50 S ribosomal subunits from Escherichia COG were incubated with an endonuclease specific for single-stranded nucleic acid. After incubation the sedimentation coefficient remained unaltered but gel filtration showed that loo/, of the nucleic acid in the ribosome had been degraded to small fragments. When the enzyme-treated native subunits were subsequently unfolded in EDTA their sedimentation properties were similar to controls. Thermal denaturation in the presence of formaldehyde produced small fragments of RNA with a sedimentation coefficient of 2s.Subunits which had fist been unfolded in EDTA and then incubated with enzyme were rapidly degraded to small fragments with a sedimentation coefficient of 2 S. The degraded fragments had 20 Ol0 hyperchromism on thermal denaturation which showed that the secondary structure of the RNA had remained intact. These experiments were interpreted in terms of the hairpin-loop model of RNA. They show that on the surface of the native subunit there is no interhelical single-stranded RNA which is accessible to enzyme whereas there is a significant amount of intrahelical single-stranded RNA. In the unfolded subunit the interhelical regions become exposed to attack by the enzyme.A digest of unfolded subunits was separated by gel atration on Sephadex 6-100 into three fractions. Fraction I, which eluted at the exclusion volume of the column, was mainly aggregated protein. Fraction 11, which contained 65 of the ultraviolet absorbing material and no protein, was hyperchromic and the circular dichroic spectrum was consistent with this fraction having double-helical secondary structure. Fraction I1 moved as a single, sharp boundary on sedimentation with sz0 = 1.7 S and had a molecular weight of 15700. Fraction 111, which contained 3501, of the ultraviolet absorbing material and no protein, was low molecular weight nucleotide with no secondary structure.Ribosomal RNA is thought to consist of short regions of double-helical "hairpin loops" connected by single-stranded interhelical regions [l -31. I n the ribosome, the size of the hairpins, the size of the singlestranded regions, the position of attachment of the proteins, and the way in which the ribonucleoprotein strand is folded in the native ribosome is largely unknown, although various models have been proposed [4-61. Useful information about the tertiary structure of spherical viruses and ribosomes has been obtained by studying the susceptibility of the native nucleoprotein to enzymic digestion [7-lo]. These studies have usually used pancreatic RNAse. However, degradation of ribosomal RNA with this enzyme leads eventually to a loss of the secondary structure which introduces ambiguities in the interpretation of the results, especially in the case of ribosomes [18]. A n attempt has been made to gain more precise information about the arrangement of RNA in the 50s subunit by studying the action of an endonuclease, highly specific for single-stranded RNA, on native subunits and subunits which have been unfolded in EDTA [Ill.
Analytical...