Methods

Binding Experiments. These were generally performed with the sample of potentialpolynucleotide ligand (25-200# tg in 0.3 mL) and the column matrix equilibrated with the standard buffer, 0.01 M cacodylate/0.1 mM EDTA/0.15 M NaCl/5 mM MgCl2, pH 7.0. For certain experiments, the samples were dialyzed exhaustively against a buffer in which the pH was reduced by changing the acid/salt ratio of the cacodylate or, for more acid pH, by using acetate of the same concen-tration to give the desired pH. In other experiments, the samples were dialyzed exhaustively against a buffer containing no Mg2+ and reduced levels of Na+. Comparably equilibrated columns were always employed in these cases, but such equilibration was only attained after prolonged exposure of the columns to the modified buffer (several equilibrations of 6-12 h at room temperature). Because of this, the fraction of polynucleotide ligand bound had to be determined by calculating the difference between the amount applied to the column and the amount that did not bind (see below). Direct measurement of the amount of ligand bound by eluting it off the column under ionic conditions unfavorablefor its retention proved tedius and unwieldy due to the time required. However, the material so eluted was intact, as judged from its spectral and melting properties.

The polynucleotide sample was allowed to percolate into an affinity column and incubate for 1 h at room temperature (unless otherwise noted). The column was then eluted with three successive 1-mL aliquots of the buffer (7-10 times the column volume), which was sufficient to remove any unbound polynucleotide. The amount of polynucleotide so eluted was determined spectrophotometrically on a Zeiss PMQII instru-ment. Binding experiments were performed at least twice and the values averaged. The values for the percent bound differed by no more than 10%. Control experiments were performed on a column of underivatized agarose.