angiogenesis and mucus production. All these effects were comparable with those of unfractionated heparin administration.1,2 The importance of these processes in gastric ulcer healing has been demonstrated in various studies.3–5 The results for low molecular weight heparin confirm that not only unfractionated heparin, but also its fragments, possess the property to promote ulcer healing. This study has demonstrated for the first time that the low molecular weight fragments derived from parent unfractionated heparin are responsible for the ulcer healing action in the stomach.
Low molecular weight heparins are manufactured from unfractionated heparin by controlled depolymerization using a variety of fragmentation methods.14 These processes yield products with mean molecular weights around 4000–6000 Da, with more than 60% of the polysaccharides between 2000 and 8000 Da. The comparable figures for unfractionated heparin are 12 000–14 000 Da, with less than 15% of 2000–8000 Da.15 Heparin with a molecular weight around 3000 Da, which was used in this study, has been shown to have a better anticancer action than heparin of other molecular sizes.16,17 The anticoagulant property of heparin depends on the presence of a specific pentasaccharide sequence (molecular weight about 1700 Da) with a high affinity to antithrombin,18 which enhances its inhibitory action against serine proteases. The reason why the low molecular weight heparin used in this study did not significantly affect the coagulation function may be that the molecule was the wrong size to produce an anticoagulant effect or the doses used were low and insufficient to produce this action. The anticoagulation factor-Xa activities of the doses of low molecular weight heparin used in this study were only one-quarter of that of the relative equivalent weight of unfractionated heparin which produces anticoagulant activity. This result, combined with the effect of both i.v. and i.g. unfractionated heparin,1,2 further substantiates that the ulcer healing property of heparin is independent of its anticoagulant activity. This phenomenon is in contrast to the observation in inflammatory bowel disease in which the beneficial effect of heparin may result from its anticoagulant property.19 It is envisaged that the beneficial effect of protamine, an antidote for the anticoagulation action of heparin, on gastric ulcer healing3 would not be observed in inflammatory bowel disease, as the anticoagulant action of any endogenously released heparin during inflammation in the colon would be blocked by protamine.
Apart from its anticoagulant action, fragments of heparin can also potentiate the mitogenic effect of fibroblast growth factor20,21 and increase the release of hepatocyte growth factor,22 both of which are angiogenic factors. It has also been reported that low molecular weight heparin can enhance the migration of vascular endothelial cells in a concentration-dependent manner in vitro.23 Furthermore, low molecular weight heparin has been demonstrated to stimulate re-endothelialization after experimental balloon angioplasty in vivo. The animals that received low molecular weight heparin showed a significantly higher number of endothelial cells and a more intact endothelium compared to untreated animals.24 Furchgott25 demonstrated that an intact endothelium was required for a normal endothelium-dependent vasorelaxation. The re-endothelialization seems to be closely associated with the improvement of endothelial function.26–28 In the present study, low molecular weight heparin increased gastric mucosal cell proliferation and regeneration, angiogenesis and mucus production. We hypothesize that all these effects may result from its positive action on endothelial cells: the migration of endothelial cells and the re-endothelialization of injured microvessels will promote angiogenesis, and the ability of low molecular weight heparin to potentiate the mitogenic effect of fibroblast growth factor and the release of hepatocyte growth factor will facilitate this process.20–22 This action has been demonstrated in an unfractionated heparin study.1 Increased angiogenesis at the ulcer site supplies more oxygen and nutrients to support cell proliferation and regeneration, leading to re-epithelialization of the ulcer crater, as well as mucus production that contributes in part to mucosal protection and epithelial recovery in the stomach. Angiogenesis at the ulcer base can facilitate the formation of granulation tissue and supply connective tissue for the restoration of the lamina propria.29,30 In addition, endothelial cells with an intact endothelium are the source of nitric oxide release. Nitric oxide is a pivotal factor for gastric ulcer healing, which has been demonstrated in studies with unfractionated heparin.1,2 Whether or not the stimulatory action of low molecular weight heparin on angiogenesis and mucus secretion is mediated by NO requires further study, although a preliminary investigation has shown that this drug increases nitric oxide synthase activity in the gastric tissue (Y. Li et al., 2000, unpublished findings). Jorneskog et al.31 reported that low molecular weight heparin increased the healing process of chronic foot ulcers in diabetic patients by improving the capillary circulation in the ulcer margin. The present study extends this beneficial effect of heparin to gastric ulcer healing, perhaps via the stimulatory effect on angiogenesis and the improvement of blood flow in the tissue. Interestingly, heparin shows completely different effects in polymorphonuclear leucocytes and in colonic cancer cells.32,33 The actions of basic fibroblast growth factor and NO release from these cells are attenuated by heparin, thereby producing antitumour effects. From all of these studies, it is likely that the pharmacological action of heparin is dependent on the tissue type and pathophysiological situation. This may explain why heparin behaves in an opposite manner in normal tissues and cancer cells.
It was first shown by Johnson et al.34 that the subcutaneous injection of a low molecular weight fraction of heparin resulted in much higher blood levels than with either unfractionated heparin or a higher molecular weight fraction. Recently, it has been shown that low molecular weight heparins exhibit favourable pharmacokinetic properties and a bioavailability of nearly 100%. This is mainly due to a better absorption of the shorter heparin chains from the subcutaneous tissue. This phenomenon may also be implicated in the ability of i.g. low molecular weight heparin to promote ulcer healing via direct and better absorption from the gastrointestinal tract, especially at the ulcer site, where the drug is bound and absorbed. This would directly facilitate the ulcer healing process in the stomach. Indeed, i.g. low molecular weight heparin is more effective than i.v. low molecular weight heparin in the promotion of gastric mucus formation (Figure 7), demonstrating that low molecular weight heparin may stimulate mucus synthesis directly in mucus-secreting cells after absorption. This hypothesis was further confirmed in the present study as the drug was shown to increase mucus synthesis in a mucus-secreting cell line (Figure 8).
In conclusion, low molecular weight heparin can accelerate gastric ulcer healing through a more efficient absorption pathway. The ulcer healing effect is due to an increase in gastric mucosal cell proliferation and regeneration, angiogenesis and mucus production. A possible action on endothelial cells is suggested as a major underlying mechanism for the ulcer healing activity in the stomach.
This work was supported in part by CRCG and RGC grants from the University of Hong Kong and the Hong Kong Research Grant Council, respectively.