Introduction

The GDP-D-glycero-α-D-manno-heptose is a key building block of lipopolysaccharide in mycobacteria and blocking of its biosynthetic pathway leads to high antibiotic susceptibility and reduced virulence of mycobacteria. The enzymes involved in GDP-D-α-D-heptose biosynthetic pathway offer an attractive target for novel antibiotics development. The GDP-D-α-D-heptose in mycobacteria is synthesized in four steps, (1) isomerization of D-sedoheptulose 7-phosphate into D-glycero-D-α-manno-heptose-7-phosphate by GmhA enzyme (2) Phosphorylation of D-glycero-D-α-manno-heptose-7-phosphate at C1 position by HddA enzyme, which forms D-glycero-D-α-manno-heptose-1,7-bisphosphate (3) Removal of phosphate at C7 position in D-glycero-D-α-manno-heptose-1,7-bisphosphate by GmhB enzyme, which leads to D-glycero-D-α-manno-heptose-1-phosphate and (4) modification of the phosphate at C1 position to form a phosphodiester linkage with GDP by HddC enzyme and leads to GDP-D-glycero-α-D-manno-heptose (Fig. 1A). GDP-D-α-D-heptose is incorporated in the S-layer glycoproteins of mycobacterial membrane by specific precursor.

Figure 1
figure 1

(A) Schematic diagram showing various enzymes involved in M. tuberculosis GDP-D-α-D-heptose biosynthetic pathway. (B) The gene construct used for MtbGmhA enzyme expression. SIS domain represents the sugar isomerase domain. (C) The elution profile of wild type and six mutants of MtbGmhA enzymes obtained from Superdex 200 column. Inset shows the SDS-PAGE of wild type enzyme; M = Mw marker, 1 = MtbGmhA. The molecular mass of MtbGmhA was calculated from standard curve generated using elution profiles of known molecular mass proteins.

Full size image

The heptose sugar maintains the structural integrity of the outer membrane of bacteria and interacts with membrane proteins and divalent cations. The heptose sugar is observed in inner core of lipopolysaccharide of E. coli and P. aeruginosa1. The heptose moiety is involved in ionic interactions and provides barrier to the passage of detergent, dyes and antibiotics through bacterial cell wall2 and is essential for survival of P. aeruginosa3.

The Gram-positive bacterial species contain crystalline two dimensional protein arrays, known as S- layers4. In S-layers, protein clusters are covalently attached to glycan chains, made up of 20–50 identical repeating units of α-L-rhamnose and D-glycero-α-D-manno-heptose units. Glycosylation of bacterial surface proteins with heptose and rhamnose are significant and involved in many functions e. g. adherence5, evasion of host immune response and enhanced resistance to proteolytic attack.

The GDP-D-α-D-heptose biosynthetic pathway in M. tuberculosis was first discovered by Eidels and Osborn6. The M. tuberculosis is a Gram-positive organism, however exhibits both features of Gram-positive and Gram-negative bacteria7. The GmhA is the key enzyme and highly conserved in Gram-positive and Gram-negative organisms e. g. A. thermoaerophilus and C. acetobutylicum. The MtbGmhA showed 26.2% sequence identity with C-terminal region of L-glutamine-D-fructose-6-phosphateamido transferase, a ketose/aldose isomerase consists of sugar isomerase domain8. The D-glycero-α-D-manno-heptose is also observed as cellular component of A. thermoaerophilus DSM 10155, a member of the bacillus/clostridium group of Gram-positive organism9.

Crystal structures of GmhA enzymes from eight Gram-negative species have been determined10,11,12,13. The GmhA enzymes from P. aeruginosa and V. cholerae were crystallized as dimer, while other GmhA orthologues were crystallized as tetramer in the asymmetric unit. The GmhA tetramer is found in two distinct conformations, “closed” and “open”. The “open” conformation is characterized by an extended α3–β2 loop, less ordered α3′ region, and loosely packed dimer-dimer interface. In “closed” conformation, unstructured α3′ region adopts a helical structure and α3–β2 loop is positioned inward, which allowed more extensive dimer-dimer interaction.

In current study, we have performed biochemical and structural analysis of MtbGmhA enzyme. The MtbGmhA enzyme was purified and its activity was analyzed by coupled assay involving MtbGmhA, MtbHddA and MtbGmhB enzymes. The binding analysis between all three enzymes was performed using surface plasmon resonance technique. The MtbGmhA residues involved in D-sedoheptulose 7-phosphate and Zn2+ binding were identified using modeled complex and their roles in catalysis were determined using site directed mutagenesis. The circular dichroism, small angle X-ray scattering, and dynamics simulation techniques were used to analyze the MtbGmhA structure and dynamics involved in D-sedoheptulose 7-phosphate and Zn2+ binding. Our structural and biochemical analysis on MtbGmhA explained the mechanism of action, which will contribute in the development of novel antibiotics against M. tuberculosis.

Results and discussion

Purified MtbGmhA forms tetramer in solution

The Fig. 1A showed the various steps involved in GDP-D-α–D-heptose biosynthetic pathway in M. tuberculosis. The MtbGmhA enzyme (196 residues, Mw ~ 24 kDa) is the first enzyme involved in D-sedoheptulose 7-phosphate isomerization and converts it to D-glycero-D-α-manno-heptose-7-phosphate. The MtbGmhA gene was cloned in pET28a expression vector (Fig. 1B) and overexpressed in soluble fraction of Escherichia coli. The MtbGmhA enzyme was purified and eluted as tetramer (~ 100 kDa) from Superdex 200 (16/60) column, identified based on molecular mass standard (Fig. 1C, inset). The purified MtbGmhA showed more than 98% purity on SDS-PAGE (Fig. 1C, inset). The SDS-PAGE profile of 9 eluted fractions of MtbGmhA (E1–E9) are shown in Fig. S1. We have generated the six-point mutants e. g. N53G, S121G, T122G, S123G, S126G, Q173G of MtbGmhA enzyme and all mutant proteins were purified like wild type enzyme and eluted as tetramer (Fig. 1C). The primers used for gene amplification of six MtbGmhA mutants are shown in Table S1. PDBePISA server analysis also indicated that biological assembly of MtbGmhA enzyme is a tetramer12.

The GmhA enzymes from P. aeruginosa and V. cholera form tetramer in solution10,14. However, their crystallographic asymmetric unit contains dimer and functional tetramers were obtained using two-fold crystallographic symmetry. The H. pylori GmhA is the only homologue, which was observed as homodimer in solution15,16.

Small angle X-ray scattering analysis revealed the tetrameric envelope of MtbGmhA enzyme

The small angle X-ray scattering data was collected on MtbGmhA enzyme to analyze its low-solution shape in solution. The SAXS intensity profile [I(Q)] as a function of momentum transfer vector Q, Guinier plot using globular approximation, Kratky plot and pair-distribution function [P(R)] were calculated (Fig. 2A–C). The Fig. 2A showed the double logarithmic plot of small angle X-ray scattering data (log10 I(Q) ~ log10Q). These data showed no inter particulate effect or aggregation of protein, as lack of upwards or downwards points were observed, when data approached to 0 nm−1. The Fig. 2B showed the bell-shaped peak in Kratky plot, which indicated the globular shape of MtbGmhA in solution. The Guinier plot analysis yielded the radius of gyration RG ~ 2.38 ± 0.11 nm, quite similar to theoretical RG of MtbGmhA tetramer. The Fig. 2C showed the probability distribution of various interatomic vectors P(R), using small angle X-ray scattering data as a reference. These data yielded the maximum linear dimension (Dmax) ~ 6.35 nm and RG ~ 2.45 nm for MtbGmhA tetramer (Fig. 2C). The single peak in P(R) profile indicated the globular shape of MtbGmhA enzyme. Using Lysozyme data (conc. ~ 1 mg/ml), the intensity at zero angle (I0) ~ 464 a. u. was observed for 1.0 kDa in 1 h exposure. Using it as control, the relative intensity of MtbGmhA protein (conc. ~ 1.2 mg/ml) has yielded the molecular mass ~ 93.2 kDa, which corresponds to MtbGmhA tetramer (Table 1).

Figure 2
figure 2

Small angle X-ray scattering analysis on MtbGmhA enzyme. (A) SAXS intensity profile of MtbGmhA. The inset shows linear fit to Guinier region of measured data. (B) Kratky plot of MtbGmhA. (C) The P(R) curves computed for MtbGmhA indicating the frequency distribution of interatomic vectors in predominant scattering species. (D–F) Fitting of MtbGmhA tetramer (closed state) in shape restored from dummy atom modeling using small angle X-ray scattering data and shown after every 90° rotation along Y-axis. Fitting of MtbGmhA tetramer (open state) in shape restored from dummy atom modeling using small angle X-ray scattering data and shown after every 90° rotation.

Full size image
Table 1 Small angle X-ray scattering data collection and experimentally derived parameters of MtbGmhA. Rg = radius of gyration. Dmax = maximum particle dimension.
Full size table

Uniform density modeling protocol was used to restore the shape, which yielded 10 independents MtbGmhA models. All models were averaged, refined and yielded the normalized spatial disposition (NSD) ~ 1.30 ± 0.03, a measure of similarity between individual solutions. The NSD showed the high similarity between dummy residue shape of solved models. The modeled MtbGmhA tetramer in closed state fitted well into SAXS envelope (χ2 ~ 1.17) (Fig. 2D–F), compared to open state (χ2 ~ 3.29) (Fig. 2G–I). The CRYSOL program was used to compute the theoretical small angle X-ray scattering profile from MtbGmhA tetrameric model and compared with experimental small angle X-ray scattering data. These data indicated that MtbGmhA exists as monodisperse, homogeneous and closed state tetramer in solution.

Activity assay using wild type and mutant MtbGmhA enzymes showed the roles of various active site residues

Activity assay using wild type enzyme

For activity analysis, the MtbHddA and MtbGmhB enzymes were purified (data not shown), performed the coupling assay using all three enzymes and monitored the release of phosphate. As control, a phosphate standard curve was calculated using 0–50 μM phosphate, dissolved in coupling reaction buffer having no MtbGmhA enzyme. Following kinetic parameters, Km ~ 0.31 ± 0.06 mM−1, kcat ~ 0.45 ± 0.02 s−1 and catalytic efficiency (kcat/Km) ~ 1.45 mM−1 s1 were obtained for MtbGmhA enzyme (Fig. 3A, Table 2).

Figure 3
figure 3

Activity assay on wild type and six mutants of MtbGmhA enzymes. Three replicates were performed for each measurement. The MtbGmhA activity was determined using Malachite green phosphate detection kit. A coupling reaction was performed using MtbGmhA, MtbHddA and MtbGmhB enzymes and release of inorganic phosphate was monitored at 600 nm. Following saturation curves were obtained, (A) wild type MtbGmhA (B) N53G mutant (C) S121G mutant (D) T122A mutant (E) S123A mutant (F) S126A mutant and (G) Q173A mutant. The 0 to 3 mM concentrations of D-sedoheptulose 7-phosphate were used in all assays and results are given in Table 2. (H) Proposed mechanism of MtbGmhA used in isomerization of D-sedoheptulose 7-phosphate to D-glycero-D-α-manno-heptose-7-phosphate through enediol intermediate, in which Gln173 acted as a catalytic acid and forms an imine stabilized by Zn2+ ion and Glu65 acted as catalytic base.

Full size image
Table 2 Kinetic parameters of wild type and six MtbGmhA mutants obtained using Michaelis Menten curve and Lineweaver–Burk plot (in red bracket). The kinetic parameters of E. coli. GmhA, B. pseudomallei GmhA and H. pyroli GmhA are shown for comparative analysis.
Full size table

The MtbGmhA kinetic parameters were compared with other GmhA orthologues, as shown in Table 2. The catalytic efficiency of MtbGmhA enzyme was quite similar to B. pseudomallei and E. coli GmhA enzymes. However, H. pylori GmhA enzyme showed ~ 50-fold higher catalytic efficiency (kcat/Km) than MtbGmhA enzyme and other GmhA orthologues. The H. pylori GmhA was observed as dimer in solution, while tetramers were observed for GmhA enzymes from M. tuberculosis, E. coli and B. pseudomallei. It is not clear how oligomeric state of GmhA enzyme affects its catalytic efficiency.

Activity assay using six MtbGmhA mutants

We build the MtbGmhA + S7P + Zn2+ complex and identified the six residues involved in D-sedoheptulose 7-phosphate binding. we have generated the six MtbGmhA mutants, performed coupling assay and release of free Pi was monitored (Fig. 3B–G, Table 2).

As seen in Table 2, Gln173 → Ala mutation leads to ~ 24 fold decrease in catalytic efficiency and ~ fivefold decrease in binding affinity to D-sedoheptulose 7-phosphate substrate in MtbGmhA, when compared to wild enzyme. Following data e. g. Asn53 → Gly mutation (~ twofold decrease in binding affinity and ~ 3.5 fold decrease in catalytic efficiency), Ser121 → Gly mutation (~ twofold decrease in binding affinity and ~ 3.5 fold decrease in catalytic efficiency), Thr122 → Ala mutation (~ threefold decrease in binding affinity and ~ 14.5 fold decrease in catalytic efficiency), Ser123 → Ala mutation (~ threefold decrease in binding affinity and ~ 4.5 fold decrease in catalytic efficiency), Ser126 → Ala mutation (~ threefold decrease in binding affinity and ~ 4.3 fold decrease in catalytic efficiency) were observed. The Asn53 forms bifurcated hydrogen bonds with O4 and O9 atoms of D-sedoheptulose 7-phosphate and Asn53 → Gly mutation leads to ~ 3.5 fold decrease in catalytic efficiency. The Gln173 forms hydrogen bond with O2 atom of D-sedoheptulose 7-phosphate and its mutation to Ala leads to ~ 24 fold decrease in catalytic efficiency. The bar diagram showing the catalytic efficiency of wild type and six MtbGmhA mutants are shown in Fig. S2.

In Escherichia coli and P. aeruginosa GmhA structures10, the Glu68 and His183 acted as acid and base in enzyme catalysis, which converted the D-sedoheptulose 7-phosphate to D-glycero-D-α-manno-heptose-7-phosphate. In B. pseudomallei GmhA structure12, His183 was buried behind Zn2+ in the active site and Glu68 and Gln175 bind to Zn2+ in such a way, that their side chains acted as acid and base in enzyme catalysis.

In modeled MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer, the Glu66 from one monomer and Gln173 from another monomer bind to D-sedoheptulose 7-phosphate and Zn2+ and may act as acid and base in enzyme catalysis (Fig. 3H). Mutation of Gln173 → Ala has shown ~ 50 fold decrease in catalytic efficiency of MtbGmhA enzyme, indicating its involvement in enzyme catalysis. Similar mechanism has been observed in other GmhA orthologues, which contain Zn2+ in the active site.

MtbGmhA, MtbHddA and MtbGmhB enzymes interact each other in µM range

Surface plasmon resonance technique was used to analyze the binding affinity between MtbGmhA, MtbHddA and MtbGmhB enzymes. Following KD values were observed e. g. ~ 8.4 ± 0.5 µM between MtbGmhA and MtbHddA (Fig. 4A), 20.0 ± 0.9 µM between MtbHddA and MtbGmhB (Fig. 4B) and 4.9 ± 0.1 µM between MtbGmhA and MtbGmhB (Fig. 4C). These data indicated that all three enzymes bind each other in µM range and form a stable complex in GDP-D-α-D-heptose biosynthetic pathway.

Figure 4
figure 4

Interaction analysis between MtbGmhA, MtbGmhB and MtbHddA enzymes using surface plasmon resonance technique. (A) Sensogram showing the binding of MtbHddA on immobilized MtbGmhA. Five different concentrations of MtbHddA enzyme (1–16 µM) were used in this assay, which yielded the KD = 8.45 ± 0.45 µM. (B) Sensogram showing binding of MtbHddA on immobilized MtbGmhB. Five different concentrations of MtbGmhB (1–16 µM) were used in this assay, which yielded the KD = 20 ± 0.95 µM. (C) Sensogram showing the binding of MtbGmhB on immobilized MtbGmhA. Five different concentrations of MtbGmhB (4 to 20 µM) were used in current assay, which yielded KD = 4.9 ± 0.11 µM. (D) Psi-Pred program analysis on MtbGmhA sequence showing the secondary structural contents in MtbGmhA enzyme. (F) The circular dichroism data of wild type and six MtbGmhA mutants, shown in different colors. (G) Thermal denaturation profile on MtbGmhA enzyme, indicating the melting temperature ~ 53.5 °C of the enzyme.

Full size image

The genetic organization in D-α-D-heptose biosynthetic pathway was first described in A. thermoaerophilus DSM 10155, in which all enzymes were located as cluster containing genes involved in synthesis and transfer of dTDP-rhamnose9. In M. tuberculosis, the genes for two different pathways were present at different locations and involved in synthesis of glycolipids and protein glycosylation. Interestingly, in CDC 1551 strain of M. tuberculosis, the GmhA and GmhB genes were fused as single gene, probably encoding a bifunctional enzyme having isomerase and phosphatase activities.

Circular dichroism analysis revealed the secondary structures and thermal stability of MtbGmhA enzyme

The PSIPRED program analysis on MtbGmhA sequence showed the secondary structures of enzyme (Fig. 4D). The Far-UV CD data on MtbGmhA and its six mutants were collected in 260–200 nm range (Fig. 4E). The K2D program has yielded the ~ 46% α-helix, ~ 19% β-sheet and ~ 35% random coil structures of wild type MtbGmhA, quite similar to structures observed in six MtbGmhA mutants (Fig. 4E, Table 3). Various secondary structure prediction programs have yielded quite similar secondary structures in MtbGmhA enzyme, as observed in CD analysis (Table S2).

Table 3 Secondary structural contents in six mutants of MtbGmhA obtained using circular dichroism spectroscopy.
Full size table

For thermal stability analysis, the mean residue ellipticity (θ222) data was collected on MtbGmhA enzyme in 25–90 °C range with 10 °C step (Fig. 4F). The helical structure of MtbGmhA enzyme was quite stable till 42 °C and disordered at 65 °C. A melting temperature, Tm ~ 53.5 °C was obtained, which indicated the high thermostability of MtbGmhA enzyme.

The modeled MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer showed the active site involved in D-sedoheptulose 7-phosphate and Zn2+ binding

The MtbGmhA model (1–196 residues) was obtained using I-TASSER (Iterative Threading ASSEmbly Refinement) server, which used the PDB-2X3Y (Crystal structure of B. pseudomallei GmhA tetramer in closed state12) as the best input template (RMSD = 0.79, id1 = 0.39, id2 = 0.39, Conv = 0.98, Z-score = 3.2). The I-TASSER server also used the PDB-2I2W (Crystal structure of Escherichia Coli phosphoheptose isomerase in open state10) as another template for MtbGmhA modeling. However, current template did not yield suitable MtbGmhA model, as having following parameters (RMSD = 2.1, id1 = 0.35, TM-score = 0.82, Conv = 0.89). The D-sedoheptulose 7-phosphate and Zn2+ ions were docked into MtbGmhA tetramer using GLIDE module of Schroedinger program (Fig. 5A). A docking score of -5.44 and X-score of 8.1 were observed in docking analysis. The P. seudomallei GmhA structures (PDB-2X3Y and PDB-2XBL) were used as reference to validate the results obtained in docking analysis of both ligands.

Figure 5
figure 5

(A) The MtbGmhA + D-sedoheptulose 7-phosphate (S7P) + Zn2+ complex model showing the α-helices (in cyan), β-sheets (in magenta), loops (in orange). The D-sedoheptulose 7-phosphate and Zn2+ ion are shown in red colors. (B) Electrostatic surface diagram of MtbGmhA monomer showing the D-sedoheptulose 7-phosphate and Zn2+ ion in the active site (blue positive surface). (C) LigPlot analysis of MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ complex showing the hydrogen bonds and van der Waals interactions between D-sedoheptulose 7-phosphate and Zn2+ with active site residues of MtbGmhA enzyme. The green dashed lines showed the hydrogen bonds and residues in “arc with spikes” are involved in van der waals interactions with ligands. (D) MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer showing each monomer in different color and D-sedoheptulose 7-phosphate and Zn2+ ligands in the active site of each monomer. (E) Electrostatic surface diagram of MtbGmhA tetramer showing four catalytic clefts (blue, positive charge), which accommodate the D-sedoheptulose 7-phosphate and Zn2+ in their binding pockets. (F) Interface between two MtbGmhA monomers, in which Glu66 residue (cyan) from one monomer and Zn2+ and Gln173 residue from another monomer binds to O1-C1 and O2-C2 of D-sedoheptulose 7-phosphate respectively and proposed as involved in isomerization of D-sedoheptulose 7-phosphate to D-glycero-D-α-manno-heptose-7-phosphate.

Full size image

The MtbGmhA monomer consists of central five stranded parallel β-sheets, flanked by eight α-helices and forms a helix-beta-helix sandwich (Fig. 5A). Overall fold of MtbGmhA enzyme was quite similar to flavodoxin-type nucleotide-binding motif, as observed in other GmhA orthologues. The Electrostatic surface of MtbGmhA monomer (Fig. 5B) showed the positively charged surface at catalytic cleft, which accommodates the D-sedoheptulose 7-phosphate and Zn2+ ion. The LIGPLOT v.4.5.3 analysis of MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ complex (Fig. 5C) showed that D-sedoheptulose 7-phosphate forms hydrogen bonds with Asn53 and Gly55 residues of β1–α3 helix region, Ser121, Thr122, Ser123 and Ser126 residues from β3–α6 helix region, Gln173 residue from α8 helix in MtbGmhA enzyme. The Zn2+ ion forms hydrogen bond with Gln173 of MtbGmhA, which stabilize the D-sedoheptulose 7-phosphate in the active site. The C. jejuni and P. aeruginosa GmhA enzymes also adopt the closed structures14 and metal binding sites were observed in both enzymes. It appears that Zn2+ binding induces the closed state conformation of GmhA and more suitable for D-sedoheptulose 7-phosphate binding and catalysis.

Since MtbGmhA is observed as tetramer in solution, we build the MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer using the same PDB-2X3Y as input template (Fig. 5D). The MtbGmhA tetramer forms a compact structure, in which four α-helices form hexagons, tilted at 120° and may help in structural stabilization. The α4 and α5 helices of each MtbGmhA monomer facing each other and forms a compact groove. 50% of α1 helix of one monomer interacts with α1 helix of the neighboring monomer. The Electrostatic surface of the MtbGmhA tetramer (Fig. 5E) showed the positively charged surface at catalytic pockets involved in D-sedoheptulose 7-phosphate and Zn2+ ion binding. Overall surface of MtbGmhA tetramer was negatively charged, except D-sedoheptulose 7-phosphate and Zn2+ binding pockets. The loop regions connecting β1–α3 and β3–α6 strands were involved in formation of D-sedoheptulose 7-phosphate binding pocket in MtbGmhA enzyme. The Fig. 5F showed the Glu66 residue of one monomer (cyan) bind to O1-C1 of D-sedoheptulose 7-phosphate and Gln173 residue of another monomer (Pink) involved in binding to O2–C2 of D-sedoheptulose 7-phosphate and Zn2+ and involved in enzyme catalysis.

Differences between MtbGmhA and other known GmhA enzymes

Sequence alignment and active site analysis

The Fig. 6A showed the sequence alignment of MtbGmhA enzyme with 11 GmhA orthologues. The secondary structures of MtbGmhA model were placed on the top of sequence alignment. The Asn53, Gly55, Ser121, Thr122, Ser123, Ser126 residues (*) involved in D-sedoheptulose 7-phosphate and Zn2+ binding were found quite conserved. The active site residues e. g. Gln173 from one monomer and Glu66 from another monomer (shown as #), which bind to C1-O1 and C2-O2 of D-sedoheptulose 7-phosphate were also quite conserved. Other MtbGmhA residues involved i in substrate binding e. g. Asn53 (fully conserved), Gly55 (G → T in 2YVA, G → S in 3TRJ), Ser121 (S → T in 3TRJ), Thr122 (fully conserved), Ser123 (S → R in 2YVA), Ser126 (fully conserved ), Gln173 (fully conserved) and Glu66 (E → S in 2YVA, E → Q in 5BY2 and E → K in 3TRJ) have shown degree of conservation with 11 other GmhA orthologues. In Francisella tularensis, Ser121 was replaced with Thr121, however not involved in catalytic cleft formation15,16. Similarly, Ser123 → Arg mutation in DiaA helps in timely initiation of chromosomal replication during cell cycle. In DiaA structure, important residues involved in enzyme catalysis were Ser52, Ala53, Arg71, Pro72, Asn83, Lys101, Leu190, and Phe19117.

Figure 6
figure 6

(A) Multiple sequence alignment of MtbGmhA with eleven GmhA orthologues, which showed highly conserved (red shade), semi-conserved (red letter) and dispersive residues (black letter). The residues involved in D-sedoheptulose 7-phosphate binding (*), in Zn2+ binding and enzyme catalysis (#) are shown above the sequence alignment. (B) Structural superposition of 11 GmhA orthologues on MtbGmhA model. These structures are 1TK9, 2X3Y, 5LU7, 1X92, 3BJZ, 5BY2, 5IOL, 3TRJ, 1X94, 1I2W. The 2YVA is not a MtbGmhA orthologue, but a homologue with different function. Major conformational changes are highlighted with dashed circles.

Full size image

In addition, other MtbGmhA residues (Gly52, Gly54, Ala60, His62, Arg75, Asn95, Asp115, T116 (T → V in 11 enzymes), L117, G124, A132, A136, G148, Gly151, Gly152, Asp159, Pro165, Ile172, Glu174, His176) were fully conserved in sequence alignment with 11 GmhA orthologues.

Structural superposition

Structural superposition of 11 GmhA orthologues on MtbGmhA model have yielded following r.m.s. deviation e. g. 1TK9 (Crystal structures of two phosphate isomerases from C. jejuni)14 (0.65 Å, 146 Cα atoms), 2X3Y (Crystal structure of GmhA from B. pseudomallei)12 (0.27 Å, 168 Cα atoms), 5LU7 (Crystal structure of heptose isomerase GmhA mutants)18 (0.15 Å, 168 Cα atoms), 1X92 (Crystal structure of Pseudomonas Aeruginosa phosphoheptose isomerase in complex with reaction product D-glycero-D-mannopyranose-7-Phosphate)10 (0.56 Å, 148 Cα atoms), 3BJZ10 (0.75 Å, 133 Cα atoms), 2YVA17 (Crystal structure of E. coli DiaA) (0.76 Å, 147 Cα atoms), 5BY211 (Sedoheptulose 7-phosphate isomerase from Colwellia psychrerythraea strain 34H) (0.63 Å, 147 Cα atoms), 5IOL (Crystal structure of Nucleoside Diphosphate kinase from Schistosoma mansoni)19 (0.53 Å, 147 Cα atoms), 3TRJ (Crystal structure of phosphoheptose isomerase from Francisella tularensis)15,16, 1X94 (Crystal structure of two putative phosphoheptose isomerase from Vibrio cholerae)14 (0.64 Å, 116 Cα atoms), 1I2W (Crystal structure of Bacillus licheniformis BS3 class-A beta-lactamase and acyl-enzyme adduct formed with cefoxtin)20 (0.69 Å, 128 Cα atoms) (Fig. 6B). Major changes were observed in the loop regions connecting α3–β2, α8–β5, β5–α6 strands and minor changes in other loops of MtbGmhA enzyme. The β1–α3 loop is involved in D-sedoheptulose 7-phosphate and Zn2+ binding and found quite conserved in 11 GmhA orthologues and showed least conformation change. Major conformational changes were observed in α4 and α7 helices of E. coli GmhA enzyme (2I2W, open state) to MtbGmhA, when compared to structures of 10 GmhA orthologues.

Dynamic simulation on MtbGmhA tetramer as (1) Apo (2) D-sedoheptulose 7-phosphate bound and (3) D-sedoheptulose 7-phosphate + Zn2+ bound state has revealed the dynamics involved in ligand recognition.

1 ns dynamics simulation was performed on MtbGmhA tetramer in (i) Apo (ii) D-sedoheptulose 7-phosphate bound and (iii) D-sedoheptulose 7-phosphate + Zn2+ bound state and analyzed the dynamics involved in D-sedoheptulose 7-phosphate and Zn2+ binding (Table 4). All three MtbGmhA tetramers obtained after dynamics simulation showed good stereochemistry and lie in allowed regions of Ramachandran plot (Figs. S3–S5).

Table 4 Dynamics simulation parameters of Apo, D-sedoheptulose 7-phosphate (D-S7P) bound and D-sedoheptulose 7-phosphate + Zn2+ (D-S7P + Zn2+) bound MtbGmhA tetramer.
Full size table

Dynamics simulation on MtbGmhA tetramer in absence of ligand

To investigate the effect of ligand binding on dynamics fluctuation of MtbGmhA tetramer, we have performed the dynamics simulation on MtbGmhA tetramer in absence of ligand. Superposition of simulated MtbGmhA tetramer (Green) on starting structure (Grey) have yielded RMSD ~ 1.3 Å for 705 Cα atoms, indicating quite similar structure, except minor changes were observed in the loop regions of protein (Fig. 7A). High B-factors (~ 40–60 Å2) were observed for amino acid stretches, 20–30, 50–60, 70–80, 105–115, 120–130, 150–155 and 160–170 of MtbGmhA tetramer (Fig. 7B). The initial RMSD of MtbGmhA tetramer was ~ 0.0 and increased slightly (~ 0.2 nm) during 1 ns simulation (Fig. 7G). The radius of gyration (Rg) in MtbGmhA tetramer was quite stable during simulation period and found ~ 2.6 nm (Fig. 7H).

Figure 7
figure 7

(A) Dynamic simulation on apo MtbGmhA tetramer. The simulated MtbGmhA tetramer (Green) is superposed on starting structure (Grey). (B) Plot showing the B-factor (Y-axis) and residue number (X-axis) of simulated MtbGmhA tetramer. (C) Dynamic simulation on MtbGmhA + D-sedoheptulose 7-phosphate tetramer. The simulated MtbGmhA + D-sedoheptulose 7-phosphate tetramer (red) is superposed on starting structure (grey). (D) The B-factor (Y-axis) and residue number (X-axis) of simulated MtbGmhA + D-sedoheptulose 7-phosphate tetramer. (E) Dynamic simulation on MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer. The simulated complex (blue) is superposed on starting structure (grey). (F) The B-factor (Y-axis) and residue number (X-axis) of simulated MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer. (G) The plot showing the backbone RMSD ~ time for apo, D-sedoheptulose 7-phosphate bound and D-sedoheptulose 7-phosphate + Zn2+ bound MtbGmhA tetramer. (H) The plot showing the radius of gyration (Rg) ~ time, which showed the changes in degree of compactness of MtbGmhA tetramer after D-sedoheptulose 7-phosphate and Zn2+ binding.

Full size image

Dynamics simulation on MtbGmhA tetramer in complex with D-sedoheptulose 7-phosphate

1 ns dynamics simulation was performed on MtbGmhA + D-sedoheptulose 7-phosphate tetramer to analyze the effect of substrate binding on MtbGmhA structure. Superposition of the simulated MtbGmhA + D-sedoheptulose 7-phosphate structure (red) on starting complex structure (grey) have yielded the RMSD ~ 1.1 Å for 740 Cα atoms (Fig. 7C). It showed that overall structure of MtbGmhA tetramer remains unperturbed after dynamics simulation, except minor changes in the loop regions of enzyme. The D-sedoheptulose 7-phosphate substrate was quite stable during simulation period. To examine, whether structural deviations occurred in local regions, or throughout the whole structure, we computed the B-factor plot for MtbGmhA tetramer during entire simulation period (Fig. 7D). The amino acid stretches 20–30, 70–80 and 150–155 showed the higher B-factor (~ 40–60 Å2) (Fig. 7D). The MtbGmhA residues e.g., Asn53, Gly55 residues involved in C5-O5 binding and Ser121, Thr122, Ser123, Ser126 residues are involved in C7-PO4 binding have shown significant decrease in B factor compared to wild-type MtbGmhA tetramer. The active site Glu66 and Gln173 residues of MtbGmhA showed low B-factor < 20 Å2 compared to 20–60 Å2 observed in wild type enzyme. These residues bind to C1–O1 and C2–O2 of D-sedoheptulose 7-phosphate and involved in catalysis. Other residues, 20–30, 70–80 and 150–155 showed higher B-factors, though not involved directly in substrate binding.

Figure 7G showed the RMSD of MtbGmhA sampled through dynamics simulation from starting structure. Minor structural changes were observed in 1 ns simulation and reached to final ~ 7 Å. The radius of gyration (Rg) of complex was quite stable and found ~ 2.2 nm (Fig. 7H). These values were quite similar to Rg of MtbGmhA tetramer obtained in small angle X-ray scattering analysis (Rg ~ 2.38 nm). These data showed that D-sedoheptulose 7-phosphate binding to MtbGmhA tetramer enhanced the overall stability of enzyme.

Dynamics simulation on MtbGmhA tetramer in complex with D-sedoheptulose 7-phosphate and Zn2+

Dynamics simulation on MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer was performed to examine the combined effect of Zn2+ and D-sedoheptulose 7-phosphate binding to MtbGmhA. Figure 7E showed the superposition of simulated MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ (Blue) on starting complex structure (Grey) (RMSD ~ 0.5 Å for 752 Cα atoms). Overall structure of MtbGmhA tetramer was quite stable except minor difference in the loop region of the enzyme. The D-sedoheptulose 7-phosphate and Zn2+ ions were quite stable in four subunits of MtbGmhA tetramer, except minor movement of Zn2+ ions occurred in different subunits. It appears that Zn2+ binding further enhanced the stability of MtbGmhA + D-sedoheptulose 7-phosphate tetramer and essential for D-sedoheptulose 7-phosphate binding by MtbGmhA enzyme.

We have computed the B-factor plot for MtbGmhA + D-sedoheptulose 7-phosphate + Zn2+ tetramer during the entire simulation period (Fig. 7F). The amino acid stretches, 20–30, 70–80, 120–130 and 150–160 showed the higher B-factor (~ 40–60 Å2) (Fig. 7D). The RMSD increased slightly till 20 ps and remained stable throughout 1 ns simulation (~ 0.2 Å) (Fig. 7G). The radius of gyration (Rg) was quite stable and found ~ 2.2 nm (Fig. 7H).

Conclusion

In current study, we have performed the structural and biochemical analysis of MtbGmhA enzyme involved in D-α-D-heptose biosynthetic pathway, critical for the development of novel antibiotics against M. tuberculosis. The MtbGmhA forms a tetramer in solution and small angle X-ray scattering analysis also yielded the tetrameric envelope of the enzyme. The MtbGmhA catalyzes the D-sedoheptulose 7-phosphate isomerization with 1.45 mM−1 s−1 rate and binds the MtbHddA and MtbGmhB enzymes in µM range. Site directed mutagenesis have identified the roles of various active site residues involved in D-sedoheptulose 7-phosphate binding and catalysis.

The circular dichroism analysis on wild type and six MtbGmhA mutants showed quite similar secondary structures and high thermostability of the enzyme. The small angle X-ray scattering analysis have revealed de novo SAXS envelope, which fitted well with modeled MtbGmhA tetramer in closed conformation. Dynamics simulation on Apo, D-sedoheptulose 7-phosphate bound and D-sedoheptulose 7-phosphate + Zn2+ bound MtbGmhA tetramer showed that ligand binding enhanced the overall stability of the enzyme. In MtbGmhA tetramer, asymmetric domain movement occurred, compared to isolated monomer. The D-sedoheptulose 7-phosphate binding restricts the domain movement and kept the enzyme in the active conformation. The small angle X-ray scattering analysis indicated that MtbGmhA adopts a compact globular conformation, usually observed in crystal structure. This compact conformation of MtbGmhA is catalytically relevant, not only for isomerization of D-sedoheptulose 7-phosphate, but also in binding to other enzymes involved in D-α-D-heptose biosynthesis pathway. Our structural and biochemical analysis on MtbGmhA have provided a new insight into mechanism, which will be critical for novel antibiotics development against M. tuberculosis. As MtbGmhA is required to maintain the permeability in M. tuberculosis, current knowledge can be implied in other Gram-positive organisms.

Materials and methods

Expression and purification

The MtbGmhA gene (Rv0113) was amplified from H37Rv genomic DNA by polymerase chain reaction and cloned into pET28a expression vector (Novagen) using NdeI and HindIII restriction sites. The pET28a-MtbGmhA plasmid was confirmed by restriction-digestion and gene sequencing analysis. The pET28a-MtbGmhA plasmid was transformed in E. coli BL21(DE3) cells and cells were grown at 37 °C in luria bertani media (50 µg/ml kanamycin as antibiotic) till OD600 ~ 0.6–0.7. The cell culture was induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside and grown for another 4 h. The MtbGmhA protein overexpressed in soluble fraction of the cell.

The cells were harvested by centrifugation at 10,000 × g for 10 min at 4 °C and suspended in lysis buffer containing (25 mM Tris/HCl pH 8.0, 500 mM NaCl, 10% (v/v) Glycerol, 3 mM β-mercaptoethanol, 1 mM Phenylmethylsulfonyl fluoride, 1 mM Benzamidine hydrochloride, 10 mM Imidazole and 0.2 mg/ml Lysozyme). The cells were homogenized, disrupted by sonication and lysate was centrifuged at 25,000 × g to collect the supernatant. The supernatant was loaded on Ni–NTA column (GE Healthcare), pre-equilibrated with buffer-A (25 mM Tris/HCl pH 8.0, 300 mM NaCl, 10% Glycerol, 3 mM β-mercaptoethanol, 1 mM Phenylmethyl sulfonyl fluoride and 1 mM Benzamidine hydrochloride). The column was washed with 0–25 mM gradient of imidazole and eluted the protein in buffer-A + 250 mM imidazole. The eluted protein was concentrated and loaded on Superdex 200 column (GE healthcare Ltd) pre-equilibrated with buffer (50 mM Tris–HCl pH 8.0, 150 mM NaCl, 10% Glycerol and 3 mM β-mercaptoethanol). The peak fractions were pooled and concentrated using 10 kDa cutoff ultracentrifugal device (Millipore, USA). The purified MtbGmhA was analyzed on 12% SDS-PAGE and mass spectrometry. Protein concentration was determined using absorbance at 280 nm and stored at − 80 °C. The recombinant MtbGmhA contains 216 residues e.g. 6 residues from 6xHis tag and 14 residues from vector at N-terminal and 196 residues of MtbGmhA enzyme.

QuickChange II XL site directed mutagenesis protocol (Stratagene) was used to generate the six mutants of MtbGmhA. All mutants were purified with protocol used for wild type MtbGmhA purification21. All MtbGmhA mutants were confirmed by gene sequencing.

Small-angle X-ray scattering analysis

The small angle X-ray scattering data was collected by using SAXSpace instrument (Anton Paar, IMTECH, India) at 10 °C using line collimation optics. The MtbGmhA enzyme (conc. ~ 1.2 mg/ml in buffer ~ 25 mM Tris–HCl pH 8.0, 150 mM NaCl and 3 mM β-mercaptoethanol) was used for small angle X-ray scattering data collection. The small angle X-ray scattering data on Lysozyme (conc. ~ 5.0 and 3.4 mg/ml in buffer ~ 40 mM sodium acetate buffer pH 3.8, 150 mM NaCl) was collected for beam intensity estimation and comparison with MtbGmhA data. After data collection, the MtbGmhA enzyme was checked on SDS-PAGE analysis to see if any degradation occurred during data collection. SAXStreat software was used for beam position correction. 120 µl of MtbGmhA was loaded in quartz capillary flow cell and exposed to X-ray beam (λ = 1.5418 Å).

The small angle X-ray scattering data was collected in triplicate and averaged. SAXSquant 4.2.4 software was used to remove the buffer contribution and desmearing the line collimation. The small angle X-ray scattering data were processed using ATSAS 2.8.3 suite and ScÅtter22 and data containing scattering intensity (I) as a function of scattering vector Q, [where Q = (4π/λ) sinθ, θ is the scattering angle and λ is wavelength] were obtained. The Kratky plot [I(Q)Q2 vs Q] was obtained from intensity profile. The radius of gyration (Rg) and radius of cross section (Rc) were estimated from scattering intensity at low Q region using Guinier analysis. Additionally, using RG and RC values, the length of an ellipsoidal structure L were estimated using the L = [12(RG2 − RC2)] ½ relationship. The GNOM45 program was used to estimate the shape and size of scattering entities, which consider both low and high Q data. Indirect Fourier transformation of scattering data over measured Q range was computed as a pairwise distribution function of interatomic vectors23. This analysis also provided RG and Io from the second moment and start of P(R) as well as Dmax (maximum diameter).

Ten independent models of MtbGmhA were generated using DAMMIF program v. 1.1.224, averaged the aligned model and filtered at a given cutoff volume using DAMAVER v. 5.0 program25. The MtbGmhA tetramer was superimposed on SAXS envelope using the SUPCOMB v. 2.3 program26. The SAXS curve was calculated from the atomic model using the CRYSOL v. 2.8.3 program27. The PyMOL v. 2.3 program28 was used for structural visualization and to generate all superimposed structures. The curve fitting and data plotting were performed using Graphpad-Prism 5 program29.

Enzyme assay

The activity of wild type and six mutants of MtbGmhA were determined by coupling reaction using MtbHddA, MtbHddA and MtbGmhB enzymes and release of Pi was monitored (as described in DeLeon et. al.30). We prepared the 200 µl of reaction mixture containing (0.31 nM MtbGmhA, 0.11 nM MtbHddA and 0.4 nM of MtbGmhB) in 20 mM HEPES buffer pH 8.0, 10 mM MgCl2, 10 mM KCl. 10 mM ATP and 0–3 mM D-sedoheptulose 7-phosphate were added in the reaction mixture and incubated at room temperature for 30 min. The Pi concentration was measured using 50 μl of Pi colorlock Gold and 0.5 μl of Accelerator (Innova Biosciences Ltd) in the reaction mixture. After 2 min, 20 μl of stabilization reagent was added in each sample and absorbance was taken at 630 nm. All reactions were performed in triplicate and 11 different concentrations of D-sedoheptulose 7-phosphate were used for each reaction. The kinetic parameters were determined by fitting the data to Michaelis–Menten equation using GraphPad Prism version 6.0.2 (GraphPad Sofware, La Jolla).

$$v = \frac{{k_{cat} *E_{t} \left[ S \right] }}{{\left( {K_{m} + \left[ S \right]} \right)}}$$

Binding analysis

The D-sedoheptulose 7-phosphate binding to MtbGmhA was performed using Autolab Esprit Surface Plasmon Resonance equipment31. The gold surface [self-assembled monolayer of 11-mercaptoundecanoic acid (11-MUA)] was first activated by EDC (N-ethyl-N dimethyl amino propyl carbodiimide, conc. ~ 0.2 M)—NHS (N-hydroxysuccinimide, conc ~ 0.05 M) coupling. 50 µM of MtbGmhA was immobilized on gold surface using 20 mM sodium acetate buffer, pH 4.2. After immobilization, the surface was blocked with 100 mM ethanolamine pH 8.5. The binding experiment was performed at 25 °C in running buffer (20 mM HEPES pH 7.5, 150 mM NaCl, 5 mM MgCl2, and 2% Glycerol). The D-sedoheptulose 7-phosphate substrate was injected at 30 µl/min for 5 min, followed by dissociation for 5 min. After each experiment, the sensor surface was regenerated using 50 mM NaOH.

For interaction analysis between MtbGmhA, MtbHddA and MtbGmhB, 50 µM of MtbGmhB was immobilized on gold surface activated by EDC-NHS coupling using 20 mM sodium acetate buffer, pH 4.2. The surface was blocked with 100 mM Ethanolamine pH 8.5. 50 µM of MtbGmhA was injected at 30 µl/min for 5 min followed by dissociation for 5 min. Similar protocol was used for interaction analysis between MtbHddA and MtbGmhB and MtbGmhA and MtbHddA enzymes. All experiments were performed at 25 °C. The sensograms were collected, processed, analyzed using BIAevaluation (GE Healthcare)32 and fitted with Langmuir 1:1 model using formula

$$R_{eq} = \frac{{R_{max} *K_{A } * C_{A} }}{{K_{A} *C_{A} + 1}}$$

where Req is binding response at steady state, Rmax is maximal binding response, CA is the concentration of analyte and KA is equilibrium association constant.

Circular dichroism analysis

Circular dichroism analysis was performed to estimate the secondary structure of MtbGmhA enzyme33. A far UV spectrum on MtbGmhA was collected using Chirascan spectropolarimeter (Applied Photophysics) with quartz cuvette of 1 mm path length. The MtbGmhA enzyme was transferred in Sodium phosphate buffer, pH 7.5. The circular dichroism data was collected in 200–260 nm range at 25 °C. The spectra from baseline was subtracted from protein spectra and obtained values were averaged for each dataset. The mean residue ellipticity (θ) was calculated from the observed spectra, as a function of wavelength. The K2D2 program was used to estimate the secondary structure of MtbGmhA34. Similar protocol was used for CD data collection on all six MtbGmhA mutants. The theoretical secondary structure prediction on MtbGmhA was performed using various programs e.g., SOPMA35, CFSSP36, GOR37, PHD38, SIMPA9639, DSC40, HNN41, RaptorX42, JPred43 and PsiPred44.

For thermal stability analysis, the circular dichroism spectra of MtbGmhA was recorded at 222 nm from 25 to 90 °C range in 10 °C step. The melting temperature was calculated from scattering profile using polynomial fitting, taking temperature corresponding to half denaturation45. A graph of (ΔA/ΔT) versus (Tavg) was plotted using Sigma Plot program. A bell-shaped melting curve was observed, and peak corresponds to melting temperature of the protein.

Multiple sequence alignment

The MtbGmhA sequence was retrieved from UniProt (P9WGG1) database and sequences of 11 GmhA orthologues were retrieved from RCSB database46 The MtbGmhA sequence was aligned with 11 GmhA orthologues using MultAlin47 and ESPript 3.048 programs.

Molecular modeling, docking and dynamics simulation

I-TASSER server was used to model the MtbGmhA tetramer and was ranked by the C-score and TM-score49. The MtbGmhA sequence (1–196 residues) was given as input and server retrieved the template proteins of similar folds from Protein Data Bank by LOMETS (Local Meta-Threading-Server)50. Additional steps were performed to remove the steric clashes and refined the global topology of the MtbGmhA models.

I-TASSER server identified the PDB-2X3Y (Crystal structure of GmhA tetramer in closed state from B. pseudomallei12) as the best template for MtbGmhA modeling (RMSD = 0.79, id1 = 0.39, id2 = 0.39, Conv = 0.98, Z-score = 3.2). The best MtbGmhA model was obtained having following parameters (C-score = 0.21 and TM-score = 0.88 ± 0.07), which indicated a reliable model with correct global topology. The I-TASSER server also yielded PDB-2I2W (Crystal structure of Escherichia Coli Phosphoheptose Isomerase in open state10) as another template for MtbGmhA modeling, however following parameters (RMSD = 2.1, TM id1 = 0.35, TM-score = 0.82, Conv = 0.89) did not warrant a reliable MtbGmhA model in “open state”.

Docking of D-sedoheptulose 7-phosphate and Zn2+ into MtbGmhA tetramer was performed using GLIDE module of the SchroeÖdinger-9.4 program51. The MtbGmhA + S7P + Zn2+ complex was generated using induced fit protocol of SchrÖdinger-9.4 program. Defaults parameters were used except Extra Precision (EP) scoring function was used in docking calculation.

The dynamics simulation of (i) Apo (ii) D-sedoheptulose 7-phosphate bound and (iii) D-sedoheptulose 7-phosphate + Zn2+ bound models of MtbGmhA tetramer were performed using GROMACS package 4.5 (GROningen MAchine for Chemical Simulations) using CHARMM all-atom force field parameters (Table 4)52. The MtbGmhA tetramer was solvated in the dodecahedron box at 1 nm distance from the boundary. Simple point charge water molecules were used for solvation together with Na+ and Cl− ions to neutralize the total charge of the system. Energy minimization was performed using 50,000 steps of steepest descent minimization. To maintain the 300 K constant temperature, protein and non-protein atoms were coupled to their temperature baths using V-scale thermostat using 100 ps NVT. The whole system was kept for 100 ps NPT equilibration and 1 bar pressure. For all simulations, the bond lengths and water molecules were restrained using the SETTLE and LINCS53 algorithms respectively.

The 1 nm cutoff was used for the treatment of Van der Waals interactions and long-range electrostatic interactions and simulated using PME54 method with 0.16 FF grid spacing and 4th order B-spline interpolation for the reciprocal sum space. Periodic boundary conditions were applied in all directions. Rigid body displacements and rotations were removed from all trajectories. The PRODRG server55 was used to generate the topology of the D-sedoheptulose 7-phosphate. PyMol and Plot256 programs were used for generating the figures and all simulation trajectories. The MtbGmhA tetramers obtained after dynamics simulation exhibited good stereochemistry. The structural validation on all models was performed using various servers e.g., ERRAT57, Verify3D58, and PROCHECK59.