MAGNESIUM AND CATALYTIC RNA 277
seems to be important because it is held in position by four hydrogen bonds
to G 8 and A 9 and stacks onto the G 10.1 – C 11.1 pair. However, the G12X variant
is not rescued by addition of sparingly soluble guanine and only modestly
rescued by a more soluble analog, 7 - deazaguanine (see Figure 6.14A ).
Authors of reference 42 presented the following conclusions from their
study of hammerhead bases using base rescue techniques: (1) Each of the
groups on the base - pairing face of C 3 forms interactions important for cataly-
sis; (2) comparison to the ground - state X - ray crystallographic structure (PDB:
1MME) indicates that these interactions develop only in the transition state;
(3) bases at A 9 , A 13 , and A 14 provide stacking interactions stabilizing both the
ground state (solid - state crystal structure) and transition state (solution base
rescue experiments); and (4) enhanced base rescue at position A13 by a meth-
ylated base (2 - methyladenine) may indicate the importance of hydrophobic
interactions for transition state stabilization. The reference 42 authors believe
that the base rescue technique can provide effi cient, inexpensive testing of
structure – function relationships in hammerheads because only one RNA
derivative per position needs to be prepared, and commercially available bases
and base derivatives can then be tested for rescue. They found in the two
studies described here (references 41 and 42 ) that only fi ve of the 14 tested
abasic hammerhead variants could be rescued by exogenous base addition.
Additional limitations of the base rescue technique include the diffi culty in
attaining saturation with poorly soluble exogenous bases and the possibility
of observing new non - wild - type binding modes with added bases that would
not be found for the base originally removed.
Scott and Uhlenbeck published a hammerhead X - ray crystallographic
structure in 1998 (PDB: 359D)^43 that showed a terbium(III), Tb(III), ion com-
peting with a particular Mg 2+ found to bind in the same location in the PDB:
301D crystal described above. The Tb(III) ion interacted with conserved resi-
dues guanosine, G 5 , and adenosine, A 6 , in the catalytic core in a similar manner
to the magnesium ion in the PDB: 301D structure. The site is approximately
10 Å from the cleavage site in the hammerhead catalytic core, in the so - called
uridine turn comprised of residues C 3 , U 4 , G 5 , and A 6. The Tb(III) ion appeared
to interact with the base - pairing face of G 5 and was not within binding distance
of any phosphate residues. The closest contact is from N 1 of G 5 to the Tb(III)
at 3.8 Å. Herschlag ’ s experiments that removed the G 5 base, described above,
confi rmed that this residue was essential for catalysis.^41 Other experiments by
the Scott group in solution showed that if the Tb(III) ion was added after
substrate cleavage had begun in the presence of magnesium ion, cleavage
stopped. If higher concentrations of magnesium ion were then added, cleavage
resumed. This experiment could be duplicated for the HH16, HH α 1 and HH8
hammerhead constructs. The crystal structure reported as PDB: 359D in refer-
ence 43 used the same construct as had been used in previous hammerhead
solid - state structures, that of RNA 6 (see Figure 6.10 ).
The reference 43 researchers put forth two theories to explain their results,
both assuming that Tb(III), binding more tightly to the hammerhead construct,