Probing the Evolution of New Specificities in Aminoacyl-Trna Synthetases

Home , Elongation factor P, Initiation factor, Peptidyl transferase, Release factor, Start codon

Probing the Evolution of New Specificities in Aminoacyl-tRNA Synthetases

Presented in Partial Fulfillment of the Requirements for the Master’s Degree in the Graduate School of The Ohio State University

Marla S. Gilreath, B.S.

Graduate Program in Biochemistry

The Ohio State University

2011

Master’s Committee:

Dr. Venkat Gopalan, Advisor

Dr. Michael Ibba Copyright by

Marla S. Gilreath

2011 ABSTRACT

Bacterial elongation factor P (EF-P) is a poorly understood soluble protein that has been shown to enhance the first step of peptide bond formation through an interaction with the ribosome and initiator tRNA. Homologous proteins have been found in both archaeal and eukaryotic systems, known as aIF5A and eIF5A, respectively. eIF5A, which was recently shown to increase translation elongation rates, is post-translationally modified at a highly conserved lysine residue through the addition of the rare amino acid hypusine. A similar pathway was recently elucidated for EF-P, in which EF-P is post- translationally modified by the enzymes PoxA and YjeK at lysine 34, corresponding to a homologous site of hypusination in a/eIF5A. As a paralog of class II LysRS, PoxA catalyzes the addition of lysine onto EF-P, but is incapable of modifying tRNA. YjeK is a

2,3-(β)-lysine aminomutase and is responsible for converting lysine to β-lysine, which

PoxA was recently shown to recognize as a preferred substrate for EF-P modification.

The amino acid binding pockets of LysRS and PoxA are highly conserved, with the exception of two residues, Gly465 and Ala229 of Geobacillus stearothermophilus

LysRS and Ala298 and Ser76 of Salmonella Typhimurium PoxA. Despite their substantial active site similarity, PoxA exhibits a significantly higher KM value for activation of lysine as compared to LysRS. This suggests that the two divergent residues in the active site determine the specificity for substrate recognition and binding, as well

ii as optimal enzymatic activity, of PoxA. To investigate the mechanisms of α- versus β- amino acid recognition in the divergent evolution of LysRS and PoxA, three amino acid replacements were made in the LysRS active site. Kinetic parameters for ATP/PPi exchange reactions were determined for wild type, A233S, and G469A LysRS and aminoacylation reactions were carried out to further characterize the activity of each variant. Results indicate that while A233S behaves like the wild type, G469A and

A233S/G469A significantly decrease the ability of LysRS to form stable lysyl- adenylates. A233S was able to shift the substrate specificity of LysRS to recognize (S)-β- lysine, indicating that few active-site substitutions are necessary to facilitate changes in the substrate specificity of an aaRS.

iii

ACKNOWLEDGEMENTS

I would like to extend a sincere thank you to Dr. Michael Ibba for fostering my graduate education in biochemistry and expanding my knowledge of experimental techniques. I am truly grateful for the time I spent learning from Dr. Ibba and my fellow

Ibba Lab members, in particular Dr. Hervé Roy, who patiently facilitated my training as a protein biochemist. I would also like to thank Dr. Venkat Gopalan for holding the position as my advisor of record.

VITA

June 2004 ...... Cumberland Valley High School

May 2008 ...... B.S. Biochemistry, University of Delaware

2008 to present ...... Graduate Research Assistant, Department of

Microbiology, The Ohio State University

PUBLICATIONS

Roy, H., Zou, S.B., Bullwinkle, T., Wolfe, B., Gilreath, M., Forsyth, C., Navarre, W.W., and Ibba, M. The tRNA synthetase paralog PoxA modifies elongation factor P with (R)-β-lysine. Nature Chemical Biology (In press)

Banerjee, R., Chen, S., Dare, K., Gilreath, M., Praetorius-Ibba, M., Raina, M., Reynolds, N., Rogers, T.E., Roy, H., Yadavalli, S.S., and Ibba, M. (2009) tRNA: Cellular barcodes for amino Acids. FEBS Lett. 584: 387-395.

FIELDS OF STUDY

Major Field: Biochemistry

TABLE OF CONTENTS

ABSTRACT...... ii

ACKNOWLEDGEMENTS...... iv

VITA...... v

PUBLICATIONS...... v

FIELDS OF STUDY ...... v

TABLE OF CONTENTS...... vi

LIST OF TABLES...... vii

LIST OF FIGURES ...... viii

INTRODUCTION ...... 1

MATERIALS AND METHODS...... 26

RESULTS ...... 30

DISCUSSION...... 38

OUTLOOK ...... 41

REFERENCES ...... 45

LIST OF TABLES

Table 1. Concentrations of purified proteins…………………………………………….31

Table 2. Steady-state kinetics of lysine activation……………………………………….34

vii

LIST OF FIGURES

Figure 1. aaRS-catalyzed aminoacylation reaction...... 2

Figure 2. Aminoacylation and co-translational insertion of an amino acid...... 3

Figure 3. Overview of mRNA-encoded protein synthesis...... 5

Figure 4. Structural comparison of EF-P with tRNA and other ribosome-binding proteins

...... 8

Figure 5. Crystal structure of EF-P bound to the 70S ribosome...... 12

Figure 7. Proteins aIF5a and eIF5A from M. jannaschii and L. mexicana...... 15

Figure 8. Structural comparison of EF-P and eIF5A ...... 16

Figure 9. Post-translational modification pathways of eIF5A and EF-P...... 17

Figure 10. Current model for the post-translational modification of EF-P...... 19

Figure 11. PoxA-catalyzed in vitro lysylation of EF-P...... 20

Figure 12. Molecular docking model comparison of LysRS and PoxA active sites...... 21

Figure 13. Aminoacylation of LysRS wild type and variants...... 32

Figure 14. Representative Michaelis-Menten curves for wild type (A), G469 (B), and

A233S (C) ...... 33

Figure 15. Amino acid specificity of LysRS wild type and A233S with (S)-β-lysine and

(R)-β-lysine ...... 36

viii

INTRODUCTION

mRNA-encoded Protein Synthesis

First posed as the “coding problem” by biologists in the late 1950s, translation involves the conversion of genetic information from DNA to protein by following a set of rules deemed the genetic code. During translation, the process of protein synthesis, nucleotides in a messenger RNA (mRNA) are decoded to direct the incorporation of amino acids into a growing peptide chain. Each set of three consecutive nucleotides, referred to as codons, either directs the incorporation of a specific amino acid or signals a stop to translation. At least one codon represents each of the 20 amino acids, with multiple codons existing for the more frequently used amino acids. The same genetic code is universally employed by all organisms for protein synthesis, although some differences do occur, for example, in the DNA of mitochondria and in other rare instances (1).

The accurate translation of mRNA to protein requires the assistance of adaptor molecules capable of binding to both the mRNA codon and the cognate amino acid to be incorporated. These substrates for translation are termed aminoacyl-tRNAs (aa-tRNAs), and there exists one tRNA with the correct anticodon region for each amino acid of the genetic code. The 20 aa-tRNAs can be divided into three categories: translation substrates

(canonical elongator aa-tRNAs, initiator aa-tRNAs, or non-canonical aa-tRNAs), misacylated translation substrates, or nontranslation substrates (2).

Cognate amino acids are esterified onto the 3’-end of the mature transfer RNA

(tRNA) through a reaction catalyzed by aminoacyl-tRNA synthetases (aaRSs). This aminoacylation reaction occurs in two steps: first the amino acid is activated with ATP to form an aminoacyl-adenylate (aaAMP), and second the activated amino acid is transferred to its associated tRNA (3). Each aa-tRNA is synthesized by a unique aaRS molecule, with cognate pairing of the tRNA and amino acid directed by preferential binding of the cognate amino acid and selective editing of near-cognate amino acids (3).

The mRNA must be faithfully translated during protein synthesis, and the fidelity of this process is based on both the availability of aa-tRNAs composed of correct amino acid:tRNA pairs, as well as the accurate selection of aa-tRNAs on the ribosome (3, 4).

Aminoacyl-tRNA Synthetases (AaRSs)

AaRSs enzymatically attach amino acids to their cognate tRNAs in a manner that assures a high level of specificity. Forming correctly paired aa-tRNAs occurs in a two- step reaction (Figure 1) (4).

aa + aaRS + ATP aaAMP + aaRS + PPi (1)

aaAMP + aaRS + tRNA aa-tRNA + aaRS + AMP (2)

Figure 1. aaRS-catalyzed aminoacylation reaction.

In the first step of this reaction, the aminoacyl-adenylate (aaAMP) is formed through attack of the α-carboxylate of the amino acid on the α-phosphate group of ATP. The terminal ribose of the tRNA then nucleophilically attacks the aaAMP to form aa-tRNA and AMP is released (5). Once released from the aaRS, elongation factor-Tu (EF-Tu) then escorts the amino acid to the ribosome for addition to the growing peptide chain

(Figure 2).

Figure 2. Aminoacylation and co-translational insertion of an amino acid. Reproduced from (6).

There are two structurally distinct classes of aaRSs, I and II, with 10 members in each class. The classes are divided based on catalytic core structure and the manner in which the aaRS catalyzes the formation of the aminoacyl-adenylate (7). Each of the 20 amino acids is charged by either a class I or II synthetase, with the exception of lysine. A lysyl-tRNA synthetase (LysRS) exists in both class I and II. LysRS1 is found in most archaea and some bacteria, while LysRS is present in all eukaryotes, most bacteria and

3 some archaea. Only a few organisms contain both LysRS1 and LysRS (4). Both LysRSs are capable of recognizing lysine and tRNALys to complete the lysylation reaction; however, the mode of lysine activation is distinct between the two synthetases. LysRS forms a lysyladenylate after binding ATP and lysine, but LysRS1 must also bind tRNALys to perform the charging reaction (8).

Although their main function is to charge amino acids for protein synthesis, aaRSs have roles in addition to translation. Recently it was discovered that binding of aa- tRNAs to EF-Tu is not necessarily an irreversible commitment to protein synthesis (9), indicating that aa-tRNAs may deliver amino acids to various other biological pathways.

These alternative cellular routes for activated amino acids include lipid modification pathways, cyclodipeptide formation and antibiotic biosynthesis (10, 11). Furthermore, the aaRS super family contains several known aaRS paralogs that mimic either the core or the appended domains of existing aaRSs (Roy et al, in press). AaRS paralogs generally retain some specificity for their canonical substrates, amino acid or tRNA, regardless of whether they function in translational pathways or have adapted a function outside of protein synthesis (12). Known paralogs function in trans to enhance aaRS activity, while others assist in tRNA and aaRS passage between the cytoplasm and nucleus, amino acid or antibiotic biosynthesis, or translation regulation (12).

The Ribosome

The bacterial 70S ribosome is a ribonucleoprotein machine composed of two subunits, which contribute different functions to the process of translating information from mRNA to form specific proteins (13, 14). Each subunit is composed of both RNA and protein and performs a detailed task in protein synthesis. To establish the sequence of amino acids in the peptide being synthesized, the small subunit (30S) recognizes the correct interactions between the anticodon of tRNA and the codons of the mRNA being translated. The large ribosomal subunit (50S) contains the site of peptide bond formation, known as the peptidyl-transferase center (PTC) (13). Together the small and large subunits contain three distinct binding sites for tRNA in the different stages of translation. The three stages of translation, initiation, elongation and termination, are illustrated in Figure 3.

Figure 3. Overview of mRNA-encoded protein synthesis. Reproduced from (13).

Initiation involves the joining of the small and large ribosomal subunits, correct positioning of the 70S ribosome at the mRNA start codon, and incorporation of the initiator tRNA in the peptidyl site (P-site) of the ribosome with the aid of initiation factors (IFs). Elongation of the peptide chain can then occur and requires two elongation factors, both of which hydrolyze GTP (1, 15). EF-Tu monitors the initial interaction between a charged tRNA and escorts the aa-tRNA to the ribosome as the ternary complex aa-tRNAGTPEF-Tu (16). Codon recognition allows EF-Tu to hydrolyze GTP and EF-

Tu is then released to bind another incoming aa-tRNA. Elongation factor G (EF-G) acts to accelerate the movement of the two bound tRNAs by binding in or near the acceptor site (A-site). The conformational change induced by EF-G binding and the subsequent

GTP hydrolysis advances the tRNA from the A-site to the P-site, creating an empty A- site for the next aa-tRNA and forcing the initiator tRNA in the P-site into the exit site (E- site) (1, 13, 16). A complete cycle of translation elongation can be seen in Figure 3. As the mRNA is read 5’ to 3’, amino acids are incorporated in this manner until a stop codon is reached and termination occurs. Both mRNA and the peptide chain are then released, the ribosome dissociates and protein synthesis is complete.

Elongation Factor P

Bacterial elongation factor P (EF-P) is a highly conserved soluble protein known to interact with the ribosome and initiator tRNA to enhance protein synthesis. EF-P was first distinguished from other translation factors by its ability to stimulate peptide bond

fMet formation between fMet-tRNAi and puromycin (17). EF-P shares no amino acid

6 similarities with other known translation factors (18) and substitution experiments with

EF-G and EF-Ts revealed that other elongation factors were incapable of stimulating the peptidyl transferase reaction to the extent of EF-P (19). Additionally, measuring the release of 35S from f[35S]Met-tRNAribosomeaminoacyloligonucleotide complexes during peptide bond synthesis indicated that EF-P does not equally enhance synthesis for all amino acids. This led to the hypothesis that EF-P may act specifically on amino acids with lower incorporation efficiencies, such as alanine or glycine, to normalize the rate of amino acid addition to a growing peptide chain (19).

Building upon these initial discoveries, EF-P has since been characterized as a tRNA mimic due to its similarity in both size and shape with the tertiary structure of tRNA. Thermus thermophilus EF-P exists as a monomer of 20 kDa under physiological conditions and is composed of 16 β-strands, which fold to form three distinct domains.

As seen in Figure 4, these three domains very closely resemble the L-shape of tRNA. The arm composed of domains I and II is 65 Å long and 23Å wide and the second arm, formed by domains II and III, is 53Å long and 25 Å wide (20). The N-terminal portion of

Escherichia coli EF-P shares similarity to the L18 and L27 ribosomal proteins, which are capable of cross-linking to puromycin to stimulate peptide bond synthesis (18).

Figure 4. Structural comparison of EF-P with tRNA and other ribosome-binding proteins. (A and B) EF-P from Thermus thermophilus at 1.65-Å resolution. (C)

Saccharomyces cerevisiae tRNAPhe. (D) T. thermophilus EF-G. (E) Ribosome recycling factor and (F) Release factor 2, both from E. coli. Reproduced from (20).

In vivo, EF-P associates with the 70S ribosome and 30S and 50S subunits and is found at 0.1 copies per 70S ribosome. The fraction of bound EF-P is known to decrease as a function of the number of ribosomes in the polyribosome fraction, consistent with a role in the initial stages of protein synthesis (21).

To further evaluate its role in protein synthesis, extensive crystallography of EF-P bound to the ribosome was performed (22). Despite its classification as a tRNA mimic, it was determined that EF-P does not bind to the ribosome in a traditional tRNA binding site, but rather adjacent to the P-site, between the P-site and E-site. EF-P spans both the

30S and 50S subunits, and makes contact with the initiator tRNA near the anticodon stem loop on the 30S subunit (Figure 5; 22). Only one major conformational change takes place in the ribosome upon EF-P binding, positioning the L1 stalk in the E-site to interact with EF-P. Neither the initiator tRNA in the P-site nor the PTC exhibit conformational changes in the presence of EF-P, indicating that EF-P most likely has an indirect effect on peptide bond formation. Overall, the crystal structure suggests a role for EF-P in facilitating an interaction with the tRNA backbone and providing stability for fMet-

fMet tRNAi in the P-site, while preventing premature movement of the initiator tRNA to the E-site. Although it is thought that this interaction shifts initiation toward the first elongation step of protein translation (22), the specific mechanism of action of EF-P has yet to be determined.

Figure 5. Crystal structure of EF-P bound to the 70S ribosome. Reproduced from (22).

EF-P and Bacterial Stress Response

In an attempt to define EF-P’s mechanism of action, studies with antibiotic resistance markers and [35S]-methionine claimed that EF-P is a vital component of eubacterial protein translation systems (23). These preliminary experiments presented a lethal phenotype and an 80% loss of peptide-bond forming activity in the absence of the efp gene (23). However, while systematically disrupting non-essential genes in E. coli to create the Keio collection, it was found that a strain of E. coli could tolerate the deletion of efp without a loss of viability (24).

Additionally, the efp gene became of recent interest during a screening for

Salmonella mutants resistant to S-nitroso-glutathione. This work noted the linkage of efp with two poorly characterized genes, poxA and yjeK, as seen in Figure 6, and these genes are necessary for tRNA and lysine biosynthesis, respectively (12). A comparative genomics study combining phylogenetic pattern analysis and physical clustering confirmed that the poxA, yjeK, and efp genes are functionally linked in a number of distinct bacterial species, supporting the hypothesis that PoxA and YjeK operate in the same pathway as EF-P (25). The poxA and yjeK genes display genetic epistasis, and poxA, yjeK, and efp interact in a pathway that is necessary for Salmonella virulence and resistance to a variety of cellular stressors (12).

Figure 6. Genetic map of the genes poxA, yjeK, and efp. Reproduced from (12).

Building upon this finding, Navarre et al (2010) then determined that EF-P is not essential for global protein production, but is instead required to maintain motility and cellular integrity under conditions of low osmolarity and antibiotic stress (12). A second study focusing on the Salmonella Typhimurium genes integral for survival in the swine gastric environment found similar sensitivities of poxA mutants to a variety of chemical and stress conditions, such as membrane detergents, chelators, antibiotics, toxic cations, and ionizing and oxidizing reagents (26), as well as an increase in metabolic activity (12,

26). The authors of this study suggested that the increased cellular sensitivity and metabolism in major catabolic pathways could be due to a disturbance of a general stress 12 response, which would cause unregulated growth during suboptimal conditions.

Furthermore, SDS-PAGE and sequencing analysis of poxA mutant protein profiles indicated differential protein expression, implying that PoxA regulates selective translation of some mRNA transcripts and therefore influences a number of bacterial processes (26).

Bacterial cells possess a regulatory mechanism that halts RNA synthesis under conditions of environmental stress or nutrient limitation (27). This stress response mechanism, termed the stringent response, allows cells to adjust to starvation conditions by downregulating expression of genes involved in the transcriptional apparatus and upregulating expression of genes required for amino acid biosynthesis. Amino acid starvation results in a large increase of deacylated tRNA, which readily binds to empty ribosomal A-sites. This triggers the RelA-dependent synthesis of (p)ppGpp, an alarmone that signals to the cell that it is in unfavorable growth conditions (28). The susceptibility of poxA mutants to harsh environmental conditions and differential protein expression in both E. coli and Salmonella suggests that PoxA and EF-P are involved in an adaptive stress response. It is possible that EF-P blocks the translational effects of ppGpp, allowing protein synthesis to continue from a particular subset of ribosomes to keep the cell alive. Conversely, EF-P may operate in a completely separate stress response or have no impact on ppGpp levels. Further studies are still necessary to determine the specific conditions under which EF-P is required and how this protein is involved in bacterial stress pathways, such as a potential role in the stringent response.

Eukaryotic Initiation Factor 5A

Unlike EF-P, the homologous proteins in both archaeal and eukaryotic cells, known as aIF5A and eIF5A, respectively are essential for cell viability (29). eIF5A was first characterized as an initiation factor due to its ability to stimulate the endpoint of methionyl-puromycin synthesis (30), and it has been noted that eIF5A is not a typical initiation factor, as it may promote the translation of only a subset of specific mRNAs

(31). Extensive work with Saccharomyces cerevisiae, including mutational analysis of

TIF51A and TIF51B, the genes encoding eIF5A, as well as polysome association profiles and pull-down assays showed that eIF5A functionally interacts with Elongation Factor 2

(eEF2) to increase translation elongation rates (30). A second more recent study with S. cerevisiae focused on narrowing down the translational stage in which eIF5A is most active. Through polysome profile analysis and by incorporating [35S]-methionine into a

“rapid depletion” yeast strain, authors concluded that eIF5A functions to stimulate the formation of the first peptide bond in global protein production (32).

In mammalian cells, eIF5A is found in both the nucleus and cytoplasm and may interact with the nuclear import system due to its role as a cofactor of the Rev transactivator protein of human immunodeficiency protein type 1 as well as the Rex protein of human T-cell leukemia virus type 1 (18). Fully active eIF5A is required for the passage of cells through the G1/S phase of the cell cycle (33).

As seen in Figures 7 and 8, eIF5A is composed of two β-barrel domains in the shape of a straight bar and Domains I and II of eIF5A correspond to domains I and II (or

II and III), respectively, of EF-P. EF-P shares 84% sequence similarity with aIF5A and

64% with eIF5A (33), and 42% of the T. thermophilus EF-P amino acids are conserved or semiconserved in the eIF5As (20). Residues Lys-29, Gly-31, Gly-33, and Ala-35 of T. thermophilus are completely conserved in all EF-Ps/eIF5As and are located on the loop connecting β3 and β4 in both protein structures (20).

Figure 7. Proteins aIF5a and eIF5A from M. jannaschii and L. mexicana, respectively.

Adapted from (33).

Figure 8. Structural comparison of EF-P and eIF5A. (A) Ribbon diagrams of T. thermophilus EF-P (blue) and M. jannaschii eIF5A (yellow), superimposed. (B)

Conservation of amino acid residues between EF-P and eIF5A. Adapted from (20).

Post-translational Modification of EF-P/eIF5A

Of greatest note, the proteins in the EF-P/eIF5A super family conserve a lysine residue at the tip of the loop connecting β3 and β4. To form fully active eIF5A, this lysine residue is post-translationally modified through the addition of the rare amino acid hypusine, which is only found in this unique modification pathway (34). The hypusination reaction takes place in two steps, shown in Figure 9. In the first step, the 4- aminobutyl moiety of the polyamine spermidine is transferred to the ε-amino group of the highly conserved lysine residue to form deoxyhypusine in a reaction catalyzed by deoxyhypusine synthase. The second step, catalyzed by deoxyhypusine hydroxylase, involves the hydroxylation of deoxyhypusine to form fully active eIF5A modified with a hypusine residue (29). Hypusinated eIF5A is required for mammalian cell proliferation

(35), and both type 1 and type 2 diabetes disease phenotypes have been correlated with conditions where eIF5A is not completely modified (36).

Figure 9. Post-translational modification pathways of eIF5A and EF-P. Adapted from

(12).

Although the enzymes deoxyhypusine synthase and deoxyhypusine hydroxylase are not found in the prokaryotic genome, a similar two-step post-translational modification pathway was recently elucidated for EF-P (12). To form fully functional

EF-P, the enzymes PoxA and YjeK post-translationally modify lysine 34, which corresponds to the site of hypusination in a/eIF5A (although some organisms encode a second copy of EF-P with an arginine residue in the place of the conserved lysine (18,

25)).

PoxA is a paralog of class II lysyl-tRNA synthetase (LysRS) and shares 32% identity and 50% similarity to the C-terminal domain of this enzyme (12). As mentioned previously, LysRS is the enzyme responsible for catalyzing the transfer of L-α-lysine to its cognate tRNA, and despite such homology, PoxA is incapable of transferring a lysine residue onto tRNALys. In the EF-P modification pathway, shown in Figure 10, PoxA catalyzes the addition of a lysine residue onto lysine 34 of EF-P in an ATP-dependent manner (12). YjeK, a 2,3-(β)-lysine aminomutase, is the second enzyme involved in this post-translational modification pathway and is responsible for converting L-lysine to (R)-

β-lysine. The current model of EF-P post-translational modification indicates that YjeK alters the lysine substrate either before or after its ligation to EF-P (12). Although much is already known about the EF-P modification pathway, the exact structure of fully modified EF-P is still uncertain.

Figure 10. Current model for the post-translational modification of EF-P. Reproduced from (12).

Studies with [14C]-lysine, ATP, and purified PoxA and EF-P proteins indicate that

PoxA is capable of charging EF-P with lysine (12). Shown in Figure 11, EF-P is rapidly labeled with [14C]-lysine in the presence of PoxA. Mass spectrometry results confirmed that EF-P is modified with a mass of 128 Da, consistent with the mass of lysine or β- lysine, as well as the previous mass reported for in vivo modified EF-P (12, 21).

Furthermore, PoxA was incapable of modifying the mutant protein EF-P K34A, again confirming that the conserved K34 residue is the location of EF-P post-translational modification.

Figure 11. PoxA-catalyzed in vitro lysylation of EF-P. Adapted from (12).

Kinetic analysis of PoxA activation of lysine revealed surprisingly high KM and low kcat values relative to E. coli LysRS, making it unlikely that lysine is the actual physiological substrate of PoxA. To gain a better understanding for the basis of this discrepancy, a comparison of the LysRS class II and PoxA active sites was performed.

Structural Analysis of the PoxA Active Site

A molecular docking model comparison of LysRS and PoxA active sites (12, Roy et al., in press), as shown in Figure 12, reveals that the residues and geometry of the two active sites are highly conserved, with the exception of Gly465 and Ala229 of

Geobacillus stearothermophilus LysRS, which correspond to Ala298 and Ser76 of

Salmonella Typhimurium PoxA. PoxA exhibits a wider active site cavity than its counterpart LysRS and the model indicates that both enzyme active sites are capable of accommodating a L-lysyl-adenylate, with the ε-NH2 group of the lysine positioned in a 20 conserved acidic pocket. The peptide GKG was modeled into the active site as a mimic of the conserved Lys34 of EF-P (12). EF-P is expected to occupy a position in the PoxA active site distinct from that of a 3’-hydroxyl group of a tRNA in a class II aaRS (12).

This indicates that while the L-lysyl-adenylate intermediate behaves similarly in both

PoxA and LysRS, EF-P and tRNA most likely bind in a different manner (12).

Figure 12. Molecular docking model comparison of LysRS and PoxA active sites. The

L-lysyl adenylate binds LysRS and PoxA in a similar manner. Crystal structures are from

G. stearothermophilus LysRS and Salmonella enterica Typhimurium PoxA. The residues differing between LysRS and PoxA are indicated in red. Reproduced from (12).

Through mutational analysis of 11 Salmonella PoxA active site amino acids, the following residues were found to be critical in vivo for PoxA activity: E78, E116, E244,

R100, E251, and R303 (12). The residues E78, E116, and E251 form an acidic pocket in 21 which the ε-NH2 group of lysine is positioned in both PoxA and LysRS. Residues R100,

E244, and R303 are conserved in both LysRS and PoxA and are necessary for substrate binding and aminoacyl-adenylate formation (12). Because these six residues are required for in vivo PoxA activity, it is likely that they play key roles in substrate binding and specificity.

Despite the substantial similarity between their active sites, PoxA is not capable of modifying EF-P with lysine as efficiently as LysRS. This suggests that the two non- conserved residues, G465 and A229 of LysRS and A298 and S76 of PoxA, determine the specificity for substrate recognition and binding, as well as optimal enzymatic activity, in

PoxA. To mimic the conserved residues of the LysRS active site, PoxA variants S76A and A294G (E. coli PoxA numbering) were created and tested for enzymatic activity by measuring kinetic parameters of ATP/PPi exchange reactions. While the S76A variant abolished activity, the A294G mutant decreased the KM for lysine by 25-fold (Roy et. al. in press). This led researchers to believe that the two non-conserved residues are highly important in not only the process of substrate selectivity but also in terms of optimizing enzymatic activity. Because PoxA is capable of activating lysine and the A294G mutant increased the KM for lysine, it is thought that the actual physiological substrate of PoxA is a compound very similar to lysine.

To further elucidate the physiological substrate for PoxA-catalyzed EF-P modification, enrichment studies were performed using cell-free extracts. Once metabolite extracts were prepared following solvent-based extraction protocols (37), TLC and liquid chromatography over a silica column were employed to fractionate the extract

14 compounds. By doping the resulting extracts with [ C]-lysine and testing for ATP/PPi exchange activity with both LysRS and PoxA, it was found that the substances producing optimal activity for these two enzymes did not co-fractionate, indicating different ideal substrates for LysRS and PoxA. Mass spectrometry was used to identify the mass of the modified residue at Lys34 of EF-P, and it was discovered that a moiety with a mass consistent with that of lysine is attached in this position (Roy et al, in press). This compound was thought to most likely be (R)-β-lysine, the product of Yjek’s activity in the lysine fermentation pathway. To test this hypothesis, kinetic parameters were measured for PoxA-catalyzed ATP/PPi exchange reactions with (R)-β-lysine and it was found that (R)-β-lysine is 100-fold more active than all other substrates tested (Roy et al, in press). This, together with the extensive enrichment studies, indicates that (R)-β-lysine is indeed the cognate substrate of PoxA-catalyzed EF-P modification.

Also shown through this work, (R)-β-lysine is a poor substrate for activation by

LysRS, as compared to α-lysine. Thus, PoxA adapted a separate method of substrate recognition than its paralog LysRS, indicating divergent evolution of the two enzymes from ancestral aminoacyl-tRNA synthetases. Also the modest, yet significant, difference in substrate usage indicates molecular evolution of α- vs. β-amino acid specificity. β- amino acids are less abundant than α-amino acids and are often by-products of metabolic pathways; they have also been isolated from meteorites (38). β-Amino acids are also extremely structurally diverse, as they can exist as R or S stereoisomers at either the alpha

(C2) or beta (C2) carbon. Because they are not encoded in natural proteins but are excellent mimics, β-amino acids are currently of interest in peptidomimetics and drug 23 discovery research (39). β-Amino acids are structural components of some naturally occurring antibiotics, such as subtulene A found in Bacillus subtilis (40). Also, β-peptides or α/β-peptides have been used in the development of foldamers with striking biological activities. β-Peptides and α/β-peptides are resistant to proteases, unlike α-peptides, which are rapidly degraded in vivo. The ability of β-peptides and α/β-peptides to mimic α- helical host-defense proteins has implicated these alternative peptides in the synthesis of antimicrobial agents and other biomedically relevant molecules (41).

The emerging significance of β-amino acids in drug discovery focuses interest on engineering aaRSs for use with unnatural amino acids. It is possible to site-specifically incorporate unnatural amino acids into proteins through the use of a carefully engineered aaRS capable of uniquely acylating “orthogonal” suppressor tRNA (42). The following study will focus on an active-site engineering approach to further determine the basis for

α- vs. β-amino acid specificity in enzymatic systems. Through site-directed mutagenesis, the roles of the two non-conserved residues between PoxA and LysRS will be further established. B. cereus LysRS was used in this work due to its active site similarity with

Geobacillus stearothermophilus LysRS, which is the most widely characterized active- site model. B. cereus LysRS and G. stearothermophilus LysRS have 77.9% sequence identify and the two organisms not only share the two highly conserved active site residues (G465 and A229 , G. stearothermophilus numbering) but also retain the 11 residues found to be critical for Salmonella PoxA activity in vivo (12). The creation and kinetic evaluation of the LysRS mutants A233S, G469A and A233S/G469A may potentially yield a shift in active site specificity from a α-amino acid to that of a β-amino 24 acid, thus mimicking PoxA in its substrate recognition. Gaining insight into the mechanism of the EF-P-PoxA-YjeK pathway, together with the structure of PoxA’s physiological substrate, will help define the function of EF-P by providing a link between the mutant phenotypes observed in Salmonella, the basis for the requirement of a β- amino acid, and what is currently known about classical protein translation systems. The knowledge of aaRS paralogs and the proteins they modify gained from this work will provide further insight into the roles of aaRSs outside of charging tRNA for protein synthesis. This research will also serve as a model for other virulence and stress response systems, and ultimately lead to a better grasp on novel antibiotic resistance mechanisms.

MATERIALS AND METHODS

Strain, plasmids and general methods

B. cereus lysS-encoded LysRS cloned into the pTYB1 vector was excised and ligated into the pQE31 vector to produce the wild type template for all subsequent experiments. Sets of two primers containing 33 nucleotides each were used to create the LysRS variants.

The following primers were used to create the two single mutants: A233S 5’

TTATATATGCGTATCTCTATTGAGCTACATTTA3’ and 5’ TTATATATGCGTATC

TCTATTGAGCTACATTTA 3’; G469A 5’ CCTACGGGTGGATTAGCAATCGGTAT

TGATCGT 3’ and 5’ CCTACGGGTGGATTAG CAATCGGTTATTGATCGT 3’. Using

Pfu Turbo DNA polymerase, site-directed mutagenesis PCR reactions were performed in accordance with Strategene protocol suggestions. Following the PCR protocol, products were digested with DpnI and transformed into XL1-Blue cells. Each point mutation was confirmed by DNA sequencing.

Substrates

A 100 mg/mL stock of α-lysine was made from Acros Organics L(+)-Lysine monohydrochloride 99+%. (R)-β-Lysine was provided by the laboratory of Dr. Craig

Forsyth Lab, Department of Chemistry, The Ohio State University. (S)-β-Lysine was a

26 gift from Professor Michael Thomas, Department of Bacteriology, University of

Wisconsin-Madison.

Lysyl-tRNA synthetase purification

The B. cereus lysS-encoded LysRS and mutants cloned into the pQE31 vector were expressed in E. coli XL1-Blue cells. Transformants were grown at 37 °C in 1 liter of LB broth supplemented with ampicillin (100 µg/mL) to cell density OD600 = 0.6. Expression of lysS was induced through the addition of 0.5 mM IPTG and cells were grown for an additional 4.5 h. Cells were harvested at 4 °C through centrifugation at 5000 xg for 5 min. Pelleted cells were washed with 0.1 M Tris-HCl pH 8 and stored at -80°C until purification.

Following a His-tagged protein purification protocol, 1 L of pelleted cells were first resuspended in 10 mL Buffer 1 [25 mM Tris-HCl (pH 8), 300 mM NaCl, 10% (v/v) glycerol and 5 mM imidazole] supplemented with protease inhibitor cocktail (Roche

Diagnostics). Cells were disrupted by passing through a French Pressure cell (pass 3 times at 1000 psig) and cell debris was pelleted by centrifigation at 75,000 xg for 45 minutes at 4 °C. Supernatant was then loaded onto a column composed of 3 mL of Talon metal affinity resin (BD Bioscience) at 4 °C, pre-equilibrated with Buffer 1. The column was then washed with 250 mL Buffer 1 at 4 °C. Ten column fractions of 1.5 mL were eluted with 25 mM Tris-HCl pH 8, 300 mM NaCl, 10 % (v/v) glycerol and 250 mM imidazole at room temperature. Aliquots of each fraction were analyzed by SDS-PAGE and fractions 1-4 were pooled and dialyzed overnight in 50 mM Tris-HCl pH 8, 1 mM

MgCl2, 10 mM β-mercaptoethanol, and 10 % (v/v) glycerol at 4°C. Proteins were

27 redialyzed for 4 hours at 4 °C against storage buffer [50 mM Tris-HCl (pH 8), 1 mM

MgCl2, 10 mM β-mercaptoethanol, and 50 % (v/v) glycerol] and stored at -20 °C until use.

Active-site titration

Active-site titration was performed in 100 mM Hepes (pH 7.2), 30 mM KCl, 10 mM

14 MgCl2, 4 mM ATP, 50 µM [ C]-lysine, and 1 unit of inorganic pyrophosphatase.

Following preincubation at 37 °C for 1-2 min, the reaction was initiated through addition of 1 µM LysRS to the reaction mixture. After a 10 min reaction at 37 °C, samples were placed on ice to terminate the reaction. The entire 50 µL reaction volume was then spotted onto nitrocellulose filters presoaked in 50 mM Hepes (pH 7.2), 15 mM KCl, and

5 mM MgCl2. Filters were washed with 5 mL buffer containing 50 mM Hepes (pH 7.2),

15 mM KCl, and 5 mM MgCl2 and dried for 25 minutes at 80 °C. The amount of complex retained was quantified through scintillation counting (Ultima Gold, Packard

Instrument Co.).

ATP-PPi Exchange

The ATP-PPi exchange reaction was carried out in 100 mM Hepes (pH 7.2), 30 mM KCl,

32 15 mM MgCl2, 10 mM NaF, 3 mM ATP, 5 mM DTT, 3 mM [ P]PPi, 37 nM-4 µM

LysRS, and a range of lysine concentrations from 42.3 µM-50 mM for α-lysine or 7 mM

(S)-β-lysine or (R)-β-lysine. The reaction mixture was preincubated for 2 minutes and the reaction was initiated with the addition of amino acid. Aliquots of 10 -15 µL were quenched by addition of quenching solution (5.6 % percholric acid, 1 % Norit A, and 75

28 mM PPi) at different timepoints. ATP absorbed on charcoal was filtered through 3MM

Whatman GF/C filter disks, washed three times with 5 mL of water and once with 5 mL

100% ethanol, and dried at 80 °C for 20 minutes. The amount of [32P]ATP formed was quantified through scintillation counting (Ultima Gold, Packard Instrument Co.) and KM and kcat values were determined for each protein.

Aminoacylation

Aminoacylation was performed at 37 °C in 100 mM Hepes (pH 7.2), 30 mM KCl, 10

14 mM MgCl2, 2 mM ATP, 50 µM [ C]-lysine, 10 mg/mL E. coli total tRNA, and 500 nM

LysRS. At 2, 4, 8, and 12 min, 9 µL aliquots were quenched on 3 MM Whatman GF/C filter disk presoaked in 5 % TCA (w/v). Following the final timepoint, filter disks were washed 3 times for 5 min at room temperature in 5 % (v/v) TCA solution, and rinsed in

100 % ethanol prior to baking at 80 ° C for 20 min. The level of [14C]-lysyl-tRNALys was then quantified through scintillation counting (Ultima Gold, Packard Instrument Co.).

RESULTS

Site-directed mutagenesis of LysRS active site

Based on structural comparison of the LysRS and PoxA active sites, the two non- conserved residues, G465 and Ala229 of G. stearothermophilus LysRS and A298 and

S76 of S. enterica Typhimurium PoxA, were chosen for close examination of their ability to differentiate between α- and β-amino acids. All studies were performed with B. cereus

LysRS, as its active site most closely resembles that of G. stearothermophilus, with which previous modeling work was performed (12). Through site-directed mutagenesis protocols, as described in Materials and Methods, the LysRS single mutants A233S,

G469A, and the double mutant A233S/G469A were created (B. cereus LysRS numbering, homologous to Gly465 and Ala229 of Geobacillus stearothermophilus). The replacements were intended to make the active site of LysRS more similar to PoxA, and to potentially observe a shift in the substrate and stereospecificity of the LysRS active site from α-lysine to (R)-β-lysine. Sequencing reactions confirmed the correct mutations were made successfully and wild type protein along with the three variants were expressed and purified as discussed in Materials and Methods.

Bradford assays along with active-site titration experiments were performed to estimate the concentration of each protein (Table 1). Active-site titration reactions provide a more accurate protein concentration based on the percentage of active enzyme,

30 and therefore, results from active-site titration experiments were used to calculate reaction volumes for wild type and A233S proteins for all subsequent experiments.

However, G469A and A322S/G469A proteins did not show activity in active-site titration assays presumably due to unstable binding of [14C]-lysyl-adenylates. The concentrations determined through Bradford assays were used for G469A and A233S/G469A in all further calculations.

LysRS Enzyme Conc. (Bradford Assay, µM) Conc. (Active-site Titration, µM) Wild Type 96.5 36.3 A233S 67.5 14.7 G469A 93.5 ND a A233S/G469A 106.5 ND a

Table 1. Concentrations of purified proteins, as determined through Bradford and active- site titration assays. a Concentration from active-site titration could not be determined as the lysyl-adenylate is unstable in the G469A and A233S/G469A active sites.

Determination of LysRS variant activity through aminoacylation

Aminoacylation reactions were performed with wild type and mutant proteins to qualitatively evaluate changes in LysRS catalytic activity due to active-site residue replacement. Wild type, A233S, G469A, and A233S/G469A proteins were evaluated for their ability to charge E. coli total tRNA with [14C]-lysine, as described in Materials and

Methods. A representative charging curve can be seen in Figure 13.

5.00 4.50

4.00 3.50 3.00 Wild Type 2.50 2.00 A233S 1.50 G469A [14C] Lysine (µM) [14C] Lysine 1.00 A233S/G469A 0.50 0.00 0 5 10 15 Time (min)

Figure 13. Aminoacylation of LysRS wild type and variants. Data represent the average of three independent experiments.

Discrimination of lysine through ATP/PPi Exchange

The effect of each replacement on the active site of LysRS was determined through the ATP/PPi exchange reaction, with α-lysine as a substrate. By measuring the kinetic parameters of this reaction, it was possible to determine the effect of each mutation on the ability of the LysRS active site to correctly form a α-lysyl-adenylate.

Representative Michaelis-Menten curves for wild type, A233S, and G469A can be seen in Figure 14 and KM, kcat, and kcat/KM values for each protein can be found in Table 2.

Kinetic parameters could not be determined for A233S/G469A, as the lysyl-adenylate is presumed to be unstable in the altered active site.

Figure 14. Representative Michaelis-Menten curves for wild type (A), G469 (B), and

A233S (C). ATP/PPi reactions were performed with α-lysine.

-1 -1 KM (µM) kcat (min ) kcat/KM (min /µM) Wild Type 162 ± 71 2540± 325 16 A233S 210 ± 63 5091 ± 492 25 G469A 3922 ± 1201 148 ± 10 a 3.8 x 10-2 a A233S/G469A NDb

Table 2. Steady-state kinetics of lysine activation by B. cereus LysRS wild type and mutant proteins. Errors correspond to the standard deviation of three independent experiments. a Estimate based on concentration determined from Bradford Assay (187

µM). b Kinetic parameters for the LysRS variant A233S/G469A could not be determined through ATP/PPi exchange due to protein/adenylate instability.

Comparison of the catalytic efficiency (kcat/KM ) of A233S and G469A with wild type LysRS indicated that the active site replacements decreased the efficiency of the variants. The KM for α-lysine increased for both A233S (210 µM) and G469A (3922 µM), with G469A no longer active in the assay under physiological conditions. The A233S variant displayed a 2-fold increase in kcat, while the kcat value for G469A decreased 17- fold, as compared to wild type. As mentioned previously, kinetic parameters for the double mutant A233S/G469A could not be confidently determined due to variable results with ATP/PPi exchange. Overall, ATP/PPi exchange with α-lysine indicates that the

G469A and A233S/G469A replacements interfere with α-lysine binding and catalysis, while the A233S replacement is not detrimental to LysRS catalytic activity.

To explore the effects of active-site replacements on the ability of LysRS to activate noncognate forms of lysine, ATP/PPi exchange reactions were performed with

(S)-β-lysine and (R)-β-lysine. Both stereoisomers of β-lysine were used to determine whether creation of the LysRS variants shifted the specificity of the LysRS active site to more closely resemble that of PoxA. While (R)-β-lysine is believed to be the physiological substrate of PoxA, the ability of LysRS to activate (S)-β-lysine was also tested, as it was possible that mutation altered the substrate specificity but not the stereospecificity of substrate recognition by LysRS. Due to low activity with both forms of β-lysine, it was not possible to determine steady-state kinetic parameters; however, some activity was detected for wild type and the A233S variant. Charging curves with both (S)-β-lysine and (R)-β-lysine for these proteins can be seen in Figure 14.

Wild Type 400 350 300

250 200

150 S-B-Lysine [ATP] (µM) [ATP] 100 R-B-Lysine 50 0 0 5 10 15 20 25

Time (min)

A233S 600 550 500 450 400 350 300 250 S-B-Lysine

[ATP] (µM) [ATP] 200 150 R-B-Lysine 100 50 0 0 5 10 15 20 25

Time (min)

Figure 15. Amino acid specificity of LysRS wild type and A233S with (S)-β-lysine and

(R)-β-lysine. Formation of ATP was monitored in the presence of 7 mM β-lysine. A

LysRS concentration of 4 µM was used for each experiment. Curves represent average values from three independent experiments.

As seen in Figure 15, wild type and the A233S variant proteins exhibit some activity with both (S)-β-lysine and (R)-β-lysine, with both proteins stimulating ATP/PPi exchange activity to a greater extent with (S)-β-lysine. Although activity with (R)-β- lysine was comparable for wild type and the A233S mutant, stimulation of lysyl- adenylate formation with (S)-β-lysine was significantly greater for A233S. This indicates that the A233S replacement shifts the substrate specificity of the LysRS active site to better accommodate β-lysine. Because neither protein showed substantial activity with

(R)-β-lysine, it is most likely that the replacement was not capable of altering the stereospecificity of the LysRS active site. Neither the G469A nor the A233S/G469A replacements showed stimulation of the ATP/PPi exchange reaction with β-lysine.

Wild type and A233S were able to successfully complete the aminoacylation reaction over a course of 12 min. Variants G469A and A233S/G469A showed substantially less activity and were incapable of charging E. coli total tRNA with [14C]- lysine in a similar time frame. Of interest, A233S showed a substantially higher initial velocity than wild type. This finding was also true of the 2-fold higher kcat value observed for A233S, and could be due to a slight variation in activity from different protein preparations. Overall, the results of aminoacylation experiments were consistent with data from active-site titration and ATP/PPi exchange experiments performed with the four LysRS proteins. Taken together, these data indicates that while the A233S mutation is not detrimental to LysRS activity, the G469A and A233S/G469A replacements significantly affect LysRS in its ability to efficiently form the lysyl-adenylate and charge tRNA with its cognate amino acid substrate.

DISCUSSION

Recent analysis of PoxA activity in ATP/PPi exchange reactions and enrichment assays concluded that the physiological substrate for PoxA-mediated EF-P post- translational modification is β-lysine (Roy et al., in press). Also, PoxA active-site mutagenesis revealed that replacement of Ala294 with glycine to mimic the conserved

LysRS active-site residue increased PoxA’s specificity for lysine. This suggests that

PoxA must have adapted a separate form of substrate recognition from class II LysRS, the enzyme from which it is derived, and indicates a role for unnatural amino acids in translational pathways. To further investigate the divergent evolution of PoxA, as well as the specificity for a β- vs. α-amino acid, a series of active-site residue replacements were made in B. cereus LysRS. Through the creation of the variants A233S, G469A and

A233S/G469A and evaluation in ATP/PPi exchange and aminoacylation reactions, the effect of the active-site mutations on the two non-conserved residues between LysRS and

PoxA could be seen.

Interpretations of the effect of the mutations on kinetic parameters indicate that alteration of the LysRS active site decreases the ability of LysRS to stimulate lysyl- adenylate formation during ATP/PPi exchange. The A233S replacement slightly increases the KM value for α-lysine while increasing the kcat value. A233S was also fully capable of charging tRNA in aminoacylation assays, further illustrating the absence of an effect this

38 mutation has on LysRS enzymatic activity. The G469A replacement was capable of performing the ATP/PPi exchange reaction; however, the concentration of enzyme and substrate required to achieve Michaelis-Menten kinetics was not physiologically relevant.

Both the binding constant and enzymatic turnover were drastically decreased in ATP/PPi exchange reactions and G469A presented significantly less activity during aminoacylation reactions. This suggests that G469A is the more detrimental mutation and therefore may be the more significant of the two non-conserved residues in determining activity. The combined A233S/G469A double mutant was incapable of forming a stable lysyl-adenylate during ATP/PPi exchange and could not charge E. coli total tRNA with

[14C]-lysine during aminoacylation. Therefore, these results suggest that these replacements abolish the ability of the LysRS active site to recognize and activate lysine, as seen previously with the PoxA S76A replacement (Roy et al., in press).

Through investigating whether the active site replacements increased the ability of

LysRS to accommodate a β-amino acid into its active site, it was shown that altering the two non-conserved residues did not increase the selectivity for a β-amino acid over a α- amino acid. Of the three variants, A233S was the only one to increase the ability of

LysRS to activate (S)-β-lysine, suggesting that the A233 residue could play a larger role in substrate selectivity than G469. Stimulation of the ATP/PPi exchange reaction was greater with (S)-β-lysine for A233S as compared to wild type, indicating that the alanine to serine replacement aids the enzyme in selecting for this form of β-lysine. This result is similar to the previously studied PoxA A294G replacement, in that a single residue replacement allowed the active site to better accommodate a different substrate. Because

39 activity with (R)-β-lysine was comparable between wild type and A233S, the substitution of serine in this position was not enough to alter the stereospecificity of the LysRS active site. Neither G469A nor A233S/G469A were capable of catalyzing lysyl-adenylate formation with either enantiomer of β-lysine, indicating that these LysRS replacements are unstable enzymes and are physiologically irrelevant.

From the site-directed mutagenesis results seen with the B. cereus LysRS active site, it can be concluded that replacement of residues A233 and G469 with those found in the conserved PoxA positions is not substantial enough to shift the specificity of the

LysRS active site from a α-amino acid to a β-amino acid. Overall, the A233S variant was found to be more active but less specific than the other variants, due to its ability to interact with (S)-β-lysine and (R)-β-lysine. With the slight change in substrate specificity

(but not stereospecificity) seen with the A233S variant, further active-site engineering efforts are needed to fully convert the preference of the LysRS active site to a β-amino acid.

OUTLOOK

Key advances have recently been made to determine the structure and function of

EF-P in its interaction with modifying enzymes PoxA and YjeK. With the structure of fully modified EF-P now known, the focus can shift to understanding the means by which

PoxA has adapted substrate specificities and functions distinct from that of the well- known members of the aaRS super family. The work presented in this thesis will contribute to defining the roles of critical residues of the PoxA active site in substrate recognition, as well as pose potential explanations as to how PoxA has evolved the ability to interact with EF-P and β-lysine and not tRNALys or α-lysine.

A logical next step to gain insight into the interworkings of the EF-P-PoxA-YjeK modification pathway is to further investigate the divergent evolution of substrate recognition for PoxA. To define the basis for PoxA’s substrate recognition, crystallographic studies will be carried out with several forms of PoxA and varying substrates. The crystal structure of PoxA with β-lysine will be determined, as well as

PoxA A294G (the variant found to increase specificity for α-lysine) with β-lysine, to evaluate the active-site differences between LysRS and PoxA that would alter specificity for α-lysine. Results of crystallographic studies will then be used to perform kinetic analyses of PoxA variants to further define the molecular basis for β-lysine specificity.

Additionally, to evaluate how EF-P recognition of PoxA mimics that of tRNALys and LysRS, mutational analysis of the residues along the EF-P:PoxA interface will be performed. Through a comparison with the tRNAAsp:AspRS system, and aminoacylation experiments, it will be possible to quantify the contribution of each residue to elucidate the interactions required for β-lysine versus α-lysine recognition. These studies will help to explain how new interactions within the active sites of aaRS paralogs have developed to exhibit preferential binding of substrates other than tRNA.

Another means of understanding the evolution of EF-P as a tRNA mimic is to investigate the role of its homologue, EF-P2. EF-P2 also exists in some bacterial organisms, either in addition to EF-P or in place of it. Because EF-P2 contains an arginine in place of a lysine at residue 34, the site of EF-P post-translational modification, it is possible that EF-P2 acts in a uniquely modified form or even as an unmodified protein. Based on these differences, it will be beneficial to gain information on the functional state of EF-P2, as well as to investigate the potential role of EF-P2 in translational pathways. This can be accomplished through testing for phenotypes produced in conditions similar to those evaluated with EF-P (such as hypoosmolarity and antibiotic stress), as well as to observe effects of EF-P2 on virulence in mouse models of infection. This information will aid researchers in evaluating the role of EF-P/EF-P2 in organisms lacking the PoxA/YjeK pathway.

As a second means of assessing the role of EF-P in translation systems, future work will focus on determining the action of modified EF-P on the ribosome. Although previous reports have detailed the structure of EF-P bound to the 70S ribosome (22), this

42 work was limited in that it did not include information on how modified EF-P associates with the different ribosomal sites. By judging the affinity of β-lysine-EF-P for the ribosome, it will be possible to detect a mode of ribosomal binding for β-lysine-EF-P. By comparing how modified and unmodified EF-P form initiation complexes with other ribosomal proteins, as well as determining what components are necessary for optimal

EF-P activity, it will be possible to gain insight into the role of EF-P in peptide bond formation.

Based on findings from this proposed work, researchers can then work to further define EF-P in its role in translation initiation and/or elongation. It is still unclear as to whether the homologous protein in eukaryotic systems, eIF5A, operates to enhance formation of the first peptide bond or aids in the incorporation of amino acids during translation elongation. Likewise, the role for EF-P is equally unknown. Measuring polysome/monosome ratios as well as monitoring 70S initiation complex formation in the presence of modified and unmodified EF-P will pinpoint the effect of EF-P on initiation.

Additionally, the role of EF-P in elongation will be studied by employing poly(Phe) synthesis assays. Furthermore, polysome profiling and in vivo reporter assays will show if the EF-P-PoxA-YjeK pathway favors processing of a subset of particular mRNA elements as opposed to acting on global protein synthesis. Overall, future work will be directed to provide knowledge of the possible mechanism by which EF-P influences translation, and largely contribute to a better understanding of EF-P across many species of bacteria. Ultimately we will gain a greater grasp on roles for aaRSs outside of

43 aminoacylating tRNA for protein synthesis, antibiotic resistance and virulence models, as well as post-translational gene regulation in bacteria.

REFERENCES

1. Alberts B JA, Lewis J, et al. (2002) Molecular Biology of the Cell (New York: Garland Science) 4 Ed. 2. Ibba M & Soll D (2004) Aminoacyl-tRNAs: setting the limits of the genetic code. Genes Dev 18(7):731-738. 3. Ling J, Reynolds N, & Ibba M (2009) Aminoacyl-tRNA synthesis and translational quality control. Annu Rev Microbiol 63:61-78. 4. Ataide SF & Ibba M (2004) Discrimination of cognate and noncognate substrates at the active site of class II lysyl-tRNA synthetase. Biochemistry 43(37):11836- 11841. 5. Eriani G, Delarue M, Poch O, Gangloff J, & Moras D (1990) Partition of tRNA synthetases into two classes based on mutually exclusive sets of sequence motifs. Nature 347(6289):203-206. 6. Ibba M & Soll D (2001) The renaissance of aminoacyl-tRNA synthesis. EMBO Rep 2(5):382-387. 7. Cusack S, Berthet-Colominas C, Hartlein M, Nassar N, & Leberman R (1990) A second class of synthetase structure revealed by X-ray analysis of Escherichia coli seryl-tRNA synthetase at 2.5 A. Nature 347(6290):249-255. 8. Ibba M, et al. (1999) Substrate recognition by class I lysyl-tRNA synthetases: a molecular basis for gene displacement. Proc Natl Acad Sci U S A 96(2):418-423. 9. Ling J, et al. (2009) Resampling and editing of mischarged tRNA prior to translation elongation. (Translated from eng) Mol Cell 33(5):654-660 (in eng). 10. Banerjee R, et al. (2010) tRNAs: cellular barcodes for amino acids. FEBS Lett 584(2):387-395. 11. Vetting MW, Hegde SS, & Blanchard JS (2010) The structure and mechanism of the Mycobacterium tuberculosis cyclodityrosine synthetase. Nat Chem Biol 6(11):797-799. 12. Navarre WW, et al. (2010) PoxA, yjeK, and elongation factor P coordinately modulate virulence and drug resistance in Salmonella enterica. Mol Cell 39(2):209-221. 13. Steitz TA (2008) A structural understanding of the dynamic ribosome machine. Nat Rev Mol Cell Biol 9(3):242-253. 14. Korostelev A & Noller HF (2007) The ribosome in focus: new structures bring new insights. Trends Biochem Sci 32(9):434-441. 15. Shoji S, Walker SE, & Fredrick K (2009) Ribosomal translocation: one step closer to the molecular mechanism. ACS Chem Biol 4(2):93-107.

16. Pingoud A, Urbanke C, Krauss G, Peters F, & Maass G (1977) Ternary complex formation between elongation factor Tu, GTP and aminoacyl-tRNA: an equilibrium study. Eur J Biochem 78(2):403-409. 17. Glick BR & Ganoza MC (1975) Identification of a soluble protein that stimulates peptide bond synthesis. Proc Natl Acad Sci U S A 72(11):4257-4260. 18. Ganoza MC, Kiel MC, & Aoki H (2002) Evolutionary conservation of reactions in translation. Microbiol Mol Biol Rev 66(3):460-485, table of contents. 19. Glick BR, Chladek S, & Ganoza MC (1979) Peptide bond formation stimulated by protein synthesis factor EF-P depends on the aminoacyl moiety of the acceptor. Eur J Biochem 97(1):23-28. 20. Hanawa-Suetsugu K, et al. (2004) Crystal structure of elongation factor P from Thermus thermophilus HB8. Proc Natl Acad Sci U S A 101(26):9595-9600. 21. Aoki H, et al. (2008) Interactions of elongation factor EF-P with the Escherichia coli ribosome. FEBS J 275(4):671-681. 22. Blaha G, Stanley RE, & Steitz TA (2009) Formation of the first peptide bond: the structure of EF-P bound to the 70S ribosome. Science 325(5943):966-970. 23. Aoki H, Dekany K, Adams SL, & Ganoza MC (1997) The gene encoding the elongation factor P protein is essential for viability and is required for protein synthesis. J Biol Chem 272(51):32254-32259. 24. Baba T, et al. (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2:2006 0008. 25. Bailly M & de Crecy-Lagard V (2010) Predicting the pathway involved in post- translational modification of elongation factor P in a subset of bacterial species. Biol Direct 5:3. 26. Bearson SM, Bearson BL, Brunelle BW, Sharma VK, & Lee IS (2011) A Mutation in the poxA Gene of Salmonella enterica Serovar Typhimurium Alters Protein Production, Elevates Susceptibility to Environmental Challenges, and Decreases Swine Colonization. Foodborne Pathog Dis. 27. Cashel M & Gallant J (1969) Two compounds implicated in the function of the RC gene of Escherichia coli. Nature 221(5183):838-841. 28. Wendrich TM, Blaha G, Wilson DN, Marahiel MA, & Nierhaus KH (2002) Dissection of the mechanism for the stringent factor RelA. Mol Cell 10(4):779- 788. 29. Park MH (2006) The post-translational synthesis of a polyamine-derived amino acid, hypusine, in the eukaryotic translation initiation factor 5A (eIF5A). J Biochem 139(2):161-169. 30. Saini P, Eyler DE, Green R, & Dever TE (2009) Hypusine-containing protein eIF5A promotes translation elongation. Nature 459(7243):118-121. 31. Park MH, Joe, Y. A., Kang, K. R., Lee, Y.B., and Wolff, E. C. (1996) The polyamine-derived amino acid hypusine: its post-translational formation in eIF- 5A and its role in cell proliferation. Amino Acids 10(2):109-121. 32. Henderson A & Hershey JW (2011) Eukaryotic translation initiation factor (eIF) 5A stimulates protein synthesis in Saccharomyces cerevisiae. (Translated from eng) Proc Natl Acad Sci U S A 108(16):6415-6419 (in eng). 46

33. Gentz PM, Blatch GL, & Dorrington RA (2009) Dimerization of the yeast eukaryotic translation initiation factor 5A requires hypusine and is RNA dependent. FEBS J 276(3):695-706. 34. Park MH, Wolff EC, & Folk JE (1993) Is hypusine essential for eukaryotic cell proliferation? Trends Biochem Sci 18(12):475-479. 35. Park MH, Nishimura K, Zanelli CF, & Valentini SR (2010) Functional significance of eIF5A and its hypusine modification in eukaryotes. Amino Acids 38(2):491-500. 36. Maier B, et al. (2010) The unique hypusine modification of eIF5A promotes islet beta cell inflammation and dysfunction in mice. J Clin Invest 120(6):2156-2170. 37. Rabinowitz JD & Kimball E (2007) Acidic acetonitrile for cellular metabolome extraction from Escherichia coli. Anal Chem 79(16):6167-6173. 38. Daniels DS, Petersson EJ, Qiu JX, & Schepartz A (2007) High-resolution structure of a beta-peptide bundle. J Am Chem Soc 129(6):1532-1533. 39. Steer DL, Lew RA, Perlmutter P, Smith AI, & Aguilar MI (2002) Beta-amino acids: versatile peptidomimetics. Curr Med Chem 9(8):811-822. 40. Thasana N, et al. (2010) Bacillus subtilis SSE4 produces subtulene A, a new lipopeptide antibiotic possessing an unusual C15 unsaturated beta-amino acid. FEBS Lett 584(14):3209-3214. 41. Horne WS & Gellman SH (2008) Foldamers with heterogeneous backbones. Acc Chem Res 41(10):1399-1408. 42. Liu DR, Magliery TJ, Pastrnak M, & Schultz PG (1997) Engineering a tRNA and aminoacyl-tRNA synthetase for the site-specific incorporation of unnatural amino acids into proteins in vivo. Proc Natl Acad Sci U S A 94(19):10092-10097.