Overview of inflammatory bowel disease pathogenesis

Inflammatory bowel disease (IBD) represents a difficult and 
challenging condition for patients, clinicians and basic investigators alike. Its 
etiology and pathogenesis are still unclear in spite of extensive investigations 
that have yielded a wealth of clinical. epidemiological, biochemical, bacteriological 
and immunological data on Crohn 's disease and ulcerative colitis. 
Although the precise mechanism(s) responsible for the intestinal inflammatory 
process remain to be defined, enough information has been assembled to 
hypothesize which components are likely to be important for this probably 
multifactorial disease. A consistent association between class I or II histocompatibility 
antigens and either Crohn's disease or ulcerative colitis has yet to be 
found. Nevertheless, ample epidemiological studies leave no doubt about the 
high frequency of familial clustering, and it must be determined whether this 
phenomenon translates a true genetic predisposition or a common environmental 
exposure, or both. Immune events occurring in the gastrointestinal tract 
are unquestionably linked to the pathogenesis of IBD, but it is unknown which 
are primary or secondary in nature. While most immune abnormalities detected 
in patients with established disease are likely to represent secondary events, these 
are no less important, as they probably contribute to the perpetuation of gut 
inflammation and tissue damage. This does not exclude that IBD is due to a 
primary defect of intestinal immunity, but this may no longer be detectable at 
the time of clinical manifestations. The answer to the question of wh1ch of the 
various intestinal immune abnormalities is central to pathogenesis must wait for 
additional research. Whether immune responses to the luminal flora, antigen 
processing mechanisms, antibody production, immunoregulation, cytotoxic activity, 
cytokine and mediator release are defective or disregulated is under intense 
investigation. It is likely that several of these events are involved, but they may 
interact in a complex and unpredictable fashion. lt is almost certain that there 
are various initiating and secondary events, and different immune mechanisms 
share relatively few common pathways for damaging the intestine, eg, cytokines, 
arachidonic acid metabolites, and oxidants. Perseverance in the study of these 
substances is finally yielding promising new approaches to the manipulation of 
immune and inflammatory responses chat cause bowel destruction. Future drugs 
may consist of combinations of highly specific inhibitors, antagonists or receptor 
blockers, that may selectively block one or several steps of the inflammatory 
cascade which is chronically active in the intestine of affected individuals. 
Therefore, we may soon face a situation not too dissimilar from what we have 
recently witnessed for peptic ulcer disease. The specific cause of IBD may still be 
beyond our comprehension, but a better understanding of its pathogenesis al lows 
us to put highly effective therapies within reach.

A NY A1TEMPT TO DISCUSS, MUCII less define, the pathogenesis of inflammatory bowel disease (IBO) is a formidable and challenging task.Indeed, too many questions abouL the cause and mechanisms of IBD go presently unanswered.As a consequence, trying to establish definitively how this entity comes about, and why it persists chronically m the gastrointestinal tract of a sizeable portion of the world's population, appears to be beyond present capabilities and constitutes a major scientific problem.Although both Crohn's disease and ulcerative colitis, two generally well defined and distmct clinical entities, share the denomination of IBO, IBD is common ly referred to and thought of as a single entity.This erroneous concept derives &om the numerous clmical, epidemiological, pathological and lherapeutic similarities between Crohn's disease and ulcerative colitis, but also from ignorance about these two diseases.As a matter of fact it is not even certain whether either one is a single entity, or whether each one represents an umbrella which covers a variety of other disorders with shared protean clinical manifestations ( L ).This is particularly true for Crohn's disease; the possibility that it may represent a syndrome rather than a distinct illness has recently become the focus of attention (2).Even assummg that Crohn's disease and ulcerative colitis are clearly separable entities, it is unknown whether either is caused by one or more agents, and whether single or multiple pathogenetic events are implicated in triggering and perpetuating the mechanism(s) of intestinal tissue injury.

POSSIBLE PATHOGENIC COMPONENTS IN IBD
In spite of th is confusing situation, a careful and nbiective analysis of the knowledge accumulated durmg the past two decades has narrowed the spectrum of causes and mechanisms that arc likely to have d irect or indirect relevance to the pathogenesis of IBD.
As a result, it 1s now generally assumed that a variet 7 of stimuli ca n trigger m susceptible individuals an immune response that will ulumately result m the local ( intestinal) mflammatory reaction rcsponsih le for the clinical and pathological manifestations whteh lead to the d iagnosis of Crohn's disease o r ulc.crativc colitis.A series of possible pathogentc components .irelisted in Table l, and because it ts presently impossible to single out any one of them as the defimte cause of IBO, each deserves to be considered.Infectious agents: Smc.c t h e very begmning of the mvest1gatton of the cause of IBO, infectious agents have been considered pnme causative candidates, mostly due to the similarities of ulcerative colitis lll coliudes of proved infectious etiolO!,'Y, such as shigellos1s, salmonellos1s, etc.Over the years, a large numher of microbiological agents  (3 .4).Later, the so-called 'tr,m,.m1ss1blc agents' became the focu, <:i attention, based on evidence of repttl• nvc mJucunn of granuloma fonnari,m in mouse footpads ( 5), cytopath1c effect in tissue culture (6), and lymphomasm nude mice (7).All have hcen found later to be nonspecific and Jue to no11-rcpl1c.atingtoxic substances (8-10).
M ycohactcria, which have long Ix-en con sidered a possible cause fo r Cmhn's disease because of the presence of granu lomatous inflammation in the gu t, have had a recent resu rgence: m popularity ( 11 ).
A considcrahle amoun t of data hal'e  an enhanced or decreased cellmediated immunity to mycobacterial antigens (17).Finally, preliminary studies do not show an elevation serum antibody titres to the mycobacterial cross-reacting heat shock proteins (HSP65) (unpublished daca), and there is no evidence for an increase of gamma/ delta T cell receptor-positive (TCRl) cells in intestinal lesions of Crohn's disease (18) (personal communication), as might be expected in an active mycobaCleria infection (.19).Therefore, in spite of che homology anJ nnmunological cross-reactivity between mycobacterial and human heat shock proteins (20), and the postulated role of these stress proteins in inflammation (21 ), the probability that some type of mycobacteria is responsible for Crohn's disease is inadequately substantiated.

Evidence for mycobacterial involvement in Crohn's disease
Even though a specific bacterial, viral or fungal agent has yet to be iJentified, the role of microorganisms, and in particular the intestinal bacterial flora in the pathogenesis of IBO cannot be dismissed.The simple presence of a massive amount of antigens and mitogens of bacterial origin in the gut lumen must be taken into account hecause of its unquestionable impact on the local immune system and intrinsic.inflammatory propcnies (Table 3 ).Endotoxin, a component of the cell wall of Gram-negative bacteria, is one of the most potent immunomodulatory substances (22).Enternbacterial common antigen is present in all enrerobacteriaccac and cross-reacts with intestinal epithelial cells (23 ), and an enhanced humoral and cellular immune response to this antigen b detected in IBO patients (24,25).Peptidoglycans from Gram-positive baccerial cell wall can cause subacute inflammation when injected in to the bowel wall (26), and intraluminal administration of the bacterial chemotactic peptide forrnyl-methionyl-leucylphenylalanine can induce colitis in expenmental animals (27).A variety of additional products from the gut flora may exist with similar properties.At present it is impossible to determine what the exact role might be for these potent immunomodulatory and pro-  (34).
A hypothesis that presently enjoys much popularity nmong clinical immunologists is that of 'molecular mimicry' associated with a 'hit-and-run event' (35).In this concept, a primary agent, such ns a virus, a bacterium or a foreign protein may attack a specific organ, triggering a local immune response, directed agamst the aggressor.This noxious offender may be eliminated or JcstroyeJ, but 111 doing so che immune system is forced to produce antibo<l ies or generate cells sensimed to some of the offender's antigens which share the same molecular configuration with suhscances of the host's tissues (molecular mimicry).As a consequence, the immune system is now able to recognize and attack normal cells,  triggering an autoaggressive reaction even though the initial culprit has comp lere l y disappeared (hit-and-run event).This theory has been suggested to explain some diseases such as glutensensitive enteropathy, ankylosingspondylitts and Reiter's syndrome, based on evidence of molecular mimicry of the wheat protein A-gliadin with the El B protein of the human adenovirus A<l-12 for celiac disease (36), and the HLA-B27 allele with Klebsiella pneumoniae nitrogenase for the mentioned arthropathies (37).If a slmilar mechanism exists in IBO, then the task of the investigators would be that of discovering wh ich primary agent may trigger an immune response against what intestinal tissue or cell.However, not only do we not know what agent may havl!triggered an autoimmune response, but we are also not sure what the target of such a response 1s m spite of many having been proposed (Table 4).Fecal and howel extracts have been used m the initial stud ies, hut the crudeness of these preparations and the variability and inconsistency of results make the reported results difficult to evaluate (38).The pn:sence of immunity to enterobacterial common antigen has already been mentioned, but this 1s not specific for IBD (39).The isolation of purified and well characterized intestinal antigens has added more scientific credibility co the investigation of autoimmune phenomena in IBO.The detection of intestinal mucosa!T cells reactive again~t gut-specific epithelial cell -assoc1ated components lends some support to the existence of cyrocox1c celb aimed at intestinal cells ( 40) .
Similarly, the isolation of lgG tissuebound antibodies in ulcerative colitis directed cowards a unique colonic antigen (Mr 40,000 protein) speaks in favour of a specific form of autoimmunity in at least one form of IBD.Additional support for the potennal value of this colonic autoant1gen has been provided recently by the demonstration that the expression of Mr 40,000 is restricted to the colon, biliary tree and skin, locations which perfectly match the sites of clinical manifescacions of ulceracive coliris (41).A recent report describing a new type of ancineucrophil cytoplasmic anubody in ulcerative colitis (42), and the already mentioned sharing of biochemical and 1mmunolog1cal characteristics between heat shock proteins and bacterial stress proteins (20), further increase the need to investigate these novel findings and study their relevance to autoimmune events in IBO.By the time of obvious clinical signs and symptoms of IBO, it may be impossible to recognize an earlier autoimmune response, but this does not diminish the importance of subsequent aurmmmune reacuons that may be cntical in explaining the relapsing nature of intestinal inflammation.Abnormal immune response: The par-cic1pat1on of the immune system in the pathogenesis of chronic inflammatory diseases is nor restricted to autoimmune responses, in which a functional immune system incorrectly interprets the hose's antigens as foreign substances.The immune system itself may be intrinsically abnormal and mountadefecnvc or inappropriate response against an appropnate target.Due to the complexity of the immune system and the multipltc1ty of its cellular and soluble components, the potential for a defective step can reside at the level of any of a myriad of cells, anubodies and cyrnkines.The purpose of this discussion is nor to review all reported anomalies of systemic and intestinal immunity in IBO, since excellent references are available (43).Rather, 1t 1s appropnate to seres..~ the unquestionable role chat intestinal immunity plays in mediating the inflammatory process of Crohn's disease anJ ulcerative colitis (44).A variety of findings have been reported, including abnormalines of T cell activation, antibody produwon, cytokine act1v1t1es, complement acuvacion, ere (40.45-49).On one end a maJor challenge facing the invemgators of IBO 1s to try to separate primary from secondary mucosa!immune abnormalities.On the other end, as discussed for autoimmunity, chis challenge would be a trivial point, as different immune events are probably important at different stages of the disease.Indeed, most of chem could be implicated as mediators of ussue inJury, as will be discussed later.Psychoneuroimmunology: Clinical observations have long called attention to the incnguing relauonship between stressful events oflife and the beginning of Crohn's disease or ulcerative colms, and the modulation of the clinical course ofIBO by social factors.After the transient trend of purely psychosomatic theories, sc 1ent1fic support for a demonstrable physical link between the environment and the different orgamof the body has been found in combined studies of the neuroendocrme anJ neuro1mmune systems (50).The mutual functional relationship between the nervous, endocrine, and immune systems is now well established, and they probably form the anatomical, physiological and cellular bases explaining why and how life events may modulate intestinal immune respome, in health and disease (51).It 1s far tcxi premature co define complete pathway, and postulate hypothe~cs for IBO.Nevertheless, this area of investigation is beginning to be vigorously pursued, and may yield information of sc1entiltc and practical value in the near future.

POSSIBLE MECHANISMS OF T ISSUE DAMAGE IN IBO
ln any disease the study of the mechanisms of nssue damage holds the key to treatment, as drugs or other measures can he devised to block, neutralize or inhibit active agent,, revert the pathl)log1cal process, and restore nonnality.As already indicated, autoimmune phenomena and abnormalities of intestinal immunity are likely contributors to intestinal inflammation in Crohn's disease and ulcerative colitis.However, the capacity for modulating the immune system is limited at present, being restricted to the use of drugs with variable degrees of specificity for the different mononuclear cell subsets (52).Upon activation these cells produce a variety of soluble mediators, most of which are well defined functionally and biologically (53).These biological activities are broad and extremely potent, makingcytokines very pertinent to and directly responsible fo r many of the pathological manifestations of tissue damage.A dramatic example of their action and the beneficial result of their modulation has been recently described in an animal model.ln mice undergoing graft-versus-host disease the severe intestinal involvement and related mortality was almost entirely abolished by administration of antibodies to tumour necrosis factor ( 54).
Therefore, there is little doubt that activated mononuclear cell-derived soluble factors can amplify and maintain intestinal inflammation (55).This strongly justifies the intense study of cytokines which is currently being pursued in IBO, by measuring their levels in the inflamed mucosa (Table 6), as well as assessing their effect on the local immune cells (56).Non immune factors are also present an<l obviously active in inflamed mcestinal tissue.Proscaglandins have received much attention as 'proinflammatory' substances, bur more recently their action is believed to be more of a cytoprotectivc nature, whereas thromboxanes an<l especially leukotrienes are currently held as responsible for tissue destruction and inflammation, as discussed elsewhere.More and more attention is being devoted to the investigation of the increasingly important role of polymorphonuclear neutrophils in tissue damage (57).Neutrophils are abundant in active lesions of both human IBO and experimental animal colitis, and one of their actions is the generation of oxygen free radicals, which is also discussed elsewhere.It is quite possible chat during inflammation additional unidentified substances are released that possess tissue-damaging potential.

Pathogenesis of IBD -An overview
Gascroenterologists and surgeons, forced co deal with the reality of the clinical manifestations of IBO but lacking basic knowledge of its cause, have obtained a reasonable degree of success in creating patients affected by Crohn's disease and ulcerative colitis.In some skeptical minds, the need for basic information about the nature of these diseases may even seem irrelevant to practical treatment.Nevertheless, both practising clinicians and basic investigators have the obligation to share knowledge and experience, trying to help and complement each other.The process of mcorporatmg basic knowledge into clinical therapy is alway~ painstaking and slow, but <luring the past decade enough new data on gut immunology and inflammatory mediators have been gathered that future therapeutic intervention for 180 may derive from this newly acquired body of mformacion.In the following paragraphs, an attempt will be made to show how this may be so.
In relation to the temporal appearance of clinical manifestations, treatment of IBO can be classified as early or late.Early treatment implies the actual prevention of disease, or its detection before florid clinical manifestations become evident.To do so it is imperative co detect or measure parameters that reflect an enhanced susceptibility to IBO or a higher than normal chance of acquiring the disease.Unfortunately, no 'markers' are available, although some candidates are under investigation (Table 7).
The continued study of immunogenetics is exploring new single genes or a combination of them which may show a consistent association with Crohn's disease or ulcerative colitis, as detailed elsewhere.The recently  (58).A similar situation may exist for the recently described am inuclear cywplasmic antibodies preferentially found in ulcerative colitis (42).Further studies of intestinal m ucin abnormalities may show an intriguing assoc iation with ulcerative colitis (59), while increased intestina l permeabil ity may be holding clues to the development of Crohn's Jisease (60).The existence of a well defined high risk environment may suffice to predict a higher than normal occurrence of IBO.same is true for several of the proinflammatory cytokines, including mterle ukin-1, incerleukin-6, tumour necrosis factor, p latelet act ivating factor, etc.In vitro and in vivo experiments using these strntegies a re actually being conductetl in several Laborator ies.With the realization of the potent tissue damaging potential of oxygen-derived free radicals, oxygen radical scavengers a re being screened in animal models and clinical trials.One of the actions of 5-aminosal icylic acid, t he active principle of sulphasalazine and one of the most effective drugs for the treatment of IBO, appears to be through its oxygen radical scavenger activity (61 ).New ant ibiotics are constantly being generatctl that may prove efficacious in some cases with an infectious cause or complication.

TABLE 1
Possible pathogenic components In

TABLE 4 Possible targets of autoimmune phenomena In Inflammatory bowel disease
Thus, to understand what mechanisms are crucial to gut tts• sue damage 1s obviously of fundamental importance m IBO.For practical pur• poses, rhc candidate mechanisms can

TABLE 6 Cytokine activities in the intestinal mucosa of inflammatory bowel disease• Experimental conditions Cytokine Unstimulated Stimulated
?•tevels in organ culture, cell culture, cell extract or mRNA compared to levels from histologlcally normal control Intestinal mucosa I In inactive ulcerative colitis