- Email: [email protected]
A comparative analysis of two models of colitis in rats
GASTROENTEROLOGY
1992;102:1524-1534
A Comparative Analysis of Two Models of Colitis in Rats TAMAKI YAMADA, STEVEN MARSHALL, and MATTHEW B. GRISHAM Departments of Physiology and Biophysics Medical Center, Shreveport, Louisiana
D. SPECIAN,
and Cellular Biology and Anatomy,
Two models of colitis produced in rats that have received significant attention over the past few years are the acetic acid and trinitrobenzene sulfonic acid (TNBS) models. The objective of this study was to quantify and compare the temporal relationship among mucosal permeability, epithelial injury, and inflammation induced by acetic acid, ethanol (vehicle), ethanol plus TNBS (unbuffered, pH l.O), and ethanol plus TNBS (pH 7.4). Data obtained show that the inflammation induced by these four irritants results from caustic injury to the colonic epithelium and interstitium as measured by the rapid and dramatic increases in mucosal permeability and tissue water content as well as by histological analysis. The injurious nature of TNBS was confirmed in a separate series of studies showing that buffered TNBS (pH 7.4), in the absence of ethanol, is toxic to cultured rat intestinal epithelial cell monolayers. Only after 1-2 days of the initial insult, were signs of classical inflammation observed, including increases in colonic myeloperoxidase activity (neutrophil infiltration) and colon weight as well as hyperemia and mucosal ulcerations. Although ethanol plus TNBS (pH 1.0 or 7.4) tended to produce higher mucosal permeabilities (epithelial cell injury) at l-2 weeks after the enemas than acetic acid or ethanol groups, only the ethanol plus TNBS (pH 7.4)permeabilities were found to be significantly enhanced. In addition, all four groups showed significant elevations in coionic myeloperoxidase activity and colon weight at 1-2 weeks after enema. It is suggested that these models of colitis are useful to study events that occur at the time of inflammation and repair. However, these models may have significant limitations in understanding events that initiate inflammation of the intestine in human inflammatory bowel disease. major limitation in investigating the pathogenetic mechanisms responsible for the mucosal injury observed in inflammatory bowel disease (IBD)
A
ROBERT
Louisiana State University
has been the relative paucity of relevant animal models. Two models of colitis produced in rats that have received significant attention over the past few years are the acetic acid and trinitrobenzene sulfonic acid (TNBS) models.‘-8 The mechanism by which acetic acid produces the diffuse colitis is thought to involve nonspecific, acid-induced injury to the colonic mucosa that is followed by an acute inflammatory response. Apparently, the protonated form of the acid is required to induce the colitis, because neither HCI (pH 2.3)nor sodium acetate (pH 7.0)are effective in eliciting the inflammatory response.gs’o However, there is some evidence to suggest that acetic acid may promote other pathophysiological events (e.g., fluid and electrolyte secretion) using noncytotoxic concentrations of the acid.‘l Recent studies show that intrarectal administration of the hapten TNBS, in the presence of a mucosal barrier breaker such as ethanol, produces an acute and possibly chronic colitis in unsensitized rats.5 The mechanism(s) by which buffered or unbuffered TNBS in the presence of ethanol initiates inflammation in unsensitized animals is unclear; however, it has been suggested to involve macrophage-mediated recognition and lysis of TNBS-modified autologous cells within the mucosa.12 Beck et al. reported that some of the colonic injury induced by ethanol plus TNBS (pH 1.0) may be attenuated by prior tolerization of the animals to TNBS, suggesting cell-mediated tissue injury.13 Although these data suggest an immune celldependent component of TNBS-induced colonic inflammation, the mechanism(s) for ethanol plus TNBS-induced injury and inflammation may be more complex than previously appreciated. For example, the barrier breaker ethanol is an extremely potent proinflammatory solvent a10ne.14 Furthermore, we have shown that TNBS is metabolized by certain colonic enzymes and substrates to yield both proinflammatory and cytotoxic oxidants that could 0 1992 by
the American Gastroenterological 0016-5085/92/$3.00
Association
COLONIC INFLAMMATION
May 1992
initiate colonic inflammation.15 Although both acetic acid and ethanol plus TNBS are widely used as models of colitis in rats, there has been no systematic study that quantitatively compares the temporal relationship between mucosal injury and inflammation in both models of colitis. The objective of this study was to quantify and compare the temporal relationship among mucosal permeability, epithelial injury, and inflammation induced by the intrarectal administration of acetic acid, ethanol (vehicle), ethanol plus TNBS (unbuffered, pH 1.0) and ethanol plus TNBS (pH 7.4). Materials and Methods Induction
of Colitis
Male Sprague-Dawley rats (300-350 g) were fasted overnight with access to water ad libitum. A total of 129 rats were randomized into five major groups consisting of a saline control group as well as an acetic acid, ethanol, ethanol plus TNBS (E + TNBS; unbuffered, pH l.O), and E + TNBS (buffered, pH 7.4) group. Diffuse colitis was induced with acetic acid as described previously.3 Briefly, the animals were lightly anesthetized with Ethrane (enflurane; Anaquest Co., Memphis, TN), and 1 mL of either saline (controls) or 4% (vol/vol) acetic acid (pH 2.3) was slowly administered into the lumen of the colon via a s-cm polyethylene tubing (PE-240) fitted onto a 1-mL syringe. After a 30-second period of exposure, excess fluid was withdrawn, and 1.5 mL phosphate-buffered saline (PBS) was introduced to flush the colon. Colitis was induced in a third set of rats by intrarectal administration of 0.25 mL 50% (vol/vol) ethanol using the l-mL syringe fitted with the 5-cm catheter. Finally, the fourth and fifth groups of anesthetized rats were treated intrarectally with 0.25 mL TNBS solution (120 mg/mL TNBS in 50% unbuffered ethanol, pH 1.0) as described by Morris et al.’ or a TNBS solution (120 mg/mL TNBS in 50% ethanol adjusted to pH 7.4 with NaOH).
Mucosal Permeability At various times after the enemas, rats were anesthetized with 120 mg/kg of Na-5-ethyl-l(l’-methyl-propyl)-2-thio-barbiturate (Inactin; Byk Gulden, Konstanz, Germany). The animals were tracheotomized, and the right femoral artery was cannulated for arterial pressure recording and blood sampling. The right femoral vein was also cannulated for the injection of radioisotope marker. A midline abdominal incision was performed. The colon was isolated and cannulated at both the splenic flexure and the rectum using Silastic tubing (ID, 0.025 and 0.25 in, respectively; Dow Corning). The luminal contents were removed by perfusion with warm (37°C) Tyrode’s solution. Both renal pedicles were ligated to prevent rapid excretion of the radioisotope marker into the urine. The isolated, perfused colon was returned to the abdominal cavity, and the abdominal wall was closed to minimize dehydration of the colon segment during the experiment. Body temperature was maintained at 37°C with a thermistor-controlled infra-
IN RATS
1525
red heat lamp. Mucosal permeability was determined using the blood-to-lumen clearance of [Yrlethylenediaminetetraacetic acid (EDTA) as described previous1y.‘6 [Yr]EDTA, 100 pCi, was injected via the femoral vein followed by a l-hour equilibration period, during which the intestinal lumen was perfused with Tyrode’s solution but no clearances measurements were taken. The perfusate was sampled every 10 minutes for 40 minutes for the appearance of [“CrJEDTA. Blood samples (0.3 mL) were taken at 40 minutes for use as reference counts. Blood-tolumen clearance of [“Cr]EDTA was calculated using the following formula:
Clearance
(EDTA) = ~‘~~
,
bl
where Cpe. and Cbl represent counts per minute per milliliter of [“Cr]EDTA in the luminal perfusate and blood, respectively; Q represents the luminal perfusion rate (0.5 mL/min), and W represents the dry weight of the perfused segment of the colon. Mucosal permeability was determined from the mean of four clearance values.
TNBS-Mediated
Epithelial
Cell Cytotoxicity
Normal rat epithelial cells (IEC-18) were purchased from American Type Culture Collection (Rockville, MD). Cells were grown and maintained using Dulbecco’s modified Eagle medium (DMEM) containing 4.5 g/L glucose, 5% fetal bovine serum, and 0.1 U/mL insulin. All studies used confluent monolayers that possessed 40-15 population doublings. Confluent monolayers (approximately 2 X lo5 cells/well) were incubated with DMEM containing 15 pCi [51Cr]NaCr0,/mL for 18 hours. The wells were washed (3X) with PBS (pH 7.4) to remove unincorporated radioactivity. ‘ICr-labeled monolayers were then overlayed with 0.5 mL Dulbecco’s phosphate-buffered saline (DPBS) or 0.5 mL of varying concentrations of TNBS in DPBS (pH 7.4) and incubated for 2 hours at 37°C. After the incubation period, the plate was centrifuged for 3 minutes at 100 X g, and the supernatant was removed and counted for 5*Cr activity. An aliquot (0.5 mL) of 2N NaOH was then added to each well to lyse the remaining cells. After a 60-minute incubation period, the fluid was removed and counted for Yr activity. Cytotoxicity was expressed as the percent specific cytotoxicity (51Cr release) calculated using the following formula:
% Specific Cytotoxity
A-B
= C_
X 100,
where A is the mean counts per min (cpm) in the supernatant of samples containing TNBS; B is the mean cpm in the supernatants of samples containing DPBS; and C is the total cpm available in each well (i.e., the sum of the supernatant plus lysate). Each data point represents the mean of triplicate determinations from two different experiments.
Tissue Analyses After the determinations of mucosal permeability, the animals were killed with pentobarbital, and the colons
1526
YAMADA ET AL.
GASTROENTEROLOGY
Table 1. Time Course Study Groups
ofGrossFindings:
Control Gross findings Acetic acid ETOH TNBS (pH 1.0) TNBS (pH 7.4) “P < *P < “P < dP <
0.05 0.05 0.05 0.05
Comparison
zh
0 -
1 wk
48 h
2.0f o.o* 2.4+ O.Zb 1.0I?0.0 0.8+ 0.2
5.0z!I 0.0" 3.2+ 0.3 4.8A O.Zd 3.8+ 0.8
zwk
2.0iz0.0 3.4f 0.2 3.3Ik0.9 2.6f 0.6
2.0+ 3.0+ 3.4f 3.2+
0.0 0.6 0.7 0.4
compared with other groups at the same time point. compared with TNBS (pH 1.0) and TNBS (pH 7.4) groups at the same time point. compared with ETOH and TNBS @H 7.4) groups at the same time point. compared with ETOH group at the same time point.
were excised and opened longitudinally. Gross mucosal injury was assessed using a minor modification of the grading scale of Morris et a1.5 The mucosal appearance was graded as follows: normal mucosa (grade 0); localized hyperemia but no ulcers or scars (grade 1);linear ulcer or ulcer scar with no significant inflammation (grade 2); linear ulcer or ulcer scar with inflammation at one site (grade 3); two or more sites of ulceration and/or inflammation (grade 4); and two or more major sites of inflammation and ulceration or one major site of inflammation and ulceration extending >I cm along the length of the colon (grade 5). Inflammation was defined as regions of bowel wall thickening and hyperemia. The length and weight of each colon was then recorded, and the tissue was divided longitudinally into two strips for wet-to-dry and myeloperoxidase (MPO) determinations. Colonic MPO activity was determined by the method of Grisham et al. using 3,3’-5,5’-tetramethylbenzidine as the electron-donating substrate.17 For histological analysis, two random colonic samples were obtained from each group; fixed by immersion in a solution consisting of 2% fresh paraformaldehyde, 2.5% glutaraldehyde, and 0.1 mol/L phosphate buffer, pH 7.4, at 4OC; and sliced into 2-&mm pieces and fixed overnight. The tissue was dehydrated to 95% ethanol and embedded in glycol methacrylate (JB-4; Polysciences Inc., Warrington, PA). JB-4 blocks were cured overnight in
Table 2. Time Course Study of Mucosal Permeability: (pH 7.4) Groups Control Mucosal permeability (mL - min-’ - 100 g dry wt-‘)
Acetic acid
0.07 * 0.03 [n = 15) -
ETOH
-
TNBS (pH 1.0)
-
TNBS (pH 7.4)
-
"P < *P < "P < dP <
Among Acetic Acid, ETOH, TNBS (pH 1.O], and TNBS [pH 7.4) 6h
1.6k 0.2" 1.0rt0.0 1.0f 0.0 1.0It0.0
Vol. 102, No. 5
a vacuum desiccator at room temperature, and semithin (I-2-pm) sections were cut on glass knives and stained with toluidine blue. For some experiments, the degree and extent of acute vs. chronic inflammation was scored by an individual unaware of the treatment protocol. Scoring criteria were based on the method of Sartor et al. in which acute and chronic inflammation was graded on a 0-4+ scale.18 Briefly, components of the acute inflammatory score consisted of edema and neutrophil infiltration, whereas the chronic inflammatory response was based on the appearance of mononuclear cells (lymphocytes and macrophages). Statistical
Results are expressed as the mean -t SEM. Data were analyzed using one-way analysis of variance (ANOVA) procedures, and statistical differences between and within groups were confirmed using least significant differences.
Results Gross inspection the intrarectal
E + TNBS
Comparison zh
Analyses
of the colonic
administration
mucosa
after
of acetic acid, ethanol,
(pH l.O), or E + TNBS
(pH 7.4) showed
Among Acetic Acid, ETOH, TNBS (pH l.O), and TNBS 6h
48 h
5.21+ 1.03"** 1.26f 0.52 1.18+ 0.44 (n = 7) (n = 6) (n = 6) 1.23f 0.32' 0.69+ 0.30' 0.10+ 0.01 (n = 5) (n = 5) (n = 5) 2.35 f 0.68a 1.51+ 0.67 1.21+ 0.40 (n = 9) (n = 10) (n = 9) 10.49+ o.88a*c 5.13+ 2.08a'c 0.68+ 0.15 (n = 5) (n = 4) (n = 6)
0.05compared with control. 0.05compared with ETOH and TNBS @H 1.0) groups at the same time point. 0.05compared with other groups at the same time point. 0.05 compared with ETOH and acetic acid groups at the same time point.
1 wk
0.18f 0.05 (n = 5) 0.14!I0.39 (n = 6) 0.92+ 0.33 (n = 11) 1.07+ 0.67 (n = 5)
2 wk
0.16k 0.03 (n = 5) 0.08zk0.03 (n = 5) 1.13+ 0.42 (n = 10) 2.25f 0.57d (n = 5)
May 1992
COLONIC
INFLAMMATION
IN RATS
1527
Figure 1. (A-C) Rat colonic tissue exposed to 4% acetic acid at pH 2.3 for 30 seconds (2-S-pm plastic sections stained with toluidine blue; bar = 50 pm). (A) Two hours after acetic acid the epithelial surface is undergoing massive erosion. The lumen is filled with mucus and exfoliated cells but there is only a modest infiltration of defense cells. (B) Two days after exposure to acetic acid, the epithelial barrier has reformed, with prominent crypt epithelial cells and a flattened layer of cells covering the colonic surface. Granulocytes are still prominent in the underlying lamina propria. (C) Two weeks after insult, the mucosa has been totally repaired, with little evidence of injury.
noticeable hyperemia and exfoliation of the epithelial lining 2 and 6 hours after enema (Table 1). Maximal mucosal injury and inflammation (e.g., ulcerations, bowel wall thickening, and hyperemia) were evident 48 hours after enema for most groups. The extent of the mucosal surface area involved was similar for all four groups, i.e., approximately 20%-30% of the colon, and the area was localized primarily in the descending portion. By 1-2 weeks the colonic mucosa of the acetic acid and ethanol animals appeared healed; however, the bowel wall appeared thickened, and evidence for scar formation and a
slight hyperemia was evident in all four groups. Injury induced by E + TNBS (pH 1.0) was more variable than in the other groups, with some animals showing significant injury and inflammation 2 weeks after enema, whereas other animals showed only moderate inflammation at this time point. Table 2 shows the effects of the intrarectal administration of acetic acid, ethanol, E + TNBS (pH 1.0) and E + TNBS (pH 7.4) on mucosal permeability at different times after the enemas. It was determined in preliminary studies that the values obtained for saline-treated rats at all time points after the enema
1528
YAMADA
ET AL.
GASTROENTEROLOGY
Vol. 102, No. 5
Figure 2. (A-C) Rat colonic tissue exposed to 50% ethanol intrarectally as a control for the TNBS-ethanol group @+pm plastic sections stained with toluidine blue; bar = 50 urn). (A) Two hours after exposure to ethanol, the mucosa shows large-scale erosions of the epithelium, with only the base of the crypts remaining intact. Inflammatory cells have not yet infiltrated into the lamina propria. (E) Two days after insult, portions of the mucosa are only beginning to recover. In this region, epithelial cells have begun to repopulate the surface epithelium, but the process is not yet complete. (C) Two weeks after the insult, the mucosa appears normal.
were not significantly different from each other; they were thus grouped together as one control group. All four treatments produced rapid and significant increases in mucosal permeability 2 hours after the enema when compared with controls. Throughout the Zweek time course, mucosal permeabilities gradually decreased so that at 2 weeks, the values for the acetic acid and ethanol groups were not different from controls. Although the permeabilities in the E + TNBS (1.0) and E + TNBS (pH 7.4) groups remained elevated at 2 weeks, only the permeability in the E +
TNBS (pH 7.4) group was found to be significantly higher than the ethanol or acetic acid group. The large increases in mucosal permeabilities 2 hours after the intrarectal application of the four agents appeared to correlate well with extensive mucosal injury, as observed by light microscopy (Figures l-4). Histological inspection showed marked epithelial cell injury 2 hours after the enema for all four groups (Figures lA, 2A, 3A, and 4A). The epithelium was eroded from the colonic surface and down into the crypts. By 48 hours (Figures lB, 2B, 3B, and
May 1992
COLONIC
INFLAMMATION
IN RATS
1529
Figure 3. (A-D) Rat colonic tissue exposed to TNBS at pH 1.0 in 30% ethanol intrarectally (23-pm plastic sections stained with toluidine blue; bar = 50 pm). (A) The pathology associated with exposure to TNBS differs from that seen after exposure to the vehicle alone. Erosion of the epithelium appears to begin at the base of the crypts, rather than with the surface epithelium. By 2 hours after insult, the basal half of the crypts have been defoliated. (B) By 48 hours after insult, the mucosal injury is present in some areas but largely healed in others. This micrograph illustrates a transition from a normal appearing epithelial barrier and an adjacent area devoid of any crypt epithelial cells. (C) Mucosal injury 2 weeks after exposure to TNBS-ethanol (pH 1.0) is variable: in some regions of the bowel, there is a total absence of epithelium, with the mucosal area instead occupied by defense cells and bacteria. (0) Other areas of the mucosa are nearly normal with an intact epithelial barrier and a limited population of defense cells in the interstitium.
1530 YAMADA
ET AL.
GASTROENTEROLOGY
Vol. 102, No. 5
Figure 4. (A-D) Rat colonic tissue exposed to TNBS at pH 7.4 in 30% ethanol intrarectally (2-3-Bm plastic sections stained with toluidine blue; bar = 30 pm). (A) The pathology associated with exposure to neutralized TNBS differs from that seen after exposure to acidified TNBS. Erosion of the epithelium appears to begin at the mucosal surface, with the upper half of the crypt defoliated by 2 hours after exposure. (B) By 48 hours after insult, the mucosal injury is evident throughout the samples with a massive infiltration of defense cells. The epithelium has been ravaged, and only remnants of the former crypt structure remain. (Cl Two weeks after exposure to TNBS-ethanol (pH 7.4), the mucosal repair is variable, although not as variable as after exposure to acidified TNBS. Although areas of the mucosa still have breaks in the epithelial barrier, the crypt structure is intact and populated with columnar cells. (D] The majority of the mucosa has restituted, although the architecture of the mucosa has not returned entirely to normal. Chronic inflammatory cells are apparent in the interstitium.
COLONIC INFLAMMATION
May 1992
after injury, crypt cells were recolonizing the surface of the colon, although these cells remained flattened in appearance. The crypt epithelium became more robust as the cells increased in density. Two weeks after injury (Figures 1C and ZC), the coionic mucosa appeared relatively normal in the acetic acid and ethanol groups with much less evidence of inflammatory cell infiltrate and injury. The pathology 2 weeks after enema for E + TNBS (pH 1.0) was more variable than for acetic acid, ethanol, or E + TNBS (pH 7.4). For example, some regions of the bowel showed areas of total exfoliation of the colonic epithelium accompanied by extensive destruction of the mucosal interstitium 2 weeks after treatment with E + TNBS (pH 1.0) (Figure 3C), whereas other areas showed only minor injury and inflammation (Figure 3D). Intrarectal administration of E + TNBS (pH 7.4) produced more consistent injury 2 weeks after the enema, although some variability remained (Figure 4C and D). As predicted, the acute inflammatory response predominated within the first 48 hours whereas a more chronic inflammation appeared at l-2 weeks in all groups after the enema (Table 3). Also, the extent of injury appeared to be confined primarily to the mucosa for the acetic acid and ethanol groups whereas both E + TNBS groups showed transmural inflammation and injury at 48 hours and l-2 weeks. Granulomas were present in some of the E + TNBS samples but absent in the acetic acid and
IN RATS
1531
4B)
Table 3. Comparison of Microscopic Findings in Rats Given Acetic Acid, ETOH, TNBS (pH l.O), or TNBS (pH 7.4)
Acetic acid 2h 48 h 2 wk ETOH 2h 48 h 2 wk TNBS (pH 1.0) 2h 48 h 2 wk TNBS (pH 7.4) zh 48 h 2wk
Acute inflammation
Chronic inflammation
++ + _
++ _
+ + -
+ +
+ + +
+++ +++
+ ++ +
++++ +++
_
_
_
_
Extent of injury
Granuloma
Mucosal Mucosal -
_ _
Mucosal Mucosal Mucosal
_ _ _
Mucosal Transmural Transmural
+
Mucosal Transmural Transmural
_
-t
NOTE. Acute and chronic inflammation were scored on a O-4+ scale. Components of acute inflammation consisted of edema and neutrophil infiltration, whereas chronic inflammation was based on presence of mononuclear cells (lymphocytes and macrophages present).
1
100
10
TNBS
Concentration
(mM)
Figure 5. TNBS-mediated cytotoxicity toward intestinal epithelial cells. ‘Q-labeled rat intestinal epithelial cells (IEC-11; 2 x 10’ cells/well) were exposed to varying concentrations of TNBS (pH 7.4) for 2 hours at WC. Cytotoxicity was quantified as the percent specific release of “Cr. Each data point represents the mean f SD of triplicate samples from two different experiments.
ethanol groups. Whereas it is well accepted that ethanol and acetic acid are injurious to most tissues, it is uncertain whether TNBS (pH 7.4), in the absence of ethanol, injures intestinal epithelial cells. In a separate series of studies we found that neutralized TNBS is very toxic to intestinal epithelial cell monolayers, possessing an ED,, of approximately 5 mmol/L (Figure 5). The large increases in mucosal permeability and epithelial cell injury also appeared to parallel roughly the increases in tissue water content as measured by wet-to-dry ratios (Table 4). We found significant increases in wet-to-dry ratios 2 hours after the enema for all four groups and at 6 hours for acetic acid, ethanol, and E + TNBS (pH 7.4) when compared with controls. Wet-to-dry ratios of the ethanol group were significantly higher than those of all other groups 2 hours after the enema. Although mucosal permeability (epithelial injury) and wet-to-dry ratios increased quite dramatically 2 and 6 hours after the enema, we did not observe the signs of classic inflammation until 48 hours after the enema. For example, maximal MPO activity (i.e., neutrophil infiltration) did not occur until 48 hours to 1 week after the acetic acid enema, whereas maximal MPO values were obtained at 48 hours for all other groups with the E + TNBS (pH 7.4) group showing significantly more colonic MPO than the acetic acid group (Table 5). Interestingly, MPO values at l-2 weeks were significantly elevated for all four groups compared with controls, suggesting the presence of granulocytes during colonic mucosal repair. Neutrophil infiltration was confirmed by histological inspection of the tissue from all four groups 48 hours after the enema (Figures lB, ZB, 3B, and 4B).
1532
YAMADA ET AL.
Table 4.
GASTROENTEROLOGY
Time Course Study of Wet-Dry Ratios: Comparison Among Acetic Acid, ETOH, (pH 1.O),and TNBS (pH 7.4) Groups Control
Wet-dry Acetic ETOH TNBS TNBS
Vol. 102,No. 5
ratio acid (pH 1.0) (pH 7.4)
6.16zt0.23 -
2h
7.58+ 9.94+ 7.43f 7.72+
6h
0.60" 0.80"** 0.22O 0.48'=
48 h
7.98f 0.66" 7.7310.80" 7.51+ 0.44O 6.63f 0.22
6.30+ 6.56+ 6.62+ 6.30f
0.16 0.12 0.14 0.16
I wk
6.42+ 0.12 6.91+ 0.21' 6.17t-0.12 5.89f 0.18
2 wk
6.93f 0.17 6.34f 0.26 6.67+ 0.17 6.03+-0.24
“P < 0.05 compared with control. *P < 0.05compared with other groups at the same time point. “P < 0.05compared with ETOH and TNBS (pH 1.0) groups at the same time point.
Although the MPO values for all four groups remained elevated 2 weeks after the enema, large numbers of neutrophils were not as apparent at 2 weeks than at 48 hours (Figures lC, 2C, 3C, and 4C). In a pattern very similar to colonic MPO, colon weight was significantly increased 48 hours and 1 and 2 weeks after the intrarectal application of the four agents (Table 6). Colon weights from the E + TNBS (pH 1.0)group were significantly increased at 1 and 2 weeks compared with all other groups and the acetic acid group, respectively.
two models, we determined the temporal relationship among mucosal permeability, epithelial injury, and inflammation. Data obtained in this study confirmed the supposition that the inflammation induced by acetic acid, ethanol, or E + TNBS (pH 1.0 or 7.4) results from an initial, caustic injury to the coionic epithelial and interstitium, as measured by the rapid and dramatic increases in mucosal permeability and tissue water content as well as by histological appearance (Tables 2 and 3; Figures l-4). Only after 1-2 days after the initial insult, did we observe signs of classical inflammation, i.e., increases in colonic MPO activity (neutrophil infiltration), colon weight, hyperemia, and ulcerations (Tables 1, 4, and 5; Figures l-4). Interestingly, we found that even after 1-2 weeks after the induction of colitis, all four groups (acetic acid, ethanol, and E + TNBS, pH 1.0 and 7.4) showed significant amounts of inflammation as measured by increases in MPO activity, colon weight, and gross mucosal injury. These data are interesting in view of the fact that histological inspection of the tissue revealed fewer neutrophils at 1-2 weeks than would be expected from the MPO data. This apparent discrepancy may be explained by the fact that the MPO assay is not specific for neutrophils per se, because eosinophils, monocytes, and, to a lesser extent, macrophages all possess significant hemoprotein peroxidase as well. Similar to what has been reported by Morris et a1.5 and Allgayer et a1.,7 we found that the combination
Discussion Intrarectal administrations of either acetic acid or TNBS in ethanol are widely used as models of IBD because of their simplicity, adaptability to small animals, reproducibility, and noninvasive nature. Although most models of IBD (e.g., acetic acid, TNBS, and immune complex) require an initial, injurious insult to the colonic mucosa via the use of organic acids (acetic acid) or solvents (ethanol, formalin), it has been suggested that the TNBS-induced model of colitis may be the more relevant model of IBD because this model involves the use of an immunologic hapten and because the acute mucosal injury produced by the barrier breaker, ethanol, resolves quickly and is followed by a more chronic phase of inflammation.5 Because of the importance in establishing a relevant model of IBD in small animals and because no information is available directly comparing these
Table 5. Time Course Study (pH 7.4)Groups
of CoJonic MPO Activity: Control
Colonic MPO activity (u/g dry M) Acetic acid ETOH TNBS (pH 1.0) TNBS (pH 7.4)
214 f 21 -
Comparison
2h
221+57 241 f18 17Ok 22 385 +31*
6h
267+47 225 + 33 214 k 34 336f 35
“P < 0.05compared with control. *P < 0.05 compared with other groups at the same time point. “P < 0.05 compared with acetic acid group at the same time point.
Among
Acetic
Acid, ETOH, TNBS (pH l.O), and TNBS
48 h
1009 f 95a 1272 k200' 1276+191' 1214 _+ill".'
I wk
1680 -c260' 1002 f196O 929 k184O 855+243"
2 wk
774 + 382' 768 f 242a 814 k 164" 1260+267"
COLONIC INFLAMMATION
May 1992
IN RATS
1533
Table 6. Time Course Study of Colon Weight: Comparison Among Acetic Acid, ETOH, TNBS (pH l.O), and TNBS (pH 7.4) Groups Control
Colon weight [g dry wt/cm) Acetic acid ETOH TNBS (pH 1.0) TNBS (pH 7.4)
zh
6h
48 h
0.021+ 0.001 0.023t 0.001 0.018+ 0.001 0.036f 0.024zk0.002 0.021* 0.001 0.038+ 0.022+ 0.001 0.027+ O.OOZatb 0.032f 0.021+ 0.002 0.017rf: 0.001 0.033+
0.003' 0.003" 0.002' 0.005"
1 wk
zwk
0.032t 0.001' 0.029t 0.024k 0.001 0.031k 0.041+_0.002°.b 0.036A 0.028z!z 0.002" 0.031f
O.OO1° 0.002" 0.002"~c o.oo2a
“P < 0.05compared with control. bP < 0.05 compared with other groups at the same time point. “P < 0.05 compared with acetic acid group at the same time point.
of ethanol and TNBS (buffered or unbuffered) produced more colonic injury at l-2 weeks than acetic acid or ethanol alone (Table 2). However, we found that only E + TNBS (pH 7.4) group produced consistently and significantly more epithelial cell injury (mucosal permeability) than the acetic acid or ethanol groups. The reasons for this discrepancy are unclear; however, there are several possibilities. First, it may be that sex and/or strain differences account for some of the observed differences. Beagly et al. have shown that the strain of the animal may be very important in producing a chronic inflammation in the rat colon.” Second, it may be that differences in TNBS preparations account for at least some of these differences. We have found that different lots of TNBS from the same vendor may produce different degrees of injury and inflammation using the same sex and strain of rats. Finally, our data suggest the variability in response to E + TNBS (pH 1.0) is much greater than for the other agents (Figure 3C and D). It should be noted that the intrarectal application of acetic acid or ethanol produced very consistent injury and inflammation from day to day and animal to animal (data not shown). The mechanism(s) by which acetic acid produces inflammation in the rat colon appear to involve the entry of the lipid soluble form (protonated) of the acid into the epithelium, where it dissociates to liberate protons within the intracellular space. It is this massive intracellular acidification that most likely accounts for the epithelial injury observed, because intraluminal introduction of HCl (pH 2.3) did not produce injury or inflammationg*” (data not shown). Our studies also showed the dramatic recuperative powers of the colonic mucosa indicated by the rapid re-epithelialization, which appeared to correlate well with the decrease in mucosal permeability after the initial insult (Table 2; Figures l-4). The intraluminal administration of ethanol alone or E + TNBS (pH 1.0 or 7.4) produced inflammation in a time very similar to that observed with acetic acid. The proinflammatory activity of ethanol towards the colon was originally described by Wallace
et a1.14who showed that dilute solutions of ethanol produce rapid and severe injury to the colonic mucosa. We found that the intrarectal administration of 50% ethanol produced an acute inflammatory response that resolved over the subsequent 2 weeks (Table 2; Figure 2). The mechanism(s) by which E + TNBS induces colitis appear more complex than those for acetic acid or ethanol. Although TNBS is in fact a classic hapten, it is generally thought that animals need to be sensitized to a hapten before an immunologic response is possible. Furthermore, TNBS is not innocuous. For example, we have shown in this study that TNBS (pH 7.4) in the absence of ethanol is very toxic to intestinal epithelial cell monolayers in vitro (Figure 9, suggesting that some of the inflammation induced by the intrarectal or intramural administration of TNBS may result from TNBSmediated tissue injury. The mechanism by which TNBS mediates epithelial cell injury remains undefined; however we have recently shown that the rat colon as well as rat colonocytes are capable of metabolizing TNBS to produce large quantities of reactive oxygen metabolites such as superoxide, hydrogen peroxide, and hydroxyl radical in vitro.” Formation of these oxidants within the mucosa would be extremely injurious and could initiate an inflammatory response. Indeed, the production of reactive oxygen metabolites by cyclical reduction and spontaneous oxidation reactions (redox cycling) of the nitro functional groups on aromatic compounds is well characterized and has been suggested as one mechanism to account for side effects of metronidazole.20p21 Taken together, our data suggest that all four agents are effective in injuring the colonic mucosa and inducing an acute colitis in rats. However, it appears that the combination of ethanol and TNBS may produce a more chronic mucosal injury and inflammation than acetic acid and ethanol alone. We suggest any one of the models of colitis are useful to study those events that occur at the time of inflammation such as (arachidonate metabolism, granulocyte infiltration, and metabolism etc) or during repair. However, the use of these models of colitis
1534
YAMADA ET AL.
may have significant limitations in understanding those immunologic events that initiate the acute and chronic inflammatory episodes. For example, the inflammation and tissue injury observed in human IBD is most probably a result of some sort of inappropriate immunologic activation (e.g., autoimmune, infectious agent, etc.), whereas the inflammation induced by the intrarectal application of acetic acid, ethanol, or E + TNBS is, at least initially, a result of extensive mucosal injury. Thus, the mechanisms by which inflammation (and mucosal injury) are achieved in the human disease may be very different than those in the experimental models. References 1.
MacPherson BR, Pfeiffer CJ. Experimental production of diffuse colitis in rats. Digestion 1978;17:135-150. 2. Sharon P, Stenson WF. Metabolism of arachidonic acid in acetic acid colitis in rats. Similarity to human inflammatory bowel diseases. Gastroenterology 1985;88:55-63. 3. Sekizuka E, Grisham MB, Li M, Deitch EA, Granger DN. Inflammation-induced intestinal hyperemia in the rat: role of neutrophils. Gastroenterology 1988;95:1528-1534, 4. Fedorak RN, Empey LR, MacArthur C, Jewel1 LD. Misoprostol provides a colonic mucosal protective effect during acetic acid-induced colitis in rats. Gastroenterology 1990;98:615625. 5. Morris GP, Beck PL, Herridge MS, Depew WT, Szewczuk MR, Wallace JL. Hapten-induced model of chronic inflammation and ulceration in the rat colon. Gastroenterology 1989;96: 795-803. 6. Wallace JL. Release of platelet-activating factor (PAF) and accelerated healing induced by a PAF antagonist in an animal model of chronic colitis. Can J Physiol Pharm 1988;66:422425. 7. Allgayer H, Deschryver K, Stenson WF. Treatment with 16,16’-dimethyl prostaglandin E, before and after induction of colitis with trinitrobenzene sulfonic acid in rats decreases inflammation. Gastroenterology 1989;96:1290-1300. 8. Wallace JL, MacNaughton WK, Morris GP, Beck PL. Inhibition of leukotriene synthesis markedly accelerates healing in a rat model of inflammatory bowel disease. Gastroenterology 1989;96:29-36. 9. Zeitlin IJ, Norris AA. Animal models of colitis. British Society of Gastroenterology/Smith Kline and French Laboratories workshop. Stanstead Abbotts, England, 1983:70-73.
GASTROENTEROLOGY
Vol. 102, No. 5
10.
Strober W. Animal models of inflammatory bowel diseasean overview. Dig Dis Sci 1985;3O(Suppl):3S-10s. 11. Bustos-Fernandez L, Gonzalez E, de Paolo IL, Celener D, de Furuya KO. Organic anions induce colonic secretion. Dig Dis 1976;21:329-332. 12. Kunin S, Gallily R. Recognition and lysis of altered-self cells by macrophages. Modification of target cells by 2,4,6-trinitrobenzene sulfonic acid. Immunology 1983;48:265-272. 13. Beck PL, Morris, GP, Wase AW, Szewczuk, M, Wallace JL. Immunological manipulation of disease progression in a rat model of chronic inflammatory disease of the colon. In: MacDermott RP, ed. Inflammatory bowel disease: current status and future approaches. Amsterdam: Elsevier Science, 1988:201-206. 14. Wallace JL, Whittle BJR, Boughton-Smith NK. Prostaglandin protection of rat colonic mucosa from damage induced by ethanol. Dig Dis Sci 1985;30:866-876. 15. Grisham MB, Volkner C, Tso P, and Yamada T. Metabolism of trinitrobenzene sulfonic acid by the rat colon produces reactive oxygen species. Gastroenterology 1991;101:540-547. 16. von Ritter C, Sekizuka E, Grisham MB, Granger DN. The chemotactic peptide N-formyl-methionyl-leucyl-phenylalanine increases mucosal permeability in the distal ileum of the rat. Gastroenterology 1988;95:651-655. 17. Grisham MB, Benoit JN, Granger DN. Assessment of leukocyte involvement during ischemia and reperfusion of intestine. Methods Enzymol 1990;186:729-742. 18. Sartor RB, Cromartie WJ, Powell DW, and Schwab JH. Granulomatous enterocolitis induced in rats by purified bacterial cell wall fragments. Gastroenterology 1985;89:587-595. 19. Beagly KW, Black CA, Elson CO. Strain differences in susceptibility to TNB-induced colitis (abstr). Gastroenterology 1991;100:A560. 20. Rao DNR, Harman L, Motten A, Schreiber J, Mason RP. Generation of radical anions of nitrofurantoin, misoidazole, and metronidazole by ascorbate. Arch Biochem Biophys 1987;255: 419-427. 21. Rao DNR, Mason RP. Generation of nitro radical anions of some %nitrofurans, 2- and 5-nitroimidazoles by norepinephrine, dopamine, and serotonin. J Biol Chem 1987;262:1173111736.
Received July 9, 1990. Accepted September 23,1991. Address requests for reprints to: Matthew B. Grisham, Ph.D., Department of Physiology and Biophysics, Louisiana State University Medical Center, 1501Kings Highway, P.O. Box 33932, Shreveport, Louisiana 71130-3932. Supported by grants from the National Institutes of Health (DK 39168, DK 33720), the Crohn’s and Colitis Foundation of America, and by funds from the Feist Research Foundation.
1992;102:1524-1534
A Comparative Analysis of Two Models of Colitis in Rats TAMAKI YAMADA, STEVEN MARSHALL, and MATTHEW B. GRISHAM Departments of Physiology and Biophysics Medical Center, Shreveport, Louisiana
D. SPECIAN,
and Cellular Biology and Anatomy,
Two models of colitis produced in rats that have received significant attention over the past few years are the acetic acid and trinitrobenzene sulfonic acid (TNBS) models. The objective of this study was to quantify and compare the temporal relationship among mucosal permeability, epithelial injury, and inflammation induced by acetic acid, ethanol (vehicle), ethanol plus TNBS (unbuffered, pH l.O), and ethanol plus TNBS (pH 7.4). Data obtained show that the inflammation induced by these four irritants results from caustic injury to the colonic epithelium and interstitium as measured by the rapid and dramatic increases in mucosal permeability and tissue water content as well as by histological analysis. The injurious nature of TNBS was confirmed in a separate series of studies showing that buffered TNBS (pH 7.4), in the absence of ethanol, is toxic to cultured rat intestinal epithelial cell monolayers. Only after 1-2 days of the initial insult, were signs of classical inflammation observed, including increases in colonic myeloperoxidase activity (neutrophil infiltration) and colon weight as well as hyperemia and mucosal ulcerations. Although ethanol plus TNBS (pH 1.0 or 7.4) tended to produce higher mucosal permeabilities (epithelial cell injury) at l-2 weeks after the enemas than acetic acid or ethanol groups, only the ethanol plus TNBS (pH 7.4)permeabilities were found to be significantly enhanced. In addition, all four groups showed significant elevations in coionic myeloperoxidase activity and colon weight at 1-2 weeks after enema. It is suggested that these models of colitis are useful to study events that occur at the time of inflammation and repair. However, these models may have significant limitations in understanding events that initiate inflammation of the intestine in human inflammatory bowel disease. major limitation in investigating the pathogenetic mechanisms responsible for the mucosal injury observed in inflammatory bowel disease (IBD)
A
ROBERT
Louisiana State University
has been the relative paucity of relevant animal models. Two models of colitis produced in rats that have received significant attention over the past few years are the acetic acid and trinitrobenzene sulfonic acid (TNBS) models.‘-8 The mechanism by which acetic acid produces the diffuse colitis is thought to involve nonspecific, acid-induced injury to the colonic mucosa that is followed by an acute inflammatory response. Apparently, the protonated form of the acid is required to induce the colitis, because neither HCI (pH 2.3)nor sodium acetate (pH 7.0)are effective in eliciting the inflammatory response.gs’o However, there is some evidence to suggest that acetic acid may promote other pathophysiological events (e.g., fluid and electrolyte secretion) using noncytotoxic concentrations of the acid.‘l Recent studies show that intrarectal administration of the hapten TNBS, in the presence of a mucosal barrier breaker such as ethanol, produces an acute and possibly chronic colitis in unsensitized rats.5 The mechanism(s) by which buffered or unbuffered TNBS in the presence of ethanol initiates inflammation in unsensitized animals is unclear; however, it has been suggested to involve macrophage-mediated recognition and lysis of TNBS-modified autologous cells within the mucosa.12 Beck et al. reported that some of the colonic injury induced by ethanol plus TNBS (pH 1.0) may be attenuated by prior tolerization of the animals to TNBS, suggesting cell-mediated tissue injury.13 Although these data suggest an immune celldependent component of TNBS-induced colonic inflammation, the mechanism(s) for ethanol plus TNBS-induced injury and inflammation may be more complex than previously appreciated. For example, the barrier breaker ethanol is an extremely potent proinflammatory solvent a10ne.14 Furthermore, we have shown that TNBS is metabolized by certain colonic enzymes and substrates to yield both proinflammatory and cytotoxic oxidants that could 0 1992 by
the American Gastroenterological 0016-5085/92/$3.00
Association
COLONIC INFLAMMATION
May 1992
initiate colonic inflammation.15 Although both acetic acid and ethanol plus TNBS are widely used as models of colitis in rats, there has been no systematic study that quantitatively compares the temporal relationship between mucosal injury and inflammation in both models of colitis. The objective of this study was to quantify and compare the temporal relationship among mucosal permeability, epithelial injury, and inflammation induced by the intrarectal administration of acetic acid, ethanol (vehicle), ethanol plus TNBS (unbuffered, pH 1.0) and ethanol plus TNBS (pH 7.4). Materials and Methods Induction
of Colitis
Male Sprague-Dawley rats (300-350 g) were fasted overnight with access to water ad libitum. A total of 129 rats were randomized into five major groups consisting of a saline control group as well as an acetic acid, ethanol, ethanol plus TNBS (E + TNBS; unbuffered, pH l.O), and E + TNBS (buffered, pH 7.4) group. Diffuse colitis was induced with acetic acid as described previously.3 Briefly, the animals were lightly anesthetized with Ethrane (enflurane; Anaquest Co., Memphis, TN), and 1 mL of either saline (controls) or 4% (vol/vol) acetic acid (pH 2.3) was slowly administered into the lumen of the colon via a s-cm polyethylene tubing (PE-240) fitted onto a 1-mL syringe. After a 30-second period of exposure, excess fluid was withdrawn, and 1.5 mL phosphate-buffered saline (PBS) was introduced to flush the colon. Colitis was induced in a third set of rats by intrarectal administration of 0.25 mL 50% (vol/vol) ethanol using the l-mL syringe fitted with the 5-cm catheter. Finally, the fourth and fifth groups of anesthetized rats were treated intrarectally with 0.25 mL TNBS solution (120 mg/mL TNBS in 50% unbuffered ethanol, pH 1.0) as described by Morris et al.’ or a TNBS solution (120 mg/mL TNBS in 50% ethanol adjusted to pH 7.4 with NaOH).
Mucosal Permeability At various times after the enemas, rats were anesthetized with 120 mg/kg of Na-5-ethyl-l(l’-methyl-propyl)-2-thio-barbiturate (Inactin; Byk Gulden, Konstanz, Germany). The animals were tracheotomized, and the right femoral artery was cannulated for arterial pressure recording and blood sampling. The right femoral vein was also cannulated for the injection of radioisotope marker. A midline abdominal incision was performed. The colon was isolated and cannulated at both the splenic flexure and the rectum using Silastic tubing (ID, 0.025 and 0.25 in, respectively; Dow Corning). The luminal contents were removed by perfusion with warm (37°C) Tyrode’s solution. Both renal pedicles were ligated to prevent rapid excretion of the radioisotope marker into the urine. The isolated, perfused colon was returned to the abdominal cavity, and the abdominal wall was closed to minimize dehydration of the colon segment during the experiment. Body temperature was maintained at 37°C with a thermistor-controlled infra-
IN RATS
1525
red heat lamp. Mucosal permeability was determined using the blood-to-lumen clearance of [Yrlethylenediaminetetraacetic acid (EDTA) as described previous1y.‘6 [Yr]EDTA, 100 pCi, was injected via the femoral vein followed by a l-hour equilibration period, during which the intestinal lumen was perfused with Tyrode’s solution but no clearances measurements were taken. The perfusate was sampled every 10 minutes for 40 minutes for the appearance of [“CrJEDTA. Blood samples (0.3 mL) were taken at 40 minutes for use as reference counts. Blood-tolumen clearance of [“Cr]EDTA was calculated using the following formula:
Clearance
(EDTA) = ~‘~~
,
bl
where Cpe. and Cbl represent counts per minute per milliliter of [“Cr]EDTA in the luminal perfusate and blood, respectively; Q represents the luminal perfusion rate (0.5 mL/min), and W represents the dry weight of the perfused segment of the colon. Mucosal permeability was determined from the mean of four clearance values.
TNBS-Mediated
Epithelial
Cell Cytotoxicity
Normal rat epithelial cells (IEC-18) were purchased from American Type Culture Collection (Rockville, MD). Cells were grown and maintained using Dulbecco’s modified Eagle medium (DMEM) containing 4.5 g/L glucose, 5% fetal bovine serum, and 0.1 U/mL insulin. All studies used confluent monolayers that possessed 40-15 population doublings. Confluent monolayers (approximately 2 X lo5 cells/well) were incubated with DMEM containing 15 pCi [51Cr]NaCr0,/mL for 18 hours. The wells were washed (3X) with PBS (pH 7.4) to remove unincorporated radioactivity. ‘ICr-labeled monolayers were then overlayed with 0.5 mL Dulbecco’s phosphate-buffered saline (DPBS) or 0.5 mL of varying concentrations of TNBS in DPBS (pH 7.4) and incubated for 2 hours at 37°C. After the incubation period, the plate was centrifuged for 3 minutes at 100 X g, and the supernatant was removed and counted for 5*Cr activity. An aliquot (0.5 mL) of 2N NaOH was then added to each well to lyse the remaining cells. After a 60-minute incubation period, the fluid was removed and counted for Yr activity. Cytotoxicity was expressed as the percent specific cytotoxicity (51Cr release) calculated using the following formula:
% Specific Cytotoxity
A-B
= C_
X 100,
where A is the mean counts per min (cpm) in the supernatant of samples containing TNBS; B is the mean cpm in the supernatants of samples containing DPBS; and C is the total cpm available in each well (i.e., the sum of the supernatant plus lysate). Each data point represents the mean of triplicate determinations from two different experiments.
Tissue Analyses After the determinations of mucosal permeability, the animals were killed with pentobarbital, and the colons
1526
YAMADA ET AL.
GASTROENTEROLOGY
Table 1. Time Course Study Groups
ofGrossFindings:
Control Gross findings Acetic acid ETOH TNBS (pH 1.0) TNBS (pH 7.4) “P < *P < “P < dP <
0.05 0.05 0.05 0.05
Comparison
zh
0 -
1 wk
48 h
2.0f o.o* 2.4+ O.Zb 1.0I?0.0 0.8+ 0.2
5.0z!I 0.0" 3.2+ 0.3 4.8A O.Zd 3.8+ 0.8
zwk
2.0iz0.0 3.4f 0.2 3.3Ik0.9 2.6f 0.6
2.0+ 3.0+ 3.4f 3.2+
0.0 0.6 0.7 0.4
compared with other groups at the same time point. compared with TNBS (pH 1.0) and TNBS (pH 7.4) groups at the same time point. compared with ETOH and TNBS @H 7.4) groups at the same time point. compared with ETOH group at the same time point.
were excised and opened longitudinally. Gross mucosal injury was assessed using a minor modification of the grading scale of Morris et a1.5 The mucosal appearance was graded as follows: normal mucosa (grade 0); localized hyperemia but no ulcers or scars (grade 1);linear ulcer or ulcer scar with no significant inflammation (grade 2); linear ulcer or ulcer scar with inflammation at one site (grade 3); two or more sites of ulceration and/or inflammation (grade 4); and two or more major sites of inflammation and ulceration or one major site of inflammation and ulceration extending >I cm along the length of the colon (grade 5). Inflammation was defined as regions of bowel wall thickening and hyperemia. The length and weight of each colon was then recorded, and the tissue was divided longitudinally into two strips for wet-to-dry and myeloperoxidase (MPO) determinations. Colonic MPO activity was determined by the method of Grisham et al. using 3,3’-5,5’-tetramethylbenzidine as the electron-donating substrate.17 For histological analysis, two random colonic samples were obtained from each group; fixed by immersion in a solution consisting of 2% fresh paraformaldehyde, 2.5% glutaraldehyde, and 0.1 mol/L phosphate buffer, pH 7.4, at 4OC; and sliced into 2-&mm pieces and fixed overnight. The tissue was dehydrated to 95% ethanol and embedded in glycol methacrylate (JB-4; Polysciences Inc., Warrington, PA). JB-4 blocks were cured overnight in
Table 2. Time Course Study of Mucosal Permeability: (pH 7.4) Groups Control Mucosal permeability (mL - min-’ - 100 g dry wt-‘)
Acetic acid
0.07 * 0.03 [n = 15) -
ETOH
-
TNBS (pH 1.0)
-
TNBS (pH 7.4)
-
"P < *P < "P < dP <
Among Acetic Acid, ETOH, TNBS (pH 1.O], and TNBS [pH 7.4) 6h
1.6k 0.2" 1.0rt0.0 1.0f 0.0 1.0It0.0
Vol. 102, No. 5
a vacuum desiccator at room temperature, and semithin (I-2-pm) sections were cut on glass knives and stained with toluidine blue. For some experiments, the degree and extent of acute vs. chronic inflammation was scored by an individual unaware of the treatment protocol. Scoring criteria were based on the method of Sartor et al. in which acute and chronic inflammation was graded on a 0-4+ scale.18 Briefly, components of the acute inflammatory score consisted of edema and neutrophil infiltration, whereas the chronic inflammatory response was based on the appearance of mononuclear cells (lymphocytes and macrophages). Statistical
Results are expressed as the mean -t SEM. Data were analyzed using one-way analysis of variance (ANOVA) procedures, and statistical differences between and within groups were confirmed using least significant differences.
Results Gross inspection the intrarectal
E + TNBS
Comparison zh
Analyses
of the colonic
administration
mucosa
after
of acetic acid, ethanol,
(pH l.O), or E + TNBS
(pH 7.4) showed
Among Acetic Acid, ETOH, TNBS (pH l.O), and TNBS 6h
48 h
5.21+ 1.03"** 1.26f 0.52 1.18+ 0.44 (n = 7) (n = 6) (n = 6) 1.23f 0.32' 0.69+ 0.30' 0.10+ 0.01 (n = 5) (n = 5) (n = 5) 2.35 f 0.68a 1.51+ 0.67 1.21+ 0.40 (n = 9) (n = 10) (n = 9) 10.49+ o.88a*c 5.13+ 2.08a'c 0.68+ 0.15 (n = 5) (n = 4) (n = 6)
0.05compared with control. 0.05compared with ETOH and TNBS @H 1.0) groups at the same time point. 0.05compared with other groups at the same time point. 0.05 compared with ETOH and acetic acid groups at the same time point.
1 wk
0.18f 0.05 (n = 5) 0.14!I0.39 (n = 6) 0.92+ 0.33 (n = 11) 1.07+ 0.67 (n = 5)
2 wk
0.16k 0.03 (n = 5) 0.08zk0.03 (n = 5) 1.13+ 0.42 (n = 10) 2.25f 0.57d (n = 5)
May 1992
COLONIC
INFLAMMATION
IN RATS
1527
Figure 1. (A-C) Rat colonic tissue exposed to 4% acetic acid at pH 2.3 for 30 seconds (2-S-pm plastic sections stained with toluidine blue; bar = 50 pm). (A) Two hours after acetic acid the epithelial surface is undergoing massive erosion. The lumen is filled with mucus and exfoliated cells but there is only a modest infiltration of defense cells. (B) Two days after exposure to acetic acid, the epithelial barrier has reformed, with prominent crypt epithelial cells and a flattened layer of cells covering the colonic surface. Granulocytes are still prominent in the underlying lamina propria. (C) Two weeks after insult, the mucosa has been totally repaired, with little evidence of injury.
noticeable hyperemia and exfoliation of the epithelial lining 2 and 6 hours after enema (Table 1). Maximal mucosal injury and inflammation (e.g., ulcerations, bowel wall thickening, and hyperemia) were evident 48 hours after enema for most groups. The extent of the mucosal surface area involved was similar for all four groups, i.e., approximately 20%-30% of the colon, and the area was localized primarily in the descending portion. By 1-2 weeks the colonic mucosa of the acetic acid and ethanol animals appeared healed; however, the bowel wall appeared thickened, and evidence for scar formation and a
slight hyperemia was evident in all four groups. Injury induced by E + TNBS (pH 1.0) was more variable than in the other groups, with some animals showing significant injury and inflammation 2 weeks after enema, whereas other animals showed only moderate inflammation at this time point. Table 2 shows the effects of the intrarectal administration of acetic acid, ethanol, E + TNBS (pH 1.0) and E + TNBS (pH 7.4) on mucosal permeability at different times after the enemas. It was determined in preliminary studies that the values obtained for saline-treated rats at all time points after the enema
1528
YAMADA
ET AL.
GASTROENTEROLOGY
Vol. 102, No. 5
Figure 2. (A-C) Rat colonic tissue exposed to 50% ethanol intrarectally as a control for the TNBS-ethanol group @+pm plastic sections stained with toluidine blue; bar = 50 urn). (A) Two hours after exposure to ethanol, the mucosa shows large-scale erosions of the epithelium, with only the base of the crypts remaining intact. Inflammatory cells have not yet infiltrated into the lamina propria. (E) Two days after insult, portions of the mucosa are only beginning to recover. In this region, epithelial cells have begun to repopulate the surface epithelium, but the process is not yet complete. (C) Two weeks after the insult, the mucosa appears normal.
were not significantly different from each other; they were thus grouped together as one control group. All four treatments produced rapid and significant increases in mucosal permeability 2 hours after the enema when compared with controls. Throughout the Zweek time course, mucosal permeabilities gradually decreased so that at 2 weeks, the values for the acetic acid and ethanol groups were not different from controls. Although the permeabilities in the E + TNBS (1.0) and E + TNBS (pH 7.4) groups remained elevated at 2 weeks, only the permeability in the E +
TNBS (pH 7.4) group was found to be significantly higher than the ethanol or acetic acid group. The large increases in mucosal permeabilities 2 hours after the intrarectal application of the four agents appeared to correlate well with extensive mucosal injury, as observed by light microscopy (Figures l-4). Histological inspection showed marked epithelial cell injury 2 hours after the enema for all four groups (Figures lA, 2A, 3A, and 4A). The epithelium was eroded from the colonic surface and down into the crypts. By 48 hours (Figures lB, 2B, 3B, and
May 1992
COLONIC
INFLAMMATION
IN RATS
1529
Figure 3. (A-D) Rat colonic tissue exposed to TNBS at pH 1.0 in 30% ethanol intrarectally (23-pm plastic sections stained with toluidine blue; bar = 50 pm). (A) The pathology associated with exposure to TNBS differs from that seen after exposure to the vehicle alone. Erosion of the epithelium appears to begin at the base of the crypts, rather than with the surface epithelium. By 2 hours after insult, the basal half of the crypts have been defoliated. (B) By 48 hours after insult, the mucosal injury is present in some areas but largely healed in others. This micrograph illustrates a transition from a normal appearing epithelial barrier and an adjacent area devoid of any crypt epithelial cells. (C) Mucosal injury 2 weeks after exposure to TNBS-ethanol (pH 1.0) is variable: in some regions of the bowel, there is a total absence of epithelium, with the mucosal area instead occupied by defense cells and bacteria. (0) Other areas of the mucosa are nearly normal with an intact epithelial barrier and a limited population of defense cells in the interstitium.
1530 YAMADA
ET AL.
GASTROENTEROLOGY
Vol. 102, No. 5
Figure 4. (A-D) Rat colonic tissue exposed to TNBS at pH 7.4 in 30% ethanol intrarectally (2-3-Bm plastic sections stained with toluidine blue; bar = 30 pm). (A) The pathology associated with exposure to neutralized TNBS differs from that seen after exposure to acidified TNBS. Erosion of the epithelium appears to begin at the mucosal surface, with the upper half of the crypt defoliated by 2 hours after exposure. (B) By 48 hours after insult, the mucosal injury is evident throughout the samples with a massive infiltration of defense cells. The epithelium has been ravaged, and only remnants of the former crypt structure remain. (Cl Two weeks after exposure to TNBS-ethanol (pH 7.4), the mucosal repair is variable, although not as variable as after exposure to acidified TNBS. Although areas of the mucosa still have breaks in the epithelial barrier, the crypt structure is intact and populated with columnar cells. (D] The majority of the mucosa has restituted, although the architecture of the mucosa has not returned entirely to normal. Chronic inflammatory cells are apparent in the interstitium.
COLONIC INFLAMMATION
May 1992
after injury, crypt cells were recolonizing the surface of the colon, although these cells remained flattened in appearance. The crypt epithelium became more robust as the cells increased in density. Two weeks after injury (Figures 1C and ZC), the coionic mucosa appeared relatively normal in the acetic acid and ethanol groups with much less evidence of inflammatory cell infiltrate and injury. The pathology 2 weeks after enema for E + TNBS (pH 1.0) was more variable than for acetic acid, ethanol, or E + TNBS (pH 7.4). For example, some regions of the bowel showed areas of total exfoliation of the colonic epithelium accompanied by extensive destruction of the mucosal interstitium 2 weeks after treatment with E + TNBS (pH 1.0) (Figure 3C), whereas other areas showed only minor injury and inflammation (Figure 3D). Intrarectal administration of E + TNBS (pH 7.4) produced more consistent injury 2 weeks after the enema, although some variability remained (Figure 4C and D). As predicted, the acute inflammatory response predominated within the first 48 hours whereas a more chronic inflammation appeared at l-2 weeks in all groups after the enema (Table 3). Also, the extent of injury appeared to be confined primarily to the mucosa for the acetic acid and ethanol groups whereas both E + TNBS groups showed transmural inflammation and injury at 48 hours and l-2 weeks. Granulomas were present in some of the E + TNBS samples but absent in the acetic acid and
IN RATS
1531
4B)
Table 3. Comparison of Microscopic Findings in Rats Given Acetic Acid, ETOH, TNBS (pH l.O), or TNBS (pH 7.4)
Acetic acid 2h 48 h 2 wk ETOH 2h 48 h 2 wk TNBS (pH 1.0) 2h 48 h 2 wk TNBS (pH 7.4) zh 48 h 2wk
Acute inflammation
Chronic inflammation
++ + _
++ _
+ + -
+ +
+ + +
+++ +++
+ ++ +
++++ +++
_
_
_
_
Extent of injury
Granuloma
Mucosal Mucosal -
_ _
Mucosal Mucosal Mucosal
_ _ _
Mucosal Transmural Transmural
+
Mucosal Transmural Transmural
_
-t
NOTE. Acute and chronic inflammation were scored on a O-4+ scale. Components of acute inflammation consisted of edema and neutrophil infiltration, whereas chronic inflammation was based on presence of mononuclear cells (lymphocytes and macrophages present).
1
100
10
TNBS
Concentration
(mM)
Figure 5. TNBS-mediated cytotoxicity toward intestinal epithelial cells. ‘Q-labeled rat intestinal epithelial cells (IEC-11; 2 x 10’ cells/well) were exposed to varying concentrations of TNBS (pH 7.4) for 2 hours at WC. Cytotoxicity was quantified as the percent specific release of “Cr. Each data point represents the mean f SD of triplicate samples from two different experiments.
ethanol groups. Whereas it is well accepted that ethanol and acetic acid are injurious to most tissues, it is uncertain whether TNBS (pH 7.4), in the absence of ethanol, injures intestinal epithelial cells. In a separate series of studies we found that neutralized TNBS is very toxic to intestinal epithelial cell monolayers, possessing an ED,, of approximately 5 mmol/L (Figure 5). The large increases in mucosal permeability and epithelial cell injury also appeared to parallel roughly the increases in tissue water content as measured by wet-to-dry ratios (Table 4). We found significant increases in wet-to-dry ratios 2 hours after the enema for all four groups and at 6 hours for acetic acid, ethanol, and E + TNBS (pH 7.4) when compared with controls. Wet-to-dry ratios of the ethanol group were significantly higher than those of all other groups 2 hours after the enema. Although mucosal permeability (epithelial injury) and wet-to-dry ratios increased quite dramatically 2 and 6 hours after the enema, we did not observe the signs of classic inflammation until 48 hours after the enema. For example, maximal MPO activity (i.e., neutrophil infiltration) did not occur until 48 hours to 1 week after the acetic acid enema, whereas maximal MPO values were obtained at 48 hours for all other groups with the E + TNBS (pH 7.4) group showing significantly more colonic MPO than the acetic acid group (Table 5). Interestingly, MPO values at l-2 weeks were significantly elevated for all four groups compared with controls, suggesting the presence of granulocytes during colonic mucosal repair. Neutrophil infiltration was confirmed by histological inspection of the tissue from all four groups 48 hours after the enema (Figures lB, ZB, 3B, and 4B).
1532
YAMADA ET AL.
Table 4.
GASTROENTEROLOGY
Time Course Study of Wet-Dry Ratios: Comparison Among Acetic Acid, ETOH, (pH 1.O),and TNBS (pH 7.4) Groups Control
Wet-dry Acetic ETOH TNBS TNBS
Vol. 102,No. 5
ratio acid (pH 1.0) (pH 7.4)
6.16zt0.23 -
2h
7.58+ 9.94+ 7.43f 7.72+
6h
0.60" 0.80"** 0.22O 0.48'=
48 h
7.98f 0.66" 7.7310.80" 7.51+ 0.44O 6.63f 0.22
6.30+ 6.56+ 6.62+ 6.30f
0.16 0.12 0.14 0.16
I wk
6.42+ 0.12 6.91+ 0.21' 6.17t-0.12 5.89f 0.18
2 wk
6.93f 0.17 6.34f 0.26 6.67+ 0.17 6.03+-0.24
“P < 0.05 compared with control. *P < 0.05compared with other groups at the same time point. “P < 0.05compared with ETOH and TNBS (pH 1.0) groups at the same time point.
Although the MPO values for all four groups remained elevated 2 weeks after the enema, large numbers of neutrophils were not as apparent at 2 weeks than at 48 hours (Figures lC, 2C, 3C, and 4C). In a pattern very similar to colonic MPO, colon weight was significantly increased 48 hours and 1 and 2 weeks after the intrarectal application of the four agents (Table 6). Colon weights from the E + TNBS (pH 1.0)group were significantly increased at 1 and 2 weeks compared with all other groups and the acetic acid group, respectively.
two models, we determined the temporal relationship among mucosal permeability, epithelial injury, and inflammation. Data obtained in this study confirmed the supposition that the inflammation induced by acetic acid, ethanol, or E + TNBS (pH 1.0 or 7.4) results from an initial, caustic injury to the coionic epithelial and interstitium, as measured by the rapid and dramatic increases in mucosal permeability and tissue water content as well as by histological appearance (Tables 2 and 3; Figures l-4). Only after 1-2 days after the initial insult, did we observe signs of classical inflammation, i.e., increases in colonic MPO activity (neutrophil infiltration), colon weight, hyperemia, and ulcerations (Tables 1, 4, and 5; Figures l-4). Interestingly, we found that even after 1-2 weeks after the induction of colitis, all four groups (acetic acid, ethanol, and E + TNBS, pH 1.0 and 7.4) showed significant amounts of inflammation as measured by increases in MPO activity, colon weight, and gross mucosal injury. These data are interesting in view of the fact that histological inspection of the tissue revealed fewer neutrophils at 1-2 weeks than would be expected from the MPO data. This apparent discrepancy may be explained by the fact that the MPO assay is not specific for neutrophils per se, because eosinophils, monocytes, and, to a lesser extent, macrophages all possess significant hemoprotein peroxidase as well. Similar to what has been reported by Morris et a1.5 and Allgayer et a1.,7 we found that the combination
Discussion Intrarectal administrations of either acetic acid or TNBS in ethanol are widely used as models of IBD because of their simplicity, adaptability to small animals, reproducibility, and noninvasive nature. Although most models of IBD (e.g., acetic acid, TNBS, and immune complex) require an initial, injurious insult to the colonic mucosa via the use of organic acids (acetic acid) or solvents (ethanol, formalin), it has been suggested that the TNBS-induced model of colitis may be the more relevant model of IBD because this model involves the use of an immunologic hapten and because the acute mucosal injury produced by the barrier breaker, ethanol, resolves quickly and is followed by a more chronic phase of inflammation.5 Because of the importance in establishing a relevant model of IBD in small animals and because no information is available directly comparing these
Table 5. Time Course Study (pH 7.4)Groups
of CoJonic MPO Activity: Control
Colonic MPO activity (u/g dry M) Acetic acid ETOH TNBS (pH 1.0) TNBS (pH 7.4)
214 f 21 -
Comparison
2h
221+57 241 f18 17Ok 22 385 +31*
6h
267+47 225 + 33 214 k 34 336f 35
“P < 0.05compared with control. *P < 0.05 compared with other groups at the same time point. “P < 0.05 compared with acetic acid group at the same time point.
Among
Acetic
Acid, ETOH, TNBS (pH l.O), and TNBS
48 h
1009 f 95a 1272 k200' 1276+191' 1214 _+ill".'
I wk
1680 -c260' 1002 f196O 929 k184O 855+243"
2 wk
774 + 382' 768 f 242a 814 k 164" 1260+267"
COLONIC INFLAMMATION
May 1992
IN RATS
1533
Table 6. Time Course Study of Colon Weight: Comparison Among Acetic Acid, ETOH, TNBS (pH l.O), and TNBS (pH 7.4) Groups Control
Colon weight [g dry wt/cm) Acetic acid ETOH TNBS (pH 1.0) TNBS (pH 7.4)
zh
6h
48 h
0.021+ 0.001 0.023t 0.001 0.018+ 0.001 0.036f 0.024zk0.002 0.021* 0.001 0.038+ 0.022+ 0.001 0.027+ O.OOZatb 0.032f 0.021+ 0.002 0.017rf: 0.001 0.033+
0.003' 0.003" 0.002' 0.005"
1 wk
zwk
0.032t 0.001' 0.029t 0.024k 0.001 0.031k 0.041+_0.002°.b 0.036A 0.028z!z 0.002" 0.031f
O.OO1° 0.002" 0.002"~c o.oo2a
“P < 0.05compared with control. bP < 0.05 compared with other groups at the same time point. “P < 0.05 compared with acetic acid group at the same time point.
of ethanol and TNBS (buffered or unbuffered) produced more colonic injury at l-2 weeks than acetic acid or ethanol alone (Table 2). However, we found that only E + TNBS (pH 7.4) group produced consistently and significantly more epithelial cell injury (mucosal permeability) than the acetic acid or ethanol groups. The reasons for this discrepancy are unclear; however, there are several possibilities. First, it may be that sex and/or strain differences account for some of the observed differences. Beagly et al. have shown that the strain of the animal may be very important in producing a chronic inflammation in the rat colon.” Second, it may be that differences in TNBS preparations account for at least some of these differences. We have found that different lots of TNBS from the same vendor may produce different degrees of injury and inflammation using the same sex and strain of rats. Finally, our data suggest the variability in response to E + TNBS (pH 1.0) is much greater than for the other agents (Figure 3C and D). It should be noted that the intrarectal application of acetic acid or ethanol produced very consistent injury and inflammation from day to day and animal to animal (data not shown). The mechanism(s) by which acetic acid produces inflammation in the rat colon appear to involve the entry of the lipid soluble form (protonated) of the acid into the epithelium, where it dissociates to liberate protons within the intracellular space. It is this massive intracellular acidification that most likely accounts for the epithelial injury observed, because intraluminal introduction of HCl (pH 2.3) did not produce injury or inflammationg*” (data not shown). Our studies also showed the dramatic recuperative powers of the colonic mucosa indicated by the rapid re-epithelialization, which appeared to correlate well with the decrease in mucosal permeability after the initial insult (Table 2; Figures l-4). The intraluminal administration of ethanol alone or E + TNBS (pH 1.0 or 7.4) produced inflammation in a time very similar to that observed with acetic acid. The proinflammatory activity of ethanol towards the colon was originally described by Wallace
et a1.14who showed that dilute solutions of ethanol produce rapid and severe injury to the colonic mucosa. We found that the intrarectal administration of 50% ethanol produced an acute inflammatory response that resolved over the subsequent 2 weeks (Table 2; Figure 2). The mechanism(s) by which E + TNBS induces colitis appear more complex than those for acetic acid or ethanol. Although TNBS is in fact a classic hapten, it is generally thought that animals need to be sensitized to a hapten before an immunologic response is possible. Furthermore, TNBS is not innocuous. For example, we have shown in this study that TNBS (pH 7.4) in the absence of ethanol is very toxic to intestinal epithelial cell monolayers in vitro (Figure 9, suggesting that some of the inflammation induced by the intrarectal or intramural administration of TNBS may result from TNBSmediated tissue injury. The mechanism by which TNBS mediates epithelial cell injury remains undefined; however we have recently shown that the rat colon as well as rat colonocytes are capable of metabolizing TNBS to produce large quantities of reactive oxygen metabolites such as superoxide, hydrogen peroxide, and hydroxyl radical in vitro.” Formation of these oxidants within the mucosa would be extremely injurious and could initiate an inflammatory response. Indeed, the production of reactive oxygen metabolites by cyclical reduction and spontaneous oxidation reactions (redox cycling) of the nitro functional groups on aromatic compounds is well characterized and has been suggested as one mechanism to account for side effects of metronidazole.20p21 Taken together, our data suggest that all four agents are effective in injuring the colonic mucosa and inducing an acute colitis in rats. However, it appears that the combination of ethanol and TNBS may produce a more chronic mucosal injury and inflammation than acetic acid and ethanol alone. We suggest any one of the models of colitis are useful to study those events that occur at the time of inflammation such as (arachidonate metabolism, granulocyte infiltration, and metabolism etc) or during repair. However, the use of these models of colitis
1534
YAMADA ET AL.
may have significant limitations in understanding those immunologic events that initiate the acute and chronic inflammatory episodes. For example, the inflammation and tissue injury observed in human IBD is most probably a result of some sort of inappropriate immunologic activation (e.g., autoimmune, infectious agent, etc.), whereas the inflammation induced by the intrarectal application of acetic acid, ethanol, or E + TNBS is, at least initially, a result of extensive mucosal injury. Thus, the mechanisms by which inflammation (and mucosal injury) are achieved in the human disease may be very different than those in the experimental models. References 1.
MacPherson BR, Pfeiffer CJ. Experimental production of diffuse colitis in rats. Digestion 1978;17:135-150. 2. Sharon P, Stenson WF. Metabolism of arachidonic acid in acetic acid colitis in rats. Similarity to human inflammatory bowel diseases. Gastroenterology 1985;88:55-63. 3. Sekizuka E, Grisham MB, Li M, Deitch EA, Granger DN. Inflammation-induced intestinal hyperemia in the rat: role of neutrophils. Gastroenterology 1988;95:1528-1534, 4. Fedorak RN, Empey LR, MacArthur C, Jewel1 LD. Misoprostol provides a colonic mucosal protective effect during acetic acid-induced colitis in rats. Gastroenterology 1990;98:615625. 5. Morris GP, Beck PL, Herridge MS, Depew WT, Szewczuk MR, Wallace JL. Hapten-induced model of chronic inflammation and ulceration in the rat colon. Gastroenterology 1989;96: 795-803. 6. Wallace JL. Release of platelet-activating factor (PAF) and accelerated healing induced by a PAF antagonist in an animal model of chronic colitis. Can J Physiol Pharm 1988;66:422425. 7. Allgayer H, Deschryver K, Stenson WF. Treatment with 16,16’-dimethyl prostaglandin E, before and after induction of colitis with trinitrobenzene sulfonic acid in rats decreases inflammation. Gastroenterology 1989;96:1290-1300. 8. Wallace JL, MacNaughton WK, Morris GP, Beck PL. Inhibition of leukotriene synthesis markedly accelerates healing in a rat model of inflammatory bowel disease. Gastroenterology 1989;96:29-36. 9. Zeitlin IJ, Norris AA. Animal models of colitis. British Society of Gastroenterology/Smith Kline and French Laboratories workshop. Stanstead Abbotts, England, 1983:70-73.
GASTROENTEROLOGY
Vol. 102, No. 5
10.
Strober W. Animal models of inflammatory bowel diseasean overview. Dig Dis Sci 1985;3O(Suppl):3S-10s. 11. Bustos-Fernandez L, Gonzalez E, de Paolo IL, Celener D, de Furuya KO. Organic anions induce colonic secretion. Dig Dis 1976;21:329-332. 12. Kunin S, Gallily R. Recognition and lysis of altered-self cells by macrophages. Modification of target cells by 2,4,6-trinitrobenzene sulfonic acid. Immunology 1983;48:265-272. 13. Beck PL, Morris, GP, Wase AW, Szewczuk, M, Wallace JL. Immunological manipulation of disease progression in a rat model of chronic inflammatory disease of the colon. In: MacDermott RP, ed. Inflammatory bowel disease: current status and future approaches. Amsterdam: Elsevier Science, 1988:201-206. 14. Wallace JL, Whittle BJR, Boughton-Smith NK. Prostaglandin protection of rat colonic mucosa from damage induced by ethanol. Dig Dis Sci 1985;30:866-876. 15. Grisham MB, Volkner C, Tso P, and Yamada T. Metabolism of trinitrobenzene sulfonic acid by the rat colon produces reactive oxygen species. Gastroenterology 1991;101:540-547. 16. von Ritter C, Sekizuka E, Grisham MB, Granger DN. The chemotactic peptide N-formyl-methionyl-leucyl-phenylalanine increases mucosal permeability in the distal ileum of the rat. Gastroenterology 1988;95:651-655. 17. Grisham MB, Benoit JN, Granger DN. Assessment of leukocyte involvement during ischemia and reperfusion of intestine. Methods Enzymol 1990;186:729-742. 18. Sartor RB, Cromartie WJ, Powell DW, and Schwab JH. Granulomatous enterocolitis induced in rats by purified bacterial cell wall fragments. Gastroenterology 1985;89:587-595. 19. Beagly KW, Black CA, Elson CO. Strain differences in susceptibility to TNB-induced colitis (abstr). Gastroenterology 1991;100:A560. 20. Rao DNR, Harman L, Motten A, Schreiber J, Mason RP. Generation of radical anions of nitrofurantoin, misoidazole, and metronidazole by ascorbate. Arch Biochem Biophys 1987;255: 419-427. 21. Rao DNR, Mason RP. Generation of nitro radical anions of some %nitrofurans, 2- and 5-nitroimidazoles by norepinephrine, dopamine, and serotonin. J Biol Chem 1987;262:1173111736.
Received July 9, 1990. Accepted September 23,1991. Address requests for reprints to: Matthew B. Grisham, Ph.D., Department of Physiology and Biophysics, Louisiana State University Medical Center, 1501Kings Highway, P.O. Box 33932, Shreveport, Louisiana 71130-3932. Supported by grants from the National Institutes of Health (DK 39168, DK 33720), the Crohn’s and Colitis Foundation of America, and by funds from the Feist Research Foundation.