Vrije Universiteit Brussel

Leishmania donovani tyrosyl-tRNA synthetase structure in complex with a tyrosyl adenylateanalog and comparisons with human and protozoan counterpartsBarros-Álvarez, Ximena; Kerchner, Keshia M; Koh, Cho Yeow; Turley, Stewart; Pardon, Els;Steyaert, Jan; Ranade, Ranae M; Gillespie, J Robert; Zhang, Zhongsheng; Verlinde,Christophe L M J; Fan, Erkang; Buckner, Frederick S; Hol, Wim G JPublished in:Biochimie

DOI:10.1016/j.biochi.2017.04.006

Publication date:2017

Document Version:Final published version

Link to publication

Citation for published version (APA):Barros-Álvarez, X., Kerchner, K. M., Koh, C. Y., Turley, S., Pardon, E., Steyaert, J., ... Hol, W. G. J. (2017).Leishmania donovani tyrosyl-tRNA synthetase structure in complex with a tyrosyl adenylate analog andcomparisons with human and protozoan counterparts. Biochimie, 138, 124-136.https://doi.org/10.1016/j.biochi.2017.04.006

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portalTake down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Download date: 22. Apr. 2021

lable at ScienceDirect

Biochimie 138 (2017) 124e136

Contents lists avai

Biochimie

journal homepage: www.elsevier .com/locate/b iochi

Research paper

Leishmania donovani tyrosyl-tRNA synthetase structure in complexwith a tyrosyl adenylate analog and comparisons with human andprotozoan counterparts

Ximena Barros-�Alvarez a, b, Keshia M. Kerchner a, Cho Yeow Koh a, 1, Stewart Turley a,Els Pardon c, d, Jan Steyaert c, d, Ranae M. Ranade e, J. Robert Gillespie e,Zhongsheng Zhang a, Christophe L.M.J. Verlinde a, Erkang Fan a, Frederick S. Buckner e,Wim G.J. Hol a, *

a Department of Biochemistry, University of Washington, Seattle, WA, USAb Laboratorio de Enzimología de Par�asitos, Facultad de Ciencias, Universidad de los Andes, M�erida, Venezuelac Structural Biology Brussels, Vrije Universiteit Brussel, Brussel, Belgiumd VIB-VUB Center for Structural Biology, VIB, Brussels, Belgiume Division of Allergy and Infectious Diseases, School of Medicine, University of Washington, Seattle, WA, USA

a r t i c l e i n f o

Article history:Received 26 February 2017Accepted 12 April 2017Available online 18 April 2017

Keywords:Aminoacyl-tRNA synthetaseTyrosyl adenylate analogNanobodyLeishmaniaTrypanosomatidsNeglected tropical diseases

* Corresponding author.E-mail address: wghol@u.washington.edu (W.G.J.

1 Current address: Department of Biological ScieSingapore, Singapore 117543, Singapore.

http://dx.doi.org/10.1016/j.biochi.2017.04.0060300-9084/© 2017 Elsevier B.V. and Société Française

a b s t r a c t

The crystal structure of Leishmania donovani tyrosyl-tRNA synthetase (LdTyrRS) in complex with ananobody and the tyrosyl adenylate analog TyrSA was determined at 2.75 Å resolution. Nanobodies arethe variable domains of camelid heavy chain-only antibodies. The nanobody makes numerous crystalcontacts and in addition reduces the flexibility of a loop of LdTyrRS. TyrSA is engaged in many in-teractions with active site residues occupying the tyrosine and adenine binding pockets. The LdTyrRSpolypeptide chain consists of two pseudo-monomers, each consisting of two domains. Comparing thetwo independent chains in the asymmetric unit reveals that the two pseudo-monomers of LdTyrRS canbend with respect to each other essentially as rigid bodies. This flexibility might be useful in the posi-tioning of tRNA for catalysis since both pseudo-monomers in the LdTyrRS chain are needed for chargingtRNATyr.

An “extra pocket” (EP) appears to be present near the adenine binding region of LdTyrRS. Since thispocket is absent in the two human homologous enzymes, the EP provides interesting opportunities forobtaining selective drugs for treating infections caused by L. donovani, a unicellular parasite causingvisceral leishmaniasis, or kala azar, which claims 20,000 to 30,000 deaths per year. Sequence andstructural comparisons indicate that the EP is a characteristic which also occurs in the active site ofseveral other important pathogenic protozoa. Therefore, the structure of LdTyrRS could inspire the designof compounds useful for treating several different parasitic diseases.

© 2017 Elsevier B.V. and Société Française de Biochimie et Biologie Moléculaire (SFBBM). All rightsreserved.

1. Introduction

The leishmaniases are a variety of diseases caused by more than20 Leishmania species. These protozoans are transmitted throughthe bites of infected female phlebotomine sandflies. Approximately

Hol).nces, National University of

de Biochimie et Biologie Molécul

350 million people in the tropics and sub-tropics are at risk ofinfection, with 0.9e1.3 million new cases and 20,000 to 30,000deaths annually [1,2]. Depending on the species involved, cuta-neous, mucocutaneous or visceral leishmaniasis can develop, thelatter being the most serious as far as number of deaths is con-cerned. Caused by L. donovani and L. infantum, visceral leishmani-asis (VL), or kala-azar, is fatal if left untreated in 95% of the cases. VLis characterized by irregular bouts of fever, weight loss, enlarge-ment of the spleen and liver, and anemia. About 200,000 to400,000 new cases of VL occur each year and children are the most

aire (SFBBM). All rights reserved.

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136 125

severely affected group [2,3]. Available drugs used for the treat-ment of the leishmaniases are mainly pentavalent antimonialcomplexes, amphotericin B, the aminoglycoside paromomycin, andthe alkylphosphocholine miltefosine, either as single or combina-tion treatments [4,5]. Most of the available drugs are not oral and,due to their complicated administration regimens, low efficacy forparasite elimination, poor safety profile, occurrence of drug resis-tance [4] and treatment failure, there is an urgent need to take neworal drug candidates into clinical development [1].

Protein translation is an essential cellular function in Trypano-somatid parasites. The overall fidelity of protein synthesis relies onthe accuracy of both codon-anticodon recognition and aminoacyl-tRNA synthesis [6]. Aminoacyl-tRNA synthetases (aaRSs) play acrucial role in accurately pairing each amino acid with its cognatetRNA through a two-step esterification reaction formingaminoacyl-tRNA [7]. The first, ATP dependent, step leads to theformation of an enzyme-bound aminoacyl-adenylate and an inor-ganic pyrophosphate leaving group. The second step results in the3’-esterification of the tRNA with the amino acid moiety and gen-eration of AMP as a leaving group, followed by product release [6].

Human cells have two sets of aaRS, one for cytosolic aminoacyl-tRNA synthesis and a second one, of bacterial evolutionary origin,for aminoacylation of mitochondrial tRNAs. In contrast, Trypano-somatids contain only a single set of aaRS genes, with the exceptionof three amino acids (aspartate, lysine and tryptophan) for whichtwo aaRS genes exist [8]. Hence, in these parasites 17 aaRSs have tofunction in both subcellular locations. This is a remarkable differ-ence between human and parasite aaRS. It suggests that inhibitingany of these aaRS in parasites would have the effect of hinderingprotein synthesis in the cytosol as well as in the mitochondria.

AaRSs have been recognized as validated antimicrobial drugtargets [9e11]. Anti-aaRS compounds can act by blocking thebinding site of ATP and/or amino acid, an allosteric site, tRNArecognition, or a secondary editing site [12]. Successes in targetingpathogenic aaRS are encouraging. At least two drugs, mupirocinand tavaborole, are in clinical topical use against, respectively,Staphylococcus aureus infections and onychomycosis. Mupirocinbinds both the ATP and Ile binding sites of S. aureus isoleucyl-tRNAsynthetase (IleRS), while tavaborole (Kerydin™) inhibits fungalleucyl-tRNA (LeuRS) by binding to its editing site [10,13,14]. Com-pounds targeting parasitic protozoan aaRS have been shown tointerfere with protozoan cell growth [10,15]. Also, several naturalproducts inhibit aaRS from important tropical parasites. Forinstance, Plasmodium falciparum prolyl-tRNA synthetase, lysyl-tRNA synthetase and threonyl-tRNA synthetase are inhibited bythe natural products halofuginone, cladosporin and borrelidin,respectively [12,16,17]. These compounds have also inhibitory ef-fects on malaria parasite growth in vitro or in vivo [18]. Here wefocus on tyrosyl-tRNA synthetase (TyrRS) from L donovani.

TyrRS belongs to the class I aaRSs, characterized by a Rossmannfold and two hallmark sequence motifs (“HIGH” and “KMSKS”).More specifically, TyrRS belongs to subclass Ic, together withtryptophanyl-tRNA synthetase (TrpRS), and contains an “AIDQ”motif characteristic of the ATP binding site. Progress has beenmadein inhibiting bacterial TyrRSs both with compounds identified fromnatural sources, as is the case of SB-219383 [19e21], and withsynthetized inhibitors [22e25]. However, there have been no sig-nificant advances in experimental work with parasitic TyrRSs asdrug targets [10]. A structure guided approach in the design ofdrugs for the treatment of VL would greatly benefit from theknowledge and analysis of new leishmanial TyrRSs structures, inparticular from L. donovani and L. infantum.

Recent biochemical and parasitological studies on L. donovaniTyrRS (LdTyrRS) have indicated that the enzyme has a cytoplasmicsubcellular location and that, as expected, it is essential for cell

proliferation. Interestingly, an intriguing additional function, i.e. aneffect on inflammation, has been shown in L. donovani for secretedTyrRS. The enzyme appears to be involved in attracting neutrophilsand binding to macrophages through an N-terminal ELR motiftriggering further cytokine TNF-a and IL-6 release by host macro-phages [26]. Regarding inhibition studies of LdTyrRS, the flavonoidfisetin has shown to have anti-leishmanial properties and its effecthas been ascribed to its anti-TyrRS activity [26]. This inhibitoryeffect of fisetin agrees well with the observation that this com-pound binds to the active site of L. major TyrRS (LmTyrRS) [27].

In order to provide a structural platform to assist further anti-leishmanial drug development, we report here the 2.75 Å resolu-tion crystal structure of LdTyrRS in complex with the tyrosyl ade-nylate analog TyrSA (50-O-[N-(L-tyrosyl)sulfamoyl]adenosine) and aspecific anti-LdTyrRS nanobody (NbA) used as crystallizationchaperone [28]. Nanobodies are the variable domains of camelidheavy chain-only antibodies which can substantially increase thesuccess of protein crystal growth [28]. The structure of LdTyrRS incomplex with TyrSA will in particular expand our insight into thebinding characteristics of ligands in the neighborhood of theadenine binding pocket, a feature which was not probed in theprevious structures of LmTyrRS. Analysis of the structure indicatesthe presence of promising opportunities to exploit structural dif-ferences of the parasite enzyme with both human cytosolic andmitochondrial TyrRS variants. Of particular relevance is the pres-ence of an “extra pocket” (EP) near the adenine binding site in thestructure of LdTyrRS. This pocket is absent in the two human ho-mologs and hence provides distinct opportunities to arrive at se-lective inhibitors. The EP is also present in the structure of L. majorTyrRS [27], and, in view of the considerable similarities in the activesite regions, is highly likely to occur in other Leishmania speciescausing human disease. The same holds for TyrRS from relatedTrypanosomatids, such as Trypanosoma brucei, the causative agentof human African trypanosomiasis, and T. cruzi, responsible forChagas disease. In addition, TyrRS from the malaria parasite Plas-modium falciparum [29] appears to contain a close variant of the EP.Amino acid sequence comparisons suggest that the EP occurs ineven more parasitic protozoa. Hence the EP may allow the devel-opment of selective tyrosyl-tRNA synthetase inhibitors which couldbecome new tools in tackling several major diseases caused byunicellular parasites.

2. Materials and methods

2.1. LdTyrRS expression and purification

LdTyrRS was cloned into the AVA0421 vector and expressed inE. coli for subsequent purification. A first round of Ni-NTA affinitychromatography was followed by cleavage of the N-terminal His6-tag using 3C protease (overnight at 4 �C). In a second Ni-NTA step,the cleaved LdTyrRS was purified from the N-terminally His6-tag-ged 3C protease. A size-exclusion chromatography on a Superdex200 column (Amersham Pharmacia Biotech) using SEC buffer(25 mM HEPES at pH 7.25, 500 mM NaCl, 2 mM TCEP, 5% glycerol,0.025% NaN3) was performed and a final yield of 6 mg of pureLdTyrRS per liter of E. coli culture was obtained and concentrated toabove 15 mg/mL for co-crystallization with NbA.

2.2. NbA production

LdTyrRS specific nanobodies were generated as previouslydescribed [28]. In brief, one llama (Lama glama) was immunized sixtimeswith in total 0.9mg of LdTyrRS. Four days after the final boost,blood was taken to isolate peripheral blood lymphocytes. RNA waspurified from these lymphocytes and reverse transcribed by PCR to

Table 1Crystallographic data collection and refinement statistics.

Parameters LdTyrRS�NbA�TyrSA(PDB: 5USF)

Data collectionSpace group P 65Cell dimensions: a, b, c (Å) 96.18, 96.18, 351.83Resolution (Å) 83.29e2.75

(2.85e2.75)Rmerge 0.210 (1.911)Rpim 0.069 (0.659)Observed reflections 488,525 (43,134)Unique reflections 47,547 (4632)Mean I/sI 8.7 (2.0)Multiplicity 10.3 (9.3)Completeness (%) 100.0 (100.0)CC1/2 0.995 (0.607)RefinementResolution (Å) 83.29e2.75Reflections used 45,156Rwork/Rfree 0.19/0.24Number of atomsProtein 12,283TyrSA 70Water 353

Number of residues 1600Average B-factors (Å2)Protein 66.0TyrSA 49.8Water 55.9

R.m.s. deviationsBond lengths (Å) 0.01Bond angles (�) 1.44

Ramachandran plota

Favored (%) 97Outlier (%) 0

Ligand (TyrSA)Average LLDFb �0.53Average RSRc 0.14

Values in parentheses are for the highest-resolution shell.a Ramachandran Plot statistics as reported by the wwPDB validation report.b Local ligand density fit as reported by the wwPDB validation report.c Real space R value as reported by the wwPDB validation report.

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136126

obtain cDNA. The resulting library was cloned into the phagedisplay vector pMESy4 bearing a C-terminal hexa-His tag and aCaptureSelect sequence tag (Glu-Pro-Glu-Ala). Six different familieswere selected by biopanning. For this, LdTyrRS was solid phasecoated directly on plates, LdTyrRS specific phage were recovered bylimited trypsinization. After two rounds of selection, periplasmicextracts were made and subjected to ELISA screens [28].

2.3. NbA expression and purification

NbA cloned in the pMESy4 vector that carries the pelB sequencecoding for the secretion signal peptide of PelB was expressed inE. coli for subsequent purification from the bacterial periplasm. NbApurificationwas performed as previously described by Pardon et al.[28]. Briefly, after induction, the bacterial pellet was gently resus-pended in TES buffer (200 mM Tris at pH 8.0, 0.5 mM EDTA,500 mM Sucrose) and upon incubation on ice the periplasmiccontent was recovered by centrifugation after an osmotic shockwith ice-cold 50 mM Tris at pH 8.0, 0.125 mM EDTA, 125 mM Su-crose. NbA was purified from the periplasmic content by Ni-NTAaffinity chromatography followed by size-exclusion chromatog-raphy on a Superdex 75 column (Amersham Pharmacia Biotech)using NbSEC buffer (25 mM HEPES at pH 7.25, 300 mM NaCl, 1 mMTCEP, 10% glycerol, 0.025% NaN3) and concentrated to above 10 mg/mL for co-crystallization with LdTyrRS.

2.4. Nanobody-LdTyrRS binding studies

Nanobody (Nb) binding studies were systematically carried outby native gel electrophoresis and size-exclusion chromatographyfor eight nanobodies. The purified Nb was first incubated withLdTyrRS for 30 min at 4 �C and after native gel electrophoresis thepositions of LdTyrRS, Nb and complex were analyzed. The forma-tion of the complex was also tested by comparing the elution peakson a Superdex 200 column (Amersham Pharmacia Biotech) whereLdTyrRS was run as well as the potential LdTyrRS�Nb complexformed upon 30 min incubation at 4 �C.

2.5. LdTyrRS aminoacylation assay

The IC50 of TyrSA in the LdTyrRS aminoacylation assay wasdetermined using methods as previously described [30e32].Briefly, TyrSA (tested in triplicate) was pre-incubated for 15 min atroom temperature with 0.13 nM LdTyrRS, 500 nM [3H]L-tyrosine(40 Ci/mmol), 0.2 mM ATP, 0.1 U/mL pyrophosphatase, 0.2 mMspermine, 0.2 mg/mL bovine serum albumin, 2.5 mM dithiothreitol,1 mMMgCl2, 25 mM KCl, 50 mMHEPES-KOH pH 7.6, and 2% DMSO.The reactions were started with 200 mg/mL tRNA from brewer'syeast (Roche) and incubated for 30 min at room temperaturewithout shaking. The reactions were stopped with cold 10% tri-chloroacetic acid and processed as previous described [32].

2.6. LdTyrRS�NbA�TyrSA complex crystallization

Purified LdTyrRS and NbA proteins were incubated on ice for30 min at a 1:2 M ratio followed by buffer exchange to crystalli-zation buffer (25 mM HEPES at pH 7.25, 100 mM NaCl, 1 mM TCEP-HCl, 5% glycerol, 0.025% NaN3). The complex (5 mg/mL) was thenincubated with 200 mM of TyrSA (50-O-[N-(L-tyrosyl)sulfamoyl]adenosine) on ice for 30e60 min prior to setting up the crystalli-zation tray. The crystals were obtained after 5e7 days at roomtemperature by vapor diffusion using sitting drops equilibratedagainst a reservoir containing 0.1 M sodium cacodylate pH 5.7, 22%PEG 4000. The drops contained 1 mL of LdTyrRS�NbA�TyrSA com-plex at 5 mg/mL and 1 mL of reservoir solution. After growth,

crystals were flash frozen in liquid nitrogen in cryo-solution (25%glycerol in reservoir solution) and stored until data collection.

2.7. Data collection and structure determination

Data was collected under cryogenic conditions at the StanfordSynchrotron Radiation Lightsource (SSRL) using beamline 12-2 at awavelength of 1 Å. Data processing was carried out with the pro-gram HKL2000 [33]. Initial phases were obtained by molecularreplacement using Phaser [34] with as models the structure ofL. major TyrRS�Tyrosinol ([27]; PDB: 3P0J) and a NbA homologymodel generated by Phyre2 [35]. This was followed by iterations ofmanual building and rebuilding using Coot [36] alternated withrefinement of the structure with REFMAC5 [37]. Refinement re-straints for TyrSA were obtained with the Grade web server [38].The structure validation server MolProbity [39] was usedthroughout the process for structure validation. The final datacollection and crystallographic refinement statistics are given inTable 1. Pymol [40] was used to create the figures. Coordinates andstructure factors of the LdTyrRS�NbA�TyrSA complex have beendeposited in the Protein Data Bank under the PDB ID: 5USF.

3. Results

3.1. Leishmania donovani tyrosyl-tRNA synthetase structure

The crystal structure of the LdTyrRS�NbA�TyrSA complex was

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136 127

determined at a resolution of 2.75 Å (Table 1), where NbA is ananobody to be described below, TyrSA a tyrosyl adenylate analog,and the symbol “�” indicates a non-covalent complex. The crystalscontain two copies of the LdTyrRS�NbA�TyrSA complex in theasymmetric unit (ASU). Each LdTyrRSmolecule is bound to one NbAmolecule (Fig. 1). Canonical TyrRSs are formed by two identicalmonomers, each consisting of a catalytic domain (CD) and ananticodon binding domain (ABD). Instead, the 75 kDa LdTyrRSchain is a pseudo-dimer where the structurally similar N- and C-terminal pseudo-monomers share only 23% sequence identity. Asdescribed for LmTyrRS [27], the N- and C-terminal pseudo-monomers are connected by a flexible linker between a14 and b9(Fig. 1A). The N-terminal pseudo-monomer of LdTyrRS contains thethree motifs characteristic for the catalytic domain of Class I tRNAsynthetases: “HIGH” (46HIAQ49 in LdTyrRS), “AIDQ” (182GLDQ185 inLdTyrRS) and “KMSKS” (222KMSKS226 in LdTyrRS) (Fig. 1B). Thesemotifs have been shown to be involved in the catalytic activity ofTyrRS enzymes [27,41]. The motifs are absent in the C-terminalpseudo-monomer, suggesting that this half of the molecule is notable to perform amino acid activation. This is in agreement with theobservations that TyrSA is bound to the active site in the N-terminal

Fig. 1. Domain organization of pseudo-dimeric LdTyrRS. A) The LdTyrRS pseudo-dimer strmodel). The N-terminal functional catalytic domain (CD) is shown in dark orange, the C-termthe C-terminal functional ABD in dark blue. The linker connecting the N-terminal and C-teinsertion present in Trypanosomatid TyrRSs, described previously by Larson et al. [27], in thethe functional CD. B) Domain organization of TyrRS from L. donovani, L. major, P. falciparum (to the ones in the LdTyrRS structure in part A. The characteristic motifs HIGH, AIDQ and KMfunctional anticodon binding domains are depicted. When processed by an elastase enzymEMAPII-like domain (yellow), which has cell signaling activity [29]. (EMAP stands for: “Endoterminal and C-terminal pseudo-monomers of LdTyrRS and ScTyrRS. The LdTyrRS N-termScTyrRS�tRNATyr complex (PDB ID: 2DLC) (sand). Residue F296, implied in ScTyrRS tRNATyr reA36) are depicted. The secondary structure elements in the LdTyrRS N-terminal ABD are labelN-terminal ABD.

CD of LdTyrRS, but not to the C-terminal CD. Hence, the N-terminalCD is called the “functional CD” and the C-terminal CD the “non-functional CD” (Fig. 1A).

The opposite is true for the ABD. Although both ABDs containthe sequence motifs “AC1” (244KIRQAYC250 and 578KIKKAYS584 inLdTyrRS) and “AC2” (313VSEDALK319 and 636LHPADLK642 in LdTyrRS)involved in the recognition of the tRNA anticodon arm, as describedfor LmTyrRS [27], the loop located between b7 and b8 in the N-terminal ABD is considerably shorter than the corresponding loopbetween b14 and b15 in the C-terminal pseudo-monomer. This loopis responsible for binding the anticodon base G34 of tRNATyr ac-cording to the structure of Saccharomyces cerevisiae TyrRS [42]. Theshort b7-b8 loop in the N-terminal ABD homolog is unable toengage with this base (Fig. 1C). Therefore, the N-terminal ABD ofLdTyrRS is called the “non-functional ABD” and the C-terminal ABDthe “functional ABD” (Fig. 1A, C).

The superposition of the two LdTyrRS chains A and B in theLdTyrRS�NbA�TyrSA crystals yields an overall r.m.s.d of 1.57 Å for678 Ca atoms. All four individual domains of the two chains in theasymmetric unit are highly similar, with r.m.s.d values rangingbetween 0.25 and 0.77 Å. The N- or C- terminal pseudo-monomers

ucture in complex with nanobody A (pink) and the tyrosyl adenylate analog TyrSA (CPKinal nonfunctional CD in light orange, the N-terminal non-functional ABD in light blue,rminal pseudo-monomers of LdTyrRS is shown in dark green, while the plant/plastidN-terminal ABD is shown in bright green. The ELR motif is located in the exposed a2 inPfTyrRS), human (HsTyrRS) and yeast S. cerevisiae (ScTyrRS). Domain colors correspondSKS in the functional catalytic domain as well as AC1 and AC2 in the nonfunctional ande HsTyrRS gives rise to an N-terminal TyrRS known as Mini TyrRS and a C-terminal

thelial Monocyte-Activating Polypeptide II”). C) Anticodon recognition regions of the N-inal (light blue) and C-terminal (dark blue) ABDs are shown superimposed with thecognition, and the tRNATyr anticodon bases guanine - pseudo-uridine - adenine (34G-j-ed. A black arrow points to the shortened loop between b7 and b8 in the nonfunctional

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136128

of the A and B chains are also similar to each other, with r.m.s.d.values of 0.52 Å and 0.77 Å after superposition. However, afterapplying the superposition operation of the N-terminal pseudo-monomers to the entire chains, there appears to be a change inorientation of the two C-terminal pseudo-monomers, resulting in a10.4 Å displacement at the farthest end of the C-terminal pseudo-monomer loop between b14 and b15 (Fig. 2). These comparisonsof chains A and B indicate that there is intrinsic flexibility withinthe LdTyrRS molecule, in particular between the N- and C-terminalpseudo-monomers.

When considering the LdTyrRS active site, all chain A residuesare well defined, while in chain B density for two residues (S224 andK225) belonging to the 221KMSKS226 loop is missing. Therefore wefocus our structural analysis on chain A of LdTyrRS. The KMSKS loopadopts a closed conformation in our LdTyrRS structure with thetyrosyl adenylate analog TyrSA bound (Fig. 3). Interestingly, theKMSKS loop was also found in a closed conformation in LmTyrRS,even in the absence of ATP or tyrosyl adenylate analogs [27].

Not only the KMSKS loop can have an open and closed confor-mation, also the entire active site can be in an open or closed state.These two distinct conformational states have been described byLarson et al. [27] for the active site in the LmTyrRS�Tyrosinol and

Fig. 2. Analysis of LdTyrRS flexibility through superposition of LdTyrRS chains Aand B in the asymmetric unit. Chain A is colored by domains as described in Fig. 1,while chain B is shown in grey. When superimposing the N-terminal pseudo-monomers of both LdTyrRS copies in the crystal, a shift in the C-terminal half of theenzyme is made evident especially in the farthest end of the C-terminal pseudo-monomer where a 10.4 Å displacement occurs (black arrow).

LmTyrRS�Fisetin crystal structures. In the functional catalytic do-mains of LdTyrRS�TyrSA both chains adopt the closed state (Fig. 3).In this state, the loop containing residues 146e154 is well-orderedand reaches the proximity of the sulfamoyl group of TyrSA, wherethe phosphate of tyrosyl adenylate or the a and b phosphates of ATPwould be during the catalytic reaction by TyrRS. As in the LmTyrRSclosed state, the LdTyrRS a4 and the loop connecting this helix witha5 curl over the active site to interact with residues 40e43 pre-ceding the HIGH motif (47HIAQ49 in LdTyrRS) involved in the en-zyme's catalytic activity (Fig. 3).

3.2. NbA structure and its interactions with LdTyrRS

After various failed attempts to crystallize LdTyrRS, we hy-pothesized that some flexibility in the four-domain pseudo-dimerwas impairing crystal formation. With the use of nanobodies ascrystallization chaperones we aimed to freeze LdTyrRS in aconformation more amenable for crystal growth, while the nano-body might also be able to establish favorable crystal contacts.Nanobodies are small compact single-domain fragments of theoriginal heavy-chain camelid antibodies that retain their fullantigen-binding capacity. Typically three variable loops of b-strands, referred to as the complementarity determining regions(CDRs), are responsible for binding to the antigen. The usually longCDR3 loop of nanobodies can have a special significance sincecryptic epitopes located in cavities or clefts of the antigenic proteinare sometimes recognized by it [28]. However, in the LdTyrRS�N-bA�TyrSA complex the CDR3 loop in anti-LdTyrRS llama nanobodyA (NbA) does not interact extensively with LdTyrRS as describedbelow.

We tested 4 anti-LdTyrRS nanobodies as crystallizationchaperones before obtaining high quality LdTyrRS crystals withNbA. The presence of the tyrosyl adenylate analog TyrSA wasessential as well, and its interaction with the enzyme will bedescribed later. The 1:1 LdTyrRS�NbA complex was purified by sizeexclusion chromatography (SEC) and the binding of NbA to LdTyrRSwas verified through the shift of the 280 nm absorbance elutionpeak (Fig. 4A), corroborating the formation of a larger molecularweight species. The presence of both proteins LdTyrRS and NbA inthe larger molecular weight species was confirmed by SDS-PAGE(Fig. 4B). In addition, native gel binding assays were performedsupporting the interaction between the two proteins (not shown).

NbA is a 14 kDa protein and consists of two b-sheets with aGreek key topology (Fig. 4C) held together by a disulfide bond(C22eC96). Interestingly, the CDR3 loop in NbA adopts an anti-parallel pair of b-strands connected by a short loop of 4 residues.Side chains of the b-strands of CDR3 do not interact with LdTyrRS,but the short connecting loop does. In this case, CDR2 buries alarger solvent accessible surface area (511 Å2) when interactingwith LdTyrRS than CDR3 (377 Å2) (calculated by PISA [43]).

With a buried surface area (BSA) of 1947 Å2, one NbA moleculeinteracts mainly with the non-functional CD of each LdTyrRS chainin the crystal, and makes a few additional contacts with the func-tional ABD. The corresponding estimated free energy of dissocia-tion DiG is �8 kcal/mol (calculated by PISA). The contacts betweenthe variable region of NbA and LdTyrRS are predominantly hydro-philic (Fig. 4D). Residues N31 and W33 of CDR1, residues R50 andG54 of CDR2, and, residue R102 of CDR3 (through a water mole-cule), make hydrogen bonds with LdTyrRS. Side chain carbon atomsof residues N56 (CDR2) and Y103 (CDR3) are engaged in hydro-phobic interactions with the enzyme. In addition, several framework residues, not located in CDR loops, establish polar (R19, S69and S71) and hydrophobic (Y80) interactions with the synthetase.

It is of interest to see if NbA has a stabilizing effect on LdTyrRS,thereby promoting possibly crystal growth. The comparison with

Fig. 3. Closed conformational state of the active site in the LdTyrRS�TyrSA structure. LdTyrRS in orange, TyrSA with purple carbon atoms. The various loops and motifs that arewell ordered in a closed conformation of the enzyme's active site are colored as follows: HIGH motif in red, KMSKS loop yellow, a4 - a5 loop cyan, and 146e154 loop green.

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136 129

the LmTyrRS structure is helpful here. The loop 541e571 connect-ing the non-functional CD with the functional ABD in LdTyrRS in-teracts with NbA and exhibits a well-defined density, but theequivalent loop in the three LmTyrRS crystal structures available ishalf or completely disordered [27]. Hence, it seems reasonable toconclude that NbA diminishes the flexibility of an exposed loop inLdTyrRS.

A second reason why nanobodies can promote crystal growth isby generating additional crystal contacts. In the LdTyrRS�N-bA�TyrSA complex, NbA bound to chain A establishes crystallo-graphic interactions with the other LdTyrRS copy (Chain B) with aBSA of 792 Å2. Moreover, NbA also contacts a symmetry-relatedNbA copy with a BSA of 969 Å2. The nanobody is clearly engagedin extensive interactions with various protein chains in the crystal.Most likely NbA increases the probability of crystal growth bymaking numerous crystal contacts and by diminishing the flexi-bility of an exposed loop.

3.3. TyrSA binds to LdTyrRS with its adenine ring near an extrapocket (EP)

The tyrosyl adenylate analog TyrSA (50-O-[N-(L-tyrosyl)sulfa-moyl]adenosine) (Fig. 5A) was necessary for obtaining well dif-fracting LdTyrRS crystals. As measured by aminoacylation assays,TyrSA binds tightly to LdTyrRS with an IC50 of 0.69 nM. The in-hibitor was found to occupy the active site of the functional CD(Fig. 5B) in both LdTyrRS chains in the asymmetric unit. The highaffinity of the compound for the enzyme is most probablyexplained by the large number of interactions with LdTyrRS active

site residues, limiting thereby the flexibility of this region. Thisdecrease in motility is likely responsible for the fact that TyrSAwascrucial for obtaining well diffracting crystals of the LdTyrRS�NbAcomplex.

The LdTyrRS active site contains two critical pockets: the tyro-sine binding pocket (YBP) where the tyrosyl group of TyrSA is sit-uated, and the adenine binding pocket (ABP) where the adeninemoiety of TyrSA binds (Fig. 5C). As evidenced through the use ofPoseView as part of the ProteinPlus structure-based modellingsoftware tools [44,45], the tyrosyl adenylate analog contacts manyLdTyrRS active site residues (BSA of 933 Å2), mainly throughhydrogen bonds, although a few hydrophobic interactions arepresent as well. Residues making hydrogen bonds with the TyrSAtyrosyl group in the YBP are Y36, Y163, Q167, D170 and Q185.Residues G38, A72 and F75 are responsible for the hydrophobicinteractions between enzyme and tyrosine moiety in the YBP.

A hydrogen bond is established between an oxygen atombelonging to the sulfamoyl group of TyrSA and the main chain ni-trogen atom of residue E40 of LdTyrRS. The residues responsible forhydrogen bond interactions with the ribose ring are D37, G182 andD184. In the ABP, hydrogen bonds with the adenine moiety aremade by H210 and L213, while M212 makes hydrophobic contactswith the adenine ring.

Further analysis of the active site near the ABP showed a pre-viously not reported feature: a pocket in LdTyrRS close to the TyrSAadenine moiety (Fig. 6). This pocket, called “extra pocket” (EP),which is also present in the L. major TyrRS structure, is mainly linedby residues belonging to helix a3 as well as by the side chain ofresidue H210. Helix a3 atoms contributing to the pocket (Fig. 6) are

Fig. 4. Interactions between LdTyrRS and NbA. A) SEC elution profile of purified LdTyrRS (grey) and LdTyrRS�NbA complex (orange). B) SEC fractions were run in SDS-PAGE tocorroborate the formation of the 1:1 complex. The low molecular weight present in the SEC chromatogramwas proven to be excess NbA present in the mix of LdTyrRS and NbA thatwas loaded into the SEC column. C) NbA structure (pink) showing the three CDR loops: CDR1 (cyan), CDR2 (green) and CDR3 (gold). D) LdTyrRS�NbA complex interactions. Theunderlined residues correspond to NbA. Most of the interactions are polar (doted lines). The three CDR loops are involved in interactions with the enzyme (in light orange).

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136130

the carbonyl oxygen of A48, and side chain atoms of Q49, F52, K53and N56. The EP of LdTyrRS harbors two deeply buried watermolecules. The most buried water molecule (Wat1) makeshydrogen bonds with the side chains of residues K53 and N56 andwithWat2.Wat2 engages in hydrogen bondingwith themain chaincarbonyl oxygen of A48 and with Wat1 and Wat3. The third watermolecule (Wat3) interacts with (i) the ether oxygen atom of theTyrSA ribose ring, (ii) Wat2 in the EP, and (iii) the carboxylate groupof residue D37 at the pocket entrance. The observation of this extrapocket (EP) in LdTyrRS and LmTyrRS turned out to be particularlyinteresting when comparing the TyrRS enzymes from the humanhost and a range of unicellular parasites.

3.4. The extra pocket (EP) of LdTyrRS is absent in the human TyrRSenzymes

In order to discover new opportunities for drug design, theactive sites of LdTyrRS and the two HsTyrRS enzymes werecompared (Fig. 7; for a list of TyrRS structures compared withLdTyrRS in this paper see Table 2). While the YBPs are highly similaramong the compared TyrRS structures, the analysis of the ABP re-veals interesting amino acid differences between human andparasite enzymes as also described by Larson et al. [27] on the basisof the comparison of their L. major TyrRS structure with the twohuman TyrRS enzymes. The basic conclusions of these authors are

confirmed by our current LdTyrRS structure.Interestingly, the EP offers additional exploitable differences

between the enzymes from parasite and host. When super-imposing the L. donovani and human TyrRS structures, a substantialdifference in position of a3 in the human enzymes vs. parasiteenzyme is evident (Figs. 7 and 8). Specifically, in cytosolic HsTyrRS,the helix corresponding to the LdTyrRS a3 helix is moved 3.0 Åtowards the EP of LdTyrRS (Fig. 8B), while in the mitochondrialHsTyrRS structure this helix is moved by 4.6 Å towards the EP ofLdTyrRS (Fig. 8C, D). As result of this shift, the EP is absent in thehuman enzymes.

It is worthwhile to try to discern why the EP is absent in thehuman enzymes. It appears that helix a3 side chains forming the EPin LdTyrRS are all different in the human enzymes. For instance,Q49 of LdTyrRS is equivalent to Y52 and H91 in the human cytosolicandmitochondrial enzymes, F52 of LdTyrRS to V54 and L93 in thesetwo human homologs, K53 of LdTyrRS to P55 and A94, and N56 ofLdTyrRS to K58 and G97, respectively. Actually, the L. donovaniTyrRS residues from C35 to K59, comprising helix a3, and theL. donovani TyrRS residues V206 to L213, surrounding LdTyrRSH210, are essentially all different when comparing LdTyrRS to thetwo human TyrRS enzymes (Fig. 9). As a result of all these changes,helix a3 shifts in the human enzymes compared to the L. donovanienzyme, and the EP is absent in the human homologs (Fig. 7).

Fig. 5. TyrSA binding to LdTyrRS. A) Chemical structure of TyrSA. B) LdTyrRS�NbA�TyrSA with difference electron density map calculated by omitting TyrSA, contoured at 3s(positive density in grey, negative density in red). C) General features of TyrSA binding mode. The protein surface and the two pockets, tyrosine binging pocket (YBP) and adeninebinding pocket (ABP), where the compound is bound are shown. D) Extensive hydrogen bond network in the LdTyrRS�TyrSA interaction.

Fig. 6. LdTyrRS residues forming the extra pocket (EP). Stereo view of the LdTyrRS catalytic site indicating the residues creating the EP. Most of the residues involved in the EPformation are part of helix a3, in addition to residues D37 and H210. TyrSA molecule shown with purple carbon atoms. Three water molecules, Wat1, Wat2 and Wat3 are located inthe pocket and drawn as grey spheres.

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136 131

3.5. A sequence fingerprint for the extra pocket (EP)

The seven residues forming the EP in LdTyrRS: D37, A48, Q49,F52, K53, N56 and H210 can be used as an “EP fingerprint”: D-x10-AQ-x2-FK-x2-N-z-H, where x stands for any amino acid and z for avariable large number of residues. This EP fingerprint is present inthe TyrRS sequences of all Leishmania species analyzed (Fig. 9).Hence, it is likely that the EP is present in all Leishmania TyrRSenzymes, as confirmed by a comparison of the L. major andL. donovani TyrRS structures (Fig. 7A and B). The EP fingerprint alsooccurs in other Trypanosomatids. The latter include the importanthuman pathogens Trypanosoma brucei and T. cruzi, as well as

T. vivax, causing nagana in cattle in sub-Saharan Africa. Given theabsence of the EP in the two human homologs, this difference in-dicates interesting opportunities for arriving at compounds withhigher affinity for Trypanosomatid than for the human tyrosyl-tRNA synthetases.

4. Discussion

The structure of the 75 kDa LdTyrRS in complex with TyrSA hasincreased our insights in the architecture of this unusual member ofthe tRNA synthetase family. The two pseudo-monomers compriseeach two domains. The N-terminal pseudo-monomer contains a

Fig. 7. LdTyrRS catalytic pocket surface representation and comparison to human TyrRSs. In panels A, B and E, the extra pocket (EP) is labeled. A) The LdTyrRS�TyrSA structurein complex shown with TyrSA with purple carbon atoms. B) The LmTyrRS�Tyrosinol structure with the tyrosinol molecule depicted with cyan carbon atoms (PDB: 3P0J). The EP inthe L. major enzyme is very similar to that in L. donovani TyrRS. C) HsMitoTyrRS�TyrSA structure with the TyrSA molecule in pink (PDB: 2PID). The absence of the EP is indicated witha dashed arrow. D) HsCytoTyrRS�Tyrosine structure with the tyrosine molecule with green carbon atoms (PDB: 4QBT). The lack of the EP is indicated with a dashed arrow. E) Surfacerepresentation of the LdTyrRS binding pocket with bound TyrSA molecule with purple carbon atoms. Protein carbon atoms are colored grey, nitrogens blue, oxygens red, and sulfuratoms yellow. The adenine binding pocket (ABP) and the EP are indicated. The three water molecules occupying the EP are shown as red spheres. Two of those (Wat1 and Wat2) aredeeply positioned in the EP. F) Surface representation of the HsMitoTyrRS binding pocket with bound TyrSA with pink carbon atoms (PDB: 2PID) in the same orientation as theLdTyrRS structure in E. Like in the human cytosolic enzyme (not shown in surface representation), no EP is present.

Table 2Compared structures of tyrosyl tRNA synthetases.

PDB ID Organism Ligand(s) Crystallized TyrRS Reference

5USF Leishmania donovani TyrSA and nanobody Full length This publication3P0H Leishmania major Fisetin Full length [27]3P0J Leishmania major Tyrosinol Full length [27]3VGJ Plasmodium falciparum TyrAMP Full length [29]4QBT Homo sapiens Cytosolic L-Tyrosine Truncated (1e341) [46]2PID Homo sapiens Mitochondrial TyrSA Full length [47]2DLC Saccharomyces cerevisiae Tyr-AMP analog and tRNATyr Truncated (1e364) [42]1JII Staphylococcus aureus SB-219383 Full length [20]2JAN Mycobacterium tuberculosis None Full length [48]1VBM Escherichia coli TyrSA Truncated (1e322) [41]1JLU Methanocaldococcus jannaschii L-Tyrosine and tRNATyr Full length [49]

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136132

functional catalytic domain and a non-functional anticodon-bind-ing domain, and the C-terminal pseudo-monomer a non-functionalcatalytic domain and a functional anticodon binding domain. Asimilar architecture had been observed for L. major TyrRS [27].Comparison of the two crystallographically independent chains inour structure showed that the two pseudo-monomers can flex with

respect to each other. This results in a considerable difference inposition of the functional anti-codon binding domain at the C-terminus of the second pseudo-monomer when the N-terminalpseudo-monomers are superimposed (Fig. 2). This flexibility mightbe essential in positioning tRNATyr properly with respect to thecatalytic domain during the second catalytic step of the reaction

Fig. 8. Comparison of a3 in human and Leishmania TyrRS structures. The superposition of structures was done to understand the absence of the EP in the human TyrRSs. A)LdTyrRS�TyrSA structure in orange with TyrSA molecule in purple showing the EP surface in grey. B) Superposition of LdTyrRS�TyrSA with HsMitoTyrRS�TyrSA (PDB: 2PID) in pink,HsCytoTyrRS�Tyrosine (PDB: 4QBT) in light green and LmTyrRS�Tyrosinol (PDB: 3P0J) in cyan; LdTyrRS EP surface in grey. C) A 4.6 Å and 3.0 Å shift in a3 prevents the formation ofthe EP in HsMitoTyrRS and HsCytoTyrRS, respectively. D) Another view of the superposition of the three enzymes showing the shift of helix a3 in the HsTyrRS structures.

Fig. 9. Partial amino acid sequence alignment of TyrRSs from selected species and the signature fingerprint of the extra pocket (EP). The residues involved in the EP formationin L. donovani (top line) are indicated with red boxes. When the corresponding residues in other species are identical to those of the EP fingerprint in LdTyrRS then these residues arealso enclosed in a red box. Residues L and I at positions 4 and 6 of the EP fingerprint are indicated with orange boxes. The EP was not found in the available structures of human,yeast, bacteria or archaeal (Methanocaldococcus jannaschii [49]) TyrRSs which agrees with the lack of conservation of the EP fingerprint residues. The amino acid sequence alignmentof LdTyrRS with enzymes from selected pathogenic bacterial species shows that these bacterial TyrRS do not contain an EP, and comparisons of TyrRS crystal structures from(S. aureus and M. tuberculosis [20,48]) with LdTyrRS confirm that this is indeed the case (not shown). The full species names are: Ldonovani ¼ Leishmania donovani;Lmajor ¼ Leishmania major; Lmexicana ¼ Leishmania mexicana; Tcruzi ¼ Trypanosoma cruzi; Tbrucei ¼ Trypanosoma brucei; Pfalciparum ¼ Plasmodium falciparum;Pvivax ¼ Plasmodium vivax; Cparvum ¼ Cryptosporidium parvum; Tgondii ¼ Toxoplasma gondii; Glamblia ¼ Giardia lamblia; Ehistolytica ¼ Entamoeba hystolitica;HsapCyto ¼ cytosolic Homo sapiens; HsapMito ¼ mitochondrial Homo sapiens; Scerevisiae ¼ Saccharomyces cerevisiae; Saureus ¼ Staphylococcus aureus;Mtuberculosis ¼ Mycobacterial tuberculosis; Mjannaschii ¼ Methanocaldococcus jannaschii.

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136 133

catalyzed by Leishmania TyrRS.The binding mode of TyrSA to LdTyrRS provides additional

insight as to how the tyrosyl-adenylate adduct binds to the activesite of this group of TyrRS enzymes. A new feature revealed by our

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136134

analysis of the LdTyrRS structure is the presence of an “extrapocket” (EP), close to the adenine binding pocket and filled by threemutually interacting water molecules. This EP is present as well inthe L. major TyrRS structure [27] and, based on amino acid sequencecomparisons, also in other Trypanosomatids. Several of the latter,such a T. brucei and T. cruzi, are major global human pathogenscausing death and disease in particular in low-income countries[2,3]. Since the EP is absent in the two human TyrRS enzymes, thisdifference between enzymes from parasite and human host in-dicates opportunities to arrive at inhibitors with high affinity andselectivity.

It is of interest to evaluate if the EP is present in the TyrRS fromother medically important parasites. The enzyme from the proto-zoan Plasmodium falciparum, the major causative agent of malaria,has been the subject of several studies, including the determinationof the crystal structure of cytosolic P. falciparum TyrRS (PfTyrRS)[29]. The EP fingerprint is present in PfTyrRS with two changes: theF at the fourth position in the fingerprint is in PfTyrRS an L, whichconserves the hydrophobic side chain characteristic at this position,and the N at the sixth position is an I in PfTyrRS (Fig. 9). A com-parison of the structure of LdTyrRS with the crystal structure ofP. falciparum TyrRS [29], reveals that in the latter enzyme the EP isindeed present (Fig. 10). In this case, the EP harbors only one deepwater molecule, the equivalent of Wat2 in the LdTyrRS EP. Theabsence of Wat1 is due to the fact that residue N56, in the back ofthe EP of LdTyrRS (Fig. 7B), is substituted by I80 in the P. falciparumenzyme, providing a more hydrophobic environment within thepocket and no hydrogen-bonding partner for Wat1. Based on the

Fig. 10. P. falciparum TyrRS residues forming the extra pocket (EP). Surface representatiatoms are colored grey, nitrogens blue, oxygens red and phosphorous orange. All but twosomatids. PfTyrRS harbors one deep water in the EP (equivalent to Wat2 in LdTyrRS) and a wTyrAMP carbon atoms are shown in green.

amino acid sequence alignment (Fig. 9), this same feature is presentin P. vivax, another important malaria parasite. These comparisonssuggest that there are opportunities to arrive at high affinityPfTyrRS inhibitors with good selectivity. This is extra interestingsince Kahn [18] has recently emphasized the opportunities of TyrRSas a drug target for malaria.

Turning to other protozoa causing considerable human sufferingand death across the globe, the comparison of EP fingerprints inCryptosporidium parvum, Toxoplasma gondii and Giardia lambliareveals that in these pathogens there is only a single residue dif-ference in the fingerprint: the F at the fourth position is an L, I or Ain these three parasites, respectively. Hence the EP in these pro-tozoa is even closer to that of LdTyrRS than the EP in the twoP. falciparum species. The N at the sixth position of the fingerprint ismaintained so that it is very likely that in these three species the EPis present with Wat1 deep in the pocket. Consequently, the EPmight enable the development of effective and selective TyrRS in-hibitors for treating cryptosporidiosis, toxoplasmosis and giardi-asis. Although in Entamoeba histolytica the EP fingerprint is lessperfectly conserved, with differences in the first (D to N), fourth (Fto L) and fifth (K to T) positions with respect to LdTyrRS, two ofthese three side chain changes are very conservative. Hence, anexploitable EP might also exist in the TyrRS of this pathogen.

5. Conclusions

The structure of the LdTyrRS�NbA�TyrSA complex elucidatedrevealed different mutual orientations of its two pseudo-

on of the EP in the PfTyrRS�TyrAMP crystal structure [29] (PDB: 3VGJ). Protein carbon(L76 and I80) of the residues involved in the EP formation are shared with trypano-ater molecule at the EP entrance (equivalent to Wat3 in LdTyrRS), shown as red spheres.

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136 135

monomers. The comparison of the LdTyrRS structure to both hu-man homologs showed major differences compared in the activesite, in particular the presence of an “extra pocket” in the parasiteenzyme. The latter feature is also found in the TyrRSs from otherunicellular parasites and could be a key element in the develop-ment of novel compounds for treating diseases caused by a widerange of important pathogenic protozoa.

Author contributions

Protein purification and crystallization: KMM, CYK.Nanobody generation: EP and JS.Inhibitor synthesis: ZZ and EF.Enzyme inhibition measurements: RMR, JRG, FSB.Crystallographic data collection and refinement: XB-A, ST.Structure analysis: XB-A, CLMJV, WGJH.XB-A and WGJH wrote the manuscript with input from all

authors.

Acknowledgments

We like to thank Ethan Merritt for stimulating discussions.Research reported in this publication was supported by the Na-tional Institute of Allergy and Infectious Diseases of the NationalInstitutes of Health under award number R01AI084004 (to WGJH)and R01AI097177 (to FSB and EF). We thank Instruct, part of theEuropean Strategy Forum on Research Infrastructures (ESFRI), andthe Research Foundation Flanders (FWO) for their support to theNanobody discovery. We are grateful to Nele Buys for the technicalassistance during Nanobody discovery. We acknowledge the sup-port of a Fulbright Fellowship to X.B.-A. We thank Robert Steinfeldtfor providing support for the computing environment at the Bio-molecular Structure Center of the University of Washington. Crys-tallography performed in support of the work benefitted fromremote access to resources at the Stanford Synchrotron RadiationLightsource supported by the U.S. Department of Energy Office ofBasic Energy Sciences under Contract No. DE-AC02-76SF00515 andby the National Institutes of Health (P41GM103393). The content issolely the responsibility of the authors and does not necessarilyrepresent the official views of the National Institutes of Health.

References

[1] DNDi, Leishmaniasis fact sheet, Available from: http://www.dndi.org/wp-content/uploads/2016/10/Factsheet_2015_Leish.pdf, 2015.

[2] WHO, Leishmaniasis fact sheet, september 2016, Available from: http://www.who.int/mediacentre/factsheets/fs375/en/, 2016.

[3] WHO, Control of the Leishmaniasis. Report of a Meeting of the WHO ExpertCommittee on the Control of Leishmaniases, 22 e 26 March 2010, Geneva2010, 2010. Available from: http://apps.who.int/iris/bitstream/10665/44412/1/WHO_TRS_949_eng.pdf.

[4] A.H. Fairlamb, et al., Drug resistance in eukaryotic microorganisms, Nat.Microbiol. 1 (7) (2016) 16092.

[5] S. Sundar, J. Chakravarty, Leishmaniasis: an update of current pharmaco-therapy, Expert Opin. Pharmacother. 14 (1) (2013) 53e63.

[6] M. Ibba, D. Soll, Aminoacyl-tRNA synthesis, Annu. Rev. Biochem. 69 (2000)617e650.

[7] K. Sheppard, P.M. Akochy, D. Soll, Assays for transfer RNA-dependent aminoacid biosynthesis, Methods 44 (2) (2008) 139e145.

[8] F. Charriere, et al., Dual targeting of a tRNAAsp requires two different aspartyl-tRNA synthetases in Trypanosoma brucei, J. Biol. Chem. 284 (24) (2009)16210e16217.

[9] U.A. Ochsner, et al., Aminoacyl-tRNA synthetases: essential and still promisingtargets for new anti-infective agents, Expert Opin. Investig. Drugs 16 (5)(2007) 573e593.

[10] J.S. Pham, et al., Aminoacyl-tRNA synthetases as drug targets in eukaryoticparasites, Int. J. Parasitol. Drugs Drug Resist 4 (1) (2014) 1e13.

[11] P. Schimmel, J.S. Tao, J. Hill, Aminoacyl tRNA synthetases as targets for newanti-infectives, Faseb J. 12 (15) (1998) 1599e1609.

[12] P. Fang, et al., Structural basis for full-spectrum inhibition of translationalfunctions on a tRNA synthetase, Nat. Commun. 6 (2015) 6402.

[13] T. Nakama, O. Nureki, S. Yokoyama, Structural basis for the recognition ofisoleucyl-adenylate and an antibiotic, mupirocin, by isoleucyl-tRNA synthe-tase, J. Biol. Chem. 276 (50) (2001) 47387e47393.

[14] F.L. Rock, et al., An antifungal agent inhibits an aminoacyl-tRNA synthetase bytrapping tRNA in the editing site, Science 316 (5832) (2007) 1759e1761.

[15] C.Y. Koh, et al., Structures of Trypanosoma brucei methionyl-tRNA synthetasewith urea-based inhibitors provide guidance for drug design against sleepingsickness, PLoS Negl. Trop. Dis. 8 (4) (2014) e2775.

[16] S. Khan, et al., Structural basis of malaria parasite lysyl-tRNA synthetase in-hibition by cladosporin, J. Struct. Funct. Genomics 15 (2014) 9.

[17] H. Zhou, et al., ATP-directed capture of bioactive herbal-based medicine onhuman tRNA synthetase, Nature 494 (7435) (2013) 121e124.

[18] S. Khan, Recent advances in the biology and drug targeting of malaria parasiteaminoacyl-tRNA synthetases, Malar. J. 15 (2016) 203.

[19] R.C. Greenwood, D.R. Gentry, Confirmation of the antibacterial mode of actionof SB-219383, a novel tyrosyl tRNA synthetase inhibitor from a Micro-monospora sp, J. Antibiot. 55 (4) (2002) 423e426.

[20] X. Qiu, et al., Crystal structure of Staphylococcus aureus tyrosyl-tRNA syn-thetase in complex with a class of potent and specific inhibitors, Protein Sci.10 (2001) 9.

[21] A.L. Stefanska, et al., SB-219383, a novel tyrosyl tRNA synthetase inhibitorfrom a Micromonospora sp I. Fermentation, isolation and properties,J. Antibiot. 53 (4) (2000) 345e350.

[22] P. Brown, et al., Synthetic analogues of SB-219383. Novel C-glycosyl peptidesas inhibitors of tyrosyl tRNA synthetase, Bioorg. Med. Chem. Lett. 11 (5)(2001) 711e714.

[23] L.W. Chen, et al., Metronidazole containing pyrazole derivatives potentlyinhibit tyrosyl-tRNA synthetase: design, synthesis, and biological evaluation,Chem. Biol. Drug Des. 88 (4) (2016) 592e598.

[24] R.L. Jarvest, et al., Interaction of tyrosyl aryl dipeptides with S-aureus tyrosyltRNA synthetase: inhibition and crystal structure of a complex, Bioorg. Med.Chem. Lett. 9 (19) (1999) 2859e2862.

[25] X.D. Wang, et al., 3-Aryl-4-acyloxyethoxyfuran-2(5H)-ones as inhibitors oftyrosyl-tRNA synthetase: synthesis, molecular docking and antibacterialevaluation, Bioorg. Med. Chem. 21 (17) (2013) 4914e4922.

[26] S. Anand, R. Madhubala, Twin attributes of Tyrosyl-tRNA synthetase ofLeishmania donovani: a housekeeping protein translation enzyme and aMIMIC of host chemokine, J. Biol. Chem. 291 (34) (2016) 17754e17771.

[27] E.T. Larson, et al., The double-length tyrosyl-tRNA synthetase from theeukaryote Leishmania major forms an intrinsically asymmetric pseudo-dimer,J. Mol. Biol. 409 (2) (2011) 159e176.

[28] E. Pardon, et al., A general protocol for the generation of Nanobodies forstructural biology, Nat. Protoc. 9 (3) (2014) 674e693.

[29] T.K. Bhatt, et al., Malaria parasite tyrosyl-tRNA synthetase secretion triggerspro-inflammatory responses, Nat. Commun. 2 (2011) 530.

[30] S. Shibata, et al., Selective inhibitors of methionyl-tRNA synthetase havepotent activity against Trypanosoma brucei Infection in Mice, Antimicrob.Agents Chemother. 55 (5) (2011) 1982e1989.

[31] S. Shibata, et al., Urea-based inhibitors of Trypanosoma brucei methionyl-tRNA synthetase: selectivity and in vivo characterization, J. Med. Chem. 55(14) (2012) 6342e6351.

[32] L. Pedro-Rosa, et al., Identification of potent inhibitors of the Trypanosomabrucei methionyl-tRNA synthetase via high-throughput orthogonal screening,J. Biomol. Screen 20 (1) (2015) 122e130.

[33] Z. Otwinowski, W. Minor, Processing of X-ray diffraction data collected inoscillation mode, in: C. Carter, R. Sweet (Eds.), Macromolecular Crystallog-raphy, Part A, Academic Press, 1997, pp. 307e326.

[34] A.J. McCoy, et al., Phaser crystallographic software, J. Appl. Crystallogr. 40 (Pt4) (2007) 658e674.

[35] L.A. Kelley, et al., The Phyre2 web portal for protein modeling, prediction andanalysis, Nat. Protoc. 10 (6) (2015) 845e858.

[36] P. Emsley, et al., Features and development of Coot, Acta Crystallogr. D. Biol.Crystallogr. 66 (Pt 4) (2010) 486e501.

[37] G.N. Murshudov, A.A. Vagin, E.J. Dodson, Refinement of macromolecularstructures by the maximum-likelihood method, Acta Crystallogr. D. Biol.Crystallogr. 53 (1997) 16.

[38] O.S. Smart, et al., Grade Web Server, Version 1.001, 2011. Available from:http://grade.globalphasing.org.

[39] V.B. Chen, et al., MolProbity: all-atom structure validation for macromolecularcrystallography, Acta Crystallogr. D. Biol. Crystallogr. 66 (Pt 1) (2010) 12e21.

[40] L.L.C. Schr€odinger, The PyMOL Molecular Graphics System, Version 1.7,Available from: https://www.pymol.org.

[41] T. Kobayashi, et al., Structural snapshots of the KMSKS loop rearrangement foramino acid activation by bacterial tyrosyl-tRNA synthetase, J. Mol. Biol. 346(1) (2005) 105e117.

[42] M. Tsunoda, et al., Structural basis for recognition of cognate tRNA by tyrosyl-tRNA synthetase from three kingdoms, Nucleic Acids Res. 35 (13) (2007)4289e4300.

[43] E. Krissinel, K. Henrick, Inference of macromolecular assemblies from crys-talline state, J. Mol. Biol. 372 (3) (2007) 774e797.

[44] P. Fricker, M. Gastreich, M. Rarey, Automated generation of structural mo-lecular formulas under constraints, J. Chem. Inf. Comput. Sci. 44 (3) (2004) 14.

[45] K. Stierand, P.C. Maass, M. Rarey, Molecular complexes at a glance: automatedgeneration of two-dimensional complex diagrams, Bioinformatics 22 (14)(2006) 1710e1716.

X. Barros-�Alvarez et al. / Biochimie 138 (2017) 124e136136

[46] M. Sajish, P. Schimmel, A human tRNA synthetase is a potent PARP1-activating effector target for resveratrol, Nature 519 (7543) (2015) 370e373.

[47] L. Bonnefond, et al., Crystal structure of human mitochondrial tyrosyl-tRNAsynthetase reveals common and idiosyncratic features, Structure 15 (11)(2007) 1505e1516.

[48] M.D. Hartmann, et al., The Structure of Tyrosyl-tRNA Synthetase fromMycobacterium Tuberculosis, PDB ID 2JAN, 2006.

[49] T. Kobayashi, et al., Structural basis for orthogonal tRNA specificities oftyrosyl-tRNA synthetases for genetic code expansion, Nat. Struct. Biol. 10 (6)(2003).