Indomethacin can induce cell death in rat gastric parietal cells through alteration of some apoptosis- and autophagy-associated molecules

Abstract

In clinical medicine, indomethacin (IND, a non-steroidal anti-inflammatory drug) is used variously in the treatment of severe osteoarthritis, rheumatoid arthritis, gouty arthritis or ankylosing spondylitis. A common complication found alongside the therapeutic characteristics is gastric mucosal damage. This complication is mediated through apoptosis and autophagy of the gastrointestinal mucosal epithelium. Apoptosis and autophagy are critical homeostatic pathways catalysed by caspases downstream of the gastrointestinal mucosal epithelial injury. Both act through molecular signalling pathways characterized by the initiation, mediation, execution and regulation of the cell regulatory cycle. In this study we hypothesized that dysregulated apoptosis and autophagy are associated with IND-induced gastric damage. We examined the spectra of in vivo experimental gastric ulcers in male Sprague-Dawley rats through gastric gavage of IND. Following an 18-hour fast, IND was administered to experimental rats. They were sacrificed at 3-, 6- and 12-hour intervals. Parietal cells (H⁺ , K⁺ -ATPase β-subunit assay) and apoptosis (TUNEL assay) were determined. The expression of apoptosis-signalling caspase (caspases 3, 8, 9 and 12), DNA damage (anti-phospho-histone H2A.X) and autophagy (MAP-LC3, LAMP-1 and cathepsin B)-related molecules in gastric mucosal cells was examined. The administration of IND was associated with gastric mucosal erosions and ulcerations mainly involving the gastric parietal cells (PCs) of the isthmic and upper neck regions and a time-dependent gradual increase in the number of apoptotic PCs with the induction of both apoptotic (upregulation of caspases 3 and 8) cell death and autophagic (MAP-LC3-II, LAMP-1 and cathepsin B) cell death. Autophagy induced by fasting and IND 3 hours initially prompted the degradation of caspase 8. After 6 and 12 hours, damping down of autophagic activity occurred, resulting in the upregulation of active caspase 8 and its nuclear translocation. In conclusion we report that IND can induce time-dependent apoptotic and autophagic cell death of PCs. Our study provides the first indication of the interactions between these two homeostatic pathways in this context.

Keywords: apoptosis; autophagy; gastric mucosa; indomethacin; parietal cells.

Figure 1

1.1 Gross examination of the stomach showing normal intact glistening pink mucosa in healthy‐eating (A) and fasting 18 h (B). Brown haemorrhagic superficial erosions (arrow) are demonstrated at early time IND 3 h (C), deeper at 6 h (D) and 12 h (E). 1.2. Photomicrographs of the gastric mucosal tissue. A, The gastric mucosa appears unremarkable in the control group (animals eating normally). B, The gastric mucosa appears unremarkable in the animals fasting for 18 h. C‐E, In IND‐treated animals, there are superficial erosions involving the isthmic and neck regions of the gastric mucosa, together with sloughing of overlying superficial cells (arrows) at 3 h (C). Ulceration (arrow) of the gastric mucosa occurred after 6 h (D) and 12 h (E). (PAS and haematoxylin)

Figure 2

2.1. Photomicrograph of gastric mucosa showing very few TUNEL‐positive cells on the surface epithelium in healthy‐eating (A) and in fasting 18‐h groups (B). TUNEL‐positive cells are located at the isthmic and neck regions of the mucosal glands in IND‐treated groups (arrow) after 3 h (C), exfoliated after 6 and 12 h (D,E). Note, the PCs with positive nuclei shown in the insets. Counting of stained cell nuclei and statistical analysis proved the significant and time‐dependent increase in TUNEL‐positive cells (F) after IND treatment (#: non‐significant, *P < .05, **P < .001). 2.2. Photomicrographs of the gastric mucosa showing H⁺ K⁺ ATPase ß‐subunit–positive cells (PCs, parietal cells) which are normally distributed along the entire gastric glands in the control group (A: animals eating normally) and in the animal fasting for 18 h (B). Exfoliation of these cells at the site of damage occurred at the surface after 3, 6 and 12 h (red arrows, C, D and E respectively). We counted of the cells with cytoplasmic staining cell, and there was a statistically significant time‐dependent decrease in PCs at a time starting at 6 till 12 h following IND treatment (F) (#: non‐significant, *P < .05, **P < .001, the red asterisk means decrease)

Figure 3

3.1. Photomicrographs of gastric mucosa showing cleaved caspase 3–positive cells. There are very few positive nuclei in healthy‐eating (A) and few at the gastric pit and upper part of the neck in fasting 18 h (B). IND‐treated groups showing a time‐dependent increase in cleaved caspase 3–positive cells at the site of mucosal damage (arrow) after 3, 6 and 12 h (C‐E). Note, the positively stained nuclei of PCs in the inset. Counting of stained cell nuclei and statistical analysis proved the time‐dependent and significant increase in cleaved caspase 3–positive cells in IND‐treated groups at the site of mucosal damage (F) ( #: non‐significant, *P < .05). 3.2. Photomicrographs of gastric mucosa stained by double‐labelling immunofluorescence of active caspase 3 (green dots) and PCs (red) marker showing co‐localization of active caspase 3 inside PC nucleus (arrow) in IND‐treated group, and the double labelling appears few intranuclear after 3 h (left panel) but mainly nuclear after 6 h (middle panel) and 12 h (right panel)

Figure 4

4.1. Photomicrographs of gastric mucosa showing active caspase 8–positive cells. There are few positive cells at the gastric pits and the upper part of the neck in healthy‐eating (A), fasting 18 h (B) and after 3 h of IND (C) groups. IND‐treated groups showing a time‐dependent increase in active caspase 8–positive cells at the site of mucosal damage after 6 and 12 h (D,E). Note, the positivity is cytoplasmic in healthy‐eating (A), fasting 18 h (B) but nuclear in few PCs after 3 h that increased after 6 and 12 h (D,E) as shown in the insets. 4.2. Photomicrographs of gastric mucosa stained by double‐labelling immunofluorescence of active caspase 8 (green dots) and PC (red) marker showing co‐localization of active caspase 8 inside PC cytoplasm in normal (A) and in fasting 18‐h (B) groups. In IND‐treated group, the double labelling appears few intranuclear after 3 h (C) but mainly nuclear after 6 h (D) and 12 h (E). Counting of co‐localized dots and statistical analysis (F) proved the time‐dependent and significant increase in active caspase 8 in PCs in IND‐treated groups (#: non‐significant,*P < .05, **P < .001, red asterisk = decrease and black asterisk = increase). 4.3 Double‐labelling immunofluorescence of active caspase 8 (red dots) and γ H2A.X (green dots) showing few cytoplasmic red dots in healthy‐eating (A) and fasting 18‐h (B) groups. In contrast, there is an apparent nuclear co‐localization (yellow dots pointed by arrows) of active caspase 8 and γ H2AX in IND‐treated group at 3 h (C) and 6 h (D) increased in IND 12 h (E). A perinuclear ring of co‐localization (arrowhead) appeared at IND 6‐h–treated group (4.3 G). Counting of γ H2A.X dots and statistical analysis (F) proved the time‐dependent and significant increase in γ H2A.X in PCs in IND‐treated groups (#: non‐significant, **P < .001)

Figure 5

5.1 Photomicrographs of gastric mucosa showing MAP‐LC3–positive cells mostly in the basal part and few in the luminal part of the gastric glands in the healthy‐eating group (A). However, luminal part of the gland shows an increase in the positive cells in fasting 18 h (B) and in IND 3‐h (C) but decreases in IND 6‐h (D) and IND 12‐h (E) groups. Note, the cytoplasmic positivity in PC is shown in the inset. Note, the damaged area is pointed by the arrow. 5.2 Double‐labelling immunofluorescence of LC3 (green dots) and PC marker (red) showing minimal co‐localization of MAP‐LC3 inside luminal PCs cytoplasm in normal (A) and increased in fasting 18‐h (B) groups. In IND‐treated groups, the double labelling appears maximally in 3‐h (C) and damped down in 6‐h (D) and 12‐h (E) IND‐treated groups. Counting of co‐localized dots and statistical analysis (F) proved the significant increase in LC3 and PC IF in fasting 18 h and in 3 h after IND‐treated groups and damped down in 6‐h and 12‐h IND‐treated groups (*P < .05, **P < .001, red asterisk = decrease and black asterisk = increase). 5.3. A diagram (C) showing an increase in fluorometric protein assay of cathepsin B in fasting 18 h and in 3 h after IND‐treated groups and damped down in 6‐h and 12‐h IND‐treated groups (*P < .05, **P < .001, red asterisk = decrease and black asterisk = increase)

Figure 6

6.1. Double‐labelling immunofluorescence of active caspase 8 (green dots) and MAP‐LC3 (red dots) showing the presence of few cytoplasmic co‐localization (yellow dots pointed by the arrow) in the healthy‐eating group (A). In contrast, numerous co‐localization of caspase 8 and LC3 is present in fasting 18 h (B) and IND 3 h (C) with few co‐localization in 6 h and very few in IND 12 h. Note, dots of active caspase 8 appear intranuclear (D&E). 6.2. Double‐labelling immunofluorescence of caspase 8 (red dots) and Lamp‐1 (green dots) showing few co‐localization (yellow dots pointed by the arrow) in a healthy‐eating group (A). In contrast marked co‐localization of active caspase 8 and Lamp‐1 is present in fasting 18 h (B) and in IND 3 h (C) with few co‐localization in 6 h (D) and very few in IND 12 h (E). A photograph (A) showing Western blot analysis of the whole gastric mucosa active caspase 8, anti‐phospho‐histone γ H2A.X and MAP‐LC3. 6.3.B analysis of western blotting of LC3I versus LC3II Note, the western blot analysis of the three parameters between different groups and time intervals; active caspase 8 level decreases in fasting 18 h and treated 3h (but increased in IND‐treated groups after 6 and 12 h). γ H2A.X shows little increase in fasting 18 h and more in IND‐treated groups by time intervals. MAP‐LC3 increases in fasting 18 h and in 3 h after IND‐treated group and dropped after 6‐ and 12‐h IND‐treated groups

Figure 7

Diagram summarizing the hypothesis of the mechanism of IND‐induced gastric PC damage. It is plausible to propose that modifying autophagy may have therapeutic ramifications against IND‐induced gastric mucosal damage. Taken as a whole, since IND‐induced apoptosis or autophagy seems to be allied to licence the PCs fate, optimal modulation of autophagy can be a therapeutic strategy to attenuate IND‐associated gastric damages. However, extensive large‐scale experimental work is mandated to investigate how IND could inhibitor abort autophagy could be a leading strategy to rescue gastric mucosa from damage. This effect could be due to cleavage of specific autophagic proteins or genes either directly or indirectly through upregulation of active caspase 8 (blue arrow = increase or upregulation, while red arrows = decrease or downregulation)