- Email: [email protected]
Role of mucus in the repair of gastric epithelial damage in the rat
GASTROENTEROLOGY 1986:91:603-11
Role of Mucus in the Repair of Gastric Epithelial Damage in the Rat Inhibition Agents
of Epithelial
Recovery
by Mucolytic
JOHN L. WALLACE and BRENDAN J. R. WHITTLE Department England
of Mediator Pharmacology,
Wellcome Research Laboratories,
A role for mucus in providing a microenvironment over sites of gastric damage, which is conducive to reepithelialization, has been proposed. We tested this hypothesis by examining the effects of disruption of such mucus on the recovery of epithelial integrity after damage induced by 50% ethanol. Exposure of an ex vivo chambered gastric mucosa to topically applied 50% ethanol resulted in copious release ofmucus, cellular debris, and plasma, which formed a continuous cap over the mucosal surface. Ethanol-induced gastric damage was accompanied by extensive surface epithelial cell damage and a marked decrease in transmucosal potential difference. During the 30 min after ethanol was removed from the chamber, the epithelium became reestablished and the potential difference gradually recovered to 94% of the level before ethanol treatment. However, if the mucolytic agents N-acetylcysteine (5%) or pepsin (0.5%) were added to the bathing solutions, the “mucoid cap” disintegrated and the recovery of potential difference was significantly retarded (recovering to only 51% and 52% of levels before ethanol treatment). Histologic evaluation confirmed that mucosae treated with either agent had significantly less [p < 0.005) intact epithelium at the end of the experiment. Removal of the mucoid cap with forceps caused a similar inhibition of the repair of the epithelium and the recovery of potenReceived August 29, 1985. Accepted February 28, 1986. Address requests for reprints to: Dr. J. L. Wallace, Department of Physiology, Queen’s University, Kingston, Ontario, K7L 3N6, Canada. Dr. Wallace is the recipient of a fellowship from the Medical Research Council of Canada. Part of these data were presented at the 86th Annual Meeting of the American Gastroenterologicai Association in New York, in May 1985. 0 1986 by the American Gastroenterological Association 0016-5085/86/$3.50
Beckenham,
Kent,
tial difference. Both mechanical and chemical (Nacetylcysteine) disruption of the mucoid cap resulted in a significant increase in the mucosal leakage of albumin and hemoglobin, supporting previous histologic evidence that the mucoid cap traps blood components over the damaged mucosa. These studies support the hypothesis that mucus released in response to topical application of an irritant plays an important role in the repair of epithelial damage through the process of restitution. The role of mucus in the prevention of gastric damage remains controversial (l-3), and it has been suggested that a more important role for mucus may be in the repair of acute damage. Using morphologic techniques, Morris et al. (4) demonstrated that there was a discontinuous coat of mucus on the undamaged rat stomach and a profound release of gel mucus in response to the topical application of an ulcerogenic agent. This gelatinous mucus formed a cap over the damaged region, which along with the entrapped cellulaf debris and plasma, it was proposed, could provide an environment favorable for the process of reepithelialization (restitution) to occur. Rapid repair of the epithelium appears to be dependent upon an intact basal lamina along which the cells migrate (5,6). A relatively high pH may be necessary for the maintenance of the integrity of the basal lamina, since this structure is readily degraded by acid (6,7). If mucus does play a role in providing a microenvironment over damaged regions that is conducive to the restitution process, then disruption of the mucoid cap should allow luminal acid to penetrate to and damage the basal lamina, thus impairing restitution. The present study was performed to Abbreviation
used
in this paper: NAC, 5% N-acetylcysteine.
604 WALLACE AND WHITTLE
determine the effects of disruption of the mucoid cap, either directly or by dissolution with mucolytic agents, on the repair of the gastric epithelium after damage induced by 50% ethanol. The extent of gastric epithelial damage before and after application of 50% ethanol was assessed histologically. The recovery toward control levels of transmucosal potential difference was used as an additional index of recovery of epithelial integrity, since this has been shown to closely parallel the restitution process (6,8). As a functional mucoid cap would trap plasma and blood over a damaged region, mucosal leakage of albumin, Evans blue, and hemoglobin were also measured.
Materials and Methods Male Wistar rats weighing 225-250 g were used. The rats were deprived of food, but not water, for 18-20 h before an experiment. Under pentobarbitone anesthesia (60 mg/kg i.p.) an ex vivo gastric chamber, which allowed direct observation of the gastric’ mucosa throughout the experiment, was prepared as previously described in detail (9,lO). Each experiment consisted of six sequential IO-min periods. At the beginning of each period 5.0 ml of a prewarmed (37%) solution was added to the chamber. While in the chamber, the solution was stirred by a glass paddle which turned at -200 rpm and was placed 1 cm above the mucosal surface. At the end of the period, the solution was removed from the chamber by a syringe. In the control group, 0.3 M mannitol was added to the chamber in the first two periods. In the third period 50% ethanol (vol/vol) was added to the chamber. During periods 4-6, the luminal bathing solution was 0.05 M hydrochloric acid in 0.2 M mannitol. In the test groups, either 5% N-acetylcysteine (NAC) or 0.5% pepsin (both wt/vol; Sigma Chemical Co., St. Louis, MO.) were added to the chamber in each period except the third. These compounds were dissolved in the corresponding control solution. Transmucosal potential difference was measured continuously throughout each experiment using apparatus described previously (11). At the end of the final period the exposed mucosa was photographed on color transparency film. Macroscopically visible damage was quantified planimetrically from these transparencies in a randomized, blind manner (10). The gastric pedacle was cut and the mucosa was immersed in neutral buffered formalin. Samples (10 X 2 mm) of the fundic (corpus) region of the stomach were excised and processed by routine techniques before embedding in paraffin. These samples were taken from the same regions of all mucosae; i.e., the midfundus of both the ventral and dorsal sides of the stomach. Sections (4-6 pm) were stained with hematoxylin and eosin and mounted on glass slides. Additional experiments [n = 4 per group) were performed in which the mucosa was fixed at the end of period 2 or period 3. These experiments were performed to determine the extent of epithelial damage before and immediately after application of 50% ethanol. The histo-
GASTROENTEROLOGY Vol. 91. No. 3
logic analysis was performed in a randomized manner, with the slides coded to avoid observer bias. The sections of fundic mucosa were examined and the percentage of the length of the section with an intact epithelium was calculated. Three sections (each 10 mm long) from each stomach were examined, and the overall score for the stomach was taken as the mean of the scores for each section. Regions on the sections in which hemorrhagic erosions were present were excluded from the calculation. Mucus release into the luminal solutions was determined using the Alcian blue method (12). After mixing the luminal solutions, a l.O-ml sample was added to 100 ~1 of Alcian blue reagent (10 mg/ml; Sigma) and incubated at room temperature for 12 h. The samples were then centrifuged (5 min, 1000 g) and the Alcian blue concentration in the supernatant was measured spectrophotometrically (absorbance at 620 nm). There is a linear relationship between the amount of Alcian blue precipitated and the concentration of mucous glycoprotein in the sample (12). Results are expressed as the number of micrograms of Alcian blue precipitated by 1.0 ml of sample. As N-acetylcysteine itself binds to Alcian blue, these determinations could not be performed with samples from that series of experiments. Additional experiments were performed in which Evans blue (10 mg/kg in saline; Phase Separations Ltd., Queensferry, United Kingdom) was administered intravenously 1 min after application of 50% ethanol to the chambered mucosa in period 3. Mucosal release of Evans blue following intragastric administration of 50% ethanol has previously been described (13). If mucus released in response to ethanol application did form a cap over the damaged area, then the dye should be trapped beneath such a cap. Furthermore, subsequent application of mucolytic agents should result in release of Evans blue into the luminal solution. Samples of the luminal solutions in periods 4-6 were analyzed for Evans blue concentration using a spectrophotometer (absorbance at 600 nm). It was possible that any effects of N-acetylcysteine or pepsin on epithelial intregity or recovery of potential difference might not be due to their mucolytic actions, but rather to a direct inhibition of the reepithelialization process. For instance, these substances might themselves damage the cells migrating from the gastric pits. Because of this possibility, a series of experiments was performed (n = 5) in which the mucoid cap that formed after application of 50% ethanol was removed with forceps. This procedure was performed taking meticulous care not to touch the underlying mucosa and was facilitated by the use of a stereomicroscope (Zeiss Opti 1, Carl Zeiss Ltd., Federal Republic of Germany) mounted over the chamber preparation. The mucus was removed 8 min after addition to the chamber of 50% ethanol (period 3) and took -2 min to complete. During the ensuing periods, 0.05 M HCl was added to the chamber, as in the control experiments. In the N-acetylcysteine experiments and the experiments in which the mucoid cap was removed with forceps, albumin and hemoglobin leakage from the chambered mucosa were measured. At the beginning of these experiments, 0.5 &i of 14C-labeled bovine serum albumin (50 &i/mg; Amersham International, Aylesbury, U.K.) dissolved in 0.25 ml of normal saline was administered
MUCUS AND RESTITUTION
September 1966
intravenously via a cannula in a femoral vein. At the end of each period, an aliquot (1.0 ml) of the luminal bathing solution was transferred to a scintillation vial and counted on a p spectrometer. Following the final period of the experiment, a sample of blood from a carotid artery (1.0 ml) was centrifuged (1 min; 9000 g) and the counts per minute in 100 ~1 of the plasma was determined. This was used to calculate the volume of plasma present in each sample of the luminal bathing solutions. Hemoglobin concentrations in l.O-ml samples of the luminal bathing solutions were measured spectrophotometrically. A few drops of a lysing reagent (Zaponin, Clay Adams, Parsippany, N.J.) were added to each sample. Absorbance at 540 nm was determined for each sample and was compared to a standard curve based on bovine hemoglobin [Sigma) for determination of the number of milligrams of hemoglobin per milliliter of sample. The accuracy of this spectrophotometric method was verified by checking the hemoglobin concentration of some samples on a hematology analyzer (Clay Adams). Statistical
1. Gastric Surface Epithelial Cell Damage Before, During, and After Application of 50% Ethanol: Effects of Mucolytic Agents Gastric luminal surface with intact epithelium (%I
Control NAC Peosin
Before ethanol
During ethanol
98.8 + 0.7 98.5 t 0.5 98.1 t 0.6
19.2 ? 5.2 21.3 lr 4.9 16.4 '_ 4.0
After ethanol 77.0 t 3.1 41.2 + 5.0" 59.0 r 1.0"
The percentage of the luminal surface with an intact epithelium was measured on coded slides using a light microscope. Regions in which hemorrhagic erosions were present were excluded from these calculations. These data are expressed as the mean ? SEM of 6-14 samples. The tissue samples were taken at 20 min (before 50% ethanol), 29 min (during the ninth minute of the period in which 50% ethanol was in the chamber), or 60 min (30 min after removal of 50% ethanol from the chamber). Groups that differ significantly from the control group: ” p < 0.001; b p < 0.005. NAC, 5% N-acetylcysteine; pepsin, 0.5% pepsin.
Analysis
All data are expressed as mean * SEM. Comparisons between groups were performed using the Student’s two-tailed t-test for unpaired data. To compare the relationship between percentage epithelial damage and the final potential difference value for each mucosa, linear regression analysis was performed. With all statistical analyses, an associated probability (p value) of ~5% was considered as significant.
Results Macroscopic
Table
605
Observations
During the first two periods of each experiment, the mucosa appeared normal and there were strands of mucus visible on the mucosal surface. The application of NAC or 0.5% pepsin during these periods had no visible effect on mucosal appearance. The addition of 50% ethanol to the chamber in the third period resulted in the development of focal patches of hyperemia. Extensive release of “mucus” (in fact, a mixture of sulfated and nonsulfated mucus, exfoliated cells, blood, and plasma components; see Reference 6) was evident during this period. This mucus formed a continuous coat over the mucosal surface, which we have subsequently referred to as the mucoid cap. Despite constant stirring of the solution in the chamber, the mucoid cap remained adherent to the mucosa. When 0.05 M hydrochloric acid was added to the chamber in periods 4-6, the hyperemic regions developed into hemorrhagic erosions. In the control group, these hemorrhagic erosions covered 16.2% -+: 2.6% (n = 6) of the glandular mucosa (determined at the end of the experiments). The extent of hemorrhagic damage in the groups treated with NAC (20.8% * 3.6%; n = 6) or pepsin (15.8% + 3.0%; n = 5) was not significantly different
from the control group. Some breaks in the mucoid cap became apparent during the final 30 min of the control experiments. In the NAC and pepsin groups, the dissolution of the mucoid cap was much more obvious. Large clumps of mucus became detached from the mucosa and were visible in the bathing solution. This mucolytic effect was particularly evident in the final 15 min of the experiments. During this time, blood clots over hemorrhagic erosions dispersed and frank bleeding into the bathing solution was observed.
Histologic
Observations
In the samples fixed at the end of the second period (i.e., before application of 50% ethanol) the mucosa appeared normal and the surface epithelium was almost completely intact (Table 1).The addition of NAC or pepsin to the luminal solutions in the first two periods did not cause any detectable damage. Application of 50% ethanol to the chambered mucosae resulted in extensive destruction of the epithelium with the resulting release of intracellular mucus. This epithelial damage involved -80% of the luminal surface, and the extent of this damage was not significantly affected by prior exposure to NAC or pepsin (Table 1). A layer of mucus, cellular debris, and blood was visible along the damaged portions of the luminal surface. In samples fixed at the end of the 60-min experiments, there were marked differences between the extent of intact epithelium in the four groups. In the control group, an average of 77% of the luminal surface had an intact epithelial cell layer, significantly (p < 0.005) more than in the NAC or pepsin
606
WALLACE AND WHITTLE
GASTROENTEROLOGY Vol. 91, No. 3
A
Mannitol
f NAC
50% ethanol
HCI + NAC
Potential difference (mV)
I
20
I
I
I
I
30
40
50
60
TIME (minutes)
---
Pepsin
Potentia! difference (mV)
20
10
40
30
50
60
TIME (minutes)
C
T
T
----
Control Mucoid cap removed (1)
Potential difference (mV)
0
10
20
30
I
I
,
40
50
60
TIME (minutes)
Figure
1. Transmucosal potential difference recordings for chambered gastric mucosae damaged with 50% ethanol. Each point represents the mean f SEM of a 5-min interval (n = 5-8 per group]. Asterisks denote points at which the test group differs significantly (* p < 0.05; ** p < 0.01) from the control group. A. In the test group, 5% N-acetylcysteine (NAC) was added to the chamber in each period except the period in which 50% ethanol was applied (minutes 21-30). B. The protocol for these experiments was the same as shown at the top of (A) except that 0.5% pepsin was added to the chamber in place of NAC. C. In the test group, the solutions added to the chamber were identical to those in the control group. Under a stereomicroscope, the mucoid cap that formed after application of 50% ethanol was meticulously peeled off the underlying mucosa at 28 min (arrow).
groups or the group in which the mucoid cap was removed with forceps (53.0% + 5.3%; p < 0.005). Transmucosal
Potential
Difierence
During the first two peripds of the control experiments the transmucosal potential difference
(PD) remained between -38 and -48 mV (Figure 1). The addition of NAC or pepsin to the control solutions had no effect on the PD. The addition of 50% ethanol to the chamber in period 3 elicited a sharp decrease in PD to --lo mV in all four groups (no significant differences). The PD remained depressed throughout the period in which ethanol was in the
September
MUCUS AND RESTITUTION
1966
chamber. During the final 30 min of the control experiments, the PD gradually recovered toward levels before ethanol treatment. At the end of the experiment the PD was not significantly different from that immediately before the application of ethanol to the mucosa. In the NAC and pepsin groups the recovery of PD was much slower, with significantly lower values than the control group at all time points during the final 30 min. By the end of the experiment the PD in the NAC and pepsin groups had recovered to only 51% + 4% and 52% + 7%, respectively, of the pre-ethanol-treatment level, both significantly lower (p < 0.01) than in the control group (94% t 6%). Direct removal of the mucoid cap with forceps had no immediate effect on PD, but had profound effects on the recovery of PD in the 30 min after removal of ethanol from the chamber (Figure 1C). In each period after removal of the cap, the PD was significantly lower (p < 0.05) than in the control group. Linear regression analysis of the data from the control, NAC, and pepsin groups revealed that there was a significant correlation between the transmucosal potential difference at the end of the experiment and the percentage of the luminal surface with an intact epithelium (r = 0.75; p < 0.01). Alcian Blue Binding During the first two periods there was a small amount of Alcian blue binding activity in the luminal solutions from control experiments (Table 2). However, the application of 50% ethanol during period 3 resulted in a marked increase (approximately sixfold; p < 0.001) in the level of mucus present in the luminal solutions. These data are in agreement with the direct observations of macroscopically visible mucus release in response to the application of 50% ethanol. During the final three Table
2. Alcian Blue Binding (~glml) in Luminal Solutions From Control and 0.5% Pepsin-
Treated Gastric Mucosae Time (min) l-10 11-20 21-30 31-40 41-50 51-60
Control 16.9 15.7 93.4 24.3 16.9 15.2
f ? t r C *
7.0 11.1 1.1 0.8 2.1 1.9
607
30
20
10
0 6
5 TIME
PERIOD
(10mm
1
Figure 2. Release of Evans blue into the luminal bathing solutions after gastric damage induced by 50% ethanol. In the control group, 0.05 M HCl in 0.2 M mannitol bathed the mucosa during periods 4-6. In the test groups, either 5% N-acetylcysteine or 0.5% pepsin was added to the control solution. Evans blue was administered intravenously (10 mgikg) 1 min after application of 50% ethanol to the chambered mucosa. Asterisks denote groups in which the leakage of Evans blue was significantly greater than in the corresponding period of the control group: * p < 0.05; ** p < 0.01; *** p < 0.001. Each bar represents the mean + SEM for four to five experiments.
the Alcian blue periods of the control experiments, binding capacity of the luminal solutions returned to the low levels observed in the first two periods. Addition of 0.5% pepsin to the luminal solutions added to the chamber in the first two periods resulted in a significant (p < 0.05) increase in Alcian blue binding compared with the control experiments. The release of mucus in response to 50% ethanol application was not significantly different from that observed in the control group. During the final three periods there was a highly significant (p < 0.001) increase in the levels of Alcian blue binding in the luminal solutions of 0.5% pepsin-treated mucosae compared with the corresponding periods in the control experiments. Evans Blue Leakage
Pepsin 62.5 51.8 76.1 36.8 37.7 43.4
* 2 2 2 * f
16.5” 11.9” 15.4 l.Zb 1.5b 2.7b
Data are expressed as the mean 2 SEM for four experiments. Groups that differ significantly from the control group a p < 0.05; b p < 0.001. The composition of the luminal solutions was: 0.3 M mannitol during the first two periods, 50% ethanol during the third period, and 0.05 M HCl in 0.2 M mannitol during the final three periods. In the test group, pepsin (0.5%) was added to the luminal solutions in each IO-min period except the third.
Evans blue leakage from the mucosa was observed within 1 min of its intravenous administration (during the period when ethanol was in the chamber). The major proportion of the dye was trapped beneath the layer of mucus that had been released after addition of 50% ethanol to the chamber. When breaks in the mucus cap occurred, it was released into the luminal bathing solution. Hence, the concentration of Evans blue in the recovered luminal solution could be used as a crude index of the mucolytic activity of the bathing solution. In the control group, some of the dye leaked into the
608
WALLACE AND WHITTLE
GASTROENTEROLOGY Vol. 91, No. 3
2.5 t-
*
??
2.0
N-Acetylcysteine
/n=6)
1.5
Hemoglobin leakage hd 1.0
1
2
3
4
5
6
TIME PERIOD
luminal solutions during periods 4-6 (Figure 2). The greatest leakage of Evans blue was observed when acid was first applied to the mucosa (period 4), with progressively less release in the ensuing periods. In the NAC and pepsin groups, there was significantly more release of Evans blue during these final two periods, coinciding with the observed disruption of the mucoid cap.
Figure
3. Release of hemoglobin into the luminal solutions in control and N-acetylcysteine experiments. In the test group, 5% N-acetylcysteine was added to the control solutions in each period except the third, when 50% ethanol was added to the chamber. See Materials and Methods for detailed description of the protocol for these experiments. Asterisks denote time periods in which the N-acetylcysteine group had significantly greater release of hemoglobin than the control group: * p < 0.05; * * p < 0.01.
The removal of the mucoid cap with forceps resulted in a significant increase in both hemoglobin and 14C-albumin leakage during the final four periods (Table 3). There was an immediate release of blood that had been trapped within the mucoid cap (period 3) as well as a second phase of hemoglobin release during the final two periods, which coincided with an observed renewal of bleeding from the ethanol-induced erosions.
Albumin and Hemoglobin Leakage During the first two periods there was a small amount of leakage of both 14C-albumin and hemoglobin (Figures 3 and 4). Addition of NAC to the control solutions caused a slight, significant (p < 11.05) increase in 14C-albumin leakage in period 2, but had no effect on hemoglobin leakage. The application of 50% ethanol to the chamber in period 3 caused a marked increase in both 14C-albumin and hemoglobin leakage in the control and NAC groups. 14Calbumin release was further increased with the application of acid to the mucosa in period 4. In the control group, the release of 14C-albumin decreased toward control levels in the final two periods. In the NAC group, there was significantly more 14C-albumin released into the bathing solutions during the final two periods. While acid was present in the chamber in periods 4-6, the leakage of hemoglobin in the control group decreased. In fact, the hemoglobin release in periods 5 and 6 was not significantly different from that observed before ethanol-induced damage. In the NAC group the hemoglobin leakage increased sharply in periods 5 and 6, corresponding to the observed dispersion of blood clots. This release of hemoglobin was significantly greater than that observed during the same periods in the control group.
Discussion The topical application of 50% ethanol to the chambered rat stomach resulted in a rapid release of copious mucus, plasma, and cellular debris which formed a continuous coat over the mucosal surface. The hypothesis that this mucoid cap plays an important role in the repair of acute damage (4) is supported by the results of the present study. The histologic data indicated that the mucolytic agents significantly retarded the restitution process. Both N-acetylcysteine and pepsin also caused a significant inhibition of the recovery of transmucosal PD after 50% ethanol-induced damage. The recovery of PD after damage has been reported to closely parallel the restitution process both in vitro and in vivo (6,610). Furthermore, the significant correlation between the histologic data and the PD data supports the hypothesis that PD recovery is a good index of recovery of epithelial integrity. Neither NAC nor pepsin increased the extent of epithelial damage induced by ethanol, as determined histologically, nor did they cause a change in PD or an increase in the hemorrhagic damage area. During the second period, NAC did cause a small, but significant (p < 0.05), increase in 14C-albumin leakage, but this increase was less than one-tenth
September
MUCUS AND RESTITUTION
1986
609
30
Figure
4. Release of Y-albumin into the luminal solutions in control and Nacetylcysteine (NAC) experiments. In the test group, 5% NAC was added to the control solutions in each period except the third, when 50% ethanol was added to the chamber. Asterisks denote time periods in which the NAC group had significantly greater release of radiolabeled albumin than the control group: * p < 0.05; ** p < 0.01; *** p < 0.001.
20 14C-Albumin
leakage(~1)
1
2
3
4
5
6
TIMEPERIOD
that caused by the application to the mucosa of 50% ethanol. These agents were selected for use in this study because they have been shown to have mucolytic properties in other models (14-17) and because their mechanisms of mucolytic action differ (18). Their mucolytic activity was confirmed in the present study by direct observations of dissolution of the mucoid cap, by their ability to increase the release of Evans blue trapped beneath the cap, and, for pepsin, by the Alcian blue binding studies. Our hypothesis that the mucolytic actions of NAC and pepsin accounted for their inhibition of epithelial recovery is supported by the finding that mechanical disruption of the mucoid cap with forceps caused a similar impairment of reepithelialization and inhibition of the recovery of PD. The data on Evans blue and 14C-albumin leakage support the hypothesis that the mucoid cap traps plasma-rich fluid. Morris and Wallace (6) and Ito and Lacy (19) noted the presence of fibrin in the mucus released following topical application of ethanol. This cap could provide a mixing barrier similar to that previously proposed as a model to explain the resistance of the mucosa to acid digestion (20). The relatively high pH of this cap could also provide a microenvironment conducive to the rapid repair of the epithelium by cell migration. The process of restitution was first noted in vivo after ethanolinduced mucosal damage (6). A previous study had noted the scarcity of epithelial damage 30 min after aspirin-induced mucosal damage (21). Similar observations have since been reported in other in vivo (1,19,22,23) and in vitro models (8,24,25) and this subject has recently been reviewed by Silen and Ito (26).It is possible that a relatively high pH at the mucosal surface may be prerequisite for restitution to occur. Vracko (5) proposed that an intact basal
lamina is necessary for epithelial cell migration to occur. Morris and Wallace (6) demonstrated that luminal acid destroyed the basal lamina in some regions of an ethanol-damaged mucosa and that restitution did not occur in those regions. Svanes et al. (27)reported that recovery of electrophysiologic parameters and epithelial continuity after damage induced by hyperosmolar sodium chloride was inhibited if the pH of the luminal solution was <3. Recently, Black et al. (7) further demonstrated the susceptibility of the exposed gastric basal lamina to acid injury, with extensive damage observed after a lo-min exposure to concentrations of acid of 0.02or 0.05 M. Hence, the effects of mucolytic agents on recovery of epithelial integrity in the present study
Table
Mechanical Disruption of the Mucoid Cap on MucosaJ Release of Hemoglobin and
3. Effects of
“C-Albumin Time (min] l-10
11-20 21-30 31-40 41-50 51-60
Hemoglobin Control 0.4k 0.3t 0.7k 0.9f 0.52 0.6k
0.1 0.1 0.1 0.3 0.3 0.2
(mg) Test
0.3k 0.1 0.32 0.1 4.0k 0.6" 1.12 0.3 1.82 O.Zb 2.1-r0.20
‘%-Albumin Control 2.4r+0.5 2.9t 0.7 13.8+ .I.5 23.5I!2.7 11.6t 1.3 6.7z!z 0.6
(~1) Test
3.1+ 0.8 3.2k 0.6 84.4z!z 21.9" 86.52 21.2' 55.72 15.0' 41.9? 12.4'
In the test group, the mucoid cap was peeled from the mucosa during min 28-30. Fifty percent ethanol was applied to the chambered mucosae during the third lo-min period. Results are expressed as the mean + SEM of seven experiments in the control group and five experiments in the test group. Hemoglobin release into the luminal solutions is expressed in terms of milligrams of hemoglobin released during the lo-min period. ‘%-albumin leakage is expressed in terms of microliters of serum. Groups that differ significantly from the corresponding control period: a p < 0.005; b p < 0.05; c p < 0.01.
GASTROENTEROLOGY Vol. 91. No. 3
610 WALLACE AND WHITTLE
could be explained by the disruption of the mucoid cap and the subsequent diffusion of acid to the denuded basal lamina in sufficient concentrations to cause damage, thereby retarding restitution. The mucoid cap may also protect the underlying epithelium from damage induced by other luminal agents. Lacy (28) recently reported that superficial mucosal damage induced by 70% ethanol was almost completely repaired 1 h later. Furthermore, such damage could not be reproduced by a second exposure to 70% ethanol unless the necrotic gelatinous layer (what we have referred to as the mucoid cap) produced by the first insult was removed. A possible corollary to the hypothesis tested in the present study is that agents that increase the effectiveness of a mucoid cap over damage may increase the recovery of epithelial integrity. This corollary is supported to some extent by our recent report (11) that 16,16-dimethyl prostaglandin Ea, an agent reported to stimulate mucus secretion and increase mucus gel thickness (14,16,20), significantly increased the recovery of epithelial integrity after damage induced by 50% ethanol. However, these effects of 16,16-dimethyl prostaglandin Ez may be unrelated to any effects of this agent on mucus (such effects have been disputed; see Reference 29) and perhaps may be more attributable to prevention of damage to the cells involved in the restitution process. Mucus may also play an important role in hemostasis following such acute gastric damage. The addition of NAC to the luminal solutions resulted in a significant increase in the release of hemoglobin by the chambered mucosa during the final 20 min of the experiments. It is likely that the increase in hemoglobin release was due to the mucolytic actions of NAC inasmuch as this compound has no effect on fibrin or normal blood clots (SO), but does disintegrate mixed blood and mucus clots (31). Furthermore, a similar increase in mucosal bleeding was observed when the mucoid cap was mechanically disrupted with forceps. The present study thus demonstrates that mucolytic agents can inhibit the recovery of epithelial integrity after acute gastric damage induced by ethanol. These results support the hypothesis that gastric mucus plays an important role in the repair of superficial damage (4). It is possible that such release of mucus after superficial damage induced by irritants or food present in the lumen may be a physiologic response aimed at limiting the extent of tissue damage and mucosal blood loss induced by such luminal agents and providing a favorable microenvironment for the repair of the damage through the process of restitution.
References 1. Morris GP, Harding PL. Mechanisms of mucosal recovery from acute gastric damage: the roles of extracellular mucus and cell migration. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984: 209-14. 2. McQueen S, Allen A, Garner A. Measurement of gastric and duodenal mucus gel thickness. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984:215-21. 3. Morris GP. The myth of the mucus barrier. Gastroenterol Clin Biol 1985;9:106-7. 4. Morris GP, Harding RK, Wallace JL. A functional model for extracellular gastric mucus in the rat. Virchows Arch [Cell Pathol] 1984;46:239-51. 5. Vracko R. Basal lamina scaffold: anatomy and significance for maintenance of orderly tissue structure: a review. Am J Path01 1974;77:314-38. 6. Morris GP, Wallace JL. The roles of ethanol and of acid in the production of gastric mucosal erosions in rats. Virchows Arch [Cell Pathol] 1981;38:23-38. 7. Black BA, Morris GP, Wallace JL. Effects of acid on the basal lamina of the rat stomach and duodenum. Virchows Arch [Cell Pathol] 1985;50:109-18. 8. Rutten MJ, Ito S. Morphology and electrophysiology of guinea pig gastric mucosal repair in vitro. Am J Physiol 1983;244: G171-82. 9. Mersereau WA, Hinchey EJ. Effects of gastric acidity on gastric ulceration induced by hemorrhage in the rat, utilizing a gastric chamber technique. Gastroenterology 1973;64: 1130-5. 10. Wallace JL, Morris GP, Krausse EJ, Greaves SE. Reduction by cytoprotective agents of ethanol-induced damage to the rat gastric mucosa: a correlated morphological and physiological study. Can J Physiol Pharmacol 1982;60:1686-99. 11. Wallace JL, Whittle BJR. Acceleration of recovery of epithelial integrity by 16,16-dimethyl prostaglandin EZ. Br J Pharmacol 1985;86:837-42. 12. Hall RL, Miller RJ, Peatfield AC, et al. A calorimetric assay for mucous glycoproteins using alcian blue. Biochem Sot Trans 1980;8:72. 13. Szabo S, Trier JS, Brown A, Schnoor J. Early vascular injury and increased vascular permeability in gastric mucosal injury caused by ethanol in the rat. Gastroenterology 1985;88: 228-36. 14. Bickel M, Kauffman GL. Gastric gel mucus thickness: effect of distention, 16,16-dimethyl prostaglandin E,, and carbenoxolone. Gastroenterology 1981;80:770-5. 15. Ross IR, Bahari HMM, Turnberg LA. The pH gradient across mucus adherent to rat fundic mucosa in vivo and the effect of potential damaging agents. Gastroenterology 1981;81:713-8. 16. Kerss S, Allen A, Garner A. A simple method for measuring thickness of the mucus gel layer adherent to rat, frog and human gastric mucosa: influence of feeding, prostaglandin, N-acetylcysteine and other agents. Clin Sci 1982;63:187-95. 17. Davenport HW. Protein-losing gastropathy produced by sulfhydryl reagents. Gastroenterology 1971;80:870-9. 18. Bell AE, Sellers LA, Allen A, Cunliffe WJ, Morris ER, RossMurphy SB. Properties of gastric and duodenal mucus: effects of proteolysis, disulfide reduction, bile, acid, ethanol, and hypertonicity on mucus gel structure. Gastroenterology 1985; 88:289-80. 19. Ito S, Lacy ER. Morphology of rat gastric mucosal damage,
September
20.
21.
22.
23.
24.
1986
defense, and restitution in the presence of luminal ethanol. Gastroenterology 1985;88:250-60. Allen A, Garner A. Mucus and bicarbonate secretion in the stomach and their possible role in mucosal protection. Gut 1980;21:249-62. Hingson DJ, Ito S. Effect of aspirin and related compounds on the fine structure of mouse gastric mucosa. Gastroenterology 1971;61:156-77. Lacy ER, Ito S. Ethanol-induced insult to the superficial gastric epithelium: a study of damage and rapid repair. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984:49-56. Tarnawski A, Hollander D, Stachura J, Krause WJ, Gergely H. Prostaglandin protection of the gastric mucosa against alcohol injury-a dynamic time-related process. Role of the mucosal proliferative zone. Gastroenterology 1985;88:334-52. Svanes K, Ito S, Takeuchi K, Silen W. Restitution of the surface epithelium of the in vitro frog gastric mucosa after damage with hyperosmolar sodium chloride. Gastroenterology 1982;82:1409-26.
MUCUS AND RESTITUTION
611
25. Critchlow J, Magee D, Ito S, Takeuchi K, Silen W. Requirements for restitution of the surface epithelium of frog stomach after mucosal injury. Gastroenterology 1985;88:23749. 26. Silen W, Ito S. Mechanisms for rapid re-epithelialization of the gastric mucosal surface. Ann Rev Physiol 1985;47:217-29. 27. Svanes K, Critchlow J, Takeuchi K, Magee D, Ito S, Silen W. Factors influencing reconstitution of frog gastric mucosa: role of prostaglandins. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984:33-8. 28. Lacy ER. Gastric mucosal resistance to a repeated ethanol insult. Stand J Gastroenterol 1985;2O(Suppl 110):63-72. 29. Robert A, Bottcher W, Golanska E, Kauffman GL. Lack of correlation between mucus gel thickness and gastric cytoprotection in rats. Gastroenterology 1984;86:6704. 30. Swinyard EA, Pathak MA. Surface-acting drugs. In: Gilman AG, Goodman LS, Gilman A, eds. The pharmacological basis of therapeutics. 6th ed. New York: Macmillan, 1980:951-63. 31. Risack LE. Vandevelde ME, Gobert JG. Effect of thiol derivatives on mixed mucus and blood clots in vitro. Resuscitation 1978;6:9-20.
Role of Mucus in the Repair of Gastric Epithelial Damage in the Rat Inhibition Agents
of Epithelial
Recovery
by Mucolytic
JOHN L. WALLACE and BRENDAN J. R. WHITTLE Department England
of Mediator Pharmacology,
Wellcome Research Laboratories,
A role for mucus in providing a microenvironment over sites of gastric damage, which is conducive to reepithelialization, has been proposed. We tested this hypothesis by examining the effects of disruption of such mucus on the recovery of epithelial integrity after damage induced by 50% ethanol. Exposure of an ex vivo chambered gastric mucosa to topically applied 50% ethanol resulted in copious release ofmucus, cellular debris, and plasma, which formed a continuous cap over the mucosal surface. Ethanol-induced gastric damage was accompanied by extensive surface epithelial cell damage and a marked decrease in transmucosal potential difference. During the 30 min after ethanol was removed from the chamber, the epithelium became reestablished and the potential difference gradually recovered to 94% of the level before ethanol treatment. However, if the mucolytic agents N-acetylcysteine (5%) or pepsin (0.5%) were added to the bathing solutions, the “mucoid cap” disintegrated and the recovery of potential difference was significantly retarded (recovering to only 51% and 52% of levels before ethanol treatment). Histologic evaluation confirmed that mucosae treated with either agent had significantly less [p < 0.005) intact epithelium at the end of the experiment. Removal of the mucoid cap with forceps caused a similar inhibition of the repair of the epithelium and the recovery of potenReceived August 29, 1985. Accepted February 28, 1986. Address requests for reprints to: Dr. J. L. Wallace, Department of Physiology, Queen’s University, Kingston, Ontario, K7L 3N6, Canada. Dr. Wallace is the recipient of a fellowship from the Medical Research Council of Canada. Part of these data were presented at the 86th Annual Meeting of the American Gastroenterologicai Association in New York, in May 1985. 0 1986 by the American Gastroenterological Association 0016-5085/86/$3.50
Beckenham,
Kent,
tial difference. Both mechanical and chemical (Nacetylcysteine) disruption of the mucoid cap resulted in a significant increase in the mucosal leakage of albumin and hemoglobin, supporting previous histologic evidence that the mucoid cap traps blood components over the damaged mucosa. These studies support the hypothesis that mucus released in response to topical application of an irritant plays an important role in the repair of epithelial damage through the process of restitution. The role of mucus in the prevention of gastric damage remains controversial (l-3), and it has been suggested that a more important role for mucus may be in the repair of acute damage. Using morphologic techniques, Morris et al. (4) demonstrated that there was a discontinuous coat of mucus on the undamaged rat stomach and a profound release of gel mucus in response to the topical application of an ulcerogenic agent. This gelatinous mucus formed a cap over the damaged region, which along with the entrapped cellulaf debris and plasma, it was proposed, could provide an environment favorable for the process of reepithelialization (restitution) to occur. Rapid repair of the epithelium appears to be dependent upon an intact basal lamina along which the cells migrate (5,6). A relatively high pH may be necessary for the maintenance of the integrity of the basal lamina, since this structure is readily degraded by acid (6,7). If mucus does play a role in providing a microenvironment over damaged regions that is conducive to the restitution process, then disruption of the mucoid cap should allow luminal acid to penetrate to and damage the basal lamina, thus impairing restitution. The present study was performed to Abbreviation
used
in this paper: NAC, 5% N-acetylcysteine.
604 WALLACE AND WHITTLE
determine the effects of disruption of the mucoid cap, either directly or by dissolution with mucolytic agents, on the repair of the gastric epithelium after damage induced by 50% ethanol. The extent of gastric epithelial damage before and after application of 50% ethanol was assessed histologically. The recovery toward control levels of transmucosal potential difference was used as an additional index of recovery of epithelial integrity, since this has been shown to closely parallel the restitution process (6,8). As a functional mucoid cap would trap plasma and blood over a damaged region, mucosal leakage of albumin, Evans blue, and hemoglobin were also measured.
Materials and Methods Male Wistar rats weighing 225-250 g were used. The rats were deprived of food, but not water, for 18-20 h before an experiment. Under pentobarbitone anesthesia (60 mg/kg i.p.) an ex vivo gastric chamber, which allowed direct observation of the gastric’ mucosa throughout the experiment, was prepared as previously described in detail (9,lO). Each experiment consisted of six sequential IO-min periods. At the beginning of each period 5.0 ml of a prewarmed (37%) solution was added to the chamber. While in the chamber, the solution was stirred by a glass paddle which turned at -200 rpm and was placed 1 cm above the mucosal surface. At the end of the period, the solution was removed from the chamber by a syringe. In the control group, 0.3 M mannitol was added to the chamber in the first two periods. In the third period 50% ethanol (vol/vol) was added to the chamber. During periods 4-6, the luminal bathing solution was 0.05 M hydrochloric acid in 0.2 M mannitol. In the test groups, either 5% N-acetylcysteine (NAC) or 0.5% pepsin (both wt/vol; Sigma Chemical Co., St. Louis, MO.) were added to the chamber in each period except the third. These compounds were dissolved in the corresponding control solution. Transmucosal potential difference was measured continuously throughout each experiment using apparatus described previously (11). At the end of the final period the exposed mucosa was photographed on color transparency film. Macroscopically visible damage was quantified planimetrically from these transparencies in a randomized, blind manner (10). The gastric pedacle was cut and the mucosa was immersed in neutral buffered formalin. Samples (10 X 2 mm) of the fundic (corpus) region of the stomach were excised and processed by routine techniques before embedding in paraffin. These samples were taken from the same regions of all mucosae; i.e., the midfundus of both the ventral and dorsal sides of the stomach. Sections (4-6 pm) were stained with hematoxylin and eosin and mounted on glass slides. Additional experiments [n = 4 per group) were performed in which the mucosa was fixed at the end of period 2 or period 3. These experiments were performed to determine the extent of epithelial damage before and immediately after application of 50% ethanol. The histo-
GASTROENTEROLOGY Vol. 91. No. 3
logic analysis was performed in a randomized manner, with the slides coded to avoid observer bias. The sections of fundic mucosa were examined and the percentage of the length of the section with an intact epithelium was calculated. Three sections (each 10 mm long) from each stomach were examined, and the overall score for the stomach was taken as the mean of the scores for each section. Regions on the sections in which hemorrhagic erosions were present were excluded from the calculation. Mucus release into the luminal solutions was determined using the Alcian blue method (12). After mixing the luminal solutions, a l.O-ml sample was added to 100 ~1 of Alcian blue reagent (10 mg/ml; Sigma) and incubated at room temperature for 12 h. The samples were then centrifuged (5 min, 1000 g) and the Alcian blue concentration in the supernatant was measured spectrophotometrically (absorbance at 620 nm). There is a linear relationship between the amount of Alcian blue precipitated and the concentration of mucous glycoprotein in the sample (12). Results are expressed as the number of micrograms of Alcian blue precipitated by 1.0 ml of sample. As N-acetylcysteine itself binds to Alcian blue, these determinations could not be performed with samples from that series of experiments. Additional experiments were performed in which Evans blue (10 mg/kg in saline; Phase Separations Ltd., Queensferry, United Kingdom) was administered intravenously 1 min after application of 50% ethanol to the chambered mucosa in period 3. Mucosal release of Evans blue following intragastric administration of 50% ethanol has previously been described (13). If mucus released in response to ethanol application did form a cap over the damaged area, then the dye should be trapped beneath such a cap. Furthermore, subsequent application of mucolytic agents should result in release of Evans blue into the luminal solution. Samples of the luminal solutions in periods 4-6 were analyzed for Evans blue concentration using a spectrophotometer (absorbance at 600 nm). It was possible that any effects of N-acetylcysteine or pepsin on epithelial intregity or recovery of potential difference might not be due to their mucolytic actions, but rather to a direct inhibition of the reepithelialization process. For instance, these substances might themselves damage the cells migrating from the gastric pits. Because of this possibility, a series of experiments was performed (n = 5) in which the mucoid cap that formed after application of 50% ethanol was removed with forceps. This procedure was performed taking meticulous care not to touch the underlying mucosa and was facilitated by the use of a stereomicroscope (Zeiss Opti 1, Carl Zeiss Ltd., Federal Republic of Germany) mounted over the chamber preparation. The mucus was removed 8 min after addition to the chamber of 50% ethanol (period 3) and took -2 min to complete. During the ensuing periods, 0.05 M HCl was added to the chamber, as in the control experiments. In the N-acetylcysteine experiments and the experiments in which the mucoid cap was removed with forceps, albumin and hemoglobin leakage from the chambered mucosa were measured. At the beginning of these experiments, 0.5 &i of 14C-labeled bovine serum albumin (50 &i/mg; Amersham International, Aylesbury, U.K.) dissolved in 0.25 ml of normal saline was administered
MUCUS AND RESTITUTION
September 1966
intravenously via a cannula in a femoral vein. At the end of each period, an aliquot (1.0 ml) of the luminal bathing solution was transferred to a scintillation vial and counted on a p spectrometer. Following the final period of the experiment, a sample of blood from a carotid artery (1.0 ml) was centrifuged (1 min; 9000 g) and the counts per minute in 100 ~1 of the plasma was determined. This was used to calculate the volume of plasma present in each sample of the luminal bathing solutions. Hemoglobin concentrations in l.O-ml samples of the luminal bathing solutions were measured spectrophotometrically. A few drops of a lysing reagent (Zaponin, Clay Adams, Parsippany, N.J.) were added to each sample. Absorbance at 540 nm was determined for each sample and was compared to a standard curve based on bovine hemoglobin [Sigma) for determination of the number of milligrams of hemoglobin per milliliter of sample. The accuracy of this spectrophotometric method was verified by checking the hemoglobin concentration of some samples on a hematology analyzer (Clay Adams). Statistical
1. Gastric Surface Epithelial Cell Damage Before, During, and After Application of 50% Ethanol: Effects of Mucolytic Agents Gastric luminal surface with intact epithelium (%I
Control NAC Peosin
Before ethanol
During ethanol
98.8 + 0.7 98.5 t 0.5 98.1 t 0.6
19.2 ? 5.2 21.3 lr 4.9 16.4 '_ 4.0
After ethanol 77.0 t 3.1 41.2 + 5.0" 59.0 r 1.0"
The percentage of the luminal surface with an intact epithelium was measured on coded slides using a light microscope. Regions in which hemorrhagic erosions were present were excluded from these calculations. These data are expressed as the mean ? SEM of 6-14 samples. The tissue samples were taken at 20 min (before 50% ethanol), 29 min (during the ninth minute of the period in which 50% ethanol was in the chamber), or 60 min (30 min after removal of 50% ethanol from the chamber). Groups that differ significantly from the control group: ” p < 0.001; b p < 0.005. NAC, 5% N-acetylcysteine; pepsin, 0.5% pepsin.
Analysis
All data are expressed as mean * SEM. Comparisons between groups were performed using the Student’s two-tailed t-test for unpaired data. To compare the relationship between percentage epithelial damage and the final potential difference value for each mucosa, linear regression analysis was performed. With all statistical analyses, an associated probability (p value) of ~5% was considered as significant.
Results Macroscopic
Table
605
Observations
During the first two periods of each experiment, the mucosa appeared normal and there were strands of mucus visible on the mucosal surface. The application of NAC or 0.5% pepsin during these periods had no visible effect on mucosal appearance. The addition of 50% ethanol to the chamber in the third period resulted in the development of focal patches of hyperemia. Extensive release of “mucus” (in fact, a mixture of sulfated and nonsulfated mucus, exfoliated cells, blood, and plasma components; see Reference 6) was evident during this period. This mucus formed a continuous coat over the mucosal surface, which we have subsequently referred to as the mucoid cap. Despite constant stirring of the solution in the chamber, the mucoid cap remained adherent to the mucosa. When 0.05 M hydrochloric acid was added to the chamber in periods 4-6, the hyperemic regions developed into hemorrhagic erosions. In the control group, these hemorrhagic erosions covered 16.2% -+: 2.6% (n = 6) of the glandular mucosa (determined at the end of the experiments). The extent of hemorrhagic damage in the groups treated with NAC (20.8% * 3.6%; n = 6) or pepsin (15.8% + 3.0%; n = 5) was not significantly different
from the control group. Some breaks in the mucoid cap became apparent during the final 30 min of the control experiments. In the NAC and pepsin groups, the dissolution of the mucoid cap was much more obvious. Large clumps of mucus became detached from the mucosa and were visible in the bathing solution. This mucolytic effect was particularly evident in the final 15 min of the experiments. During this time, blood clots over hemorrhagic erosions dispersed and frank bleeding into the bathing solution was observed.
Histologic
Observations
In the samples fixed at the end of the second period (i.e., before application of 50% ethanol) the mucosa appeared normal and the surface epithelium was almost completely intact (Table 1).The addition of NAC or pepsin to the luminal solutions in the first two periods did not cause any detectable damage. Application of 50% ethanol to the chambered mucosae resulted in extensive destruction of the epithelium with the resulting release of intracellular mucus. This epithelial damage involved -80% of the luminal surface, and the extent of this damage was not significantly affected by prior exposure to NAC or pepsin (Table 1). A layer of mucus, cellular debris, and blood was visible along the damaged portions of the luminal surface. In samples fixed at the end of the 60-min experiments, there were marked differences between the extent of intact epithelium in the four groups. In the control group, an average of 77% of the luminal surface had an intact epithelial cell layer, significantly (p < 0.005) more than in the NAC or pepsin
606
WALLACE AND WHITTLE
GASTROENTEROLOGY Vol. 91, No. 3
A
Mannitol
f NAC
50% ethanol
HCI + NAC
Potential difference (mV)
I
20
I
I
I
I
30
40
50
60
TIME (minutes)
---
Pepsin
Potentia! difference (mV)
20
10
40
30
50
60
TIME (minutes)
C
T
T
----
Control Mucoid cap removed (1)
Potential difference (mV)
0
10
20
30
I
I
,
40
50
60
TIME (minutes)
Figure
1. Transmucosal potential difference recordings for chambered gastric mucosae damaged with 50% ethanol. Each point represents the mean f SEM of a 5-min interval (n = 5-8 per group]. Asterisks denote points at which the test group differs significantly (* p < 0.05; ** p < 0.01) from the control group. A. In the test group, 5% N-acetylcysteine (NAC) was added to the chamber in each period except the period in which 50% ethanol was applied (minutes 21-30). B. The protocol for these experiments was the same as shown at the top of (A) except that 0.5% pepsin was added to the chamber in place of NAC. C. In the test group, the solutions added to the chamber were identical to those in the control group. Under a stereomicroscope, the mucoid cap that formed after application of 50% ethanol was meticulously peeled off the underlying mucosa at 28 min (arrow).
groups or the group in which the mucoid cap was removed with forceps (53.0% + 5.3%; p < 0.005). Transmucosal
Potential
Difierence
During the first two peripds of the control experiments the transmucosal potential difference
(PD) remained between -38 and -48 mV (Figure 1). The addition of NAC or pepsin to the control solutions had no effect on the PD. The addition of 50% ethanol to the chamber in period 3 elicited a sharp decrease in PD to --lo mV in all four groups (no significant differences). The PD remained depressed throughout the period in which ethanol was in the
September
MUCUS AND RESTITUTION
1966
chamber. During the final 30 min of the control experiments, the PD gradually recovered toward levels before ethanol treatment. At the end of the experiment the PD was not significantly different from that immediately before the application of ethanol to the mucosa. In the NAC and pepsin groups the recovery of PD was much slower, with significantly lower values than the control group at all time points during the final 30 min. By the end of the experiment the PD in the NAC and pepsin groups had recovered to only 51% + 4% and 52% + 7%, respectively, of the pre-ethanol-treatment level, both significantly lower (p < 0.01) than in the control group (94% t 6%). Direct removal of the mucoid cap with forceps had no immediate effect on PD, but had profound effects on the recovery of PD in the 30 min after removal of ethanol from the chamber (Figure 1C). In each period after removal of the cap, the PD was significantly lower (p < 0.05) than in the control group. Linear regression analysis of the data from the control, NAC, and pepsin groups revealed that there was a significant correlation between the transmucosal potential difference at the end of the experiment and the percentage of the luminal surface with an intact epithelium (r = 0.75; p < 0.01). Alcian Blue Binding During the first two periods there was a small amount of Alcian blue binding activity in the luminal solutions from control experiments (Table 2). However, the application of 50% ethanol during period 3 resulted in a marked increase (approximately sixfold; p < 0.001) in the level of mucus present in the luminal solutions. These data are in agreement with the direct observations of macroscopically visible mucus release in response to the application of 50% ethanol. During the final three Table
2. Alcian Blue Binding (~glml) in Luminal Solutions From Control and 0.5% Pepsin-
Treated Gastric Mucosae Time (min) l-10 11-20 21-30 31-40 41-50 51-60
Control 16.9 15.7 93.4 24.3 16.9 15.2
f ? t r C *
7.0 11.1 1.1 0.8 2.1 1.9
607
30
20
10
0 6
5 TIME
PERIOD
(10mm
1
Figure 2. Release of Evans blue into the luminal bathing solutions after gastric damage induced by 50% ethanol. In the control group, 0.05 M HCl in 0.2 M mannitol bathed the mucosa during periods 4-6. In the test groups, either 5% N-acetylcysteine or 0.5% pepsin was added to the control solution. Evans blue was administered intravenously (10 mgikg) 1 min after application of 50% ethanol to the chambered mucosa. Asterisks denote groups in which the leakage of Evans blue was significantly greater than in the corresponding period of the control group: * p < 0.05; ** p < 0.01; *** p < 0.001. Each bar represents the mean + SEM for four to five experiments.
the Alcian blue periods of the control experiments, binding capacity of the luminal solutions returned to the low levels observed in the first two periods. Addition of 0.5% pepsin to the luminal solutions added to the chamber in the first two periods resulted in a significant (p < 0.05) increase in Alcian blue binding compared with the control experiments. The release of mucus in response to 50% ethanol application was not significantly different from that observed in the control group. During the final three periods there was a highly significant (p < 0.001) increase in the levels of Alcian blue binding in the luminal solutions of 0.5% pepsin-treated mucosae compared with the corresponding periods in the control experiments. Evans Blue Leakage
Pepsin 62.5 51.8 76.1 36.8 37.7 43.4
* 2 2 2 * f
16.5” 11.9” 15.4 l.Zb 1.5b 2.7b
Data are expressed as the mean 2 SEM for four experiments. Groups that differ significantly from the control group a p < 0.05; b p < 0.001. The composition of the luminal solutions was: 0.3 M mannitol during the first two periods, 50% ethanol during the third period, and 0.05 M HCl in 0.2 M mannitol during the final three periods. In the test group, pepsin (0.5%) was added to the luminal solutions in each IO-min period except the third.
Evans blue leakage from the mucosa was observed within 1 min of its intravenous administration (during the period when ethanol was in the chamber). The major proportion of the dye was trapped beneath the layer of mucus that had been released after addition of 50% ethanol to the chamber. When breaks in the mucus cap occurred, it was released into the luminal bathing solution. Hence, the concentration of Evans blue in the recovered luminal solution could be used as a crude index of the mucolytic activity of the bathing solution. In the control group, some of the dye leaked into the
608
WALLACE AND WHITTLE
GASTROENTEROLOGY Vol. 91, No. 3
2.5 t-
*
??
2.0
N-Acetylcysteine
/n=6)
1.5
Hemoglobin leakage hd 1.0
1
2
3
4
5
6
TIME PERIOD
luminal solutions during periods 4-6 (Figure 2). The greatest leakage of Evans blue was observed when acid was first applied to the mucosa (period 4), with progressively less release in the ensuing periods. In the NAC and pepsin groups, there was significantly more release of Evans blue during these final two periods, coinciding with the observed disruption of the mucoid cap.
Figure
3. Release of hemoglobin into the luminal solutions in control and N-acetylcysteine experiments. In the test group, 5% N-acetylcysteine was added to the control solutions in each period except the third, when 50% ethanol was added to the chamber. See Materials and Methods for detailed description of the protocol for these experiments. Asterisks denote time periods in which the N-acetylcysteine group had significantly greater release of hemoglobin than the control group: * p < 0.05; * * p < 0.01.
The removal of the mucoid cap with forceps resulted in a significant increase in both hemoglobin and 14C-albumin leakage during the final four periods (Table 3). There was an immediate release of blood that had been trapped within the mucoid cap (period 3) as well as a second phase of hemoglobin release during the final two periods, which coincided with an observed renewal of bleeding from the ethanol-induced erosions.
Albumin and Hemoglobin Leakage During the first two periods there was a small amount of leakage of both 14C-albumin and hemoglobin (Figures 3 and 4). Addition of NAC to the control solutions caused a slight, significant (p < 11.05) increase in 14C-albumin leakage in period 2, but had no effect on hemoglobin leakage. The application of 50% ethanol to the chamber in period 3 caused a marked increase in both 14C-albumin and hemoglobin leakage in the control and NAC groups. 14Calbumin release was further increased with the application of acid to the mucosa in period 4. In the control group, the release of 14C-albumin decreased toward control levels in the final two periods. In the NAC group, there was significantly more 14C-albumin released into the bathing solutions during the final two periods. While acid was present in the chamber in periods 4-6, the leakage of hemoglobin in the control group decreased. In fact, the hemoglobin release in periods 5 and 6 was not significantly different from that observed before ethanol-induced damage. In the NAC group the hemoglobin leakage increased sharply in periods 5 and 6, corresponding to the observed dispersion of blood clots. This release of hemoglobin was significantly greater than that observed during the same periods in the control group.
Discussion The topical application of 50% ethanol to the chambered rat stomach resulted in a rapid release of copious mucus, plasma, and cellular debris which formed a continuous coat over the mucosal surface. The hypothesis that this mucoid cap plays an important role in the repair of acute damage (4) is supported by the results of the present study. The histologic data indicated that the mucolytic agents significantly retarded the restitution process. Both N-acetylcysteine and pepsin also caused a significant inhibition of the recovery of transmucosal PD after 50% ethanol-induced damage. The recovery of PD after damage has been reported to closely parallel the restitution process both in vitro and in vivo (6,610). Furthermore, the significant correlation between the histologic data and the PD data supports the hypothesis that PD recovery is a good index of recovery of epithelial integrity. Neither NAC nor pepsin increased the extent of epithelial damage induced by ethanol, as determined histologically, nor did they cause a change in PD or an increase in the hemorrhagic damage area. During the second period, NAC did cause a small, but significant (p < 0.05), increase in 14C-albumin leakage, but this increase was less than one-tenth
September
MUCUS AND RESTITUTION
1986
609
30
Figure
4. Release of Y-albumin into the luminal solutions in control and Nacetylcysteine (NAC) experiments. In the test group, 5% NAC was added to the control solutions in each period except the third, when 50% ethanol was added to the chamber. Asterisks denote time periods in which the NAC group had significantly greater release of radiolabeled albumin than the control group: * p < 0.05; ** p < 0.01; *** p < 0.001.
20 14C-Albumin
leakage(~1)
1
2
3
4
5
6
TIMEPERIOD
that caused by the application to the mucosa of 50% ethanol. These agents were selected for use in this study because they have been shown to have mucolytic properties in other models (14-17) and because their mechanisms of mucolytic action differ (18). Their mucolytic activity was confirmed in the present study by direct observations of dissolution of the mucoid cap, by their ability to increase the release of Evans blue trapped beneath the cap, and, for pepsin, by the Alcian blue binding studies. Our hypothesis that the mucolytic actions of NAC and pepsin accounted for their inhibition of epithelial recovery is supported by the finding that mechanical disruption of the mucoid cap with forceps caused a similar impairment of reepithelialization and inhibition of the recovery of PD. The data on Evans blue and 14C-albumin leakage support the hypothesis that the mucoid cap traps plasma-rich fluid. Morris and Wallace (6) and Ito and Lacy (19) noted the presence of fibrin in the mucus released following topical application of ethanol. This cap could provide a mixing barrier similar to that previously proposed as a model to explain the resistance of the mucosa to acid digestion (20). The relatively high pH of this cap could also provide a microenvironment conducive to the rapid repair of the epithelium by cell migration. The process of restitution was first noted in vivo after ethanolinduced mucosal damage (6). A previous study had noted the scarcity of epithelial damage 30 min after aspirin-induced mucosal damage (21). Similar observations have since been reported in other in vivo (1,19,22,23) and in vitro models (8,24,25) and this subject has recently been reviewed by Silen and Ito (26).It is possible that a relatively high pH at the mucosal surface may be prerequisite for restitution to occur. Vracko (5) proposed that an intact basal
lamina is necessary for epithelial cell migration to occur. Morris and Wallace (6) demonstrated that luminal acid destroyed the basal lamina in some regions of an ethanol-damaged mucosa and that restitution did not occur in those regions. Svanes et al. (27)reported that recovery of electrophysiologic parameters and epithelial continuity after damage induced by hyperosmolar sodium chloride was inhibited if the pH of the luminal solution was <3. Recently, Black et al. (7) further demonstrated the susceptibility of the exposed gastric basal lamina to acid injury, with extensive damage observed after a lo-min exposure to concentrations of acid of 0.02or 0.05 M. Hence, the effects of mucolytic agents on recovery of epithelial integrity in the present study
Table
Mechanical Disruption of the Mucoid Cap on MucosaJ Release of Hemoglobin and
3. Effects of
“C-Albumin Time (min] l-10
11-20 21-30 31-40 41-50 51-60
Hemoglobin Control 0.4k 0.3t 0.7k 0.9f 0.52 0.6k
0.1 0.1 0.1 0.3 0.3 0.2
(mg) Test
0.3k 0.1 0.32 0.1 4.0k 0.6" 1.12 0.3 1.82 O.Zb 2.1-r0.20
‘%-Albumin Control 2.4r+0.5 2.9t 0.7 13.8+ .I.5 23.5I!2.7 11.6t 1.3 6.7z!z 0.6
(~1) Test
3.1+ 0.8 3.2k 0.6 84.4z!z 21.9" 86.52 21.2' 55.72 15.0' 41.9? 12.4'
In the test group, the mucoid cap was peeled from the mucosa during min 28-30. Fifty percent ethanol was applied to the chambered mucosae during the third lo-min period. Results are expressed as the mean + SEM of seven experiments in the control group and five experiments in the test group. Hemoglobin release into the luminal solutions is expressed in terms of milligrams of hemoglobin released during the lo-min period. ‘%-albumin leakage is expressed in terms of microliters of serum. Groups that differ significantly from the corresponding control period: a p < 0.005; b p < 0.05; c p < 0.01.
GASTROENTEROLOGY Vol. 91. No. 3
610 WALLACE AND WHITTLE
could be explained by the disruption of the mucoid cap and the subsequent diffusion of acid to the denuded basal lamina in sufficient concentrations to cause damage, thereby retarding restitution. The mucoid cap may also protect the underlying epithelium from damage induced by other luminal agents. Lacy (28) recently reported that superficial mucosal damage induced by 70% ethanol was almost completely repaired 1 h later. Furthermore, such damage could not be reproduced by a second exposure to 70% ethanol unless the necrotic gelatinous layer (what we have referred to as the mucoid cap) produced by the first insult was removed. A possible corollary to the hypothesis tested in the present study is that agents that increase the effectiveness of a mucoid cap over damage may increase the recovery of epithelial integrity. This corollary is supported to some extent by our recent report (11) that 16,16-dimethyl prostaglandin Ea, an agent reported to stimulate mucus secretion and increase mucus gel thickness (14,16,20), significantly increased the recovery of epithelial integrity after damage induced by 50% ethanol. However, these effects of 16,16-dimethyl prostaglandin Ez may be unrelated to any effects of this agent on mucus (such effects have been disputed; see Reference 29) and perhaps may be more attributable to prevention of damage to the cells involved in the restitution process. Mucus may also play an important role in hemostasis following such acute gastric damage. The addition of NAC to the luminal solutions resulted in a significant increase in the release of hemoglobin by the chambered mucosa during the final 20 min of the experiments. It is likely that the increase in hemoglobin release was due to the mucolytic actions of NAC inasmuch as this compound has no effect on fibrin or normal blood clots (SO), but does disintegrate mixed blood and mucus clots (31). Furthermore, a similar increase in mucosal bleeding was observed when the mucoid cap was mechanically disrupted with forceps. The present study thus demonstrates that mucolytic agents can inhibit the recovery of epithelial integrity after acute gastric damage induced by ethanol. These results support the hypothesis that gastric mucus plays an important role in the repair of superficial damage (4). It is possible that such release of mucus after superficial damage induced by irritants or food present in the lumen may be a physiologic response aimed at limiting the extent of tissue damage and mucosal blood loss induced by such luminal agents and providing a favorable microenvironment for the repair of the damage through the process of restitution.
References 1. Morris GP, Harding PL. Mechanisms of mucosal recovery from acute gastric damage: the roles of extracellular mucus and cell migration. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984: 209-14. 2. McQueen S, Allen A, Garner A. Measurement of gastric and duodenal mucus gel thickness. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984:215-21. 3. Morris GP. The myth of the mucus barrier. Gastroenterol Clin Biol 1985;9:106-7. 4. Morris GP, Harding RK, Wallace JL. A functional model for extracellular gastric mucus in the rat. Virchows Arch [Cell Pathol] 1984;46:239-51. 5. Vracko R. Basal lamina scaffold: anatomy and significance for maintenance of orderly tissue structure: a review. Am J Path01 1974;77:314-38. 6. Morris GP, Wallace JL. The roles of ethanol and of acid in the production of gastric mucosal erosions in rats. Virchows Arch [Cell Pathol] 1981;38:23-38. 7. Black BA, Morris GP, Wallace JL. Effects of acid on the basal lamina of the rat stomach and duodenum. Virchows Arch [Cell Pathol] 1985;50:109-18. 8. Rutten MJ, Ito S. Morphology and electrophysiology of guinea pig gastric mucosal repair in vitro. Am J Physiol 1983;244: G171-82. 9. Mersereau WA, Hinchey EJ. Effects of gastric acidity on gastric ulceration induced by hemorrhage in the rat, utilizing a gastric chamber technique. Gastroenterology 1973;64: 1130-5. 10. Wallace JL, Morris GP, Krausse EJ, Greaves SE. Reduction by cytoprotective agents of ethanol-induced damage to the rat gastric mucosa: a correlated morphological and physiological study. Can J Physiol Pharmacol 1982;60:1686-99. 11. Wallace JL, Whittle BJR. Acceleration of recovery of epithelial integrity by 16,16-dimethyl prostaglandin EZ. Br J Pharmacol 1985;86:837-42. 12. Hall RL, Miller RJ, Peatfield AC, et al. A calorimetric assay for mucous glycoproteins using alcian blue. Biochem Sot Trans 1980;8:72. 13. Szabo S, Trier JS, Brown A, Schnoor J. Early vascular injury and increased vascular permeability in gastric mucosal injury caused by ethanol in the rat. Gastroenterology 1985;88: 228-36. 14. Bickel M, Kauffman GL. Gastric gel mucus thickness: effect of distention, 16,16-dimethyl prostaglandin E,, and carbenoxolone. Gastroenterology 1981;80:770-5. 15. Ross IR, Bahari HMM, Turnberg LA. The pH gradient across mucus adherent to rat fundic mucosa in vivo and the effect of potential damaging agents. Gastroenterology 1981;81:713-8. 16. Kerss S, Allen A, Garner A. A simple method for measuring thickness of the mucus gel layer adherent to rat, frog and human gastric mucosa: influence of feeding, prostaglandin, N-acetylcysteine and other agents. Clin Sci 1982;63:187-95. 17. Davenport HW. Protein-losing gastropathy produced by sulfhydryl reagents. Gastroenterology 1971;80:870-9. 18. Bell AE, Sellers LA, Allen A, Cunliffe WJ, Morris ER, RossMurphy SB. Properties of gastric and duodenal mucus: effects of proteolysis, disulfide reduction, bile, acid, ethanol, and hypertonicity on mucus gel structure. Gastroenterology 1985; 88:289-80. 19. Ito S, Lacy ER. Morphology of rat gastric mucosal damage,
September
20.
21.
22.
23.
24.
1986
defense, and restitution in the presence of luminal ethanol. Gastroenterology 1985;88:250-60. Allen A, Garner A. Mucus and bicarbonate secretion in the stomach and their possible role in mucosal protection. Gut 1980;21:249-62. Hingson DJ, Ito S. Effect of aspirin and related compounds on the fine structure of mouse gastric mucosa. Gastroenterology 1971;61:156-77. Lacy ER, Ito S. Ethanol-induced insult to the superficial gastric epithelium: a study of damage and rapid repair. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984:49-56. Tarnawski A, Hollander D, Stachura J, Krause WJ, Gergely H. Prostaglandin protection of the gastric mucosa against alcohol injury-a dynamic time-related process. Role of the mucosal proliferative zone. Gastroenterology 1985;88:334-52. Svanes K, Ito S, Takeuchi K, Silen W. Restitution of the surface epithelium of the in vitro frog gastric mucosa after damage with hyperosmolar sodium chloride. Gastroenterology 1982;82:1409-26.
MUCUS AND RESTITUTION
611
25. Critchlow J, Magee D, Ito S, Takeuchi K, Silen W. Requirements for restitution of the surface epithelium of frog stomach after mucosal injury. Gastroenterology 1985;88:23749. 26. Silen W, Ito S. Mechanisms for rapid re-epithelialization of the gastric mucosal surface. Ann Rev Physiol 1985;47:217-29. 27. Svanes K, Critchlow J, Takeuchi K, Magee D, Ito S, Silen W. Factors influencing reconstitution of frog gastric mucosa: role of prostaglandins. In: Allen A, Flemstrom G, Garner A, Silen W, Turnberg LA, eds. Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984:33-8. 28. Lacy ER. Gastric mucosal resistance to a repeated ethanol insult. Stand J Gastroenterol 1985;2O(Suppl 110):63-72. 29. Robert A, Bottcher W, Golanska E, Kauffman GL. Lack of correlation between mucus gel thickness and gastric cytoprotection in rats. Gastroenterology 1984;86:6704. 30. Swinyard EA, Pathak MA. Surface-acting drugs. In: Gilman AG, Goodman LS, Gilman A, eds. The pharmacological basis of therapeutics. 6th ed. New York: Macmillan, 1980:951-63. 31. Risack LE. Vandevelde ME, Gobert JG. Effect of thiol derivatives on mixed mucus and blood clots in vitro. Resuscitation 1978;6:9-20.