Downloaded from gut.bmj.com on 25 July 2008 896 COMMENTARIES

Crohn’s disease this observation has not been replicated ...... by other investigators, and there is no evidence of increased frequency of Crohn’s disease in the offspring of Does avium subspecies mothers versus fathers with Crohn’s disease. Even if transmission of viable paratuberculosis cause Crohn’s MAP occurs, a plausible mechanism of tissue injury and induction of chronic disease? intestinal inflammation has not been proposed. Even advocates of the theory R Balfour Sartor that MAP causes Crohn’s disease con- cede that , if present, consists ...... of a low bacterial load and that no histochemical evidence of acid fast Reassessing this persistent theory in light of advances in molecular staining in Crohn’s disease tissues is microbial detection and genetic pathogenesis of disease seen. This could be explained by a paucibacillary infection with an obligate imilarities between chronic idio- Data from this study help address intracellular, cell wall deficient bacterial 19 pathic granulomatous ileocolitis some of the controversies that have form. In this setting, inflammation Sand mycobacterial have fuelled the vigorous debate between and tissue injury must be mediated by been noted since the original descrip- committed advocates and confirmed a cell mediated immune response. tions of the clinical syndrome now sceptics that is receiving increasing However, a cellular immune response called Crohn’s disease.1–4 Interest in a attention in the scientific literature, lay to MAP has not been documented in 20 possible infectious origin of this disorder press, and internet chat rooms. The Crohn’s disease patients, despite was renewed in 1989 when Chiodini arguments in favour or opposed to this increased serological responses to MAP et al cultured apparently identical theory (table 1) have some merit but antigens in the same patients. Another Mycobacterium avium subspecies paratu- many are flawed by incomplete data and serious flaw in the MAP pathogenesis of berculosis (MAP) from three patients lack of rigorous reflection. There is no Crohn’s disease theory is the observa- with Crohn’s disease.5 This controversy doubt that a potential source of zoonotic tion that these patients respond to 21 increased in intensity following the infection exists, with widespread MAP chronic immunosuppressive therapies detection of the specific DNA insertion infections in the herds of Europe, and acquired immunosuppressive infec- sequence, IS900, of MAP in relatively North America, and Australia,13 14 excre- tions decrease disease activity as CD4 T 22 high numbers of patients with Crohn’s tion of MAP in from infected cell counts fall. In contrast, M tubercu- disease relative to and cows,15 relative resistance of intracellu- losis massively proliferates with anti- normal controls,6 and is now raging as lar MAP to widely used pasteurisation tumour necrosis factor or steroid treat- several different groups have detected techniques,16 and recovery of viable ment and M avium intracellulare thrives this organism in the food chain7 and MAP from the water supplying Los in the intestine as CD4 counts fall in water supply,8 proposed maternal-fetal Angeles.8 Moreover, the vast majority human immunodeficiency virus transmission in human milk,9 reported of studies using diverse techniques have infected patients. It is possible that long term responses to antimycobacter- detected MAP DNA or cultured this intracellular cell wall deficient MAP ial antibiotic combinations,10 and even organism in higher frequency from may not replicate well despite immuno- cultured viable M paratuberculosis in tissues of patients with Crohn’s disease suppression, but this issue has never blood samples of Crohn’s disease than from those with ulcerative colitis been studied by in vitro investigation or patients.11 and other disorders, although the in animals with Johne’s disease. In the Additional data to support an associa- reported frequency of recovery in both study by Autschbach et al, corticosteroid tion of MAP with Crohn’s disease is Crohn’s disease and ulcerative colitis therapy was associated with lower MAP provided by Autschbach and collea- have ranged from 0% to 100%.17 These detection rates.12 gues12 in this issue of Gut (see page results are consistent with two possibil- The most irrefutable evidence that a 944). This carefully performed and well ities: either MAP infection could cause microbial agent causes a disease is long controlled study used nested polymerase Crohn’s disease in a subset of patients term remission of clinical manifesta- chain reaction (PCR) to detect the IS900 that are either selectively exposed to this tions and an altered natural history of insertion element of MAP in 52% of organism or who are genetically suscep- disease following clearance of the infec- Crohn’s disease resected tissues versus tible to infection or, alternatively, this tion. In vitro sensitivity analyses show 2% of ulcerative colitis and 5% of mostly relatively common dietary organism that clinical isolates of MAP are not non-inflammatory control tissues. This may selectively colonise (or a dead responsive to traditional anti-M tubercu- study provides novel data regarding the organism selectively lodge in) the ulcer- losis agents, and therefore lack of effi- prevalence of MAP in various pheno- ated mucosa of Crohn’s disease patients cacy with isoniazid, ethambutol, and types of Crohn’s disease by showing but not initiate or perpetuate intestinal rifampicin treatment for two years with slightly higher detection of IS900 DNA inflammation. Molecular fingerprints a three year follow up23 does not detract in colonic (66.7%) compared with distal show that genotypes of bovine and from this theory. However, reports of ileal (40.5%) tissues and decreased human isolates are not similar but efficacy of combinations of clarithromy- detection rates with corticosteroid use. instead indicate that human and ovine cin or azithromycin, rifabutin, and a In addition, these authors reported () strains are more closely variety of other agents in 58–82% of weak associations with perianal involve- related.18 Maternal/fetal transmission of Crohn’s disease patients10 24 are also not ment and a shorter duration of disease MAP has been proposed following cul- definitive due to the uncontrolled nat- but no correlation with patient sex, age ture of MAP from breast milk of two ure of these studies, the small number at diagnosis, stricturing versus penetrat- patients with Crohn’s disease.9 of patients treated, the variable treat- ing phenotype, or presence of granulo- However, the frequency of positive ment regimens, and the fact that these mas. cultures in human milk is uncertain, antibiotics, particularly clarithromycin,

www.gutjnl.com Downloaded from gut.bmj.com on 25 July 2008 COMMENTARIES 897

Table 1 Arguments for and against a Mycobacterium avium subspecies MAP analogous to Helicobacter pylori in paratuberculosis (MAP) causation of Crohn’s disease peptic ulcer disease, gastritis, and gastric cancer, where host genetics and micro- Pros bial virulence factors determine (1) Clinical and pathological similarities between Johne’s and Crohn’s diseases34 immune responses that mediate clinical (2) Presence in food chain (milk, meat) and water supplies78 5633 disease in a small minority of patients (3) Increased detection of MAP in Crohn’s disease tissues by culture, PCR, FISH (4) Positive blood cultures of MAP in Crohn’s disease patients11 exposed to a widespread infectious (5) Increased serological responses to MAP in Crohn’s disease patients20 34 agent? Are we repeating the mistake of 9 (6) Detection of MAP in human breast milk by culture and PCR H pylori where the scientific establish- (7) Progression of cervical lymphadenopathy to distal ileitis in a patient with MAP infection35 (8) Therapeutic responses to combination antituberculosis therapy that include macrolide antibiotics10 24 ment resisted a new theory that chal- Cons lenged established paradigms of peptic (1) Differences in clinical and pathological responses in Johne’s and Crohn’s diseases4 ulcer disease until overwhelming clin- 36 (2) Lack of epidemiological support of transmissible infection ical evidence made such resistance (3) No evidence of transmission to humans in contact with animals infected with MAP (4) Genotypes of Crohn’s disease and bovine MAP isolates not similar18 untenable? Well designed clinical, (5) Variability in detection of MAP by PCR (0–100% in Crohn’s disease and ulcerative colitis tissues)8 microbiological, and mechanistic experi- and serological testing37 ments are urgently needed to defini- (6) No evidence of mycobacterial cell wall by histochemical staining tively settle this still unresolved debate. (7) No worsening of Crohn’s disease with immunosuppressive agents or HIV infection (8) No documented cell mediated immune responses to MAP in patients with Crohn’s disease20 To establish a causal relationship (9) No therapeutic response to traditional antimycobacterial antibiotics23 between MAP and Crohn’s disease, we need to determine if clearance of MAP MAP, Mycobacterium avium subspecies paratuberculosis; PCR, polymerase chain reaction; FISH, selectively changes the natural history fluorescent in situ hybridisation; HIV, human immunodeficiency virus. of disease in an infected subset of patients, perform definitive investiga- tions of cellular immune responses to have a broad spectrum of activity Crohn’s disease is that defective func- this organism in Crohn’s disease and against commensal enteric . tion of this gene results in ineffective control patients, determine if NOD2/ Moreover, these studies and a yet to be clearance of intracellular MAP infection CARD15 and other microbial signalling published ongoing controlled trial in and in decreased luminal a defensin pathways influence intracellular MAP Australia using these agents in Crohn’s secretion that permits increased muco- infection and clearance, and review disease patients are flawed by not sal adherence and epithelial invasion of results of ongoing large multi-institu- assessing IS900 DNA in biopsy speci- ingested organisms. Defective clearance tional studies to detect MAP in shared mens by PCR and serological responses of intracellular MAP by innate immune coded tissues by various molecular and to MAP before and after therapy, so that cells, including , could culture methods. These studies need to clinical results can be correlated with explain the seemingly paradoxical ther- be designed and conducted by estab- the presence of tissue MAP and its apeutic response of some Crohn’s lished investigators who bring no pre- clearance with treatment. Selective disease patients to granulocyte-macro- determined biases to this contentious responses in those patients with detec- phage colony stimulating factor.31 topic. If MAP is responsible for a subset tible MAP colonisation that clear infec- However, the phenotypic information of Crohn’s disease, public health mea- tion with antibiotic treatment would provided by Autschbach and collea- sures must be implemented to eliminate strongly imply a causal relationship of gues12 argues against an association of the source of infection in our food chain the infection. NOD2/CARD15 with MAP infection, as and food processing practices must be Our evolving molecular understand- MAP was more commonly detected in modified. In addition, the medical com- ing of gene/environmental interactions colonic than ileal disease and was not munity must develop ways to efficiently offers an opportunity to reassess the more frequently found in early onset or and cost effectively screen for MAP MAP causation theory of Crohn’s dis- stricturing disease.12 Likewise, patients infection and develop methods to effi- ease in a new light. NOD2/CARD15 is an with extensive ileocolitis responded bet- ciently clear this organism from infected intracellular receptor for muramyl ter to macrolide antibiotics and rifabu- tissues, possibly through a combination dipeptide (MDP), the smallest immuno- tin than did those with isolated ileal of effective antibiotics and immuno- logically active component of bacterial disease.10 These results directly contrast stimulants that enhance innate clear- peptidoglycan. Ligation of MDP by with the strong association of NOD2/ ance responses. If there is no evidence of NOD2/CARD15 activates nuclear factor CARD15 polymorphisms with early a causal association of MAP and Crohn’s kB. This pathway may contribute to onset ileal Crohn’s disease with a disease, we need to direct resources to clearance of intracellular bacterial infec- stricturing phenotype.32 other avenues of research. This contro- tion25 and secretion of a defensins by Crohn’s disease certainly has environ- versy has persisted far too long and Paneth cells, which constitutively mental and host genetic influences that needs to be expeditiously resolved. express NOD2/CARD15.26 The three interact to cause clinically evident dis- most common polymorphisms of this ease. It is equally clear that MAP is Gut 2005;54:896–898. gene are found in 25–35% of Caucasian widely present in our food chain and doi: 10.1136/gut.2004.055889 27 Crohn’s disease patients and lead to that the DNA of this organism can be Correspondence to: Dr R Balfour Sartor, UNC defective nuclear factor kB activation by recovered from the intestine of Crohn’s Department of Medicine/Division of MDP.28 29 Expression of the common disease patients. Although existing data Gastroenterology and Hepatology, CB #7032, truncation mutation of NOD2/CARD15 do not compellingly implicate MAP as a Room 7309 Biomolecular Bldg (MBRB), Chapel is associated with defective clearance of causal agent in Crohn’s disease, neither Hill, NC 27599-7032, USA; [email protected] invasive salmonella infection in epithe- do they definitively exclude this possi- Conflict of interest: None declared. lial cells.25 In addition, NOD2/CARD15 bility. We must determine whether MAP mutations in Crohn’s disease are asso- infection causes human disease, which REFERENCES ciated with diminished mucosal a is unlikely in my opinion, or whether 30 1 Crohn BB, Ginzburg L, Oppenheimer GD. defensin expression. Thus an attractive this environmental contaminant inno- Regional ileitis: a pathologic and clinical entity. explanation linking NOD2/CARD15 to cently lodges in ulcerated mucosa. Is JAMA 1932;99:1323–9.

www.gutjnl.com Downloaded from gut.bmj.com on 25 July 2008 898 COMMENTARIES

2 Dalzeil TK. Chronic intestinal enteritis. BMJ 15 Grant IR. Mycobacterium paratuberculosis and 27 Hugot JP, Chamaillard M, Zouali H, et al. 1913;2:1068–9. milk. Acta Vet Scand 2003;44:261–6. Association of NOD2 leucine-rich repeat variants 3 Greenstein RJ. Is Crohn’s disease caused by a 16 Lund BM, Gould GW, Rampling AM. with susceptibility to Crohn’s disease. Nature mycobacterium? Comparisons with , of milk and the heat resistance of 2001;411:599–603. , and Johne’s disease. Lancet Infect Mycobacterium avium subsp. paratuberculosis: a 28 Girardin SE, Boneca IG, Viala J, et al. Nod2 is a Dis 2003;3:507–14. critical review of the data, Int J Food Microbiol general sensor of peptidoglycan through muramyl 4 Chiodini RJ. Crohn’s disease and the 2002;77:135–45. dipeptide (MDP) detection. J Biol Chem mycobacterioses: a review and comparison of two 17 Sartor RB. Microbial influences in inflammatory 2003;278:8869–72. disease entities. Clin Microbiol Rev bowel disease: Role in pathogenesis and 29 Bonen DK, Ogura Y, Nicolae DL, et al. Crohn’s 1989;2:90–117. clinical implications. In: Sartor RB, Sandborn WJ, disease-associated NOD2 variants share a 5 Chiodini RJ, Van Kruiningen HJ, Thayer, WR, et eds. Kirsner’s inflammatory bowel diseases, 6th signaling defect in response to lipopolysaccharide al. Possible role of mycobacteria in inflammatory edn. Amsterdam: Elsevier Publishers, and peptidoglycan. Gastroenterology bowel disease. I. An unclassified Mycobacterium 2004:138–62. 2003;124:140–6. species isolated from patients with Crohn’s 18 Motiwala AS, Strother M, Amonsin A, et al. 30 Wehkamp J, Harder J, Weichenthal M, et al. disease. Dig Dis Sci 1984;29:1073–9. Molecular epidemiology of Mycobacterium avium NOD2 (CARD15) mutations in Crohn’s disease 6 Sanderson JD, Moss MT, Tizard ML, et al. subsp. paratuberculosis: evidence for limited are associated with diminished mucosal Mycobacterium paratuberculosis DNA in Crohn’s strain diversity, strain sharing, and identification alpha-defensin expression. Gut disease tissue. Gut 1992;33:890–6. of unique targets for diagnosis, J Clin Microbiol 2004;53:1658–64. 7 Millar D, Ford J, Sanderson J, et al. IS900 PCR to 2003;41:2015–26. 31 Dieckgraefe BK, Korzenik JR. Treatment of active Crohn’s disease with recombinant human detect Mycobacterium paratuberculosis in retail 19 Chiodini RJ, Van Kruiningen HJ, Thayer WR, granulocyte- colony-stimulating supplies of whole pasteurized cows’ milk in et al. Spheroplastic phase of mycobacteria factor. Lancet 2002;360:1478–80. England and Wales. Appl Environ Microbiol isolated from patients with Crohn’s disease. J Clin 32 Ahmad T, Tamboli CP, Jewell D, et al. Clinical 1996;62:3446–52. Microbiol 1986;24:357–63. 8 Mishina D, Katsel P, Brown ST, et al. On the relevance of advances in genetics and 20 Olsen I, Wiker HG, Johnson E, et al. Elevated etiology of Crohn disease. Proc Natl Acad Sci U S A pharmacogenetics of IBD. Gastroenterology antibody responses in patients with Crohn’s 1996;93:9816–20. 2004;126:1533–49. disease against a 14-kDa secreted 9 Naser SA, Schwartz D, Shafran I. Isolation of 33 Hulten K, El-Zimaity HM, Karttunen TJ, et al. purified from Mycobacterium avium subsp. Mycobacterium avium subsp paratuberculosis Detection of Mycobacterium avium subspecies paratuberculosis. Scand J Immunol from breast milk of Crohn’s disease patients. paratuberculosis in Crohn’s diseased tissues by in 2001;53:198–203. Am J Gastroenterol 2000;95:1094–5. situ hybridization. Am J Gastroenterol 21 Hanauer SB, Feagan BG, Lichtenstein GR, et al. 10 Gui GP, Thomas PR, Tizard ML, et al. Two-year- 2001;96:1529–35. outcomes analysis of Crohn’s disease treated with Maintenance infliximab for Crohn’s disease: the 34 Naser SA, Hulten K, Shafran I, et al. Specific rifabutin and macrolide antibiotics. J Antimicrob ACCENT I randomised trial. Lancet seroreactivity of Crohn’s disease patients against Chemother 1997;39:393–400. 2002;359:1541–9. p35 and p36 antigens of M. avium subsp. 11 Naser SA, Ghobrial G, Romero C, et al. Culture 22 James SP. Remission of Crohn’s disease after paratuberculosis. Vet Microbiol of Mycobacterium avium subspecies human immunodeficiency virus infection. 2000;77:497–504. paratuberculosis from the blood of patients with Gastroenterology 1988;95:1667–9. 35 Hermon-Taylor J, Barnes N, Clarke C, et al. Crohn’s disease. Lancet 2004;364:1039–44. 23 Thomas GA, Swift GL, Green JT, et al. Controlled Mycobacterium paratuberculosis cervical 12 Autschbach F, Eisold S, Hinz U, et al. High trial of antituberculous chemotherapy in Crohn’s lymphadenitis, followed five years later by prevalence of Mycobacterium avium subspecies disease: a five year follow up study. Gut terminal ileitis similar to Crohn’s disease. BMJ paratuberculosis IS900 DNA in gut tissues from 1998;42:497–500. 1998;316:449–53. individuals with Crohn’s disease. Gut 24 Shafran I, Kugler L, el-Zaatari FA, et al. Open 36 Sandler RS, Loftus EV. Epidemiology of 2005;54:944–9. clinical trial of rifabutin and clarithromycin inflammatory bowel diseases. In: Sartor RB, 13 Jubb TF, Galvin JW. Effect of a test and control therapy in Crohn’s disease. Dig Liver Dis Sandborn WJ, eds. Kirsner’s inflammatory bowel program for bovine Johne’s disease in Victorian 2002;34:22–8. diseases, 6th edn. Philadelphia: WB Saunders, dairy herds 1992–2002. Aust Vet J 25 Hisamatsu T, Suzuki M, Reinecker HC, et al. 2003:245–62. 2004;82:228–32. CARD15/NOD2 functions as an anti-bacterial 37 Bernstein CN, Blanchard JF, Rawsthorne P, et al. 14 Raizman EA, Wells SJ, Godden SM, et al. The factor in human intestinal epithelial cells. Population-based case control study of distribution of Mycobacterium avium ssp. Gastroenterology 2003;124:993–1000. seroprevalence of Mycobacterium paratuberculosis in the environment surrounding 26 Lala S, Ogura Y, Osborne C, et al. Crohn’s paratuberculosis in patients with Crohn’s disease Minnesota dairy farms. J Dairy Sci disease and the NOD2 gene: a role for paneth and ulcerative colitis. J Clin Microbiol 2004;87:2959–66. cells. Gastroenterology 2003;125:47–57. 2004;42:1129–35.

Colitis beyond inherent basic nutrition’’.3 ...... Bacteria associated with probiotic activ- ity are most commonly lactobacilli, bifidobacteria, and streptococci but Probiotics and barrier function in colitis other non- such as some strains of Escherichia coli and P Gionchetti, K M Lammers, F Rizzello, M Campieri microorganisms such as the yeast Saccharomyces boulardii have been used ...... in IBD. Probiotic administration may exert a protective effect in colitis by Encouraging results have been obtained with probiotics in several preventing mucosal barrier disruption and influencing the extent experimental animal models of IBD.4–7 of mucosal injury In humans, probiotics are effective in the prevention of pouchitis onset and 8–10 here is strong evidence of a role for undiscovered due to technical limita- relapse. Results in ulcerative colitis the indigenous flora in driving tions, it was hypothesised that modula- are promising, both in prevention of inflammatory responses in inflam- tion of an abnormal microflora in these relapse and treatment of mild to mod- T 11–13 matory bowel disease (IBD) in geneti- patients by introducing high titres of erate attacks. Results in Crohn’s cally predisposed individuals.1 For years, ’’protective’’ bacteria might overwhelm disease are not yet clear because of researchers have tried in vain to identify the ’’aggressive’’ strain(s) and inhibit its conflicting data and the limited number a specific pathogen as the cause of these deleterious effects. On this basis, pro- of well performed studies.14–16 chronic intestinal inflammatory disor- biotic treatment was proposed as a Efforts are being made by many ders but the possibility that one or more therapeutic approach.2 researchers to unravel the precise bacterial agents are responsible cannot Probiotics are defined as ‘‘living mechanisms by which probiotic bacteria be ruled out. Considering the implica- organisms which, on ingestion in cer- and their metabolic products (short tions of a pathogen in IBD, as yet tain numbers, exert health benefits chain fatty acids, vitamins) exert their

www.gutjnl.com