Analysis of the puromycin reaction. The ribosomal exclusion principle for AcPhe-tRNA binding re-examined

Protein synthesis factors (RF1, RF2, RF3, RRF, and tmRNA) and peptidyl-tRNA hydrolase rescue stalled ribosomes at sense codons

During translation, ribosomes stall on mRNA when the aminoacyl-tRNA to be read is not readily available. The stalled ribosomes are deleterious to the cell and should be rescued to maintain its viability. To investigate the contribution of some of the cellular translation factors on ribosome rescuing, we provoked stalling at AGA codons in mutants that affected the factors and then analyzed the accumulation of oligopeptidyl (peptides of up to 6 amino acid residues, oligopep-)-tRNA or polypeptidyl (peptides of more than 300 amino acids in length, polypep-)-tRNA associated with ribosomes. Stalling was achieved by starvation for aminoacyl-tRNA(Arg4) upon induced expression of engineered lacZ (β-galactosidase) reporter gene harboring contiguous AGA codons close to the initiation codon or at internal codon positions together with minigene ATGAGATAA accompanied by reduced peptidyl-tRNA hydrolase (Pth). Our results showed accumulations of peptidyl-tRNA associated with ribosomes in mutants for release factors (RF1, RF2, and RF3), ribosome recycling factor (RRF), Pth, and transfer-messenger RNA (tmRNA), implying that each of these factors cooperate in rescuing stalled ribosomes. The role of these factors in ribosome releasing from the stalled complex may vary depending on the length of the peptide in the peptidyl-tRNA. RF3 and RRF rescue stalled ribosomes by "drop-off" of peptidyl-tRNA, while RF1, RF2 (in the absence of termination codon), or Pth may rescue by hydrolyzing the associated peptidyl-tRNA. This is followed by the disassembly of the ribosomal complex of tRNA and mRNA by RRF and elongation factor G.

Fluctuations of transfer RNAs between classical and hybrid states

Adjacent transfer RNAs (tRNAs) in the A- and P-sites of the ribosome are in dynamic equilibrium between two different conformations called classical and hybrid states before translocation. Here, we have used single-molecule fluorescence resonance energy transfer to study the effect of Mg(2+) on tRNA dynamics with and without an acetyl group on the A-site tRNA. When the A-site tRNA is not acetylated, tRNA dynamics do not depend on [Mg(2+)], indicating that the relative positions of the substrates for peptide-bond formation are not affected by Mg(2+). In sharp contrast, when the A-site tRNA is acetylated, Mg(2+) lengthens the lifetime of the classical state but does not change the lifetime of the hybrid state. Based on these findings, the classical state resembles a state with direct stabilization of tertiary structure by Mg(2+) ions whereas the hybrid state resembles a state with little Mg(2+)-assisted stabilization. The antibiotic viomycin, a translocation inhibitor, suppresses tRNA dynamics, suggesting that the enhanced fluctuations of tRNAs after peptide-bond formation drive spontaneous attempts at translocation by the ribosome.

Effect of glycosaminoglycans on peptide bond formation in bacterial ribosomes

1. A cell-free system derived from E. coli has been used in this study. The process of peptide bond formation was assessed with the aid of the puromycin reaction, which is catalyzed by peptidyltransferase. 2. This reaction is inhibited by heparin, in contrast, this reaction is activated by hyaluronic acid. 3. The presence of heparin decreases the percentage of formed initiation complex (complex C), but hyaluronic acid, chondroitin sulphate and keratan sulphate have no effect on the formation of complex C. 4. From other types of glycosaminoglycans, only hyaluronic acid increases the stability of active complex C.

Puromycin reaction of the A-site bound peptidyl-tRNA

AcPhe2-tRNA(Phe) which appears in ribosomes after consecutive binding of AcPhe-tRNA(Phe) at the P sites and EF-Tu-directed binding of Phe-tRNA(Phe) at the A sites is able to react quantitatively with puromycin in the absence of EF-G. One could readily explain this fact to be the consequence of spontaneous translocation. However, a detailed study of kinetics of puromycin reaction carried out with the use of viomycin (inhibitor of translocation) and the P-site test revealed that, apart from spontaneous translocation, this peptidyl-tRNA could react with puromycin being located at the A site. This leads to the conclusion that the transpeptidation reaction triggers conformational changes in the A-site ribosomal complex bringing the 3'-end of a newly synthesized peptidyl-tRNA nearer to the peptidyl site of peptidyltransferase center. This is detected functionally as a highly pronounced ability of such a peptidyl-tRNA to react with puromycin.

Puromycin reaction for the A site-bound peptidyl-tRNA

AcPhe2-tRNA(Phe) synthesized in 70S ribosomes after consecutive binding of AcPhe-tRNA(Phe) at the P sites and EF-Tu-directed binding of Phe-tRNA(Phe) at the A sites is able to react quantitatively with puromycin in the absence of EF-G. A detailed study of the kinetics of the puromycin reaction, its comparison with that of spontaneous translocation, the use of antibiotic viomycin as an effective inhibitor of spontaneous translocation revealed that, besides spontaneous translocation, this peptidyl-tRNA could react with puromycin being located at the A site. This leads to the conclusion that the transpeptidation reaction per se triggers conformational changes in the ribosomal complex bringing the 3'-end of a newly synthesized peptidyl-tRNA nearer to the peptidyl-site of the peptidyltransferase center. This is detected functionally as the ability of such an A site bound peptidyl-tRNA to react with puromycin. This reaction is highly pronounced at elevated (25 degrees C) temperature but can be hardly detected at 0 degrees C.

Kinetics of inhibition of peptide bond formation on bacterial ribosomes

A cell-free system derived from Escherichia coli has been used in order to study the kinetics of inhibition of peptide bond formation with the aid of the puromycin reaction in solution. A similar study has been carried out earlier on a solid support matrix with the same inhibitors. We find that the overall pattern of the kinetics of inhibition is the same in the two systems. At low concentrations of inhibitor there is a competitive phase of inhibition, whereas at higher concentrations of inhibitor the type of inhibition becomes mixed noncompetitive. The values of Ki of the competitive phase in the system in solution are: 5.8 microM (amicetin), 0.2 microM (blasticidin S), 0.5 microM (chloramphenicol), and 0.5 microM (tevenel). The inhibitors amicetin, blasticidin S, and tevenel interact with the ribosome in a reaction which is slower than that of the substrate puromycin, showing clear-cut characteristics of slow-onset inhibition in both systems. Chloramphenicol, on the other hand does not easily show such a delay in solution. It interacts with the ribosome relatively faster than the other three antibiotics. Despite this, chloramphenicol too shows characteristics of time-dependent inhibition.

The function of the translating ribosome: allosteric three-site model of elongation

During the last decade, a new model for the ribosomal elongation cycle has emerged. It is based on the finding that eubacterial ribosomes possess 3 tRNA binding sites. More recently, this has been confirmed for archaebacterial and eukaryotic ribosomes as well, and thus appears to be a universal feature of the protein synthetic machinery. Ribosomes from organisms of all 3 kingdoms harbor, in addition to the classical P and A sites, an E site (E for exit), into which deacylated tRNA is displaced during translocation, and from which it is expelled by the binding of an aminoacyl-tRNA to the A site at the beginning of the subsequent elongation round. The main features of the allosteric 3-site model of ribosomal elongation are the following: first, the third tRNA binding site is located 'upstream' adjacent to the P site with respect to the messenger, ie on the 5'-side of the P site. Second, during translocation, deacylated tRNA does not leave the ribosome from the P site, but co-translocates from the P site to the E site--when peptidyl-tRNA translocates from the A site to the P site. Third, deacylated tRNA is tightly bound to the E site in the post-translocational state, where it undergoes codon--anticodon interaction. Fourth, the elongating ribosome oscillates between 2 main conformations: (i), the pre-translocational conformer, where aminoacyl-tRNA (or peptidyl-tRNA) and peptidyl-tRNA (or deacylated tRNA) are firmly bound to the A and P sites, respectively; and (ii), the post-translocational conformer, where peptidyl-tRNA and deacylated tRNA are firmly bound to the P and E sites, respectively. The transition between the 2 states is regulated in an allosteric manner via negative cooperatively. It is modulated in a symmetrical fashion by the 2 elongation factors Tu and G. An elongating ribosome always maintains 2 high-affinity tRNA binding sites with 2 adjacent codon--anticodon interactions. The allosteric transition from the post- to the pre-translocational state is involved in the accuracy of aminoacyl-tRNA selection, and the maintenance of 2 codon--anticodon interactions helps to keep the messenger in frame during translation.

Similarities and differences in the inhibition patterns of thiostrepton and viomycin: evidence for two functionally different populations of P sites when occupied with AcPhe-tRNA

According to the allosteric three-site model for the ribosomal elongation cycle, the reactions from the pre- to the post-translocational state and vice versa represent allosteric transitions which are catalyzed by elongation factor (EF)-G and EF-Tu, respectively. It has been shown recently that the non-related antibiotics thiostrepton and viomycin inhibit protein biosynthesis via a surprisingly similar mechanism. Both drugs primarily block the allosteric transitions in either direction (Hausner et al. (1988) J. Biol. Chem. 263, 13103-13111). Here we show that the secondary effects of these antibiotics differ strikingly. When the P site of poly(U) programmed ribosomes is quantitatively filled with AcPhe-tRNA, thiostrepton stimulates the rate of the formation of AcPhe-puromycin 2-fold, whereas viomycin inhibits the puromycin reaction (up to 75% inhibition). The thiostrepton-dependent stimulation is only observed when the drug is given before the P site is occupied; when thiostrepton is added after pre-filling the P site, the peptidyltransferase activity is not affected, in contrast to the translocation reaction, which is blocked irrespective of whether the drug is administered before or after tRNA is bound. The effects of both drugs became distinctly more pronounced when the P sites were saturated with AcPhe-tRNA as compared to half-saturated ribosomes. We conclude that roughly one half of the ribosomes, which first bind AcPhe-tRNA to the P site, carry this ligand in a different orientation to that of the second half of the ribosome population. These two populations probably reflect the P site in the pre- and post-translocational state, respectively.

The allosteric three-site model for the ribosomal elongation cycle: features and future

The ribosome contains three binding sites for tRNA, viz., the A site for aminoacyl-tRNA (decoding site), the P site for peptidyl-tRNA, and the E site for deacylated tRNA (E for exit). The surprising finding of an allosteric linkage between the E and A sites in the sense of a negative cooperativity has three consequences: (a) it improves the proper selection of aminoacyl-tRNAs while preventing interference from noncognate aminoacyl-tRNAs in the decoding process, (b) it provides an explanation for the ribosomal accuracy without having to resort to the proofreading hypothesis, and (c) it has deepened our understanding of the mode of action of some antibiotics.

Reactivity of the P-site-bound donor in ribosomal peptide-bond formation

The puromycin reaction, catalyzed by the ribosomal peptidyltransferase, has been carried out so as to make the definition of two distinct parameters of this reaction possible. These are (a) the final degree of the reaction which gives the proportion of peptidyl (P)-site binding of the donor and (b) the reactivity of the donor substrate expressed as an apparent rate constant (kobs). This kobs varies with the concentration of puromycin; the maximal value (k3) of the kobs, at saturating concentrations of puromycin, gives the reactivity of the donor independently of the concentrations of both the donor and puromycin. k3 is also a measure of the activity of peptidyltransferase expressed as its catalytic rate constant (kcat). If we assume that the puromycin-reactive donor is bound at the ribosomal P site, we observe the following, depending on the conditions of the experiment: the proportion of P-site binding of the donor substrates AcPhe-tRNA or fMet-tRNA can be the same and close to 100%, while there is a tenfold increase in the reactivity of the donor (k3 = 0.8 min-1 versus 8.3 min-1). On the other hand there are conditions, under which the proportion of P-site binding increases from 30% to 100% while k3 remains low and equal to 0.8 min-1. Using the puromycin reaction it was also found that an increase of Mg2+ from 10 mM to 20 mM reduces the reactivity of the donor and, hence, the activity of peptidyltransferase, provided that this change in Mg2+ occurs during the binding of the donor but not when it occurs during peptide bond formation per se. The fact that the donor substrate may exist in various states of reactivity in this cell-free system raises the possibility that the rate of peptide bond formation may not be uniform during protein synthesis.

Recovery of active ribosomal complexes from cellulose nitrate membranes

The ternary Ac-[3H]Phe-tRNA-poly(U)-ribosome complex (complex C) [D. L. Kalpaxis, D.A. Theocharis, and C. Coutsogeorgopoulos (1986) Eur. J. Biochem. 154, 267-271] was used in model experiments aiming at the purification of this complex via adsorption on cellulose nitrate membranes and then desorbing the complex back into solution. The desorption was carried out at pH 7.2 in the presence of the nonionic detergent Zwittergent (ZW). The activity status of complex C was assessed with the aid of the puromycin reaction which characterizes ribosomal peptidyltransferase as part of complex C. The optimal conditions for desorbing complex C were 5 degrees C and a buffered solution containing 0.1% ZW. The kinetic constants of peptidyltransferase in the adsorbed state were kcat = 2.0 min-1, Ks = 0.4 mM. In the desorbed state, in solution, kcat = 3.4 min-1 and Ks = 0.3 mM. The method promises to be suitable for the rapid purification of ribosomal complexes containing mRNA and aminoacyl-tRNA.

Prediction of three-dimensional structure of Escherichia coli ribosomal RNA

A model for the tertiary structure of 23S, 16S and 5S ribosomal RNA molecules interacting with three tRNA molecules is presented using the secondary structure models common to E. coli, Z. mays chloroplast, and mammalian mitochondria. This ribosomal RNA model is represented by phosphorus atoms which are separated by 5.9 A in the standard A-form double helix conformation. The accumulated proximity data summarized in Table 1 were used to deduce the most reasonable assembly of helices separated from each other by at least 6.2 A. Straight-line approximation for single strands was adopted to describe the maximum allowed distance between helices. The model of a ribosome binding three tRNA molecules by Nierhaus (1984), the stereochemical model of codon-anticodon interaction by Sundaralingam et al. (1975) and the ribosomal transpeptidation model, forming an alpha-helical nascent polypeptide, by Lim & Spirin (1986), were incorporated in this model. The distribution of chemically modified nucleotides, cross-linked sites, invariant and missing regions in mammalian mitochondrial rRNAs are indicated on the model.

Crosslinking of the anticodon of P site bound tRNA to C-1400 of E.coli 16S RNA does not require the participation of the 50S subunit

Crosslinking of the 5'-anticodon base of ribosomal P site bound AcVal-tRNA to residue C-1400 of 16S RNA or to its equivalent in 18S RNA has been shown to occur on 70S or 80S ribosomes of both prokaryotes and eukaryotes [Ciesiolka, J., Nurse, K., Klein, J. and Ofengand, J. (1985) Biochemistry 24, 3233-3239]. In the present work, we show that the crosslinking rate, crosslinking yield, and site of crosslinking are all unchanged when the 50S subunit is omitted. Therefore, all of the positional features of tRNA-ribosome complexes which allow crosslinking to occur are entirely contained in the 30S subunit. Blockage of reverse transcription by crosslink formation was used to determine the site of crosslinking. This analysis revealed that RNA modifications which do not directly block base-pairing ligands sometimes allow the modified base to be transcribed, leading to doublet band formation even when there is only a single crosslink site.

The ribosomal E site at low Mg2+: coordinate inactivation of ribosomal functions at Mg2+ concentrations below 10 mM and its prevention by polyamines

Under standard conditions (Mg2+/150 mM NH4+) ribosomes can quantitatively participate in tRNA binding at Mg2+ concentrations of 12 to 15 mM. The overall poly(U)-directed Phe incorporation and the extent of tRNA binding to either P, E or A sites decrease in a parallel manner when the Mg2+ concentration is lowered below 10 mM. At 4 mM the inactivation amounts to about 80%. The coordinate inactivation of all three binding sites is accompanied by an increasing impairment of the ability to translocate A-site bound AcPhe-tRNA to the P site. The translocation efficiency is already reduced at 10 mM Mg2+, and is completely blocked at 6-8 mM. The severe inactivation seen at 6 mM Mg2+ vanishes when the polyamines spermine (0.6 mM) and spermidine (0.4 mM) are present in the assay; tRNA binding again becomes quantitative, the total Phe synthesis even exceeds that observed in the absence of polyamines by a factor of 4. In the presence of polyamines and low Mg2+ (3 and 6 mM) two essential features of the allosteric three-site model (Rheinberger and Nierhaus, J. Biol. Chem. 261, 9133 (1986] are demonstrated. 1) Deacylated tRNA is not released from the P site, but moves to the E site during the course of translocation. 2) Occupation of the E site reduces the A site affinity and vice versa (allosteric interactions between E and A sites). The quality of an in vitro system for protein synthesis can be assessed by two criteria. First, the incubation conditions must allow a near quantitative tRNA binding. Secondly, protein synthesis should proceed with near in vivo rate and accuracy. The 3 mM Mg2+/NH4+/polyamine-system seems to be the best compromise at present between these two requirements.

Evidence that the G2661 region of 23S rRNA is located at the ribosomal binding sites of both elongation factors

Alpha-sarcin cleaves one phosphodiester bond of 23S rRNA within 70S ribosomes or 50S subunits derived from E. coli. The resulting fragment was isolated and sequenced. The cleavage site was identified as being after G2661 and is located within a universally conserved dodecamer. Cleavage after G2661 specifically blocked the binding of both elongation factors, i.e. that of the ternary complex Phe-tRNA*EF-Tu*GMPPNP and of EF-G*GMPPNP, whereas all elongation-factor independent functions of the ribosome, such as association of the ribosomal subunits, tRNA binding to A and P sites, the accuracy of tRNA selection at both sites, the peptidyl transferase activity, and the EF-G independent, spontaneous translocation, were not affected at all. Control experiments with wheat germ ribosomes yielded an equivalent inhibition pattern. The data suggest that the universally conserved dodecamer containing the cleavage site G2661 is located at the presumably overlapping region of the binding sites of both elongation factors.

Tetracycline can inhibit tRNA binding to the ribosomal P site as well as to the A site

A and P sites of Escherichia coli ribosomes were titrated with AcPhe-tRNAPhe, in the absence or presence of tetracycline. The P-site location of the bound AcPhe-tRNA was assessed by means of a quantitative puromycin reaction. The results demonstrate that, in agreement with the generally held view, tetracycline exclusively inhibits the A-site binding, if the statistical number of bound acyl-tRNA molecules per ribosome does not exceed about 0.5. However, above this value the P site becomes sensitive to tetracycline as well. It follows that the tightly coupled 70S ribosomes used in functional studies appear to be functionally heterogeneous, i.e. those P sites which cannot be affected by tetracycline are preferentially occupied by AcPhe-tRNA, whereas higher concentrations of this tRNA species are required to fill tetracycline-sensitive P sites. Furthermore, the results imply that under tRNA saturation conditions the tetracycline inhibition cannot be used as an indicator for the site location of bound tRNA.