Doxycycline blocks gastric ulcer by regulating matrix metalloproteinase-2 activity and oxidative stress

Abstract

Aim: To examine the effect of doxycycline on the activity of matrix metalloproteinases (MMPs) and oxidative stress in gastric tissues of rats following gastric injury.

Methods: Gastric ulcers were generated in rats by administration of 70% ethanol, and activity of doxycycline was tested by administration 30 min prior to ethanol. Similarly, the effect of doxycycline was tested in an indomethacin-induced gastric ulcer model. The activities and expression of MMPs were examined by zymography and Western blot analysis.

Results: Gastric injury in rats as judged by elevated ulcer indices following exposure to ulcerogen, either indomethacin or ethanol, was reversed significantly by doxycycline. Indomethacin-induced ulcerated gastric tissues exhibited about 12-fold higher proMMP-9 activity and about 5-fold higher proMMP-3 activity as compared to control tissues. Similarly, ethanol induced about 22-fold and about 6-fold higher proMMP-9 and proMMP-3 activities, respectively, in rat gastric tissues. Both proMMP-9 and MMP-3 activities were markedly decreased by doxycycline in ulcerogen treated rat gastric tissues. In contrast, the reduced MMP-2 activity in ulcerated tissues was increased by doxycycline during ulcer prevention. On the other hand, doxycycline inhibited significantly proMMP-9, -2 and -3 activities in vitro. In addition, doxycycline reduced oxidative load in gastric tissues and scavenged H₂O₂ in vitro. Our results suggest a novel regulatory role of doxycycline on MMP-2 activity in addition to inhibitory action on MMP-9 and MMP-3 during prevention of gastric ulcers.

Conclusion: This is the first demonstration of dual action of doxycycline, that is, regulation of MMP activity and reduction of oxidative stress in arresting gastric injury.

Keywords: Doxycycline; Extracellular matrix; Matrix metalloproteinases; Reactive oxygen species; Tissue inhibitor of metalloproteinase.

Figure 1

Protective effect of doxycycline on gastric injury. Gastric ulcers were induced in rats by oral administration of 48 mg/kg indomethacin or 70% ethanol (low and high) doses. Doxycycline (50 mg/kg) was administered orally prior to ulcerogen treatment to assess the protective effect. Control rats received sterile water only. After 3 h, rats were sacrificed and ulcer indices were scored. Mean ulcer index for indomethacin-induced (A) and ethanol-induced (B) ulceration in each group of rats was plotted. Dox: Doxycycline; Con: Control; Indo: Indomethacin; L.EtOH: Low dose ethanol; H.EtOH: High dose ethanol. Data expressed as mean ± SE. ^aP < 0.05, ^bP < 0.001 vs control; ^cP < 0.001 vs indomethacin; ^dP < 0.001 vs high dose ethanol; ^eP < 0.01 vs low dose ethanol.

Figure 2

Effect of doxycycline on secreted and synthesized matrix metalloproteinase-2 and -9 activities during protection against indomethacin-induced ulcers. Gastric ulcer was induced by intragastric administration of 48 mg/kg indomethacin in rats, and doxycycline was given 30 min prior to indomethacin for protection studies. Gelatin zymography was performed using phosphate buffered saline (PBS) (A) and Triton X-100 (TX) (C) extracts of gastric tissue from different groups of rats. Histographic representation of arbitrary activity of pro-matrix metalloproteinase (MMP)-9 (B) and proMMP-2 (D) in PBS and TX extracts of gastric tissues, as measured by Lab Image software using the above zymogram and two other representative zymograms from independent experiments. Data expressed as mean ± SE. ^aP < 0.05, ^bP < 0.01, ^dP < 0.001 vs control; ^eP < 0.05, ^fP < 0.01 vs indomethacin.

Figure 3

Effect of doxycycline on secreted and synthesized matrix metalloproteinase-2 and -9 activities during protection of ethanol-treated gastric ulcer. Low (0.4 mL/rat) and high (0.8 mL/rat) doses of 70% ethanol were administered intragastrically to induce moderate and maximum gastric ulcers, and doxycycline was administered 30 min prior to ethanol. Gelatin zymography was performed using phosphate buffered saline (PBS) (A) and Triton X-100 (TX) (C) extracts of gastric tissues from different group of rats. Histographic representation of arbitrary activity of pro-matrix metalloproteinase (MMP)-9 (B) and proMMP-2 and MMP-2 (D) in PBS and TX extracts of gastric tissues as measured by Lab Image software using the above zymogram and two other representative zymograms from independent experiments. PBS extracts were subjected to Western blotting (E) and probed with polyclonal anti-MMP-2 and β-tubulin antibody. Fold changes of MMP-2 (F) are represented in a histogram as measured by Lab Image software using the above western blot and two other representative blots from independent experiments. Data expressed as mean ± SE. ^aP < 0.05, ^bP < 0.01, ^dP < 0.001, vs control; ^eP < 0.05, ^fP < 0.001 vs high-dose ethanol; ^gP < 0.05, ^hP < 0.01 vs low-dose ethanol of respective PBS and TX extracts. L.EtOH: Low dose ethanol; H.EtOH: High dose ethanol.

Figure 4

Effect of doxycycline on matrix metalloproteinase-3 activity during protection against indomethacin-induced gastric ulcer. Gastric ulcer was induced by intragastric administration of 48 mg/kg indomethacin in rats, and doxycycline was given 30 min prior to indomethacin for protection studies. Casein zymography was performed using phosphate buffered saline (PBS) (A) and Triton X-100 (TX) (C) extracts of gastric tissue from different groups of rats. Histographic representation of arbitrary activity of pro-matrix metalloproteinase (MMP)-3 in PBS (B) and TX (D) extracts of gastric tissues as measured by Lab Image software designed using the above zymogram and two other representative zymograms from independent experiments. Data expressed as mean ± SE. ^aP < 0.05, vs indomethacin; ^dP < 0.001 vs control.

Figure 5

Effect of doxycycline on matrix metalloproteinase and tissue inhibitors of metalloproteinase activity and expression during protection against ethanol-treated gastric ulcer. Ulcers were induced in rat using 70% ethanol at low and high doses, and protection by doxycycline was performed. Casein zymography (A) was performed using phosphate buffered saline (PBS) extract of different gastric tissues. Histographic representation (B) of arbitrary activity of pro-matrix metalloproteinase (MMP)-3 as measured by Lab Image software using the above zymograms and two other representative zymograms from independent experiments. PBS extracts of gastric tissues from different experimental groups were subjected to Western blot blotting (C) and probed with polyclonal anti-MMP-9 and -3, tissue inhibitors of metalloproteinase (TIMP)-1 and -2, and anti-β-tubulin antibodies. Fold changes of proteins as represented below each band were measured by Lab Image software using the above wester blots and two other blots from independent experiments. Data expressed as mean ± SE. ^aP < 0.05 vs low-dose ethanol; ^dP < 0.01 vs high-dose ethanol. L.EtOH: Low dose ethanol; H.EtOH: High dose ethanol.

Figure 6

Inhibitory effect of doxycycline on matrix metalloproteinase activity in rat gastric tissue extract in vitro. A: Gelatin and casein zymography of gastric tissue extract following treatment with doxycycline or 1, 10-phenanthroline at two different concentrations; B: Histographic representation of pro-matrix metalloproteinase (MMP)-3, proMMP-2 and proMMP-9 activities as measured by Lab Image software using the above zymograms and two other representative zymograms from independent experiments. Data expressed as mean ± SE. ^bP < 0.01, ^dP < 0.001 vs control phenanthroline was labeled as Phen in the figure.

Figure 7

H₂O₂ scavanging activity of doxycycline. A: Gelatin zymography of phosphate buffered saline (PBS) extract from control gastric tissue was incubated with H₂O₂ (500 mmol/L) at 37°C for 1 h. Equal amounts of PBS extract were pretreated with doxycycline (200 μmol/L) before incubation with H₂O₂. In one lane (Cat, lane 4), the extract was pretreated with catalase before reaction with H₂O₂; B: Histographic representation of pro-matrix metalloproteinase (MMP)-2 and MMP-2 activities as measured by Lab Image software using the above zymograms and two other representative zymograms from independent experiments. Veh: Vehicle; Cat: Catalase. Data expressed as mean ± SE. ^bP < 0.01, ^dP < 0.001 vs control; ^eP < 0.05, ^fP < 0.01 vs vehicle.