Most of Crohn's disease is caused by Mycobacterium avium subsp.paratuberculosis (MAP). Although not all the evidence relevant to this is yet fully published, it is compelling.
MAP is a specific cause of chronic inflammation of the intestine in many species of animals including primates 1,2. Of great importance to our understanding of its role as a human pathogen is the recognition that the disease in animals presents a broad range of histopathological types, from pluribacillary to paucimicrobial with chronic granulomatous inflammation, just like leprosy in humans 3. The paucimicrobial form of the disease occurs particularly in sheep 4,5. In such animals MAP cannot be seen in the tissues microscopically, cannot usually be cultured, IS900 PCR is only positive about half the time (depending much upon how the tissue sample is processed), and the standard veterinary immunological tests for MAP infection are unreliable or negative. These animals with paucimicrobial MAP infection have a chronic granulomatous ileocolitis like Crohn's disease in humans.
MAP can exist within the tissues of animals for years without causing clinical disease 6. Subclinical infection with MAP is widespread in Western European cattle, sheep and goats and appears to be extending south and east. MAP is harboured in wild species such as deer and rabbits 7,8. Other industrialised societies in temperate regions of the globe are similarly affected. Subclinically infected animals shed abundant MAP onto pastures from where it may find its way into water supplies 9,10. As with some other organisms, parasitisation of free-living environmental amoebae with MAP may enhance its virulence 11. MAP is robust and versatile and would not be inactivated by current water treatment procedures. Subclinically infected dairy cattle shed MAP in their milk 12. Pasteurisation at 72oC for 15 seconds which was introduced in the first half of this century to ensure the destruction of Mycobacterium bovis and Mycobacterium tuberculosis, does not ensure the destruction of the tougher MAP 13. Studies we carried out throughout central and southern England from 1990 to 1994 demonstrated intact MAP in retail pasteurised milk with a phenotype consistent with a viable, but very slow growing form of the organism 14. Delay has resulted from the understandable, but inappropriate reluctance to recognise the potential scale of this problem. Recent studies from the Department of Food Science, Queens University, Belfast 15, showed that 6 of 31 (19%) cartons of pasteurised cows' milk obtained from 16 outlets in Northern Ireland tested positive for MAP by IS900 PCR. Bactec liquid cultures were positive on all 6 cartons, with acid-fast mycobacteria visible microscopically, and a single colony appeared on a slope of Herrold's egg yolk medium. Despite early warnings (January 1992) to the Ministry of Agriculture in Britain, conditions of pasteurisation which do ensure the destruction of MAP have not been accurately identified. UHT treatment of milk (134oC for 1 second) which kills vegetative organisms cannot be assumed to kill a resilient organism like MAP which tends to exist in clumps. It is certain that human populations in regions where domestic livestock harbour MAP, are exposed to these pathogens.
With the opportunity to amplify in our efficient but intensive farming for over a hundred years, MAP has probably undergone an adaptive radiation and has taken up the intestine of animals and humans as one of its natural habitats, acquiring as it has done so an intermediate status between an environmental organism and a low-grade pathogen. Extensive studies by Dr Ivo Pavlic 16 from the Veterinary Research Institute, Brno, in the Czech Republic on 740 individual isolates of MAP from animals and humans throughout the world has so far demonstrated 23 different genotypes with ovine and bovine sub-species. Can MAP cause chronic inflammation of the intestine in humans ? This proposal was first published in 1913 17. Now at the end of the 20th Century, the increasing number of people struggling through life with Crohn's disease, and their families, have a right to demand a clear answer to this question from Gastroenterologists, some of whom at present appear tragically confused, or preoccupied with the mechanisms of inflammation.
MAP has not yet been visualised microscopically in the tissues of humans with Crohn's disease. After months or years of incubation it can however be cultured to conventional bacillary form from the inflamed intestine of about 5% of Crohn's disease patients, rather like sheep with paucimicrobial MAP infection. The detection rate in Crohn's disease rises to about 30% when IS900 PCR is applied to long term cultures 18,19. Recent data obtained by applying IS900 PCR to Bactec cultures after only 6 weeks of incubation so far demonstrate a detection rate for MAP in 6 of 6 Crohn's disease tissues and none of 10 non-IBD controls 20. IS900 PCR can be applied directly to DNA extracted from tissues. Since the demonstration six years ago of MAP in the intestine in Crohn's disease by these means 21, at least 16 other studies have been reported using a wide range of sample processing procedures and PCR methodologies, 8 of which were IS900 positive and 8 negative. In some of the negative studies the methods used were wholly inadequate for the demonstration of a low abundance pathogen in a clinical sample. Similar difficulties have occurred in the PCR detection of the causative agent in other chronic inflammatory diseases such as end-stage Lyme disease, brucellosis and tuberculoid leprosy. The methods used for tissue sample dissolution and DNA extraction are absolutely critical. Recent work on these by Dr Ann Verstocken and others at St. George's has lead to what may be an important observation. This is that treatment of a sample with 6M guanidine HCl at 50oC or prolonged incubation in SDS protease K which reliably release the DNA from other bacteria and spheroplasts, do not do so for MAP in the form that this organism exists in the tissues of sheep with paucimicrobial MAP infection and humans with Crohn's disease. Reliable access to MAP DNA in both sheep and humans required sample dissolution in SDS protease K followed by mechanical disruption by vibration in a slurry of silica and ceramic particles at 6.5 m/sec for 45 seconds using the Hybaid Ribolyser system. A nested PCR is then carried out for IS900 which is uniquely specific for MAP. Using these methods in a blinded study with Dr Jeremy Sanderson at Guy's Hospital London, we have shown unequivocally that MAP is present and detectable in endoscopic mucosal biopsies from the ileum and colon in a substantial proportion of people who do not have inflammatory bowel disease, just as is the case for many other pathogens, including H.pylori. This further demonstrates the exposure of the population to this organism. MAP is also present in mucosal biopsies from patients with Crohn's disease. Large collaborative studies between institutes, including tests on full thickness surgically resected tissues now need to be carried out using these optimised methodologies.
Researchers at the University of Central Florida and Baylor College of Medicine in Texas isolated two proteins p35 and p36 from a genomic expression library of MAP. Either or both of these were recognised by 93% of Crohn's disease sera and 25% non-IBD controls (p< .001). Recognition of both was seen in 74% of Crohn's disease and in none of the controls 22. Our own studies by Marcus Reddy and Nazira Sumar in 104 Crohn's disease patients and the same number of matched controls, have demonstrated the presence of an epitope within the sequence RHTQAVLALARRRLN in the carboxyterminal portion of p43 encoded by IS900, which is recognised in a highly significant manner (p<.0001) by IgG in Crohn's disease sera.
We do not yet know the precise phenotype of MAP in Crohn's disease. It is not that of an acid-fast mycobacterium, nor is it a spheroplast. It appears to exist in a tough protease-resistant form perhaps extensively coated by carbohydrate. It is present in low abundance, probably within macrophages and replicating only very slowly. It is unlikely to cause disease by releasing toxins or molecules which are directly granulomagenic. It is much more likely to cause an intermittent immune disregulation within the intestinal wall which, with accompanying defects in the overlying epithelial layer, results in a chronic inflammatory and allergic response to bacteria and other constituents from the intestinal lumen. Clinical improvement can be achieved by suppressing or modulating the immune response itself, or by reducing the intensity of the allergic component with elemental diets. Some clinical improvement can also come from the use of general antimicrobial agents such as metronidazole and ciprofloxacin, as demonstrated by the extensive experience of Prantera et al in this issue of the Journal. Without killing the causative organisms however, such therapeutic approaches are unlikely to achieve lasting resolution of the disease.
M.avium infections in immunocompetent hosts are difficult to eradicate. M.avium sp. are naturally resistant in vivo to many drugs used in the treatment of M. tuberculosis such as rifamycins, isoniazid and ethambutol 23. M.avium sp. are able to prevent these agents from penetrating the microbial cell, and can rapidly develop specific mutations which confer drug resistance. The recent conclusion of Thomas et al 24 that mycobacteria cannot play an important part in the pathogenesis of Crohn's disease because their patients treated with rifampicin, isoniazid and ethambutol did not apparently get better, is wholly naive and ignorant of an extensive microbiological and chemotherapeutic literature in the field. Furthermore, drugs such as isoniazid and ethambutol exert their antimycobacterial actions mostly by inhibition of cell wall component synthesis, including especially mycolic acids and the polymerisation of arabinan. Neither of these drugs are likely to be active against MAP in vivo which does not in any case, require a conventional mycobacterial cell wall. Chemically modified natural streptomyces antibiotics such as rifabutin , clarithromcyin and azithromycin, which target mycobacterial protein synthesis and which also have the particular advantage of being concentrated with macrophages, were predicted to be more active against very slowly replicating low abundance MAP causing Crohn's disease. In the absence of funding to support a randomised controlled trial, we carried out a careful outcomes analysis of the use of combination therapy with these agents which showed a highly significant improvement in disease severity parameters accompanied by healing of the inflamed intestine 25. These results have recently been supported by work carried out independently in Sydney Australia 26. New chemotherapeutic agents effective against MAP infections are needed. The recent discovery of a 'pathogenicity island' (designated 'GS') in MAP 27 has identified genes which are also promising candidates for DNA and protein-based therapeutic vaccines to assist in microbial clearance. Of overwhelming importance however, is the need for government agencies, public health institutions and the profession to recognise the true nature of the increasing problems being caused by MAP, and design and implement the necessary remedial measures.
1.Chiodini RJ, Van Kruiningen HJ, Merkal RS. Ruminant paratuberculosis (Johne's Disease): The current status and future prospects. Cornell Vet 1984,74: 218-262.
2.McClure HM, Chiodini RJ, Anderson DC, Swenson RB, Thayer WR, Coutu JA. Mycobacterium paratuberculosis (Johne's disease) in a colony of stump-tail macaques (Macaca arctoides). J Infect Dis 1987, 155:1011-1019.
3.Hermon-Taylor J, Barnes N, Clarke C, Finlayson C. Mycobacterium paratuberculosis cervical lymphadenitis, followed five years later by terminal ileitis similar to Crohn's disease. Brit.Med.J. 1998, 316: 449-453.
4. Clarke CJ, Little D. The Pathology of ovine paratuberculosis: histological and morphometric changes and correlations in the intestine and other tissues. J Comp Path 1996, 114:419-437.
5.Clarke CJ. The Pathology and Pathogenesis of Paratuberculosis in Ruminants and Other Species. J.Comp.Path. 1997, 116: 217-261.
6.Çetinkaya B, Egan K, Morgan KL. An abbattoir-based study of the prevalence of subclinical Johne's disease in adult cattle in south west England. Epidemiol.Infect 1996,116: 373-379.
7.Stehman SM. Paratuberculosis in small ruminants, deer, and South American Camelids. Vet.Clin.N.America: Food Animal Practice. 1996, 12: 441-455.
8.Grieg A, Stevenson K, Perez V, Pirie AA, Grant JM, Sharp JM. Paratuberculosis in wild rabbits (Oryctolagus cuniculus.Vet.Record 1997,140:141-143.
9.Hermon-Taylor J. Causation of Crohns Disease: the impact of clusters. Gastroenterology 1993, 104: 643-646.
10.Mishina D, Katsel P, Brown ST, Gilberts ECAM, Greenstein RJ. On the etiology of Crohn's disease. Proc.Natl.Acad.Sci USA. 1996, 93: 9816-9820.
11.Cirillo JD, Falkow S, Tompkins LS, Bermudez LE. Interaction of Mycobacterium avium with Environmental Amoebae Enhances Virulence. Infection and Imunity. 1997, 65: 3759-3767.
12.Sweeney RW, Whitlock RH, Rosenberger AE. Mycobacterium paratuberculosis cultured from milk and supramammary lymph nodes of infected asymptomatic cows. J Clin Microbiol 1992,30: 166-171.
13.Grant IR, Ball HJ, Rowe MT. Effect of high-temperature, short-time (HTST) pasteurisation on milk containing low numbers of Mycobacterium paratuberculosis. Letts in Appl.Microbiol. 1998, 26: 166-170.
14. Millar D, Ford J, Sanderson JD, Withey S, Tizard M, Doran T, Hermon-Taylor J. IS900 PCR to detect Mycobacterium paratuberculosis in retail supplies of whole pasteurised cows' milk in England and Wales. Appied & Environ.Microbiol 1996, 62: 3446-3452.
15.Irene Grant. Department of Food Science, The Queen's University Belfast, Northern Ireland, U.K. Personal communication 1998.
16. Pavlik I, Bejková L, Pavlas M, Rozsypalová Z, Kosková S. Characterization by restriction endonuclease analysis and DNA hybridization using IS900 of bovine, ovine, caprine and human dependent strains of Mycobacterium paratuberculosis isolated in various localities. Vet Microbiol 1995, 45: 311-318, and personal communication 1998.
17.Dalziel TK. Chronic interstitial enteritis. Br Med J 1913,ii: 1068-1070.
18.Moss MT, Sanderson JD, Tizard ML. Hermon-Taylor J. El-Zaatari FA, Markesich DC et al. Polymerase chain reaction detection of Mycobacterium paratuberculosis and Mycobacterium avium subsp. silvaticum in long term cultures from Crohn's disease and control tissues. Gut. 1992, 33: 1209-13.
19.Wall S, Kunze ZM, Saboor S, Soufleri J, Seechum P, Chiodini R et al. Identification of spheroplast-like agents isolated from tissues of patients with Crohn's disease and control tissues by polymerase chain reaction. J.Clin.Microbiol. 1993, 31: 1241-5.
20.Saleh Naser and Ira Shafran, Dept of Molecular Biology and Microbiology and Centre for Diagnostic and Drug Development, University of Central Florida, Orlando, Florida USA. Personal communication 1998.
21.Sanderson JD, Moss MT, Tizard MLV, Hermon-Taylor J. Mycobacterium paratuberculosis DNA in Crohn's disease tissue. Gut. 1992, 33: 890-6.
22.Naser S, Shafran I. Fenster W, El-Zaatari F. Serological Analysis Strongly Associates Mycobacterium avium subsp. paratuberculosis with Crohn's disease. Gut 1998, submitted.
23.Morris SL, Rouse DA. The genetics of multiple drug resistance in Mycobacterium tuberculosis and the Mycobacterium avium complex. Res.Microbiol 1996, 147: 68-73.
24.Thomas GAO, Swift GL, Green JT, Newcombe RG, Braniff-Mathews C, Rhodes J, Wilkinson S, Strohmeyer G, Kreuzpainter G. Controlled trial of antituberculous chemotherapy in Crohn's disease: a five year follow up study. Gut 1998, 42: 497-500.
25.Gui G, Thomas P, Tizard M, Lake J, Sanderson JD, Hermon-Taylor J. Two year outcomes analysis of Crohn's disease treated with rifabutin and macrolide antibiotics. J Antimicrob Chemotherap 1997,39: 393-400.
26.Borody T, Pearce L, Bampton PA, Leis S. Treatment of severe Crohn's disease using Rifabutin-Macrolide-Clofazimine combination: Interim Report. Gastroenterology 1998, 114 A938.
27.Tizard MLV, Bull T, Millar D, Doran T, Martin H, Sumar N, Ford J, Hermon-Taylor J. A low G+C content genetic island in Mycobacterium avium subsp paratuberculosis and M.avium subsp silvaticum with homologous genes in M.tuberculosis. Microbiology 1998, 144: in press.