292 GROUP I AND II METALS IN BIOLOGICAL SYSTEMS: GROUP II
and C 1.2 in the substrate, are highly solvent accessible in the X - ray structures.
Most of these appear near the edges of the catalytic pocket and could become
solvent protected as stems I and II move toward each other.
The reference 56 authors conclude the following: (1) Protected nucleotides
in the hydroxyl radical footprinting experiments support a global tertiary
structure similar or identical to that found in X - ray crystallographic experi-
ments; (2) the protection found for residue 15.3 (U 15.3 in RNA 6 used for
crystallography) provides strong evidence that the crystal form is also present
in solution; (3) protection at U 7 seen in the hydroxyl radical experiments dis-
agrees with its solvent accessibility in crystal structures indicating a more
compact ribozyme – substrate complex form in solution; (4) 2 ′ - deoxy G 5 sub-
stitution eliminates solvent protection (and by analogy proper hammerhead
folding); (5) since folding is much faster than catalysis, it is diffi cult to know if
the structural form found with hydroxyl radical footprinting is actually on the
hammerhead native folding pathway; (6) although specifi c metal ion binding
sites are seen in hammerhead crystal structures and seemingly confi rmed by
DeRose ’ s NMR methods, the hydroxyl radical footprinting experiments indi-
cate that many different mono - and divalent cations are equally capable of
forming a solvent - protected motif; and (7) the data support the hypothesis of
a conformational change between ground and transition states; however, the
hydroxyl radical footprinting results probably only report on the most abun-
dant species in solution — that is, the ground state.
In 2003, the Scott group published two more crystal structures showing the
relationship between the hammerhead ribozyme as a nuclease — catalytic
cleavage to produce a 5 ′ - OH terminus and a 2 ′ ,3 ′ - cyclic phosphate termi-
nus — and as a ligase — essentially the cleavage reaction in reverse to produce
a phosphodiester (PDB: 1NYI and 1Q29).^58 The hammerhead ribozyme
sequence catalyzes both self - cleavage and self - ligation, depending on which is
needed biologically. The most studied hammerhead ribozymes, and all that
have been discussed thus far, have greatly favored cleavage over ligation. It is
known that engineering a variable length crosslink between two sequentially
distant but spatially close 2 ′ - oxygen atoms of residues 11.5 (in stem II) and 2.5
(in stem I) can switch the ribozyme from being a nuclease to being a ligase.^59
The hammerhead RNA begins to favor ligation when the length of the cross-
link is increased.^60
In the fi rst reference 58 structure discussed (PDB: 1NYI), the researchers
added an extra guanosine nucleotide at the 5 ′ end of stem I that possessed an
unusual 5 ′ to 5 ′ phosphodiester linkage. Concurrently, a corresponding guano-
sine (G 11.4 ) was removed from the 5 ′ end of stem II. The crystallization proce-
dure followed corresponded to that carried out by the Scott group in previous
experiments and included freeze - trapping the crystalline state by immersion
in liquid nitrogen and maintaining the crystal at 100 K during data collection.
It was hoped that the changes in hammerhead construct design would provide
more stable crystal packing and hence a higher - resolution X - ray structure than
those of 3.0 - Å resolution achieved previously. Instead, the unusually linked