Portal de Periódicos da CAPES

Kinetics of Macrolide Action

2004; Elsevier BV; Volume: 279; Issue: 51 Linguagem: Inglês

10.1074/jbc.m401625200

1083-351X

Martin Lovmar, Tanel Tenson, Måns Ehrenberg,

Enzyme Structure and Function

Members of the macrolide class of antibiotics inhibit peptide elongation on the ribosome by binding close to the peptidyltransferase center and blocking the peptide exit tunnel in the large ribosomal subunit. We have studied the modes of action of the macrolides josamycin, with a 16-membered lactone ring, and erythromycin, with a 14-membered lactone ring, in a cell-free mRNA translation system with pure components from Escherichia coli. We have found that the average lifetime on the ribosome is 3 h for josamycin and less than 2 min for erythromycin and that the dissociation constants for josamycin and erythromycin binding to the ribosome are 5.5 and 11 nm, respectively. Josamycin slows down formation of the first peptide bond of a nascent peptide in an amino acid-dependent way and completely inhibits formation of the second or third peptide bond, depending on peptide sequence. Erythromycin allows formation of longer peptide chains before the onset of inhibition. Both drugs stimulate the rate constants for drop-off of peptidyl-tRNA from the ribosome. In the josamycin case, drop-off is much faster than drug dissociation, whereas these rate constants are comparable in the erythromycin case. Therefore, at a saturating drug concentration, synthesis of full-length proteins is completely shut down by josamycin but not by erythromycin. It is likely that the bacterio-toxic effects of the drugs are caused by a combination of inhibition of protein elongation, on the one hand, and depletion of the intracellular pools of aminoacyl-tRNAs available for protein synthesis by drop-off and incomplete peptidyl-tRNA hydrolase activity, on the other hand. Members of the macrolide class of antibiotics inhibit peptide elongation on the ribosome by binding close to the peptidyltransferase center and blocking the peptide exit tunnel in the large ribosomal subunit. We have studied the modes of action of the macrolides josamycin, with a 16-membered lactone ring, and erythromycin, with a 14-membered lactone ring, in a cell-free mRNA translation system with pure components from Escherichia coli. We have found that the average lifetime on the ribosome is 3 h for josamycin and less than 2 min for erythromycin and that the dissociation constants for josamycin and erythromycin binding to the ribosome are 5.5 and 11 nm, respectively. Josamycin slows down formation of the first peptide bond of a nascent peptide in an amino acid-dependent way and completely inhibits formation of the second or third peptide bond, depending on peptide sequence. Erythromycin allows formation of longer peptide chains before the onset of inhibition. Both drugs stimulate the rate constants for drop-off of peptidyl-tRNA from the ribosome. In the josamycin case, drop-off is much faster than drug dissociation, whereas these rate constants are comparable in the erythromycin case. Therefore, at a saturating drug concentration, synthesis of full-length proteins is completely shut down by josamycin but not by erythromycin. It is likely that the bacterio-toxic effects of the drugs are caused by a combination of inhibition of protein elongation, on the one hand, and depletion of the intracellular pools of aminoacyl-tRNAs available for protein synthesis by drop-off and incomplete peptidyl-tRNA hydrolase activity, on the other hand. Macrolides form a large group of antibiotics that target the protein synthesis machinery (1Vázquez D. Mol. Biol. Biochem. Biophys. 1979; 30: 169-175Google Scholar, 2Walsh C. Antibiotics: Actions, Origins, Resistance. American Society for Microbiology, Washington, D. C.2003: 57-63Google Scholar). All macrolides contain a 14-, 15-, or 16-membered lactone ring with sugar residues (2Walsh C. Antibiotics: Actions, Origins, Resistance. American Society for Microbiology, Washington, D. C.2003: 57-63Google Scholar). This study concerns the action of two members of the macrolide family, erythromycin and josamycin. Erythromycin, a frequently used drug in the clinic (2Walsh C. Antibiotics: Actions, Origins, Resistance. American Society for Microbiology, Washington, D. C.2003: 57-63Google Scholar), contains a 14-membered lactone ring, whereas josamycin contains a 16-membered lactone ring (Fig. 1A). The molecular mechanism of macrolide action has remained puzzling during decades of scientific studies. All macrolides bind competitively to a site in the vicinity of the peptidyltransferase center (3Di Giambattista M. Engelborghs Y. Nyssen E. Cocito C. J. Biol. Chem. 1987; 262: 8591-8597Abstract Full Text PDF PubMed Google Scholar), but direct inhibition of the peptidyltransferase reaction has been observed only for macrolides with 16-membered lactone rings (4Poulsen S.M. Kofoed C. Vester B. J. Mol. Biol. 2000; 304: 471-481Crossref PubMed Scopus (111) Google Scholar, 5Mao J.C. Robishaw E.E. Biochemistry. 1971; 10: 2054-2061Crossref PubMed Scopus (48) Google Scholar). Erythromycin, in contrast, allows synthesis of short peptides before nascent protein elongation is inhibited (5Mao J.C. Robishaw E.E. Biochemistry. 1971; 10: 2054-2061Crossref PubMed Scopus (48) Google Scholar). Macrolides also induce drop-off of peptidyl-tRNAs from translating ribosomes both in vitro (6Otaka T. Kaji A. Proc. Natl. Acad. Sci. U. S. A. 1975; 72: 2649-2652Crossref PubMed Scopus (49) Google Scholar, 7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar) and in vivo (8Menninger J.R. Otto D.P. Antimicrob. Agents Chemother. 1982; 21: 811-818Crossref PubMed Scopus (123) Google Scholar). The macrolide induced drop-off events may be toxic for bacterial cells because accumulation of peptidyl-tRNA in the cell depletes pools of free tRNA isoacceptors (9Heurgue-Hamard V. Mora L. Guarneros G. Buckingham R.H. EMBO J. 1996; 15: 2826-2833Crossref PubMed Scopus (54) Google Scholar, 10Tenson T. Herrera J.V. Kloss P. Guarneros G. Mankin A.S. J. Bacteriol. 1999; 181: 1617-1622Crossref PubMed Google Scholar, 11Heurgue-Hamard V. Dincbas V. Buckingham R.H. Ehrenberg M. EMBO J. 2000; 19: 2701-2709Crossref PubMed Scopus (49) Google Scholar). Recently, high resolution crystal structures of several macrolides in complex with the large ribosomal subunit were obtained (12Schlunzen F. Zarivach R. Harms J. Bashan A. Tocilj A. Albrecht R. Yonath A. Franceschi F. Nature. 2001; 413: 814-821Crossref PubMed Scopus (890) Google Scholar, 13Hansen J.L. Ippolito J.A. Ban N. Nissen P. Moore P.B. Steitz T.A. Mol. Cell. 2002; 10: 117-128Abstract Full Text Full Text PDF PubMed Scopus (452) Google Scholar, 14Schlunzen F. Harms J.M. Franceschi F. Hansen H.A. Bartels H. Zarivach R. Yonath A. Structure (Lond.). 2003; 11: 329-338Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar, 15Berisio R. Harms J. Schluenzen F. Zarivach R. Hansen H.A. Fucini P. Yonath A. J. Bacteriol. 2003; 185: 4276-4279Crossref PubMed Scopus (157) Google Scholar, 16Berisio R. Schluenzen F. Harms J. Bashan A. Auerbach T. Baram D. Yonath A. Nat. Struct. Biol. 2003; 10: 366-370Crossref PubMed Scopus (165) Google Scholar). Together with new biochemical data from a cell-free translation system (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar), these structures have clarified some aspects of macrolide action. The crystal structures show that all macrolides bind to the beginning of the nascent peptide exit tunnel, with their C-5 sugars extending toward the peptidyltransferase center (Fig. 1). The biochemical data suggest a common mechanism for macrolide action, based on the space that is available between the peptidyltransferase center and the ribosome bound macrolide (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar). At the same time, several questions have remained unanswered. (i) Are the rate constants for peptidyl-tRNA drop-off enhanced by macrolides, or are these events an indirect consequence of the inhibition of the peptidyltransferase reaction by the drugs? (ii) At which step in the elongation cycle does peptidyl-tRNA drop-off occur? (iii) Do macrolides inhibit translocation of peptidyl-tRNA from the A to P site? (iv) Can protein synthesis be resumed by the dissociation of a macrolide that has caused ribosomal stalling? (v) Are the growth inhibitory effects of macrolides because of direct inhibition of protein elongation on the ribosome, or are they an indirect effect of depletion of tRNA pools because of frequent drop-off events and insufficient intracellular peptidyl-tRNA hydrolase activity? In this study we have clarified some of these issues for the macrolides erythromycin and josamycin by studying their kinetic properties in a cell-free system for protein synthesis with Escherichia coli components of high purity (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar, 17Freistroffer D.V. Pavlov M.Y. MacDougall J. Buckingham R.H. Ehrenberg M. EMBO J. 1997; 16: 4126-4133Crossref PubMed Scopus (246) Google Scholar). GTP, ATP, and [3H]Met were from Amersham Biosciences. Putrescine, spermidine, phosphoenolpyruvate, myokinase, and nonradioactive amino acids were from Sigma. Pyruvate kinase was from Roche Applied Science. Erythromycin was from Sigma. Josamycin was from ICN Biomedicals. All experiments were performed in polymix buffer, containing 5 mm magnesium acetate, 5 mm ammonium chloride, 95 mm potassium chloride, 0.5 mm calcium chloride, 8 mm putrescine, 1 mm spermidine, 5 mm potassium phosphate, and 1 mm dithioerythritol (18Jelenc P.C. Kurland C.G. Proc. Natl. Acad. Sci. U. S. A. 1979; 76: 3174-3178Crossref PubMed Scopus (268) Google Scholar). The template DNAs for in vitro transcription were prepared by annealing the following oligonucleotides at the complementary sequences (underlined) and filling the gaps by PCR: forward oligonucleotide, CTCTCTGGTACCGAAATTAATACGACTCACTATAGGGAATTCGGGCCCTTGTTAACAATTAAGGAGG, and reverse oligonucleotide for MVSN, TTTTTTTTTTTTTTTTTTTTTCTGCAGATTTAGTTAGAAACCATAGTATACCTCCTTAATTGTTAACAAGGGCCCG; reverse oligonucleotide for MFSN, TTTTTTTTTTTTTTTTTTTTTCTGCAGATTTAGTTAGAAAACATAGTATACCTCCTTAATTGTTAACAAGGGCCCG. The reverse oligonucleotide used to prepare the MS mRNA encoding the N-terminal fragment of the MS2 phage coat protein was described previously (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar). In vitro transcription and purification of mRNAs containing a poly(A) tail were as described previously (19Pavlov M.Y. Ehrenberg M. Arch. Biochem. Biophys. 1996; 328: 9-16Crossref PubMed Scopus (91) Google Scholar). Components of the translation system were purified as described previously (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar, 20Antoun A. Pavlov M.Y. Tenson T. Ehrenberg M.M. Biol. Proced. Online. 2004; (http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=15103398)PubMed Google Scholar). All experiments were performed in polymix buffer at 37 °C using equal volumes of preinitiated ribosomes and pre-formed ternary complexes containing EF-Tu, 1The abbreviations used are: EF, elongation factor; RP-HPLC, reverse phase-high pressure liquid chromatography; PT, peptidyltransferase. GTP, and aminoacyl-tRNA. The reactions were quenched by adding 10 volumes of 15% formic acid to 1 volume of reaction mixture, and the peptides were analyzed as described previously (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar). The initiation mixture, containing ribosomes (1.6 μm, ∼50% active in peptidyl transfer), [3H]fMet-tRNAfMet (2 μm), mRNA 3.2 μm, IF2 (0,3 μm), IF1 (0.6 μm) and IF3 (0.6 μm), was preincubated for 10 min at 37 °C to allow for formation of initiated ribosome complexes. The elongation mixture, containing EF-G (1.6 μm), EF-Tu (24 μm), EF-Ts (0.24 μm), and tRNAbulk (∼0.18 mm), the relevant aminoacyl-tRNA synthetases (0.1 units/μl) (defined in Ref. 21Ehrenberg M. Bilgin N. Kurland C.G. Spedding G. Ribosomes and Protein Synthesis: A Practical Approach. IRL Press at Oxford University Press, Oxford1990: 101-129Google Scholar), peptidyl-tRNA hydrolase (1.12 μm), and amino acids (160 μm each), was preincubated for 8 min at 37 °C to allow for formation of ternary complexes. In addition, both mixtures contained ATP (1 mm), GTP (1 mm), phosphoenolpyruvate (10 mm), myokinase (3 μg/ml), and pyruvate kinase (50 μg/ml). Kinetic Experiments with Quench-flow Techniques—Preinitiated ribosomes, formed in the absence or presence of josamycin, were rapidly mixed with the elongation mixture in a quench-flow instrument (KinTek Corp.), and the reaction was quenched with formic acid after different incubation times. The composition of peptides was determined by RP-HPLC, as described in Ref. 7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar. Association Rate Constant for Josamycin—Josamycin at different concentrations (2, 3, 4, and 6 μm) was added to preinitiated ribosomes to start the incubation. One volume of elongation mix was added to 1 volume of reaction mix at each incubation time, and after 10 s the reaction was quenched with formic acid. The association rates were estimated from the fraction of tri-peptide-forming ribosomes. Ratio between the Association Rate Constants for Erythromycin and Josamycin—An experiment, where josamycin and erythromycin competed for ribosome binding, was carried out by adding mixtures of josamycin and erythromycin to preinitiated ribosomal complexes. The josamycin concentration was 100 μm and the erythromycin concentration was varied between 0.5 and 16 μm. After 20 s of incubation, 1 volume of elongation mix was added to 1 volume of reaction mix, and the reaction was quenched by formic acid after 10 s. The extent of tri-peptide formation was analyzed as in the previous case. Macrolide Dissociation Rate Constants from Chase Experiments—To determine the dissociation rate constants for josamycin and erythromycin, in a first experiment initiated ribosome complexes were formed in the presence of 4 μm josamycin during 10 min. Then 300 μm erythromycin was added to start the chase. At varying incubation times, 1 volume of elongation mixture was added to 1 volume of reaction mixture. The reaction was quenched with formic acid after 10 s, and the amount of tri-peptide formed was analyzed by RP-HPLC as in the previous experiments. The fraction of ribosomes competent in tri-peptide formation increased with a rate determined by the first order compounded rate constant kobs1. In a second experiment, the initiated ribosome complexes were formed with 4 μm erythromycin by incubation for 10 min at 37 °C. Then 250 μm josamycin was added to start the chase. The reaction was stopped, and the extent of tri-peptide formation was analyzed as described above. The fraction of ribosomes competent in tri-peptide formation decayed with a rate determined by the compounded rate constant kobs2. Together with the association rate constants for josamycin (ka,J) and erythromycin (ka,E), the estimates of kobs1 and kobs2 were used to estimate the dissociation rate constants for josamycin (kd,J) and erythromycin (kd,E), as described below. Peptidyl-tRNA Drop-off—Preinitiated ribosome complexes were mixed with translation factors including peptidyl-tRNA hydrolase. The reaction was quenched with formic acid after varying incubation times, and the amounts of peptides on the ribosome (pellet) and peptides originating from drop-off events (supernatant) were analyzed by RP-HPLC as described previously (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar). Translation and Drop-off Model—To estimate the compounded rate constants k1–k5 in Fig. 4 from the translation and drop-off experiments, we described the model with ordinary differential Equations 1–5, dc1(t)/dt=−k1⋅c1(t)dc2(t)/dt=k1⋅c1(t)−(k2+k4)⋅c2(t)c1(0)=1dc3(t)/dt=k2⋅c2(t)−(k3+k5)⋅c3(t)c2(0)=c3(0)=c4(0)=c5(0)=0dc4(t)/dt=k3⋅c3(t)dc5(t)/dt=k4⋅c2(t)+k5⋅c3(t)Equations 1–5 where c1 is the concentration of initiation complex normalized to the concentration of active ribosomes; c2 and c3 are the normalized concentrations of di-peptide complexes before and after translocation, respectively. c4 is the normalized concentration of tri-peptide complex, and c5 is the normalized concentration of peptidyl-tRNA that has dissociated from ribosomes in drop-off events. In the translation experiments (Fig. 2) we could not separate c2 from c3, and therefore rate constants k2 and k3 were combined to k23 when fitting the model parameters to the experimental data as shown in Equation 6, k23=k2k3k2+k3(Eq. 6) The equation system was solved analytically using Maple 7 (Waterloo Maple Inc.), and the solution was used to fit the parameters (k1 to k5) to the experimental data. The fitting was performed using the Marquardt algorithm (22Marquardt D.W. J. Soc. Indust. Appl. Math. 1963; 11: 431-441Crossref Google Scholar) implemented in Origin 7 (OriginLab Corp.). Rate Constants for Macrolide Binding to and Dissociation from Active Ribosomes—It has been suggested (23Dinos G.P. Kalpaxis D.L. Biochemistry. 2000; 39: 11621-11628Crossref PubMed Scopus (19) Google Scholar, 24Dinos G.P. Connell S.R. Nierhaus K.H. Kalpaxis D.L. Mol. Pharmacol. 2003; 63: 617-623Crossref PubMed Scopus (27) Google Scholar) that macrolide antibiotics bind to the ribosome in two steps, according to Reaction 1, k1k12S+R⇄C1⇄C2k−1k21Reaction 1 where S is free antibiotic; R is free ribosome; C1 is an initial weak complex, and C2 is a strong complex formed between the ribosome and the drug. When the total drug concentration s0 is much larger than the total ribosome concentration r0, then the kinetics of drug binding is described by two coupled linear differential equations, and the solution is always the sum of two exponentially decaying terms and one constant term. When k-1 >> k12, the first binding step equilibrates much faster than the subsequent strong binding of the drug. Elimination of the fast variable (25Elf J. Ehrenberg M. Genome Res. 2003; 13: 2475-2484Crossref PubMed Scopus (318) Google Scholar) can then be used to simplify the slow binding step according to Reaction 2 (26Dinos G. Synetos D. Coutsogeorgopoulos C. Biochemistry. 1993; 32: 10638-10647Crossref PubMed Scopus (20) Google Scholar). Kk12S+R⇄C1⇄C2k21Reaction 2 Formation of C2 can then be approximated by Equation 7, c2(t)=c¯(1−e−kobst)(Eq. 7) c2(t) and c̄ are the current and equilibrium concentrations of C2, respectively. The compounded rate constant kobs is given by Equation 8, kobs=k12⋅s0s0+K+k21(Eq. 8) The binding of josamycin at different concentrations to the ribosome (Fig. 3A) can be described as a single step reaction, and the equilibration rate constant kobs is perfectly linear in the josamycin concentration (Fig. 3A, inset). Thus in this concentration interval, s0 << K and s0k12 >> k21, meaning that Equation 8 approximates to Equation 9. kobs=kas0=(k12/K)s0(Eq. 9) A putative pre-equilibrium complex C1 for ribosome bound josamycin must have a dissociation constant K larger than 5 × 10-5m to be compatible with the observed linear dependence of kobs on s0 (Fig. 3A, inset). To enhance the precision of the evaluation by accounting for the decrease of concentration of free josamycin, the data points in Fig. 3A were not fitted directly to Equation 7 but instead to Equation 10, tripeptides(%)=100(1−j0(e(r0−j0)ka,Jt−1)−j0+r0e(r0−j0)ka,Jt)(Eq. 10) where j0 is the total josamycin concentration, and r0 is the total concentration of active ribosomes. When the two macrolide antibiotics, josamycin (J) and erythromycin (E), compete for the same binding site on the ribosome, their binding kinetics is described by Reaction 3, provided that the criteria for elimination of fast variables are fulfilled as in Reaction 2. Kd,JkJJ+R⇄CJI⇄CJ2Kd,JKd,EkEE+R⇄CE1⇄CE2Kd,EReaction 3 In experiments at high concentrations j0 and e0 of J and E, respectively, the probability PJ that a ribosome becomes strongly bound to J rather than to E is given by Equations 11 and 12, PJ=ka,J⋅j0ka,J⋅j0+ka,E⋅e0(Eq. 11) so that 1PJ=1+ka,Eka,J⋅e0J0(Eq. 12) The probability, PE, that the ribosome binds strongly to E instead, follows from conservation of probability as given in Equation 13. PE=ka,E⋅e0ka,J⋅j0+ka,E⋅e0(Eq. 13) The compounded rate constants ka,J and ka,E are defined in analogy with ka in Equation 9 as given in Equation 14, ka,J=kJKJ and ka,E=kEKE(Eq. 14) The effective association rate constant ka,J or ka,E approximates either one of the association rate constants for drug binding in the pre-equilibrium complex C1 multiplied with the probability that it continues to the strong complex C2 instead of dissociating. The approximation is valid in the limit of high dissociation probability (k-1 >> k12), which brings Reaction 1 to Reaction 2. The experiment in Fig. 3B was designed with josamycin and erythromycin concentrations in large excess over the ribosome concentration. The incubation time (20 s) was long enough to allow for formation of a strong binding complex with either one of the drugs, as verified in experiments where the extent of binding was monitored as a function of time. The incubation time was also short enough to make exchange of drugs between their strong binding sites negligible (see Table I and below). The slope of the straight line in Fig. 3B was used to obtain the ratio between ka,E and ka,J according to Equation 12.Table IJosamycin and erythromycin ribosome binding kineticsJosamycinErythromycinka (μm-1 s-1)0.0325 ± 0.00071.00 ± 0.04kd (s-1)0.18 × 10-3 ± 0.02 × 10-310.8 × 10-3 ± 0.7 × 10-3KD (nm)5.5 ± 0.510.8 ± 0.3 Open table in a new tab When ribosomes have been equilibrated with one macrolide and then another macrolide is added, there will be an exchange of macrolides strongly bound to the ribosome, and this can be monitored experimentally. In addition, when both macrolides are present in large excess over ribosomes as well as over their respective dissociation constants and the affinity of each drug to its strong binding site is large compared with its affinity to the pre-equilibrium site, the exchange kinetics takes a particularly simple form. To see this, assume that all ribosomes are bound to a macrolide at all times. Assume further that kJ >> kd,J and kE >> kd,E so that the ribosomes are in one of their strong complexes CJ2 or CE2. Under those conditions, the full reaction dynamics can be accounted for as transitions between CJ2 and CE2 complexes (Reaction 5). kJECJ2⇄CE2kEJReaction 4 The compounded rate constant kJE is the dissociation rate constant kd,J for J, leaving complex CJ2 multiplied with the probability PE that J will be replaced by E. Similarly, the compounded rate constant kEJ is the dissociation rate constant kd,E for E, leaving complex CE2 multiplied by the probability PJ that E is replaced by J (Equation 15). kJE=kd,J⋅PE and kEJ=kd,E⋅P(Eq. 15) The time variation of the concentration cJ2 is determined by the differential Equation 16, dcJ2dt=−kJEcJ2+kEJcE2(Eq. 16) The concentration cE2 of complex CE2 can be eliminated by the conservation relation cJ2 + cE2 = r0. With the relaxation rate kobs = kJE + kEJ, this gives Equation 17. dcJ2dt=−kobs⋅cJ2+kEJ⋅r0(Eq. 17) Two chase experiments can be used to determine the dissociation rate constants kd,J and kd,E. In the first, ribosomes are pre-equilibrated with J and then E is added (relaxation rate kobs1, see Fig. 3C). In the second experiment, ribosomes are pre-equilibrated with E and then J is added (relaxation rate kobs2, see Fig. 3D). The dissociation rate constants follow from Equation 18, kobs1=kd,JPE1+kd,EPJ1kobs2=kd,JPE2+kd,EPJ2(Eq. 18) which has the solution shown in Equation 19, kd,J=kobs1PJ2−kobs2PJ1PJ2PE1−PJ1PE2 and kd,E=kobs2PE1−kobs1PE2PJ2PE1−PJ1PE2(Eq. 19) The approximations leading to Equation 19 require that the C2 complexes have much higher affinity than the C1 complexes and that the ribosomes are fully saturated with a macrolide at all times. We have estimated the pre-equilibriums for josamycin and erythromycin to be at least a factor of 10,000 or 100, respectively, weaker than their strong binding states, so that these criteria are fulfilled in the experiments in Fig 3, C and D. For the analysis in this section we have assumed for simplicity that the total drug concentrations are always much larger than the ribosome concentration. In some cases, the drug excess was only 3-fold, and here we have accounted for the disappearance of drugs from their free state by ribosome binding in our analysis. Josamycin and erythromycin belong to two different subclasses of the macrolide family and have been reported to inhibit protein synthesis in principally different ways (4Poulsen S.M. Kofoed C. Vester B. J. Mol. Biol. 2000; 304: 471-481Crossref PubMed Scopus (111) Google Scholar). This study elucidates their modes of action with kinetic data from a cell-free translation system with pure E. coli components (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar, 19Pavlov M.Y. Ehrenberg M. Arch. Biochem. Biophys. 1996; 328: 9-16Crossref PubMed Scopus (91) Google Scholar). Inhibition of Di- and Tripeptide Formation by Josamycin—It was proposed from the crystal structure of the 50 S ribosomal subunit in complex with carbomycin A, a macrolide very similar to josamycin (Fig. 1A), that the isobutyrate residue of the drug occupies the same site as the side chain of the amino acid on the aminoacyl-tRNA in the A site of the large subunit (Fig. 1C) (13Hansen J.L. Ippolito J.A. Ban N. Nissen P. Moore P.B. Steitz T.A. Mol. Cell. 2002; 10: 117-128Abstract Full Text Full Text PDF PubMed Scopus (452) Google Scholar). This is in line with results from biochemical experiments, showing that carbomycin A inactivates peptidyl transfer in an artificial system with puromycin as the acceptor and N-acetyl-Phe-tRNAPhe as the donor (4Poulsen S.M. Kofoed C. Vester B. J. Mol. Biol. 2000; 304: 471-481Crossref PubMed Scopus (111) Google Scholar, 5Mao J.C. Robishaw E.E. Biochemistry. 1971; 10: 2054-2061Crossref PubMed Scopus (48) Google Scholar). In contrast, experiments in a more realistic system with heteropolymeric mRNAs show that both fMet-Val-tRNAVal and fMet-Gly-Ile-tRNAIle can be formed on ribosomes in complex with josamycin (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar). This suggests that the inhibitory action of josamycin can be affected by the identities of the amino acids on the tRNAs in the A and P sites as well as by the length of the amino acid chain on the peptidyl-tRNA in the P site. To study this phenomenon further, we prepared two different mRNAs: one encoding fMet-Phe-Ser and the other encoding fMet-Val-Ser. Ribosomes, initiated with either one of these mRNAs in the absence or presence of josamycin, were rapidly mixed in a quench-flow instrument with all components necessary for the formation fMet-Val-Ser or fMet-Phe-Ser tri-peptides. The amounts of di- and tri-peptides at different incubation times were quantified in an HPLC system with on-line radiometry, and the results are shown in Fig. 2. The kinetics of the four reactions could in each case be approximated by a scheme containing only one rate constant for di-peptide formation (k1) and one combined rate constant for tri-peptide formation (k23), including both the rate constant for translocation (k2) and second peptidyl transfer rate constant (k3) (Fig. 4). In the absence of josamycin, k1 was 54 s-1 for fMet-Phe formation and 61 s-1 for fMet-Val formation. The rate constant k23 was 5.7 s-1 for fMet-Phe-Ser formation and 2.3 s-1 for fMet-Val-Ser formation in the absence of josamycin. In the presence of josamycin, k1 was about 0.06 s-1 for fMet-Phe formation and 14 s-1 for fMet-Val formation. The rate constant k23 was too slow to be detected both for fMet-Phe-Ser and fMet-Val-Ser formation in the presence of josamycin. We also studied the effects of erythromycin on the rate constants for di- and tri-peptide formation in the same two cases, and we found no inhibition of any rate constant for either of the two mRNAs (data not shown). In summary, the rate of di-peptide formation was inhibited about 1000-fold by josamycin in the formation of fMet-Phe, and only about 5-fold in the formation of fMet-Val. The rate of tri-peptide formation was reduced to virtually zero by josamycin, but the rates of both di- and tri-peptide formation were unaffected by erythromycin for both mRNAs. This means that ribosomes that contain josamycin can be separated from those that do not contain a macrolide or contain erythromycin by their ability to form the tri-peptide fMet-Phe-Ser. This kinetic difference between ribosomes containing josamycin and ribosomes that do not was used to obtain the complete binding kinetics of both josamycin and erythromycin, as will be described next. Kinetics of Josamycin and Erythromycin Binding to the Ribosome—Josamycin and erythromycin bind competitively to the ribosome, but they block protein synthesis at different lengths of the nascent peptide (7Tenson T. Lovmar M. Ehrenberg M. J. Mol. Biol. 2003; 330 (and references therein): 1005-1014Crossref PubMed Scopus (324) Google Scholar). A ribosome that contains josamycin cannot form an fMet-Phe-Ser pept