Abstract
Antimitochondrial antibodies (AMA) are the serum hallmark of primary biliary cirrhosis (PBC). However, AMA-positivity can be found in non-PBC sera when lower dilutions are used, thus raising issues about the specificity and sensitivity of the test. AMA reacts primarily with the lipoylated domains of pyruvate dehydrogenase-E2 (PDC-E2) which is highly conserved across species, including bacteria. We studied 77 serum samples, including 24 from patients with anti-PDC-E2-positive PBC and 53 controls (16 with autoimmune hepatitis (AIH), 10 with primary sclerosing cholangitis (PSC), and 27 healthy individuals) for their reactivities at serial dilutions (1:10, 1:20 and 1:40) against Escherichia coli DH5 alpha lysate overexpressing human PDC-E2 using immunoblotting (IB). A murine anti-human PDC-E2 monoclonal antibody (mAB) was used as control. We further studied positive sera using adsorption with a synthetic E. coli peptide sharing similarity with human PDC-E2. Finally, we verified whether a unique buffer for E. coli preparation could reduce non-specific serum reactivity. Results demonstrated that 100% of anti-PDC-E2-positive PBC and up to 38% of control sera at 1:10 dilution recognized E. coli PDC-E2 at IB while dilution tests indicated that the overall potency of PBC reactivity was 100-fold higher compared to controls. In fact, a subgroup (20-38%) of non-PBC sera were positive at low titers but lost the reactivity when absorbed with the synthetic E. coli peptide. Finally, our unique buffer reduced the reactivity of non-PBC sera as measured by ELISA. In conclusion, we demonstrated that weak cross-reactivity with E. coli PDC-E2 occurs in non-PBC sera at lower dilutions and that such reactivity is not due to AMA-positivity. The use of a specific buffer might avoid the risk of false positive AMA determinations when E. coli-expressed recombinant antigens are used.