Inflammatory mediators in inflammatory bowel disease : Clues for designer therapy

WALLACE. Inflammatory mediators in inflammatory bowel disease: Clues designer therapy. Can J Gastroenterol 1993;7(2):121-126. The etiology Ii inflammatory bowel disease (IBO) is poorly understood. Regardless of the 'tiating events, it is likely that soluble mediators of inflammation play a critical de in the pathogenesis of the mucosal injury characteristic of this group of · s. These mediators arc responsible for recruiting and activating granuloes. Thus, it is possible that these mediators may be rational targets for the velopment of novel anti-inflammatory drugs for the treatment of intestinal ammation. Recent advances in research into the roles of leukotrienes, thromne, prostaglandins, platelet-activating factor, interleukinI and interleukinin the pathogenesis of IBO are reviewed.

ge nesis of inflammatory bowel diseases (lBD) are more clearly understood, the therapy for chis group of d iseases will remain somewhat empirical.The most frequently cited theory fo r the etiology of lBD is chat an infectious agent or a dietary antigen gains access to the lam ina propria, perhaps through an abnormally permeahle epithelium, and initiates an inflammatory response (Figure 1).le is also widely held that through a gencLic disposition, the immune response to such a challenge is inadequate or uncontrolled.Regardless of the initiating facto r, 1t is clear that much of the tissue necrosis which characterizes IBO is produced by the hose inflammatory response itself.For example, recent studies utilizing an experimental model of colitis demonstrated that up ro 80% of the colonic epithelial injury associated with the coli tis could be inhibited by pretreating the animals with a monoclonal antibody which prevented granulocyce migration out of blood vessels ( 1 ).Various inflammatory mediators contribute significantly to chis injury by promoting the rccruitmen L of granulocyces co the site of injury, as well as by affecting the function of immune cells (T and B lymphocytes).In recent years, a great deal of

LEUKOTRIENES
Of the multitude of inflammatory mediators, perhaps the one for which there is the most compelling evidence for a role in the pathogenesis of lBO is leukotriene (LT) 134.L TB4 is a potent chemotaxin for neutrophils and can activate neutrophils to release oxygenderived free radicals and proteases (2, 3 ).LTB4 also exerts immunomodulatory effects, stimulating the proliferation of B lymphocytes ( 4) and the production of tumour necrosis factor, interferon gamma and interleukin-2 by human mononuclear cells (J ).In several animal models of IBO, LTB4 synthesis has been shown to be greatly elevated soon after initiation of the inflammatory response (5-7), and in one model L TB4 synthesis was shown to be elevated for as long as five weeks after induction of inflammation (5).LTB4 has also been shown to exacerbate 122 colonic injury induced by intracolonic ethanol (8).
In humans, L TB4 synthesis is increased in ulcerative colitis and Crohn's disease patients by about SOfold above normal levels (9).These tissue concentrations are within the range necessary to induce neutrophil chemotaxis; such a role is supported by the finding that 78 to 90% of the chemotactic activity present in samples of inflamed bowel was lipid in nature, co-eluted with LTB4 on high performance liquid chromatography, and could be removed by preincubation of the extract with an antibody directed against L TB4 (10).It has been suggested that one of the mechanisms through which 5-ASA or prednisone reduce colonic inflammation is by inhibiting LT synthesis, since these compounds were shown to reduce significantly rectal dialysate L TB4 levels and clinical symptoms in a parallel manner (11 ).Interestingly, there was a subgroup of patients in this study who were unresponsive to therapy with 5-ASA and prednisone.These patients were found to have significantly higher pretherapy rectal dialysate L TB4 levels than the patients who did respond to therapy, suggesting that high levels of L TB4 in rectal dialysates may be predictive of how a patient will respond to anti-inflammatory therapy.
In recent years a number of specific inhibitors of LT synthesis have become available (Figure 2) and have been shown to accelerate healing in animal models of lBD (5,12,13).In each case the compound had to be administered early in the course of colitis for a beneficial effect to be observed.The ability of compounds to accelerate healing correlated well with their ability to inhibit colonic LT synthesis ( 14).LT synthesis inhibitors also significantly reduced the incidence of adhesions and diarrhea and improved the weight gain after induction of colitis (5, 13 ).
The results of clinical trials with a specific inhibitor of LT synthesis have recently been reported.Laursen et al (l 5) reported that oral administration of the 5-lipoxygenase inhibitor, zileuton, inhibited rectal L TB4 synthesis by approximately 67% for a period of 4 to 8 h.The compound was then assessed in a randomized, double.blind, placebo-controlled trial over a period of 28 days (16).Patients treated with zi leuron showed significant improvement compared with placebotreated controls, but not with panents created with sulphasalazine.Unfortu• nately, the twice-a-day dosing regimen employed in this trial was not sufficient to achieve 24 h inhibition of colonic LT synthesis, so it is still unclear if effective inhibition of LT synthesis will represent an alternative approach to the treatment of active IBO.

THROMBOXANE
Thromboxane is derived from arachidonic acid via the acLions of cyclooxygenase and chromboxane synthetase (Figure 2).The principal eel lular source of this mediator is the platelet, although it is also made b, macrophages and neutrophils.It is a potent vasoconstrictor and activatorol platelets, and can induce lesions in the upper gastrointestinal tract when ad• ministered systemically (l 7).lnteres1 in the role of thromboxane in IBD wru first stimulated by reports that throm• boxane production by cultured rectal biopsies from Crohn's disease patients ( 18) and th romboxane levels in rectal dialysates &om ulcerative colitis and Crohn's disease patients were markedl1 elevated above controls ( 19).Using a rat model of colitis, Vi laseca et al (201 demonstrated that colonic production of many eicosanoids was elevated; how, ever, while the production of most of the eicosanoids gradually returned towards control levels, thromboxane synthesis remained elevated through, out the 21-day study period.When the rats were treated with 5-ASA or pred• nisone at doses that significantl1 reduced the severity of injury, they observed a significant decrease in colonic thromboxane synthesis.They then as, sessed the effects of two thromboxane synthetase inhibitors in their model.Both compounds significantly in, hibited the production of throm• boxane, and significantly reduced the severity of the colitis.The authors sug, gested chat the balance between pros, taglandin and thromboxane synthesu

PROSTAGLANDINS
Prostagland111s have traditionally l,ecn classified as inflammatory med i-1tors based on their abili ty to potenuate edema formation in the presence if other med iators, such as bradykinin and histamine, and because of their hyperalgesic effects in various inflammatory diseases.In fact, there is considerable evidence to suggest that, for the most part, prostaglanJ111s exert anti-inflammato ry effects, and in the case of IBD, prostaglandins may play a critical role in 'dampening' the inflammatory rc~ponse.As outlined in Table 1, prostaglandins exert anti-inflammatory ef-lL'Ctli at a number of cellular targets.For example, prostaglandins of th e E and I ;.:nes inhibit neutrophil activation and )Ccretion as well as 111h1biting their adherence and migration (21)(22)(23).Pros-1.1glandinshave a lso been shown to inhibit the production of other 111flam-matory mediators, inclu<ling LTB.t synthesis hy neutrophils (24) and rumour necrosis factor synthesis by macrophages (25 ).
The evidence that prostaglandins exert anti-inflammatory effects m IBO comes primarily from an imal studies.A number of groups have demonstrated that exogenous prostaglandins are capable of reducing the severity of in-Jury and the extent of granulocyte infiltration in experimental models (5, 26-30).Pretreatment with prosta-gland111 Ez was shown co dose-dependently reduce the seventy of colon ic injury in the rat induced by intracolonic 1nst1 llation of 30% ethanol (26).Allgayer and Stenson demonstrated that administration of a prosta-gland111 E2 analogue after induction of colitis in the rat with trinitrobenzene sulphonic acid rcsulte<l 111 a significant decrea e in the migration of granulocytes into the affected region over the 24 h period that fo llowed (27).Using the acetic ac1<l model of acute coitus in the rat, Yamada et al (29) recently reported that misoprostol, a prosraglandin E1 analogue, significantly reduced the infiltranon of neutroph ils into the colon and the changes in epithe lial permeabili ty.They suggested that the prosraglandin protected the prolifera-nve zone of the colonic ep1thel1um, allowing for more rapid establishment of an intact epitheli um after the initia l injury (31 ).The second type of ev1-<lence for an anti-inflammatory role of endogenous prosraglamlins b the evidence that administration of nonstero1dal anti-inflammatory <lrugs to rats with colins results in profound exacerbation of the 111jury, and s1gn1ficant mortality (a consequence of intest111al perforation) (12).This exacerbanon of colitis was observed with doses of nonsteroidal anu-inflammatory drugs (NSAIOs) that did not produce gc1stro-111test1nal 111jury in healthy rats and was not attnhutable to increased LT synthestS.
There is also some clinical evidence consistent with these animal studies, including numerous anecdotal reports of exacerbation of IBO in patients ingesting NSAIDs (32).There has been only one clinical trial of a prostagland in analogue in IBO.Gold111 and Rachmi lewitz (33) reported the failure of a prostaglandin Ez analogue to improve the course of ulcerative colitis.ln fact, they had to terminate the study prematurely because of a high incidence of d iarrhea in the patients treated with the prostaglandin.Perhaps the potennal utility of prostaglandms 111 the treatment of lBD deserves further

INTERLEUKIN-I AND INTERLEUKIN-8
In recent years, there has heen a grl•ar Je.il o( interest 111 the poss1hle wlc of a number of cytokmcs 111 the patho-grncst~ of !BD.Perhaps the greatest 111reresr has focused on interleukin (IL)-1 anJ more recently on lL-8.IL-1 is a ple1otrop1c molecule produced by numerous types of cells, mcluding macrophages and ma.st cells (34).Among ns many actions, IL-l upregulates rhe expressllll1 of aJhes1on molecules on the vascu lar enJothclium (35) and st imulates endnthcl1al synthesis of lL-8.IL-8 is one of rhc most potent ncutrnphil chemoraxms yet described (36).IL-8 increases the expression of aJhcsum molecules, such as the, (32-111regnn family (CD11/CD18) on neutrophils.Thus, IL-I and lL-8 act m concert to tncrcase the recruitment li neutrophils to a sire of mflammation.lL-1 synthesis b not detectable m samples ot normal colonic tissue, bur can be de tee ted m samples taken shl)rtly after rhe inducrton of colitis m nus (37) or rabhus (38).These change\ in IL-l synthesis occurred he fore detectable changes m e1cosan01d synthesis ( 38), suggesting that IL-I might be the rrigger for the production of these other inflammatory mcdtawrs.T reatment of rats or rabbits with an IL-I receptor antagonist significantly rcJuccd the seventy of the neuosis ,mJ inflammation wlrn.:harc charnctcnst1c of their model nf colitts {38-40).
T he role of IL-I anJ IL-8 111 human !Bl) has been 111vest1garcd by a number of groups m recent years.Ligumsky ct al ( 41) reported that IL-l synthesis was markeJly elevated in colonic samples from patients with IBO, particularly when the ttssue samples were taken from patients w1ch active Jisease.Mahida ct al ( 4 2) a lso demonstrated increases in I L-1 synthesis m human IBO, and further dcmonstrateJ that IL-I synthesis could be mhibitcJ by incubating the tissue samples with 5-ASA.The source of I L-1 production in the colon has also beenexammed.More recently, Mahida et al ( 43) demonstrated that IL-8 synthesis was elevated in colonic samples from patients with ulcerative colitis, but was not e levated in samples from pattents with Crohn's disease, even tho ugh both groups of patients had comparable disease acttvity.
PAF synthesis by the rat colon is markedly elevated after induction of colitis (47,48).Treatment with one of three PAF antagonists resulted in a significant reJucnon in the seventy of colonic damage and infla mmation in one of these rat mode ls ( 49).In another rat model, treatment with either of two srrucwrally unrelated PAF antagonists produceJ a 43 to 53% reduction in neutrophil infiltration mco the colon in a rat model of colt tis ( 49).
Eliak1m et a l (50) extended these studies to human IBO.Colonic biopsies from patients with ulcerative coltci, produced significantly more PAF than control biopsies when incubated m vitro and sttmulated with calcium ionophore.PAF synthesis could also l't stimulated by anc1-lgE, suggesting that mast cells or ocher lgE-bmJmg cell, were the primary sources of PAF.P.'\f levels m the stool of patients wtth Crohn's Jiscase were recently reported to be markedly elevated above contro, levels (51).Although a number of specific PAF receptor anrago111sts han been available for over five years, cltrn, cal trials of these compounJs m me have not been reportcJ.

IMPLICATIONS FOR FUTURE
THERAPY Unul recently, the study of the rob of vanous eicosanmds, cytokmes anu PAF in the pathogenesis of IBD ha! been limited to animal mo<lek Hl111ever, specific and safe pharmacologkal probes arc now becoming availabl, which will permit similar studies to~ performed in man .The results of trial• w1rh suc h compounds arc eagerh awaireJ.SruJ1es with these probes w1! provide information concern mg thl relative impo rtance of each of these classes of mediators anJ, hopefully, w1! represent a n ew avenue fo r che therari of inflammatory diseases of the mtc1tine.H owever, it is a lways possible chat these agents may prove ro be tl) spec ific to be effective in a mulnfac, tonal disease such as IBO.lndeeJ, drul."such as corticosteroids and 5-ASA ma, be effecttve 111 the treatment of !BC because of their lack of specifici ty.B1 inh1b1ting the synthesis of a number o. inflammatory meJiacors, as well a~ ex, errmg other activ1t1es (cg, free raJ1c; scavenging), these compounds ma, have more profound effects in 180 char a drug wh ich is targeteJ very specifical, ly a t a single mflammacory mediator.a course, questams of th ii.rype can onlt be a nswe red when appropriate cliniu trials have been performed.ACKNOWLEDGEMENTS: Dr Wallaa is an MRC of Cannda Sc1cnt1~t and an Ak berta Heritage Foundation for MeJ1cal Research Scholar He is suprxirteJ hy granr from the MRC anti the Canadian FounJJ.non for llem, and Colitis.

Figure 2 )
Figure 2) Cascade of inflammatory mediators m IBD