1054 Chemical Reviews, 1998, Vol. 98, No. 3
UpOC6H4-p-NO2 than to that of UpOC6H4-p-NO2. This expectation exists because the 2′-thiol group has pKa ) 8.2 by kinetic and thermodynamic measure- ments,240 but the 2′-hydroxyl group has pKa ) 12.5 by kinetic measurements241 and pKa ) 13.9 by thermodynamic measurements.242 Yet, RNase A does not appear to catalyze the cleavage of 2′-deoxy-2′-thio-
UpOC6H4-p-NO2.243 appears not to be nucleotides, UpA is
Likewise, 2′-deoxy-2′-thio-UpU a substrate.244 Among 2′-oxo cleaved faster by RNase A than
is UpOC6H4-p-NO2 or deoxy-2′-thio-UpA was
Accordingly, 2′- and its interac-
RNase A, the values of kcat and cleavage of this 2′-thiol nucleotide l e a s t 1 0 5 - f o l d l o w e r t h a n a r e t h o s kcat/Km for the analogue are at for the cleavage e 243 of UpA. unclear. The basis for such poor catalysis is Nonetheless, because His119 has been
identified as the acid for the cleavage reaction, it seems reasonable to put forth His12 as the base.
The rate enhancements conferred by His12 and His119 agree with those expected for general acid/ base catalysis by these residues. For example, sup- pose a water molecule were to replace the imidazole lost in the H12A and H119A variants. The rate enhancements then derived from the Brønsted equa- tion are
kwild-type k H119A
)( )( Ka + H Ka K a H3O i s 1 2 ) H i s 1 1 9 K a H 2 O ) R
where pKaHis12 ) 5.8 and pKaHis119 ) 6.2,245 and pKaH3O+ ) -1.7 and pKaH2O ) 15.7. The Brønsted equation therefore predicts that general base cataly- sis provides a 107.5-fold rate enhancement, and general acid catalysis provides a 109.5R-fold rate enhancement. Values of R and tend to be ap- proximately 0.5 for proton transfers between oxygen and nitrogen.246 Thus, the rate enhancements pre- dicted with this simple model are similar to those observed by experiment.
His119 has also been replaced by an asparagine residue.46 This substitution decreases the affinity of the enzyme for the rate-limiting transition state by 102-fold during the cleavage of poly(C) and UpA. An asparagine residue, unlike an alanine residue, can donate a hydrogen bond to the leaving group in the transition state. One interpretation of these data is that such a hydrogen bond can enhance the affinity of the enzyme for the transition state by 102-fold.
Finally, the results of experiments in imidazole buffer (but in the absence of enzyme) have been used to argue for a different role for His119 in catalysis by RNase A. Specifically, RNase A has been pro- posed to catalyze RNA cleavage via a triester mech- anism.225 In this mechanism, His119 is proposed to both protonate a nonbridging oxygen of the phos-
phate anion and deprotonate this same oxygen in a phosphorane intermediate. The evidence for and against a triester mechanism in the buffer-catalyzed cleavage of RNA has been a subject of recent re- views.247,248 Some textbooks (cf. refs 249 and 250) present the triester mechanism as the one operating in the enzymic active site. The results of at least three experiments on the enzyme itself provide direct evidence against this view. First, wild-type RNase A and the H119A variant cleave UpOC6H4-p-NO2 at the same rate.154 These data preclude the participa- tion of His119 in the formation or breakdown of a phosphorane, at least during the cleavage of UpOC6H4- p-NO2.251 Second, catalysis by RNase A has small thio effects, which are rate effects upon substitution of a nonbridging phosphoryl oxygen with sulfur.252,253 These data have been used to argue against the triester mechanism,254 although correlation of the thio effects with the chirality of the enzymic transi- tion state and considerations of the identity of the rate-limiting transition state somewhat weaken this argument.247,251 Third, kinetic isotope effect data on the cleavage of 18O-labeled UpOCH2C6H4-m-NO2 by RNase A are inconsistent with a triester mechanism. Rather, these data support a concerted mechanism in which the transition state is slightly associative.255
Why does RNase A not use the triester mechanism? In the active site of RNase A, the desolvated side chains of His12 and His119 are aligned to interact simultaneously as a base and acid with a bound, desolvated substrate (Figure 6). Such an alignment of two imidazolyl groups is implausable in imidazole buffer and improbable in an enzyme mimic. Thus, the enzyme can access a reaction coordinate that is relatively unavailable in nonenzymic systems.
Early chemical modification work suggested that Lys41 contributes to catalytic activity.256 This find- ing was confirmed when a variant in which Lys41 is replaced by an arginine residue was shown to have approximately 2% of the activity of the wild-type enzyme for C>p hydrolysis.153 These studies dem- onstrated the importance, but not the role, of Lys41 in catalysis.
The catalytic role most commonly attributed to Lys41 is to stabilize the excess negative charge that accumulates on the nonbridging phosphoryl oxygens in the transition state during RNA cleavage (Figure 7). It has been assumed that this stabilization occurs by Coulombic interactions.153,236,257,258 But, it has also been proposed that the stabilization occurs by way of a short, strong hydrogen bond involving the partial transfer of a proton from Lys41. 259
To probe the role of Lys41 in catalysis, cysteine elaboration was used to introduce nonnatural amino acid residues at position 41.155 Specifically, Lys41 was replaced by a cysteine residue, which was then alkylated with five different haloalkylamines. In the resulting enzymes, high values of kcat/Km for poly(C) cleavage correlate with low values of side chain pKa. The presence of an amidino side chain, which can donate a second hydrogen bond, does not enhance activity. An enzyme with a quaternary amino group