An approach to sequence-specific antibody proteases. The use of haptens mimicking both a transition state and a distorted ground state

Isolation of antigens and antibodies by affinity chromatography

Antibody-antigen binding constants are commonly strong enough for an effective affinity purification of antibodies (by immobilized antigens) or antigens (by immobilized antibodies) to work out a straightforward purification method. A drawback is that antibodies are large protein molecules and subject to denaturation under conditions required for the elution from the complex. Structures of antigens can vary but usually antigens are also equally subject to similar problems. The lability of the components can sometimes make the procedure sophisticated, but usually in all cases it is possible to find a satisfactory approach. In certain cases, specific interactions of the Fc part of antibodies are more facile to exploit for their purification.

Evolution of catalytic antibody repertoire in autoimmune mice

We have attempted to efficiently obtain catalytic antibodies (catAbs) with amidase/esterase activity in the expanded sequence space of the antibody repertoire. In doing so, we used an autoimmune mouse strain, MRL/lpr, that is known to produce enhanced levels of autoantibodies. We applied different types of haptens, such as, and, that are supposed to mimic the transition state of the substrate in the ester/amide hydrolysis. Among them, hapten (2) could not be used, as it was readily broken down after synthesis. Upon immunization with hapten (1), catAbs preferentially evolved in MRL/lpr mice, but this did not happen upon immunization with haptens (3) and (4). Independently, immunization to MRL/lpr mice with successfully elicited the catAbs with the ability to activate vitamin B(6) prodrugs. The common observation seen in these two cases is that most of the catAbs derived from MRL/lpr mice by hapten (1) and half of them by hapten (5) had a Lys at H95, which is at the junctional N region between the V(H) and J(H) gene segments. Despite the conservation of Lys (H95), analyses of the N-region and utilization of the D gene segment in the heavy chain gene showed that these catAbs were from several independent clones of the same family. Studies of site-directed mutagenesis suggest that, in the catAbs elicited from hapten (1), a Lys (H95) and a His (L91) are involved in the catalytic function. Both residues are known to interact with the phosphonate moiety of hapten (1). Such studies also suggest that, in the catAbs elicited from hapten (5), a Lys (H95) and a His (H35) are involved in the catalytic function. These basic amino acids seem to be important for binding to the phosphonate hapten, as they were not changed even after extensive evolution following multiple mutations. By contrast, in normal BALB/c mice, immunization of hapten (1) resulted in eliciting catAbs in lower yield and the majority were the non-catAbs, whose sequences were quite different from those of the catAbs from MRL/lpr mice. They were clonally related to one another and most of them originated from a single clone. The positions of the interacting key residues in the CDRs that interact with the phosphorus moiety strongly differ between our catAbs and other reported catAbs with esterase/amidase activity, which were elicited by the phosphonate/phosphonamidate haptens from normal mice. Further comparison of antibodies elicited by the phosphorus haptens, such as DNA, RNA, phosphocholine, and phosphotyrosine, indicated that none of them had sequence similarity in the basic amino acids and their positions in the CDRs, except for one example, which is anti-DNA antibody elicited from C3H-lpr mice. Analysis based on the classification of canonical structures of the antibodies again suggested that our catAbs derived from MRL/lpr mice belong to an unusual class that is not listed in the literature. Taken together, the above evidence suggests that the unique catalytic subsets that existed in the initial repertoire in the MRL/lpr mice could effectively be captured by the phosphonate haptens through the interaction with the Lys at H95. In the BALB/c mice, however, another noncatalytic subset with an ability to bind only to a moiety other than the phosphonate moiety alternatively evolved, because of the lowest abundance or elimination of the catalytic subsets.

DNA hydrolysis by monoclonal anti-ssDNA autoantibody BV 04-01: origins of catalytic activity

Monoclonal anti-DNA autoantibody BV 04-01 catalyzed hydrolysis of DNA in the presence of Mg2+ ions. DNA hydrolyzing activity was associated with BV 04-01 IgG, Fab, and SCA 04-01 proteins. Pronounced cleavage specificity for both ss and dsDNA was observed with efficient hydrolysis of the C-rich region of the oligonucleotide A7C7ATATAGCGCGT7 as well as preference for cleavage within CG-rich regions of double-stranded DNA. Data on specificity of ssDNA hydrolysis and kinetic data obtained from wild-type SCA 04-01 and two SCA 04-01 mutants (L32Phe and L27dHis) were used to model the catalytically active antibody site utilizing the previously resolved X-ray structure of (dT)3 liganded Fab 04-01. The resulting model suggested that BV 04-01 activates the target phosphodiester bond by induction of conformational strain. In addition, the antibody-DNA complex contained a potential Mg2+ ion coordination site composed of the L32Tyr and L27dHis amino acid side chains and a DNA 3'-phosphodiester group. Induction of strain and metal coordination could be constituents of a mechanism by which this antibody catalyzed DNA hydrolysis. Sequence data for BV 04-01 VH and VL genes suggested that the proposed catalytic antibody active site was germ-line encoded. This observation suggests the hypothesis that catalytic activity might represent an important but unspecified function of some antibody molecules.