Studies of the novel ketolide ABT-773: transport, binding to ribosomes, and inhibition of protein synthesis in Streptococcus pneumoniae.

Macrolide resistance in Streptococcus pneumoniae has been associated with two main mechanisms: target modification by Erm methyltransferases and efflux by macrolide pumps. The ketolide ABT-773, which has a 3-keto group and no L-cladinose sugar, represents a new class of drugs with in vitro activity against a variety of resistant bacteria. Several approaches were undertaken to understand how ABT-773 was able to defeat resistance mechanisms. We demonstrated tighter ribosome binding of ABT-773 than erythromycin. We also showed that ABT-773 (i) accumulated in macrolide-sensitive S. pneumoniae at a higher rate than erythromycin, (ii) was able to bind with methylated ribosomes, though at lower affinities than with wild-type ribosomes, and (iii) accumulated in S. pneumoniae strains with the efflux-resistant phenotype.

Domain structure analysis of elongation factor-3 from Saccharomyces cerevisiae by limited proteolysis and differential scanning calorimetry.

Elongation-factor-3 (EF-3) is an essential factor of the fungal protein synthesis machinery. In this communication the structure of EF-3 from Saccharomyces cerevisiae is characterized by differential scanning calorimetry (DSC), ultracentrifugation, and limited tryptic digestion. DSC shows a major transition at a relatively low temperature of 39 degrees C, and a minor transition at 58 degrees C. Ultracentrifugation shows that EF-3 is a monomer; thus, these transitions could not reflect the unfolding or dissociation of a multimeric structure. EF-3 forms small aggregates, however, when incubated at room temperature for an extended period of time. Limited proteolysis of EF-3 with trypsin produced the first cleavage at the N-side of Gln775, generating a 90-kDa N-terminal fragment and a 33-kDa C-terminal fragment. The N-terminal fragment slowly undergoes further digestion generating two major bands, one at approximately 75 kDa and the other at approximately 55 kDa. The latter was unusually resistant to further tryptic digestion. The 33-kDa C-terminal fragment was highly sensitive to tryptic digestion. A 30-min tryptic digest showed that the N-terminal 60% of EF-3 was relatively inaccessible to trypsin, whereas the C-terminal 40% was readily digested. These results suggest a tight structure of the N-terminus, which may give rise to the 58 degrees C transition, and a loose structure of the C-terminus, giving rise to the 39 degrees C transition. Three potentially functional domains of the protein were relatively resistant to proteolysis: the supposed S5-homologous domain (Lys102-Ile368), the N-terminal ATP-binding cassette (Gly463-Lys622), and the aminoacyl-tRNA-synthase homologous domain (Glu820-Gly865). Both the basal and ribosome-stimulated ATPase activities were inactivated by trypsin, but the ribosome-stimulated activity was inactivated faster.

Cellular accumulation, localization, and activity of a synthetic cyclopeptamine in fungi.

A novel synthetic cyclopeptamine, A172013, rapidly accumulated by passive diffusion into Candida albicans CCH442. Drug influx could not be totally facilitated by the membrane-bound target, beta-(1,3)-glucan synthase, since accumulation was unsaturable at drug concentrations up to 10 microg/ml (about 1.6 x 10(-7) molecules/cell), or 25x MIC. About 55 and 23% of the cell-incorporated drug was associated with the cell wall and protoplasts, respectively. Isolated microsomes contained 95% of the protoplast-associated drug, which was fully active against glucan synthesis in vitro. Drug (0.1 microg/ml) accumulation was rapid and complete after 5 min in several fungi tested, including a lipopeptide/cyclopeptamine-resistant strain of C. albicans (LP3-1). The compound penetrated to comparable levels in both yeast and hyphal forms of C. albicans, and accumulation in Aspergillus niger was 20% that in C. albicans. These data indicated that drug-cell interactions were driven by the amphiphilic nature of the compound and that the cell wall served as a major drug reservoir.

Identification and kinetic analysis of a functional homolog of elongation factor 3, YEF3 in Saccharomyces cerevisiae.

Yeast and other fungi contain a soluble elongation factor 3 (EF-3) which is required for growth and protein synthesis. EF-3 contains two ABC cassettes, and binds and hydrolyses ATP. We identified a homolog of the YEF3 gene in the Saccharomyces cerevisiae genome database. This gene, designated YEF3B, is 84% identical in protein sequence to YEF3, which we will now refer to as YEF3A. YEF3B is not expressed during growth under laboratory conditions, and thus cannot rescue growth of YEF3A deletion strains. However, YEF3B can take the place of YEF3A in vivo when expressed from the YEF3A or ADH1 promoters. The products of the YEF3A and YEF3B genes, EF-3A and EF-3B, respectively, were expressed from the ADH1 promoter and purified. Both factors possessed basal and ribosomal-stimulated ATPase activity, and had similar affinity for yeast ribosomes (103 to 113 nM). K(m) values for ATP were similar, but the Kcat values differed significantly. Ribosome-dependent ATPase activity of EF-3A was more efficient than EF-3B, since the Kcat and Kcat/K(m) values for EF-3A were about two-fold higher; however, the difference in Kcat/K(m) values between the two factors was small for basal ATPase activity.

Molecular basis of clarithromycin activity against Mycobacterium avium and Mycobacterium smegmatis.

Clarithromycin, the 6-O-methyl derivative of erythromycin, is approved for treatment of Mycobacterium avium infections and for prophylaxis in patients at risk. Since clarithromycin is more active against mycobacteria than the parent compound, erythromycin, we evaluated the interaction of erythromycin and clarithromycin with cells and ribosomes isolated from M. avium and Mycobacterium smegmatis. The MIC of clarithromycin was 32 and 64 times lower than that of erythromycin for M. smegmatis and M. avium, respectively. The cellular uptake rate for clarithromycin was two- to five-fold faster than for erythromycin, and cell-associated clarithromycin reached a plateau two-fold higher than that of erythromycin after 3 h. Energy was not required for uptake. Fractionation of cell-associated clarithromycin yielded 12% in the walls, 21% bound to ribosomes, with the remainder being lost during work-up. In addition, three- to six-fold more clarithromycin was associated with the isolated cell integument compared with erythromycin. The Kd for clarithromycin binding to ribosomes was 2.9- and 3.5-fold tighter for M. smegmatis and M. avium, respectively, than for erythromycin, due mainly to a slower off-rate. The log partition coefficients of the non-ionized form (log Pu) for clarithromycin and erythromycin were 3.24 and 2.92, respectively. Thus clarithromycin is more hydrophobic than erythromycin. This would favour more rapid diffusion within and across hydrophobic regions of the cell integument, since once a solute saturates a membrane the net flux across the membrane must equal the net flux within the membrane as dictated by diffusion. We conclude that the lower MIC of clarithromycin for M. avium and M. smegmatis is due to a combination of increased cellular uptake, the major factor, possibly through a peripheral hydrophobic layer, and increased binding affinity to ribosomes.

Novel antifungal agents which inhibit lanosterol 14alpha-demethylase in Candida albicans CCH442.

We have identified four non-azole inhibitors of lanosterol 14a-demethylase in Candida albicans CCH442. The most potent compound, A-39806, had IC50 values for ergosterol inhibition of 0.9 microM (0.3 mg/L) and 1.9 microM (0.6 mg/L) in whole cell and cell-free extract assays, respectively. A-39806 demonstrated broad in-vitro antifungal activity against several Candida species as well as against Cryptococcus albidus and Aspergillus niger. In-vitro antifungal activity was also demonstrated against a fluconazole-resistant clinical isolate of C. albicans.

Inhibition of 2,3-oxidosqualene-lanosterol cyclase in Candida albicans by pyridinium ion-based inhibitors.

The N-(4E,8E)-5,9,13-trimethyl-4,8,12-tetradecatrien-1- ylpyridinium and N-(4E,8E)-5,9,13-trimethyl-4,8,12-tetradecatrien-1- ylpicolinium cations were evaluated for their ability to inhibit 2,3-oxidosqualene-lanosterol cyclase activity in Candida albicans. Both compounds inhibited fungal growth, were fungicidal, and resulted in the accumulation of squalene epoxide concurrent with a decrease in ergosterol, monomethyl sterols, and lanosterol, as was expected for the specific inhibition of 2,3-oxidosqualene-lanosterol cyclase activity. These compounds are electron-poor aromatic mimics of a monocyclized transition state or high-energy intermediate formed from oxidosqualene, which may explain their selective action.

Design, synthesis and in vitro evaluation of pyridinium ion based cyclase inhibitors and antifungal agents.

The design, synthesis and in vitro biological evaluation of pyridinium ion based inhibitors of oxidosqualene cyclase enzymes are reported. N-Alkyl- and N-prenylpyridinium ions have been found to be potent and specific inhibitors of Candida albicans oxidosqualene-lanosterol cyclase and to exhibit antifungal activity. The ability of pyridinium ions to inhibit the C. albicans cyclase increases with increasing structural resemblance to a putative monocyclized species formed during the course of the cyclization process. The N-(4E,8E)-5,9,13-trimethyl-4,8,12-tetradecatrien-1- ylpyridinium cation 1 inhibits the C. albicans enzyme at concentrations more than 100-fold lower than does the directly analogous piperidinium derivative 4.

Antifungal drug targets: Candida secreted aspartyl protease and fungal wall beta-glucan synthesis.

The incidence of severe, life-threatening fungal infections has increased dramatically over the last decade. Unfortunately, in practice the arsenal of antifungal drugs is limited to flucytosine, a few approved azoles, and polyenes, mainly amphotericin B. This situation is rather precarious in view of the extended spectrum of fungi causing severe disease in immunocompromised patients, development of resistance to some of the currently used agents, and the minimal fungicidal activity of the azoles. Although lagging behind the need for new antifungal agents, the study of fungal biochemistry, physiology, and genetics has undergone a resurgence to new heights of activity, thus providing a framework on which to build drug discovery programs in several new areas, two of which will be discussed in detail: the biology of Candida albicans secreted aspartyl protease with respect to inhibitor discovery, evaluation, and possible clinical utility; and the fungal cell wall beta-glucans with respect to the mechanism and regulation of synthesis and target sites for drug inhibition.

Kinetics of beta-1,3 glucan interaction at the donor and acceptor sites of the fungal glucosyltransferase encoded by the BGL2 gene.

Formation of branched glucan, glucan-glucan cross links, and glucan-chitin cross links most likely involves the action of fungal wall glucanases and transglycosylases. We developed an HPLC assay using radiolabeled substrates in order to study the kinetics of interaction of donor and acceptor molecules with a glucosyltransferase present in the cell walls of both Saccharomyces cerevisiae and Candida albicans. Purified transferase first forms an activated intermediate from a donor beta-1,3 glucan, releasing free disaccharide. The activated intermediate is transferred, in the presence of an appropriate acceptor beta-1,3 glucan, yielding a linear glucan containing a beta-1,6 linkage at the transfer site [Yu, L., Goldman, R., Sullivan, P., Walker, G. & Fesik, S. W. (1993) J. Biomol. NMR 3, 429-441]. An apparent Km of 0.41 mM for the acceptor site was determined using laminaritetraose as the acceptor. An apparent Km of 31 mM for the donor site was determined using increasing concentrations of laminaripentaose, and monitoring formation of laminaribiose. The enzyme functioned as a glucanase at low concentrations of acceptor molecules, with excess H2O competing for reaction at the activated donor site, thus resulting in hydrolysis. However, as the concentration of acceptor increased, the reaction shifted from hydrolysis to glucosyltransfer. The reaction appeared specific for beta-1,3 glucan as acceptor, in as much as no transfer was detected when either hexa-N-acetyl-chitohexaose or maltooligosaccharides were used as acceptors. The roles of such an enzymic activity in cell wall metabolism is discussed in terms of repair, cross linking and incorporation of newly synthesized chains of beta-1,3 glucan into the previously existing cell wall structure.

Synthesis of yeast cell wall glucan and evidence for glucan metabolism in a Saccharomyces cerevisiae whole cell system.

The synthesis and metabolism of yeast cell wall glucan were studied using a Saccharomyces cerevisiae construct in which radiolabelled galactose is metabolized to UDP-glucose and preferentially incorporated into glucan. Greater than 85% of the incorporated radiolabel was found within insoluble cell wall material. Our study also demonstrated that radiolabelled wall glucan is released from cells growing exponentially, and that the released radiolabel is reutilizable low molecular mass material. Size exclusion chromatography and enzymic analysis indicate that laminaribiose comprises approximately 50% of the released fraction. This is consistent with in vitro findings that laminaribiose is a by-product of a newly identified glucosyltransferase (R. P. Hartland, G. W. Emerson & P. A. Sullivan, 1991, Proc R Soc Lond B 246, 155-160) associated with fungal cell walls. Our results also suggest that pre-existing glucan undergoes less metabolic processes than newly synthesized material as evidenced by a decrease in released radiolabel over time. Pulse double labelling of glucan and total cellular protein indicate that glucan metabolism and protein synthesis (ps) are not tightly coupled although they do parallel each other during exponential growth. Inhibitors of glucan synthesis (gs) decrease the glucan to protein ratio. Measurement of ps allows normalization for non-specific decreases in the rate of cell wall synthesis due to general cessation of growth. Cilofungin and papulacandin B, two putative inhibitors of gs, inhibited galactose incorporation into glucan and thus showed a decrease in the glucan to protein ratio, although ps was affected.(ABSTRACT TRUNCATED AT 250 WORDS)

Macrolide transport in Escherichia coli strains having normal and altered OmpC and/or OmpF porins.

[(14)C]Erythromycin and [(14)C]azithromycin uptake rates were studied in Escherichia coli strains containing normal OmpC and OmpF porins (strain MRC 106) and altered OmpC porins due to small insertions (strains RAM121 and OC1555) or deletions (strain RAM122) in the ompC alleles and altered OmpF porins due to small ompF deletions (strains OC1555 and PLB3255). Strains RAM121 and RAM122 also lacked OmpF porins in their outer membrane. The porin mutants demonstrated a 2- to 1224-fold increase in macrolide transport and a concurrent 3- to 530-fold decrease in MIC when compared to the parent strain, MCR106. Both strains OC1555 and PLB3255 had enhanced permeability to 1-N-phenylnaphthylamine indicating increased permeability of the outer membrane to hydrophobic molecules. The macrolides, at 2 to 1000 times MIC, failed to displace the cationic probe polymyxin; therefore, drug entry by a self-promoted mechanism was not indicated. Since >95% of macrolide is protonated and thus hydrophilic (logP(i)=-0.89) at neutral pH, the bulk of drug entry may be via the porin channels.

Anti-Candida activity of cispentacin: the active transport by amino acid permeases and possible mechanisms of action.

Cispentacin tranport into Candida albicans CCH442 was via a specific inducible proline permease and other amino acid permeases. Drug entry was also dependent upon the proton motive force. The apparent Km and Vmax for drug uptake under induced conditions were 0.4 mM and 7 nmol/microliter/min, respectively, with cellular accumulation in the mM range. Cispentacin uptake was competitively inhibited by L-proline with an apparent Ki of 75 microM. Cispentacin did not charge to transfer-RNA or incorporate into protein; however, the compound did inhibit in vivo incorporation of [14C]lysine into protein and [3H]adenine into RNA as well as in vitro [14C]proline charging to transfer-RNA. Cispentacin did not inhibit amino acid biosynthesis in vivo but did elevate levels of several amino acids possibly by interfering with self-regulatory mechanisms.

A non-azole inhibitor of lanosterol 14 alpha-methyl demethylase in Candida albicans.

6-Hydroxy-N-methyl-N-(2-[4-phenylphenyl] ethyl)-1,2,3,4-tetrahydro-1- napthalene methanamine (A60586), a new non-azole inhibitor of ergosterol biosynthesis in Candida albicans ATCC62376 has been identified. In whole cells A60586 produced a dose related reduction of [14C]acetate incorporation into ergosterol and a concurrent increase in the radiolabelling of 4,4-dimethylated sterols. Similar observations were made with [14C]mevalonic acid lactone labelled cell free extracts. The IC50s for inhibition of ergosterol in the whole cell and cell free systems were 22 microM (10 mg/L) and 7.8 microM (3.5 mg/L), respectively. Analysis by gas chromatography of sterols from cells previously incubated at 37 degrees C for 24 h with A60586 (200 mg/L) confirmed the presence of lanosterol and 14 alpha-methyl fecosterol. These data indicate that A60586, inhibits the demethylation of the C-14 methyl group of lanosterol. The MIC of A60586 for several candida strains ranged from 12.5 to 50 mg/L, and against Cryptococcus albidus and Aspergillus niger ranged from 50 to 100 mg/L. The best in-vitro activity of A60586 was against Torulopsis glabrata (MIC range = 3.12 to 50 mg/L). The membrane permeabilizing effect of this compound (50% leakage of [14C]aminoisobutyric acid at 70 mg/L A60586) may have contributed to its in-vitro antifungal activity.

Application of a fluorogenic substrate in the assay of proteolytic activity and in the discovery of a potent inhibitor of Candida albicans aspartic proteinase.

A fluorescent method for monitoring the activity of the secreted Candida carboxyl (aspartic) proteinase (EC 3.4.23.6) was developed using a fluorogenic substrate based on resonance energy transfer. The fluorescent assay was used to monitor proteinase production, purification, and inhibition. The Km for the fluorogenic substrate, 4-(4-dimethylaminophenylazo)benzoyl-gamma-aminobutyryl-Ile-His-Pro - Phe-His-Leu-Val-Ile-His-Thr- [5-(2-aminoethyl)amino]naphthalene-1-sulfonic acid, was found to be 4.3 microM at the optimum pH of 4.5. Reaction products were separated by reverse-phase high-performance liquid chromatography and identified by amino acid analysis or by 252Cf plasma desorption mass spectrometry. Cleavage of the fluorogenic substrate was between the histidine-threonine residues, releasing the fluorescent product, threonine-[5-(2-aminoethyl)amino]naphthalene-1-sulfonic acid. Proteolytic activity was expressed as nanomoles of fluorescent product released at 22 degrees C/60 min, pH 4.5, and the release of 0.9 nmol product was equivalent to one hemoglobin proteolytic unit (O.D.A700 increase of 0.100) produced at 37 degrees C/60 min, pH 3.5. The aspartic proteinase inhibitor pepstatin had an IC50 of 27 nM when tested in a dose-response study with the purified enzyme. The apparent Ki for pepstatis was 2.9 nM. Several synthetic inhibitors of the enzymes were identified with IC50's in the nanomolar range. The most potent compound, A70450, was characterized as a fast, tight-binding inhibitor having an IC50 of 1.3 nM and apparent Ki of 0.17 nM.

Inhibition of lipopolysaccharide synthesis in Agrobacterium tumefaciens and Aeromonas salmonicida.

Lipopolysaccharide (LPS) synthesis was inhibited, new lipid A metabolites accumulated, and growth ceased, when the plant pathogen Agrobacterium tumefaciens and the fish pathogen Aeromonas salmonicida were treated with an antibacterial agent which specifically inhibits CTP:CMP-3-deoxy-manno-octulosonate cytidylyltransferase (CMP-KDO synthase). The new lipid A metabolites were purified by chromatography on DEAE-cellulose and chemically analysed. Metabolites isolated from both bacterial species contained glucosamine and phosphate in a 1:1 molar ratio, and 3-OH-C14:0 was the major fatty acid present (1 mol and 1.4 mol per mol glucosamine for A. tumefaciens and A. salmonicida, respectively). Inhibition of LPS synthesis by CMP-KDO synthase inhibitor had no effect on the initial kinetics of A. tumefaciens attachment to cultured carrot cells, but did inhibit cell aggregation normally induced by bacterial cellulose synthesis. Bacteria treated with inhibitor remained viable and able to synthesize protein at 15% the rate of control cells, indicating that the lack of cellulose-induced aggregation was not due to the inability of bacteria to make protein, but rather the inability to respond normally to the bacterial-plant cell interaction.

Role of an energy-dependent efflux pump in plasmid pNE24-mediated resistance to 14- and 15-membered macrolides in Staphylococcus epidermidis.

We have elucidated a new mechanism for bacterial resistance to the 14-membered macrolides oleandomycin and erythromycin and the 15-membered macrolide azithromycin. Plasmid pNE24, previously isolated from a clinical specimen of Staphylococcus epidermidis, was characterized as causing resistance to 14-membered but not 16-membered macrolides by a mechanism suggested to involve reduced antibiotic permeation of bacterial cells (B. C. Lampson, W. von David, and J. T. Parisi, Antimicrob. Agents Chemother. 30:653-658, 1986). Our recent investigations have demonstrated that S. epidermidis 958-2 containing plasmid pNE24 also contains an energy-dependent macrolide efflux pump which maintains intracellular antibiotic concentrations below those required for binding to ribosomes. Thus, when strain 958-2 was pretreated with the inhibitor carbonyl cyanide m-chlorophenylhydrazone (CCCP), macrolide accumulated at the same rate and to the same extent as in CCCP-treated or untreated control cells lacking plasmid pNE24 (strain 958-1). In contrast, macrolide did not accumulate in energy-competent strain 958-2 but did accumulate to levels equal to those of ribosomes immediately following CCCP addition. Furthermore, intracellular macrolide was excreted and bacteria resumed growth when CCCP but not macrolide was removed from the growth medium. As expected, the 16-membered macrolide niddamycin accumulated to the same level in energy-competent strains 958-1 and 958-2 at the same rapid rate. Macrolide incubated with lysates prepared from both strains or recovered from cells of strain 958-2 was unmodified and bound to ribosomes from strains 958-1 and 958-2 with identical affinities and kinetics, thus precluding a role for ribosome or drug alteration in the resistance mechanism. We conclude that the presence of plasmid pNE24 results in specific energy-dependent efflux of 14- and 15-membered macrolides.

Erythromycin and azithromycin transport into Haemophilus influenzae ATCC 19418 under conditions of depressed proton motive force (delta mu H).

The effect of collapsing the electrochemical proton gradient (delta mu H) on [3H]erythromycin and [14C]azithromycin transport in Haemophilus influenzae ATCC 19418 was studied. The proton gradient and membrane potential were determined from the distribution of [2-14C]dimethadione and rubidium-86, respectively. delta mu H was reduced from 124 to 3 mV in EDTA-valinomycin-treated cells at 22 degrees C with 150 mM KCl and 0.1 mM carbonyl cyanide m-chlorophenylhydrazone. During the collapse of delta mu H, macrolide uptake increased. Erythromycin efflux studies strongly suggested that this increase was not due to an energy-dependent efflux pump but was likely due to increased outer membrane permeability. These data indicated that macrolide entry was not a delta mu H-driven active transport process but rather a passive diffusion process.

Mechanism of mupirocin transport into sensitive and resistant bacteria.

Pseudomonic acid A (mupirocin) blocks protein synthesis in bacteria by inhibition of bacterial isoleucyl-tRNA synthetase. [16, 17-3H]mupirocin, isolated from a methionine auxotroph of Pseudomonas fluorescens, was used to study transport of this antibiotic into sensitive and resistant strains of Bacillus subtilis, Staphylococcus aureus, and Escherichia coli. The transport of mupirocin into sensitive bacteria was energy independent and temperature dependent (decreased uptake at lower temperatures), indicating non-carrier-mediated passive diffusion. Uptake was also saturable with time or increasing antibiotic concentration. The saturable intracellular binding site, most likely the target isoleucyl-tRNA synthetase as determined by the amount of bound mupirocin (2,700 to 3,100 molecules per cell), caused concentration of the antibiotic within the cell. E. coli transformed with a plasmid containing ileS overproduced the target enzyme and demonstrated greater accumulation of mupirocin than a strain containing a control plasmid. The concentrations needed to half saturate (Kd) these binding sites in B. subtilis and S. aureus were 35 and 7 nM, respectively. In gram-positive organisms trained for mupirocin resistance, uptake was not saturable with increasing antibiotic concentration, and intra- and extracellular concentrations of drug equilibrated with time. Kinetic analysis of crude isoleucyl-tRNA synthetase from trained and untrained B. subtilis strains revealed differences in apparent Ki for mupirocin (resistant strain SB23T, Ki = 71.1 nM; sensitive strain SB23, Ki = 33.5 nM), while the Km for isoleucine remained unchanged (2.7 to 2.9 microM). A Km of 0.4 micromolar isoleucine and Ki of 24 nM mupirocin was demonstrated for isoleucyl-tRNA synthetase from sensitive S. aureus 730a, while no isoleucyl-tRNA synthetase activity was detected in extracts of resistance-trained S. aureus 3000 even at 40 micromolar isoleucine, suggesting instability of the enzyme. Free isoleucine pools differed between sensitive (0.26 micromolar) and resistance-trained (1.06 micromolar) S. aureus. Our results demonstrate that (i) mupirocin enters cells by passive diffusion, (ii) mupirocin concentrates in sensitive bacteria due to binding to isoleucyl-tRNA synthetase, and (iii) resistance to mupirocin involves restricted access to the binding site of isoleucyl-tRNA synthetase.