Antibodies against Mycobacterium paratuberculosis in

Transcription

1 Q J Med 1996; 89: Antibodies against Mycobacterium paratuberculosis in Crohn's disease R.S. WALMSLEY, J.P. IBBOTSON\ H. CHAHAL 1 and R.N. ALLAN From the Gastroenterology Unit, Queen Elizabeth Hospital, Birmingham, and 1 Department of Infection, University of Birmingham, Birmingham, UK Received 8 August 1995; Accepted 6 September 1995 Summary Until recently the investigation of serological responses to mycobacteria in patients with Crohn's disease has been hindered by the considerable degree of cross-reactivity between antigens of M. paratuberculosis, and other mycobacterial subspecies. We evaluated the serological response of Crohn's disease patients to a recently identified species-specific 18 kda protease-resistant antigen corresponding to M. paratuberculosis bacterioferritin. The 18 kda antigen was purified from M. paratuberculosis as previously described. Serum was obtained from 40 patients with Crohn's disease, 15 with ulcerative colitis, 25 coeliac patients, and 21 normal blood donors. Antibody levels were measured by enzyme-linked immunosorbent assay (ELISA), with anti-human IgA and IgG alkaline phosphatase conjugate. Antibody titres were expressed as the dilution giving 1/3 of the plateau binding value of a standard positive serum (MT/3). Disease activity of the Crohn's disease cases was assessed using the Harvey-Bradshaw index. There was no statistically significant elevation of the mean IgG or IgA MT/3 titres of Crohn's disease patients over controls. No patients had antibody titres greater than two standard deviations above the mean control MT/3 titres, and there was no significant correlation between Crohn's disease activity and level of antibody titres. These findings make it unlikely that M. paratuberculosis is of primary pathogenic importance in Crohn's disease. Introduction There is continued interest in the role of mycobacteria in Crohn's disease, particularly following the recent isolation of Mycobacterium paratuberculosis 1 from Crohn's disease tissues and the identification of the M. paratuberculosis specific DNA sequence IS900 in 13 to 65% of Crohn's disease samples. 2 ' 3 Debate continues as to the pathogenic role of these organisms. Efforts to detect a specific immune response to M. paratuberculosis have been hindered by the extensive cross-reactivity between mycobacterial species. 4 ' 5 Recent work has shown that a protease-resistant M. paratuberculosis extract antigen (Antigen D) gives 100% sensitivity and 90% specificity for bovine paratuberculosis. 6 Elsaghier et al. used Western blotting to characterize the antibody response in M. avium ss. avium (MAA) and M. paratuberculosis infected BALB/c and C57/BL6 mice. 7 They analysed three major antigens and found that a 24 kda antigen was specific to M. paratuberculosis 203 infection, a 38 kda antigen was specific to MAA B2 infection and an 18 kda proteinaseresistant antigen was specific to MAA A2- and M. paratuberculosis 203-infected BALB/c mice. 7 The same workers went on to study responses to these three M. paratuberculosis-speaftc antigens in inflammatory bowel disease patients and found that 84% of Crohn's disease patients had IgG antibodies to at least one of the antigens. 8 We therefore investigated the IgG and IgA response of a different population of inflammatory bowel disease patients to the most discriminating of the antigens, namely the 18 kda protease-resistant antigen corresponding to mycobacterial bacterioferritin, to which positive titres were detected in 53% of Address correspondence to Dr R.N. Allan, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH Oxford University Press 1996

2 218 R.S. Walmsley etal. Crohn's disease patients in the original study. In addition, we attempted to correlate further the results with disease activity. log 10 MT/3 dilutions of Crohn's disease patients were compared to the Harvey-Bradshaw scores by Spearman's rank correlation. Methods Antigen The 18kDa bacterioferritin (donated by Dr A. Elsaghier and Professor Y. Ivanyi, Tuberculosis and Related Infections Unit, Royal Postgraduate Medical School, London) was purified from M. tuberculosis soluble extract. 8 Patients Sera were collected prospectively from 40 Crohn's disease patients. An assessment of their disease activity was made using the Harvey-Bradshaw simple index 9 on the same day. Sera were also collected from 15 patients with ulcerative colitis, 25 with coeliac disease, and 21 healthy blood donors who acted as controls. ELISA The ELISA was performed as originally described by Elsaghier et a/. 8. Ninety-six-well microtitre plates were coated with 50 jal/well of antigen (concentration of 2 ig/ml), washed with 0.05% PBS Tween and then blocked with 3% dried milk in 0.05% Tween. Plates were incubated at 37 C for 1 h, washed with PBS Tween and 100 (il/well of diluted serum (in 2% dried milk) at dilutions of 1 :10, 1 :40, 1 :160 and 1 :640, duplicating each sample. After incubation at 37 C for a further hour and further washing, 100 jj/well of goat antihuman IgA or IgG alkaline phosphatase conjugate (Sigma), diluted 1 in 1000 in dried milk was added. Incubation for 1 h at 37 C was followed by repeated washing and addition of 100 il/well of substrate (diethanolamine, Sigma, 1 mg/ml) and the reaction was stopped with 50 (xl 4 M NaOH. Optical density was read at 450 nm. All samples were coded, and analyses were performed blinded. Statistical analysis The optical densities of the dilutions for each sample were plotted against the log 10 dilution, and results were expressed as the log 10 dilution giving 1/3 of the plateau binding value of a standard positive serum (MT/3). Sera were classified as positive if they exceeded the mean log 10 MT/3 dilution of the control group by more than two standard deviations. The mean log 10 MT/3 dilutions were compared using a two-tailed unpaired t-test and mean Results Antibody titres to 18 kda antigen The mean log 10 MT/3 antibody titres and standard deviations for the four patient groups of Crohn's disease, ulcerative colitis, coeliac disease and normal controls, for both IgA and IgG, are summarized in Table 1. There was no significant elevation in mean MT/3 antibody titres for Crohn's disease, ulcerative colitis or coeliac patients over normal controls for either antibody isotype. There were also no patients with antibody titres greater than two standard deviations above the mean control MT/3 titres. Figures 1 and 2 show the distribution and mean values for the IgA and IgG assays. Disease activity There was no significant correlation between disease activity of the Crohn's disease patients and either IgA or IgG antibody titres. Discussion The similar clinical features of Johne's disease of ruminants, caused by Mycobacterium paratuberculosis, and Crohn's disease in humans have been appreciated for over 80 years. 10 There has been little histological evidence to identify these bacteria in Crohn's disease tissues, and it has proved extremely difficult to culture them from such tissues. It has been suggested that they might exist as spheroblast cell-wall-deficient forms. 11 Chiodini and his colleagues isolated M. paratuberculosis from a number of patients with Crohn's disease 12 and subsequently induced a granulomatous enteritis in goats following enteral inoculation with one of these strains. 13 More recently, the use of the polymerase chain reaction to identify the presence of mycobacterial species in human tissues has led to a resurgence of interest in the possible pathogenic role of M. paratuberculosis. Table 1 Crohn's disease Ulcerative colitis Coeliac disease Normal controls Mean (SD) log 10 MT/3 antibody titres IgG 2.33 (0.445) 2.27 (0.405) 2.49 (0.471) 2.65 (0.425) IgA 1.71 (0.536) 1.75 (0.49) 1.42 (0.452) 1.84(0.629)

3 M. paratuberculosis and Crohn's disease r S 2 f Xi 1 «1.5 s o o> Crohn's Disease Ulcerative Colitis Coeliac Controls Figure 1. IgA antibodies to 18 kda MAP antigen. Lines show mean values (n = 39 CD, 17 UC, 25 coeliac, 14 controls). O.O "" 3 - t i t! «1.5 - «1 - Crohn's Disease Ulcerative Colitis Coeliac Controls Figure 2. IgG antibodies to 18 kda MAP antigen. Lines show mean values (n = 33 CD, 12 UC, 21 coeliac, 20 controls). Groups in the UK, USA and Japan have found copies of the highly specific M. paratuberculosis DNA insertion sequence IS900 in 10% to 65% of colonic and lymph-node tissues from Crohn's disease patients. 2 ' 3 ' 14 ' 15 Other groups have been unable to confirm these findings 16 ' 17 but have been criticized for using fixed tissues rather than frozen specimens. It is not uncommon, however, to find evidence of other mycobacteria in addition to M. paratuberculosis in the same tissue samples, as well as in previously unidentified cultures from Crohn's patients. 14 ' 18 It is therefore difficult to determine whether M. paratuberculosis is acting as primary pathogens or, like other mycobacteria, merely non-pathogenic commensals. This question could be resolved by the demonstration of an immune response specifically aimed at M. paratuberculosis. Attempts to elicit cell-mediated immune responses against mycobacteria have suggested a generalized activation of lymphocytes derived from gut-associated lymphoid tissue or peripheral blood, 19 ' 20 with no specific response to M. paratuberculosis. Serological studies have been

4 220 R.S. Walmsleyetal. hindered by cross-reactivity of antigens between subspecies of mycobacteria 4 giving rise to conflicting results. Thayer etal., although reporting a statistically significantly higher antibody level to the protoplasmic antigen of M. paratuberculosis in Crohn's disease compared with controls, also found 52.5% crossreactivity between M. paratuberculosis and M. kansasii, and 39% with M. tuberculosis. 4 Bruneilo and colleagues also confirmed this cross reactivity, with 50% of patients suffering from active tuberculosis giving positive ELISA results against a protoplasmic M. paratuberculosis antigen, but found that only 4% of Crohn's disease patients gave a positive result. 21 Other workers have found no evidence of sensitization to M. paratuberculosis antigens in Crohn's disease 22 ' 23 although Stainsby ef al. found high antibody titres to a range of mycobacteria in controls as well as inflammatory bowel disease patients, suggesting contact with environmental mycobacteria in all populations. 5 Our study does not confirm the earlier findings of Elsaghier et al. using the same 18 kda antigen which seems to be species-specific and immunogenic in paratuberculosis-infected sheep 6 and experimentally infected BALB/c mice. 7 They found highly significant differences in antibody titres between Crohn's disease patients and controls but no difference between ulcerative colitics and controls with positive antibody titres (i.e. greater than 2 SD above the mean control antibody titre) in 53% of Crohn's cases and 10% of ulcerative colitis cases. 8 One explanation of the difference in the results of the two studies is that the antigen is not species-specific and that the Italian population from which their Crohn's disease patients were selected had been exposed to different environmental mycobacteria. We also tested the IgG response of a patient with Crohn's disease for 5 years who had undergone three operations including a terminal ileal resection prior to his referral. Histological review of further resection specimens revealed a number of acid-fast bacilli. The tissue was not sent for culture. His mean log 10 MT/3 titre was 2.18, well within the 95% confidence interval for the control population. This implies that, if the 18 kda antigen is highly species-specific, either M. paratuberculosis is not the organism involved in his disease, or else it is not the pathogenic cause of Crohn's disease. The patient is now in clinical remission after 12 months of anti-tuberculosis therapy. The negative results of recent trials of antituberculous therapy in Crohn's disease 24 and the results of this study suggest that M. paratuberculosis is unlikely to be the causative agent of Crohn's disease and is more likely to be a non-pathogenic commensal. Acknowledgements We thank Dr A. Elsaghier and Professor Y. Ivanyi of the Tuberculosis and Related Infections Unit, Royal Postgraduate Medical School, London, for donation of the 18 kda antigen. References 1. Wall S, Kunze ZM, Saboor S, Soufleri I, Seechurn P, Chiodini R. 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: Fidler HM, Thurrell W, Johnson NM, Rook CA, McFadden JJ. Specific detection of Mycobacterium paratuberculosis DNA associated with granulomatous tissue in Crohn's disease. Gut 1994; 35: Sanderson JD, Moss MT, Tizard ML, Hermon- Taylor J. Mycobacterium paratuberculosis DNA in Crohn's disease tissue. GuM992; 33: Thayer WR, Coutu JA, Chiodini RJ, Van Kruiningen HJ, Merkal RS. Possible role of mycobacteria in inflammatory bowel disease. Mycobacterial antibodies in Crohn's disease. Dig Dis Sci 1984; 29(12): Stainsby KJ, Lowes JR, Allan RN, Ibbotson JP. Antibodies to Mycobacterium paratuberculosis and nine species of environmental mycobacteria in Crohn's disease and control subjects. Guf\ 993; 34: Sugden EA, Brooks BW, Young NM, etal. Chromatographic purification and characterization of antigens A and D from Mycobacterium paratuberculosis and their use in enzyme linked immunosorbent assays for diagnosis of paratuberculosis in sheep. J Clin Microbiol 1991; 29(8): Elsaghier A, Nolan A, Allen B, Ivanyi J. Distinctive western blot antibody patterns induced by infection of mice with individual strains of the Mycobacterium avium complex. Immunology 1992; 76: Elsaghier A, Prantera C, Moreno C, Ivanyi J. Antibodies to Mycobacterium paratuberculosis-speofic protein antigens in Crohn's disease. Clin Exp Immunol 1992; 90: Harvey RF, Bradshaw JM. A simple index of Crohn's disease activity. /.ancem980; 1: Dalziel TK. Chronic interstitial enteritis. BrMedJ 1913; 2: Burnham WR, Lennard-Jones JE, Stanford JL, Bird RG. Mycobacterium as a possible cause of inflammatory bowel disease. Lancef\97%; 2: Chiodini RJ, Van Kruiningen HJ, Thayer WR, Merkal RS, Coutu JA. Possible role of mycobacteria in inflammatory bowel disease. An unclassified mycobacterium species isolated from patients with Crohn's disease. Dig Dis Sci 1984; 29(12): Van Kruiningen HJ, Chiodini RJ, Thayer WR, Coutu JA, Merkal RS. Experimental disease in infant goats induced by a mycobacterium isolated from a patient with Crohn's disease. A preliminary report. Dig Dis Sci 1986; 31: Yokoyama Y, Seunaga K, Okazaki K, Yamamoto Y. Mycobacteria in the intestine of patients with inflammatory bowel disease. Gastroenterology 1994; 106:A794.

5 M. paratuberculosis and Crohn's disease Ibbotson JP, Fidler H, Chahal H, Mussaddeq Y, Allan RN, McFadden JJ. Detection by polymerase chain reaction of Mycobacterium paratuberculosis specific DNA in Crohn's disease tissues. CuM 994; 35:S Rosenberg WMC, Bell Jl, Jewell DP. Mycobacterium paratuberculosis DNA cannot be detected in Crohn's disease tissues. Castroenterology 1991; 100:A Rowbotham DS, Mapstone NP, Trejdosiewicz LK, Howdle PD, Quirke P. Mycobacterium paratuberculosis cannot be detected by the polymerise chain reaction in Crohn's disease tissue. CuM993; 34:S Moss MT, Sanderson JD, Tizard ML, Hermon-Taylor J, el-zaatari FA, Graham DY. Polymerase chain reaction detection of Mycobacterium paratuberculosis and Mycobacterium avium subsp silvaticum in long term cultures from Crohn's disease and control tissues. CuM992; 33: Ibbotson JP, Lowes JR, Chahal H, Caston JSH, Life P, Kumararatne DS, etal. Mucosal cell-mediated immunity to mycobacterial, enterobacterial and other microbial antigens in inflammatory bowel disease. Clin Exp Immunol 1992; 87: Seldenrijk CA, Drexhage HA, Meuwissen SCM, Meijer CJLM. T-cellular immune reactions (in macrophage inhibition factor assay) against Mycobacterium paratuberculosis, Mycobacterium kansasii, Mycobacterium tuberculosis, Mycobacterium avium in patients with chronic inflammatory bowel disease. Cuf 1990; 31: Brunello F, Pera A, Martini S, etal. Antibodies to Mycobacterium paratuberculosis in patients with Crohn's disease. Dig Dis Sci 1991; 36: Tanaka K, Wilks M, Coates PJ, Farthing MJC, Walker-Smith JA, Tabaqchali S. Mycobacterium paratuberculosis and Crohn's disease. Cuf 1991; 32: Cho S, Brennan PJ, Korelitz Bl, Graham DY. Mycobacterial aetiology of Crohn's disease: serologic study using common mycobacterial antigens and a species specific glycolipid antigen from Mycobacterium paratuberculosis. Cut 1986; 27: Swift GL, Srivastava ED, Stone R, ef al. Controlled trial of anti-tuberculous chemotherapy for two years in Crohn's disease. CuM994; 35:363-8.

6