X-ray crystallographic studies of aminoacyl-tRNA synthetase: a target for the development of new antibiotics.

Antibiotics are one of the corner pieces of our healthcare system, as they are at the frontline in the prevention against and treatment of infections with pathogenic bacteria. Unfortunately, the inevitable emergence and spread of antibiotic resistant bacteria has hampered their clinical use. Together with a lack of the discovery and development of novel antibacterial classes, the worldwide overconsumption and misusage of the available antibiotics has aided into the selection and spread of multi-resistant bacterial strains for which last-resort antibiotics need to be used or even worse no treatments are available. To contain the continuous increase in antibiotic resistance in pathogenic bacterial strains, it is essential that new antibiotic classes are discovered and developed for clinical use.

The family of aminoacyl-tRNA synthetases (aaRSs) are recognized as an excellent target for the development of new antibiotics. aaRSs are essential for cell viability, as they charge the correct amino acid to the corresponding tRNA molecule, via a two-step mechanism. In the first step the amino acid is activated by a nucleophilic attack on the α-phosphate of ATP generating an aminoacyl-AMP intermediate and pyrophosphate. In the second step the activated amino acid is transferred to the 2′ or 3′ -hydroxyl of the ribose of the terminal tRNA adenosine.

aaRSs are already clinically validated as an antibacterial target as mupirocin, a bacterial-selective isoleucyl-tRNA synthetase (ileRS) is already available on the market for the topical treatment of skin infections. Next to mupirocin, a large number of natural and synthetic aaRS inhibitors have been discovered. The largest group of the inhibitors described so far are the non-hydrolysable analogues of aa-AMP, where the labile phosphodiester linkage has been replaced with a hydrolytically stable linkage. Among them, 5'-O-(N-L-aminoacyl)-sulfamoyl adenosine (aaSA) analogues have proven to be potent nanomolar inhibitors in vitro. Although having a strong affinity for the target aaRS, these compounds lack in vivo efficacy, have limited species selectivity, are chemically unstable and are prone to enzymatic modification.

In the first three research chapters of this work, we investigated how the chemical and pharmacological parameters of the aaSA scaffold can be optimized. To improve the chemical stability, our collaborators generated 3-deazaadenosine (3DA) analogs (chapter three). To fully assess the consequence of substituting this single atom, we have performed a comprehensive family wide analysis of the compounds evaluating the aaS3DA analogues against both class I and class II aaRS representatives. The in-vitro assay results point to a clear class bias, as the inhibitory activity of the class II-targeting inhibitors was severely affected by the substitution, while the class I aa3DA inhibitors reporting an limited reduced inhibitory activity compared to the native aaSA. X-ray structures of aspRS, leuRS, ileRS and tyrRS were solved in complex with the synthesized molecules providing insight in the obtained activity data. In a similar vein a previous publication1 reported that pyrimidine substituted analogs are good ileRS inhibitors. We therefore investigated if this approach could be extrapolated to other class I aaRS (chapter four). Despite sharing the same active site architecture, differences in inhibitory response to the various pyrimidine bases have been observed between class I family members. Structural analysis shows that subtle difference in base-aaRS interactions are responsible for this behavior. Additionally we evaluated if these compounds are able to circumvent the natural resistance mechanism against aaSA, acetylation of the primary amine by N-acetyltransferases rimL and YhhY (chapter five). Biochemical data shows that adenine-containing nucleoside analogs are more prone to degradation by rimL over the pyrimidine bases. This is in contrast to YhhY, which prefers the latter over the natural base. Structural data of the compounds in complex with rimL is obtained and a model of YhhY is generated, to explain the base selectivity.

IleRS has been the target aaRS of many academic and industrial efforts yielding a large plethora of different inhibitors. Unfortunately, the lack of structural information on inhibitor:aaRS interactions hampered the further developed of these compounds. Using a wide variety of ileRS constructs we identified a crystallization condition which allowed us to solve the structures of ileRS in complex with inhibitors (chapter six). Besides a detailed understanding of the effect of various inhibitors on their target aaRS, we fortuitously unraveled a new tri-lobed architecture of an ileS2 type anticodon-binding domain (ACBD) that has important consequences on our understanding of the evolution of these ancient enzyme family (chapter seven).

In summary, the obtained structures of inhibitor:aaRS complexes provide an excellent starting point for further structure-based drug design of improved aaRS inhibitors.