Two susceptible (Bcg') mouse strains, BALB/c and C57BL/6, were compared by Western blot (immunoblot) analysis for their immunoglobulin G response to 14-day-old BCG culture filtrate (CF) following intravenous infection with live Mycobacterium bovis BCG. The two strains demonstrated a completely different antibody repertoire. BALB/c antibodies were directed against a wide range of CF antigens between 20 and about 100 kilodaltons (kDa), with a preferential recognition of the 65-kDa heat shock protein and the 32-kDa fibronectin-binding protein. C57BL/6 sera, on the other hand, showed a much more restricted antibody pattern, almost exclusively directed against three antigens with estimated molecular sizes of 37, 38, and 40 kDa. Whereas the 37-and 38-kDa antigens were also recognized by BALB/c mice, the 40-kDa antigen was very intensely stained by C57BL/6 sera only. F1 mice had the restricted antibody pattern of C57BL/6 after one injection of BCG and had a hybrid BALB/c-C57BL/6 phenotype following a boost injection of BCG 2 months after the initial infection. Analysis of seven recombinant inbred strains derived from the BALB/c x C57BL/6 cross and of congenic mice differing in major histocompatibility complex-coding chromosome 17 fragments suggests that a gene in the K-IA region of the H-2 locus is associated with the preferential recognition of certain CF antigens. Inoculation with the same dose of killed BCG failed to elicit an antibody response to these filtrate antigens.

351-366, 1987). Spleen cells from BCG-sensitized mice produce significant amounts of gamma interferon (IFN-,y) in response to this P32 protein.

C57BL/10 and C57BL/6 mice (H-2b); B10 congenic mice with f, k, p, q, r, and s H-2 haplotypes; B10 mice with recombinant g2, o2, a, h2, h4, i5, and bq1 H-2 haplotypes; and B6 mice with major histocompatibility complex (MHC) mutant bm1 and bm13 (class I) and bm12 (class II) haplotypes were infected intravenously with 4 x 10(6) CFU of live Mycobacterium bovis BCG and examined by Western immunoblot analysis for serum antibodies against BCG culture filtrate antigens, following a boost injection with live BCG or with BCG culture filtrate. Parental B10 and B6 mice reacted very intensely with three culture filtrate protein bands with estimated molecular masses of 37, 38, and 40 kDa. Response against the 40-kDa protein was stronger following a boost injection with live BCG than following a boost with culture filtrate. Sera from mice with f, p, i5, bm1, and bm13 haplotypes reacted strongly, with both the 37-38- and 40-kDa antigens, and sera from mice with q and bq1 haplotypes showed a somewhat weaker reaction. Sera from mice with r, s, and bm12 haplotypes reacted against the 37-38-kDa antigen but not against the 40-kDa antigen, and sera from mice with the h2 haplotype reacted only with the 40-kDa antigen but not with the 37-38-kDa antigen. Sera from mice with the k, g2, o2, a, and h4 haplotypes showed, at most, a very weak reaction with the 37-38- and 40-kDa antigens. These results demonstrate that MHC genes profoundly affect the antibody repertoire used against culture filtrate antigens in mice infected with live M. bovis BCG. In particular, as shown in mice with the recombinant H-2 haplotype and in class II mutant bm12 mice, the I-A heterodimer controls the recognition of the immunodominant 40-kDa antigen. By using crossed immunoelectrophoresis, this 40-kDa antigen was identified as antigen 88 according to the reference system of Closs et al. for BCG antigens.

One liter of strongly Australia antigen positive plasma was subjected to a modified Cohn fractionation and samples from the various steps of the process were assayed quantitatively for Australia antigen. More than 85% of the original quantity of Australia antigen was recovered from fraction IV. Intermediate amounts, about 2%, were found in fraction III and in the washing fluids after the fibrinogen wash. The final fibrinogen product had no detectable Australia antigen. About 1 % of the antigen was recovered from the albumin. γ-Globulin before treatment with DEAE-Sephadex had a low content, about 0.1% of the original amount of antigen, whereas DEAE-Sephadex treated material (final gammaglobulin) was free of demonstrable Au-antigen. The findings are discussed in relation to clinical reports on hepatitis risks connected with the various products.

No abstract