A study of secondary and tertiary solution structure of yeast tRNA(Asp) by nuclear magnetic resonance. Assignment of G.U ring NH and hydrogen-bonded base pair proton resonances.

The 270-MHz spectra of yeast tRNA(Asp) in H2O solutions containing Mg2+ show clearly resolved resonances in the region from -15 to -9.5 ppm. Resonances between -15 and -11.5 ppm from the hydrogen-bonded protons of the acceptor stem and anticodon arm decrease in intensity with increasing temperature and disappear by 75 degrees C. Simultansously, four well-resolved resonances between -11.2 and -10.3 ppm also decrease in intensity and disappear. Because of this behavior and their positions these resonances have been assigned to the four ring NH protons of G.U base pairs 5 and 30 in the acceptor stem and anticodon arm which are thereby shown not to be hydrogen bonded by normal Watson-Crick hydrogen bonds. The five G.C base pair resonances of the T psi C arm remain visible above 70 degrees C after all other resonances have disappeared. The high-temperature tRNA spectrum agrees well with that of the isolated T psi C hairpin and CCA half-molecule fragments, each of which contains the same five hydrogen-bonded proton resonances. The root-mean-square error between the observed and calculated resonance positions for the hydrogen-bonded base pair protons of these three arms is 0.19 ppm. The dihydrouridine stem is expected to have two A.U Watson-Crick base pairs and no B.C base pairs. However, it does not contribute any hydrogen-bonded resonances to the nuclear magnetic resonance (NMR) spectrum below -11.5 ppm. This suggests that even at 35 degrees C this helix is not hydrogen bonded in a normal manner. In the region below -11.4 ppm there are three additional proton resonances melting earlier than the rest which cannot be assigned to a particular helix of the cloverleaf. We suggest that these resonances arise from hydrogen-bonded protons involved in stabilizing tertiary structure.