GASTROENTEROLOGY

1992:103:188-196

Chemiluminescence Assay of Mucosal Reactive Oxygen Metabolites in Inflammatory Bowel Disease

NICOLA J. SIMMONDS, ROS E. ALLEN, TIM R. J. STEVENS, R. NIALL M. VAN SOMEREN, DAVID R. BLAKE, and DAVID S. RAMPTON Inflammation England

Group, Gastrointestinal

Science Research

Previous studies suggesting increased reactive oxygen metabolite (ROM) production in inflammatory bowel disease have been restricted to peripheral blood and isolated intestinal phagocytes. In the current study, chemiluminescence and the effect of various scavengers, enzymes, and enzyme inhibitors were used to show that ROMs account for the increased production of oxidants by colorectal mucosal biopsy specimens in inflammatory bowel disease. Luminol-amplified chemiluminescence was increased in active ulcerative colitis [macroscopic grade 1: 25 photons - mg-’ - min -10m3(median), 8-47 (95% confidence intervals), n = 40; grade 2:89,65156, n = 30; grade 3: 247, 133-562,n = 131 and Crohn’s disease [mild: 9, 3-84, n = 6; severe: 105, 25-789 (range), n = 51 compared with normal-looking mucosa (ulcerative colitis: 0.8, 0.4-1.4, n = 22, P < 0.01; Crohn’s disease: 0.8,0.1-2, n = 6, P < 0.05) and controls (0.6, 0.04-1.4, n = 52, P < 0.01). In ulcerative colitis, luminol chemiluminescence correlated with microscopic inflammation (Spearman’s p = 0.74, P = 0.0001) and was decreased by sodium azide (-89%, P < 0.05), taurine (--31%, P < 0.05), catalase (-23%, P < 0.05), and dimethyl sulfoxide (-29%, P < 0.05). Superoxide dismutase and oxypurinol decreased lucigenin chemiluminescence in ulcerative colitis by -63% (P< 0.05) and -27% (P < 0.05),respectively. Luminol chemiluminescence chemiluminescence correlated with lucigenin (Spearman’s p = 0.62, P = 0.0001) and myeloperoxidase activity (Spearman’s p = 0.72, P = 0.003). These results suggest that neutrophil-derived oxidants (superoxide, hydrogen peroxide, hydroxyl radical, and hypochlorite) are generated in colorectal mucosa in active inflammatory bowel disease and support the hypothesis that production of such metabolites by neutrophils is of major pathogenetic importance.

Unit, London Hospital Medical College, London,

eactive oxygen metabolites (ROMs) have been implicated in many inflammatory disorders, including those of the gastrointestinal tract.lm3 Potential sources of ROMs in the gastrointestinal tract include the xanthine, amine, and aldehyde oxidase systems of gut epithelial cells4 and the NADPH oxidase of resident macrophages. The influx of neutrophils and monocytes associated with inflammation can produce further ROMs via the enzymes of the respiratory burst (NADPH oxidase and myeloperoxidase) and those involved in prostaglandin and leukotriene metabolism.5 The role of ischemia in the pathogenesis of inflammatory bowel disease (IBD) has received support recently.6 Ischemia results in the conversion of xanthine dehydrogenase to the oxidase with formation of superoxide (OJ and hydrogen peroxide (H,O,) on reperfusion. Neutrophils are attracted to the site and produce further ROMs and hence tissue damage.4 In patients with IBD, studies on peripheral blood have shown increased oxidative metabolism in monocytes but conflicting results in neutrophils.7-‘7 This may in part be explained by differences in the cell separation procedure and stimulus used“j but also suggests that events in peripheral blood may not reflect those in the gut mucosa. Although enhanced mucosal synthesis of eicosanoids, which could result in increased production of ROMs, has been shown in IBD both in vivo and in vitr0’8,‘g and phagocytes isolated from the inflamed gut produce more ROMs, 20-22there appears to be no direct studies examining the production of ROMs by the inflamed intestinal mucosa itself. We have therefore investigated the production of oxidants by colorectal mucosal biopsy specimens using luminol- and

R

0 1992bytheAmerican Gastroenterological 0016-5085/92/$3.00

Association

July

REACTIVE OXYGEN METABOLITES IN IBD

1992

SQD

H20+02

187

horse heart).‘* Catalase was inactivated by heating the powder to 100°C for 2 hours and then leaving the solution in PBS at room temperature for 24 hours. Inactivation was confirmed using an assay based on the horseradish peroxidase (type II)-mediated oxidation of phenol red by H,0,.25 Reagents for myeloperoxidase staining were obtained from Sigma in kit form. Patients

Sodium azide Figure 1. Schematic representation of production of ROMs by a neutrophil to show the sites of action of the enzymes, enzyme inhibitors, and ROM scavengers used to identify the ROM produced in UC. The ROMs interacting with luminol to produce chemiluminescence are surrounded by an ellipse. Catalysis is represented by the solid arrow and inhibition or scavenging by the broken arrow. SOD catalyzes the dismutation of O,- to H,O, and 0,. Catalase catalyzes the conversion of H,O, to H,O and 0,. DMSO scavenges OH., taurine scavenges OCl-, and sodium azide inhibits myeloperoxidase, thus decreasing the formation of oc1-.

lucigenin-amplified chemiluminescence, relating it to disease activity, smoking habits, and treatment with aminosalicylates or steroids. The addition of various enzymes, enzyme inhibitors, and scavengers (Figure 1)was used to show that the oxidants measured were ROMs and to assess the contribution of different ROMs to the chemiluminescence observed.

Materials and Methods Reagents Chemicals were obtained from Sigma Chemical Co., Poole, Dorset, England. Luminol was made up as a stock solution in dimethyl sulfoxide (DMSO) at a concentration of 50 mg/mL, stored at 4”C, and diluted to a final concentration of 300 pmol/L in Dulbecco’s phosphate-buffered saline (PBS) with added glucose (5 mmol/L) on the day of the experiment. Lucigenin was dissolved in PBS at 300 pmol/L and stored at 4’C. Fresh solutions were made weekly, and 5 mmol/L glucose was added on the day of the experiment. Superoxide dismutase (SOD, from bovine erythrocytes; final concentration, 300 U/mL), catalase (from bovine liver, 3000 U/mL), taurine (20 mmol/L), and sodium azide (1 mmol/L) were all dissolved in PBS on the day of the experiment. DMSO was used at a final concentration of 5%. Oxypurinol was dissolved initially in 1 mmol/L NaOH and then used at a final concentration of lo-* mmol/L in PBS. Heat-inactivated SOD was prepared by heating a solution in PBS in a boiling water bath for 2 hours.‘3 Inactivation was confirmed using the xanthine/ xanthine oxidase (grade 1, from buttermilk) reaction as measured by the reduction of cytochrome c (type III, from

Biopsy specimens were obtained from patients undergoing routine sigmoidoscopy or colonoscopy. Each patient gave informed consent, and the study was approved by the Tower Hamlets Health Authority Ethical Committee. Control subjects included patients undergoing routine investigation for lower gastrointestinal symptoms or follow-up for polyps or cancer whose biopsy specimens had normal mucosal appearance. The diagnoses of Crohn’s disease (CD) and ulcerative colitis (UC) were based on conventional clinical, radiological, endoscopic, and histological criteria. Characteristics of subjects are given in Table 1. Disease activity was determined from the colorectal mucosal appearances at endoscopy. In CD, the mucosa was described as normal, mildly inflamed (hyperemic or with aphthoid ulcers only), or severely inflamed (frank ulcers and/or bleeding). In UC, appearances were graded according to standard criteria’” (0, normal; 1, mucosal edema or loss of vascular pattern; 2, contact bleeding; and 3, spontaneous bleeding). Measurement

of Chemiluminescence

Luminol and lucigenin react with oxidants, such as ROM, to form 3-aminophthalate and N-methylacridone. The excited electrons in these compounds revert to their ground state with the emission of energy as light (chemiluminescence), which can be detected by the photomultiplier tubes of a scintillation counter.z7 Biopsy specimens were washed immediately in preoxygenated PBS with added glucose (5 mmol/L), pH 7.4. They were then transferred to scintillation vials containing 1 mL

Table

1. Patient

Characteristics Control

No. Age [yr

(ransell

Sex (M/F) Smokers

52 53 (18-79)

UC 105

CD 17

48 (17-87)

38 (22-64)

53/52 8

8/9 10

16

11

1

0

85

9

0

38

0

0

34

7

0

12

3

28/24 16

Drugs None

5-ASA/SASP Topical steroids Oral steroids Azathioprine Metronidazole Disease extent

0

0

Distal 40 Left-sided 13 Subtotal 16 Total 36

3

Small bowel Ileocolonic Colonic

2 5 10

188

SIMMONDS ET AL.

GASTROENTEROLOGY Vol. 103, No. 1

minescence (arbitrarily defined as >50 X lo3 photons per minute after subtraction of background) after the initial measurement, and the vials were recounted immediately. The effects of these substances were compared with those of added PBS or inactivated enzyme, as appropriate, on specimens of comparable activity from the same patient. All studies were performed within 2.5 hours of the initial count. Control experiments in which repeated baseline counts were done failed to show any significant differences in the chemiluminescence observed over this time (data not shown). At the end of the experiment, the samples were blotdried and weighed and then placed in formalin pending histology or snap-frozen before staining for myeloperoxidase activity. 1

10

loo

LUMINOL CONCENTRATION (@,I) Figure 2. Chemiluminescence

observed using different concentrations of luminol. Results are shown as photons *mg-’ . min. 1Om3after subtraction of background. Each line represents the mean of two biopsy specimens from one patient.

of preoxygenated luminol or lucigenin, at a concentration of 300 ymol/L, in PBS with glucose (pH 7.4) and counted in a Packard Tri Carb 1600 CA liquid scintillation analyzer operated in the out-of-coincidence mode for 5 minutes. All vials were counted before the addition of the biopsy specimen. Chemiluminescence is expressed as photons - mg-’ min. 10e3 after subtraction of background. Luminol was used at a concentration of 300 umol/L to maximize the sensitivity of the assay. In vitro experiments using xanthine oxidase have shown that the maximal chemiluminescence response was obtained at this concentration.” Preliminary experiments using different concentrations of luminol suggested that 300 pmol/L luminol would give the best sensitivity (Figure 2). Because the quantum yield of lucigenin is comparable with that of luminol,2g it was decided to use both chemiluminigenic probes at the same concentration. Repeatability3’ was assessed using the results from patients in whom two biopsy specimens were obtained from the same site at the same time. Because the differences between biopsy specimens were related to the mean values, the coefficient of repeatability was calculated on log transformed values. Results, given as the limits of agreement of the percentage difference of the second reading from the first, were as follows: Juminol-amplified chemiluminescence: control (43 pairs), 85%-1087%; inactive UC (grades 0 and 1, 46 pairs), 86%-971%; active UC (grades 2 and 3, 35 pairs), 86%-661%; and CD (14 pairs), 93%-264%; Jucigenin-amplified chemiluminescence: control (20 pairs), 77%-276%; UC (27 pairs), 73%-262%; and CD (15 pairs), 72%-276%. Although the scatter of paired biopsy specimens from the same patient is wide, a lo-fold intrapatient variation remains small in the context of a 400-fold increase in chemiluminescence in patients with active UC compared with controls. To identify the ROMs observed in UC, enzyme inhibitors, enzymes, or ROM scavengers (Figure 1) were added to biopsy specimens showing measurably increased chemilu-

Histological

Assessment

Each biopsy specimen was assessed blindly by two independent observers, unaware of the macroscopic grading or the result of chemiluminescence, according to the severity of changes in the crypts, goblet cells, and inflammatory cells in the lamina propria. The scoring system was adapted from a similar system used in other studies31 and is summarized in Table 2. Staining

for Myeloperoxidase

Activity

Sections were fixed in glutaraldehyde-acetone fixative for 10 minutes and then stained according to the Sigma procedure no. 391, based on the myeloperoxidasemediated oxidation of diaminobenzidine, using double the stated reaction times. Each section was scored by an observer blinded to the macroscopic grade and the result of chemiluminescence and was given an arbitrary score (O-6).

Table 2. Histological in UC crypts Normal Single inflammatory Cryptitis Crypt abscesses Goblet cell depletion Normal Slight decrease Moderate decrease Marked increase Mononuclear cells Normal Slight increase Moderate increase Marked increase Neutrophils Normal Slight increase Moderate increase Marked increase

Assessment

of MucosaJ Biopsies

0

cells

Adapted with permission.3’

2 3 0

2

0 1

2

REACTIVE

July 1992

NS w

pco.oo1

pco.oo1

pco.013

InI-

7

??

1

.

8

Y

CONTROL

0

1 ETDOSCOPIC

I

I

2

3

OXYGEN

METABOLITES

IN IBD

189

controls and correlated with the degree of inflammation assessed macroscopically (Figure 3). Median values (95% confidence intervals in parentheses) expressed as photons 0mg-’ - min -10m3for each group were as follows: control, 0.6 (0.04-1.4); grade 0, 0.8 (0.4-1.4); grade 1, 25 (8-47); grade 2, 89 (65-156); and grade 3, 247 (133-562) (P = 0.0001, Kruskal-Wallis). Similarly, there were significant differences between patients with CD and controls (Figure 4): control, 0.6 (0.04-1.4); normal, 0.8 (0.1-2); mild inflammation, 9 (3-84); and severe inflammation, 106 (25-789) (P = 0.0001, Kruskal-Wallis). Lucigeninamplified chemiluminescence was also increased in UC (Figure 5) [control, 5.2 (4.2-8.4); grade 0, 10 (361); grade 1, 37 (2-85); grade 2,24 (4-29); and grade 3, 69 (range, 42-96) (P = 0.03, Kruskal-Wallis)] and CD (Figure 6) [control, 5.2 (4.2-8.4); normal, 12 (4-38); mild, 8 (5-66); and severe, 94 (range, 61-120) (P = 0.03, Kruskal-Wallis)], Luminoland lucigenin-amplified chemiluminescence were well correlated (Spearman’s p = 0.62, P = 0.0001).

,

NS

, , p=O.O04

,,

NS

,

GRADE

Figure 3. The production of luminol-amplified chemiluminescence by colorectal mucosal biopsy specimens in UC according to the macroscopic grade assessed endoscopically” and controls. Results are shown as log photons - mg-’ . min-’ after subtraction of background. Points along the horizontal axis represent patients in whom chemiluminescence was less than or equal to background. Each point represents the mean of two or more specimens from the same patient. The horizontal bars represent the medians in each group.

. . .

Statistics Differences among groups were assessed using the Kruskal-Wallis test and differences between groups by the Mann-Whitney U test using the Bonferroni adjustment to correct for multiple comparisons.32 P values < 0.01 for the differences between groups were considered significant. Correlation with microscopic appearances and with myeloperoxidase activity were assessed using Spearman’s rank correlation test. The effect of smoking and of taking oral aminosalicylates or steroids on the chemiluminescence within a particular macroscopic group was assessed using the Mann-Whitney LJ test with corrections for multiple comparisons. The effect of inhibitors was determined by using the Wilcoxon signed rank test for paired variables. Results Chemiluminescence According Macroscopic Grading

cence

In endoscopically was significantly

to

active UC, chemiluminesincreased compared with

-P;

* .

8

I

COSTROL

SORMAL

I

I

MILD

ESDOSCOI’IC

SEVERE GRADE

Figure 4. The production of luminol-amplified chemiluminescence by colorectal mucosal biopsy specimens in CD according to the macroscopic grade assessed endoscopically and controls. Results are shown as log photons - mg-’ . min-’ after subtraction of background. Points along the horizontal axis represent patients in whom chemiluminescence was less than or equal to background. Each point represents the mean of two or more specimens from the same patient. The horizontal bars represent the medians in each group.

190

SIMMONDS ET AL.

GASTROENTEROLOGY Vol. 103, No. 1

6r

: 8

1

24 CONTROL

4

1

0

3

2

ENDOSCOPIC

Chemiluminescence According Microscopic Grading

0 0

0

(Figure 7) (Spearman’s

Chemiluminescence According Myeloperoxidase Activity

--i0

22 MILD

ENDOSCOPIC

p = 0.74,

to

Chemiluminescence was positively correlated with myeloperoxidase activity in a group of control patients (n = 8) and patients with IBD (n = 10; Spearman’s p = 0.72, P = 0.003) (Figure 8). With Drugs

The results of luminol-amplified chemiluminescence within a particular macroscopic grade were analyzed according to smoking habits (Table 3) and use of aminosalicylates (Table 4) or steroids (Table 5) at the time of the investigation. No significant differences were found when the results were analyzed in this way. However, the numbers in some of the groups were small, making a type II error possible.

0

NORMAL

cence was less than or equal to background. Spearman’s p = 0.74, P = 0.0001.

Effect of Smoking and Treatment

0

CONTROL

SCORE

Chemiluminescence according to the microscopic disease activity3’ in UC. Results are shown as log photons *mg-’ +min- 1 after subtraction of background. Points along the horizontal axis represent patients in whom chemilumines-

sessed histologically P = 0.0001).

to

0

12 HISTOLOGICAL

In UC, chemiluminescence was positively correlated with the degree of inflammation as as-

:

O4

GRADE

Figure 5. The production of lucigenin-amplified chemiluminescence by colorectal mucosal biopsy specimens in UC according to the macroscopic grade assessed endoscopicallyz6 and controls. Results are shown as log photons - mg-’ . min-’ after subtraction of background. Each point represents the mean of two or more specimens from the same patient. The horizontal bars represent the medians in each group. Groups significantly differed (P = 0.03, Kruskal-Wallis),

-4-

t

SEVERE GRADE

Figure 6. The production of lucigenin-amplified chemiluminescence by colorectal mucosal biopsy specimens CD according to the macroscopic grade assessed endoscopically and controls. Results are shown as log photons. mg-’ - min after subtraction of background. Each point represents the mean of two or more specimens from the same patient. The horizontal bars represent the medians in each group. Groups significantly differed (P = 0.03, Kruskal-Wallis).

Effect of Inhibitors Luminol-amplified chemiluminescence was significantly decreased by the addition of sodium azide, which inhibits myeloperoxidase (-89%, P < 0.05). There were smaller but significant decreases with catalase, which breaks down H,O,, (-23%, P < 0.05); DMSO, a hydroxyl radical (OH a) scavenger (-29%, P < 0.05); and taurine, a hypochlorite scavenger (-31%, P < 0.05). SOD decreased lucigenin-amplified chemiluminescence (-63%, P < 0.05) but not

July 1992

REACTIVE

.

. 0

. X X

X

.

.

0

so[_ / , , , , , 0

1

2

4

3

MYELOPEROXIDASE

5

6

ACTIVITY

Figure 8. Chemiluminescence according to myeloperoxidase activity. Results are shown as log photons - mg-’ . min-’ after subtraction of background. Points along the horizontal uxis represent patients in whom chemiluminescence was less than or equal to background. 0, Control patients; 0, patients with UC; X, patients with CD. Spearman’s p = 0.72, P = 0.003.

luminol-amplified chemiluminescence. Oxypurinol decreased lucigenin-amplified chemiluminescence (-27%, P < 0.05) (Figure 9). Discussion Most of the evidence for a role for ROMs in the pathogenesis of IBD is indirect. Peripheral blood monocytes from patients with IBD show increased oxidative metabolism in response to a variety of stimuli, but results in peripheral blood neutrophils have been conflicting.7-‘6 Antioxidant defenses, such as SOD3”,34and glutathione,35 are decreased, and drugs that are effective in the treatment of IBD, salicylazosulfapyridine (SASP) and aminosalicylic acid (5-

OXYGEN

METABOLITES

IN IBD

191

ASA), have been shown to scavenge OH - and hypochlorite (OCl-), to inhibit myeloperoxidase, and to protect cells from the oxidative damage produced by activated neutrophils both in culture and in animal models.364” Preliminary uncontrolled results on the use of liposome-encapsulated SOD and CuZnSOD in the treatment of CD have been promising,47 and stimulants of neutrophil function, such as formylmethionyl-leucyl-phenylalanine48~4g and phorbol12-myristate-13-acetate,50 have been used to produce animal models of colitis. The only direct evidence so far is the demonstration of increased production of ROMs in response to various stimuli by phagocytic cells isolated from the inflamed bowel of patients with active IBD20-22and the evidence of increased lipid peroxides, suggesting oxidative damage, in rectal biopsy specimens from such patients51 However, studies on isolated intestinal macrophages have failed to show any differences using unstimulated cells”; also, measurement of lipid peroxides using the thiobarbituric acid reactio?l is difficult to interpret, because it also measures endoperoxides produced enzymically by the prostaglandin synthesis pathway.52 Chemiluminescence is a nonspecific but sensitive method of detecting oxidizing species. Using this technique, we have directly shown disease activityrelated production of such oxidants by inflamed colorectal mucosal biopsy specimens from patients with IBD. Experiments with various enzymes, enzyme inhibitors, and scavengers (Figure 9) confirmed that these oxidants are ROMs; they also confirm the findings of previous in vitro studies showing that lucigenin reacts directly with superoxide, whereas luminol detects O,- and H,O, predominantly via a myeloperoxidase-catalyzed reaction.2g The current experiments and the correlation of luminol-amplified chemiluminescence with myeloperoxidase activity also suggest that the neutrophil-derived oxidants O,-, H,O,, OH - , and OCl- are all produced in the colorectal mucosa of patients with IBD.

Table 3. Effect of Smoking on Observed Chemiluminescence in Controls and Patients With UC Smokers

n Control UC0 UC1 UC2 UC3

16 3 7 2 1

Nonsmokers

Chemiluminescence (photons * mg . min 0.8 27 16 45 5.6

(0.04-2.1) (0.6-50) (0.3-129) (1.6-88) -

* 10~~)

n 36 19 33 28 12

Chemiluminescence (photons. mg-’ . min. 0.4 0.7 33 107 294

(~0.2 to 1.6) (-1.3 to 1.4) (7-55) (39-169) (133-628)

IO-~)

P NS NS NS NS NS

NOTE. Chemiluminescence, given as median (95% confidence intervals or range where n < 6), was determined within each endoscopic macroscopic grade according to whether patients smoked or not. The numerical suffix to UC indicates the macroscopic grade (see Materials and Methods, reference 26). Data on patients with CD are not given because the numbers in each group were too small to draw any conclusions.

192

GASTROENTEROLOGY Vol. 103, No. 1

SIMMONDS ET AL.

Table 4. Effect ofAminosalicyJate Treatment on Observed Chemiluminescence in Patients With LJC Patients on SASP/5-ASA

n UC0

20

UC1 UC2 UC3

29 24 12

Patients not on SASP/S-ASA

Chemiluminescence (photons - mg-’ *min. lOma) 0.8 (0.4-1.4) 40 137 294

Chemiluminescence (photons. mg-’ *min * 10~~)

n 2 11 4 1

(4-120) (65-209) (94-628)

137 36 30 161

(65-209) (6-82) (2-66) -

P NS NS NS NS

NOTE. Chemiluminescence, given as median (95% confidence intervals or range where n < 6), was determined within each endoscopic macroscopic grade according to whether patients were taking SASP or 5-ASA or not. The numerical suffix to UC indicates the macroscopic grade (see Materials and Methods, reference 26). Data on patients with CD are not given because the numbers in each group were too small to draw any conclusions.

Neutrophils are the most likely candidates for the excessive production of ROMs observed in IBD because they are the main site of myeloperoxidase activity and account for up to 20% of the cells found in the lamina propria in active IBD.15 Although peripheral blood monocytes and isolated intestinal macrophages have been shown to undergo an increased respiratory burst in patients with active IBD, monocytes lose their myeloperoxidase as they mature into tissue macrophages,53 making it unlikely that they are responsible for the chemiluminescence produced by colorectal biopsy specimens. Studies of animal models of colitis have confirmed that myeloperoxidase activity is related to neutrophil infiltration and not to the presence of histiocytes.54 Whereas the marked decrease in chemiluminescence caused by the addition of sodium azide (Figure 9) is probably largely caused by its inhibition of myeloperoxidase and, therefore, the peroxidase-catalyzed reaction of ROMs with lumino1,27~2gazide is a relatively nonspecific agent, also inhibiting other heme proteins such as catalase and quenching singlet oxygen and OH - .55 The relatively small inhibition of chemiluminescence observed with taurine (Figure 9), a hypochlorite scavenger,44 suggests that hypochlorite is detected in this system but that the

Table 5. Effect of Steroid

Treatment

on Observed

peroxidase-catalyzed reaction of oxidants with luminol makes a greater contribution to the observed chemiluminescence. It is unlikely that we were observing a nonspecific toxic effect of azide, because experiments measuring the luminol-amplified chemiluminescence of isolated latex-stimulated neutrophils have shown that the inhibitory effect of 1 mmol/L azide was much less marked (20%) when it was added after the stimulus than when it was added before (1OO%).5”Such a finding would be difficult to explain if azide at this concentration had a purely cytotoxic effect. Although, theoretically, singlet oxygen can be formed by the reaction of OCl- and H,O, under alkaline conditions, there is as yet no convincing evidence of its production under physiological conditions; attempts to detect its presence using more specific inhibitors are hampered by the fact that they are oxidized by OC1-.52 It is also unlikely, in our studies, that quenching of OH. accounts for the reduction in chemiluminescence caused by the addition of sodium azide, because the OH - scavenger, DMSO, had a much smaller inhibitory effect. Any inhibition of catalase by Na azide would be expected to cause an increase in chemiluminescence, as a result of decreased breakdown of H,O,, rather than the observed decrease.

Chemiluminescence

in Patients

Patients on steroids

n UC0 UC1 UC2 UC3

10 27 13 10

Chemiluminescence (photons * mg-’ - min. IO-~) 10 36 136 294

(-2.6 to 41) (3.9-73) (35-209) (94-628)

With UC

Patients not on steroids

n 12 13 17 3

Chemiluminescence (photons. mg-’ . min. 10~~) 0.8 16 88 161

(-1.3 to 1.4) (0.7-82) (38-164) (49-1015)

P NS NS NS NS

NOTE. Chemiluminescence, given as median (95% confidence intervals or range where n < 6), was determined within each endoscopic macroscopic grade according to whether patients were taking steroids or not. The numerical suffix to UC indicates the macroscopic grade (see Materials and Methods, reference 26). Data on patients with CD are not given because the numbers in each group were too small to draw any conclusions.

REACTIVE

July 1992

LUMINOL CL

12.0 ;

199

5

80

g

60

9

40

@ -120 t

CATALASE SOD (“=“) (n=7)

TAURINE

DMSO (“=‘)sODIUM AZIDE (n=7)

(n=lO)

JCIGENIN

CL_

_I SOD (“76) OXYPURINOL (n=S)

Figure 9. The effect of inhibitors on chemiluminescence in UC. Results are expressed as the percentage change in chemiluminescence on adding the inhibitor compared with that of adding the appropriate control; the actions of SOD and catalase were compared with those of adding heat-inactivated enzyme and those of DMSO, taurine, sodium azide, and oxypurinol to added PBS. The vertical bars represent 95% confidence intervals.

Neutrophil production of OCI- may be harmful in several ways. Hypochlorite is a powerful oxidant that may act directly on membrane-associated targets or indirectly by forming less reactive chloramines, which are then able to diffuse across the plasma membrane and attack cytosolic components.57 Hypochlorite is also involved in the inactivation of protective antiproteases and activation of collagenases, hence transforming the oxidizing potential of OCI- into enzyme-catalyzed degradation of proteins and collagens.57r58 The findings of low levels of mucosal trypsin inhibitor5’ and increased levels of circulating leucocyte elastase”’ and colorectal mucosal collagenase61 in patients with IBD may reflect the activity of OCl- in patients with active disease. A role for neutrophil-derived OCI- is also supported by the therapeutic efficacy of drugs known to scavenge OCl- and inhibit myeloperoxidase in both IBD and in animal models of IBD.3g,42-46 The decrease in lucigenin-amplified, but not luminol-amplified, chemiluminescence using SOD suggests that any effect of SOD in the luminol-dependent system is masked by the catalytic effect of myeloperoxidase. Although O,- does react directly with 1umino16’ and is thought to be a crucial intermediate in the high-quantum yield luminescence of lumino16” 4% of O,- radicals react to bring about a net oxidation of luminol. In addition, SOD has been shown to be inactivated by H,0,.64 Our findings are also consistent with the demonstration that SOD in-

OXYGEN

METABOLITES

IN IBD

193

hibits lucigenin-amplified chemiluminescence of neutrophils stimulated by phorbol myristate acetate, but not luminol-amplified chemiluminescence.65 The role of O,- production by xanthine oxidase in reperfusion injury, both in attracting neutrophils to the site of injury and in mediating injury via OH. and myeloperoxidase activity, is well established.“6067Preliminary studies on the use of the xanthine oxidase inhibitor, allopurinol, in patients with pouchitis6’ after colectomy for UC suggest that the xanthine/xanthine oxidase system may have a pathophysiological role in IBD. The small decrease observed on the addition of oxypurinol suggests that xanthine oxidase may also play a role; studies are currently in progress to elucidate this further. Of the other candidate ROMs, the small inhibitory effect of catalase on chemiluminescence (Figure 9) argues against a major role for H,O,. However, the large molecular weight of catalase may have prevented its access into the tissue after addition to the incubate; although H,O, itself is a stable oxidant and crosses cell membranes easily, neutrophils consume the bulk of this metabolite and only a small proportion of the H,O, produced appears in the extracellular pooL6’ H,O, as a ROM is relatively benign and is probably only harmful as a precursor of the more toxic ROMs, OH - and 0C1-.58 However, direct administration of high concentrations of H,O, to human colonic mucosa is known to cause inflammation7’ and pretreatment of endothelial cells with H,O, has been shown to increase adherence of neutrophils, an effect that can be inhibited with 5-ASA.71 The small decrease in chemiluminescence produced by DMSO (Figure 9) an OH. scavenger, suggests that OH - is one of the ROMs detected in colorectal mucosal biopsy specimens from patients with UC. Although there is no convincing direct evidence for the production of OH - by stimulated neutrophils, several in vivo and whole tissue studies have implicated OH. in neutrophil-mediated cell damage.58 However, lowering OH. production with DMSO or desferrioxamine had no effect on the severity of inflammation in an animal model of colitis7’; also, whereas the therapeutically active drugs 5-ASA and 4-ASA73 both scavenge OCll and inhibit myeloperoxidase,45 only 5-ASA scavenges OH - ,41This makes it unlikely that OH* plays a major pathogenetic role in IBD. In view of the strong negative association between UC and smoking,74 reduced mucosal ROM production might have been predicted in the patients who smoked. In the event, there was little difference in either controls or patients (Table 3) although the numbers of smokers with UC studied were too small to exclude an effect. Although high concentrations of nicotine and cotinine inhibit the production of

194

SIMMONDS ET AL.

ROMs from stimulated neutrophils, this effect is unlikely to be of clinical relevance in UC, because plasma concentrations of these compounds after a cigarette are much lower than those found to be inhibitory.75 The beneficial effects of 5-ASA are thought to be caused, at least in part, by its ROM scavenging properties5* We postulated that patients taking oral SASP or 5-ASA show less chemiluminescence for a given macroscopic grade than patients not taking these drugs. In fact, no such differences were observed (Table 4). The numbers of patients not taking aminosalicylates were small in both groups: further studies are necessary to confirm these findings. Although there is a change in the molecular structure of 5-ASA when it interacts with oxidants, the resulting compound does not seem to react with luminol to produce chemiluminescence because 5-ASA is a powerful inhibitor of luminol-dependent chemiluminescence of stimulated neutrophils in vitro.” In vivo, 5-ASA may control inflammation by scavenging oxidants and acting as an alternative substrate for myeloperoxidase.45 However, under the experimental conditions used, the concentration of 5-ASA will be much lower than that found in vivo and may allow detection of oxidants that would not have been present in vivo. Treatment with steroids seemed to have no effect on the chemiluminescence observed (Table 5). Again, numbers are too small to exclude a type II error, but this data would fit with the in vitro demonstration of lack of antioxidant activity of steroids, except at very high concentrations.76 In conclusion, we have shown increased production of ROMs by the colorectal mucosa of patients with IBD, with the amount generated correlating well with disease activity assessed macroscopically, microscopically, and by staining for myeloperoxidase. Final proof of a pathogenetic role for ROMs in IBD, however, awaits the demonstration that specific antioxidant therapy in vivo is beneficial. With this end in mind, studies using our simple in vitro system to evaluate potential new antioxidant treatments are in progress. References 1. 2.

3.

4.

5.

Braganza JM. Towards antioxidant therapy for gastrointestinal disease. Curr Med Lit Gastroenterol 1989;8:99-106. Parks DA, Bulkley GB, Granger DN. Role of oxygen-derived free radicals in digestive tract diseases. Surgery 1983;94:415422. Cross CE, Halliwell B, Borish ET, Pryor WA, Ames BN, Saul RL, McCord JM, Harman D. Oxygen radicals and human disease. Ann Intern Med 1987;107:526-545. Grisham MB, Granger DN. Neutrophil-mediated mucosal injury. Role of reactive oxygen metabolites. Dig Dis Sci 1988;33:6S-15s. Blake DR, Allen RE, Lunec J. Free radicals in biological sys-

GASTROENTEROLOGY Vol.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17. 18. 19.

20.

21. 22.

23.

103, No. I

terns-a review oriented to inflammatory processes, Br Med Bull 1987;43:371-385. Wakefield AJ, Dhillon AP, Rowles PM, Sawyerr AM, Pittilo RM, Lewis AAM, Pounder RE. Pathogenesis of Crohn’s disease: multifocal gastrointestinal infarction. Lancet 1989;2: 1057-1062. Ward M, Eastwood MA. The nitroblue tetrazolium test in Crohn’s disease and ulcerative colitis. Digestion 1976;14:179183. Verspaget HW, Mieremet-Ooms MAC, Weterman IT, Pena AS. Partial defect of neutrophil oxidative metabolism in Crohn’s disease. Gut 1984;25:849-853. Verspaget HW, Elmgreen J, Weterman IT, Pena AS, Riis P, Lamers CBHW. Impaired activation of the neutrophil oxidative metabolism in chronic inflammatory bowel disease. Stand J Gastroenterol 1986;21:1124-1130. Worsaae N, Staehr Johansen K, Christensen KC. Impaired in vitro function of neutrophils in Crohn’s disease. Stand J Gastroenterol 1982;17:91-96. Faden H, Rossi TM. Chemiluminescent response of neutrophils from patients with inflammatory bowel disease. Dig Dis Sci 1985;30:139-142. Okabe N, Kuriowa A, Fujita K, Shibuta T, Yao T, Okamura M. Immunological studies on Crohn’s disease. vi. Increased chemiluminescent response of peripheral blood monocytes. J Clin Lab Immunol 1986;21:11-15. Suematsu M, Suzuki M, Kitahora T, Mura S, Suzuki K, Hibi T, Watanabe M, Nagata H, Asakura H, Tsuchiya M. Increased respiratory burst of leukocytes in inflammatory bowel disease -the analysis of free radical generation by using chemiluminescence probe. J Clin Lab Immunol1987;24:125-128. Kitahora T, Suzuki K, Asakura H, Yoshida T, Suematsu M, Watanabe M, Tsuchiya M. Active oxygen species generated by monocytes and polymorphonuclear cells in Crohn’s disease. Dig Dis Sci 1988;33:951-955. Williams JG, Hughes LE, Hallett MB. Toxic oxygen metabolite production by circulating phagocytic cells in inflammatory bowel disease. Gut 1990;31:187-193. Kelleher D, Feighery C, Weir DG. Chemiluminescence by polymorphonuclear leucocyte subpopulations in chronic inflammatory bowel disease. Influence of the cell separation procedure. Digestion 1990;45:158-165. Curran FT, Allan RN, Keighley MRB. Superoxide production by Crohn’s disease neutrophils. Gut 1991;32:399-402. Rampton DS, Hawkey CJ. Prostaglandins and ulcerative colitis. Gut 1984;25:1399-1413. Stenson WF. Leukotriene B4 in inflammatory bowel disease. In: Goebell H, Peskar BM, Malchow H, eds. Inflammatory bowel diseases-basic research and clinical implications. Lancaster, England: MTP, 1988:143-152. Mahida YR, Wu KC, Jewel1 DP. Respiratory burst activity of intestinal macrophages in normal and inflammatory bowel disease. Gut 1989;30:1362-1370. Verspaget H, Beeken W. Mononuclear phagocytes in the gastrointestinal tract. Acta Chir Stand Suppl 1985;525:113-126. Williams JG. Phagocytes, toxic oxygen metabolites and inflammatory bowel disease: implications for treatment. Ann R Co11 Surg Engl 1990;72:253-262. Heikkila RE. Inactivation of superoxide dismutase by diethyldithiocarbamate. In: Greenwald RA, ed. CRC handbook of methods for oxygen radical research. Boca Raton, FL: CRC, 1985:387-390.

24. Fridovich I. Cytochrome c. In: Greenwald RA, ed. CRC handbook of methods for oxygen radical research. Boca Raton, FL: CRC, 1985:312-315. 25. Pick E, Keisari Y. A simple calorimetric method for the mea-

July

26.

27.

28.

29.

30.

31.

32. 33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

REACTIVE OXYGEN METABOLITES IN IBD

1992

surement of hydrogen peroxide produced by cells in culture. J Immunol Methods 1980;38:161-170. Baron JH, Connell AM, Lennard-Jones JE. Variation between observers in describing mucosal appearances in proctocolitis. Br Med J 1964;1:89-92. Allen RC. Phagocyte oxygenation activities: quantitative analysis based on chemiluminescence. In: Scholmerich J, Andreeson R, Kapp A, Ernst M, Wood WG, eds. Bioluminescence and chemiluminescence. New perspectives. New York: Wiley, 1986:13-22. Wilhelm J, Vilim V. Variables in xanthine oxidase-initiated luminol chemiluminescence: implications for chemiluminescence measurements in biological systems. Anal Biochem 1986;158:201-210. Allen RC. Biochemiexcitation: chemiluminescence and the study of biological oxygenation reactions. In: Adam W, Cilento G, eds. Chemical and biological generation of excited states. New York: Academic, 1982309-344. Bland JM, Altman DC. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-310. Saverymuttu SH, Camilleri M, Rees H, Lavender JP, Hodgson HJF, Chadwick VS. Indium Ill-granulocyte scanning in the assessment of disease extent and disease activity in inflammatory bowel disease. A comparison with colonoscopy, histology, and fecal indium 111-granulocyte excretion. Gastroenterology 1986;90:1121-1128. Godfrey K. Statistics in practice. Comparing the means of several groups. N Engl J Med 1985;313:1450-1456. Mulder TPJ, Verspaget HW, Janssens AR, Pena AS, Griffioen G. Weterman IT, Lamers CBHW. Decreased levels of anti-oxidant proteins in the intestinal mucosa ofpatients with inflammatory bowel disease (IBD) (abstr). Gut 1990;31:A465. Verspaget HW, Pena AS, Weterman IT, Lamers CBHW. Diminished neutrophil function in Crohn’s disease and ulcerative colitis identified by decreased oxidative metabolism and low superoxide dismutase content. Gut 1988;29:223-228, Inauen W, Bilzer M, Rowedder E, Halter F, Lauterburg BH. Decreased glutathione (GSH) in colonic mucosa of patients with inflammatory bowel disease: mediated by oxygen free radicals (abstr)? Gastroenterology 1988;94:A199. Miyachi Y, Yoshioka A, Imamura S, Niwa Y. Effect of sulphasalazine and its metabolites on the generation of reactive oxygen species. Gut 1987;28:190-195. Suematsu M, Suzuki M, Miura S, et al. Sulfasalazine and its metabolites attenuate respiratory burst of leukocytes-a possible mechanism of anti-inflammatory effects. J Clin Lab Immunol 1987;23:31-33. Williams JG, Hallett MB. Effect of sulphasalazine and its active metabolite, 5-aminosalicylic acid, on toxic oxygen metabolite production by neutrophils. Gut 1989;30:1581-1587. Aruoma 01, Wasil M, Halliwell B, Hoey BM, Butler J, The scavenging of oxidants by sulphasalazine and its metabolites. A possible contribution to their anti-inflammatory effects? Biochem Pharmacol 1987;36:3739-3742. Gionchetti P. Guarnieri C. Campieri M, et al. Scavenger effect of sulfasalazine, 5-aminosalicylic acid, and olsalazine on superoxide radical generation. Dig Dis Sci 1990;36:174-178. Allgayer H. Hofer P, Schmidt M, Bohne P, Kruls W, Gugler R. Different radical scavenger properties of 5-amino-salicylic acid (5-ASA) and its derivatives. A comparative electron spin resonance (ESR) study (abstr). Gastroenterology 1990;98:A156. Grisham MB, Ryan E, Von Ritter C. 5-aminosalicyclic acid scavenges hydroxyl radical and inhibits myeloperoxidase activity (abstr). Gastroenterology 1987;92:A1416. Dull BJ, Salata K, Van Langenhove A, Goldman P. 5-aminosalicylate: oxidation by activated leukocytes and protection of

44.

45.

46.

47.

48.

49. 50.

51.

52.

53. 54.

55.

56.

57. 58.

59.

60.

61. 62.

63.

64.

195

cultured cells from oxidative damage. Biochem Pharmacol 1987;36:2467-2472. Dallegri F, Ottonello L, Ballestrero F, Ferrando F, Patrone F. Cytoprotection against neutrophil derived hypochlorous acid: a potential mechanism for the therapeutic action of s-aminosalisylic acid in ulcerative colitis. Gut 1990;31:184-186. Von Ritter C, Grisham MB, Granger DN. Sulfasalazine metabolites and dapsone attenuate formyl-methionyl-leucyl-phenylalanine-induced mucosal injury in rat ileum. Gastroenterology 1989;96:811-816. Craven PA, Pfanstiel J, Saito R, DeRubertis FR. Actions of sulfasalazine and 5-aminosalisylic acid as reactive oxygen scavengers in the suppression of bile acid-induced increases in colonic epithelial cell loss and proliferative activity. Gastroenterology 1987;92:1998-2008. Emerit J, Pelletier S, Tosoni-Verlignue D, Mollet M. Phase II trial of copper zinc superoxide dismutase (CuZnSOD) in treatment of Crphn’s disease. Free Radic Biol Med 1989;7:145-149. Chester JF, Ross JS, Malt RA, Weitzman SA. Acute colitis produced by chemotactic peptides in rats and mice. Am J Physiol 1985;121:284-290. LeDuc LE, Nast CC. Chemotactic peptide-induced acute colitis in rabbits. Gastroenterology 1990;98:929-935. Fretland DJ, Widomski DL, Anglin CP, Levin S, Gaginella TS. Colonic inflammation in the rodent induced by phorbol-12myristate-13-acetate: a potential model of inflammatory bowel disease: effect of SC-41930 (abstr). Gastroenterology 1990;98:A449. Ahnfelt-Ronne I, Nielsen OH, Christensen A, Langholz E, Binder V, Riis P. Clinical evidence supporting the radical scavenger mechanism of 5-aminosalicylic acid. Gastroenterology 1990;98:1162-1169. Halliwell B, Gutteridge JMC. Free radicals in biology and medicine. 2nd ed. Oxford, England: Clarendon, 1989:218-234. 385-387. Babior BM. Oxidants from phagocytes: agents of defense and destruction. Blood 1984;64:959-966. Krawisz JE, Sharon P, Stenson WF. Quantitative assay for acute intestinal inflammation based on myeloperoxidase activity. Gastroenterology 1984;87:1344-1350. Nurcombe HL, Edwards SW. Role of myeloperoxidase in intracellular and extracellular chemiluminescence of neutrophils. Ann Rheum Dis 1989:48:56-62. Hallett MB, Campbell AK. Two distinct mechanisms for stimulation of oxygen-radical production by polymorphonuclear phagocytes. Biochem J 1983;216:459-465. Weiss SJ. Tissue destruction by neutrophils. N Engl J Med 1989:320:365-376. Winterbourn CC. Neutrophil oxidants: production and reactions. In: Das DK, Essman WB, eds. Oxygen radicals: systemic events and disease processes. Basel, Switzerland: Karger, 1990:31-70. Playford RJ, Freeman T, Quinn C, Calam J. Deficiency of a mucosal trypsin inhibitor in patients with ulcerative colitis. Gut 1990;31:A1166. Adeyemi EO, Neumann S, Chadwick VS, Hodgson HJF, Pepys MB. Circulating human leucocyte elastase in patients with inflammatory bowel disease. Gut 1985;26:1306-1311, Sturzaker HG. Hawley PR. Collagenase activity in rectal and colonic biopsies. Proc R Sot Med 1975;68:33. Merenyi G, Lind J, Eriksen TE. The reactivity of superoxide (O,..) and its ability to induce chemiluminescence with luminol. Photochem Photobiol 1985;41:203-208. Miller EK. Fridovich I. A demonstration that O,- is a crucial intermediate in the high quantum yield luminescence of luminol. Free Radic Biol Med 1986;2:107-110. Hodgson EK. Fridovich I. The interaction of bovine superox-

196

65.

66. 67. 68.

69.

70. 71.

72.

SIMMONDS ET AL,

ide dismutase with hydrogen peroxide: inactivation of the enzyme. Biochemistry 1975;14:5294-5299. Allen RC. Phagocytic leukocyte oxygenation activities and chemiluminescence: a kinetic approach to analysis. Methods Enzymol 1986;133:449-493. Parks DA. Oxygen radicals: mediators of gastrointestinal pathophysiology. Gut 1989;30:293-298. McCord JM. Oxygen-derived free radicals in postischaemic tissue injury. N Engl J Med 1985;312:159-163. Levin KE, Pemberton JH, Phillips SF, Zinsmeister AR, Pezim ME. Effect of a xanthine oxidase inhibitor (allopurinol) in patients with pouchitis after ileal pouch-anal anastomosis (abstr). Gut 1990;31:A1168. Test ST, Weiss SJ. Quantitative and temporal characterization of the extracellular H,O, pool generated by human neutrophils. J Biol Chem 1984;259:399-405. Bilotta JJ, Waye JD. Hydrogen peroxide enteritis: the “snow white” sign. Gastrointest Endosc 1989;35:428-430. Inauen W, Granger DN, Kvietys PR. Hydrogen peroxide-induced neutrophil adherence to vascular endothelium: effects of 5-aminosalicylic acid, hydrocortisone and a monoclonal antibody against the adhesion molecule CDll/CD18 (abstr). Gastroenterology 1990;98:A658. Keshavarzian A, Morgan G, Sedghi S, Gordon JH, Doria M.

GASTROENTEROLOGY Vol. 103, No. 1

73. 74. 75.

76.

Role of reactive oxygen metabolites in experimental colitis, Gut 1990;31:786-790. Ireland I, Jewel1 DP. Sulphasalazine and the new salicylates. Eur J Gastroenterol Hepatol 1989;1:43-50. Calkins BM. A meta-analysis of the role of smoking in inflammatory bowel disease. Dig Dis Sci 1989;34:1841-1854. Srivatsava ED, Hallett MB, Rhodes J. Effect of nicotine and cotinine on the production of oxygen free radicals by neutrophils in smokers and non-smokers. Hum Toxic01 1989;8:461463. Levine PH, Hardin JC, Scoon KL, Krinsky NI. Effect of corticosteroids on the production of superoxide and hydrogen peroxide and the appearance of chemiluminescence by phagocytosing polynorphnuclear leukocytes. Inflammation 1981;5:19-27.

Received May 13, 1991. Accepted January 7, 1992. Address requests for reprints to: N. J. Simmonds, M.D., Inflammation Group, Gastrointestinal Science Research Unit, London Hospital Medical College, 26 Ashfield Street, London, El 2AJ, England. Supported by the Hilden Charitable Fund (N.J.S.), the National Association for Colitis and Crohn’s disease (T.R.J.S.), and the Arthritis and Rheumatism Council (D.R.B.).