Refocusing of B-cell responses following a single amino acid substitution in an antigen

Abstract

Intranasal immunization of BALB/c strain mice was carried out using baculovirus-derived human chorionic gonadotrophin (hCG) beta-chain, together with Escherichia coli heat-labile enterotoxin. Gonadotrophin-reactive immunoglobulin A (IgA) was induced in a remote mucosal site, the lung, in addition to a systemic IgG response. The extensive sequence homology with luteinizing hormone (LH) results in the production of LH cross-reactive antibodies when holo-hCG is used as an immunogen. In contrast to wild-type hCGbeta, a mutated hCGbeta-chain containing an arginine to glutamic acid substitution at position 68 did not induce the production of antibodies which cross-react with LH. Furthermore, the epitopes utilized in the B-cell response to the mutated hCGbeta shifted away from the immunodominant region of the parent wild-type molecule towards epitopes within the normally weakly immunogenic C terminus. This shift in epitope usage was also seen following intramuscular immunization of rabbits. Thus, a single amino acid change, which does not disrupt the overall structure of the molecule, refocuses the immune response away from a disadvantageous cross-reactive epitope region and towards a normally weakly immunogenic but antigen-unique area. Similar mutational strategies for epitope-refocusing may be applicable to other vaccine candidate molecules.

Figure 1

A computer-generated space-filled model of hCG highlighting residue hCGβR68. The α-subunit is indicated in dark grey and the β-subunit is shown in pale grey with the N terminus of the β-subunit and residue βR68 indicated in black. Human CG is a member of the cysteine-knot family in which a triple-loop structure for each subunit is held together by three disulphide bonds in a knot-like conformation, with loops 1 and 3 comprising one end of each subunit. The α- and β-subunits are oriented opposite to each other such that loops 1 and 3 of each subunit form either end of a cigar-like structure which is stabilized by a fourth short loop structure of the β-subunit, the seatbelt., Note that the CTP (β:110–145) is absent from the structure.

Figure 2

Analysis of the conformation of BAChCGβ and BAChCGβ (R68E) with different hCG-specific mAb: 1 µg/ml of pregnancy-urine-derived holo-hCG, pregnancy-urine-derived hCGβ, BAChCGβ, BAChCGβ(R68E), synthetic CTP or pituitary-derived LH were coated on to ELISA plates and tested for reactivity with the hCG-specific and irrelevant (anti-TT) mAb.

Figure 3

Serum IgG responses of BALB/c mice (n = 5) immunized intranasally with LT mixed non-covalently with holo-hCG, hCGβ, BAChCGβ, or BAChCGβ(R68E). The antigen-specific IgG in a 1:100 dilution of the sera was determined in an ELISA on plates coated with hCG holohormone, LH, hCGβ, BAChCGβ, BAChCGβ(R68E) and CTP. Each symbol indicates the result for one serum sample read at A₄₀₅ with the mean values indicated with a bar. Mice immunized with LT alone gave an OD of 0·2, indicated by the continuous line.

Figure 4

Specificity of antisera (diluted 1:500) from rabbits immunized with holo-hCG (rabbit hCG-TT), BAChCGβ (rabbit BAChCGβ-ΤΤ), or BAChCGβ(R68Ε) (two rabbits, BAChCGβ (R68E)) covalently coupled to TT. The antigen-specific IgG responses were determined in an ELISA on a plate coated with CTP or LH. Each bar represents serum from one rabbit read at A₄₀₅.

Figure 5

Secretory IgA responses in lung lavages of BALB/c mice immunized intranasally with LT or LT mixed with holo-hCG, hCGβ, BAChCGβ, or BAChCGβ(R68E). The lavages were centrifuged to separate mucus before the total IgA was determined using ELISA. Samples were subsequently examined for specific antigen recognition using ELISA. Each symbol represents an individual mouse with the result for a 1:5 dilution of the lung lavage measured at A₄₀₅ normalized to their total IgA content. The mean values are indicated by the bars. Lung lavages from mice immunized with LT alone gave a background value of 0·11 units, indicated by the continuous line.