Structures of glyceraldehyde 3‐phosphate dehydrogenase in Neisseria gonorrhoeae and Chlamydia trachomatis

Abstract

Neisseria gonorrhoeae (Ng) and Chlamydia trachomatis (Ct) are the most commonly reported sexually transmitted bacteria worldwide and usually present as co‐infections. Increasing resistance of Ng to currently recommended dual therapy of azithromycin and ceftriaxone presents therapeutic challenges for syndromic management of Ng‐Ct co‐infections. Development of a safe, effective, and inexpensive dual therapy for Ng‐Ct co‐infections is an effective strategy for the global control and prevention of these two most prevalent bacterial sexually transmitted infections. Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) is a validated drug target with two approved drugs for indications other than antibacterials. Nonetheless, any new drugs targeting GAPDH in Ng and Ct must be specific inhibitors of bacterial GAPDH that do not inhibit human GAPDH, and structural information of Ng and Ct GAPDH will aid in finding such selective inhibitors. Here, we report the X‐ray crystal structures of Ng and Ct GAPDH. Analysis of the structures demonstrates significant differences in amino acid residues in the active sites of human GAPDH from those of the two bacterial enzymes suggesting design of compounds to selectively inhibit Ng and Ct is possible. We also describe an efficient in vitro assay of recombinant GAPDH enzyme activity amenable to high‐throughput drug screening to aid in identifying inhibitory compounds and begin to address selectivity.

1. INTRODUCTION

Sexually transmitted infections (STIs) are among the most common causes of disease worldwide.1 Infections caused by two prominent gram‐negative bacteria, Neisseria gonorrhoeae (Ng) and Chlamydia trachomatis (Ct), can cause severe long‐term reproductive complications, such as pelvic inflammatory disease, endometritis, tubal inflammation, ectopic pregnancy, and infertility, and can increase the risk for HIV acquisition.2, 3, 4 Ng colonizes and infects the urogenital, anorectal, and oropharyngeal mucosal sites in both men and women. Men with gonorrhea typically exhibit symptomatic urethritis, while women are more likely to be asymptomatic with silent damage to their reproductive tract.5 The obligate intracellular bacterium Ct is one of the most prevalent human pathogens and is associated with both genital tract and ocular pathologies.6, 7 Ct has the highest incidence of infection among all reportable infectious diseases in the US, with almost two million new cases reported every year.8, 9 While both sexes can be infected with Ct, young women are disproportionately at risk for severe and chronic reproductive pathologies.2, 4

The rapid emergence of acquired antimicrobial resistant Ng strains has severely compromised the ability to treat gonorrhea. Resistance to nearly all previously and currently recommended antibiotics heightens the need for novel drugs that are effective against these so‐called “superbugs.”10, 11 Azithromycin, cephalosporin, and fluoroquinolone resistance in Ng have been reported worldwide.11, 12 Ceftriaxone, the last remaining first‐line antimicrobial treatment against Ng infections, is now under threat with the reported isolation of ceftriaxone‐resistant Ng strains.13, 14, 15 The repertoire of available drugs to treat resistant Ng is limited. Consequently, the development of new antimicrobials to address and treat these pathogens is an emerging international public health priority. Furthermore, because of the high rates of Ng and Ct co‐infection as well as the frequent syndromic management of STIs, the identification of new agents dually effective against Ng and Ct is desirable. New protein targets must be identified and characterized as avenues for future dually efficacious antimicrobial development.

Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) participates in numerous critical pathways in bacteria and eukaryotic cells. The best‐characterized role of GAPDH is in glycolysis, wherein it reduces nicotinamide adenine dinucleotide (NAD⁺ to NADH) by oxidizing glyceraldehyde‐3‐phosphate (GAP) to 1, 3 bisphosphoglycerate. Available evidence suggests that Neisseria GAPDH may influence intracellular processes of host cells to the advantage of the invading organism.16 Likewise, previous studies indicate that the expression of Chlamydia GAPDH as one of the glycolytic and pentose phosphate pathway enzymes is essential for Ct growth.17 On this basis, cell‐penetrating GAPDH inhibitors may affect the proliferation of Ng and Ct and could be a reasonable drug target.

In addition to being a reasonable drug target, GAPDH has been shown to be present on the surface of the bacterial cell membrane in both Gram‐positive and Gram‐negative bacteria.18 Indeed, surface‐exposed GAPDH has been proposed as a potential vaccine target for some bacteria. Surface‐expressed GAPDH (GapA‐1) is responsible for the adherence of Neisseria meningitidis (Nm) and other bacteria to epithelial and endothelial tissues.16 The lack of surface‐expressed GapA‐1 does not affect the growth of Nm in vitro, but an Nm GapA‐1 deficient mutant demonstrated a consequential reduction of adhesion, which suggests that GapA‐1 contributes to meningococcal pathogenesis. It is noteworthy that only surface GapA‐1 expression was reduced without deleting GapA‐2 levels.16 Ng also relies on adherence to host surfaces for subsequent tissue colonization and pathogenesis.19, 20 However, the specific role of NgGAPDH in the underlying molecular mechanisms of Ng adherence to host cells is unknown. Because GapA‐1 is well conserved across Neisseria species, it is conceivable that surface expressed Ng GapA‐1 behaves similarly to Nm GapA‐1 and promotes Ng adhesion.18 Thus, the structure of NgGAPDH could also help inform the molecular basis of GapA‐1 binding to mucosal surfaces or future vaccine development.

Collectively, these direct and indirect biological studies support the potential of GAPDH as a drug target for Ng and Ct. As a roadmap for the selection of compound libraries to identify new lead compounds that specifically target these enzymes instead of human GAPDH (HsGAPDH), we determined the structures of NgGAPDH and CtGAPDH by X‐ray crystallography. To identify potential new drug leads, we developed a fast, robust, and simple luminescence‐based assay for screening of inhibitor libraries against recombinant GAPDH enzymes.

2. RESULTS AND DISCUSSION

2.1. Crystal structures of NgGAPDH and CtGAPDH

We solved the crystal structures of NgGAPDH and CtGAPDH, bound to NAD⁺, to 2.5 and 2.4 Å resolution, respectively. NgGAPDH crystallizes with eight copies in the asymmetric unit, which form two tetramers. The model of the various chains starts between the N‐terminal His‐tag and I7 and extends through residue G332 to I334. While all residues could be built for most chains, for chains C and E the density was too weak to build residues in loops between K61 and V93. All eight copies are very similar with root mean square deviations (RMSDs) for Cα atoms of 0.25–0.45 Å. CtGAPDH crystallizes with four copies that form one tetramer in the asymmetric unit. The model covers part of the N‐terminal His‐tag through N342. The four subunits are also very similar, with RMSDs for Cα atoms below 0.3 Å.

The overall and quaternary structures are in accordance with other previously described structures for GAPDH (Figure 1).21, 22 In both structures, GAPDH crystallizes as a tetramer consisting of four subunits commonly designated as O, P, Q and R that bind one NAD⁺ molecule each and form three non‐equivalent interfaces along perpendicular axes P, Q, and R.23 As with other GAPDH enzymes, NgGAPDH and CtGAPDH exhibited mixed α/β topology in which a large β‐sheet of protomers O/P and Q/R form intermolecular interactions (Figure 1). Dimers O/P and Q/R interact via loop residues, and thus the homotetramer could more accurately be described as a dimer of homodimers (Figure 1). As expected, based on homologous structures, the catalytic residue C151 (NgGAPDH) or C150 (CtGAPDH) resides in proximity to the nicotinamide ring of the NAD⁺ cofactor.

Figure 1.

Quaternary architecture of NAD⁺‐bound of NgGAPDH (left) and CtGAPDH (right) with the ribbon representations of the O, P, Q, and R protomers colored in pink, gold, gray and blue, respectively. Dimers O/P and Q/R predominantly form via interactions of their β‐sheet, whereas loop residues are responsible for generation of the homotetramer (or dimer of dimers). The bound NAD⁺ molecules are depicted as ball‐and‐stick representations inside solvent‐excluded molecular surfaces

2.2. Structural comparisons of NgGAPDH, CtGAPDH, and HsGAPDH

Specific inhibition of NgGAPDH and CtGAPDH without affecting HsGAPDH is paramount since GAPDH is highly conserved and essential for human cell functions. NgGAPDH and CtGAPDH superimpose with RMSDs for Cα atoms of around 1.0 Å. Similarly, both NgGAPDH and CtGAPDH superimpose with HsGAPDH (PDB: 4WNC) with RMSDs for Cα atoms of around 1.0 Å. A structure‐based sequence alignment shows that NgGAPDH and CtGAPDH share 44 and 56% identity with HsGAPDH, respectively, whereas the Ng and Ct enzymes share 47% amino acid identity with each other (Figure 2).

Figure 2.

Structure‐based alignment of NgGAPDH, CtGAPDH, and HsGAPDH sequences. Secondary structure elements for NgGAPDH and HsGAPDH are shown above and below the alignment, respectively. Strictly conserved residues are boxed with a red background and conserved residues are in red font. The position of the F37 insertion site in HsGAPDH is highlighted in yellow. Residues causing the loss of the 3₁₀A helix in CtGAPDH are highlighted in green. Residues predicted to interact with GAPDH‐001 are boxed in green

Analysis of the 3D‐crystal structures of NgGAPDH and CtGAPDH compared to the HsGAPDH structure revealed key differences between the HsGAPDH and NgGAPDH active sites that may be used for specific structure‐based drug design (SBDD). The most promising difference for selectivity for Ng/CtGAPDH over HsGAPDH was previously described within the active site of other bacterial GAPDH structures.24 This is an L to PF insertion at position 36–37 in the HsGAPDH that changes the local backbone conformation (L35 in NgGAPDH and L34 in CtGAPDH) (Figure 3). This site appears quite promising for the development of NAD⁺‐competitive inhibitors of the bacterial enzymes, while providing selectivity over the human enzyme.24

Figure 3.

NAD⁺‐binding sites of NgGAPDH, CtGAPDH, and HsGAPDH (left to right). The side chains of residues forming hydrogen bonds (black lines in red circles) with NAD⁺ are shown, as well as sites causing backbone distortions in NG (L35) and CT (L34, V79) compared to human (F37 insertion)

Further analysis of the hydrogen bond network between the NAD⁺ cofactor and NgGAPDH, CtGAPDH, and HsGAPDH structures shows that a number of amino acid residues which directly interact with the substrate (NAD⁺) are conserved in all three enzymes, for example, D35, S122, and D316 using the numbering on HsGAPDH (Figure 3). However, a residue change from P to V at position 79 in the CtGAPDH structure causes the loss of the 3₁₀A helix secondary structure, preventing the formation of a hydrogen bond between the N6 amine group of the NAD⁺ adenosine moiety and the CtGAPDH backbone carbonyl group of residue 77 (R80 in HsGAPDH marked X; Figure 3b). This could be exploited with medicinal chemistry using suitable chemical functional groups for SBDD to drive selectivity for CtGAPDH. We also note that this change is proximal to the L/PF difference at position 36–37 (human numbering) described above, and that NAD⁺‐competitive inhibitors that bind here could also incorporate design features targeting the loss of the 3₁₀A helix in CtGAPDH.

2.3. GAPDH‐001 inhibition of NgGAPDH and CtGAPDH enzyme activity

An assay to determine NgGAPDH and CtGAPDH activity was developed using a light‐emission assay to detect NADH generated by the reaction (NAD(P)H‐Glo™ Detection System (Promega, Madison, WI). The GAPDH inhibitor, 2‐phenoxy‐1, 4‐naphthoquinone (GAPDH‐001), was used to validate the assay and its chemical synthesis was previously described.25 Inhibition of Trypanosoma brucei GAPDH (TbGAPDH) by GAPDH‐001, as defined by the IC₅₀ (the concentration at which enzymatic activity is reduced by 50%), was previously reported as 7.2 μM. GAPDH‐001 was reported as non‐competitive (mixed) with NAD⁺ in TbGAPDH.26, 27 We therefore determined the IC₅₀ values of GAPDH‐001 against NgGAPDH and CtGAPDH at four concentrations of NAD⁺. The IC₅₀ values against NgGAPDH measured as 12.4 μM, 7.4 μM, 5.8 μM, and 3.9 μM in the presence of 100 μM, 10 μM, 5 μM, and 1 μM NAD⁺ (Figure 4a, Table 1). The IC₅₀ values against CtGAPDH measured as 13.3 μM, 5.0 μM, 5.0 μM, and 3.6 μM in the presence of the same respective concentrations of NAD⁺ (Figure 4b, Table 1). The small increase in IC₅₀ with increased NAD⁺ suggests non‐competitive (mixed) affinity of NAD⁺ and GAPDH‐001 for GAPDH binding.

Figure 4.

Plots of GAPDH‐001 inhibition of recombinant NgGAPDH (left) and CtGAPDH (right) in the presence of varying concentration of NAD⁺. IC₅₀ value of GAPDH‐001 against NgGAPDH and CtGAPDH for each concentration of NAD⁺ was calculated using GraphPad prism 6 (GraphPad software, San Diego, CA)

Table 1.

NgGAPDH and CtGAPDH IC₅₀ ± SD values in the presence of varying concentrations of NAD⁺

IC50 (μM)	100 μM NAD+	10 μM NAD+	5 μM NAD+	1 μM NAD+
NgGAPDH	12.4 ± 2.0	7.4 ± 0.7	5.8	3.9 ± 2.5
CtGAPDH	13.3 ± 0.5	5.0 ± 1.1	5.0	3.6 ± 1.4

2.4. GAPDH‐001 docking into NgGAPDH, CtGAPDH and HsGAPDH

We attempted but were unable to obtain co‐crystallographic structures for NgGAPDH or CtGAPDH in the presence of inhibitor GAPDH‐001, likely due to weak affinity for the compound. Thus, GAPDH‐001 was docked into available TbGAPDH structures (PDB:2X0N chain R) with GlideXP as a proof of principle. Top scoring poses (GlideXP score = −4.7) were consistent with the binding mode previously described (Figure 5a).26 Based on this result, in‐silico docking experiments were performed on NgGAPDH, CtGAPDH, and HsGAPDH (Figure 5b–d, Table 2) using GlideXP.28 Top ranking poses were consistent with previously described results for TbGAPDH, where the catalytic cysteine side‐chain is in position for nucleophile attack of the thiolate to the quinone.26 The hypothesis that a covalent bond may be formed between the cysteine and GAPDH‐001 is not completely supported by our observed IC₅₀ values, which are slightly affected by dilution.26 Our observation that the IC₅₀ of GAPDH‐001 for NgGAPDH slightly increased as the NAD⁺ concentration increased (Table 1) would be more consistent with its previously reported binding mechanism.26 This is supported by our docking results, where the predicted inhibitor‐binding site overlaps with the pocket occupied by the nicotinamide group of NAD⁺ (Figure 5(e)). The GAPDH‐001 pocket further overlaps with site Ps, one of two phosphate‐binding sites of GAPDH that are highly conserved across species.28, 29 Accordingly, our docking results indicate that the putative binding mode of GAPDH‐001 to NgGAPDH, CtGAPDH, and HsGAPDH is conserved. It is noteworthy that GAPDH‐001 docking succeeded on the NAD⁺‐bound PDB structure for HsGAPDH and CtGAPDH, but for NgGAPDH it required optimization of the binding pocket by minimizing the protein structure with the ligand in place (RMSD over 333 superimposed Cα atom pairs to the crystal structure: 0.28 Å) prior to unconstrained docking. Initial docking failure could be explained by the limitations of rigid‐body docking, which does not account for backbone plasticity of the receptor. Indeed, the only difference in the putative GAPDH‐001 binding site between NgGAPDH compared to CtGAPDH and HsGAPDH is a T181‐G‐D loop in NgGAPDH that corresponds to T‐A‐T in CtGAPDH and HsGAPDH (Figure 2). In our docking experiments, the side‐chains of the T‐A‐T loop in CtGAPDH and HsGAPDH provide favorable, non‐specific hydrophobic interactions with the ligand. This specific interaction was lost in NgGAPDH. It is thus possible that such a dynamic makes the binding more labile in NgGAPDH than CtGAPDH and HsGAPDH, due to decreased interactions to lock the compound in place. In any case, this inhibitor is not predicted to give selectivity of inhibition of NgGAPDH and CtGAPDH compared to HsGAPDH. The search for selective inhibitors should focus on compounds that bind nearer the residues that differ between NgGAPDH/CtGAPDH and HsGAPDH.

Figure 5.

Putative‐binding mode of GAPDH‐001 to (a) TbGAPDH, (b) NgGAPDH, (c) CtGAPDH, and (d) HsGAPDH. The NgGAPDH T181‐G‐D loop is indicated with a green arrow. (e) The prediction from docking that GAPDH‐001 would bind the pocket occupied by the nicotinamide group of NAD⁺. Shown is CtGAPDH with NAD⁺ bound in the crystal structure conformation and GAPDH‐001 superimposed in green

Table 2.

Docking GlideXP score

Structure	NAD+	GAPDH‐001
NgGAPDH	−6.3	−4.5
CtGAPDH	−8.9	−5.1
HsGAPDH	−7.4	−5.0

3. CONCLUSIONS

Antimicrobial resistant STIs are a major and urgent public health concern. An important aspect of addressing this issue involves the identification of potential molecular targets for rational drug development programs. Based on information derived from crystal structures, the GAPDH enzymes from Ct and Ng could be successful drug targets. Comparing Ng druggable targets to homologous proteins in other major STI pathogens such as Ct using structural analysis and enzyme activity data are important first steps towards identifying effective new therapeutics. We have identified differences in amino acid residues within the active site of the bacterial enzymes relative to the human homolog that could be explored for selective inhibition. To complement structural studies, enzyme activity assays that provide reliable and easily observable readouts will be needed to directly measure the efficacy of various compounds and accelerate drug discovery. A non‐colorimetric assay technique was developed to measure enzyme activity of recombinant GAPDH. The assay relies on a direct measurement of a product of the GAPDH catalytic reaction, NADH, via a luminescence readout with NAD(P)H‐Glo™ (Promega). The average background to noise ratio was high (greater than 100‐fold), which allows for reliable analytical determination of enzyme activity and inhibition. The advantages of the new assay platform include easy readouts and high sensitivity since it requires low amounts of recombinant enzyme (nanomolar concentrations). New effective therapeutics that can be applied alone or in combination with currently used drugs are paramount to reducing public health issues related to multidrug resistant STI pathogens.

4. MATERIALS AND METHODS

4.1. Bioinformatics

HsGAPDH (UniProt: O14556) and orthologs from Ng (UniProt: B4RPP8) and Ct (UniProt: P0CE13) were identified by BLASTP. Comparisons of amino acid identities and alignments of their predicted amino acid sequences were generated using ESPript.30 Docking experiments were performed with Schroedinger release 2018‐04 using the Glide protocol. X‐ray coordinates for NgGAPDH, CtGAPDH, and HsGAPDH were retrieved from the Protein Data Bank (PDB 5VMT, 60K4, and 4WNC, respectively). Coordinates for GAPDH‐001 were downloaded from ChEMBL (ID 256155). Ligand structures were prepared with LigPrep using default parameters. Protein structures were pre‐processed and the hydrogen‐bonding network optimized using the default parameters of the Protein Preparation Wizard tool. Structures were refined by energy minimization with the OPLS3e force field until heavy atoms converged within 0.3 A RMSD from the starting coordinates and refined structures were manually inspected. Chains H, D, and C, respectively, of the NgGAPDH, CtGAPDH, and HsGAPDH prepared structures were chosen to generate the receptor grid. The grid was centered on the co‐crystalized ligand, with inner and outer box sized 10 A and 30 A, respectively. The Van der Waals radii of ligand atoms were scaled to 0.8 and the partial charge cutoff set to 0.15. Docking and scoring of poses were performed with Extra‐Precision Glide (GlideXP), including post‐docking minimization, with up to five poses saved per ligand for visual inspection. No explicit constraints to guide the ligand docking were added. For the docking of GAPDH‐001 into NgGAPDH, the ligand was placed in the CtGAPDH‐bound conformation prior to protein preparation and refinement. GlideXP scores for NAD binding to the native structures were obtained by setting the ligand sampling option to “refine only.”

4.2. Protein expression and purification

GAPDH from Ng was cloned, expressed, and purified by the Seattle Structural Genomics Center for Infectious Disease (SSGCID) using previously published protocols (Table 3).31 The full‐length enzyme (Uniprot: B4RPP8) encoding amino acids 1–334 was PCR‐amplified using a genomic DNA template from Ng. The full‐length GAPDH enzyme (Uniprot: P0CE13) from Ct was amplified in the same manner using genomic DNA template. The amplified products were extracted, purified and cloned into a ligation‐independent cloning (LIC) pET‐14b derived, N‐terminal His‐tag expression vector pBG1861 with a T7 promoter.32 The ligation products were then transformed into chemically competent Escherichia coli GC‐5 host cells for insert incorporation into the vector. Plasmid DNA was purified from the subsequent colonies and further transformed in chemically competent E. coli BL‐21(DE3) R3 Rosetta cells. Cells were grown and harvested for small‐scale expression testing, and soluble protein pellets were upscaled using 2 L cultures in an auto‐induction media in a LEX Bioreactor (Epiphyte3 Inc., Toronto, Ontario, Canada).33

Table 3.

Macromolecule production information

Source organism	Neisseria gonorrhoeae	Chlamydia trachomatis
DNA source	ATCC® 700825D‐5™	gDNA, strain D/UW‐3‐Cx
Forward primer	5′CTCACCACCACCACCACCATATGAGCATCAAAGTAGCGATTAACG‐3’	5′CTCACCACCACCACCACCATATGAG AATTGTGATTAATGGTTTTGG‐3′
Reverse primer	5′ATCCTATCTTACTCACTTAGATTTTGCCTGCGAAGTATTCCAA‐3′	5′ATCCTATCTTACTCACTTATTTAGAG TTTTCTTGTACGTACTCTA‐3′
Cloning vector	pBG1861	pBG1861
Expression vector	pBG1861	pBG1861
Complete amino acid sequence of the construct produced	MSIKVAINGFGRIGRLALRQIEKAHGIEVAAVNDLTPAEMLLHLFKYDSTQGRFQGTELKDDAIVVNGREIKVFANPNPEELPWGELGVDVVLECTGFFTNKTKAEAHIRAGARKVVISAPGGNDVKTVVYGVNQDILDGSETVISAASCTTNCLAPMAAVLQKEFGVVEGLMTTIHAYTGDQNTLDAPHRKGDLRRARAAALNIVPNSGAAKAIGLVIPELNGKLDGSAQRVPVATGSLTELVSVLERPATKEEINAAMKAASSESYGYNEDQIVSSDVVGIEYGSLFDATQTRVMTVGGKQLVKTVAWYDNEMSYTCQLVRTLEYFAGKI	MRIVINGFGR IGRLVLRQILKRNSPIEVV AINDLVAGDLLTYLFKYDSTHGSFAPQA TFSDGCLVMGERKVHFLAEKDVQKLPW KDLDVDVVVESTGLFVNRDDVAKHLDS GAKRVLITAPAKGDVPTFVMGVNHQQF DPADVIISNASCTTNCLAPLAKVLLDNF GIEEGLMTTVHAATATQSVVDGPSRKD WRGGRGAFQNIIPASTGAAKAVGLCLP ELKGKLTGMAFRVPVADVSVVDLTVKL SSATTYEAICEAVKHAANTSMKNIMYY TEEAVVSSDFIGCEYSSVFDAQAGVALN DRFFKLVAWYDNEIGYATRIVDLLEYV QENSK

The proteins were purified by a two‐step protocol involving immobilized metal affinity chromatography (IMAC) and size‐exclusion chromatography (SEC). All chromatography work was performed on an ÄKTAexplorer (GE Healthcare, Chicago, IL) using automated IMAC and SEC programs in adherence to prior established procedures.34 The final SEC was performed on a HiLoad 26/600 Superdex 75 (GE Healthcare) using an elution buffer of 500 mM NaCl, 25 mM HEPES, 5% glycerol, 0.025% azide, and 2 mM DTT pH 7.0. Fractions reflecting the protein peak were eluted as a single peak consistent with a monomeric enzyme. Fractions were pooled and analyzed for the soluble presence of the proteins of interest (MW: NG: 37 kDa, CT: 32 kDa) using SDS‐PAGE. The NgGAPDH fractions were concentrated to 32.15 mg/ml using an Amicon purification system (Millipore, Burlington, MA) for a total of 1.8 ml of concentrated enzyme in 200 μl aliquots. The CtGAPDH enzyme was concentrated to 17.29 mg/ml with a total of 1.2 ml concentrated enzyme. Enzymes were flash‐frozen in liquid nitrogen and stored at −80°C. The clones and purified proteins are available for order, free of charge, through the SSGCID website. https://apps.sbri.org/SSGCIDTargetStatus/Target/NegoA.00617.a, https://apps.sbri.org/SSGCIDTargetStatus/Target/ChtrB.00839.a

The protein batch number associated with the SSGCID identifier for the NgGAPDH is PS38018 and PW38497 for CtGAPDH.

4.3. GAPDH enzyme assay

Enzymatic activity of NgGAPDH and CtGAPDH was measured in 96‐well plates. The assay buffer consisted of 40 mM Tris–HCL pH 8.0, 2 mM EDTA, 1 mg/ml BSA, 10 mM K₂HPO₄, 10 mM MgCl₂, and 0.9 M KCl. The total reaction volume was 20 μl and contained 6 nM (NgGAPDH) or 10 nM (CtGAPDH) enzyme, 100 μM GAP, and 100 μM, 10 μM, 5 μM, or 1 μM of NAD⁺. The assay was validated for reproducibility and reliability against previously described GAPDH inhibitor GAPDH‐001 that was synthesized as described.25 GAPDH‐001 was tested for inhibition of the enzymes at three‐fold serial dilutions starting from 30 μM. The assay was incubated shaking at 37°C for 90 min. The activity of NgGAPDH and CtGAPDH in the presence of serial concentrations of GAPDH‐001 was measured by adding 20 μl of NAD(P)H‐Glo™ (Promega) to each well, incubating at room temperature for 10 min, and then reading with an Envision Plate Reader (Perkin Elmer, Waltham, MA). Dose response analysis was conducted with GraphPad Prism 6 (GraphPad Software, San Diego, CA). The assay was repeated on different days.

4.4. Crystallization

The enzymes were crystallized using SSGCID‐established protocols.31 Single crystals were obtained through vapor diffusion in sitting drops directly from the sparse matrix screens. For both proteins, NAD⁺ was added both to the crystallization set up and the cryo solution. Crystallization results and procedures are summarized below in Table 4.

Table 4.

Crystallization

Source organism	Neisseria gonorrhoeae	Chlamydia trachomatis
PDB ID	5VMT	6OK4
Method	Vapor diffusion, sitting drop	Vapor diffusion, sitting drop
Plate type	96 well, Compact 300, Rigaku	96 well, Compact 300, Rigaku
Temperature (K)	289.0	287.0
Protein concentration	21.2 mg/ml	17.29 mg/ml
Buffer composition of protein solution	25 mM HEPES pH 7.0, 500 mM NaCl, 5% glycerol, 2 mM DTT, 0.025% Azide + 3mM NAD	25 mM HEPES pH 7.0, 500 mM NaCl, 5% glycerol, 2 mM DTT, 0.025% Azide + 4mM NAD
Composition of reservoir solution	JCSG+ screen condition A9: 25% (w/v) PEG 3350, 0.2 M ammonium chloride	MCSG1 screen condition B9: 20% (w/v) PEG‐3350, 0.2 M magnesium chloride
Volume and ratio of drop	0.4 μl protein plus 0.4 μl reservoir	0.4 μl protein plus 0.4 μl reservoir
Volume of reservoir	80 μl microliter	80 μl microliter
Cryoprotectant	20% (v/v) ethylene glycol + 3mM NAD+	20% (v/v) ethylene glycol + 4 mM NAD+

4.5. Data collection and structure determination

Data collection and processing is detailed below in Table 5. All data was reduced with XDS/XSCALE.35 The diffraction images are available on the Integrated Resources for Reproducibility in Macromolecular Crystallography (http://proteindiffraction.org/) with doi: 10.18430M365VMT, doi: 10.18430M366OK4.36

Table 5.

Data collection and processing

Source organism	Neisseria gonorrhoeae	Chlamydia trachomatis
PDB ID	5VMT	6OK4
Ligand	NAD+	NAD+
Diffraction source	APS beamline 21‐ID‐F	APS beamline 21‐ID‐F
Wavelength (Å)	0.97872	0.97872
Temperature (K)	100	100
Detector	Rayonix MX‐300	Rayonix MX‐300
Crystal‐detector distance (mm)	320	300
Rotation range per image (°)	1	1
Total rotation range (°)	180	150
Exposure time per image (s)	1	1.5
Space group	C 2	P212121
a, b, c (Å)	142.45, 132.10, 156.00	86.47, 135.29, 137.67
α, β, γ (°)	90, 94.561, 90	90, 90, 90
Mosaicity	0.171	0.066
Resolution range (Å)	50.0‐2.5	50.0‐2.4
Total no. of reflections	378,801	398,110
No. of unique reflections	99,272	63,802
Completeness (%)	99.8	99.9
Redundancy	3.8	6.2
Overall B factor from Wilson plot (Å2)	41.89	37.01

4.6. Structure solution and refinement

Both structures were solved with molecular replacement using MoRDa. For NgGAPDH and CtGAPDH structures, 4QX6 (71% identity to NgGAPDH) and 3PYM (60% identity to CtGAPDH) were used as starting models, respectively.37 Iterative rounds of manual model building and automated refinement were carried out using Coot and Phenix.38, 39 The structures were quality checked with Molprobity.40 Structure refinement data is provided in Table 6.

Table 6.

Structure solution and refinement

Source organism	Neisseria gonorrhoeae	Chlamydia trachomatis
PDB ID	5VMT	6OK4
Resolution range (Å)	45.49‐2.50	49.68–2.40
No. of reflections, working set	97,261	55,983
No. of reflections, test set	1994	3566
Final R cryst	0.165	0.172
Final R free	0.219	0.208
No. of non‐H atoms
Protein	19,030	9931
Ligand	352	176
Water	646	430
R.m.s. deviations
Bonds (Å)	0.007	0.004
Angles (°)	0.936	0.624
Average B factors (Å2)
Protein	54.28	39.41
Ligand	52.75	46.09
Water	43.97	36.88
Ramachandran plot
Most favored (%)	96.95	96.31
Allowed (%)	3.01	3.39

Barrett KF, Dranow DM, Phan IQ, et al. Structures of glyceraldehyde 3‐phosphate dehydrogenase in Neisseria gonorrhoeae and Chlamydia trachomatis . Protein Science. 2020;29:768–778. 10.1002/pro.3824