Solid phase synthesis and binding affinity of peptidyl transferase tra
http://www.100md.com
《核酸研究医学期刊》
1 Department of Molecular Biophysics and Biochemistry and 2 Department of Chemistry, Yale University, PO Box 208114, New Haven, CT 06520-8114, USA, 3 Dharmacon, Inc., Lafayette, CO 80026, USA and 4 Department of Chemistry, St Olaf College, Northfield, MN 55057, USA
*To whom correspondence should be addressed. Tel: +1 203 432 9772; Email: strobel@csb.yale.edu
ABSTRACT
All living cells are dependent on ribosomes to catalyze the peptidyl transfer reaction, by which amino acids are assembled into proteins. The previously studied peptidyl transferase transition state analog CC-dA-phosphate-puromycin (CCdApPmn) has important differences from the transition state, yet current models of the ribosomal active site have been heavily influenced by the properties of this molecule. One significant difference is the substitution of deoxyadenosine for riboadenosine at A76, which mimics the 3' end of a P-site tRNA. We have developed a solid phase synthetic approach to produce inhibitors that more closely match the transition state, including the critical P-site 2'-OH. Inclusion of the 2'-OH or an even bulkier OCH3 group causes significant changes in binding affinity. We also investigated the effects of changing the A-site amino acid side chain from phenylalanine to alanine. These results indicate that the absence of the 2'-OH is likely to play a significant role in the binding and conformation of CCdApPmn in the ribosomal active site by eliminating steric clash between the 2'-OH and the tetrahedral phosphate oxygen. The conformation of the actual transition state must allow for the presence of the 2'-OH, and transition state mimics that include this critical hydroxyl group must bind in a different conformation from that seen in prior analog structures. These new inhibitors will provide valuable insights into the geometry and mechanism of the ribosomal active site.
INTRODUCTION
Ribosomes are the macromolecular machines responsible for synthesizing proteins in all living cells. The large ribosomal subunit (50S in prokaryotes) contains the site of catalysis, the peptidyl transferase center. The reaction substrates include a peptidyl-tRNA, charged with the growing peptide chain bound to a tRNA binding site on the ribosome, termed the P-site, and an aminoacyl-tRNA, charged with a single amino acid bound to a second location on the ribosome, termed the A-site. Peptide bond formation involves aminolysis of the P-site ester by the A-site -amino group (Fig. 1). A large body of biochemical work, in conjunction with the recently determined high resolution crystal structures of the 50S subunit has lead to a substantial chemical insight regarding ribosomal function (1–4).
Figure 1. The peptidyl transferase reaction, including the theoretical transition state. Peptidyl transferase occurs when the -amino group of the A-site amino acid (in this case, tyrosine) nucleophilically attacks the carboxyl carbon of the P-site nascent peptide’s C-terminal amino acid. The resulting transition state contains a tetrahedral carbon with a single oxyanion. Subsequent release of the carboxyl carbon from the P-site tRNA yields the reaction products, a deacylated P-site tRNA and an N + 1 length peptide linked to the A-site tRNA.
Several previous biochemical studies of the peptidyl transferase center have focused on the interaction between the 50S subunit and small molecules that act as peptidyl transferase substrates or inhibitors, as well as amino-acylated tRNAs (5–11). Such studies identified regions of the 23S rRNA closely associated with the peptidyl transferase center, including the A and P loops. Other studies have addressed the stereochemistry, composition and significance of the amino acid acceptors (1). The importance of specific substrate functional groups, particularly the 2'-OH and 3'-OH of A76, for activity as amino acid donors or acceptors and during translocation has also been investigated (12–16).
Most of these early biochemical studies analyzed individual ribosomal substrates, or inhibitors mimicking individual substrates. Welch and colleagues synthesized an inhibitor that simulated the simultaneous binding of both A-site and P-site charged tRNA fragments, intended to mimic the reaction’s transition state (17). This inhibitor, CC-dA-phosphate- puromycin (CCdApPmn), consists of the antibiotic puromycin (Pmn) covalently linked via a bridging phosphoramidate to the trinucleotide CCdA (Fig. 2A). The puromycin portion of the molecule, which is a dimethyl-adenosine linked by a non-hydrolyzable amide bond to the amino acid methyl-tyrosine, binds in the A-site and imitates the 3' end of an amino-acylated A-site tRNA. The CCdA trinucleotide binds in the P-site and corresponds to the CCA end of a P-site tRNA. The bridging phosphoramidate group was designed to mimic the theoretical tetrahedral transition state of the peptidyl transferase reaction (17).
Figure 2. Peptidyl transferase transition state analogs. (A) CCdApPmn. The puromycin portion of the molecule mimics the terminal adenosine and amino acid of an amino-acylated A-site tRNA. The trinucleotide CCdA mimics the 3' end of a P-site tRNA, and the tetrahedral phosphate mimics the tetrahedral carbon of the transition state. Significant differences between the actual transition state and this molecule include the absence of a 2'-OH from the P-site A and the distribution of the negative charge between two oxygen atoms. (B) Improved transition state analogs. Six new molecules were synthesized, varying in their P-site A76 2' functional group (R1) and their A-site amino acid R group (R2). (C) The structure of CCdApPmn bound in the peptidyl transferase center of 50S ribosomal subunits from H.marismortui. CCdApPuro is shown in green, and nearby rRNA bases are shown in gray. rRNA base numbers are H.marismortui numbering with E.coli equivalents in parentheses. One of the non-bridging oxygens of the tetrahedral phosphate is within 2.8 ? of the 2' carbon of the P-site A76, shown in red. The methyl tyrosine portion is stacked between A2486 and U2487.
Transition state analogs can be extremely useful tools for examining the mechanism by which enzymes catalyze their reactions (18). Consistent with this expectation, a number of critical insights into ribosomal catalysis have been achieved by exploring CCdApPmn binding. The tight interaction of CCdApPmn with the ribosome and its inhibition of peptidyl transferase supported the nucleophilic attack based reaction scheme by demonstrating that a mimic of the transition state can bind in the active site, and that the A-site and P-site are sufficiently proximal to simultaneously accommodate this molecule (17). Subsequent structural studies on the 50S ribosomal subunit used CCdApPmn to identify the location of the peptidyl transferase center. The observation that no portion of any ribosomal protein was within 18 ? of the tetrahedral phosphate provided evidence that the ribosomal RNA is the catalytic agent of the ribosome (19).
The positioning of CCdApPmn within the active site has led to detailed mechanistic models of ribosome function involving specific rRNA functional groups (2,19–21). For example, in the co-crystal structure, the positioning of A2451 relative to one of the non-bridging phosphate oxygens of the tetrahedral phosphate (Fig. 2C) was said to suggest a possible role for this nucleotide as a general acid/base, or it suggested involvement of A2451 in stabilization of the transition state oxyanion (19). Later, studies showing that CCdApPmn binding is pH independent were used to discredit a model for ribosomal catalysis that includes oxyanion stabilization (22). However, there are some serious caveats to any mechanistic conclusions based upon the CCdApPmn inhibitor.
Despite its clear utility for structural mapping of the peptidyl transferase center, CCdApPmn is not an ideal transition state analog. It has become apparent that differences between CCdApPmn and the transition state are likely to be functionally important. For example, in contrast to the single oxyanion of the actual transition state, the charge of CCdApPmn is delocalized between the two non-bridging oxygens. Also, because the two non-bridging oxygens are chemically identical, the tetrahedral center of CCdApPmn is achiral, which creates stereochemical ambiguity in CCdApPmn’s mimicry of the transition state. Perhaps the most significant problem with the inhibitor is the substitution of deoxyadenosine in the position corresponding to A76 of the P-site tRNA. tRNAs containing a deoxyadenosine at position 76 are inactive as P-site substrates (13). In addition to eliminating a potentially important functional group from the 2' position of the adenosine, this change also results in a different sugar pucker in the ribose ring. This problem is illustrated by the crystal structure of CCdApPmn bound in the peptidyl transferase center, wherein one of the non-bridging phosphate oxygens is within 2.8 ? of the deoxyadenosine A76 C2' (Fig. 2C). This oxygen is within the van der Waals radius of the position expected to be occupied by the omitted 2'-OH, suggesting that the phosphate cannot occupy the same position in the context of a ribo-adenosine at the A76 P-site (19,22).
The conformation of CCdApPmn in crystals of the Haloarcula marismortui 50S subunit also has some critical differences from a model of the transition state that was extrapolated from the positions of independently placed A-site and P-site substrates. The differences are particularly evident with regard to the position of the tetrahedral phosphate (20). This model suggests that the original assignment of the two non-bridging oxygens of CCdApPmn to the oxyanion and the growing peptide chain were reversed. Instead of pointing toward the N3 of A2451 and being within hydrogen bonding distance, the oxyanion is pointing away from A2451 and is not in obvious hydrogen bonding distance of any base. This argues against a direct stabilizing interaction between the oxyanion and the ribosomal RNA. These factors bring into question the relevance of CCdApPmn’s conformation within the peptidyl transferase center and suggest important mechanistic information could be obtained if improved inhibitors could be developed.
The synthesis of transition state analogs with a 2'-OH on the P-site A76 calls for an improved synthetic scheme. The solution phase synthetic route used to prepare CCdApPmn involved the coupling of CCdAp with puromycin. The 3'-phosphate of CCdAp was activated with 1-ethyl-3-carbodiimide which facilitated nucleophilic attack by the -amino group of puromycin or by water. Welch et al. (17) elected to substitute the terminal A with 2'-deoxyadenosine because the adenosine 2'-OH would preferentially attack the activated 3'-phosphate to produce a trinucleotide with a cyclic phosphate instead of puromycin. This eliminated some of the potential competing reactions, and while the yields were low, it was possible to generate a sufficient amount of the inhibitor for biochemical and structural studies (17). However, this synthetic approach is not a general scheme that could be applied readily for the preparation of variants of the original inhibitor.
In order to generate improved transition state analogs and eliminate the problems associated with established synthetic methods, we have employed a solid phase synthetic strategy (Fig. 3). This approach is compatible with 2'-bis(2-acetoxyethoxy)methyl (ACE) protecting group chemistry (23), thus making it possible to include the 2'-OH of A76 that was omitted in the original inhibitor. We designed two unique A-site nucleosides coupled to conventional polystyrene solid supports. A family of inhibitors based upon CCdApPmn was prepared by coupling standard nucleoside phosphoramidites to the supports. These compounds have made it possible to examine interactions with the amino acid side chain and explore interactions with the P-site ribose substituent. Here we report the solid phase organic synthesis of peptidyl transferase inhibitors and the characterization of inhibitor binding to the ribosomal peptidyl transferase center.
Figure 3. Solid phase synthesis of transition state analogs. The initial reaction shown here is the coupling of TMS protected puromycin aminonucleoside with the acyl chloride of L-hydroxy acids where R1 = CH3 (alanine analogs) or CH2C6H5 (phenylalanine analogs). Standard protection and deprotection conditions afforded compounds 5a and 5b in an overall yield of 42–48% from the starting nucleoside. Polymer bound puromycin was obtained by coupling compounds 5a and 5b to N-terminus polystyrene supports using standard techniques. The fluoride labile silyl protecting group was removed with TEMED/HF in MeCN prior to the first phosphoramidite reaction. Three separate adenosine phosphoramidites were followed by cytidine phosphoramidites to generate the desired addition of CCA to the A-site puromycin analog. Oxidation, deprotection and cleavage reactions were all conducted using established protocols.
MATERIALS AND METHODS
General
Solvents were reagent grade and dried prior to use. Polystyrene solid support was purchased from Amersham. All other chemicals were purchased from Aldrich and used without further purification. 1H-NMR spectra were recorded on a 400 MHz spectrometer using CDCl3 as an internal standard. HPLC was carried out on a C18 reversed-phase column (Rainin), which was eluted with a 100 mM triethylammonium acetate/acetonitrile gradient. HRMS data were obtained from the Keck Foundation Biotechnology Resource Center at Yale University. Mass spectrometry was performed using a Micromass LCT electrospray time of flight spectrometer.
TBDMS protected L-lactic acid (1a) and L-phenyl lactic acid (1b)
To a stirred solution of either L-lactic acid or L-phenyl lactic acid (3.44 mmol) with imidazole (10.3 mmol) in 10 ml of anhydrous dimethylformamide was added 1.04 g (6.88 mmol) of tert-butyldimethylsilyl chloride. The reaction was allowed to proceed overnight at room temperature after which time the solution was diluted with hexanes and washed once with 50 ml of water, once with 50 ml of saturated NaHCO3 and once with 50 ml of brine. The organic layer was dried over Na2SO4 and the solvent was removed under reduced pressure to yield colorless oils. Compound 1a (0.88 g, 2.7 mmol) 80% yield; compound 1b (1.1 g, 2.9 mmol) 85% yield. The products were used in subsequent reactions without further purification.
Compound 1a. 1H-NMR CDCl3 0.034 (s, 3H, Si-CH3), 0.070 (s, 3H, Si-CH3), 0.243 (s, 6H, Si-CH3), 0.866 (s, 9H, Si-C(CH3)3), 0.909 (s, 9H, Si-C(CH3)3), 1.36 (d, 3H, CH3), 4.22 (q, 1H, CH).
Compound 1b. 1H-NMR CDCl3 –0.001 (s, 3H), 0.128 (s, 3H), 0.453 (s, 3H), 0.473 (s, 3H), 1.02 (s, 9H), 1.14 (s, 9H), 3.09 (m, 1H), 3.28 (m, 1H), 4.49 (m, 1H), 7.46 (m, 5H).
3'-N-(TBDMS-hydroxy)-amidopuromycin (2a and 2b)
The TBMDS protected hydroxy acids (1a, 1b) were activated for coupling by adding oxalyl chloride (0.68 ml, 7.8 mmol) to a stirred solution of either TBDMS protected L-lactic acid (1a) (1.66 g, 5.2 mmol) or L-phenyl lactic acid (1b) (2.0 g, 5.2 mmol) in 5 ml of anhydrous methylene chloride with catalytic N,N-dimethylformamide (27). After 2 h at room temperature, this solution was added dropwise to a solution of TMS protected puromycin aminonucleoside prepared by adding trimethylsilyl chloride (0.5 ml, 3.9 mmol) to a solution of puromycin aminonucleoside (0.5 g, 1.17 mmol) in 5 ml of anhydrous pyridine (room temperature, 2 h). The N-acylation reaction was complete after 1 h at room temperature upon which time the reactions were quenched by adding 0.78 g (4 mmol) of citric acid in 10 ml of methanol followed by 5 ml of 50 mM K2CO3 in methanol. Evaporation of the solvent followed by purification on silica with a 0–5% methanol in CH2Cl2 gradient yielded products 2a and 2b as white solids. Compound 2a (0.46 g, 0.95 mmol) 81% yield; compound 2b (0.47 g, 0.84 mmol) 72% yield.
Compound 2a. 1H-NMR CDCl3 0.104 (s, 6H, Si-CH3), 0.911 (s, 9H, Si-C(CH3)3), 1.37 (d, 3H, CH3), 3.83 (d, 1H, H5'), 3.97 (d, 1H, H5'), 4.22 (m, 2H, unresolved CH, H4'), 4.41 (m, 1H, H3'), 4.65 (m, 1H, H2'), 5.89 (d, 1H, H1'), 7.62 (d, 1H, NHCO), 8.09 (s, 1H, H8), 8.16 (s, 1H, H2); MS calc. for C21H36N6O5SiNa: 503.2414, found M+Na 503.2405. Rf = 0.2 (silica, 5% MeOH/CH2Cl2).
Compound 2b. 1H-NMR CDCl3 –0.25 (s, 3H, Si-CH3), –0.08 (s, 3H, Si-CH3), 0.87 , 2.85 (q, 1H, CH2), 3.05 (dd, 1H, CH2), 3.80 (m, 1H, H5'), 4.00 (m, 1H, H5'), 4.12 (m, 1H, H4'), 4.32 (m, 1H, H3'), 4.38 (m, 1H, H2'), 4.47 (m, 1H, CH), 5.82 (d, 1H, H1'), 7.18 (m, 5H, phe), 8.06 (s, 1H, H8), 8.20 (s, 1H, H2); MS calc. for C27H41N6O5Si: 557.2907, found M+H 557.2908. Rf = 0.3 (silica, 1:1 CH2Cl2/EtOAc).
5'-Dimethoxytrityl-2'-acetyl-3'-N-(TBDMS-hydroxy)-amidopuromycin (3a and 3b)
4,4-Dimethoxytrityl chloride (0.63 g, 1.88 mmol) was added to a stirred solution of 3'-N-(TBDMS-L-lactic acid)-amidopuromycin (2a) or 3'-N-(TBDMS-L-phenyllactic acid)-amidopuromycin (2b) (0.6 g, 1.25 mmol) in anhydrous pyridine at 0°C. The solution was slowly warmed to room temperature. After 3 h, TLC (silica, 5% MeOH/CH2Cl2) showed complete conversion to a faster moving product (Rf = 0.3–0.4). The solution was cooled to 0°C and dimethylaminopyridine (0.02 g, 0.017 mmol) and acetic anhydride (0.6 ml, 6.6 mmol) were added. The reactions continued overnight at room temperature. The solvent was removed under reduced pressure and the products were isolated as white solids by flash chromatography (silica, 1:1 hexanes/ethyl acetate). Compound 3a (0.74 g, 0.90 mmol) 72% yield; compound 3b (0.83 g, 0.92 mmol) 74% yield.
Compound 3a. 1H-NMR CDCl3 –0.039 (s, 3H, Si-CH3), 0.055 (s, 3H, Si-CH3), 0.869 (s, 9H, Si-C(CH3)3), 1.30 (d, 3H, CH3), 2.13 (s, 3H, COCH3), 3.33 (q, 1H, H5'), 3.41 (dd, 1H, H5'), 3.50 (broad s, 6H, N(CH3)2), 3.74 (s, 6H, OCH3), 4.12 (m, 2H, unresolved H4', H3'), 5.14 (m, 1H, CH), 5.72 (dd, 1H, H2'), 6.15 (d, 1H, H1'), 6.74 (m, 4H, DMT), 6.83 (d, 1H, NHCO), 7.27 (m, 4H, DMT), 7.38 (m, 2H, DMT), 7.90 (s, 1H, H8), 8.29 (s, 1H, H2); MS calc. for C44H56N6O8SiNa: 847.3827, found M+Na 847.3840. Rf = 0.2 (silica, 1:1 hexanes/EtOAc).
Compound 3b. 1H-NMR CDCl3 –0.28 (s, 3H, Si-CH3), –0.21 (s, 3H, Si-CH3), 0.83 (s, 9H, Si-C(CH3)3), 1.97 (s, 3H, COCH3), 2.89 (q, 1H, CH2), 2.97 (dd, 1H, CH2), 3.28 (m, 1H, H5'), 3.38 (m, 1H, H5'), 3.57 (broad s, 6H, N(CH3)2), 3.75 (s, 6H, OCH3), 4.07 (m, 1H, H4'), 4.23 (m, 1H, H3'), 5.14 (m, 1H, CH), 5.28 (m, 1H, H2'), 6.57 (m, 1H, NHCO), 6.75 (m, 4H, DMT), 7.22 (m, 12H, unresolved DMT, phe), 7.37 (m, 2H, DMT), 7.95 (s, 1H, H8), 8.38 (s, 1H, H2); MS calc. for C50H61N6O8Si: 901.4320, found M+H 901.4321. Rf = 0.2 (silica, 1:1 hexanes/EtOAc).
2'-Acetyl-3'-N-(TBDMS-hydroxy)-amidopuromycin (4a and 4b)
To a solution of 5'-dimethoxytrityl-2'-acetyl-3'-N-(TBDMS-lactic acid)-amidopuromycin (3a) (0.74 g, 0.9 mmol) or 5'-dimethoxytrityl-2'-acetyl-3'-N-(TBDMS-phenyl lactic acid)-amidopuromycin (3b) (0.94 g, 1.04 mmol) in anhydrous CH2Cl2 at 0°C was added dropwise trifluoroacetic acid (0.4 ml, 5 mmol). After 30 min the bright orange solution was extracted with 1 x 30 ml 30% NaHCO3 and 1 x 30 ml H2O. The organic phase was dried over Na2SO4 and the solvent removed under reduced pressure. Flash chromatography (silica, 2% MeOH/CH2Cl2) yielded products 4a and 4b, respectively, as crystalline white solids. Compound 4a (0.43 g, 0.78 mmol) 91% yield; compound 4b (0.58 g, 0.97 mmol) 92% yield.
Compound 4a. 1H-NMR CDCl3 0.126 (s, 3H, Si-CH3), 0.137 (s, 3H, Si-CH3), 0.945 (s, 9H, Si-C(CH3)3), 1.38 (d, 3H, CH3), 2.12 (s, 3H, COCH3), 3.50 (broad s, 6H, N(CH3)2), 3.78 (m, 1H, H5'), 4.01 (m, 1H, H5'), 4.26 (m, 2H, H4', H3'), 4.95 (m, 1H, CH), 5.73 (m, 1H, H2'), 6.05 (d, 1H, H1'), 7.13 (d, 1H, NHCO), 7.81 (s, 1H, H8), 8.30 (s, 1H, H2); MS calc. for C23H38N6O6SiNa: 545.2520, found M+Na 545.2518. Rf = 0.1 (silica, 1:1 hexanes/EtOAc).
Compound 4b. 1H-NMR CDCl3 –0.10 (s, 3H, Si-CH3), –0.01 (s, 3H, Si-CH3), 0.91 (s, 9H, Si-C(CH3)3), 1.94 (s, 3H, COCH3), 2.97 (m, 2H, CH2), 3.44 (broad s, 6H, N(CH3)2), 3.74 (m, 1H, H5'), 3.95 (m, 1H, H5'), 4.08 (d, 1H, H4'), 4.39 (t, 1H, H3'), 4.87 (dd, 1H, CH), 5.41 (m, 1H, H2'), 5.96 (d, 1H, H1'), 6.76 (d, 1H, NHCO), 7.20 (m, 5H, phe), 7.82 (s, 1H, H8), 8.33 (s, 1H, H2); MS calc. for C29H43N6O6Si: 599.3013, found M+H 599.3005. Rf = 0.3 (silica, 5% MeOH/CH2Cl2).
5'-Succinyl-2'-acetyl-3'-N-(TBDMS-hydroxy)-amidopuromycin (5a and 5b)
A mixture of 2'-acetyl-3'-N-(TBDMS-lactic acid)-amidopuromycin (4a) (0.43 g, 0.82 mmol) or 2'-acetyl-3'-N-(TBDMS-phenyl lactic acid)-amidopuromycin (4b) (0.56 g, 0.93 mmol), succinic anhydride (0.15 g, 1.5 mmol or 0.93 g, 9.35 mmol), dimethylaminopyridine (0.17 g, 1.4 mmol) in anhydrous pyridine (10 ml) was stirred at room temperature for 2 h. TLC (silica, 5% methanol/CH2Cl2) showed clean complete reactions. Following removal of the solvent in vacuo, the residues were dissolved in CH2Cl2 and washed twice with 30 ml of 10% (w/v) citric acid, once with 30 ml of brine and once with 30 ml of H2O. The organic phase was dried over Na2SO4 and the solvent removed under reduced pressure. Chromatography on silica (0–5% methanol in CH2Cl2) afforded products 5a and 5b as crystalline white solids. Compound 5a (0.49 g, 0.79 mmol) 96% yield; compound 5b (0.57 g, 0.81 mmol) 87% yield.
Compound 5a. 1H-NMR CDCl3 0.116 (s, 6H Si-CH3), 0.933 (s, 9H, Si-C(CH3)3), 1.34 (d, 3H, CH3), 2.16 (s, 3H, COCH3), 2.65 (m, 4H, CH2CH2), 3.49 (broad s, 6H, N(CH3)2), 4.11 (m, 1H, H5'), 4.21 (m, 2H, H5', H4'), 4.51 (dd, 1H, H3'), 5.26 (m, 1H, CH), 5.68 (dd, 1H, H2'), 6.06 (s, 1H, H1'), 7.07 (d, 1H, NHCO), 7.96 (s, 1H, H8), 8.34 (s, 1H, H2); MS calc. for C27H42N6O9SiNa: 645.2680, found M+Na 645.2682. Rf = 0.2 (silica, 5% MeOH/CH2Cl2).
Compound 5b. 1H-NMR CDCl3 –0.09 (s, 3H Si-CH3), 0.00 (s, 3H Si-CH3), 0.90 (s, 9H, Si-C(CH3)3), 1.97 (s, 3H, COCH3), 2.63 (s, 4H, CH2CH2), 2.97 (m, 2H, CH2), 3.44 (broad s, 6H, N(CH3)2), 4.00 (m, 1H, H5'), 4.19 (m, 1H, H5'), 4.40 (m, 1H, H4'), 4.50 (m, 1H, H3'), 5.25 (m, 2H, H2', CH, unresolved), 5.94 (d, 1H, H1'), 6.75 (d, 1H, NHCO), 7.14 (m, 5H, phe), 7.91 (s, 1H, H8), 8.35 (s, 1H, H2); MS calc. for C33H47N6O9Si: 699.3174, found M+H 699.3168. Rf = 0.2 (silica, 5% MeOH/CH2Cl2).
Derivatization of solid support
Nucleosides 5a and 5b were coupled to LCCA-polystyrene solid supports (1.0 g) (Amersham) by dissolving 5a or 5b (0.2 mmol) in anhydrous pyridine (10 ml) with triethylamine (0.08 ml), dimethylaminopyridine (0.012 g, 0.1 mmol) and 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (0.38 g, 2 mmol). The reactants along with the support were placed in 40 ml conical glass tubes and rocked gently at room temperature for 24 h. The reactions were filtered through a sintered glass funnel and washed once with 20 ml of pyridine, once with 20 ml of methanol and once with 20 ml of CH2Cl2 and then air dried. Nucleoside loading was determined by reacting a small portion of the support (5 mg) with 50 mM K2CO3 in methanol and measuring the optical absorbance at 275 nm. This gave an estimated nucleoside loading of 92 μmol/g for compound 5a and 147 μmol/g for compound 5b. High loadings could be employed in this case because the synthetic targets were quite short. Unreacted sites on the support were capped by adding 5 ml of 1 M acetic anhydride, 1 M 2,6-lutidine in tetrahydrofuran and 5 ml of 1 M N-methyl imidazole in tetrahydrofuran and rocking gently at room temperature for 3 h. The support was then filtered and washed with 20 ml of methanol and 20 ml of CH2Cl2 and air dried.
Solid phase synthesis of inhibitors
The puromycin derivatized support (80 mg) was packed into oligo synthesis columns (Twist column; Glen Research) and placed on a 394 synthesizer. TEAHF desilylation reagent was delivered at a flow rate of 1 ml/min for 1 min followed by a 29 min wait step during which the TEAHF reagent sat in the synthesis column. This delivery and wait was repeated 15 times for a total reaction time of 8 h and a total reagent consumption of 16 ml of TEAHF. Following TBDMS deprotection, the support in the column was washed with acetonitrile, the support was coupled to either A, dA or 2'-O-methyl-A phosphoramidites and extended by Cs. Couplings followed standard procedures (24).
2'-ACE deprotection and purification of inhibitors
Bis(2-acetoxyethoxy)methyl orthoester protecting groups were removed from the 2' hydroxyls of the transition state analogs as described (23). Analogs were purified by HPLC on a Microsorb C18 column (Varian) eluted with a 100 mM triethyl amine acetate (pH 7.0)/acetonitrile gradient (0–50% acetonitrile in 60 min; retention time = 20–25 min). Each product was confirmed by mass spectrometry and the proper mobility confirmed by gel electrophoresis. Mass spectra: CCdApOPmnphe: calc. for C49H63N17O24P3+ = 1365.31 Da; found m/z = 1365.23. CCApOPmnphe: calc. for C49H63N17O25P3+ = 1381.31 Da; found m/z = 1381.46. CCAomepOPmnphe: calc. for C50H61N17O25P3+ = 1395.32 Da; found m/z = 1395.38. CCdApOPmnala: calc. for C43H55N17O24P32+ (M+2H) = 1291.28 Da; found m/z =645.63, corresponding to 1291.26 Da. CCApOPmnala: calc. for C43H55N17O25P32+ (M+2H) = 1307.28 Da; found m/z = 653.60, corresponding to 1307.20 Da. CCAomepOPmnala: calc. for C44H57N17O25P32+ (M+2H) = 1321.29 Da; found m/z = 660.62, corresponding to 1321.24 Da.
Purification of ribosomes
Escherichia coli 50S ribosomal subunits were prepared from MRE600 cells as described (25) with the following modifications. Spin times were adjusted for a Beckman Ti 70 rotor. Following the first wash step, the ribosomes were resuspended in 20 mM Tris–HCl, pH 7.6, 0.5 M NH4Cl, 1.5 mM magnesium acetate, 0.5 mM EDTA, 7 mM 2-mercaptoethanol and incubated at 4°C for 4 h to dissociate the subunits. All subsequent wash steps were done with this low magnesium buffer. 50S and 30S particles were isolated by zonal centrifugation in a Beckman Ti 14 rotor on a 10–40% sucrose gradient (450 ml) in 20 mM Tris–HCl (pH 7.6), 60 mM NH4Cl, 5.2 mM magnesium acetate, 0.25 mM EDTA, and 3.0 mM 2-mercaptoethanol (45 000 r.p.m., 4 h 15 min, 4°C). 50S subunits were stored at –80°C in 70 mM NH4Cl, 30 mM KCl, 7 mM MgCl2, 0.5 mM EDTA, 1 mM DTT and 50 mM Tris–HCl (pH 7.6). Prior to all binding assays, ribosomes were activated by incubation at 42°C for 10 min in 200 mM KCl, 20 mM MgCl2, 50 mM Tris–HCl (pH 8.0).
Chemical modification of rRNA
23S rRNA was modified with 1-cyclohexyl-3-(2-morpholinoethyl)carbodiimide metho-p-toluenesulfonate (CMCT) essentially as described (8). 50S ribosomal subunits (10 nM) were incubated with varying concentrations of the transition state analogs at 0°C for 2 h in 200 mM KCl, 20 mM MgCl2, 33% methanol and 50 mM Tris–HCl (pH 8.0). Solid CMCT was added to each 100 μl reaction to a final concentration of 42 mg/ml and allowed to react at 0°C for 60 min. Reactions were quenched with 600 μl of ethanol, and the RNA was extracted. The extent of CMCT modification at U2585 was analyzed by primer extension as described (17). Inhibitor binding reduces accessibility of U2585 to CMCT, resulting in a decrease in band intensity. Loss of band intensity as a function of inhibitor concentration was used to determine the dissociation constants (Kd) of each inhibitor using the equation:
I = Isat + (I0 – Isat) / (1 + / Kd)
where I is band intensity, Isat is band intensity when inhibitor is saturating, and I0 is band intensity in the absence of inhibitor. All band intensities were internally normalized for overall extent of CMCT reactivity with the 23S rRNA and for gel loading, and subsequently normalized relative to I0, the band intensity with no inhibitor present.
RESULTS AND DISCUSSION
Studies on CCdApPmn have lead to structural and mechanistic models of the ribosomal active site. However, all inferences drawn from these studies are dependent on the behavior of a molecule that is missing an essential chemical group located close to the peptidyl transferase reaction center.
Inhibitor synthesis
Considering both the value and the shortcomings of CCdApPmn as a mimic of the peptidyl transferase transition state, we set out to prepare a set of analogs that would improve this interesting molecule (Fig. 2B). In particular, we wanted to include a hydroxyl group and other functional groups at the 2' position of A76. We also wanted the synthesis to be modular, so it could be easily modified to generate a series of related molecules. Toward this goal we redesigned the synthetic approach used for inhibitor preparation (Fig. 3). We settled upon a solid phase strategy that provided a maximum level of modularity for inhibitor design.
In order to ensure rapid, high yielding phosphoramidite coupling reactions during standard solid phase organic synthesis of the inhibitors, we chose to synthesize two new puromycin analogs with the -amine substituted with an -hydroxyl protected with a fluoride labile group. Although this substitution is not ideal from the perspective of transition state analogy, the -hydroxyl functional group substitution on puromycin has been shown in previous studies not to be detrimental to binding (26). A key step for this synthetic route was the preparation of TBDMS protected L-hydroxy acids linked to puromycin aminonucleoside (Fig. 3). Previous reports have shown that a reactive acyl chloride can be generated under mild conditions from a readily prepared TBDMS protected hydroxy ester (27). We were able to isolate in near quantitative yield both TBDMS protected L-lactic acid (1a) and TBDMS protected L-phenyl lactic acid (1b) that were of suitable purity for the conversion to the acyl chloride without subsequent purification. The 3'-amino group of the commercially available puromycin aminonucleoside proved to be satisfactorily reactive towards the acyl chloride generated in situ and yielded the desired amide linkage at the 3' position in acceptable yields (26,28,29). Transient protection of the 5' and 2' hydroxyls of puromycin amino nucleoside with trimethylsilyl chloride prior to the addition of the acyl chloride allowed the use of excess reagent to facilitate complete conversion of the starting nucleoside. Transient 5',2' protection of the TBDMS hydroxy puromycin derivatives (2a, 2b) under standard conditions, followed by coupling to a polystyrene solid support, made the two target resins suitable for coupling to a variety of commercially available phosphoramidites. This approach has several fundamental advantages over the solution phase synthesis of previous peptidyl transferase inhibitors. The silyl protection of the -hydroxyl group is readily removed under mild conditions. The use of fully protected phosphoramidites ensured a minimal number of side reactions, which improved the yield relative to the solution phase approach. Most importantly, the reaction scheme is modular. It can readily accommodate a variety of amino acids linked to the A-site puromycin, and the chemical composition of the nucleotide residues in the P-site sequence are easily varied during solid phase synthesis.
Inhibitor design
We synthesized a series of novel inhibitors based upon CCdApPmn as a lead compound (17). Each had CCA trinucleotides, simulating the 3' end of a P-site tRNA, linked through a phosphate to a puromycin-like moiety (Fig. 2B). Unlike CCdApPmn, which has a phosphoramidite linkage, all of these molecules possessed phosphodiester linkages because our use of conventional solid phase oligoribonucleotide synthesis required substitution of the puromycin -amino group with a hydroxyl moiety. The family of new molecules varied at the 2' position of A76, possessing the three chemical groups -H, -OH, -OCH3, and at the amino acid side chain on the puromycin-like moiety. One set of inhibitors had phenylalanine as the A-site R group, rather than methyl tyrosine as in CCdApPmn. The second set had alanyl side chains as the A-site R groups. In all, six new transition state analogs were synthesized: CCdApOPmnphe, CCApOPmnphe, CCAomepOPmnphe, CCdApOPmnala, CCApOPmnala and CCAomepOPmnala (Table 1).
Table 1. Binding affinities for each of the transition state inhibitors
With these molecules in hand, we determined how the modifications in the new inhibitors affect their affinity for the ribosomal peptidyl transferase center (Fig. 4). Chemical modification of the 23S rRNA by CMCT was used to monitor the binding of the inhibitors in the ribosomal active site, as has been used to measure binding of other peptidyl transferase ligands, including CCdApPmn (8,17). Protection of several sites in domain V of 23S rRNA (U2506, U2584 and U2585) from modification by CMCT has been linked to binding of tRNA in the P-site, and was also observed upon CCdApPmn binding (7,17). We measured the protection of U2585 from CMCT modification in varying concentrations of the inhibitors to determine their binding affinities. The complete set of binding constants is listed in Table 1.
Figure 4. Binding of transition state mimics in the peptidyl transferase center of E.coli based upon chemical modification of the 23S rRNA. Molecules bound in the active site protect U2585 from modification by CMCT, resulting in the disappearance of a CMCT-dependent reverse transcriptase stop. (A) A typical gel of reverse transcripts of 23S rRNA from ribosomes modified by CMCT in the presence of increasing concentrations of CCdApOPmnphe and (B) CCApOPmnphe. (C) U2585 band intensities were normalized for overall CMCT reactivity and gel loading and then relative to the intensity for no inhibitor present. Binding curves were fit to these data and dissociation constants were derived from the fits.
Two modifications to CCdApPmn were made for synthetic purposes: the substitution of the puromycin -amino group with a hydroxyl, and the substitution of phenylalanine for methyl-tyrosine. The first question we wished to address was: do these background changes in the inhibitor have a significant effect on their binding to the ribosome? CCdApOPmnphe, the molecule that is most similar to CCdApPmn, but includes both background modifications, exhibited a Kd of 53 nM, in close agreement with the previously published figure for CCdApPmn (70 nM) (17). CCdApPmn had a similar binding affinity when tested under our binding conditions (data not shown). The correspondence of the binding affinities of CCdApPmn and CCdApOPmnphe suggests that it is probably valid to assume that the background changes made to CCdApPmn have little, if any, effect on the binding of these molecules in the ribosomal active site. Alternatively, the similar binding constants may result from two inverse energetic effects that nearly cancel each other. If this were the case, then these differences would be consistent throughout the series of new inhibitors and these effects are unlikely to have a major influence on how these molecules are binding in the ribosome.
Once the legitimacy of comparing the behavior of the new set of transition state analogs with that of CCdApPmn was established, we wished to determine how the variations within this family of molecules affected their ribosomal affinity. Of primary interest was the involvement of the A76 2'-OH, which is absent from CCdApPmn. Surprisingly, the molecule with riboadenosine at position 76, CCApOPmnphe, had a Kd of 331 nM, 6-fold weaker than the deoxyadenosine version (G = –1 kcal/mol). No significant decomposition of these molecules was observed after incubation with ribosomes for 24 h, ruling out the possibility that the increased Kd resulted from degradation (data not shown).
A molecule containing a 2'-OCH3 at A76, CCAomepOPmnphe, was also synthesized and tested for binding. No significant protection of U2585 was observed at any concentration of CCAomepOPmnphe up to 5000 nM, indicating a Kd 5000 nM. This corresponds to at least a 100-fold decrease in binding (G –2.5 kcal/mol).
The structural conformation of the deoxy inhibitor cannot for steric reasons be assumed by the ribo-version of the inhibitor, but that the deoxy inhibitor binds with greater affinity. This suggests that the deoxy inhibitors assume a conformation that is energetically more stable, but less biologically relevant, at least in the case of the aromatic side chain. The presence of the 2'-OH forces CCApOPmnphe out of this conformation, reducing binding affinity. Presumably, the ribo-variant has assumed a different conformation within the PTC than the original inhibitor, a conformation that is more likely to reflect the actual transition state. The presence of a 2'-OCH3, which is considerably larger than a hydroxyl group, precludes binding altogether. Apparently the additional bulk of an O-methyl group cannot be accommodated in the local region of the peptidyl transferase center by any variation of a conformational change. The P-site A76 is expected to participate in a Type I A-minor interaction with A2450/C2501 (E.coli) (30). This interaction involves a close approach of the ribose sugar against the A-C pair and hydrogen bonding with the A76 2'-OH. A 2'-OCH3 on this base would be completely incompatible with this interaction.
Considering the rather unexpected observation that inclusion of a 2'-OH on A76, which makes the analog more similar to the actual transition state, decreases binding affinity, we sought an explanation from the 50S crystal structure with CCdApPmn bound in the active site. The structure shows that the methyl tyrosine side chain is stacked between two bases of the rRNA. This interaction could significantly affect binding of the molecule to the ribosome and influence the orientation of the phosphoramidate linkage. Previous studies showed that the amino acid side chain of puromycin provides a substantial contribution to the efficacy of puromycin as an inhibitor of peptidyl transferase, but that analogs containing non-aromatic side chains can function as inhibitors, particularly if the nucleic acid portion of the molecule is expanded to mimic more of the amino-acyl tRNA (31,32). Stacking of the methyl tyrosine, while energetically important for binding of the A-site portion of the inhibitor, may limit flexibility of nearby regions of the molecule and force the phosphate into the incorrect conformation observed for CCdApPmn.
To test the role of the A-site amino acid side chain in binding transition state analogs in the active site, we generated a set of transition state analogs that had alanine substituted for phenylalanine in the puromycin moiety (Fig. 2B). In particular, we wanted to determine how changing the amino acid and eliminating the stacking interaction would affect the ribosome’s ability to accommodate 2'-OH substitution at A76. When the amino acid side chain is replaced with alanine, the observed binding constants were 250 and 280 nM for CCdApOPmnala and CCApOPmnala, respectively. Again, no binding was observed with the 2'-O-methyl substituted variant CCAomepOPmnala up to 5000 nM. The reduction in binding affinity from CCdApOPmnphe to CCdApOPmnala is likely due to the loss of the stacking interactions between the aromatic phenylalanine side chain and A2451 and C2452 (E.coli) of the 23S rRNA. Significantly, the binding of the A76 riboadenosine and deoxyadenosine analogs were approximately the same for the alanyl derivatives. Based only on the comparison of inhibitors containing riboadenosine and deoxyriboadenosine, it could be argued that steric clash between the phosphate oxygen and the 2'-hydroxyl only affects binding in the case of an aromatic A-site amino acid side chain and is not a general effect. The elimination of the phenylalanine stacking interaction may increase the conformational flexibility of the entire molecule, allowing the tetrahedral phosphate to move in a way that would accommodate the 2'-hydroxyl at A76. However, the elimination of binding upon inclusion of an O-methyl group, regardless of the identity of the amino acid side chain, indicates that additional flexibility afforded to non-aromatic A-site substrates cannot compensate for the additional steric bulk of a 2'-O-methyl group. There is a finite space in the vicinity of the A76 2'-carbon available for occupancy by various functional groups. This space must not be substantially larger than the hydroxyl group found in native substrates given that the only moderately larger O-methyl group cannot be accommodated. Because the hydroxyl group must nearly fill the available space, its presence will certainly have a significant effect on the conformation of bound substrates or transition state analogs, even in the case of non-aromatic A-site side chains.
CONCLUDING REMARKS
We have presented a synthetic scheme for efficiently producing transition state analogs of the peptidyl transferase reaction. One important development compared with previously studied transition state analogs is the inclusion of the hydroxyl group at A76 of the P-site tRNA. Although the binding affinities are not improved, these molecules will be of particular interest in structural studies, to help refine the mechanism of the ribosome active site during chemistry. There are substantial steric constraints placed on the tetrahedral phosphate by the now present 2'-OH of A76. Therefore, we expect a significant rearrangement of the conformation of the phosphate from that seen when CCdApPmn is bound in the ribosomal active site. Such structures will be more relevant to the reality of peptidyl transferase catalysis by the ribosome.
In addition to allowing inclusion of the critical A76 2'-OH, our synthetic scheme imparts modularity to the production of inhibitors, allowing the synthesis of a large variety of related molecules. It particularly affords flexibility in the tetrahedral phosphate and P-site portions of the molecules. Among other changes, we will be able to synthesize transition state analogs with chiral centers at their tetrahedral phosphates. This will make it possible to resolve the stereochemical ambiguity of the peptidyl transferase transition state.
ACKNOWLEDGEMENTS
We thank Mark Parnell and Kevin Huang for comments on the manuscript. We thank the laboratory of Andrew D. Miranker for assistance with mass spectrometry. This work was supported by American Cancer Society Beginning Investigator grant RSG-02-052-GMC to S.A.S., a Research Scholar Grant PF-013-01-GMC from the American Cancer Society to G.W.M. and a National Science Foundation Predoctoral Fellowship to J.S.W.
REFERENCES
Chladek,S. and Sprinzl,M. (1985) The 3'-end of transfer-RNA and its role in protein biosynthesis. Angew. Chem. Int. Ed., 24, 371–391.
Yarus,M. and Welch,M. (2000) Peptidyl transferase: ancient and exiguous. Chem. Biol., 7, R187–R190.
Green,R. and Noller,H.F. (1997) Ribosomes and translation. Annu. Rev. Biochem., 66, 679–716.
Ban,N., Nissen,P., Hansen,J., Moore,P.B. and Steitz,T.A. (2000) The complete atomic structure of the large ribosomal subunit at 2.4 ? resolution. Science, 289, 905–920.
Barta,A., Steiner,G., Brosius,J., Noller,H.F. and Kuechler,E. (1984) Identification of a site on 23s ribosomal-RNA located at the peptidyl transferase center. Proc. Natl Acad. Sci. USA, 81, 3607–3611.
Moazed,D. and Noller,H.F. (1987) Chloramphenicol, erythromycin, carbomycin and vernamycin B protect overlapping sites in the peptidyl transferase region of 23S ribosomal RNA. Biochimie, 69, 879–884.
Moazed,D. and Noller,H.F. (1989) Interaction of tRNA with 23S rRNA in the ribosomal A, P and E sites. Cell, 57, 585–597.
Moazed,D. and Noller,H.F. (1991) Sites of interaction of the CCA end of peptidyl-tRNA with 23S rRNA. Proc. Natl Acad. Sci. USA, 88, 3725–3728.
Bocchetta,M., Xiong,L.Q. and Mankin,A.S. (1998) 23S rRNA positions essential for tRNA binding in ribosomal functional sites. Proc. Natl Acad. Sci. USA, 95, 3525–3530.
Samaha,R.R., Green,R. and Noller,H.F. (1995) A base pair between tRNA and 23S rRNA in the peptidyl transferase centre of the ribosome. Nature, 377, 309–314. Erratum in: Nature (1995), 378, 419.
Kim,D.F. and Green,R. (1999) Base-pairing between 23S rRNA and tRNA in the ribosomal A site. Mol. Cell, 4, 859–864.
Takai,K., Horie,N., Yamaizumi,Z., Nishimura,S., Miyazawa,T. and Yokoyama,S. (1994) Recognition of UUN codons by two leucine tRNA species from Escherichia coli. FEBS Lett., 344, 31–34.
Quiggle,K., Kumar,G., Ott,T.W., Ryu,E.K. and Chladek,S. (1981) Donor site of ribosomal peptidyltransferase: investigation of substrate specificity using 2'(3')-O-(N-acylaminoacyl)dinucleoside phosphates as models of the 3' terminus of N-acylaminoacyl transfer ribonucleic acid. Biochemistry, 20, 3480–3485.
Ringer,D., Quiggle,K. and Chladek,S. (1975) Recognition of the 3' terminus of 2'-O-aminoacyl transfer ribonucleic acid by the acceptor site of ribosomal peptidyltransferase. Biochemistry, 14, 514–520.
Wagner,T., Cramer,F. and Sprinzl,M. (1982) Activity of the 2' and 3' isomers of aminoacyl transfer ribonucleic acid in the in vitro peptide elongation on Escherichia coli ribosomes. Biochemistry, 21, 1521–1529.
Bhuta,A., Quiggle,K., Ott,T., Ringer,D. and Chladek,S. (1981) Aminoacyl derivatives of nucleosides, nucleotides and polynucleotides. 33. Stereochemical control of ribosomal peptidyltransferase reaction—role of amino-acid side-chain orientation of acceptor substrate. Biochemistry, 20, 8–15.
Welch,M., Chastang,J. and Yarus,M. (1995) An inhibitor of ribosomal peptidyl transferase using transition-state analogy. Biochemistry, 34, 385–390.
Wolfenden,R., Rammler,D.H. and Lipmann,F. (1964) On the site of esterification of amino acids to soluble RNA. Biochemistry, 3, 329–338.
Nissen,P., Hansen,J., Ban,N., Moore,P.B. and Steitz,T.A. (2000) The structural basis of ribosome activity in peptide bond synthesis. Science, 289, 920–930.
Hansen,J.L., Schmeing,T.M., Moore,P.B. and Steitz,T.A. (2002) Structural insights into peptide bond formation. Proc. Natl Acad. Sci. USA, 99, 11670–11675.
Ramakrishnan,V. (2002) Ribosome structure and the mechanism of translation. Cell, 108, 557–572.
Parnell,K.M., Seila,A.C. and Strobel,S.A. (2002) Evidence against stabilization of the transition state oxyanion by a pKa-perturbed RNA base in the peptidyl transferase center. Proc. Natl Acad. Sci. USA, 99, 11658–11663.
Scaringe,S.A. (2001) RNA oligonucleotide synthesis via 5'-silyl-2'-orthoester chemistry. Methods, 23, 206–217.
Scaringe,S.A., Wincott,F.E. and Caruthers,M.H. (1998) Novel RNA synthesis method using 5'-O-silyl-2'-O-orthoester protecting groups. J. Am. Chem. Soc., 120, 11820–11821.
Rodnina,M.V. and Wintermeyer,W. (1995) GTP consumption of elongation factor Tu during translation of heteropolymeric mRNAs. Proc. Natl Acad. Sci. USA, 92, 1945–1949.
Katunin,V.I., Muth,G.W., Strobel,S.A., Wintermeyer,W. and Rodnina,M.V. (2002) Important contribution to catalysis of peptide bond formation by a single ionizing group within the ribosome. Mol. Cell, 10, 339–346.
Wissner,A. and Grudzinskas,C.V. (1978) Reaction of tert-butyldimethylsilyl esters with oxalyl chloride-dimethylformamide: preparation of carboxylic acid chlorides under neutral conditions. J. Org. Chem., 43, 3972–3974.
Kelly,S.E. and LaCour,T.G. (1992) A one pot procedure for the synthesis of alpha-hydroxy amides from the corresponding alpha-hydroxy acids. Synth. Commun., 22, 859–869.
Trost,B.M. and McDougal,P.G. (1982) Synthesis of optically active verrucarinic acid derivatives. Tetrahedron Lett., 23, 5497–5500.
Nissen,P., Ippolito,J.A., Ban,N., Moore,P.B. and Steitz,T.A. (2001) RNA tertiary interactions in the large ribosomal subunit: the A-minor motif. Proc. Natl Acad. Sci. USA, 98, 4899–4903.
Fong,K.L. and Vince,R. (1978) Peptidyl transferase substrate specificity with nonaromatic aminoacyl analogues of puromycin. J. Med. Chem., 21, 792–795.
Symons,R.H., Harris,R.J., Clarke,L.P., Wheldrake,J.F. and Elliott,W.H. (1969) Structural requirements for inhibition of polyphenylalanine synthesis by aminoacyl and nucleotidyl analogues of puromycin. Biochim. Biophys. Acta, 179, 248–250.(Joshua S. Weinger1, David Kitchen3, Step)
*To whom correspondence should be addressed. Tel: +1 203 432 9772; Email: strobel@csb.yale.edu
ABSTRACT
All living cells are dependent on ribosomes to catalyze the peptidyl transfer reaction, by which amino acids are assembled into proteins. The previously studied peptidyl transferase transition state analog CC-dA-phosphate-puromycin (CCdApPmn) has important differences from the transition state, yet current models of the ribosomal active site have been heavily influenced by the properties of this molecule. One significant difference is the substitution of deoxyadenosine for riboadenosine at A76, which mimics the 3' end of a P-site tRNA. We have developed a solid phase synthetic approach to produce inhibitors that more closely match the transition state, including the critical P-site 2'-OH. Inclusion of the 2'-OH or an even bulkier OCH3 group causes significant changes in binding affinity. We also investigated the effects of changing the A-site amino acid side chain from phenylalanine to alanine. These results indicate that the absence of the 2'-OH is likely to play a significant role in the binding and conformation of CCdApPmn in the ribosomal active site by eliminating steric clash between the 2'-OH and the tetrahedral phosphate oxygen. The conformation of the actual transition state must allow for the presence of the 2'-OH, and transition state mimics that include this critical hydroxyl group must bind in a different conformation from that seen in prior analog structures. These new inhibitors will provide valuable insights into the geometry and mechanism of the ribosomal active site.
INTRODUCTION
Ribosomes are the macromolecular machines responsible for synthesizing proteins in all living cells. The large ribosomal subunit (50S in prokaryotes) contains the site of catalysis, the peptidyl transferase center. The reaction substrates include a peptidyl-tRNA, charged with the growing peptide chain bound to a tRNA binding site on the ribosome, termed the P-site, and an aminoacyl-tRNA, charged with a single amino acid bound to a second location on the ribosome, termed the A-site. Peptide bond formation involves aminolysis of the P-site ester by the A-site -amino group (Fig. 1). A large body of biochemical work, in conjunction with the recently determined high resolution crystal structures of the 50S subunit has lead to a substantial chemical insight regarding ribosomal function (1–4).
Figure 1. The peptidyl transferase reaction, including the theoretical transition state. Peptidyl transferase occurs when the -amino group of the A-site amino acid (in this case, tyrosine) nucleophilically attacks the carboxyl carbon of the P-site nascent peptide’s C-terminal amino acid. The resulting transition state contains a tetrahedral carbon with a single oxyanion. Subsequent release of the carboxyl carbon from the P-site tRNA yields the reaction products, a deacylated P-site tRNA and an N + 1 length peptide linked to the A-site tRNA.
Several previous biochemical studies of the peptidyl transferase center have focused on the interaction between the 50S subunit and small molecules that act as peptidyl transferase substrates or inhibitors, as well as amino-acylated tRNAs (5–11). Such studies identified regions of the 23S rRNA closely associated with the peptidyl transferase center, including the A and P loops. Other studies have addressed the stereochemistry, composition and significance of the amino acid acceptors (1). The importance of specific substrate functional groups, particularly the 2'-OH and 3'-OH of A76, for activity as amino acid donors or acceptors and during translocation has also been investigated (12–16).
Most of these early biochemical studies analyzed individual ribosomal substrates, or inhibitors mimicking individual substrates. Welch and colleagues synthesized an inhibitor that simulated the simultaneous binding of both A-site and P-site charged tRNA fragments, intended to mimic the reaction’s transition state (17). This inhibitor, CC-dA-phosphate- puromycin (CCdApPmn), consists of the antibiotic puromycin (Pmn) covalently linked via a bridging phosphoramidate to the trinucleotide CCdA (Fig. 2A). The puromycin portion of the molecule, which is a dimethyl-adenosine linked by a non-hydrolyzable amide bond to the amino acid methyl-tyrosine, binds in the A-site and imitates the 3' end of an amino-acylated A-site tRNA. The CCdA trinucleotide binds in the P-site and corresponds to the CCA end of a P-site tRNA. The bridging phosphoramidate group was designed to mimic the theoretical tetrahedral transition state of the peptidyl transferase reaction (17).
Figure 2. Peptidyl transferase transition state analogs. (A) CCdApPmn. The puromycin portion of the molecule mimics the terminal adenosine and amino acid of an amino-acylated A-site tRNA. The trinucleotide CCdA mimics the 3' end of a P-site tRNA, and the tetrahedral phosphate mimics the tetrahedral carbon of the transition state. Significant differences between the actual transition state and this molecule include the absence of a 2'-OH from the P-site A and the distribution of the negative charge between two oxygen atoms. (B) Improved transition state analogs. Six new molecules were synthesized, varying in their P-site A76 2' functional group (R1) and their A-site amino acid R group (R2). (C) The structure of CCdApPmn bound in the peptidyl transferase center of 50S ribosomal subunits from H.marismortui. CCdApPuro is shown in green, and nearby rRNA bases are shown in gray. rRNA base numbers are H.marismortui numbering with E.coli equivalents in parentheses. One of the non-bridging oxygens of the tetrahedral phosphate is within 2.8 ? of the 2' carbon of the P-site A76, shown in red. The methyl tyrosine portion is stacked between A2486 and U2487.
Transition state analogs can be extremely useful tools for examining the mechanism by which enzymes catalyze their reactions (18). Consistent with this expectation, a number of critical insights into ribosomal catalysis have been achieved by exploring CCdApPmn binding. The tight interaction of CCdApPmn with the ribosome and its inhibition of peptidyl transferase supported the nucleophilic attack based reaction scheme by demonstrating that a mimic of the transition state can bind in the active site, and that the A-site and P-site are sufficiently proximal to simultaneously accommodate this molecule (17). Subsequent structural studies on the 50S ribosomal subunit used CCdApPmn to identify the location of the peptidyl transferase center. The observation that no portion of any ribosomal protein was within 18 ? of the tetrahedral phosphate provided evidence that the ribosomal RNA is the catalytic agent of the ribosome (19).
The positioning of CCdApPmn within the active site has led to detailed mechanistic models of ribosome function involving specific rRNA functional groups (2,19–21). For example, in the co-crystal structure, the positioning of A2451 relative to one of the non-bridging phosphate oxygens of the tetrahedral phosphate (Fig. 2C) was said to suggest a possible role for this nucleotide as a general acid/base, or it suggested involvement of A2451 in stabilization of the transition state oxyanion (19). Later, studies showing that CCdApPmn binding is pH independent were used to discredit a model for ribosomal catalysis that includes oxyanion stabilization (22). However, there are some serious caveats to any mechanistic conclusions based upon the CCdApPmn inhibitor.
Despite its clear utility for structural mapping of the peptidyl transferase center, CCdApPmn is not an ideal transition state analog. It has become apparent that differences between CCdApPmn and the transition state are likely to be functionally important. For example, in contrast to the single oxyanion of the actual transition state, the charge of CCdApPmn is delocalized between the two non-bridging oxygens. Also, because the two non-bridging oxygens are chemically identical, the tetrahedral center of CCdApPmn is achiral, which creates stereochemical ambiguity in CCdApPmn’s mimicry of the transition state. Perhaps the most significant problem with the inhibitor is the substitution of deoxyadenosine in the position corresponding to A76 of the P-site tRNA. tRNAs containing a deoxyadenosine at position 76 are inactive as P-site substrates (13). In addition to eliminating a potentially important functional group from the 2' position of the adenosine, this change also results in a different sugar pucker in the ribose ring. This problem is illustrated by the crystal structure of CCdApPmn bound in the peptidyl transferase center, wherein one of the non-bridging phosphate oxygens is within 2.8 ? of the deoxyadenosine A76 C2' (Fig. 2C). This oxygen is within the van der Waals radius of the position expected to be occupied by the omitted 2'-OH, suggesting that the phosphate cannot occupy the same position in the context of a ribo-adenosine at the A76 P-site (19,22).
The conformation of CCdApPmn in crystals of the Haloarcula marismortui 50S subunit also has some critical differences from a model of the transition state that was extrapolated from the positions of independently placed A-site and P-site substrates. The differences are particularly evident with regard to the position of the tetrahedral phosphate (20). This model suggests that the original assignment of the two non-bridging oxygens of CCdApPmn to the oxyanion and the growing peptide chain were reversed. Instead of pointing toward the N3 of A2451 and being within hydrogen bonding distance, the oxyanion is pointing away from A2451 and is not in obvious hydrogen bonding distance of any base. This argues against a direct stabilizing interaction between the oxyanion and the ribosomal RNA. These factors bring into question the relevance of CCdApPmn’s conformation within the peptidyl transferase center and suggest important mechanistic information could be obtained if improved inhibitors could be developed.
The synthesis of transition state analogs with a 2'-OH on the P-site A76 calls for an improved synthetic scheme. The solution phase synthetic route used to prepare CCdApPmn involved the coupling of CCdAp with puromycin. The 3'-phosphate of CCdAp was activated with 1-ethyl-3-carbodiimide which facilitated nucleophilic attack by the -amino group of puromycin or by water. Welch et al. (17) elected to substitute the terminal A with 2'-deoxyadenosine because the adenosine 2'-OH would preferentially attack the activated 3'-phosphate to produce a trinucleotide with a cyclic phosphate instead of puromycin. This eliminated some of the potential competing reactions, and while the yields were low, it was possible to generate a sufficient amount of the inhibitor for biochemical and structural studies (17). However, this synthetic approach is not a general scheme that could be applied readily for the preparation of variants of the original inhibitor.
In order to generate improved transition state analogs and eliminate the problems associated with established synthetic methods, we have employed a solid phase synthetic strategy (Fig. 3). This approach is compatible with 2'-bis(2-acetoxyethoxy)methyl (ACE) protecting group chemistry (23), thus making it possible to include the 2'-OH of A76 that was omitted in the original inhibitor. We designed two unique A-site nucleosides coupled to conventional polystyrene solid supports. A family of inhibitors based upon CCdApPmn was prepared by coupling standard nucleoside phosphoramidites to the supports. These compounds have made it possible to examine interactions with the amino acid side chain and explore interactions with the P-site ribose substituent. Here we report the solid phase organic synthesis of peptidyl transferase inhibitors and the characterization of inhibitor binding to the ribosomal peptidyl transferase center.
Figure 3. Solid phase synthesis of transition state analogs. The initial reaction shown here is the coupling of TMS protected puromycin aminonucleoside with the acyl chloride of L-hydroxy acids where R1 = CH3 (alanine analogs) or CH2C6H5 (phenylalanine analogs). Standard protection and deprotection conditions afforded compounds 5a and 5b in an overall yield of 42–48% from the starting nucleoside. Polymer bound puromycin was obtained by coupling compounds 5a and 5b to N-terminus polystyrene supports using standard techniques. The fluoride labile silyl protecting group was removed with TEMED/HF in MeCN prior to the first phosphoramidite reaction. Three separate adenosine phosphoramidites were followed by cytidine phosphoramidites to generate the desired addition of CCA to the A-site puromycin analog. Oxidation, deprotection and cleavage reactions were all conducted using established protocols.
MATERIALS AND METHODS
General
Solvents were reagent grade and dried prior to use. Polystyrene solid support was purchased from Amersham. All other chemicals were purchased from Aldrich and used without further purification. 1H-NMR spectra were recorded on a 400 MHz spectrometer using CDCl3 as an internal standard. HPLC was carried out on a C18 reversed-phase column (Rainin), which was eluted with a 100 mM triethylammonium acetate/acetonitrile gradient. HRMS data were obtained from the Keck Foundation Biotechnology Resource Center at Yale University. Mass spectrometry was performed using a Micromass LCT electrospray time of flight spectrometer.
TBDMS protected L-lactic acid (1a) and L-phenyl lactic acid (1b)
To a stirred solution of either L-lactic acid or L-phenyl lactic acid (3.44 mmol) with imidazole (10.3 mmol) in 10 ml of anhydrous dimethylformamide was added 1.04 g (6.88 mmol) of tert-butyldimethylsilyl chloride. The reaction was allowed to proceed overnight at room temperature after which time the solution was diluted with hexanes and washed once with 50 ml of water, once with 50 ml of saturated NaHCO3 and once with 50 ml of brine. The organic layer was dried over Na2SO4 and the solvent was removed under reduced pressure to yield colorless oils. Compound 1a (0.88 g, 2.7 mmol) 80% yield; compound 1b (1.1 g, 2.9 mmol) 85% yield. The products were used in subsequent reactions without further purification.
Compound 1a. 1H-NMR CDCl3 0.034 (s, 3H, Si-CH3), 0.070 (s, 3H, Si-CH3), 0.243 (s, 6H, Si-CH3), 0.866 (s, 9H, Si-C(CH3)3), 0.909 (s, 9H, Si-C(CH3)3), 1.36 (d, 3H, CH3), 4.22 (q, 1H, CH).
Compound 1b. 1H-NMR CDCl3 –0.001 (s, 3H), 0.128 (s, 3H), 0.453 (s, 3H), 0.473 (s, 3H), 1.02 (s, 9H), 1.14 (s, 9H), 3.09 (m, 1H), 3.28 (m, 1H), 4.49 (m, 1H), 7.46 (m, 5H).
3'-N-(TBDMS-hydroxy)-amidopuromycin (2a and 2b)
The TBMDS protected hydroxy acids (1a, 1b) were activated for coupling by adding oxalyl chloride (0.68 ml, 7.8 mmol) to a stirred solution of either TBDMS protected L-lactic acid (1a) (1.66 g, 5.2 mmol) or L-phenyl lactic acid (1b) (2.0 g, 5.2 mmol) in 5 ml of anhydrous methylene chloride with catalytic N,N-dimethylformamide (27). After 2 h at room temperature, this solution was added dropwise to a solution of TMS protected puromycin aminonucleoside prepared by adding trimethylsilyl chloride (0.5 ml, 3.9 mmol) to a solution of puromycin aminonucleoside (0.5 g, 1.17 mmol) in 5 ml of anhydrous pyridine (room temperature, 2 h). The N-acylation reaction was complete after 1 h at room temperature upon which time the reactions were quenched by adding 0.78 g (4 mmol) of citric acid in 10 ml of methanol followed by 5 ml of 50 mM K2CO3 in methanol. Evaporation of the solvent followed by purification on silica with a 0–5% methanol in CH2Cl2 gradient yielded products 2a and 2b as white solids. Compound 2a (0.46 g, 0.95 mmol) 81% yield; compound 2b (0.47 g, 0.84 mmol) 72% yield.
Compound 2a. 1H-NMR CDCl3 0.104 (s, 6H, Si-CH3), 0.911 (s, 9H, Si-C(CH3)3), 1.37 (d, 3H, CH3), 3.83 (d, 1H, H5'), 3.97 (d, 1H, H5'), 4.22 (m, 2H, unresolved CH, H4'), 4.41 (m, 1H, H3'), 4.65 (m, 1H, H2'), 5.89 (d, 1H, H1'), 7.62 (d, 1H, NHCO), 8.09 (s, 1H, H8), 8.16 (s, 1H, H2); MS calc. for C21H36N6O5SiNa: 503.2414, found M+Na 503.2405. Rf = 0.2 (silica, 5% MeOH/CH2Cl2).
Compound 2b. 1H-NMR CDCl3 –0.25 (s, 3H, Si-CH3), –0.08 (s, 3H, Si-CH3), 0.87 , 2.85 (q, 1H, CH2), 3.05 (dd, 1H, CH2), 3.80 (m, 1H, H5'), 4.00 (m, 1H, H5'), 4.12 (m, 1H, H4'), 4.32 (m, 1H, H3'), 4.38 (m, 1H, H2'), 4.47 (m, 1H, CH), 5.82 (d, 1H, H1'), 7.18 (m, 5H, phe), 8.06 (s, 1H, H8), 8.20 (s, 1H, H2); MS calc. for C27H41N6O5Si: 557.2907, found M+H 557.2908. Rf = 0.3 (silica, 1:1 CH2Cl2/EtOAc).
5'-Dimethoxytrityl-2'-acetyl-3'-N-(TBDMS-hydroxy)-amidopuromycin (3a and 3b)
4,4-Dimethoxytrityl chloride (0.63 g, 1.88 mmol) was added to a stirred solution of 3'-N-(TBDMS-L-lactic acid)-amidopuromycin (2a) or 3'-N-(TBDMS-L-phenyllactic acid)-amidopuromycin (2b) (0.6 g, 1.25 mmol) in anhydrous pyridine at 0°C. The solution was slowly warmed to room temperature. After 3 h, TLC (silica, 5% MeOH/CH2Cl2) showed complete conversion to a faster moving product (Rf = 0.3–0.4). The solution was cooled to 0°C and dimethylaminopyridine (0.02 g, 0.017 mmol) and acetic anhydride (0.6 ml, 6.6 mmol) were added. The reactions continued overnight at room temperature. The solvent was removed under reduced pressure and the products were isolated as white solids by flash chromatography (silica, 1:1 hexanes/ethyl acetate). Compound 3a (0.74 g, 0.90 mmol) 72% yield; compound 3b (0.83 g, 0.92 mmol) 74% yield.
Compound 3a. 1H-NMR CDCl3 –0.039 (s, 3H, Si-CH3), 0.055 (s, 3H, Si-CH3), 0.869 (s, 9H, Si-C(CH3)3), 1.30 (d, 3H, CH3), 2.13 (s, 3H, COCH3), 3.33 (q, 1H, H5'), 3.41 (dd, 1H, H5'), 3.50 (broad s, 6H, N(CH3)2), 3.74 (s, 6H, OCH3), 4.12 (m, 2H, unresolved H4', H3'), 5.14 (m, 1H, CH), 5.72 (dd, 1H, H2'), 6.15 (d, 1H, H1'), 6.74 (m, 4H, DMT), 6.83 (d, 1H, NHCO), 7.27 (m, 4H, DMT), 7.38 (m, 2H, DMT), 7.90 (s, 1H, H8), 8.29 (s, 1H, H2); MS calc. for C44H56N6O8SiNa: 847.3827, found M+Na 847.3840. Rf = 0.2 (silica, 1:1 hexanes/EtOAc).
Compound 3b. 1H-NMR CDCl3 –0.28 (s, 3H, Si-CH3), –0.21 (s, 3H, Si-CH3), 0.83 (s, 9H, Si-C(CH3)3), 1.97 (s, 3H, COCH3), 2.89 (q, 1H, CH2), 2.97 (dd, 1H, CH2), 3.28 (m, 1H, H5'), 3.38 (m, 1H, H5'), 3.57 (broad s, 6H, N(CH3)2), 3.75 (s, 6H, OCH3), 4.07 (m, 1H, H4'), 4.23 (m, 1H, H3'), 5.14 (m, 1H, CH), 5.28 (m, 1H, H2'), 6.57 (m, 1H, NHCO), 6.75 (m, 4H, DMT), 7.22 (m, 12H, unresolved DMT, phe), 7.37 (m, 2H, DMT), 7.95 (s, 1H, H8), 8.38 (s, 1H, H2); MS calc. for C50H61N6O8Si: 901.4320, found M+H 901.4321. Rf = 0.2 (silica, 1:1 hexanes/EtOAc).
2'-Acetyl-3'-N-(TBDMS-hydroxy)-amidopuromycin (4a and 4b)
To a solution of 5'-dimethoxytrityl-2'-acetyl-3'-N-(TBDMS-lactic acid)-amidopuromycin (3a) (0.74 g, 0.9 mmol) or 5'-dimethoxytrityl-2'-acetyl-3'-N-(TBDMS-phenyl lactic acid)-amidopuromycin (3b) (0.94 g, 1.04 mmol) in anhydrous CH2Cl2 at 0°C was added dropwise trifluoroacetic acid (0.4 ml, 5 mmol). After 30 min the bright orange solution was extracted with 1 x 30 ml 30% NaHCO3 and 1 x 30 ml H2O. The organic phase was dried over Na2SO4 and the solvent removed under reduced pressure. Flash chromatography (silica, 2% MeOH/CH2Cl2) yielded products 4a and 4b, respectively, as crystalline white solids. Compound 4a (0.43 g, 0.78 mmol) 91% yield; compound 4b (0.58 g, 0.97 mmol) 92% yield.
Compound 4a. 1H-NMR CDCl3 0.126 (s, 3H, Si-CH3), 0.137 (s, 3H, Si-CH3), 0.945 (s, 9H, Si-C(CH3)3), 1.38 (d, 3H, CH3), 2.12 (s, 3H, COCH3), 3.50 (broad s, 6H, N(CH3)2), 3.78 (m, 1H, H5'), 4.01 (m, 1H, H5'), 4.26 (m, 2H, H4', H3'), 4.95 (m, 1H, CH), 5.73 (m, 1H, H2'), 6.05 (d, 1H, H1'), 7.13 (d, 1H, NHCO), 7.81 (s, 1H, H8), 8.30 (s, 1H, H2); MS calc. for C23H38N6O6SiNa: 545.2520, found M+Na 545.2518. Rf = 0.1 (silica, 1:1 hexanes/EtOAc).
Compound 4b. 1H-NMR CDCl3 –0.10 (s, 3H, Si-CH3), –0.01 (s, 3H, Si-CH3), 0.91 (s, 9H, Si-C(CH3)3), 1.94 (s, 3H, COCH3), 2.97 (m, 2H, CH2), 3.44 (broad s, 6H, N(CH3)2), 3.74 (m, 1H, H5'), 3.95 (m, 1H, H5'), 4.08 (d, 1H, H4'), 4.39 (t, 1H, H3'), 4.87 (dd, 1H, CH), 5.41 (m, 1H, H2'), 5.96 (d, 1H, H1'), 6.76 (d, 1H, NHCO), 7.20 (m, 5H, phe), 7.82 (s, 1H, H8), 8.33 (s, 1H, H2); MS calc. for C29H43N6O6Si: 599.3013, found M+H 599.3005. Rf = 0.3 (silica, 5% MeOH/CH2Cl2).
5'-Succinyl-2'-acetyl-3'-N-(TBDMS-hydroxy)-amidopuromycin (5a and 5b)
A mixture of 2'-acetyl-3'-N-(TBDMS-lactic acid)-amidopuromycin (4a) (0.43 g, 0.82 mmol) or 2'-acetyl-3'-N-(TBDMS-phenyl lactic acid)-amidopuromycin (4b) (0.56 g, 0.93 mmol), succinic anhydride (0.15 g, 1.5 mmol or 0.93 g, 9.35 mmol), dimethylaminopyridine (0.17 g, 1.4 mmol) in anhydrous pyridine (10 ml) was stirred at room temperature for 2 h. TLC (silica, 5% methanol/CH2Cl2) showed clean complete reactions. Following removal of the solvent in vacuo, the residues were dissolved in CH2Cl2 and washed twice with 30 ml of 10% (w/v) citric acid, once with 30 ml of brine and once with 30 ml of H2O. The organic phase was dried over Na2SO4 and the solvent removed under reduced pressure. Chromatography on silica (0–5% methanol in CH2Cl2) afforded products 5a and 5b as crystalline white solids. Compound 5a (0.49 g, 0.79 mmol) 96% yield; compound 5b (0.57 g, 0.81 mmol) 87% yield.
Compound 5a. 1H-NMR CDCl3 0.116 (s, 6H Si-CH3), 0.933 (s, 9H, Si-C(CH3)3), 1.34 (d, 3H, CH3), 2.16 (s, 3H, COCH3), 2.65 (m, 4H, CH2CH2), 3.49 (broad s, 6H, N(CH3)2), 4.11 (m, 1H, H5'), 4.21 (m, 2H, H5', H4'), 4.51 (dd, 1H, H3'), 5.26 (m, 1H, CH), 5.68 (dd, 1H, H2'), 6.06 (s, 1H, H1'), 7.07 (d, 1H, NHCO), 7.96 (s, 1H, H8), 8.34 (s, 1H, H2); MS calc. for C27H42N6O9SiNa: 645.2680, found M+Na 645.2682. Rf = 0.2 (silica, 5% MeOH/CH2Cl2).
Compound 5b. 1H-NMR CDCl3 –0.09 (s, 3H Si-CH3), 0.00 (s, 3H Si-CH3), 0.90 (s, 9H, Si-C(CH3)3), 1.97 (s, 3H, COCH3), 2.63 (s, 4H, CH2CH2), 2.97 (m, 2H, CH2), 3.44 (broad s, 6H, N(CH3)2), 4.00 (m, 1H, H5'), 4.19 (m, 1H, H5'), 4.40 (m, 1H, H4'), 4.50 (m, 1H, H3'), 5.25 (m, 2H, H2', CH, unresolved), 5.94 (d, 1H, H1'), 6.75 (d, 1H, NHCO), 7.14 (m, 5H, phe), 7.91 (s, 1H, H8), 8.35 (s, 1H, H2); MS calc. for C33H47N6O9Si: 699.3174, found M+H 699.3168. Rf = 0.2 (silica, 5% MeOH/CH2Cl2).
Derivatization of solid support
Nucleosides 5a and 5b were coupled to LCCA-polystyrene solid supports (1.0 g) (Amersham) by dissolving 5a or 5b (0.2 mmol) in anhydrous pyridine (10 ml) with triethylamine (0.08 ml), dimethylaminopyridine (0.012 g, 0.1 mmol) and 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (0.38 g, 2 mmol). The reactants along with the support were placed in 40 ml conical glass tubes and rocked gently at room temperature for 24 h. The reactions were filtered through a sintered glass funnel and washed once with 20 ml of pyridine, once with 20 ml of methanol and once with 20 ml of CH2Cl2 and then air dried. Nucleoside loading was determined by reacting a small portion of the support (5 mg) with 50 mM K2CO3 in methanol and measuring the optical absorbance at 275 nm. This gave an estimated nucleoside loading of 92 μmol/g for compound 5a and 147 μmol/g for compound 5b. High loadings could be employed in this case because the synthetic targets were quite short. Unreacted sites on the support were capped by adding 5 ml of 1 M acetic anhydride, 1 M 2,6-lutidine in tetrahydrofuran and 5 ml of 1 M N-methyl imidazole in tetrahydrofuran and rocking gently at room temperature for 3 h. The support was then filtered and washed with 20 ml of methanol and 20 ml of CH2Cl2 and air dried.
Solid phase synthesis of inhibitors
The puromycin derivatized support (80 mg) was packed into oligo synthesis columns (Twist column; Glen Research) and placed on a 394 synthesizer. TEAHF desilylation reagent was delivered at a flow rate of 1 ml/min for 1 min followed by a 29 min wait step during which the TEAHF reagent sat in the synthesis column. This delivery and wait was repeated 15 times for a total reaction time of 8 h and a total reagent consumption of 16 ml of TEAHF. Following TBDMS deprotection, the support in the column was washed with acetonitrile, the support was coupled to either A, dA or 2'-O-methyl-A phosphoramidites and extended by Cs. Couplings followed standard procedures (24).
2'-ACE deprotection and purification of inhibitors
Bis(2-acetoxyethoxy)methyl orthoester protecting groups were removed from the 2' hydroxyls of the transition state analogs as described (23). Analogs were purified by HPLC on a Microsorb C18 column (Varian) eluted with a 100 mM triethyl amine acetate (pH 7.0)/acetonitrile gradient (0–50% acetonitrile in 60 min; retention time = 20–25 min). Each product was confirmed by mass spectrometry and the proper mobility confirmed by gel electrophoresis. Mass spectra: CCdApOPmnphe: calc. for C49H63N17O24P3+ = 1365.31 Da; found m/z = 1365.23. CCApOPmnphe: calc. for C49H63N17O25P3+ = 1381.31 Da; found m/z = 1381.46. CCAomepOPmnphe: calc. for C50H61N17O25P3+ = 1395.32 Da; found m/z = 1395.38. CCdApOPmnala: calc. for C43H55N17O24P32+ (M+2H) = 1291.28 Da; found m/z =645.63, corresponding to 1291.26 Da. CCApOPmnala: calc. for C43H55N17O25P32+ (M+2H) = 1307.28 Da; found m/z = 653.60, corresponding to 1307.20 Da. CCAomepOPmnala: calc. for C44H57N17O25P32+ (M+2H) = 1321.29 Da; found m/z = 660.62, corresponding to 1321.24 Da.
Purification of ribosomes
Escherichia coli 50S ribosomal subunits were prepared from MRE600 cells as described (25) with the following modifications. Spin times were adjusted for a Beckman Ti 70 rotor. Following the first wash step, the ribosomes were resuspended in 20 mM Tris–HCl, pH 7.6, 0.5 M NH4Cl, 1.5 mM magnesium acetate, 0.5 mM EDTA, 7 mM 2-mercaptoethanol and incubated at 4°C for 4 h to dissociate the subunits. All subsequent wash steps were done with this low magnesium buffer. 50S and 30S particles were isolated by zonal centrifugation in a Beckman Ti 14 rotor on a 10–40% sucrose gradient (450 ml) in 20 mM Tris–HCl (pH 7.6), 60 mM NH4Cl, 5.2 mM magnesium acetate, 0.25 mM EDTA, and 3.0 mM 2-mercaptoethanol (45 000 r.p.m., 4 h 15 min, 4°C). 50S subunits were stored at –80°C in 70 mM NH4Cl, 30 mM KCl, 7 mM MgCl2, 0.5 mM EDTA, 1 mM DTT and 50 mM Tris–HCl (pH 7.6). Prior to all binding assays, ribosomes were activated by incubation at 42°C for 10 min in 200 mM KCl, 20 mM MgCl2, 50 mM Tris–HCl (pH 8.0).
Chemical modification of rRNA
23S rRNA was modified with 1-cyclohexyl-3-(2-morpholinoethyl)carbodiimide metho-p-toluenesulfonate (CMCT) essentially as described (8). 50S ribosomal subunits (10 nM) were incubated with varying concentrations of the transition state analogs at 0°C for 2 h in 200 mM KCl, 20 mM MgCl2, 33% methanol and 50 mM Tris–HCl (pH 8.0). Solid CMCT was added to each 100 μl reaction to a final concentration of 42 mg/ml and allowed to react at 0°C for 60 min. Reactions were quenched with 600 μl of ethanol, and the RNA was extracted. The extent of CMCT modification at U2585 was analyzed by primer extension as described (17). Inhibitor binding reduces accessibility of U2585 to CMCT, resulting in a decrease in band intensity. Loss of band intensity as a function of inhibitor concentration was used to determine the dissociation constants (Kd) of each inhibitor using the equation:
I = Isat + (I0 – Isat) / (1 + / Kd)
where I is band intensity, Isat is band intensity when inhibitor is saturating, and I0 is band intensity in the absence of inhibitor. All band intensities were internally normalized for overall extent of CMCT reactivity with the 23S rRNA and for gel loading, and subsequently normalized relative to I0, the band intensity with no inhibitor present.
RESULTS AND DISCUSSION
Studies on CCdApPmn have lead to structural and mechanistic models of the ribosomal active site. However, all inferences drawn from these studies are dependent on the behavior of a molecule that is missing an essential chemical group located close to the peptidyl transferase reaction center.
Inhibitor synthesis
Considering both the value and the shortcomings of CCdApPmn as a mimic of the peptidyl transferase transition state, we set out to prepare a set of analogs that would improve this interesting molecule (Fig. 2B). In particular, we wanted to include a hydroxyl group and other functional groups at the 2' position of A76. We also wanted the synthesis to be modular, so it could be easily modified to generate a series of related molecules. Toward this goal we redesigned the synthetic approach used for inhibitor preparation (Fig. 3). We settled upon a solid phase strategy that provided a maximum level of modularity for inhibitor design.
In order to ensure rapid, high yielding phosphoramidite coupling reactions during standard solid phase organic synthesis of the inhibitors, we chose to synthesize two new puromycin analogs with the -amine substituted with an -hydroxyl protected with a fluoride labile group. Although this substitution is not ideal from the perspective of transition state analogy, the -hydroxyl functional group substitution on puromycin has been shown in previous studies not to be detrimental to binding (26). A key step for this synthetic route was the preparation of TBDMS protected L-hydroxy acids linked to puromycin aminonucleoside (Fig. 3). Previous reports have shown that a reactive acyl chloride can be generated under mild conditions from a readily prepared TBDMS protected hydroxy ester (27). We were able to isolate in near quantitative yield both TBDMS protected L-lactic acid (1a) and TBDMS protected L-phenyl lactic acid (1b) that were of suitable purity for the conversion to the acyl chloride without subsequent purification. The 3'-amino group of the commercially available puromycin aminonucleoside proved to be satisfactorily reactive towards the acyl chloride generated in situ and yielded the desired amide linkage at the 3' position in acceptable yields (26,28,29). Transient protection of the 5' and 2' hydroxyls of puromycin amino nucleoside with trimethylsilyl chloride prior to the addition of the acyl chloride allowed the use of excess reagent to facilitate complete conversion of the starting nucleoside. Transient 5',2' protection of the TBDMS hydroxy puromycin derivatives (2a, 2b) under standard conditions, followed by coupling to a polystyrene solid support, made the two target resins suitable for coupling to a variety of commercially available phosphoramidites. This approach has several fundamental advantages over the solution phase synthesis of previous peptidyl transferase inhibitors. The silyl protection of the -hydroxyl group is readily removed under mild conditions. The use of fully protected phosphoramidites ensured a minimal number of side reactions, which improved the yield relative to the solution phase approach. Most importantly, the reaction scheme is modular. It can readily accommodate a variety of amino acids linked to the A-site puromycin, and the chemical composition of the nucleotide residues in the P-site sequence are easily varied during solid phase synthesis.
Inhibitor design
We synthesized a series of novel inhibitors based upon CCdApPmn as a lead compound (17). Each had CCA trinucleotides, simulating the 3' end of a P-site tRNA, linked through a phosphate to a puromycin-like moiety (Fig. 2B). Unlike CCdApPmn, which has a phosphoramidite linkage, all of these molecules possessed phosphodiester linkages because our use of conventional solid phase oligoribonucleotide synthesis required substitution of the puromycin -amino group with a hydroxyl moiety. The family of new molecules varied at the 2' position of A76, possessing the three chemical groups -H, -OH, -OCH3, and at the amino acid side chain on the puromycin-like moiety. One set of inhibitors had phenylalanine as the A-site R group, rather than methyl tyrosine as in CCdApPmn. The second set had alanyl side chains as the A-site R groups. In all, six new transition state analogs were synthesized: CCdApOPmnphe, CCApOPmnphe, CCAomepOPmnphe, CCdApOPmnala, CCApOPmnala and CCAomepOPmnala (Table 1).
Table 1. Binding affinities for each of the transition state inhibitors
With these molecules in hand, we determined how the modifications in the new inhibitors affect their affinity for the ribosomal peptidyl transferase center (Fig. 4). Chemical modification of the 23S rRNA by CMCT was used to monitor the binding of the inhibitors in the ribosomal active site, as has been used to measure binding of other peptidyl transferase ligands, including CCdApPmn (8,17). Protection of several sites in domain V of 23S rRNA (U2506, U2584 and U2585) from modification by CMCT has been linked to binding of tRNA in the P-site, and was also observed upon CCdApPmn binding (7,17). We measured the protection of U2585 from CMCT modification in varying concentrations of the inhibitors to determine their binding affinities. The complete set of binding constants is listed in Table 1.
Figure 4. Binding of transition state mimics in the peptidyl transferase center of E.coli based upon chemical modification of the 23S rRNA. Molecules bound in the active site protect U2585 from modification by CMCT, resulting in the disappearance of a CMCT-dependent reverse transcriptase stop. (A) A typical gel of reverse transcripts of 23S rRNA from ribosomes modified by CMCT in the presence of increasing concentrations of CCdApOPmnphe and (B) CCApOPmnphe. (C) U2585 band intensities were normalized for overall CMCT reactivity and gel loading and then relative to the intensity for no inhibitor present. Binding curves were fit to these data and dissociation constants were derived from the fits.
Two modifications to CCdApPmn were made for synthetic purposes: the substitution of the puromycin -amino group with a hydroxyl, and the substitution of phenylalanine for methyl-tyrosine. The first question we wished to address was: do these background changes in the inhibitor have a significant effect on their binding to the ribosome? CCdApOPmnphe, the molecule that is most similar to CCdApPmn, but includes both background modifications, exhibited a Kd of 53 nM, in close agreement with the previously published figure for CCdApPmn (70 nM) (17). CCdApPmn had a similar binding affinity when tested under our binding conditions (data not shown). The correspondence of the binding affinities of CCdApPmn and CCdApOPmnphe suggests that it is probably valid to assume that the background changes made to CCdApPmn have little, if any, effect on the binding of these molecules in the ribosomal active site. Alternatively, the similar binding constants may result from two inverse energetic effects that nearly cancel each other. If this were the case, then these differences would be consistent throughout the series of new inhibitors and these effects are unlikely to have a major influence on how these molecules are binding in the ribosome.
Once the legitimacy of comparing the behavior of the new set of transition state analogs with that of CCdApPmn was established, we wished to determine how the variations within this family of molecules affected their ribosomal affinity. Of primary interest was the involvement of the A76 2'-OH, which is absent from CCdApPmn. Surprisingly, the molecule with riboadenosine at position 76, CCApOPmnphe, had a Kd of 331 nM, 6-fold weaker than the deoxyadenosine version (G = –1 kcal/mol). No significant decomposition of these molecules was observed after incubation with ribosomes for 24 h, ruling out the possibility that the increased Kd resulted from degradation (data not shown).
A molecule containing a 2'-OCH3 at A76, CCAomepOPmnphe, was also synthesized and tested for binding. No significant protection of U2585 was observed at any concentration of CCAomepOPmnphe up to 5000 nM, indicating a Kd 5000 nM. This corresponds to at least a 100-fold decrease in binding (G –2.5 kcal/mol).
The structural conformation of the deoxy inhibitor cannot for steric reasons be assumed by the ribo-version of the inhibitor, but that the deoxy inhibitor binds with greater affinity. This suggests that the deoxy inhibitors assume a conformation that is energetically more stable, but less biologically relevant, at least in the case of the aromatic side chain. The presence of the 2'-OH forces CCApOPmnphe out of this conformation, reducing binding affinity. Presumably, the ribo-variant has assumed a different conformation within the PTC than the original inhibitor, a conformation that is more likely to reflect the actual transition state. The presence of a 2'-OCH3, which is considerably larger than a hydroxyl group, precludes binding altogether. Apparently the additional bulk of an O-methyl group cannot be accommodated in the local region of the peptidyl transferase center by any variation of a conformational change. The P-site A76 is expected to participate in a Type I A-minor interaction with A2450/C2501 (E.coli) (30). This interaction involves a close approach of the ribose sugar against the A-C pair and hydrogen bonding with the A76 2'-OH. A 2'-OCH3 on this base would be completely incompatible with this interaction.
Considering the rather unexpected observation that inclusion of a 2'-OH on A76, which makes the analog more similar to the actual transition state, decreases binding affinity, we sought an explanation from the 50S crystal structure with CCdApPmn bound in the active site. The structure shows that the methyl tyrosine side chain is stacked between two bases of the rRNA. This interaction could significantly affect binding of the molecule to the ribosome and influence the orientation of the phosphoramidate linkage. Previous studies showed that the amino acid side chain of puromycin provides a substantial contribution to the efficacy of puromycin as an inhibitor of peptidyl transferase, but that analogs containing non-aromatic side chains can function as inhibitors, particularly if the nucleic acid portion of the molecule is expanded to mimic more of the amino-acyl tRNA (31,32). Stacking of the methyl tyrosine, while energetically important for binding of the A-site portion of the inhibitor, may limit flexibility of nearby regions of the molecule and force the phosphate into the incorrect conformation observed for CCdApPmn.
To test the role of the A-site amino acid side chain in binding transition state analogs in the active site, we generated a set of transition state analogs that had alanine substituted for phenylalanine in the puromycin moiety (Fig. 2B). In particular, we wanted to determine how changing the amino acid and eliminating the stacking interaction would affect the ribosome’s ability to accommodate 2'-OH substitution at A76. When the amino acid side chain is replaced with alanine, the observed binding constants were 250 and 280 nM for CCdApOPmnala and CCApOPmnala, respectively. Again, no binding was observed with the 2'-O-methyl substituted variant CCAomepOPmnala up to 5000 nM. The reduction in binding affinity from CCdApOPmnphe to CCdApOPmnala is likely due to the loss of the stacking interactions between the aromatic phenylalanine side chain and A2451 and C2452 (E.coli) of the 23S rRNA. Significantly, the binding of the A76 riboadenosine and deoxyadenosine analogs were approximately the same for the alanyl derivatives. Based only on the comparison of inhibitors containing riboadenosine and deoxyriboadenosine, it could be argued that steric clash between the phosphate oxygen and the 2'-hydroxyl only affects binding in the case of an aromatic A-site amino acid side chain and is not a general effect. The elimination of the phenylalanine stacking interaction may increase the conformational flexibility of the entire molecule, allowing the tetrahedral phosphate to move in a way that would accommodate the 2'-hydroxyl at A76. However, the elimination of binding upon inclusion of an O-methyl group, regardless of the identity of the amino acid side chain, indicates that additional flexibility afforded to non-aromatic A-site substrates cannot compensate for the additional steric bulk of a 2'-O-methyl group. There is a finite space in the vicinity of the A76 2'-carbon available for occupancy by various functional groups. This space must not be substantially larger than the hydroxyl group found in native substrates given that the only moderately larger O-methyl group cannot be accommodated. Because the hydroxyl group must nearly fill the available space, its presence will certainly have a significant effect on the conformation of bound substrates or transition state analogs, even in the case of non-aromatic A-site side chains.
CONCLUDING REMARKS
We have presented a synthetic scheme for efficiently producing transition state analogs of the peptidyl transferase reaction. One important development compared with previously studied transition state analogs is the inclusion of the hydroxyl group at A76 of the P-site tRNA. Although the binding affinities are not improved, these molecules will be of particular interest in structural studies, to help refine the mechanism of the ribosome active site during chemistry. There are substantial steric constraints placed on the tetrahedral phosphate by the now present 2'-OH of A76. Therefore, we expect a significant rearrangement of the conformation of the phosphate from that seen when CCdApPmn is bound in the ribosomal active site. Such structures will be more relevant to the reality of peptidyl transferase catalysis by the ribosome.
In addition to allowing inclusion of the critical A76 2'-OH, our synthetic scheme imparts modularity to the production of inhibitors, allowing the synthesis of a large variety of related molecules. It particularly affords flexibility in the tetrahedral phosphate and P-site portions of the molecules. Among other changes, we will be able to synthesize transition state analogs with chiral centers at their tetrahedral phosphates. This will make it possible to resolve the stereochemical ambiguity of the peptidyl transferase transition state.
ACKNOWLEDGEMENTS
We thank Mark Parnell and Kevin Huang for comments on the manuscript. We thank the laboratory of Andrew D. Miranker for assistance with mass spectrometry. This work was supported by American Cancer Society Beginning Investigator grant RSG-02-052-GMC to S.A.S., a Research Scholar Grant PF-013-01-GMC from the American Cancer Society to G.W.M. and a National Science Foundation Predoctoral Fellowship to J.S.W.
REFERENCES
Chladek,S. and Sprinzl,M. (1985) The 3'-end of transfer-RNA and its role in protein biosynthesis. Angew. Chem. Int. Ed., 24, 371–391.
Yarus,M. and Welch,M. (2000) Peptidyl transferase: ancient and exiguous. Chem. Biol., 7, R187–R190.
Green,R. and Noller,H.F. (1997) Ribosomes and translation. Annu. Rev. Biochem., 66, 679–716.
Ban,N., Nissen,P., Hansen,J., Moore,P.B. and Steitz,T.A. (2000) The complete atomic structure of the large ribosomal subunit at 2.4 ? resolution. Science, 289, 905–920.
Barta,A., Steiner,G., Brosius,J., Noller,H.F. and Kuechler,E. (1984) Identification of a site on 23s ribosomal-RNA located at the peptidyl transferase center. Proc. Natl Acad. Sci. USA, 81, 3607–3611.
Moazed,D. and Noller,H.F. (1987) Chloramphenicol, erythromycin, carbomycin and vernamycin B protect overlapping sites in the peptidyl transferase region of 23S ribosomal RNA. Biochimie, 69, 879–884.
Moazed,D. and Noller,H.F. (1989) Interaction of tRNA with 23S rRNA in the ribosomal A, P and E sites. Cell, 57, 585–597.
Moazed,D. and Noller,H.F. (1991) Sites of interaction of the CCA end of peptidyl-tRNA with 23S rRNA. Proc. Natl Acad. Sci. USA, 88, 3725–3728.
Bocchetta,M., Xiong,L.Q. and Mankin,A.S. (1998) 23S rRNA positions essential for tRNA binding in ribosomal functional sites. Proc. Natl Acad. Sci. USA, 95, 3525–3530.
Samaha,R.R., Green,R. and Noller,H.F. (1995) A base pair between tRNA and 23S rRNA in the peptidyl transferase centre of the ribosome. Nature, 377, 309–314. Erratum in: Nature (1995), 378, 419.
Kim,D.F. and Green,R. (1999) Base-pairing between 23S rRNA and tRNA in the ribosomal A site. Mol. Cell, 4, 859–864.
Takai,K., Horie,N., Yamaizumi,Z., Nishimura,S., Miyazawa,T. and Yokoyama,S. (1994) Recognition of UUN codons by two leucine tRNA species from Escherichia coli. FEBS Lett., 344, 31–34.
Quiggle,K., Kumar,G., Ott,T.W., Ryu,E.K. and Chladek,S. (1981) Donor site of ribosomal peptidyltransferase: investigation of substrate specificity using 2'(3')-O-(N-acylaminoacyl)dinucleoside phosphates as models of the 3' terminus of N-acylaminoacyl transfer ribonucleic acid. Biochemistry, 20, 3480–3485.
Ringer,D., Quiggle,K. and Chladek,S. (1975) Recognition of the 3' terminus of 2'-O-aminoacyl transfer ribonucleic acid by the acceptor site of ribosomal peptidyltransferase. Biochemistry, 14, 514–520.
Wagner,T., Cramer,F. and Sprinzl,M. (1982) Activity of the 2' and 3' isomers of aminoacyl transfer ribonucleic acid in the in vitro peptide elongation on Escherichia coli ribosomes. Biochemistry, 21, 1521–1529.
Bhuta,A., Quiggle,K., Ott,T., Ringer,D. and Chladek,S. (1981) Aminoacyl derivatives of nucleosides, nucleotides and polynucleotides. 33. Stereochemical control of ribosomal peptidyltransferase reaction—role of amino-acid side-chain orientation of acceptor substrate. Biochemistry, 20, 8–15.
Welch,M., Chastang,J. and Yarus,M. (1995) An inhibitor of ribosomal peptidyl transferase using transition-state analogy. Biochemistry, 34, 385–390.
Wolfenden,R., Rammler,D.H. and Lipmann,F. (1964) On the site of esterification of amino acids to soluble RNA. Biochemistry, 3, 329–338.
Nissen,P., Hansen,J., Ban,N., Moore,P.B. and Steitz,T.A. (2000) The structural basis of ribosome activity in peptide bond synthesis. Science, 289, 920–930.
Hansen,J.L., Schmeing,T.M., Moore,P.B. and Steitz,T.A. (2002) Structural insights into peptide bond formation. Proc. Natl Acad. Sci. USA, 99, 11670–11675.
Ramakrishnan,V. (2002) Ribosome structure and the mechanism of translation. Cell, 108, 557–572.
Parnell,K.M., Seila,A.C. and Strobel,S.A. (2002) Evidence against stabilization of the transition state oxyanion by a pKa-perturbed RNA base in the peptidyl transferase center. Proc. Natl Acad. Sci. USA, 99, 11658–11663.
Scaringe,S.A. (2001) RNA oligonucleotide synthesis via 5'-silyl-2'-orthoester chemistry. Methods, 23, 206–217.
Scaringe,S.A., Wincott,F.E. and Caruthers,M.H. (1998) Novel RNA synthesis method using 5'-O-silyl-2'-O-orthoester protecting groups. J. Am. Chem. Soc., 120, 11820–11821.
Rodnina,M.V. and Wintermeyer,W. (1995) GTP consumption of elongation factor Tu during translation of heteropolymeric mRNAs. Proc. Natl Acad. Sci. USA, 92, 1945–1949.
Katunin,V.I., Muth,G.W., Strobel,S.A., Wintermeyer,W. and Rodnina,M.V. (2002) Important contribution to catalysis of peptide bond formation by a single ionizing group within the ribosome. Mol. Cell, 10, 339–346.
Wissner,A. and Grudzinskas,C.V. (1978) Reaction of tert-butyldimethylsilyl esters with oxalyl chloride-dimethylformamide: preparation of carboxylic acid chlorides under neutral conditions. J. Org. Chem., 43, 3972–3974.
Kelly,S.E. and LaCour,T.G. (1992) A one pot procedure for the synthesis of alpha-hydroxy amides from the corresponding alpha-hydroxy acids. Synth. Commun., 22, 859–869.
Trost,B.M. and McDougal,P.G. (1982) Synthesis of optically active verrucarinic acid derivatives. Tetrahedron Lett., 23, 5497–5500.
Nissen,P., Ippolito,J.A., Ban,N., Moore,P.B. and Steitz,T.A. (2001) RNA tertiary interactions in the large ribosomal subunit: the A-minor motif. Proc. Natl Acad. Sci. USA, 98, 4899–4903.
Fong,K.L. and Vince,R. (1978) Peptidyl transferase substrate specificity with nonaromatic aminoacyl analogues of puromycin. J. Med. Chem., 21, 792–795.
Symons,R.H., Harris,R.J., Clarke,L.P., Wheldrake,J.F. and Elliott,W.H. (1969) Structural requirements for inhibition of polyphenylalanine synthesis by aminoacyl and nucleotidyl analogues of puromycin. Biochim. Biophys. Acta, 179, 248–250.(Joshua S. Weinger1, David Kitchen3, Step)