Research work · 2000

pANCA—IS THIS YOUR FINAL ANSWER?

Lawrence J. Saubermann

Inflammatory Bowel Diseases

Abstract

Colonic bacteria express an ulcerative colitis pANCA-related protein epitope. Cohavy O, Bruckner D, Gordon LK, Misra R, Wei B, Eggena ME, Targan SR, Braun J. J Infect Immun 2000;68:1542-8. Approximately 60-70% of individuals with ulcerative colitis (UC) and 20-30% of Crohn's disease (CD) subjects can be demonstrated to be seropositive for perinuclear antineutrophil cytoplasmic antibody (pANCA). Cross-reactive bacterial epitopes for pANCA were evaluated in this investigation by Cohavy et al. Initially, colonic biopsies were taken from three individuals with CD and were carefully placed in different bacterial growth conditions to obtain individual colonies. Aerobic, microaerophilic, and anaerobic growing conditions were used to generate colonic bacteria isolates to test as potential antigens for previously identified pANCA-specific monoclonal antibodies (Fab 5-3 and Fab 5-2). The isolated bacteria were identified by their 16s rRNA sequence and cell lysates made for use in pANCA-related testing. Bacterial isolates were only obtained from the anaerobic culture conditions, and only one (Fab 5-3) of the two pANCA monoclonal antibodies demonstrated specific reactivity for the purposes of the experiments. Reactive bands were identified from two subjects, and included four Escherichia coli isolates with a 35 kDa band. Also, one of these subjects was identified to have a 100 kDa doublet and an associated 48 kDa species on the lysate from a Bacteroides isolate. The 35 kDa and the 100 kDa proteins were then isolated and subjected to sequence analysis. The 35 kDa protein demonstrated homology to the outer membrane porin, OmpC, of E. coli. Sequencing of the 100 kDa protein revealed a newly identified protein sequence. The authors then proceeded to test the sera of 31 UC patients and 10 healthy controls for immunoreactivity against the E. coli OmpC antigen using an enzymelinked immunosorbent assay (ELISA)-based approach. Eight UC subjects had high titer pANCA measurements and were separated out for analysis of seroreactivity for the OmpC antigen. The results demonstrated a significantly higher mean absorbance for the overall UC group versus control group; however, statistical significance of seropositivity between the groups was not reached, including the UC subjects with high titer pANCA. Comments: Although the antibody staining of pANCA is recognized in a number of different autoimmune-related disease states, the increased and reproducible incidence of this marker in UC subjects has been of great interest and debate. There are a number of potential roles for pANCA in this disorder. One possibility is that pANCA represents a nonspecific marker with increased prevalence in UC. Data to support this possibility include the findings that pANCA is not related to duration, extent, colectomy status, or disease activity (Inflammatory Bowel Diseases 1999;5:61-7). The same data argue that pANCA is probably not directly involved in the pathophysiology of UC through an antibody-mediated disease process. Alternatively, pANCA as a marker may represent an antibody that shows reactivity with an autoantigenic or bacterial antigenic epitope associated with UC. In support of this role are the findings that pANCA is associated with treatment-resistant disease, patients who may require early surgery, postsurgical pouchitis, increased incidence in family members, human leukocyte antigen (HLA) alleles, and production by colonic mucosal B cells (Inflammatory Bowel Diseases 1999;5:61-67). In addition, the enteric bacteria-associated interleukin-10 (IL-10) and T-cell receptor-a -/- (TCR-a knockout) mouse models of colitis have been demonstrated to have increased pANCA levels, as well as the preferential absorption of pANCA-positive sera by the enteric bacteria (J Clin Immunol 1998;18:153-60; Gastroenterology 1997;113:1828-35). Previously, the authors of the above manuscript used a phage display library, as well as the same serologic approach, to identify potential antigenic epitopes for their pANCA monoclonal antibodies. pANCA has been demonstrated to recognize a histone HI (with a PKKAK motif) as well as a mycobacterial homologue of histone H1 called HupB (J Autoimmun 2000;14: 83-97; Infect Immun 1999;67:6510-17). The correlation of the mycobacterial HupB appeared to be greater in the CD subjects than in the UC subjects. The authors have also recently shown that these same pANCA antibodies demonstrated reactivity to mast cells as well as other neuroendocrine cells in a number of different tissue sites (Clin Immunol 2000;94:42-50). In addition, another independent group has identified the high mobility group nonhistone chromosomal proteins, HMG1 and HMG2, as pANCA epitopes (Clin Exp Immunol 1997;107: 135-40). These proteins appear to be involved in chromatin structure and are affected by viral infections. As with the OmpC antigen, noted above, serologic testing has yielded inconclusive results to verify these antigens as uniquely associated with UC pathogenesis. The above manuscript by Cohavy et al. identifies new potential bacterial epitopes for pANCA. However, to fully interpret whether these newly described epitopes of E. coli and Bacteroides species are truly reactive and distinct for pANCA will require larger and more thorough serologic testing, including the use of “inflammatory” bowel controls rather than only “healthy” control subjects. It would also be advantageous to detect the bacteria from colonies grown from subjects with UC rather than CD, and to demonstrate specific reactivity by western blot in more than just two individuals. Another potential contributing factor may be that different bacteria colonize different regions of the colon, and further colonic sampling may indicate similar or different species involved and help to confirm the findings. In conclusion, although a number of intriguing autoantigenic and bacterial antigenic epitopes have been identified to react with pANCA, further investigations will be required to achieve a final answer as to their direct relevance to the ongoing disease process in UC.

DOI: 10.1002/ibd.3780060414