Gut 1996; 39:407-415 407
Evaluating the antioxidant potential of new treatments disease a for inflammatory bowel using Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from rat model of colitis
A D Millar, D S Rampton, C L Chander, A W D Claxson, S Blades, A Coumbe, J Panetta, C J Morris, D R Blake
Abstract It is now well recognised that reactive oxygen Background-Reactive oxygen species species (ROS) such as superoxide (02%), the may mediate tissue injury in inflamma- hydroxyl radical (OH), hydrogen peroxide tory bowel disease. Aminosalicylates have (H202), hypochlorous acid (HOCl) and antioxidant activity and the antioxidants, oxidant derivatives, such as N-chloramines superoxide dismutase and allopurinol, are (RNHCI), are produced in excess by the ofreported benefit in inflammatory bowel inflamed mucosa in inflammatory bowel disease. disease (IBD) and may be pathogenic.`q The Aim-To develop a convenient technique predominant sources of ROS in the inflamed for testing the antioxidant potential of mucosa are probably activated mucosal standard and novel therapeutic agents for phagocytic leucocytes5 and episodes of use in inflammatory bowel disease. ischaemia reperfusion6 (Fig 1). Increased Methods-Amplified chemiluminescence arachidonic acid metabolism7 may also lead to was used to measure reactive oxygen excess mucosal ROS production. The resulting species production by colonic biopsy oxidant stress may overwhelm the endogenous specimens from rats with acetic acid defences that regulate ROS production during induced colitis and to assess the in vitro normal metabolism, particularly if there are effect of conventional antioxidants, comparatively low tissue levels of endogenous standard therapies and proposed novel antioxidants, as in the colonic mucosa.8 therapies for inflammatory bowel disease. Consequent tissue injury by ROS9 results from Results-The model was validated by their direct reaction with carbohydrates, demonstrating that the profile ofeffects on lipids, proteins, and DNA, stimulation of chemiluminescence ofacetic acid induced phospholipase A2,'0 5-lipoxygenase,," and colitis biopsy specimens given by con- neutrophil chemotaxis1,2 and activation of http://gut.bmj.com/ ventional antioxidants (sodium azide, transcription factors involved in cytokine catalase, copper-zinc superoxide dismu- release and cell growth, such as NFKB,l3 c-fos, tase, dimethyl sulphoxide, N-acetyl- c-myc and c-jun.'4 cysteine and ascorbate) and standard It has been suggested that the efficacy of therapies (5-aminosalicylate and hydro- current standard treatments is related to their cortisone) resembled that previously antioxidant actions.4 5-ASA is a potent anti- reported using biopsy specimens from oxidant and reduces mucosal ROS production on September 30, 2021 by guest. Protected copyright. ulcerative colitis. Human recombinant by inflamed human colorectal biopsy tissue in The Gastrointestinal Science manganese superoxide dismutase did not vitro,' as well as mucosal lipid peroxidation in A D Millar alter chemilumnescence. Two novel ulcerative colitis (UC) in vivo.'5 Hydro- D S Rampton compounds, LY231617 (10 mM) and cortisone and other glucocorticoids do not and Bone and Joint amflutizole (20 mM), reduced chemilumi- directly scavenge ROS but their anti- Research Units nescence by 98% (n=5, p=0.009) and 88% inflammatory effects include inhibition of C L Chander (n=5, p=0.03), respectively. neutrophil function.'6 A W D Claxson similarity of the albeit uncontrolled, trials have S Blades Conclusions-The Recent, A Coumbe chemiluminescence responses of colonic suggested that specific antioxidant treatment C J Morris biopsy specimens from acetic acid may be therapeutically effective in IBD. In D R Blake induced colitis and ulcerative colitis to a corticosteroid resistant Crohn's disease (CD) a The London Hospital range of conventional antioxidants and combination of superoxide dismutase and Medical College, standard treatments suggests that this desferrioxamine was reportedly effective, London model is a useful method for testing the while allopurinol has been used successfully in Lilly Research antioxidant potential of new therapies for acute and chronic pouchitis.'8 Antioxidants Laboratories, inflammatory bowel disease. The anti- have been used successfully in experimen- Lilly Corporate oxidant actions of dimethyl sulphoxide, tal models of intestinal inflammation.'9 20 Center, Indianapolis, ascorbate, and the novel compounds, Compounds with antioxidant activity should Indiana, USA amflutizole and LY231617 in this model therefore be investigated as potential treat- J Panetta suggest that these agents merit further ments for IBD. There are many compounds Correspondence to: assessment in the treatment of in- with known antioxidant activity and many new Dr A D Millar, Gastrointestinal Research flammatory bowel disease. potential therapies whose ability to reduce Unit, (Gut 1996; 39: 407-415) ROS production by inflamed colon is 26 Ashfield Street, London El 2AJ. unknown. It would therefore be useful to have Accepted for publication Keywords: ulcerative colitis, antioxidant, acetic acid, a method for screening potential treatments for 9 May 1996 aminosalicylates, free radicals, superoxide dismutase. this action. 408 Millar, Rampton, Chander, Claxson, Blades, Coumbe, Panetta, Morris, Blake
CF3 CH2NHEt
NH2 QI Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from z/ . X(CH3)3C _1 C(CH3)3 NHS COOH HO
Amflutizole LY 231617 Figure 2: The chemical structures ofamflutizole and LY231617.
Poole, Dorset, unless otherwise stated. The CuZn superoxide dismutase was from human erythrocytes (specific activity: 3610 U/mg protein). Catalase was from bovine liver attached to 4% agarose beads (specific acti- ( Ischaemia reperfusion ) vity: 150000-200000 U/g agarose). Rh-Mn superoxide dismutase (specific activity: 3500 Figure 1: Schematic representation ofthe major sources ofROS production by the inflamed mucosa in inflammatory bowel disease. Enzymes are shown in italics and three examples of U/mg protein) was provided by Bender, Ges exogenous scavengers in plain text. Superoxide produced by activated neutrophils or mbH, Vienna, Austria. Amflutizole (3-(3-tri- ischaemia reperfusion dismutates to hydrogen peroxide, catalysed by superoxide dismutase. fluoromethyl) -4 - aminoisothiazole-5-carboxy- Superoxide reduces Fe' to Fe'+ thusfacilitating the Haber- Weiss reaction by which hydroxyl irons are produced. Hydrogen peroxide is catalysed by myeloperoxidase to lic acid) and LY231617 (2,6-bis(1,1-dimethyl- hypochlorous acid. ethyl)-4-[[(1-ethyl)amino]methyl]phenol hy- drochloride) (Fig 2) were obtained from Lilly Recent findings from this laboratory have Research Laboratories, Indianapolis, USA. shown that ROS production by rectal biopsy Table I shows the compounds used and their specimens from patients with UC and CD can postulated mechanisms of antioxidant activity. be measured in vitro using a chemilumi- Luminol (5-amino-2,3-dihydrophthalazine- nescence technique, and that addition of 1,4-dione) was maintained as stock solution conventional antioxidants reduces the measur- (50 mg in 1 ml ofDMSO) for up to one month. able ROS.' Biopsy material from patients with Luminol and lucigenin (N,N'-dimethyl- active UC is, however, too limited for screening 9,9'-biacridinium dinitrate) were prepared multiple compounds. prior to the experiment in Dulbecco's
Acetic acid induced colitis in rats resembles phosphate buffered saline (D-PBS) with added http://gut.bmj.com/ UC in histology, eicosanoid production,2' and calcium (1.13 mM) and glucose (5 mM) and response to sulphasalazine.22 Inflamed colonic oxygenated for 10 minutes with 95% 02, 5% mucosa in acetic acid induced colitis is also CO2 to. maintain tissue viability, following known to produce excess ROS and anti- which the pH was adjusted to 7A4 with NaOH oxidants will reduce measurable ROS23: in or HC1 (1M). these studies, however, mucosal scrapings
rather than biopsy specimens were used and no on September 30, 2021 by guest. Protected copyright. comparison was made with human material, INDUCTION OF EXPERIMENTAL COLITIS IN RATS thus limiting the comparability with IBD. Acetic acid induced colitis was induced in male This study used full thickness colonic biopsy Wistar rats (150-200 g, Charles River, UK) specimens from acetic acid induced colitis using a modification of the method described and validated the screening technique by by MacPherson and Pfeiffer.26 The animals comparing results with previous studies using were fasted for 16 hours with access to water biopsy specimens from patients with active ad libitum. Each rat was sedated by brief UC, in which the conventional antioxidants respiration of 20-25% C02 followed by (sodium azide, taurine, dimethyl sulphoxide, anaesthesia with 35 mg/kg intraperitoneal N-acetylcysteine, ascorbate and the enzymes pentobarbital. An infant feeding tube (Pennine catalase and CuZn superoxide dismutase) and Healthcare FT-1608/40, outside diameter 2 two standard therapies for IBD (5-amino- mm) was inserted into the colon to 8 cm and salicylate (5-ASA) and hydrocortisone), were 2 ml of acetic acid (3% v/v in 0.90/o saline) or evaluated.' 24 25 In addition, we studied the saline alone (control animals) infused into the effects ofpotential new antioxidant approaches colon. The acetic acid/saline was retained in with human recombinant manganese super- the colon for 30 seconds, after which fluid was oxide dismutase (Rh-Mn superoxide dismu- withdrawn. The rats were killed at 24 hours by tase) and two novel antioxidants, LY231617 CO2 asphyxiation. and amflutizole, whose structures are shown in Figure 2. ASSESSMENT OF COLITIS Methods Macroscopic scoring At postmortem laparotomy, 6 cm of colon REAGENTS extending proximally from 2 cm above the anal All chemical reagents were of analytical grade margin was removed, split longitudinally, and were obtained from Sigma Chemical, pinned out on card, and the macroscopic Antioxidant potential ofnew treatmentsfor inflammatory bowel disease 409
TABLE I Study compounds Soluble in Drug/compound aqueous Chemiluminescence Control (study concentration) Antioxidant action (reference) solution amplifier solution Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from Conventional antioxidants Sodium azide (1 mM) Myeloperoxidase `/OH '` + Luminol D-PBS Catalase (3000 U/ml) 1H202 + Luminol HI* enzyme Taurine (20 mM) HOCI 56, + Luminol D-PBS CuZn superoxide dismutase (30/300 U/mi) 02-- + Lucigenin HI* enzyme Dimethylsulphoxide (1/5/10%) OH. 58 + Luminol D-PBS N-Acetylcysteine (20 mM) t Glutathione peroxidase/M H0C146 + Luminol D-PBS Ascorbate (20 mM) IHOCl47/thiyl/peroxyl/02 6 + Luminol D-PBS Standard treatments in IBD 5-ASA (1/10/20 mM) OH 59/02 - 601HOC1 6'/Fe chelator 62 + Luminol D-PBS Hydrocortisone (1 mM) Neutrophil function'6 + Luminol D-PBS Proposed treatments in IBD Rh-Mn superoxide dismutase (300 U/mI) 02 + Lucigenin HI* enzyme Amflutizole (1/10/20 mM) Inhibits xanthine oxidase52 - Luminol D-PBS LY231617 (1/10 mM) Not known - Luminol D-PBS
*=Heat inactivated, l=scavenges or inhibits, Ilstimulates. All compounds and controls were dissolved in Dulbecco's phosphate buffered saline (D-PBS). For most compounds control was D-PBS alone.
appearances of the colonic mucosa scored on released in the form of photons, which are a scale adapted from Morris et al27 ranging detected by the photomultiplier tubes and from 0-4: (0) no macroscopic change, (1) photon detectors that comprise the scintilla- mucosal erythema alone, (2) mild mucosal tion counter.28 Lucigenin responds more oedema, slight bleeding or small erosions, (3) specifically to superoxide than luminol29 and moderate oedema, bleeding ulcers or erosions, was therefore used to detect changes in chemi- (4) severe ulceration/erosions, oedema, and luminescence induced by superoxide dis- tissue necrosis. Intermediate values reflected mutase. Luminol and lucigenin were used at intermediate appearances. the same concentration (300 ,uM) as used in previous studies using human biopsy specimens.' Microscopic scoring Full thickness biopsy specimens were taken In some experiments randomly distributed full at random by cutting cross sectional strips thickness biopsy specimens were fixed in 10% (median (IQR) wet weight; 44 (31-78) mg) formol saline prior to wax embedding, of the dissected, inflamed colon. Biopsy sectioning, and staining with haematoxylin and specimens from colons of macroscopic score eosin. Biopsy specimens (n=27) from 14 -2 were used for drug studies. Such samples http://gut.bmj.com/ animals were scored by a histopathologist were initially placed in pre-oxygenated (95% blinded to the macroscopic score and chemi- 02, 5% CO2 for 10 minutes) D-PBS, with luminescence responses of the sections, using added calcium (1 13 mM) and glucose (5 mM) the scoring system shown in Table II. at ambient temperature and then transferred to 300 ,uM luminol or lucigenin immediately prior to assessment of the chemiluminescence CHEMILUMINESCENCE ASSAY response, which was measured in a Packard Tri on September 30, 2021 by guest. Protected copyright. Luminol and lucigenin react with oxidants, Carb 1600 CA liquid scintillation counter such as ROS, to form 3-aminophthalate and analyser operated in 'the out-of-coincidence' N-methylacridone, respectively. Electrons in mode for two minutes. the luminol and lucigenin are raised to higher energy levels during the reaction with oxidants. On reverting to the ground state, energy is ASSESSMENT OF EFFECTS OF TEST COMPOUNDS ON CHEMILUMINESCENCE RESPONSE All compounds and appropriate controls were TABLE II Histological assessment offull thickness biopsy prepared prior to the experiment and adjusted specimensfrom acetic acid colitis in rats to pH 7.4 by addition of 1M NaOH or HCI. Neutrophil infiltrate Controls for these experiments were vehicle None 0 Slight increase 1 Marked increase 2 (D-PBS) for all compounds except CuZn Epithelium (0-2) Lamina propria (0-2) superoxide dismutase, Rh-Mn superoxide Muscularis mucosa (0-2) dismutase, and catalase, in which heat Submucosa (0-2) Muscularis propria (0-2) inactivated enzyme was used (Table I). Heat Serosa (0-2) inactivated enzymes were prepared by heating Fibrin deposition Absent 0 Present 1 in D-PBS at 100C for two hours.30 Mucosa Submucosa Submucosal neutrophil margination Absent 0 Present 1 Water soluble compounds Submucosal oedema Nil 0 Patchy 1 Confluent 2 The effect of test compounds on the chemi- Epithelial necrosis luminescence response of rat acetic acid Nil 0 Localised 1 Extensive 2 Epithelial ulceration induced colitis biopsy specimens was com- Absent 0 Present 1 (0-1) pared with the response of a paired biopsy to Maximum score 20 the control for that compound. The chemi- luminescence response was measured in 1-8 ml 410 Millar, Rampton, Chander, Claxson, Blades, Coumbe, Panetta, Morris, Blake
of 300 ,uM luminol or lucigenin and then Effect ofwater soluble compounds immediately repeated after addition of 200 ,ul For each compound at each concentration a of the test compound, at 10 times the final pair of biopsy specimens from each of ¢5 rats concentration, or appropriate control. After was used to assess the percentage change in counting, samples were blotted, weighed, and chemiluminescence. The chemiluminescence Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from in some experiments, placed in 10% formol counts before and after addition of test saline for subsequent histological examination. compound are given as ti and t2, respectively, and for the specimen treated with control, cl and c2, respectively. To take account of small Water insoluble compounds changes in control values during the assay, the LY231617 and amflutizole were insoluble in change in chemiluminescence produced by the aqueous solution at the required con- test compound on one biopsy specimen was centrations. Suspensions were therefore pre- adjusted by the change in chemiluminescence pared in 10% acacia dissolved in D-PBS, produced by the control solution with its pair. oxygenated for 10 minutes in 95% 02, 5% CO2 Accordingly, the percentage change in chemi- and adjusted to pH 7-4 with 1M NaOH. luminescence response induced by the test Acacia is a demulcent gum used to suspend or compound (/A%t) is calculated as follows: emulsify water insoluble pharmacological A%t=((t2XCl/C2)-tl ) X 100. agents. Oxygenation of these suspensions was The results are expressed as median % required as the biopsy specimens were (IQR). The difference in chemiluminescence transferred to the acacia suspensions between response between biopsy specimens treated the chemiluminescence counts rather than with test compounds and controls is calculated remaining in the vial as with the water soluble by comparing the percentage change in chemi- agents. After the initial chemiluminescence luminescence induced by the test agent and count, as described above, five specimens were control using the Wilcoxon signed rank test for placed together in 10% acacia with or without paired variables. A p value of less than 0.05 the addition of the test compound in (two tailed) was taken as significant. suspension. The specimens were grouped to To investigate the variability ofthe change in avoid disparity in dosage exposure due to chemiluminescence response produced by heterogeneity in the suspension. Incubation vehicle alone, the coefficient ofvariation of the was carried out for 30 minutes in a shaking response to D-PBS (control for most of the water bath at 37°C. At the end of this time the compounds tested) was calculated for two to samples were removed, washed twice in six biopsies from each of 10 animals and the oxygenated D-PBS to remove excess acacia, median (IQR) expressed for the coefficient of and luminol amplified chemiluminescence variation of the luminol amplified chemilumi- recounted. As a comparator, 5-ASA was also nescence response to D-PBS in all i0. ` assessed in acacia. http://gut.bmj.com/ Effect ofwater insoluble compounds CALCULATIONS AND STATISTICS Results are expressed as the mean percentage change in chemiluminescence for each group Chemiluminescence response ofbiopsy specimens of biopsy specimens exposed to the test Luminol and lucigenin amplified chemilumi- compound after correction of the initial count nescence is expressed as the number of for the percentage change induced by acacia on September 30, 2021 by guest. Protected copyright. photons/min/mg wet weight of tissue after alone (control). No range is given because subtraction of the background count. The grouping the biopsy specimens together background count was taken as the average resulted in loss of animal identity. The chemi- photons/min for two vials of each 12 analysed luminescence counts before and after exposure containing 300 ,uM luminol or lucigenin in to the test compound were compared using the oxygenated D-PBS alone. Comparisons Mann-Whitney U test. A p value of less than between groups of untreated control animals, 0.05 (two tailed) was taken as significant. saline treated animals, and acetic acid induced To investigate the variability of the chemi- colitis were made using the Kruskal-Wallis and luminescence response induced by incubation Dunn's multiple comparison tests. ofbiopsy specimens in 10% acacia, the median To assess the variability of the chemilumi- coefficient of variation was calculated for eight nescence response of colonic biopsy specimens groups of five biopsies from five rats incubated from acetic acid induced colitis, the coefficient in 10% acacia alone that were used as controls of variation was calculated from the chemi- in experiments assessing the effect of water luminescence response in luminol produced by insoluble compounds. two to six biopsy specimens from each of 22 rats.31 Comparisons between animals are expressed as the median and interquartile Results range (IQR) of the coefficient of variation. A similar calculation for the response to lucigenin VALIDATION OF THE ACETIC ACID INDUCED was made in two to six specimens from each COLITIS MODEL of 10 animals. The macroscopic appearances, the histo- Chemiluminescence response in acetic acid induced logical score, and the chemiluminescence colitis compared with controls response were compared using Spearman's We compared the chemiluminescence response rank correlation coefficient. of 71 colonic biopsy specimens from 23 rats Antioxidant potential ofnew treatmentsfor inflammatory bowel disease 41
with acetic acid induced colitis to eight non- inflamed colonic biopsy specimens from four .c 5.05-0 . cE rats given intracolonic 0.9% saline and 22 1b0 biopsy specimens from 10 untreated control rats (Fig 3). The median (IQR) for each of a5 3.0 Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from these groups was 17 449 (6498-38 880) -. E i.. photons/min/mg, 74 (0-154), and 0 (0-66) *>m. .. respectively. For lucigenin amplified chemi-
luminescence, 26 colonic biopsy specimens ,. 1 1. 1. from 10 rats with acetic acid induced colitis 1. [.4_ ... ai ...... a . .. . .a . were compared with eight specimens from four 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 rats given 0.9% saline and to 10 specimens Macroscopic score from seven untreated controls (Fig 3). The Figure 4: Correlation ofluminol amplified median (IQR) for these groups was 819 chemiluminescence with the macroscopic score in 71 biopsy (518-1471) photons/min/mg, 347 (157-759), specimensfrom 23 rats pre-treated with 3% acetic acid induced colitis at 24 hours. Data shown as macroscopic and 143 (48-227), respectively. Biopsy speci- score against the log ofthe chemiluminescence, showing the mens from acetic acid induced colitis produced regression line, y=0-3x+3-4, r= +0.5, p=0 0001, significantly more luminol amplified chemilumi- Spearman 's rank correlation coefficient. nescence than specimens from saline treated and untreated controls and more lucigenin amplified chemiluminescence than untreated controls inflamed rat colons and the chemilumi- (p<0.001) (Fig 3). There was no difference in nescence response of tissue from the same the chemiluminescence response of colonic specimens. The macroscopic and histological tissue from the two control groups. score were also positively correlated (r=+0-7, p=0.0001; data not shown). Chemiluminescence related to the severity of inflammation Variability ofthe initial chemiluminescence To assess the relation between ROS pro- response duction as detected by chemiluminescence and The median (IQR) of the coefficient of inflammation in the acetic acid induced colitis variation of the initial chemiluminescence model we compared the luminol amplified response of two to six specimens from 22 rats, chemiluminescence response of full thickness with acetic acid induced colitis, to luminol was biopsy specimens with the macroscopic colonic 66 (39-83)% and of two to six specimens from appearances and the histological grade of 10 rats to lucigenin was 29 (17-49)%. These inflammation in rats with acetic acid induced variations are small in comparison to the colitis. Figures 4 and 5 show significant percentage increase in luminol and lucigenin http://gut.bmj.com/ positive correlations between the macroscopic amplified chemiluminescence response of and histological scores, respectively, of acetic acid induced colitis biopsy specimens compared with control (mean, +20 000% and +500%, respectively). * I.I* ..1 Variability ofthe change in chemiluminescence on September 30, 2021 by guest. Protected copyright. E 5.0 after addition ofDulbecco's PBS cE The coefficient of variation of the change in 4.0 i luminol amplified chemiluminescence result- ,a0 ing from incubation in D-PBS (control for cD most of the experiments) calculated from the
0 3 D 0 0 2.0 1- a W.. t. .- C._ : * CD 0 ,|, : :.- * :' / 1.047 I D4.' 0 .-. : _SK z : - 01 0 4.0.B- )/ * : -/ :: / 1 1 1 1 A.~~~~I- AAC Saline Control --AAC Sle Control CL .. Luminol- * - . s ~L . 1 I . Ia 1 .a . and counts Figure 3: Log transformed luminol lucigenin amplified chemiluminescence infull 0 2 4 6 8. 10 12 14 16 .18 thickness colonic biopsy specimensfrom rats pre-treated with 3% intracolonic acetic acid (AAC) or 0O9% saline (Saline), assessed at 24 hours and in untreated control animals Histologicl score (Control). For luminol 71 colonic biopsy specimensfrom 23 rats with acetic acid induced Figure 5: Correlation ofluminol amplified colitis were compared with eight colonic biopsy specimensfromfour rats given intracolonic chemiluminescence with the histological score (see Table II) 0 9% saline and 22 biopsy specimensfrom 10 untreated controls. For lucigenin amplified in 27 biopsy specimensfrom 14 rats pre-treated with 3% chemiluminescence, 26 specimensfrom 10 animals with acetic acid induced colitis were acetic acid induced colitis at 24 hours showing the regression compared with eight specimensfromfour animals given 0 9% saline and 10 specimensfrom line, y=O01x+3.5, r= +07, p=00004, Spearman's rank seven untreated controls. *p<0.001, Dunn's multiple comparison test. correlation coefficient.-S^,...... S.. .:|: l:|e @ @ 412 Millar, Rampton, Chander, Claxson, Blades, Coumbe, Panetta, Morris, Blake
10 animals in which there were two or more chemiluminescence response in acetic acid control experiments was 18 (11-41)%. induced colitis biopsy specimens (maximum -88 (-89 to -70)% at 20 mM 5-ASA, p=003) with a limited dose response producing an Variability ofthe change in chemiluminescence estimated IC50 of 4 mM (Fig 7). Hydro- Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from after incubation ofacetic acid induced colitis cortisone did not alter luminol amplified biopsy specimens in 10% acacia chemiluminescence in either acetic acid The coefficient of variation for the change in induced colitis or UC biopsy specimens24 chemiluminescence of eight groups of five (Fig 7). biopsy specimens each, incubated in 10% acacia alone (controls), was 48%, which reflects the RESPONSE OF BIOPSY SPECIMENS FROM ACETIC variability of this system compared with incuba- ACID INDUCED COLITIS TO POTENTIAL NEW tion in D-PBS. TREATMENTS FOR IBD Rh-Mn superoxide dismutase had no dem- onstrable antioxidant activity in this system, COMPARISON OF RESPONSE OF BIOPSY SPECIMENS altering the lucigenin amplified and luminol FROM ACETIC ACID INDUCED COLITIS AND UC amplified chemiluminescence response, re- TO CONVENTIONAL ANTIOXIDANTS spectively by -9 (-17 to +3)% and +17 (+10 Significant reductions in luminol amplified to +37)%, respectively. However, two novel chemiluminescence in the acetic acid induced water insoluble antioxidant compounds, colitis model were observed with sodium azide LY231617 (10 mM) and amflutizole (20 mM), (-83 (-85 to -73)% p=0.03), catalase (-43 produced noticeable reductions in luminol (-47 to -41)%, p=003), 10% (1.28 M) amplified chemiluminescence after incubation DMSO (-67 (-78 to -50%, p=0.03), and of the specimens in suspensions of the ascorbate (-54 (-69 to -35)%, p=0.01) and in compounds in 10% acacia (means, -98%, lucigenin amplified chemiluminescence with p=0.009 and -88%, p=0.028, respectively) CuZn superoxide dismutase (-26 (-32 to (Fig 8). 5-ASA (20 mM), which had previously -15)%, p=0.03) (Fig 6). N-acetylcysteine pro- been shown to reduce luminol amplified duced no significant change in luminol ampli- chemiluminescence in aqueous solution, re- fied chemiluminescence. The profile of duced the chemiluminescence by a similar responses in acetic acid induced colitis dem- amount (mean, 83% p=0.002) when tested in onstrate a striking resemblance to that pre- acacia. viously seen in UC.1 24 25 With one compound, taurine, a small, but significant, reduction was produced in UC biopsy specimens' but not in Discussion acetic acid induced colitis. These findings show that full thickness colonic biopsy specimens from rats with acetic acid http://gut.bmj.com/ induced colitis produce increased levels of COMPARISON OF RESPONSE OF BIOPSY ROS, as detected by amplified chemilumi- SPECIMENS FROM ACETIC ACID INDUCED nescence, compared with control tissue. In COLITIS AND UC TO STANDARD TREATMENTS addition, we have found that the level of ROS As for UC specimens, 5-ASA produced a production detected by luminol amplified significant reduction in luminol amplified chemiluminescence correlates with the grade on September 30, 2021 by guest. Protected copyright.
1 Acetic acid colitis W Ulcerative colitis t 0) 125 * a) C) 100 cn a) c 75 E 50 * .)_ 25 * a, 0 Ecci 0 ._ -25 -50 a) 0) 0 -75 * 0~ a, -100 0 -125 l 1 1 3000 3000 20 20 30 300 300 1 5 5 10 20 20 20 20 Azide Catalase Taurine CuZn DMSO NAC Ascorbate (mM) (U/ml) (mM) superoxide (%) (mM) (mM) dismutase (U/ml)
Figure 6: The percentage change in chemiluminescence response ofinflamed biopsy specimensfrom acetic acid induced colitis to conventional antioxidants compared with results using mucosal biopsy specimensfrom active UC (tdata redrawnfrom references' 24 25). Chemiluminescence was measured before and after exposure to test compound and compared with a second biopsy specimen exposed to vehicle (n S5; *denotes p 60 activity in the present short-term in vitro experiments in this model or with UC biopsy M Acetic acid colitis W Ulcerative colitis t 40 specimens.24 However, other studies, using C) prolonged incubation with dexamethasone in Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from a) vitro, have shown inhibition of superoxide gen- C; 20 U) eration by stimulated phagocytes33: the thera- peutic effect ofhydrocortisone in vivo may thus 0 be partly due to an indirect antioxidant action by inhibition of neutrophil function.'6 -20 Of the conventional antioxidants examined, .) the greatest inhibition occurred with azide, a c potent inhibitor of myeloperoxidase,28 which 0' -40 also inhibits other haem proteins such as 0) * catalase and quenches singlet oxygen and the am -60 hydroxyl ion.34 Although the antioxidant action of azide could result from a cytotoxic effect, 0~ -80 T azide has been shown to reduce the luminol 1 amplified chemiluminescence of stimulated neutrophils to a greater effect when added -100 before the stimulus (100%) than after (20%),35 an effect that would not occur if the reduction -120 in chemiluminescence was purely due to cyto- 1 10 20 20 1 1 toxicity. 5-Aminosalicylate Hydrocortisone (mM) (mM) Superoxide is produced by the activated neutrophil, partly as the result of NADPH Figure 7: The percentage change in the luminol amplified chemiluminescence response of inflamed biopsy specimensfrom acetic acid induced colitis to standard treatments in IBD activity,36 and as a result of episodes of compared with results using mucosal biopsiesfrom active UC (tdata redrawnfrom ischaemia reperfusion, which may be im- references1 24 25). Chemiluminescence was measured before and after exposure to test portant in the pathogenesis of IBD37 (Fig 1). compound and compared with a second biopsy specimen exposed to vehicle (n S; *denotes Superoxide generation has been located to the p colitis,'9 suggesting that hydrogen peroxide in both UC and CD.`3 The competitive and its metabolites may contribute to colonic xanthine oxidase inhibitor, allopurinol, has inflammation. An antioxidant effect for been of reported benefit in pouchitis.'8 In the catalase can be demonstrated in both acetic present studies amflutizole was at least as acid induced colitis and UC biopsy speci- effective as 5-ASA in reducing the chemilumi- Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from mens,' further supporting the suggestion that nescence response of biopsy specimens from ROS produced by neutrophils are important in acetic acid induced colitis. Xanthine oxidase IBD. inhibition with amflutizole should, therefore, N-acetylcysteine replenishes intracellular be further explored as a therapeutic option in stores of glutathione, is a weak scavenger of IBD; again initial studies using topical treat- superoxide and hydroperoxide but reacts ment in distal UC would be desirable. avidly with hypochlorous acid and the hydroxyl In conclusion, this in vitro system using radical.45 However, N-acetylcysteine was in- biopsy specimens taken from rats with acetic effective in reducing ROS production by acetic acid colitis provides a convenient method for acid induced colitis or UC biopsy specimens the screening of the antioxidant potential of under the present experimental conditions.24 new treatments for IBD. Agents that reduce The physiological level of ROS is regulated ROS production by inflamed human colorectal by both enzymatic systems and direct oxygen biopsy specimens, such as CuZn superoxide scavengers, such as ascorbate, which reacts dismutase, catalase, and DMSO also amelio- rapidly with superoxide, the hydroxyl radical,46 rate inflammation in acetic acid induced and hypochlorous acid.47 Ascorbate levels are colitis. Previous workers have suggested that decreased in the plasma in IBD.48 ` In this the ROS production in both UC' and acetic investigation ascorbate produced similar re- acid induced colitis'9 22 is largely mediated by ductions in chemiluminescence in acetic acid neutrophils, and the results here with azide, induced colitis and human UC biopsy speci- catalase, and CuZn superoxide dismutase mens.24 However, ascorbate is readily oxidised support this proposal. The additional evidence to the potentially damaging ascorbate radical in of an antioxidant effect with the xanthine oxi- the presence of iron,50 present in the bowel dase inhibitor, amflutizole, and previous lumen in millimolar quantities.3 This reaction evidence that allopurinol reduces ROS could limit the potential usefulness of as- production by inflamed colonic biopsy speci- corbate as a treatment in UC. mens from acetic acid induced colitis22 The predominant antioxidant action of suggests that, at least in this animal model, DMSO in vitro is scavenging of the hydroxyl ischaemia reperfusion may also contribute to ion.5 As the hydroxyl ion is extremely reactive, mucosal oxidant stress. the local concentration of any scavenger would The limited antioxidant effect of antioxidant probably have to be prohibitively high to enzymes in both acetic acid induced colitis and compete with biological molecules in the UC biopsy specimens' may be because of http://gut.bmj.com/ vicinity of its production. This may explain poor tissue penetration or inactivation by evidence from in vivo studies in which oral 5% hypochlorous acid54 produced by activated (0-64 M) DMSO failed to improve acetic acid neutrophils; these factors may restrict their induced colitis.22 Nevertheless, the antioxidant therapeutic potential. effects of DMSO obtained here, and in UC Of the agents assessed in this study, di- biopsy specimens,l suggest that evaluation of methyl sulphoxide, ascorbate, amflutizole, and its potential efficacy when given in enema LY23 1617 seem most suitable for further on September 30, 2021 by guest. Protected copyright. formulation in high concentration to patients evaluation in IBD. with distal UC may be worthwhile. This work was supported by grants from Eli-Lilly and Co, USA The hypochlorite scavenger taurine had little and Bender and Co, Ges mbH, Vienna, Austria. effect on ROS production by UC biopsy specimens' and no effect in this model, 1 Simmonds NJ, Allen RE, Stevens TRJ, van Someren RNM, possibly because the reaction product of hypo- Blake DR, Rampton DS. Chemiluminescence assay of mucosal reactive metabolities in inflammatory bowel chlorous acid and taurine, taurine-N- disease. Gastroenterology 1992; 103: 186-96. chloramine, is itself an oxidant.29 2 Sedghi S, Fields JZ, Klamut M, et al. Increased production of luminol enhanced chemiluminescence by the inflamed The novel agent, LY23 1617, protects colonic mucosa in patients with ulcerative colitis. Gut against cerebral ischaemia in an animal model 1993; 34: 1191-7. 3 Babbs CF. Oxygen radicals in ulcerative colitis. Free Rad in vivo and ameliorates hydrogen peroxide Biol Med 1992; 13: 169-82. induced neuronal damage in vitro.5' The 4 Grisham MB. Oxidants and free radicals in inflammatory antioxidant shown here suggests bowel disease. Lancet 1994; 344: 859-61. potent activity 5 Repine JE, Eaton JW, Anders MW, Hoidal JR, Fox RB. that it should be further evaluated as treatment Generation ofhydroxyl radical by enzymes, chemicals and in UC, perhaps initially in an enema or sup- human phagocytes in vitro. Detection with the anti- inflammatory agent dimethyl sulphoxide. J Clin Invest pository formulation to achieve the necessary 1979; 64: 1642-51. 6 Grisham MB, Granger DN. Neutrophil-mediated mucosal intraluminal concentration (10 mM). injury. Role of reactive oxygen metabolites. Dig Dis Sci We studied another antioxidant, amflutizole, 1988; 33 (suppl): 6-15S. 7 Hawkey CJ, Rampton DS. Prostaglandins and the which is a non-competitive inhibitor of gastrointestinal mucosa: are they important in its xanthine oxidase and has also shown efficacy function, disease, or treatmlent? Gastroenterology 1985; 89: 1162-82. in a model of neuronal ischaemia.52 Ischaemia 8 Grisham MB, MacDermott RP, Deitch EA. Oxidant reperfusion with increased superoxide gener- defense mechanisms in the human colon. Inflammation 1990; 14: 669-80. ation through xanthine oxidase activation36 9 Aruoma 01, Kaur H, Halliwell B. Oxygen free radicals and may have a pathogenic role in IBD. Multifocal human diseases. JfRoy Soc Health 1991; 111: 172-7. 10 Chakraborti 5, Gurtner GH, Michael JR. Oxidant-mediated vascular microinfarction is a feature of CD37 activation of phospholipase A2 in pulmonary endo- and microvascular thrombi have been detected thelium. AmJPhysiol 1989; 257: L430-7. Antioxidant potential ofnew treatmentsfor inflammatory bowel disease 415 11 Reindeau D, Denis D, Choo LY, Nathaniel DJ. Stimulation 38 Oshitani N, Kitano A, Okabe H, Nakamura S, Matsumoto of 5-lipoxygenase activity under conditions which T, Kobayashi K. Location ofsuperoxide anion generation promote lipid peroxidation. Biochem J 1989; 263: in human colonic mucosa obtained by biopsy. Gut 1993; 565-72. 34: 936-8. 12 Zimmerman BJ, Grisham MB, Granger DN. Role of 39 Mulder TP, Verspaget HW, Janssens AR, deBruin PA, Pena oxidants in ischaemialreperfusion-induced granulocyte AS, Lamers CB. Decrease in two intestinal copper/zinc Gut: first published as 10.1136/gut.39.3.407 on 1 September 1996. Downloaded from infiltration. Am JfPhysiol 1990; 258: G185-G90. containing proteins with antioxidant function in 13 Schreck R, Rieber P, Baeuerle PA. Reactive oxygen inflammatory bowel disease. Gut 1991; 32: 1146-50. intermediates as apparently widely used messengers in the 40 Niwa Y, Somiay K, Michelson AM, Puget K. Effect of activation of NFkappaB transcription factor and HIV-1. liposomal-encapsulated superoxide dismutase on active EMBOJ 1991; 10: 2247-58. oxygen-related disorders. Free Radic Res Commun 1985; 14 Amstad PA, Krupitza G, Cerutti PA. Mechanism of c-fos 1: 137-53. induction by active oxygen. Cancer Res 1992; 52: 41 Gorecki M, Beck Y, Hartman JR, et al. Recombinant human 3952-60. superoxide dismutates: production and potential 15 Ahnfelt-Ronne I, Nielsen OH, Christensen A, Langholz E, therapeutical uses. Free Radic Res Commun 1991; 12-13 Binder V, Riis P. Clinical evidence supporting the radical Pt 1: 401-10. scavenger mechanism of 5-aminosalicylic acid. 42 Parizada B, Weber MM, Nimrod A. Protective effects of Gastroenterology 1990; 98: 1162-9. human recombinant MnSOD in adjuvant arthritis and 16 Baltch AL, Hammer MC, Smith RP, et al. Comparison of bleomycin-induced lung fibrosis. Free Radic Res Commun the effect of three adrenal corticosteroids on human 1991; 15: 297-301. granulocyte function against Pseudomonas aeruginosa. J7 43 Susuki M, Grisham MB, Granger DN. Leukocyte- Trauma 1986; 26: 523-33. endothelial adhesive interactions: role of xanthine 17 Emerit J, Pelletier S, Likforman J, Pasquier C, Thuillier A. oxidase-derived oxidants. J Leukoc Biol 1991; 50: Phase II trial ofsuperoxide dismutase (CuZn SOD) in the 488-94. treatment of Crohn's disease. Free Radic Res Commun 44 Bilota JJ, Wayne JD. Hydrogen peroxide enteritis: the "snow 1991; 12-13 (Pt 2): 563-9. white" sign. Gastrointest Endosc 1989; 35: 428-30. 18 Levin KE, Pemberton JH, Phillips SF, Zinsmeister AR, 45 Aruoma OI, Halliwell B, Hoey BM, Butler J. The Pezim ME. Role of oxygen free radicals in the etiology of antioxidant action of N-acetylcysteine: its reaction with pouchitis. Dis Colon Rectum 1992; 35: 452-6. hydrogen peroxide, hydroxyl radical, superoxide and 19 Keshavarzian A, Sedghi S, Kanofsky J, et al. Excessive hypochlorous acid. Free Rad Biol Med 1989; 6: 593-7. production of reactive oxygen metabolites by inflamed 46 Nishikimi M. Oxidation of ascorbic acid with superoxide colon: analysis by chemiluminescence probe. Gastro- anion generated by the xanthine-xanthine oxidase system. enterology 1992; 103: 177-85. Biochem Biophys Res Commun 1975; 63: 463-8. 20 Yoshikawa T, Takahashi S, Kondo M. Possible role of free 47 Halliwell B, Wasil M, Grootveld M. Biologically significant radicals in chronic inflammation in the gut. EXS 1992; scavenging of the myeloperoxidase-derived oxidant 62: 353-8. hypochlorous acid by ascorbic acid. FEBS Letts 1987; 21 Sharon P, Stenson WF. Metabolism of arachonic acid in 213: 15-8. acetic acid. Similarity to human inflammatory bowel 48 Pettit SH, Shaffer JL, Johns CW, Bennett RJM, Irving MH. disease. Gastroenterology 1985; 88: 55-63. Ascorbic acid absorption in Crohn's disease. Studies 22 Keshavarzian A, Morgan G, Sedghi S, Gordon JH, using L-[carbonyl-14C] ascorbic acid. Dig Dis Sci 1989; Doria M. Role of reactive oxygen metabolites in 34: 559-60. experimental colitis. Gut 1990; 31: 786-90. 49 Femandez-Banares F, Mingorance MD, Esteve M. 23 Keshavarzian A, Haydek J, Zabihi R, Doria M, D'Astice M, Essential antioxidants in active idiopathic colitis. In Sorensen JRJ. Agents capable of eliminating reactive Goebell H, Ewe K, Malchov H, Koebel C, eds. oxygen species: catalase, WR-2721, Cu(II)2(3,5-DIPS)4 Inflammatory bowel disease. Progress in basic research and decrease experimental colitis. Dig Dis SCi 1992; 37: clinical implications. Lancaster: Kluwer Academic, 1991: 1866-73. 404. 24 Simmonds NJ. Reactive oxygen metabolites and inflammatory 50 Rowley DA, Halliwell B. Formation of hydroxyl radicals bowel disease [Doctor of Medicine, Thesis]. Oxford: from hydrogen peroxide and iron salts by superoxide- and 1992. ascorbate dependent mechanisms: relevance to the 25 Simmonds NJ, Blake DR, Rampton DS. The effects of pathology of rheumatoid disease. Clin Sci 1983; 64: aminosalicylates and potential new drugs on mucosal 649-53. reactive oxygen metabolite production in ulcerative 51 Clemens JA, Saunders RD, Ho PP, Phebus LA, Panetta JA. colitis. Gut 1992; 33 (suppl): S7. The antioxidant LY231617 reduces global ischaemic http://gut.bmj.com/ 26 MacPherson BR, Pfeiffer CJ. Experimental production of neuronal injury in rats. Stroke 1993; 24: 722-3. diffuse colitis in rats. Digestion 1978; 17: 135-50. 52 Phillis JW, Sen S, Cao X. Amflutizole, a xanthine oxidase 27 Morris GP, Beck PL, Herridge MS, Depew WT, Szewczuk inhibitor, inhibits free radical generation in the ischemicd MR, Wallace JL. Hapten-induced model of chronic reperfused rat cerebral cortex. Neurosci Letts 1994; 169: inflammation and ulceration in the rat colon. 188-90. Gastroenterology 1989; 96: 795-803. 53 Dhillon AP, Anthony A, Sim R, et al. Mucosal capillary 28 Allen RC. Phagocyte oxygenation activities: quantitative thrombi in rectal biopsies. Histopathology 1992; 21: analysis based on chemiluminescence. In: Schoimerich J, 172-33. Andreeson R, Knapp A, Ernst M, Woods WG, eds. 54 Sharanov BP, Govorova NJ, Lyzlova SN. Serum protein Bioluminescence and chemiluminescence. New York: Wiley, degradation by hypochlorite. Biochem Int 1989; 19: 1986: 13-22. 27-35. on September 30, 2021 by guest. Protected copyright. 29 Halliwell B, Gutteridge JMC. Free radicals in biology and 55 Klebanoff S. Oxygen metabolites from phagocytes. In: Gallin medicine. (2nd ed.) Oxford: Clarendon, 1989. J, Goldstein I, Snyderman R, eds. New York: Raven Press, 30 Heikkila R. Inactivation of superoxide dismutase by 1992: 541-88. diethylcarbamate. In: Greenwald R, ed. CRC handbook of 56 Dallegri F, Ottonello L, Ballestrero F, Ferrando F, Patrone methods for oxygen radical research. Boca Raton, Florida: F. Cytoprotection against neutrophil derived CRC Press, 1985: 387-90. hypochlorous acid: a potential mechanism for the 31 Snedecor GW. Statistical methods applied to agriculture and therapeutic action of 5-aminosalicylate in ulcerative biology. 5th ed. Iowa, USA: The Iowa State University colitis. Gut 1990; 31: 184-6. Press, 1956: 62. 57 Green TR, Fellman JH, Eicher AL, Pratt KL. Antioxidant 32 Staerk-Laursen L, Stokhoim M, Bukhave K, Rask-Madsen role and subcellular location of hypotaurine and taurine J, Lauritsen K. Disposition of 5-aminosalicylic acid by in human neutrophils. Biochim Biophys Acta 1991; 1073: olsalazine and three mesalazine preparations in patients 91-7. with ulcerative colitis: comparison ofintraluminal colonic 58 Beilke MA, Collins-Lech G, Sohnle PG. Effects of dimethyl concentrations, serumn values, and urinary excretion. Gut sulphoxide on the oxidative function of human 1990; 31: 1271-6. neutrophils. J Lab Clin Med 1987; 110: 91-6. 33 Maridonneau-Parini I, Errasfa M, Russo-Marie F. 59 Miyachi Y, Yashioka A, Imamura S, Niwa Y. Effect of Inhibition of 02-generation by dexamethasone is sulphasalazine and its metabolites on the generation of mimicked by lipocortin I in alveolar macrophages. Jf Clin reactive oxygen species. Gut 1987, 28: 190-5. Invest 1989; 83: 1936-40. 60 Craven PA, Pfanstiel J, Saito R, DeRubertis FR. Actions of 34 Nurcombe HL, Edwards SW. Role of myeloperoxidase in sulphasalazine and 5-aminosalicylic acid as reactive intracellular and extracellular chemiluminescence of oxygen scavengers in the suppression ofbile acid induced neutrophils. Ann Rheum Dis 1989; 48: 56-62. increases in colonic epithelial loss and proliferative 35 Hallett MB, Campbell AK. Two distinct mechanisms activity. Gastroenterology 1987; 92: 1998-2008. for stimulation of oxygen-radical production by 61 Aruoma OI, Wasil M, Halliwell B, Hoey BM, Butler J. The polymorphonuclear phagocytes. Biochem J 1983; 216: scavenging of oxidants by sulphasalazine and its 459-65. metabolites. A possible contribution to their anti- 36 McCord JM. Oxygen-derived free radicals in postischaemic inflammatory effects? Biochem Pharmacol 1987; 36: tissue injury. NEnglJMed 1985; 312: 159-63. 3739-42. 37 Wakefield AJ, Sawyerr AM, Dhillon AP, et al. Pathogenesis 62 Grisham MB. Effect of 5-aminosalicylic acid in ferrous of Crobn's disease: multifocal gastrointestinal infarction. sulphate-mediated damage to deoxyribose. Biochem Lancet 1989; li: 1057-62. Pharmacol 1990; 39: 2060-3.