Chemistry Reference
In-Depth Information
4.2
Basic Strategy
Antibodies are modular proteins belonging to the immunoglobulin superfamily.
They contain two heavy and two light chain polypeptides, which are assembled at
the genetic level by combining different V (= variable), D (= diversity), J (= joining),
and C (= constant) gene segments. The resulting germline repertoire encompasses
approximately 10
8
different sequences [4]. Upon immunization, individual germline
antibodies are selected on the basis of their affinity for the antigen, and then subse-
quently refined by somatic mutation and further rounds of affinity selection. In this
way, tight binding receptors for virtually any molecule can be evolved within a few
weeks or months.
At the protein level, antigen recognition is mediated by two identical binding pock-
ets formed by six peptide loops extending from the core
-barrel structure of the N-
terminal antibody domains [1]. These loops, also called complementarity determining
regions (CDRs), are highly variable with respect to length and amino acid sequence.
Combinatorial association of these hypervariable elements makes recognition of the
universe of antigens possible.
Structural studies show that antibody binding sites have the dimensions of a typical
enzyme active site [1]. Moreover, they tend to be rich in tryptophans and tyrosines;
asparagines and histidines are also frequently found [5]. These residues are useful
for constructing generic binding sites that recognize chemically diverse ligands
through hydrophobic and hydrogen-bonding interactions. Electrostatic complemen-
tarity between protein and ligand, achieved through judicious placement of charged
residues within the binding pocket, is another common feature.
Given the presence of potentially reactive amino acid side chains in the confined
space of the combining site, it is not surprising in retrospect that antibodies, like
serum albumins [6], might chemically modify the ligands they bind. In fact, stoichio-
metric reactions between antibodies and labile esters have been well known for
some time [7], and germline antibodies have been found to catalyze various other
transformations [8]. Moreover, recent work has shown that virtually any antibody
has the potential to generate toxic oxidants capable of destroying antigens [9]. In
general, however, the likelihood of finding catalysts for specific reactions of interest
in the primary immunological repertoire is low, and the activities observed tend to be
modest.
The probability of identifying antibodies with catalytic properties can be increased if
a molecule carrying chemical information about a particular reaction mechanism is
used to induce the immune response. As originally proposed by Jencks [10], a stable
analog of the transition state of the target reaction would be an ideal immunogen.
Antibodies that evolve to bind such a compound tightly would also be expected to sta-
bilize the true transition state and thus speed the conversion of substrate into product.
The feasibility of this approach is now well established: antibody catalysts for diverse
transformations, extending from hydrolytic reactions to pericyclic processes, have
been elicited in response to transition state analogs [2, 3]. These catalysts exhibit
Michaelis-Menten kinetics and achieve significant rate accelerations, usually between
b
