1050 Chemical Reviews, 1998, Vol. 98, No. 3
to locate specific sites on double-stranded DNA.174,175 The rapid cleavage of single-stranded DNA by BAL 31 nuclease has been interpreted as arising from facilitated diffusion.176 The backbone of RNA, like that of DNA, could allow for the one-dimensional diffusion of proteins. 177
The facilitated diffusion of a protein along RNA has been demonstrated with RNase A.178 Evidence for facilitated diffusion has been obtained using an RNA/ DNA chimera. Specifically, a uridine nucleotide is cleaved more quickly by RNase A if it is flanked by a long stretch of poly(dA) than if it is flanked by a short stretch. This advantage is lost if the salt concentration is high, as expected from a Coulombic interaction between the cationic enzyme and an anionic nucleic acid. Facilitated diffusion may enable cytotoxic homologues of RNase A (see section XII) to use the poly(A) tail of mammalian mRNA’s as a runway, leading the enzymes to the pyrimidine nucleotides in the indispensable coding region.
D. Processive Catalysis
T45A variants are distributive, as revealed by 31P NMR and order-of-addition experiments. In contrast, the cleavage of poly(A) by both variants is proces- sive. 146,147
For a substrate to be acted on processively, it must contain a repeating structural motif. Poly(C), poly- (U), and poly(A) have repeating motifs, such as a ribosyl group, phosphoryl group, and base. Yet, none of these polymers is cleaved processively by wild-type RNase A. The distributive behavior of RNase A is likely to arise from the opposing specificities of the B1 subsite (which does not bind adenine74,81) and the B2 and B3 subsites (which bind cytosine and uracil only weakly143-145). Inducing RNase A to degrade poly(A) processively requires simply changing the specificity of the B1 subsite to match that of the B2 and B3 subsites. This change results in variants that bind (at the B1 position) and cleave a polymer that can remain bound (at the B2 and B3 positions) after catalysis has occurred (Figure 4). Making RNase A into a processive enzyme effected a new paradigm: a processive enzyme has subsites, each specific for a repeating motif within a polymeric substrate. 146
“Distributive” enzymes bind a polymeric substrate, catalyze a chemical reaction, and release to solvent a polymeric product. In contrast, “processive” en- zymes bind a polymeric substrate and catalyze a series of identical chemical reactions along that polymer before releasing it to solvent. Many enzymes that catalyze the synthesis and degradation of nucleic acids do so processively.179 The cleavage of poly(C) and poly(U) by wild-type RNase A and the T45G and
Early work on the kinetics of catalysis by RNase A used substrates that were either ill-defined het- erogeneous strands of RNA (for example, “yeast RNA”180) or nucleoside 2′,3′-cyclic phosphodiesters, which are the products rather than the substrates of the germinal transphosphorylation reaction (see 181
Figure 4. Putative mechanism for the processive cleavage of poly(A) by T45G RNase A and T45A RNase A.146 B1 subsite in these variants can accommodate an adenine base.