| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
REVIEWS |
a The Skaggs Institute for Chemical Biology, The Scripps Research Institute, Beckman Center, La Jolla, California 92037, USA
b Cubist Pharmaceuticals, Inc., Cambridge, Massachusetts 02139, USA
ABSTRACT
Because resistance has developed to mainline antibiotics, including vancomycin, new antibiotics are now being aggressively sought. For this purpose, aminoacyl tRNA synthetases are being pursued as targets for new drugs. These enzymes are universal and are essential for cell viability. The key to their usefulness lies in being able to find drugs that inhibit a pathogen synthetase but not its human cell counterpart. The possibility for species-specific inhibition was originally demonstrated with a natural product and has now been demonstrated with prototypical drugs that are based on the structure of an intermediate of the aminoacylation reaction. Efficacy of a rationally designed inhibitor has been shown in vivo with a pathogen infection established in an animal model. Although many challenges remain, these early results suggest that synthetases will continue to be of major interest for development of new anti-infectives.—Schimmel, P., Tao, J., Hill, J. Aminoacyl tRNA synthetases as targets for new anti-infectives. FASEB J. 12, 1599–1609 (1998)
Key Words: antibiotic resistance • translation apparatus • pathogen-specific inhibitor • vancomycin • mainline antibiotics
THE AMINOACYL tRNA synthetases are essential proteins found in all living organisms (1–3). The enzymes catalyze the attachment of amino acids to transfer RNAs (tRNAs)2 and thereby establish the rules of the genetic code by virtue of matching the nucleotide triplet of the tRNA anticodon with its cognate amino acid. They have emerged as leading targets for the development of new antibiotics. The goal is to develop inhibitors of the activity of one or more of these enzymes from prominent pathogens such as Enterococcus faecalis, Staphylococcus aureus, Streptococcal pneumoniae, Helicobacter pylori, Mycobacterium tuberculosis, Candida albicans, and others. The pursuit of this goal is motivated in large part by the widespread appearance of the resistance of infectious pathogens to mainline antibiotics.
EMERGENCE OF ANTIBIOTIC RESISTANCE
Since the turn of the century, the life expectancy of residents of the United States has almost doubled, expanding from about 40 years in 1900 to the mid-70s today. Many factors account for the decrease in mortality rates, but particularly significant among them was the development of antibiotics, starting with the discovery of penicillin by Alexander Fleming in 1928. Penicillin blocks an essential transpeptidation step in cell wall biosynthesis in infectious organisms such as S. aureus by forming an penicilloyl-enzyme intermediate. It was a true `wonder drug', until it was found that microorganisms have powerful adaptive mechanisms to develop multiple resistance mechanisms. These mechanisms include hydrolysis of the drug's ß-lactam ring by ß-lactamase, alterations in penicillin binding proteins, decreased cell entry, and active efflux of the drug (4).
The lesson learned with penicillin has been repeated over and over with the development of each new antibiotic, including methicillin, the quinolones, and vancomycin, among many others (
Fig. 1).
Vancomycin is now the drug of choice to treat potentially lethal infections from S. aureus (5), an organism that is one of the most common causes of nosocomial infections (6). But vancomycin resistance has started to emerge (7, 8) and, given the history with penicillin and other wonder drugs, the development of widespread vancomycin-resistant organisms is only a matter of time ( Fig. 2).
|
|
These circumstances have motivated intense efforts to develop new antibiotics that overcome the emerging vancomycin-resistant bacteria. The cellular targets for existing and potentially new antibiotics are diverse; all are essential for cell growth. They generally can be divided into two groups: those targets that are unique to infectious microorganisms and have no homologs in mammalian cells, and those that have homologs in mammals. Penicillin and vancomycin are examples of the former; each inhibits a step in the synthesis of a cell wall structure that is not found in mammals. Such antibiotics have the advantage that the likelihood of interfering with a mammalian cell biochemical step is lessened.
All other antibiotics are directed at targets that provide the same essential functions in the pathogen and in the host cell. Thus, to avoid toxic side effects, these antibiotics must differentiate between the pathogen target and its homolog in humans. The number of potential targets in this group is far greater, given that so many essential functions are conserved through evolution and use homologous components.
ANTIBIOTICS DIRECTED AT THE TRANSLATION APPARATUS
Of those found in all living organisms, the components of the translation apparatus are prominent among the targets for antibiotics that have been developed so far ( Table 1).
These antibiotics are natural products that include aminoglycosides, macrolides, and tetracyclines among others. The 30S and 50S ribosomal subunits are prominent among the targets of these drugs, which block protein synthesis in specific pathogens but not in mammalian cells. These antibiotics can sometimes be used to treat infections by organisms that are insensitive to or resistant to major antibiotics such as penicillin, methicillin, etc. Most have limitations (lack of oral availability, some toxicity in humans) and, for most of them (for example, erythromycin), resistant organisms have appeared (9).
|
In general, targets in translation for potential new drug development include the ribosomes, specific ribosomal proteins, ribosomal RNAs, and factors for chain initiation, elongation, translocation, and termination. For some of these, such as ribosome-based targets, in vitro assays used in drug screening can be cumbersome. For others, such as factors for elongation and translocation, their close structural relationships to other members of the G-factor superfamily raise concerns about whether cross-reactions of drug candidates with one of the many mammalian cell G-factor proteins would make the problem of avoiding toxicity difficult.
One of the translation-directed antibiotics listed in 
Table 1 is a natural product that is targeted to a specific tRNA synthetase. Pseudomonic acid (mupirocin) is synthesized by Pseudomonic fluorescens (10) and is an inhibitor of isoleucyl-tRNA synthetases from Gram-positive infectious pathogens, including S. aureus, S. epidermidis, and S. saprophyticus and Gram-negative organisms such as Haemophilus influenzae, Neisseria gonorrhoeae, and Neisseria meningitidis. It is active against clinical isolates of S. aureus that are resistant to penicillin, streptomycin, methicillin, and other mainstream antibiotics (11). Pseudomonic acid has an approximately 8000-fold selectivity for pathogen vs. mammalian isoleucyl-tRNA synthetase (12). But the drug's lack of systemic bioavailability has limited its use to external applications, i.e., to infections of the skin (13).
[Other known natural product inhibitors directed against synthetases and their apparent amino acid specificities include borrelidin (threonine) (14, 15), furanomycin (isoleucine) (16), granaticin (leucine) (17), indolmycin (tryptophan) (18), ochratoxin A (phenylalanine) (19), and cispentacin (proline) (20). Even though none of these cell growth inhibitors has been developed into an antibiotic for one reason or another, they illustrate the diversity of natural products that have been identified as inhibitors of tRNA synthetases.]
The experience with pseudomonic acid raised the question of whether, more generally, pathogen-specific inhibitors might be developed against any of the 20 tRNA synthetases. The appeal of these enzymes is that they are soluble, relatively stable, easy to express from recombinant genes, purify in large amounts, and straightforward to assay by one or more methods. In addition, X-ray structures are available for over half of the synthetases (21–24), and much is known about the mechanism and chemistry of the aminoacylation reaction. Based on this information, the rational design of inhibitors is realistic and might be accomplished with schemes of chemical synthesis that are far easier than required by the complex structures of many natural products.
tRNA SYNTHETASES AS TARGETS FOR ANTI-INFECTIVES
Rationale
Typically in prokaryotes there is just one synthetase for each amino acid, thus eliminating problems associated with enzyme isotypes that have slight variations in their sequences (25, 26). These variations might make one isotype insensitive to inhibitors directed at another isotype for the same enzyme activity. Because the active sites of these enzymes are constructed with one of two basic architectures (27–31), a chemistry platform that produces inhibitors for each of these architectures might be applied reiteratively to generate inhibitors for each of the enzymes. In addition, a platform of this sort might also yield drugs that are active simultaneously on two or more synthetases. The main disadvantage of the synthetases is the potential for toxic effects due to insufficient specificity of a drug for the pathogen as opposed to its human counterpart. Pseudomonic acid is a complex natural product, and it is not obvious that inhibitors rationally based on analogs of synthetase substrates or intermediates will have the desired specificity.
Aminoacylation reaction and synthetase architectures
Most synthetases catalyze the aminoacylation reactions in two steps:
 |
 |
The 20 synthetases are divided into two classes of 10 enzymes each, based on their active site architectures. Class I enzymes have a Rossmann nucleotide binding fold, similar to that found in the dehydrogenases (35–37). This architecture consists of an alternating pattern of ß-strands and 
-helices. This group of related enzymes was originally assigned through the identification of a signature sequence, a 12 amino acid sequence that ends in the tetrapeptide HIGH (27, 29). A second sequence—KMSKS—also characterizes this class of enzymes (28). These sequence motifs are critical elements in the structure of the site for adenylate synthesis. Another distinguishing characteristic of class I enzymes is that they approach the end of the tRNA acceptor helix from the minor groove side and catalyze initial attachment of the amino acid to the 2'-OH at the end of the tRNA chain (21, 38).
Class II enzymes have an entirely different active site design, comprised of a seven-stranded ß-structure with three -helices (31, 39). Three sequence elements—motifs 1, 2, and 3—characterize this group of enzymes (30). These elements, which are highly degenerate, form critical parts of the site for adenylate synthesis. The class II synthetases approach the end of the tRNA acceptor stem from the major groove side, opposite of that seen with the class I enzymes (39–41). With one possible exception, they catalyze attachment of amino acids to the 3'-OH at the end of the tRNA acceptor stem, also the opposite of that found with class I enzymes.
Sequences of over 300 tRNA synthetases are now known (42). These sequences show that the two architectures that characterize these enzymes' active sites are ancient, being found in early organisms such as Aquifex aeolicus, which is near the base of the evolutionary tree of life (43). With the exception of lysyl-tRNA synthetase (44), all of the enzymes sequenced to date have remained fixed in the same class throughout evolution. Based on sequence analyses, no evidence supports the idea that the two classes split out of the same ancestral structure (2, 3). Thus, if the two classes have a common ancestor, then separation must have occurred quite early. The strong conservation of synthetase structure throughout evolution means that a drug platform directed at one type of active site architecture could be applicable broadly to many pathogen synthetases.
Pathogen-specific inhibitors based on reaction intermediates
The most obvious chemistry platforms for synthetase-directed drugs are those based on the structures of the substrates and intermediates of the aminoacylation reactions catalyzed by these enzymes. The main issue is whether such compounds would be sufficiently selective for the pathogen synthetase so as not to interfere with human cell function. As a rule of thumb, selectivity of greater than 100-fold is desirable.
The aminoacyl adenylate intermediate ( Fig. 3)
is bound tightly to the synthetase, with dissociation constants typically in the nanomolar range (45). Dissociation constants for amino acids and ATP are higher, generally by two to three orders of magnitude. Thus, analogs based on the adenylate intermediate have the potential for the tight binding needed for a drug.
|
For class I enzymes, the site for adenylate synthesis is encoded by a Rossmann nucleotide binding fold that is split by an insertion known as connective polypeptide 1 (CP1) (46, 47). [In some class I enzymes, this insertion contains a second catalytic site that is used to clear misactivated amino acids in a tRNA-dependent editing reaction (48).] An alignment of sequences in the active site region of isoleucyl-tRNA synthetases from three pathogens—S. aureus, E. faecalis, and Escherichia coli—with that of the human counterpart shows many positions on either side of the CP1 insertion where the pathogen synthetases have residues distinct from those of the human enzyme ( Fig. 4).
These differences suggest that pathogen-specific inhibitors might be obtained on the basis of these differences.
|
Examples of seven adenylate-like inhibitors of isoleucyl-tRNA synthetases are shown in Table 2
(49, 50). A scheme for the synthesis of these sorts of compounds is given in Fig. 5
(49).These inhibitors retain the isoleucyl moiety of the adenylate and replace the labile acylphosphate linkage of the adenylate with a stable sulfamate or sulfonamide linkage. The primary example of this group is CB138, which retains the adenosine moiety. Others retain the ribose ring of adenosine, but replace adenine with a tetrazole that is linked to one or two additional five- or six-member aromatic or heterocyclic rings. This kind of diversity of structures, made around a common framework, provides an opportunity to test the sensitivity of synthetase inhibition to structural variations in the adenylate analogs.
|
|
Results on inhibition of isoleucyl-tRNA synthetase from two Gram-positive organisms (S. aureus and E. facaelis), from one Gram-negative organism (E. coli), and from human cytoplasm are shown in Table 3.
All seven are excellent inhibitors of enzymes from Gram-positive and Gram-negative organisms, with apparent dissociation constants collectively in the range of 0.5 to 40 nM. Each interacts more weakly with the human enzyme. The smallest selectivity for the pathogen synthetases (7- to 10-fold) is shown by the simplest inhibitor (CB138). In contrast, CB432, for example, shows a 60- to 1100-fold discrimination in favor of the pathogen vs. the human enzyme. Although the structural basis for this high selectivity is not known, the data are consistent with the suggestion from Fig. 4that sequence differences in the active site region can be exploited.
|
Subsequent work with CB432 established that this prototypical drug arrested cell growth of S. aureus, S. pyogenes, and E. coli (permeability mutant) in culture with minimum inhibitory concentrations (MIC) of 10, 0.5, and 10 µg/ml, respectively (49, 50). Macromolecular labeling experiments with S. aureus showed that CB432 is a selective inhibitor of protein synthesis. Increases in MIC values for S. aureus grown in the presence of large concentrations of isoleucine provided evidence that the intracellular target of CB432 was isoleucyl-tRNA synthetase.
Therapeutics with a mouse model
To determine the usefulness of CB432 in a model therapeutic application, a mouse system was used (49). Mice were inoculated with lethal infectious doses of the Gram-positive pathogen S. pyogenes. (CB432 was found to be particularly effective against S. pyogenes in whole cell assays.) Subsequently, the mice were treated with various doses of the CB432 and with erythromycin, a drug known to be active against S. pyogenes. In the absence of treatment with either CB432 or erythromycin, five of five mice succumbed within 7 days ( Table 4).
In contrast, subcutaneous injections of CB432 resulted in increasing numbers of mice that survived, the total number of survivors being dependent on the dose of CB432 that was administered. At the highest dose, all of the subjects survived, with the results matching that achieved with the control (erythromycin).
|
In the example shown, the dose required for efficacy of CB432 is too high to be useful for human applications. Subsequent work established that the high dosage required is correlated with a low bioavailability, due at least in part to the binding of CB432 to serum albumin (about 98.5% was determined to be bound to albumin.) But the data do establish that a relatively simple inhibitor of isoleucyl-tRNA synthetase is effective in an animal model. These data provide a starting point for further efforts directed at improving bioavailability and overall efficacy.
Possibility of drugs that disrupt tRNA interaction
The examples above follow the tradition of targeting drugs to the catalytic sites of enzymes, which in this instance are tRNA synthetases. For many enzymes of metabolism, the binding site for the substrate and product is imbedded within the catalytic site, because the ligands themselves are small. In the case of tRNA synthetases, the ATP and amino acid binding sites are virtually inseparable from the catalytic site. But the tRNA is far larger, and only its 3'-end has to be accommodated into the active site. Discrimination of nucleotides far from the 3'-end occurs through interactions that may be as much as 60 to 75 angstroms away (51–53). These binding interactions are essential for enzyme activity.
The tRNA structure is an L-shaped molecule comprised of two domains corresponding to each arm of the L (54, 55). One arm contains the tRNA acceptor stem and the amino acid attachment site, and the other arm harbors the anticodon triplet of the code, which is about 75 angstroms from the amino acid acceptor end. To a first approximation, the synthetases also have two domains that mirror the two domains of the tRNAs. One synthetase domain makes contact with the acceptor domain of the tRNA whereas the other synthetase domain makes contact with the second arm of the tRNA (51, 56). Thus, compounds that block these protein-RNA domain–domain interactions are potential drugs.
The key parts of the RNA binding elements in some of the synthetases have been identified and rationally manipulated in such a manner as to suggest that these elements are excellent targets for drugs. For example, eukaryote tyrosyl-tRNA synthetase does not acylate bacterial tRNATyr, and vice versa. The reason for this species specificity is due in large part to the difference of a single base pair near the acceptor end of tRNATyr: C:G for eukaryote and G:C for bacterial tRNATyr. Interchange of this base pair switches the species specificity of acylation, so that the eukaryote enzyme now charges the bacterial substrate and vice versa (57). The site on the synthetase responsible for this G:C vs. C:G discrimination has been localized to a 39 amino acid segment within the CP1 insertion (vide supra) (58). Thus, a small molecule that binds to this segment in the bacterial enzyme could be a potential drug.
Similarly, some synthetases make contact with the anticodon trinucleotide (53, 59). The structural details for this nucleotide triplet binding site on a class I and in a class II synthetase have been worked out (60, 61). Point mutations in an anticodon binding site can severely reduce acylation efficiency because of the weakening of the tRNA interactions. Moreover, in one example, simple sequence manipulations within the anticodon binding site have allowed the specificity of binding to be changed from one nucleotide triplet to another (62, 63). Thus, for some synthetases the anticodon binding site is an attractive target for drug development.
Certain tRNA synthetases have an editing activity that corrects errors of amino acid activation (Eq. 1) or aminoacylation (Eq. 2) (64, 65). For example, isoleucyl-tRNA synthetase misactivates valine and forms valyl-AMP and Val-tRNAIle. However, these reaction intermediates and products are cleared by a hydrolytic activity that destroys the misactivated or mischarged products (65–67). This activity is located within a second catalytic site that, for at least two synthetases, has been localized to the CP1 insertion (48, 68–70). Whether blocking of the editing site in CP1 of a pathogen synthetase would result in cell lethal is worth investigating.
CONCLUDING REMARKS
There is no evidence that new drugs directed against tRNA synthetases per se will be less prone to the emergence of antibiotic resistance. A point mutation in a bacterial isoleucyl-tRNA synthetase resulted in sharply lower sensitivity to pseudomonic acid, for example (71). In addition, transfer of a plasmid-encoded gene encoding a drug-resistant isoleucyl-tRNA synthetase from one organism to another has also been demonstrated (72–74). But these considerations do not diminish the interest in synthetases as targets for new antibiotics, because all new antibiotics have an important role in at least temporarily controlling infections not arrested by existing drugs that are no longer effective. In addition, because the active sites fall into one of two basic architectures, the hope is that architecture-specific drugs might be developed that inhibit simultaneously two or more synthetases. For example, valyl-, isoleucyl-, and leucyl-tRNA synthetases are particularly close homologs in their active site regions and offer the possibility of a drug that might bind to at least two of these enzymes. Also, a chemistry platform that allowed simple structural manipulations of a drug directed at one synthetase—so that it now interacts with a similar but different synthetase—offers a way to adapt quickly to a resistance that may emerge.
The tRNA amido transferases found in Gram-positive organisms may also be attractive targets. In these organisms, there is no glutaminyl- or asparaginyl-tRNA synthetase. Instead, glutamyl-tRNA synthetase (or aspartyl-tRNA synthetase) catalyzes attachment of glutamate to tRNAGln (or aspartate to tRNAAsn) (32, 34). The mischarged tRNAGln (or tRNAAsn) then reacts with a specific amido transferase that converts Glu-tRNAGln to Gln-tRNAGln (or Asp-tRNAAsn to Asn-tRNAAsn). Because these amido transferases are essential for growth of a pathogen, they have obvious appeal for the development of new anti-infectives.
ACKNOWLEDGMENTS
We thank Dr. Francis P. Tally and Dr. Tongchuan Li for helpful discussions about pathogen infections.
FOOTNOTES
1 Correspondence: The Skaggs Institute for Chemical Biology, The Scripps Research Institute, Beckman Center, 10550 N. Torrey Pines Road, La Jolla, CA 92037, USA. E-mail: Schimmel{at}Scripps.edu 
2 Abbreviations: CP1, connective polypeptide 1; MIC, minimum inhibitory concentrations; tRNA, transfer RNA.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  J. G. Hurdle, A. J. O'Neill, and I. Chopra Prospects for Aminoacyl-tRNA Synthetase Inhibitors as New Antimicrobial Agents Antimicrob. Agents Chemother., December 1, 2005; 49(12): 4821 - 4833. [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Wolfe, J. A. Warrington, L. Treadwell, and M. T. Norcum A Three-dimensional Working Model of the Multienzyme Complex of Aminoacyl-tRNA Synthetases Based on Electron Microscopic Placements of tRNA and Proteins J. Biol. Chem., November 18, 2005; 280(46): 38870 - 38878. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Buddha and B. R. Crane Structures of Tryptophanyl-tRNA Synthetase II from Deinococcus radiodurans Bound to ATP and Tryptophan: INSIGHT INTO SUBUNIT COOPERATIVITY AND DOMAIN MOTIONS LINKED TO CATALYSIS J. Biol. Chem., September 9, 2005; 280(36): 31965 - 31973. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. T. Kudva, R. W. Griffin, J. M. Garren, S. B. Calderwood, and M. John Identification of a Protein Subset of the Anthrax Spore Immunome in Humans Immunized with the Anthrax Vaccine Adsorbed Preparation Infect. Immun., September 1, 2005; 73(9): 5685 - 5696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Beyer, H.-P. Kroll, R. Endermann, G. Schiffer, S. Siegel, M. Bauser, J. Pohlmann, M. Brands, K. Ziegelbauer, D. Haebich, et al. New Class of Bacterial Phenylalanyl-tRNA Synthetase Inhibitors with High Potency and Broad-Spectrum Activity Antimicrob. Agents Chemother., February 1, 2004; 48(2): 525 - 532. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kitabatake, K. Ali, A. Demain, K. Sakamoto, S. Yokoyama, and D. Soll Indolmycin Resistance of Streptomyces coelicolor A3(2) by Induced Expression of One of Its Two Tryptophanyl-tRNA Synthetases J. Biol. Chem., June 21, 2002; 277(26): 23882 - 23887. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Austin and E. A. First Catalysis of Tyrosyl-Adenylate Formation by the Human Tyrosyl-tRNA Synthetase J. Biol. Chem., April 19, 2002; 277(17): 14812 - 14820. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Szymanski and J. Barciszewski Aminoacyl-tRNA synthetases database Y2K Nucleic Acids Res., January 1, 2000; 28(1): 326 - 328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. V. Franklund and J. B. Goldberg Cloning of the Glutamyl-tRNA Synthetase (gltX) Gene from Pseudomonas aeruginosa J. Bacteriol., June 1, 1999; 181(11): 3582 - 3586. [Abstract] [Full Text]
|
|
|
|
|
|
M. T. Norcum and J. D. Dignam Immunoelectron Microscopic Localization of Glutamyl-/ Prolyl-tRNA Synthetase within the Eukaryotic Multisynthetase Complex J. Biol. Chem., April 30, 1999; 274(18): 12205 - 12208. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S.-i. Sekine, A. Shimada, O. Nureki, J. Cavarelli, D. Moras, D. G. Vassylyev, and S. Yokoyama Crucial Role of the HIGH-loop Lysine for the Catalytic Activity of Arginyl-tRNA Synthetase J. Biol. Chem., February 2, 2001; 276(6): 3723 - 3726. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Beuning and K. Musier-Forsyth Species-specific Differences in Amino Acid Editing by Class II Prolyl-tRNA Synthetase J. Biol. Chem., August 10, 2001; 276(33): 30779 - 30785. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
