Artificial Enzymes - Artificial Enzymes - page 11

different direction, to afford essentially only 38 [125]. Molecular models were consis-
tent with this change in preferred geometry of attack on the cyclic phosphate.
We developed an assay for the cleavage of uridyluridine ( 39 , UpU) by various cat-
alysts, and used it to study the cleavage of this dimeric piece of RNA [126]. We saw that
high concentrations of imidazole buffer could catalyze this cleavage, mimicking the
high effective local concentrations of imidazole in the enzyme, and concluded that
with this buffer there was sequential base, then acid, catalysis [127]. Of course, simul-
taneous catalysis by two different buffer species by a three-body collision is unlikely
unless they are linked in the same catalyst - the enzyme or the artificial enzyme.
Further work on this simple imidazole catalysis showed that, as well as cyclization of
UpU to a cyclic phosphate, 3,5-linked UpU ( 39 ) also isomerized to the 2,5-linked iso-
mer 44 [128]. Such an isomerization requires a phosphorane intermediate ( 45) and its
pseudo-rotation so that the C-3 oxygen can become the leaving group. Our studies
showed that in this buffer-catalyzed reaction the ImH + catalyst was not simply proto-
nating the leaving group - it was first protonating the phosphate anion of the UpU so
that a phosphorane intermediate could be formed in both the cleavage and isomer-
ization. We suggested that this could also be the mechanism in the enzyme ribonu-
clease itself, but this is still a matter of controversy. In any case, we soon found evi-
dence that this mechanism - proceeding through a five-coordinate phosphorus, a
phosphorane -operated with our artificial enzyme symbolized by 41 .