The procedures developed earlier (Friedrich and Woolley, preceding paper in this journal) for probing electrostatic potential with the fluorescein label were applied to transfer RNA. By using tRNA species that contain chemically reactive bases we were able to label these bases with fluorescein derivatives and thus to 'map' the electrostatic potential around the molecule. Both the electrostatic potential and the fluorescence emission anisotropy data that were obtained at the same time could be understood in terms of the well-known, paradigmatic crystal structure of tRNAPhe. However, within the distribution of the various tRNA species, tRNA,M"' appeared to occupy an extreme position, which suggests a relation between the conformation in solution and the initiation function of this molecule. Comparison with theoretical predictions by others of the electrostatic potential map of tRNA showed agreement in respect of trends, but the values of the potentials measured were orders of magnitude lower than predicted. This we attribute primarily to solvation.We have demonstrated in the preceding paper [l] that the pK, of the fluorescein label can be used as a probe of electrostatic potential; the theoretical basis of this leads to the simple relationship y = -58.2 ApK,, where y is the potential at the position occupied by the fluorophore and ApK, is the shift in pK, of the bound fluorescein compared with a chemically identical fluorescein derivative not bound to a macromolecule. The method was applied to the 16s RNA molecule, with the fluorescein label bound at its 3' end.Here we investigate the electrostatics of the tRNA molecule. This molecule allows a much more detailed investigation, as the various tRNA species contain many chemically modified bases, at known positions, that can react specifically with various derivatives of fluorescein. Since these bases are found at different locations on the molecule, comparisons can be made and an overall picture can be built up. An advantage of tRNA is that it is less easily denatured than the larger RNA molecules and solution structures may readily be compared. Further, the electrostatics of this molecule have been investigated theoretically [2 -71 so that experimental results can be interpreted in more depth.Of the more than 200 tRNA sequences determined to date, no exception to the clover-leaf structure [8,9] has been found. The X-ray structure of tRNAPhe from yeast and that of tRNA,M"' from Escherichia coli are not greatly different [lo, 111 and it is highly probable that the yeast tRNAPhe crystal Correspondence to' P. Woolley, Kemisk Institut, Aarhus Universitet, DK-8000 k h u s , DenmarkAbbreviations. FITC, fluorescein isothiocyanate; FTSC, fluorescein thiosemicarbazide; IAFl, iodacetamidofluorescein; X, 3-(3-amino-3-carboxypropyl)-uridine; Q, (4,5-cis-dihydroxy-l-cyclopenten-3-yl aminomethyl)-7-deazoguanidine; s4U, 4-thiouridine; mams2U, 5-methylaminomethyl-2-thiouridine.