Chemistry Reference
In-Depth Information
The unusually wide range of substrates accepted by these catalysts stands in contrast
to the narrow substrate scope of catalysts raised against transition state analogs. This
functional promiscuity has been ascribed to covalent capture of the antibody early in the
process of affinity maturation [53], obviating further refinement of the binding pocket
by somatic mutation. Consistent with this idea, the active site of a structurally charac-
terized antibody aldolase consists of a large hydrophobic pocket, over 11
A
˚
deep, con-
taining a single lysine [53]. The hydrophobic environment around the lysine accounts
for its low pK
a
(between 5.5 and 6.0). Interestingly, natural aldolases adopt a completely
different strategy to lower the pK
a
of their catalytic lysine, exploiting electrostatic rather
than hydrophobic interactions [54]. A further difference between these catalysts is the
relative scarcity of other functional groups in the antibody binding pocket. For example,
there is no obvious acid or base to facilitate proton transfers in the reaction. Sequestered
water molecules, or possibly a tyrosine or serine residue in the pocket, may be involved,
but additional structural work on antibody-ligand complexes is needed to sort this out.
Despite the absence of stereochemical information in the reactive immunogen, the
aldolase antibodies promote carbon-carbon bond formation with surprisingly high
selectivity. For instance, the enamine formed from acetone adds to the si face of var-
ious aldehydes with ee's in excess of 95% [53]. In other examples, Robinson annula-
tions have been carried out with high enantioselectivity [55], tertiary aldols and other
compounds have been successfully resolved [56], and enantiopure intermediates have
been prepared for the synthesis of various natural products [57, 58].
Reactive immunogens incorporating elements of transition state mimicry have de-
livered even more efficient catalysts. Compound
17
(Scheme 4.8), for example, con-
tains a tetrahedral sulfone to mimic the geometry of the acceptor site during C-C bond
formation. It was used to produce antibodies that accelerate the retro-aldol reaction of
18
with a k
cat
/K
m
of 3
10
5
M
-1
s
-1
and a rate acceleration over background (k
cat
/k
uncat
)of
10
8
[59]. These are impressive results for a catalyst never optimized by natural
selection.
2
Scheme 4.8 Hapten 17, designed to combine transition state mi-
micry and reactive immunization strategies, produced an aldolase
antibody (84G3) that promotes aldol reactions with typically higher
rates and selectivities than antibodies raised against 15. The retro-aldol
reaction of 18 is catalyzed with notable efficiency by this antibody.
