Recent experiments have presented evidence that Watson-Crick hydrogen bonds
in a base pair are not absolute requirements for efficient synthesis of th
at pair by DNA polymerase enzymes. Here we examine quantitative steady-stat
e kinetic data from several published studies involving poorly hydrogen-bon
ding UNA base analogues and adducts, and analyze the results in terms of so
lvation, hydrogen bonding, and steric effects. We propose a mechanism that
can explain the surprising lack of hydrogen-bonding requirement accompanied
by significant selectivity in pairing. This hypothesis makes use of steric
matching, enforced both by the tightly confined polymerase active site and
by the DNA backbone, as a chief factor determining nucleotide selection du
ring DNA synthesis. The results also suggest that hydrogen bonds from bases
to water (solvation) may be important in increasing the effective size of
DNA bases, which may help prevent misinsertion of small bases opposite each
other. (C) 1998 John Wiley & Sons, Inc.