Crystal structure of UDP-glucose pyrophosphorylase from Yersinia pestis, a potential therapeutic target against plague

The crystal structure of UDP-glucose pyrophosphorylase from Yersinia pestis has been determined to 2.17 Å resolution, providing a foundation for future structure-based drug-design studies.

Abstract

Yersinia pestis, the causative agent of bubonic plague, is one of the most lethal pathogens in recorded human history. Today, the concern is the possible misuse of Y. pestis as an agent in bioweapons and bioterrorism. Current therapies for the treatment of plague include the use of a small number of antibiotics, but clinical cases of antibiotic resistance have been reported in some areas of the world. Therefore, the discovery of new drugs is required to combat potential Y. pestis infection. Here, the crystal structure of the Y. pestis UDP-glucose pyrophosphorylase (UGP), a metabolic enzyme implicated in the survival of Y. pestis in mouse macrophages, is described at 2.17 Å resolution. The structure provides a foundation that may enable the rational design of inhibitors and open new avenues for the development of antiplague therapeutics.

1. Introduction  

The Gram-negative bacterium Yersinia pestis is the etiologic agent of plague, one of the most fatal diseases in recorded human history (Butler, 2014 ▸). Although clinical cases of plague are relatively uncommon today, re-emerging infections have been reported in areas of Asia, Africa and even the United States (Dai et al., 2018 ▸; Grácio & Grácio, 2017 ▸; Shi et al., 2018 ▸; Rahelinirina et al., 2017 ▸; Centers for Disease Control and Prevention, 1994 ▸). Even with the low rate of clinical cases of plague, the main concern with Y. pestis as an infectious agent continues to be its classification as a high-threat category A bioterrorism agent by the Centers for Disease Control and Prevention (Darling et al., 2002 ▸; Inglesby et al., 2000 ▸; Weant et al., 2014 ▸). Currently, early diagnosis and the rapid administration of a limited number of effective antibiotics remain the best treatment option for Y. pestis infections (Yang, 2018 ▸; Wong et al., 2000 ▸; Frean et al., 1996 ▸). However, drug-resistant isolates of Y. pestis have developed, raising concerns about potential resistance to current antibiotics (Galimand et al., 1997 ▸, 2006 ▸; Hinnebusch et al., 2002 ▸). Given the ongoing threat of its misuse as a bioterrorism agent, the discovery of novel drug targets within the Y. pestis genome, vaccine targets and other treatment options continue to warrant further investigation (Tao et al., 2017 ▸; Sharma & Pan, 2012 ▸; Stenseth et al., 2008 ▸).

Y. pestis uses a type III secretion system (T3SS) to achieve infection of the host, and a number of proteins that are components of the T3SS have been identified as attractive targets for small-molecule inhibitors and vaccines (Deng et al., 2017 ▸; Charro & Mota, 2015 ▸). Another promising approach is to target enzymes that are essential for bacterial metabolism (Ahn et al., 2014 ▸; Charusanti et al., 2011 ▸). The Y. pestis galU gene encodes a glucose-1-phosphate uridylyltransferease (UDP-glucose pyrophosphorylase, UGP) that catalyzes the reversible formation of UDP-glucose and pyrophosphate from glucose 1-phosphate and uridine 5′-triphosphate (UTP) (Klein et al., 2012 ▸). The galU gene is ubiquitous in both bacteria and eukaryotes, and the expressed UGP catalyzes the formation of the main glycosyl donor, UDP-glucose, in oligosaccharide and polysaccharide biosynthesis during carbohydrate metabolism (Frey, 1996 ▸). UDP-glucose is also essential for the synthesis of sugar moieties composing the cell wall, capsular polysaccharides, membrane-derived oligosaccharides and lipo­polysaccharides (Flores-Díaz et al., 1997 ▸). Bacterial UGPs share a high level of amino-acid sequence identity but a low level of sequence identity with their eukaryotic counterparts, indicating that they are evolutionarily divergent and unrelated (Flores-Díaz et al., 1997 ▸; Kleczkowski et al., 2004 ▸; Führing et al., 2015 ▸). In fact, no recognizable sequence similarity exists between human and Y. pestis UGPs. The significant sequence differences make bacterial UGPs attractive drug targets for specifically inhibiting the pathogenic enzyme in human hosts while avoiding off-target effects (Berbís et al., 2015 ▸).

UGPs have been linked to virulence in several bacterial species such as Shigella flexneri, Pseudomonas aeruginosa, Helicobacter pylori and Streptococcus pneumoniae (Sandlin et al., 1995 ▸; Dean & Goldberg, 2002 ▸; Kim et al., 2010 ▸; Cools et al., 2018 ▸). The galU gene has also been demonstrated to serve a functional role in the intracellular survival of Y. pestis in mouse macrophages (Klein et al., 2012 ▸). Structural studies using X-ray crystallography have revealed that many of these bacterial UGPs share a very similar three-dimensional structure in which the enzyme crystallizes as a tetramer, with each monomer consisting of a central mixed β-sheet that is surrounded by α-helices (Berbís et al., 2015 ▸). Efforts to screen for inhibitors of S. pneumoniae UGP identified a functional assay that has yielded initial compounds that display inhibitory activity, but to date no other inhibitors of bacterial UGPs have been reported (Zavala et al., 2017 ▸; Berbís et al., 2015 ▸). Here, we describe the first crystal structure of UGP from Y. pestis, which may be useful in efforts to develop inhibitors of this enzyme.

2. Materials and methods  

2.1. Cloning, expression and purification  

The Y. pestis UDP-glucose pyrophosphorylase (UGP) gene was amplified from genomic DNA by polymerase chain reaction (PCR) using oligonucleotide primers PE-3072 and PE-3073 (Table 1 ▸). The resulting PCR amplicon was then used as the template for a second PCR amplification with primers PE-277 and PE-3073 (Table 1 ▸). The amplicon from the second PCR, coding for Y. pestis UGP with an N-terminal Tobacco etch virus (TEV) protease cleavage site (ENLYFQ/S), was recombined into the entry vector pDONR221 (Life Technologies, Grand Island, New York, USA) using Gateway cloning to generate the entry clone pMG3029, and the nucleotide sequence was confirmed experimentally. The Y. pestis UGP gene was then recombined from pMG3029 into pDEST527 to generate the expression vector pMG3038. This plasmid produces N-terminally hexahistidine-tagged UGP that can be cleaved by TEV protease to yield untagged Y. pestis UGP (Raran-Kurussi et al., 2017 ▸). The protein was overexpressed in Escherichia coli BL21-CodonPlus (DE3)-RIL competent cells. A 4 l culture was grown to mid-log phase (OD₆₀₀ of ∼0.4–0.6) at 310 K in Luria broth with 100 µg ml⁻¹ ampicillin and 30 µg ml⁻¹ chloramphenicol. Overexpression of UGP was induced with 1 mM isopropyl β-d-1-thiogalactopyranoside for 4 h at 303 K. The cells were pelleted by centrifugation and stored at 193 K.

Table 1. Macromolecule-production information.

Source organism	Y. pestis
DNA source	Y. pestis strain 195/P obtained from the United States Army Medical Research Institute of Infectious Diseases (USAMRIID)
Piggyback primer (PE-277)	G GGG ACA AGT TTG TAC AAA AAA GCA GGC TCG GAG AAC CTG TAC TTC CAG
Forward primer (PE-3072)	GAG AAC CTG TAC TTC CAG TCT ATG AAG TGT TTA AAA GCC GTC ATT CC
Reverse primer (PE-3073)	GGG GAC CAC TTT GTA CAA GAA AGC TGG GTT ATT AAT CGA GTT GCT GTT TTA GCC AGG
Cloning vector	pDONR221
Expression vector	pMG3038
Expression host	E. coli BL21(DE3)-CodonPlus RIL
Complete amino-acid sequence of the construct produced	SMKCLKAVIPVAGLGTRMLPATKAIPKEMLPVVDKPLIQYIVDECVAAGVKEIVLVSHSSKNAIENHFDTSFELEAALESRVKRQLLKEIKNICPADVTIMQVRQGHAKGLGHAVLCAKSMVGDNPFIVMLPDVLLDDSTADLSKENLASMIKRFEETGHSQIMVEPVPKADVSKYGIADCGHVALAPGESTLMTAVVEKPSIAEAPSNLAVVGRYVLSKNIWPLLEKTPRGAGDEIQLTDAIAMLMQQEPVEAFHMTGKSHDCGDKLGYMKAFVTYGVRHHTEGEKFTAWLKQQLD

All purification steps were performed at 277 K. The E. coli cell paste (17 g) was resuspended in 200 ml buffer consisting of 50 mM Tris–HCl pH 7.5, 200 mM NaCl, 25 mM imidazole (lysis buffer) with four cOmplete EDTA-free protease-inhibitor cocktail tablets (Roche Molecular Biochemicals, Indianapolis, Indiana, USA). The cells were lysed by three passes through an APV-1000 homogenizer (Invensys APV Products, Albertslund, Denmark) at 69 MPa, and the lysate was centrifuged for 30 min at 30 000g. The supernatant was passed through a 0.2 µm polyethersulfone membrane and applied onto a 5 ml HisTrap column (GE Healthcare, Piscataway, New Jersey, USA) pre-equilibrated with lysis buffer. The column was washed to baseline with lysis buffer and the protein was eluted using a linear gradient from 25 mM to 1 M imidazole in lysis buffer. Fractions containing His₆-UGP were combined and concentrated with an Amicon stirred cell using a YM10 membrane (EMD Millipore, Billerica, Massachusetts, USA). The concentrated protein was diluted with 50 mM Tris–HCl pH 7.5, 200 mM NaCl buffer to achieve an imidazole concentration of 25 mM. The His₆-UGP was then digested with 5 mg polyhistidine-tagged TEV protease overnight (Kapust et al., 2001 ▸). The protein was applied onto a 5 ml HisTrap column pre-equilibrated with lysis buffer. Y. pestis UGP was isolated in the column flowthrough and concentrated. The concentrated sample was applied onto a HiPrep S200 column (GE Healthcare) equilibrated with 25 mM Tris–HCl pH 7.4, 150 mM NaCl, 2 mM tris(2-carboxyethyl)phos­phine (TCEP) buffer. The peak fractions were combined and concentrated to 1.5 ml (6.2 mg ml⁻¹ protein as estimated at 280 nm using a molar extinction coefficient of 18 825 M ⁻¹ cm⁻¹; Gasteiger et al., 2003 ▸). Aliquots were flash-frozen in liquid nitrogen and stored at 193 K.

2.2. Crystallization and data collection  

Initial crystallization screens were set up using a Gryphon crystallization robot (Art Robbins Instruments, Sunnyvale, California, USA). Sitting-drop vapor-diffusion experiments were conducted using Intelli-Plates (Hampton Research, Aliso Viejo, California, USA) and were incubated at 292 K. Using commercially available crystallization screens from Hampton Research and Molecular Dimensions (Maumee, Ohio, USA), various protein:well solution ratios were set up (2:1, 1:1 and 1:2) at a protein concentration of 6.2 mg ml⁻¹ (in 25 mM Tris–HCl pH 7.4, 150 mM NaCl, 2 mM TCEP). Crystallization experiments were monitored using a Rock Imager automated imaging system (Formulatrix, Bedford, Massachusetts, USA). Crystals were obtained under many conditions, including condition A3 of the BCS Screen from Molecular Dimensions, which consisted of 0.1 M MES pH 6.5, 30%(v/v) PEG Smear Low (polyethylene glycol 400, polyethylene glycol methyl ether 550, polyethylene glycol 600 and polyethylene glycol 1000; Chaikuad et al., 2015 ▸). This condition was optimized by performing grid screens with various concentrations of PEG Smear Low (stock solution purchased from Molecular Dimensions) at varying pH values. In addition, additive screening was performed using Additive Screen from Hampton Research (Cudney et al., 1994 ▸). The conditions used to obtain crystals for data collection are reported in Table 2 ▸. Crystals appeared within two days. A single crystal was transferred to a drop consisting of well solution supplemented with 20%(v/v) ethylene glycol and allowed to soak for 1 min, and was then retrieved with a LithoLoop (Molecular Dimensions) and flash-cooled by rapidly plunging it into liquid nitrogen. X-ray diffraction data were collected remotely from a single crystal held at 100 K on SER-CAT beamline 22-BM at the Advanced Photon Source (Table 3 ▸). The data were indexed, integrated and scaled using HKL-3000 (Minor et al., 2006 ▸).

Table 2. Crystallization.

Method	Vapor diffusion, hanging drop
Plate type	EasyXtal 15-well plate, Qiagen
Temperature (K)	292
Protein concentration (mg ml−1)	6.2
Composition of reservoir solution	0.1 M MES pH 6.0, 20%(v/v) PEG Smear Low (polyethylene glycol 400, polyethylene glycol methyl ether 550, polyethylene glycol 600, polyethylene glycol 1000)
Additive solution	0.1 M spermidine
Volume and composition of drop	5 µl; 2.5 µl protein + 2.0 µl reservoir + 0.5 µl additive
Volume of reservoir (ml)	0.5

Table 3. Data collection and processing.

Values in parentheses are for the highest resolution shell.

Diffraction source	SER-CAT beamline 22-BM, Advanced Photon Source
Wavelength (Å)	1.0000
Temperature (K)	100
Detector	Rayonix MX300-HS
Crystal-to-detector distance (mm)	150
Rotation range per image (°)	1.0
Total rotation range (°)	180
Exposure per image (s)	3
Space group	P212121
a, b, c (Å)	92.2, 99.4, 128.7
α, β, γ (°)	90.0, 90.0, 90.0
Mosaicity (°)	0.7
Resolution range (Å)	50–2.17 (2.21–2.17)
Total No. of reflections	279773
No. of unique reflections	61347 (3056)
Completeness (%)	97.6 (99.1)
Multiplicity	4.6 (4.2)
Mean I/σ(I)	17.1 (2.1)
R merge	0.050 (0.519)
R p.i.m.	0.041 (0.343)
CC1/2	0.996 (0.773)
Overall B factor from Wilson plot (Å2)	29

2.3. Structure solution and refinement  

The structure of Y. pestis UGP was determined by molecular replacement using Phaser in the PHENIX suite (Zwart et al., 2008 ▸; McCoy et al., 2007 ▸). The coordinates of E. coli UGP (PDB entry 2e3d, chain A, 73% sequence identity; Thoden & Holden, 2007a ▸) with all nonprotein atoms removed were used as a search model to locate four molecules in the asymmetric unit (Matthews coefficient of 2.28 ų Da⁻¹, solvent content 46.2%; Matthews, 1968 ▸; Weichenberger & Rupp, 2014 ▸). The structure was refined without H atoms with phenix.refine and manually rebuilt with Coot (Afonine et al., 2012 ▸; Emsley et al., 2010 ▸). Water molecules were added and inspected using Coot and refined with phenix.refine. Model validation was performed using MolProbity (Williams et al., 2018 ▸). The coordinates and structure-factor files were deposited in the Protein Data Bank under accession code 6mnu. For structural comparisons, structural homologs were identified using the PDBeFold server (Krissinel & Henrick, 2004 ▸) and alignments were performed with PyMOL (v.2.0; Schrödinger).

3. Results and discussion  

3.1. Three-dimensional structure of Y. pestis UGP  

The structure of Y. pestis UGP was determined at 2.17 Å resolution by molecular replacement and was refined to an R _work of 0.177 and an R _free of 0.241 with good geometry. Data-processing and refinement/validation statistics are summarized in Tables 3 ▸ and 4 ▸, respectively. The enzyme crystallized as a tetramer formed by a dimer of dimers, similar to those observed for other bacterial UGPs [Fig. 1 ▸(a)] (Berbís et al., 2015 ▸; Thoden & Holden, 2007a ▸). Each chain in the tetramer is structurally very similar to the others, with aligned r.m.s.d values ranging from 0.2 to 0.4 Å. The structure of the monomeric unit consists of a core twisted-mixed β-sheet that is surrounded by α-helices [Fig. 1 ▸(b)]. Fig. 2 ▸ shows the quality of the fit of the model to the electron-density map for a representative region consisting of residues Tyr215–Leu224. Assembly and interface analysis using the PDBePISA server shows buried surface areas of 2080 Ų between chains A and B and 2220 Ų between chains C and D. Stabilizing salt bridges at the interface of chains A and B are formed by Asp33 (chain A)–Lys266 (chain B), Glu72 (chain A)–His57 (chain B), Lys266 (chain A)–Asp33 (chain B), His57 (chain A)–Glu72 (chain B) and Glu72 (chain A)–His57 (chain B). Similar stabilizing interactions are observed between chains C and D. The average buried surface interface between the dimer of dimers (AB–CD) is approximately 810 Ų.

Table 4. Structure solution and refinement.

Values in parentheses are for the highest resolution shell.

Resolution range (Å)	26.71–2.17 (2.23–2.17)
No. of reflections used	61296
Final R work	0.177 (0.227)
Final R free	0.241 (0.332)
No. of non-H atoms
Protein, total	8680
Protein, chain A	2175
Protein, chain B	2169
Protein, chain C	2182
Protein, chain D	2154
Water	623
Ethylene glycol	8
Estimated coordinate error (Å)	0.26
R.m.s. deviations
Bond lengths (Å)	0.008
Bond angles (°)	0.9
Average B factors (Å2)
Protein chain A	33
Protein chain B	34
Protein chain C	38
Protein chain D	38
All protein chains	36
Water	37
Ethylene glycol	37
Ramachandran plot
Favored (%)	96.2
Allowed (%)	3.4
Outliers (%)	0.4
PDB code	6mnu

Figure 1.

(a) Stereoview of a cartoon representation of the crystal structure of Y. pestis UGP solved at 2.17 Å resolution (PDB entry 6mnu) illustrating the tetrameric unit with individual chains colored as follows: chain A, green; chain B, cyan; chain C, magenta; chain D, yellow. (b) Stereoview of a cartoon representation of chain A with secondary-structure elements colored as follows: α-helices, cyan; loops, magenta; β-strands, red.

Figure 2.

Stereoview of a representative fit of residues Tyr215–Leu224 to the final 2F _o − F _c electron-density map (blue; 2.17 Å resolution, contoured at a 1.0 r.m.s.d. level, PDB entry 6mnu).

Closely related structural homologs identified by PDBeFold include the bacterial UGPs from E. coli (PDB entry 2e3d; r.m.s.d. of 0.7 Å over 280 aligned residues, 73% sequence identity), Erwinia amylovora (PDB entry 4d48; r.m.s.d. of 0.7 Å over 278 aligned residues, 72% sequence identity), Burkholderia ambifaria (PDB entry 5ve7; r.m.s.d. of 1.3 Å over 275 aligned residues, 52% sequence identity), Helicobacter pylori (PDB entry 3juk; r.m.s.d. of 1.7 Å over 244 aligned residues, 39% sequence identity) and several others (Thoden & Holden, 2007a ▸; Benini et al., 2017 ▸). Structural overlays of the homologs reveal similar three-dimensional structures with few variations. The main structural difference that stands out in the overlay is within the region of residues Asp179–Glu191 (in Y. pestis) between Y. pestis UGP and B. ambifaria UGP, in which the loop varies in conformation between the species [labeled ‘a’ in Fig. 3 ▸(a)]. Comparison with the UGP from H. pylori shows similar variations in the same loop region [labeled ‘a’ in Fig. 3 ▸(b)] but also structural deviations when comparing the conformation of residues Thr69–Thr98 of Y. pestis UGP with that of the analogous region in H. pylori UGP [labeled ‘b’ in Fig. 3 ▸(b)]. In Y. pestis the region adopts a conformation consisting of two α-helices separated by a short loop, whereas in H. pylori a disordered region is present, as indicated by a break between residues Tyr70 and Thr79. Additionally, in Y. pestis the C-terminus extends to residue Gln293 and adopts an α-helical secondary structure, whereas the ordered structure stops at residue Leu273 in H. pylori [labeled ‘c’ in Fig. 3 ▸(b)].

Figure 3.

(a) Stereoview of superimposed monomers of Y. pestis UGP (blue; PDB entry 6mnu, chain A, 2.17 Å resolution), E. coli UGP (cyan; PDB entry 2e3d, chain A, 1.95 Å resolution), E. amylovora UGP (orange; PDB entry 4d48, chain A, 2.46 Å resolution) and B. ambifaria (tan; PDB entry 5ve7, chain A, 2.3 Å resolution). (b) Stereoview of superimposed monomers of Y. pestis UGP (blue; PDB entry 6mnu, chain A) and H. pylori UGP (magenta; PDB entry 3juk, chain A, 2.3 Å resolution).

3.2. Comparison of active sites  

The general mechanism for UDP-glucose pyrophos­phorylase enzymes proceeds via an S_N2 nucleophilic attack of the phosphoryl O atom of glucose 1-phosphate on the α-phosphate P atom of UTP (Thoden & Holden, 2007b ▸). A magnesium ion is proposed to structurally stabilize the negative charge and to properly align the phosphoryl O atom of glucose 1-phosphate for efficient nucleophilic attack. This results in the release of the α- and γ-phosphates, which yields the enzyme–UDP-glucose complex. Fig. 4 ▸(a) depicts the structure of Y. pestis UGP with the modeled position of the product, UDP-glucose, based on the superimposed structures of Y. pestis UGP and that of H. pylori UGP complexed with UDP-glucose to highlight the location of the active site in the three-dimensional structure (Kim et al., 2010 ▸). The structural overlay indicates that the active-site residues involved in binding UDP-glucose are highly conserved between the two species and suggests that structural rearrangement of a few of the side chains of the Y. pestis active-site residues would occur to facilitate binding of the ligand [Fig. 4 ▸(b)]. In H. pylori UGP, the structure shows that the residues involved in the binding interactions include direct hydrogen bonds between the side chains of Lys25, Gln102, Asp130, Asp131, Glu190 and Lys191 and the bound UDP-glucose molecule. Additionally, a direct hydrogen bond is formed between the main-chain carbonyl O atom of Val203 and the O4′ atom of the glucose moiety and between the main-chain carbonyl of A10 and the O2 atom of the uridine moiety. These residues are conserved in Y. pestis, with the exception of Asp131, which instead is a hydrophobic residue, Val133 (Table 5 ▸). A similar difference is observed in the E. coli UGP homolog, which has Val138 in the equivalent position, as well as in E. amylovora (Val138). The side chain of Asp131 in H. pylori forms a hydrogen bond to the glucose O6′ atom of UDP-glucose, and the same interaction with UDP-glucose is observed in the structure of Corynebacterium glutamicum UGP with bound UDP-glucose (PDB entry 2pa4; Thoden & Holden, 2007b ▸). Although no crystal structure is yet available for the UGP from S. pneumoniae, an aspartic acid is also found in this position based on sequence alignments (Zavala et al., 2017 ▸). As a result of the variant amino acid at this position, this hydrogen-bonding interaction would be missing in Y. pestis and E. coli UGP. However, no crystal structure is available in the PDB of a UGP homolog with valine at this position bound to UDP-glucose to examine the structural significance of this amino-acid difference. We attempted to obtain a co-crystal structure of Y. pestis bound to UDP-glucose through co-crystallization and soaking experiments, but were unsuccessful in obtaining crystals that yielded a useful diffraction data set.

Figure 4.

(a) Stereoview of chain A of Y. pestis UGP illustrating the modeled position of UDP-glucose (blue spheres) based on superimposition of the coordinates onto the structure of chain A of H. pylori UGP complexed with UDP-glucose (PDB entry 3juk). (b) Stereoview of a structural comparison of the active-site residues of Y. pestis UGP (C atoms in gray, O atoms in red, N atoms in blue) superimposed onto the active-site residues of H. pylori (chain A, C atoms in cyan) bound to UGP-glucose (C atoms in purple, phosphate atoms in orange). (c) Stereoview of a structural comparison of Y. pestis UGP active-site residues superimposed onto the active-site residues of B. ambifaria UGP (C atoms in cyan) bound to UTP (C atoms in purple, phosphate atoms in orange) (PDB entry 5ve7, chain A).

Table 5. Comparison of structurally aligned active-site residues.

Amino-acid variations with respect to Y. pestis are shown in bold.

Y. pestis UGP (PDB entry 6mnu)	H. pylori UGP (PDB entry 3juk)	E. coli UGP (PDB entry 2e3d)	E. amylovora UGP (PDB entry 4d48)	B. ambifaria UGP (PDB entry 5ve7)	C. glutamicum UGP (PDB entry 2pa4)
Ala11	Ala10	Ala16	Ala16	Ala12	Ala20
Gly12	Gly11	Gly17	Gly17	Gly13	Gly21
Thr15	Thr14	Thr20	Thr20	Thr16	Thr24
Arg16	Arg15	Arg21	Arg21	Arg17	Arg25
Lys26	Lys25	Lys31	Lys31	Lys27	Lys35
Glu27	Glu26	Glu32	Glu32	Glu28	Glu36
Gln104	Gln102	Gln109	Gln109	Gln105	Gln112
Gly109	Gly107	Gly114	Gly114	Gly110	Gly117
Asp132	Asp130	Asp137	Asp141	Asp133	Asp142
Val133	Asp131	Val138	Val138	Asp134	Asp143
Gly176	Gly171	Gly179	Gly179	Gly172	Gly180
Glu198	Glu190	Glu201	Glu201	Glu191	Glu201
Lys199	Lys191	Lys202	Lys202	Lys192	Lys202
Val211	Val203	Val214	Val214	Val204	Ala214

Structural comparison of Y. pestis UGP with the crystal structure of B. ambifaria UGP complexed with UTP (PDB entry 5ve7; Seattle Structural Genomics Center for Infectious Disease, unpublished work) also reveals a highly conserved binding site for UTP between the homologs [Fig. 4 ▸(c)]. In B. ambifaria UGP, the active-site residues involved in direct hydrogen-bonding interactions with the bound UTP molecule include the side chains of Arg17 and Gln105, and additional hydrogen-bond interactions are formed between the backbone amide N atoms of Ala12, Gly13, Thr16, Arg17 and Gly110 and UTP. These residues are conserved between the two species and also in the other bacterial homologs presented in Table 5 ▸. Based on the structural comparisons and the high sequence conservation of the active site between the bacterial UGP homologs, potential exists for the structure-guided development of broad-spectrum inhibitors that target bacterial UGPs. Recent efforts to identify bacterial UGP inhibitors have yielded several candidates, and would be of significant value in attempts to obtain the first crystal structure of a bacterial UGP bound to an inhibitor (Zavala et al., 2017 ▸).

4. Conclusions  

In conclusion, we have determined the crystal structure of a bacterial UGP from Y. pestis that exhibits highly conserved structural homology with the crystal structures of other structurally characterized bacterial UGPs. An interesting feature of Y. pestis UGP is the identity of residue Val133, which is an aspartic acid in some other bacterial UGPs that has been shown to be involved in direct hydrogen-bonding interactions with the UDP-glucose molecule. Given the shared sequence conservation of the active site between bacterial UGP homologs, it may be possible to develop broad-spectrum inhibitors to target bacterial UGPs.

Supplementary Material

PDB reference: Y. pestis UDP-glucose phosphorylase, 6mnu

Funding Statement

This work was funded by Frederick National Laboratory for Cancer Research grant HHSN261200800001E. National Institutes of Health, National Cancer Institute grant .