Evidence is presented suggesting that streptomycin binds to 16S RNA or to 30S ribosomal subunits at the same topographical site located on the RNA chain. The equally bactericidal dihydrostreptomycin binds to the same site as streptomycin but with lower affinity. The effect of drug binding to 16S RNA (measured by reconstitution inhibition) is readily reversible, while that of drug binding to 30S subunits (measured by misreading) persists after removal of the drug. Binding of the drug is not a necessary and sufficient reason for killing.The earlier publication (1) that the site of streptomycin (Sm) binding to the 30S ribosomal subunit could be the 16S RNA may be criticized from several viewpoints. On the one hand, according to stoichiometry established for dihydrostreptomycin (H2Sm)-30S binding (2, 3), too much Sm was bound; on the other hand, in view of the potential of nucleic acid for binding the drug, too little Sm was bound. Furthermore, it is unexpected that Sm should have any specificity at all (for example, stable binding to 16S and no binding to 23S RNA).Experiments to answer these criticisms were designed with every attempt being made to relate binding and physiology. We report here the results of these experiments, which (1) confirm the role of 16S RNA as drug binding site; (2) demonstrate that Sm and H2Sm are quantitatively different with regard to binding (nevertheless they remain indistinguishable with respect to bactericidal action); (3) confirm that misreading is a hysteretic effect of either Sm or H2Sm and does not require the physical presence of the drug; and (4) Preparation of Ribosomal Components. 70S ribosomes were prepared by sedimentation (270,000 X g) of the DNasetreated lysate obtained by grinding frozen cells with alumina and removing cell debris. Subunits were dissociated by dialysis in low Mg buffer (B2+) followed by large scale zonal centrifugation in a Beckman TI-15 rotor. 16S RNA was isolated, by phenol extraction, directly from the 30S subunits obtained through the zonal centrifugation. The 30S subunits were absorbed on a DEAE-cellulose column; washed with 0.25 M NH4Cl (2 column volumes), and eluted with 0.5 M NH4Cl. These subunits, which fail to produce ethanol-soluble radioactivity when tested for RNase activity on [3H]poly(uridylic acid), were the source of the total 30S protein fraction, prepared by mixing with an equal volume of [4][5][6][7][8] solution. 30S subunits from the isolated 16S RNA (3H-labeled 16S RNA extracted from a [3H]uridine grown culture, was used whenever possible) and total protein components were reconstituted by incubation in RC+ buffer following the usual procedure. For details of all manipulations concerning the ribosomal components see refs. 1 and 8.The physical characteristics of each 30S subunit or 16S RNA component preparation, and of pellets obtained from each reconstitution experiment, were determined by absorbance and/or radioactivity measurements made upon each aliquot of a fractionated sucrose density gradient (5-20% sucrose in B2+) pre...