The smaller subparticle of rabbit reticulocyte ribosomes was shown to yield core-particle and split-protein fractions on treatment with 2.5 M-NH4Cl/61 mM-MgCl2. The core-particle fraction was inactive in poly(U)-directed polyphenylalanine synthesis, but activity was restored after recombination with the split-protein fraction. Optimum ionic conditions for the reconstruction of active subparticles were found to be 0.75 M-NH4Cl/19 mM-MgCl2 at 0 degrees C. Improved extents of reconstruction were obtained when the core-particles were isolated by methods that avoided pelleting. Core-particles isolated from subparticles pretreated with either proteinases or ribonucleases had diminished capacity to become re-activated.

Larger ribosomal subparticles (L-subparticles) of rabbit ribosomes were treated with either ribonucleases (I or T1) or proteinases (trypsin or chymotrypsin), and their capacity to function in poly(U)-directed polyphenylalanine synthesis and in the puromycin reaction was investigated. The effects of pretreatment of L-subparticles on the reconstruction of active subparticles from core-particles derived by treatment with 2.75 M-NH4Cl/69 mM-MgCl2 and split-protein fractions were also examined. The protein moiety of proteinase-treated L-subparticles was analysed by one-dimensional sodium dodecyl sulphate/polyacrylamide- and two-dimensional polyacrylamide-gel electrophoresis. The introduction of 16–100 scissions in the RNA moiety had no effect on the activity of the L-subparticles in polyphenylalanine synthesis, and there was no effect on the stability of L-subparticles to high-salt shock treatment and a marginal effect on the reconstruction of L-subparticles from high-salt-shock core-particles and split-protein fractions. In contrast, L-subparticles treated with low amounts of trypsin (0.56 ng of trypsin/microgram of L-subparticle) were inactive in polyphenylalanine synthesis, and their capacity to function in partial-reconstruction experiments was diminished. Activity in the puromycin reaction was increased by 70% as a result of trypsin treatment (280 ng of trypsin/microgram of L-subparticle). At least two of the acidic proteins implicated in the translocation function were not affected by trypsin treatment. Trypsin-treated L-subparticles had lost their capacity to bind the smaller ribosomal subparticle (S-subparticle). The protein(s) needed for S-subparticle binding were shown to be present in high-salt-shock cores. At least six proteins associated with the core-particles were attack during trypsin treatment of L-subparticles. An examination of L-subparticles isolated from trypsin-treated polyribosomes showed that the amount of trypsin necessary to decrease the activity of the subparticle by 50% was about twice that needed in the treatment of L-subparticles alone. The largest protein of rabbit L-subparticles (approx. 51 000 daltons) was cleaved in a stepwise fashion by trypsin to fragments of approx. 40 000 daltons. This protein was also cleaved by chymotrypsin.

Rats injected with N6-[Me-3H]trimethyl-lysine excrete in the urine five radioactively labelled metabolites. Two of these identified metabolites are carnitine and 4-trimethylammoniobutyrate. A third metabolite, identified as 5-trimethylammoniopentanoate, is not an intermediate in the biosynthesis of carnitine; the fourth and major metabolite, N2-acetyl-N6-trimethyl-lysine, is not a precursor of carnitine. The remaining metabolite (3-hydroxy-N6-trimethyl-lysine) is converted into trimethylammoniobutyrate and carnitine by rat liver slices and into trimethylammoniobutyrate by rat kidney slices. In rat liver and kidney-slice experiments, radioactivity from DL-N6-trimethyl-[1-14C]lysine and DL-N6-trimethyl-[2-14C]lysine was incorporated into N2-acetyl-N6-trimethyl-lysine and 3-hydroxy-N6-trimethyl-lysine, but not into trimethylammoniobutyrate or carnitine. A procedure was devised to purify milligram quantities of 3-hydroxy-N6-trimethyl-lysine from the urine of rats injected chronically with N6-trimethyl-lysine (100 mg/kg body wt. per day). The structure of 3-hydroxy-N6-trimethyl-lysine was confirmed chemically and by nuclear-magnetic-resonance spectrometry [Novak, Swift & Hoppel (1980) Biochem. J. 188, 521–527]. The sequence for carnitine biosynthesis in liver is: N6-trimethyl-lysine leads to 3-hydryxy-N6-trimethyl-lysine leads to leads to 4-trimethylammoniobutyrate leads to carnitine.

The extent of homology between the nucleotide sequence of L-rRNA (the major RNA component of the larger ribosomal subparticle) of a lower eukaryote (Neurospora crassa) and an amphibian (Xenopus laevis) was investigated by utilizing rDNA (DNA coding for rRNA) of X. laevis cloned in plasmids pMB9 and pML2, and rDNA of N. crassa cloned in bacteriophage lambda. Hybridization studies revealed that sequences common to both N. crassa and X. laevis L-rRNA comprise a total of approx. 1050 /+- 200 nucleotides. The thermal stability of the X. laevis rDNA.N. crassa L-rRNA hybrid was 5 degrees C lower than that of the X. laevis rDNA.X. laevis L-rRNA duplex, indicating the presence of fewer than 10% mismatches in homologous sequences. X. laevis rDNA was analysed by means of restriction endonucleases and hybridization with 125I-labelled N. crassa L-rRNA. Most (at least 95%) of the conserved sequences were present in a 3000-base-pair fragment produced by restriction with endonucleases HindIII and BamHI. This fragment, which includes the 3′-OH terminus of the L-rRNA-coding region, was used as an adaptor in the construction of a bacteriophage-lambda recombinant. One section of the recombinant phage terminating in a HindIII-specific site was obtained from bacteriophage lambda plac5 (after restriction with endonuclease HindIII). A second section terminating in a BamHi-specific site was obtained from bacteriophage lambda 540 (after restriction with endonuclease BamHI). These two parts were joined by means of the X. laevis rDNA fragment. Further analysis of cloned rDNA by means of restriction endonucleases confirmed that conserved sequences were widely distributed throughout the 3000-base-pair fragment produced by HindIII and BamHi endonucleases. A 3400-base-pair fragment of N. crassa rDNA cloned in a bacteriophage lambda [Cox & Peden (1979) Mol. Gen. Genet. 174, 17–24] was restricted with endonucleases. The products were hybridized with 125I-labelled X. laevis L-rRNA. Conserved sequences were shown to be distributed over a range of approx. 1600–2700 base-pairs. Hence, in neither X. laevis nor N. crassa L-rRNA can be conserved sequences from a single block; instead regions of high and low (or no) homology must be intermingled. Both N. crassa rDNA and X. laevis rDNA were found to hybridize with Drosophila melanogaster L-rDNA sequences. Those rDNA fragments with sequences common to X. laevis and N. crassa L-rRNA also hybridized with D. melanogaster L-rRNA probe. Thus the same set of conserved sequences may be present in all three species.

It was shown that high-salt (2.75 M-NH4Cl/69mM-MgCl2) shock treatment at 0 degrees C of the larger subparticles (L-subparticles) of rabbit, Xenopus laevis and Neurospora crassa cytoplasmic ribosomes yielded split-protein fractions that were not only functionally equivalent but also interchangeable. Thus, although the remaining core-particles were inactive in both the puromycin reaction and in poly(U)-directed polyphenylalanine synthesis, activity was restored when they were combined with either homologous or heterologous split-protein fractions. This technique was used to prepare active hybrid L-subparticles, e.g. rabbit cores/N. crassa split-proteins, and also active hybrid ribosomes, e.g. rabbit smaller subparticle/X. laevis core-particle/rabbit split-proteins. Rabbit and X. laevis split-protein fractions labelled with 14C by reductive methylation with [14C]formaldehyde and sodium cyanoborohydride were both shown to bind to rabbit core-particles in approximate correlation with the degree of re-activation. The split-protein fractions of rabbit and X. laevis L-subparticles were analysed by two-dimensional and sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. The molecular weights (measured in sodium dodecyl sulphate gels) of the split-proteins of rabbit and X. laevis L-subparticles were found to be similar. These results demonstrate that the peptidyltransferase active centre of cytoplasmic ribosomes of eukaryotes has components that are interchangeable over a wide evolutionary range. Evidently the essential molecular architecture of the active centre is highly conserved.

1. The larger subrivosomal particles of rabbit reticulocytes retained full activity in the puromycin reaction and in poly(U)-directed polyphenylalanine synthesis after 4h at 0° C when buffered 0.5M-NH 4 Cl/10-30mM-MgCl 2 was the solvent. 2. Activity in the puromycin reaction was diminished to approx 10% after 15-30 min at 0° C when the concentration of MgCl 2 was lowered to 2mM. 3. Activity was not restored when the concentration of MgCl 2 was raised from 2mM to 10-30 mM at 0° C. However, activity was recovered as measured by both assay systems when the ribosome fraction was heated to 37° C at the higher concentrations of MgCl2. 4. Recovery of activity was noted during the course of the polyphenylalanine synthesis in 50 mM-KCl/5mM-MgCl 2 /25mM-Tris/HCl, pH 7.6, at 37° C. Re-activation was slow at 20° C and below. 5. No more than about 5% of the protein moiety of the subparticle was lost in 0.5M-NH 4 Cl on decreasing MgCl 2 concentration from 10mM to 2mM. No proteins were detected in the supernatant fractions by gel electrophoresis after ribosomes were separated by differential centrifugation. The supernatant fraction was not essential for the recovery of activity. However, at higher (e.g. 1M) concentrations of NH4Cl, proteins were split from the subparticle. 6. The loss and regain of activity found on lowering and restoring the concentration of MgCl2 at 0.5M-NH 4 Cl appears to arise from a conformational change that does not seem to be associated with a loss and regain of particular proteins. 7. A 2% decrease in E 260 was noticed when the concentration of Mg 2+ was restored, and the change in the spectrum indicated a net increase of approx. 100A-U base-pairs per subribosomal particle. 8. When the concentration of Mg 2+ was restored, s 20,w of the subparticle remained at 52± 1S until the sample was incubated at 37° C when s 20,w increased to 56 ± 1S compared with the value of 58 ± 1S for the subparticle as originally isolated.

We report the reconstruction, from a core-particle and split-protein fraction, of the larger subribosomal particle of rabbit reticulocytes. The reassembled particle was active in polyphenylalanine synthesis and in the puromycin reaction. The core-particles and split-protein fractions were obtained by treatment of the larger subparticle with salt solutions containing NH 4+ and Mg 2+ in the molar ratio 40:1 over the range 2.25-2.75 M-NH 4 Cl/56-69mM-MgCl 2 at 0° C. This treatment led to the loss of about eight proteins (approx. 17% of the protein moiety), which were found wholly or largely in the split-protein fraction as shown by two-dimensional gel electrophoresis. The core particle retained 5S rRNA and had much decreased (no more than 10% of control) ability to function in the puromycin reaction or in poly (U)-directed polyphenylalanine synthesis. Activity was recovered when the recombined core-particle and split-protein fractions were dialysed overnight at 4° C against 0.3M-NH 4 Cl/15mM-MgCl 2 /1mM-dithiothreitol/15% (v/v) glycerol/20mM-Tris/HCl, pH 7.6, and then heated for 1 h at 37° C. The recovery was 40-80% of the original activity. Raising the concentration of MgCL2 to 300 mM in 2.5 M-NH 4 CL led to the removal of seven rather than eight proteins, and the core particle remained active in the puromycin reaction. We infer that the protein retained by raising the concentration of Mg 2+ is an essential component of the peptidyltransferase centre of the ribosome.

Mg 2+ was shown to affect the conformation of rRNA over the range of 0.03-1.2M-KCl. The species studies were Escherichia coli S-rRNA and L-rRNA (the RNA moieties of the smaller and larger subribosomal particles respectively) and rabbits S-rRNA and L-rRNA. 2. The addition of Mg 2+ to rRNA in reconstitution buffer (0.35M-KCl0.01M-Tris/HCl, pH7.2) at 20° C let to an increase in bihelical secondary structure through the formation of additional (mainly A-U) base-pairs (e.g. an additional approx. 58 A-U base-pairs per molecule of E. coli S-rRNA as judged by u.v. difference spectrophotometry…

The rate of sedimentation of a particle in a sucrose solution depends on the viscosity and density of the medium. These two variables are related to the sucrose concentration and the temperature of the medium by new simple equations. These equations were used in a rapid iterative procedure that relates the distance moved by a zone in a continuous sucrose gradient to its sedimentation coefficient. It is shown by comparison with experiment that this iterative method allows the distance moved by a zone to be calculated rapidly. The method may therefore be used to optimize the separation of particles in a sucrose-gradient-centrifugation experiment. The method also allows the unknown sedimentation coefficients of several zones to be measured from a single sucrose-gradient-centrifugation experiment.

RNA was isolated from subribosomal particles of the malaria parasite Plasmodium knowlesi. The nucleotide composition (mole fraction) of the principal species was obtained (S-rRNA, 0.295A, 0.36U, 0.25G, 0.105C: L-rRNA, 0.326A, 0.31U, 0.228G, 0.144C). Ribosomal RNA was also isolated from Drosophila melanogaster. Optical properties of these A + U-rich species were measured. In all four cases analysis of the hypochromic effect revealed that adenine and uracil residues tended to form clusters along the polynucleotide chain. A substantial fraction of residues was located in bihelical regions of approx. 50% G-C base pairs or in regions of approx. 30-35% G-C base pairs. The possible evolutionary significance of these results was considered on the basis of comparison with properties of rRNA from bacteria (Escherichia coli) and a mammal (rabbit reticulocyte).

1. The c.d. (circular dichroism) of Drosophila melanogaster rRNA (42% G+C) and of G+C-rich fragments (78% G+C) obtained by partial hydrolysis of rabbit L-rRNA (the largest RNA species isolated from the large subribosomal particle) were measured and found to differ substantially. 2. To interpret these spectra a relation between c.d. of bihelical RNA and % G+C was derived, namely delta epsilonfG = AFG2+bfG+c, where deltaepsilonfG is the c.d. of RNA characterized by a mole fraction, fG, of guanine nucleotides and a, b and c are constants. 3. A frame of reference was established by studying the c.d. of a range of rRNA species, including S-rRNA (the RNA species isolated from the smaller subribosomal particle) and L-rRNA of Escherichia coli. 4. It was found for the rRNA species studied that 0.60+/-0.05 of residues appear to form bihelical secondary structure. 5. A higher helical content, 0.66+/-0.05, was found for the G+C-rich fragment of L-rRNA. The difference in the c.d. of rabbit L-rRNA and of D. melanogaster rRNA is attributable to the dependence of c.d. of the bihelical parts on %G+C. 6. The minimum in c.d. at 295 nm increases with increasing %G+C. The c.d. of rRNA was compared with that of the parent subparticle in this region of the spectrum, where high precision may be attained.

The 3′ hydroxyl end of 28S L-rRNA (major RNA species of the larger subribosomal particle) was labelled by coupling its 2-hydroxy-3-naphthoic acid hydrazine with diazotized [3H]aniline. The RNA was hydrolysed partially with ribonuclease T1 and fractionated on Sephadex G-200. The results show that a highly structured segment with 78% G+C content and a number-average molecular weight of at least 1.0×10(5)-1.8×10(5) is located at the 3′ hydroxyl end of the 28S rRNA molecule.

The spectrophotometric analysis of 45S precursor rRNA shows that it contains more G and C residues than does mature 28S or 18S rRNA. The helical content and the length of double-helical segments in 45S and 28S rRNA are similar.

1. The ability of Escherichia coli ribosomes to function in poly(U)-directed protein synthesis was measured at elevated temperatures by using thermostable supernatant factors from Bacillus stearothermophilus . The amount of polyphenylalanine synthesized at 55°C was about the same as at 37°C, but the rate of synthesis was increased approximately fivefold. At 60°C the activity of the ribosomes was halved. 2. E. coli ribosomes can sustain the loss of approx. 10% of the double-helical secondary structure of RNA without losing activity. 3. Within the active ribosome the double-helical secondary structure of the rRNA moiety is stabilized compared with isolated rRNA, as judged by enzymic hydrolysis and by measurements of E 260 . 4. The main products, over the range 0–55°C, of ribonuclease T 1 digestion of the smaller subribosomal particle of E. coli were two fragments (s 0 20,w 15S and 25.3S) of approximately one-quarter and three-quarters of the size of the intact molecule, revealing the presence of a ‘weak spot’ where intramolecular bonds appear insufficient to hold the fragments together. 5. Subribosomal particles of B. stearothermophilus were more stable to heating, by approx. 10°C, than those of E. coli , and the stabilization of double-helical secondary structure of the RNA moiety was more striking. 6. Rabbit reticulocyte ribosomes were active in poly(U)-directed protein synthesis at 45°C, and half the activity was lost after heating to 53°C. Active subribosomal particles of rabbit reticulocytes and of oocytes of Xenopus laevis , like the bacterial subribosomal particles, underwent a conformational change to a slower-sedimenting form on heating. The temperature range of the transition depended on the species. 7. Slower-sedimenting particles, whether produced by EDTA treatment or by heating, had different ‘melting’ profiles compared with active subribosomal particles, providing another indication of conformational differences. 8. Comparison of the properties of the various subribosomal particles revealed greater variation in the secondary structure of the protein moieties (judged by measurement of circular dichroism) than in the secondary structure of the RNA moieties, which appeared to have features in common.

Both rat liver and Escherichia coli rRNA in 0.1 m -sodium chloride were titrated with acid or alkali over the range pH3–7 at approx. 0°C. rRNA did not bind acid reversibly and hysteresis was observed, i.e. the plot of acid bound to rRNA against pH had the form of a loop showing that the amount of acid bound at a particular pH depended on the direction of the titration. Although the boundary curves were reproducibly followed on titration from pH7 to 3 and from pH3 to 7, points within the loop were ‘scanned’, e.g. by titration from pH7 to a point in the range pH3–4 followed by titration with alkali to pH7. It is inferred that the ‘lag’ in the release of certain bound protons is at least 1 pH unit, that at least about 9–15% of the titratable groups (adenine and cytosine residues) that are involved in this process and that the free energy dissipated in completing a cycle is approx. 4.2kJ/mol (1kcal/mol) of nucleotide involved in hysteresis. The interpretation of the ‘scanning’ curves was illustrated by means of a cycle of possible changes in the conformation of a hypothetical nucleotide sequence that allows formation of poly(A)·poly(AαH + )-like regions in acidic solutions. It is also inferred that the extent of ‘hysteresis’ might depend on the primary nucleotide sequence of rRNA as well as on secondary structure.

1. Two species of double-helical RNA isolated from mycelium of Penicillium chrysogenum were titrated with acid at 25°C and 95°C (solvent 0.1 m -sodium phosphate buffer). At 25°C denaturation occurred at about pH3. At 95°C in the denatured form cytosine residues titrated as a simple monobasic acid of p K 3.9 compared with p K ≃2.5 for the native form at 25°C. 2. On thermal denaturation in neutral and acidic solutions one species of RNA (38% rG·rC) ‘melted’ in three distinct stages, equivalent to a mixture of three species, namely one of about 25% rG·rC, another of about 33% rG·rC and a third of about 46% rG·rC: the relative proportions were 0.25:0.35:0.40. 3. On thermal denaturation in acidic solutions the increase in the fraction of ionized cytosine residues concomitant with the ‘melting’ of rG·rC base pair also affects the spectrum especially at 280nm and serves to enhance the contribution of rG·rC base pairs at this wavelength. The increment in ε (P) at 280nm on ‘melting’ an rG·rC base pair approaches 53501·mol −1 ·cm −1 depending on pH, compared with 33501·mol −1 ·cm −1 at pH7. In contrast ε (P) at 280nm is scarcely affected by ‘melting’ rA·rU base pairs or by the protonization of adenine residues. 4. Changes in the spectrum of Escherichia coli rRNA on denaturation in acidic solutions were studied to yield the mole fractions of rA·rU and rG·rC base pairs ‘melting’ at particular pH values.

Ribosomes from oocytes of Xenopus laevis possess low endogenous activity in vivo and in vitro , yet are readily stimulated by poly(U). The ease with which these ribosomes dissociate into active subparticles under conditions where polyribosomes and active monoribosomes are stable supports the view that the majority are unprogrammed.

Single ribosomes account for 90% or more of the ribosome fraction of Xenopus laevis ovaries. These ribosomes have low endogenous activity in a cell-free protein-synthesizing system but synthesize polyphenylalanine in the presence of poly U under conditions comparable with those reported for mammalian or bacterial systems. These observations agree with the view that the ribosome is deficient in mRNA.

The spectrum of RNA from the smaller and larger subparticles of rabbit reticulocyte ribosomes was studied as a function of pH, ionic strength, urea concentration and temperature. It was inferred that both RNA species form short double-helical segments of not more than about 10 base-pairs in length. Not more than about 70% of the base residues may be located in double-helical segments. RNA from the larger subparticle is richer in guanine and cytosine residues and its secondary structure is the more stable. These conclusions are based on the use of double-helical RNA from virus-like particles and of unfractionated Escherichia coli tRNA as model systems.

1. The optical rotatory dispersion and ultraviolet-absorption spectrum of ribosomal RNA in situ appear to be unchanged when the ribosome is dissociated into its RNA and protein moieties. 2. Reaction with 0·05% formaldehyde at 20° for 2hr. ‘fixes’ ribosomes so that they remain intact in 1% sodium dodecyl sulphate. 3. The RNA moiety of the ribosome undergoes a conformational change when ribosomes in 8% formaldehyde are heated at 70° for 10min. and cooled to 20°. After this treatment no double-helical character can be detected, but neither the sedimentation coefficient nor the morphology of the ribosome determined by electron microscopy is altered. 4. It is concluded that the RNA moiety of reticulocyte ribosomes is freely accessible to formaldehyde.

1. Parts of the 16s and 30s RNA species of reticulocytes are readily hydrolysed by pancreatic ribonuclease. The biological activity of the ribosomes is diminished after treatment with low concentrations of the enzyme (e.g. 1ng. of ribonuclease/2·5mg. of polyribosome fraction/ml.). A high proportion of the chain scissions are ‘hidden’ owing to the secondary structure of the RNA moiety. 2. As the concentration of ribonuclease is increased RNA is lost from the ribosome. About 20–30% of the RNA may be removed from the ribosome without altering appreciably its sedimentation coefficient or its appearance in the electron microscope. 3. The amount of RNA removed from the ribosome is not increased by raising the concentration of enzyme from about 1μg. to 2·5mg. of ribonuclease/2·5mg. of polyribosome fraction/ml., or by increasing the temperature from 0° to 30°, or by first converting the RNA moiety into a single-stranded form before exposure to ribonuclease. 4. Untreated polyribosomes aggregate at about 75°, whereas ribosomes treated with ribonuclease aggregate at about 45°. The aggregates that are found on heating ribosomes after enzymic hydrolysis contain about 40–50% of the complement of RNA of intact ribosomes. 5. From the size of the fragments of RNA isolated from RNA-depleted ribosomes it is inferred that there is one site/60–100 nucleotides that is sensitive to ribonuclease. 6. The RNA moiety of RNA-depleted ribosomes has some double-helical character as shown by the optical properties and X-ray-diffraction pattern of ribonuclease-treated ribosomes and by the ‘melting’ properties of the isolated RNA. 7. Subparticles prepared by titration with an excess of EDTA are readily hydrolysed by ribonuclease to fragments of S 20,w less than 4s, in contrast with the intact particle.

It is suggested that the location of each of the diverse ribosomal proteins in the ribosome, or in the ribosomal subparticles in their native or derived forms, is determined by the nucleotide sequence of the 16s and 23–30s RNA moieties, i.e. that the 16s and 23–30s RNA species provide a unique binding site for each species of ribosomal protein. The ways of assembling such a thread into a ribosome-like structure appear limited if the ribosome is largely stabilized by protein–protein interactions. The nucleoprotein thread was built into a structure, having the general features of a hollow cylinder, that is consistent with the known dimensions and properties of the rabbit reticulocyte ribosome. It appears possible to test the model by experiment.

The effect of exposing rabbit reticulocyte ribosomes to concentrated solutions of potassium chloride was examined by: ( a ) dialysis against 0·5 m -potassium chloride; ( b ) zone centrifugation through a sucrose gradient in 0·5 m -potassium chloride; ( c ) differential centrifugation of a solution made 0·5 m with respect to potassium chloride. The products of each treatment and their ability to support protein synthesis in a reticulocyte cell-free system, in the presence and in the absence of polyuridylic acid, were examined. The following results were found. (1) Exposing the polysomes to 0·5 m -potassium chloride was not a sufficient condition for the complete dissociation of ribosomes into subparticles; the reaction showed a concentration-dependence, implying the existence of an equilibrium between the various ribosomal species. Disturbance of the equilibrium by removing certain products, as in zone centrifuging, can lead to complete dissociation. (2) The subparticles produced by dialysis or sucrose-gradient fractionation were biologically inactive and unstable. (3) The pellet obtained by differential centrifuging consisted of subparticles, if suspended in a Mg 2+ -free buffer; addition of Mg 2+ converted about 30% of the material into heavier sedimenting species, and the preparation had 20–40% of the activity of the untreated control polysomes in the cell-free system. Addition of the 0·5 m -potassium chloride supernatant fraction resulted in further apparent reconstitution of sub-particles into ribosomes and polysomes and in a 50–100% restoration of biological activity. When both polyuridylic acid and supernatant factors were present incorporations similar to or higher than those of the control were attained.

1. The thermal denaturation of DNA from rat liver was studied spectrophotometrically. In sodium phosphate buffers denaturation led to a single-stranded form having, at 25°, about 25% of the hypochromism of the intact double helix. 2. The hypochromism of the denatured form was the same in 1m m - as in 10m m -sodium phosphate buffer and was scarcely affected by reaction with formaldehyde. The hypochromism was decreased by about 40% in the presence of 8 m -urea. 3. The hypochromism of denatured DNA at low ionic strengths was about the same as that of fragments of reticulocyte ribosomal RNA that were too short to form double-helical secondary structure and about the same as that of RNA after reaction with formaldehyde. 4. The spectrum of DNA was slightly affected by the presence of 8 m -urea or 4 m -guanidinium chloride. The differences in the spectrum of the native and denatured forms of DNA in 0·1 m -sodium phosphate buffer, in 8 m -urea–10m m -sodium phosphate buffer and in 4 m -guanidinium chloride–10m m -sodium phosphate buffer, pH7·6, were similar but not identical. 5. Denatured rat liver DNA appears to have no double-helical character at 25° in 10m m -sodium phosphate buffer, pH7·6; increasing the buffer concentration to 0·1 m leads to a more compact form in which about 40% of the residues form base pairs.

1. RNA isolated from the sub-units of rabbit reticulocyte ribosomes was hydrolysed by 0·4 n -potassium hydroxide at 20°. The probability of main-chain scission was calculated from the number-average chain length, which was obtained from S 25,w in 0·01 m -phosphate buffer. 2. The fraction, f , of the original secondary structure that the fragments re-formed at neutral pH in 4 m -guanidinium chloride, as well as in 0·01 m - and 0·1 m -phosphate buffer, was derived from changes in extinction over the range 220–310mμ on thermal denaturation. 3. The secondary structure of RNA is regarded as an assembly of hairpin loops each of 2 N + b residues on average, where N is the number of base-paired residues and b is the number of unpaired residues. 4. If chain scission takes place at random then 2 N + b =log f /log(1– p ). 5. For RNA from the smaller sub-unit 2 N + b was estimated as 25±5 residues, compared with 30±5 residues for the less stable species and 35±5 residues for the more stable species of hairpin loop of RNA from the larger sub-unit.

1. Single-stranded RNA may be regarded as an assembly of L hairpin loops each stabilized by N base pairs and each containing b unpaired residues; one loop is connected to another by c residues. 2. A theory based on the statistics of the random degradation of linear polymers was developed to relate N , b and c with the probability, p , of hydrolysing a diesterified phosphate bond. 3. The number of residues per hairpin loop, which is 2 N + b , is related to the fraction, f , of the original loops remaining intact by the equation: 2 N + b =log f /log(1– p ). 4. The theory was extended to show that the number of residues per loop may be evaluated by fractionating the RNA after hydrolysis and examining the secondary structure of each fraction. Fragments that are shorter than the hairpin loop cannot reproduce the original secondary structure. The probability that a fragment will form an intact loop increases most rapidly for fragments of between 2 N + b and 2(2 N + b )+ c residues. 5. The probability of producing a fragment capable of forming one, and only one, hairpin loop was related to N , b and c .