Monomeric Camelus dromedarius GSTM1 at low pH is structurally more thermostable than its native dimeric form

Glutathione S‒transferases (GSTs) are multifunctional enzymes that play an important role in detoxification, cellular signalling, and the stress response. Camelus dromedarius is well-adapted to survive in extreme desert climate and it has GSTs, for which limited information is available. This study investigated the structure-function and thermodynamic properties of a mu-class camel GST (CdGSTM1) at different pH. Recombinant CdGSTM1 (25.7 kDa) was expressed in E. coli and purified to homogeneity. Dimeric CdGSTM1 dissociated into stable but inactive monomeric subunits at low pH. Conformational and thermodynamic changes during the thermal unfolding pathway of dimeric and monomeric CdGSTM1 were characterised via a thermal shift assay and dynamic multimode spectroscopy (DMS). The thermal shift assay based on intrinsic tryptophan fluorescence revealed that CdGSTM1 underwent a two-state unfolding pathway at pH 1.0–10.0. Its Tm value varied with varying pH. Another orthogonal technique based on far-UV CD also exhibited two-state unfolding in the dimeric and monomeric states. Generally, proteins tend to lose structural integrity and stability at low pH; however, monomeric CdGSTM1 at pH 2.0 was thermally more stable and unfolded with lower van't Hoff enthalpy. The present findings provide essential information regarding the structural, functional, and thermodynamic properties of CdGSTM1 at pH 1.0–10.0.


Introduction
Glutathione S-transferase (GST; EC 2.5.1.18) is one of three major groups of enzymes that carry out detoxification (elimination of foreign, cytotoxic, and genotoxic compounds) and is charge cluster are highly conserved (Fig 1A). Due to unavailability of X-ray/NMR structure of CdGSTM1, the 3D structure of CdGSTM1 was modeled using 2.68 Å X-ray diffraction structure of ligand-free hGSTM1A-1A (PDBID: 1GTU) as a template. Modeled 3D structure of CdGSTM1 was superimposed on hGSTM1A-1A (Fig 1B), indicated a high degree of similarity. eliminate impurities and GSH (Fig 2A). A single sharp peak was observed at 70 mL of eluted volume which is indicated that the fractions contained a homogenous population of dimeric CdGSTM1. Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) was also performed to confirm the level of CdGSTM1 homogeneity (purity). SDS-PAGE revealed a single band at~26 kDa, thereby confirming that CdGSTM1 was highly pure (Fig 2A inset). The yield of soluble, pure CdGSTM1 protein was approximately 30 mg L -1 in a shake-flask culture. The quaternary structure of the purified protein was determined using an analytical Superdex 200 column. CdGSTM1 was eluted at a volume of 15.3 mL, which corresponds to approximately 50.58 kDa which, is similar to the calculated molecular weight (51.5 kDa) of dimeric CdGSTM1 (Fig 2B).

Effect of pH change on CdGSTM1 activity
The enzymatic activity of CdGSTM1 was measured via equilibration at pH 1.0 to 10.0, and subsequently, residual activity was evaluated using 2,4-dinitrochlorobenzene (CDNB) as a substrate at optimum pH of 6.5 [41]. As shown in Fig 3, enzymatic activity of CdGSTM1 drastically decreased below pH 4.0, probably owing to denaturation or dissociation of CdGSTM1 subunits. However, loss of enzymatic activity at acidic pH was reversible (data not shown).

Effect of pH change on CdGSTM1 conformation
Far-UV CD, fluorescence spectroscopy, and SEC were performed to investigate changes in the secondary, tertiary, and quaternary structure of CdGSTM1 at different pH values. Before the measurements, CdGSTM1 was incubated overnight at different pH values ranging from 1.0-10.0.
Intrinsic fluorescence measurements provided clear information regarding protein tertiary structural change. The primary intrinsic fluorophores in proteins are tryptophan (Trp), tyrosine, and phenylalanine. Total fluorescence spectra of CdGSTM1 at different pHs are shown in Fig 4. From the figure, it was recorded that the CdGSTM1 at pH 7.0 showed maximum fluorescence intensity at 336 nm, confirming that CdGSTM1 exists in a well-folded form. The fluorescence emission maximum of CdGSTM1 remained unchanged in the range of 5.0-9.0 pH, indicating that the tertiary structure of CdGSTM1 was intact (Fig 4, inset). However, between pH 3.0 and 5.0, the wavelength maximum of CdGSTM1 was 6.0 nm (red shifted). The red shift in the wavelength maximum signified that the tryptophan residues were exposed to the polar environment, thus indicating changes in protein tertiary structure. Moreover, the maximum wavelength remained unchanged below pH 3.0, indicating that the same microenvironment around tryptophan residues was retained. Fluorescence intensity of CdGSTM1 increased gradually with a reduction in pH because the position of certain amino acid residues changed apart from the Trp residues, which were responsible for quenching of Trp fluorescence. Such pH-dependent changes were also reported previously [42]. The shift in maximum wavelength suggests that dimeric CdGSTM1 either begins monomerising below pH 5.0 or undergoes a loss of tertiary structure interactions, which were further characterized by other advance techniques.

Changes in the secondary structure of CdGSTM1 with changes in pH
Far-UV CD was used to characterize the change in the secondary structure of CdGSTM1 with respect to change in pH. The secondary structure of CdGSTM1 at different pH values was analysed via far-UV CD (200-250 nm) measurements ( Fig 5). Far-UV CD spectra of CdGSTM1 at pH 7.0 yielded two negative minima, one at 208 and the other at 222 nm, which is a characteristic feature of an alpha helix ( Fig 5A). Changes in the negative ellipticity of CdGSTM1 at 208 and 222 nm were insignificant at pH 1.0-10.0. CDNN analysis revealed that percent secondary structure contents (alpha-helical, beta-sheet, beta-turn, and random coil) of CdGSTM1 were not significantly altered with changes in pH ( Fig 5B). Far-UV CD results indicated that the secondary structure of CdGSTM1 was remained stable with changes in pH.

The 8-anilino-1-naphthalenesulfonic acid (ANS) binding assay
Exposure of hydrophobic patches in CdGSTM1 at different pH were characterized by ANS fluorescence measurements. ANS is a hydrophobic, sensitive dye, which binds with hydrophobic regions of proteins. At pH 7.4, the CdGSTM1 displayed maximum fluorescence intensity at 518 nm, which confirms that CdGSTM1 is well-folded, while at an acidic pH (pH 2.0) the fluorescence intensity was increased by approximately two folds with maximum wavelength blue-shifted up to 10 nm ( Fig 6A). Increase in ANS fluorescence intensity and blue-shift in maximum wavelength signified that the CdGSTM1 was either partially unfolded or monomerised.

Rayleigh light scattering measurements
Rayleigh light scattering was performed to investigate pH-dependent CdGSTM1 aggregation. Light scattering of samples was monitored at 500 nm to determine the aggregation behaviour of CdGSTM1 at different pH values. As shown in Fig 6B,

Gel-permeation chromatography
Gel-permeation chromatography is used to separate proteins on the basis of size and it is also used to determine the changes in the quaternary structure of proteins. Intrinsic and ANS fluorescence analyses revealed that camel CdGSTM1 undergoes a sharp transition between pH 3.0 and 5.0, probably owing to pH-dependent monomerisation or partial unfolding. Therefore, gel-permeation chromatography was performed to distinguish the level of monomerisation at  ; Fig 7). At pH 7.0, CdGSTM1 was eluted at a single symmetrical peak (blue), corresponding with its dimeric state (Fig 2B). When CdGSTM1 was equilibrated at pH 4.0 (pink), gel-permeation chromatograms revealed two peaks, one at 16.3 mL and the other at 25.9 mL, indicating that two types of populations were obtained: one with the size of the dimeric state; the other, the monomeric state. CdGSTM1 incubated at pH 2.0 yielded peaks at 22.2 and 25.9 mL similar to those of the monomeric state (black). Gel-permeation chromatography revealed that above pH 5.0, CdGSTM1 exists in its dimeric form; however, between pH 3.0 and 5.0, it exists in a monomeric-dimeric equilibrium and below pH 3.0, it exists only in the monomeric form.

Glutaraldehyde cross-linking
Glutaraldehyde cross-linking is widely used techniques to distinguish the subunit composition (monomer, dimer and tetramer) of proteins. pH-dependent changes in the quaternary structure of CdGSTM1 were further investigated via glutaraldehyde cross-linking (Fig 8). CdGSTM1 treated at pH 2.0, 4.0, and 7.0 was passed through a Superdex 200 column to separate CdGSTM1 monomers from dimers (Fig 7). The samples of CdGSTM1 at pH 2.0, 4.0, and 7.0 before and after gel filtration were cross-linked with glutaraldehyde. The cross-linked products of CdGSTM1 before ( Fig 8A) and after ( Fig 8B)  CdGSTM1 at different pH was loaded on a 24-mL bed volume Superdex 200 column to separate monomer-dimer species of CdGSTM1. At pH 2.0, CdGSTM1 was dissociated and eluted at two peaks (22.2 and 25.9 mL). Although 150 mM NaCl was added in the gel filtration buffer to suppress protein-matrix interactions, CdGSTM1at pH 2.0 was retained in the column and eluted near the column bed volume. Similarly, CdGSTM1 at pH 4.0, where it exits in a monomer-dimer equilibrium, eluted at two peaks. The dimer eluted at the expected peak (16.3 mL), followed by the monomeric form after crossing the bed volume owing to retention in the matrix. All elution peaks obtained at pH 2.0, 4.0, and 7.0 were separately subjected to glutaraldehyde cross-linkage ( Fig 8B). The first peak at pH 2.0 displayed cross-linkage in the monomeric form (lane 1) and the second peak was very small and showed a faint band

Acrylamide quenching fluorescence
Acrylamide quenching measurements were done to characterize the Trp exposure of proteins. The exposure of Trp amino acid residues to polar solvents was identified via uncharged acrylamide quenching measurements. The present results provide a detailed insight into conformational changes in CdGSTM1 protein at different pH conditions.  Table 1. As shown in Fig 9 and Table 1 However, Trp residues were also exposed at pH 2.0, but to a lesser extent than in 6M GdnHCltreated samples. Hence, CdGSTM1 exists in a well-folded form at a native pH, slightly unfolded at low pH, and completely unfolded in the presence of 6M GdnHCl. The slight exposure of Trp residues at pH 2.0 suggests that either CdGSTM1 is partially unfolded or the subunit dissociation.

Thermal stability of CdGSTM1 at different pH, determined via the thermal shift assay
We have determined the thermal stability of CdGSTM1 at pH 1.0-10.0 in response to changes in intrinsic tryptophan fluorescence. The temperature-melting curves of the samples were obtained through gradual heating at a constant rate of 1˚C min -1 . CdGSTM1 formed visible aggregates above 75˚C between pH 5.0 and 10.0. However, below pH 3.0, no aggregates were observed up to 94˚C. Therefore, CdGSTM1 samples between pH 5.0 and 10.0 were gradually heated till 70˚C and samples below pH 3.0 were subjected to thermal denaturation till 94˚C. The emission ratio of 350 nm to 330 nm was plotted with respect to temperature at the different pH values (Fig 10). The ratio of 350/330 nm was clustered in two groups owing to a shift in λ max (6-nm red-shift) at pH 3.0.  CdGSTM1 undergoes a single thermal transition at all pHs. The tertiary structure of CdGSTM1 undergoes thermal denaturation via a two-state unfolding pathway at all pH values (Fig 10). In this study, thermal stability of CdGSTM1 was moderate between pH 1.0 and 10.0 ( Table 2). CdGSTM1 was relatively more stable at an acidic pH than at neutral and alkaline pH. Exceptionally high stability (65.2˚C) at pH 3.0 might have resulted from monomer-dimer equilibrium.

Thermodynamic and spectroscopic properties of dimeric and monomeric CdGSTM1 revealed through dynamic multimode spectroscopy
Dynamic multimode spectroscopy is an information-rich experimental technique based on far-UV CD to determine changes in secondary structure in the entire temperature range and furnishes thermodynamic data [43,44]. Therefore, the dimeric form (pH 7.0) and monomeric form (pH 2.0) of CdGSTM1 were subjected to dynamic multimode spectroscopy (DMS) to evaluate structural and thermodynamic properties in both states. The samples at pH 7.0 were constantly heated (1˚C min -1 ) from 20-75˚C, while the sample at pH 2.0 was heated up to 94˚C under identical conditions. Far-UV CD spectra of both samples were recorded between 200 and 260 nm as a function of temperature. Selected wavelengths are shown in Fig 11. Dimeric CdGSTM1 (pH 7.0) and monomeric CdHSTM1 (pH 2.0) unfold via a single transition. At pH 7.0, they undergo sharp thermal transitions and secondary structure is completely lost at 57˚C, while CdGSTM1 at pH 2.0 follows shallow thermal transition and retains nearly half of the secondary structure up to 94˚C. The thermal melting points (Tm) and van't Hoff enthalpy of CdGSTM1 at both pHs were determined via Global 3 analysis software, Applied Photophysics Ltd, UK. The Tm of dimeric CdGSTM1 at pH 7.0 was 51.5 ± 0.1˚C and van't Hoff enthalpy was 666.4 ± 9.8 kJ mol -1 . The Tm of monomeric CdGSTM1 was higher 54.2 ± 0.1˚C; however, the van't Hoff enthalpy was quite low (116.5 ± 0.8 kJ mol -1 ). The thermal shift assay also yielded a single thermal transition with Tm values of 49.0 and 54.8˚C at pH 7.0 and 2.0, respectively (Fig 10).
Changes in secondary structure conformation at different temperature are shown in Fig 12. At pH 7.0, loss of the secondary structure was observed beyond 45˚C and was completely lost at 57˚C. During thermal unfolding beyond 45˚C, a 222-nm peak shifted towards single a 225-nm peak (Fig 12A). Thermal stress at pH 7.0 led to the conversion of CdGSTM1 from an alpha helical structure into a cross-beta sheeted structure, which aggregated beyond 75˚C [45]. However, thermal denaturation of monomeric CdGSTM1 led to the conversion of the same alpha-helical structure to a random-coil structure (Fig 12B). Hence, monomeric CdGSTM1  slowly loses its secondary structure beyond 30˚C but retains nearly half of its secondary structure even up to 94˚C. Thereafter, we measured the reversibility of thermal unfolding in dimeric and monomeric CdGSTM1. As shown in Fig 13, dimeric CdGSTM1 at 20˚C is an alpha-helical protein, which has completely lost its secondary structure at 75˚C. Moreover, heat-denatured dimeric CdGSTM1 could not refold. Monomeric CdGSTM1 at pH 2.0 was also alpha-helical at 20˚C, but converted into a completely reversible random-coiled dominant structure at 94˚C. Far-UV CD spectra of heat-renatured monomeric CdGSTM1 at pH 2.0 was superimposed on the native monomeric form of CdGSTM1 at 20˚C.

Discussion
GSTs are structurally conserved and ubiquitous in virtually all organisms [22]. They are tremendously functionally divergent, which is an inevitable requisite for their role as major detoxification enzymes [46,47]. The role of GSTs in multiple metabolic pathways and in stress responses in several organisms has been extensively studied [11,48]. Moreover, GST has been used as a fusion tag for solubility enhancement, ease of purification, and other protein engineering purposes [49,50]. Most GSTs are marginally thermostable at neutral pH. However, certain GSTs exhibit relatively higher thermo-stability and wider pH stability [51][52][53]. This study evaluated thermodynamic and conformational properties of the pH-induced structural changes in recombinant CdGSTM1 via enzymatic assays, size-exclusion chromatography, intrinsic and extrinsic fluorescence, acrylamide quenching, far-UV CD, thermal shift assay, and dynamic multimode spectroscopy.
CdGSTM1 was cloned in its native, soluble state in E. coli. Subsequently, it was purified to homogeneity, using affinity (GSH-agarose) and size-exclusion chromatography. The In the heat-stressed condition, CdGSTM1 in the dimeric state loses its secondary structure completely, while the monomeric form retained nearly half of the secondary structure with a dominant random-coil conformation. The dimeric form of CdGSTM1 was unable to refold, while the monomeric form refolded completely to its secondary structure.
https://doi.org/10.1371/journal.pone.0205274.g013 molecular weight of CdGSTM1 was determined using SEC and was found to exist in the dimeric state in solution. Between pH 3.0 and 5.0 CdGSTM1 undergoes a conformational change, resulting in nearly complete loss of enzyme activity. The crystal structure of Mu-class GST revealed that Tyr7 (catalytic residue) located at the NTD of one subunit interacts with the Phe57 located at the interface loop of the second subunit [20]. Destabilization of the interface loop, in turn, affects substrate binding and catalysis [54]. Hence, dimer interface interactions stabilize the loop conformation and are essential for enzyme activity.
Intrinsic tryptophan fluorescence provides information regarding polarity and hydrophobicity of the tryptophan microenvironment. It has been extensively used as a spectral probe to investigate protein tertiary structure [55]. Generally, interior tryptophan residues in folded proteins displayed fluorescence emission maximum below 340 nm, whereas upon protein unfolding, the fluorescence emission maximum red-shifted to 350-355 nm. CdGSTM1 contains four tryptophan residues. A 3D model of CdGSTM1 revealed that two of these (Trp8 and Trp46) are located proximal to the dimer interface (data not shown). The emission maximum of CdGSTM1 at neutral pH was observed at 336 nm (Fig 4), suggesting that the tryptophan residues in the native conformation are significantly buried in the protein core. The tryptophan emission maxima were mostly unchanged from pH 5.0 to 10.0, suggesting that the tertiary structure of the protein was stable in this pH range (Fig 4). However, a sharp red-shift (6 nm) in the intrinsic emission maxima was observed between pH 3.0 and 5.0, with no further change in emission maxima below pH 3.0 (Fig 4). The red-shift in emission maxima resulted from changes in the tryptophan microenvironment from a non-polar to a polar environment during protein denaturation or dissociation of subunits of multimeric proteins or both [55]. A minor shift in emission maxima (from 336 to 342 nm) of CdGSTM1 below pH 5.0 indicated a marginal change around the tryptophan residues in CdGSTM1 at low pH. On treatment with a chemical denaturant (6 M GdnHCl), CdGSTM1 undergoes complete unfolding with an emission maxima of 360 nm (data not shown).
Loss of enzyme activity seems to coincide with red-shift in emission maxima, indicating exposure of aromatic residues, especially tryptophan, to the polar environment. These data suggest the dissociation or denaturation, or both, of dimeric CdGSTM1 below pH 5.0. An extrinsic fluorescence dye (ANS) does not fluoresce in an aqueous buffer but fluoresces when exposed to a hydrophobic environment. ANS does not bind or exhibit very low binding with the native and completely unfolded polypeptide chain; however, it binds with protein folding intermediates formed during unfolding [56]. ANS emission spectra, used to measure surface hydrophobicity, showed a blue-shift and a two-fold increase in fluorescence intensity between pH 3.0 and 5.0. Previously, ANS was found to bind the dimer interface, thereby decreasing the emission maximum from 535 nm to 490-500 nm, accompanied by an enhanced fluorescence signal [29,30,57]. The blue-shift in the ANS spectra revealed that the environment of the ANS-binding site is hydrophobic. RLS is routinely used to detect the quantity of aggregates in a solution [58,59]. An increase in surface hydrophobicity did not affect the solubility of CdGSTM1 at low pH.
The integrity of the CdGSTM1 quaternary structure was assessed using SEC and glutaraldehyde cross-linking, which revealed that CdGSTM1 dissociated into subunits at pH 4.0 and exists in a monomer-dimer equilibrium. However, at pH 2.0, it exists only as a monomer. Although 0.15 M NaCl was included in the gel filtration buffer to supress ionic interactions with the matrix, monomeric CdGSTM1 had a retarding effect on the column. Therefore, we could not determine the molecular weight of the subunit.
The dimeric form often tends to contain Trp, Tyr, and Phe (large hydrophobic amino acid residues) at the dimer interface to anchor subunits [60]. A modelled structure of CdGSTM1 revealed a conserved Phe57 residue at the dimer interface, which serves as the key of the hydrophobic lock-and-key motif (Fig 1B). The phenyl ring of Phe57 protrudes from the NTD of one subunit into a hydrophobic pocket generated within the C-terminal domain of the other subunit formed by alpha-helix 4 and 5 [36]. A mixed charge cluster (six positively and six negatively charged residues) in a 24-residue-long stretch (Arg78 to Glu101) located close to the dimer interface (data not shown) forms ionic interactions between subunits. Thus, dimeric CdGSTM1 stabilized by charge-charge interaction and hydrophobic forces. The crystal structure of rat GSTM1-1 revealed that two groups of interactions (hydrophobic lock-and-key interaction motifs and ionic interactions) formed the core of the interface contacts and stabilized it [20,26,33,54].
Far-UV CD spectroscopy has been extensively used in analysing the secondary structure of the protein under different conditions. In this study, the far-UV CD spectrum of CdGSTM1 displayed a typical alpha-helix-structure dominant structure (Fig 5A). All the secondary structures (alpha helices, beta sheets, and beta-turns) contents were almost intact between pH 1.0 and 7.0. In the alkaline region, slight conversion of the alpha helix to the random-coil was observed (Fig 5B). Although CdGSTM1 lost its activity at a low pH and changes in tertiary structure also occurred below pH 5.0, far-UV CD revealed that CdGSTM1 retains nearly all structures below pH 5.0. These data suggest that CdGSTM1 dissociates below pH 5.0 but does not denature. Monomeric CdGSTM1 was folded and was soluble but inactive owing to the requirement of a substrate interface for substrate binding.
Thermal shift assay based on intrinsic tryptophan fluorescence and thermal transition based on far-UV CD spectra of the proteins are orthogonal techniques, which have been extensively used to monitor protein stability under different conditions [61,62]. Thermal melting curves based on changes in the tertiary structure of CdGSTM1 revealed that it follows twostate thermal unfolding at pH 1.0 to 10.0. The protein was moderately thermostable, and its thermostability increased significantly below pH 3.0. Thermal stability based on secondary structure (far-UV CD) also had similar Tm values and temperature-induced unfolding pathways of dimeric and monomeric CdGSTM1 compared to those obtained with the thermal shift assay [62]. The present results show that dimeric CdGSTM1 undergoes a sharp thermal transition with a large change in van't Hoff enthalpy (666.4 ± 9.8 kJ mol -1 ), while the monomeric form undergoes broad range of thermal transitions (approximately 40˚C) with lower van't Hoff enthalpy 116.5 ± 0.8 kJ mol -1 . Dimeric CdGSTM1 attains cross-beta sheeted structures during thermal unfolding, which undergoes irreversible thermal aggregation. However, alpha helices in monomeric CdGSTM1 were gradually converted to random-coil-like structures during thermal unfolding, which was completely reversible. Similar thermal unfolding behaviour was also observed in rat GSTM2-2, which unfolds via a single transition with a Tm of 54˚C [36].
Understanding protein stability is necessary to gain insights into protein folding, structure, and function. Proteins that are marginally stabilized exceed a minimal threshold to attain a native conformation and function [63,64]. Generally, thermodynamic stabilities of most water-soluble globular proteins varies from 12 and 42 kJ/mol [64]. The protein is stabilized via a delicate balance between certain forces and interactions. Hydrophobic forces and electrostatic interactions are major factors determining protein stability. To balance optimum biological function and conformational stability, electrostatic interactions are optimised in proteins [65]. Analysis of pH-dependent protein stability helps understand electrostatic interactions in proteins and the contribution of specific charged residues in protein structure and function. The thermodynamic stability of proteins in the native and denatured state are linked to pKa groups of the side chains, which is pH-dependent. The pKa values of the residues in the native and unfolded states depend on several factors: charge-charge, H-bonds, charge-dipole, and desolvation effects. The degree of interactions between charged resides and the rest of the protein in the native or denatured forms determines the titration properties of the ionisable groups. Generally, proteins lose structure-function and stability below pH 5.0 and above pH 10.0 owing to ionization of buried ionisable residues (e.g., Tyr, His) in the folded native state. In the strong acidic or alkaline region, ionisable residues from the surface to the core of the proteins gradually get ionized and lead to protein unfolding. The relatively higher stability of monomeric CdGSTM1 compared to that of dimeric CdGSTM may have resulted from higher intra-chain interactions within the monomer. The overall data is schematically presented in Fig 14. Briefly, CdGSTM1 exists in dimeric state at pH 7.0 while it equilibrated in dimermonomer form at pH 4.0. At acidic pH of 2.0, it is populated in monomeric form. The dimeric form was having relatively less thermal stability than the monomeric form.

Multiple sequence alignment and structural analysis
The homologous mammalian sequences of GSTM1 was identified using CdGSTM1 (accession number XP_010974221.1) amino acid sequence in PSI-BLAST. Ten homologous sequences of CdGSTM1 were retrieved from database. Multiple sequence alignment was done using MAFFT Multiple Sequence Alignment and Jalview [40,66]. The output of aligned sequences was color-coded. Three-dimentional structure CdGSTM1 was modelled using Swiss-model server [67] using human glutathione S-transferase hGSTM1a-1a (PDB: 1GTU) as a template. The 3D structure of GSTM1 was analyzed using PyMOL software [68].
Active fractions were pooled and subjected to size-exclusion chromatography to eliminate minor impurities and reduced glutathione. Superdex 75 16/600 GL prepacked column was equilibrated with 50 mM phosphate buffer containing 150 mM NaCl at pH 7.5. The active fraction of GSH-agarose elute was pooled and loaded on a column using a 5 mL superloop on an AKTA purification system. The purity of the gel filtration peak was analysed with a 4-15% mini-PROTEAN pre-casted TGX gel (Bio-rad) and stained with Coomassie Brilliant Blue G-250. The concentration of purified CdGSTM1was estimated at spectrophotometrically at 280 nm, with a molar extinction coefficient of 39,880 M −1 cm −1 .
The molecular weight of the CdGSTM1 was calculated using a Superdex 200 increase 10/ 300 GL Prepacked Tricorn Column, calibrated with five proteins: thyroglobulin (669 kDa), ferritin (440 kDa), aldolase (158 kDa), conalbumin (75 kDa), and ovalbumin (44 kDa). Purified CdGSTM1 (2.5 mg mL -1 ) in 50 mM Phosphate buffer containing 150 mM NaCl at pH 7.0 was applied on the column using a 500 μL superloop. To determine the quaternary structure at different pH, 0.25 mg CdGSTM1 at different pH (2.0, 4.0, and 7.0) were loaded on the column pre-equilibrated with buffer containing 150 mM NaCl at the aforementioned three values of pH.

GST assay
CdGSTM1 activity was determined spectrophotometrically based on the increase in UV absorbance at 340 nm at 25˚C owing to the conjugation of GSH with 1-chloro-2,4 dinitrobenzene (CDNB) in accordance with a previously reported method [41], with an extinction coefficient of 9.6 mM -1 cm -1 . The reaction was carried out in 1 mL of 0.1 M potassium phosphate buffer (pH 6.5), 1.0 mM GSH, and 1.0 mM CDNB dissolved in ethanol and the enzyme preparation. One unit of CdGSTM1 activity was defined as the amount of GST that catalyses the formation of 1.0 μmole of thioether per min.

Fluorescence spectroscopy
Total and tryptophan fluorescence emission spectra were measured at 25˚C in a Cary Eclipse Fluorescence Spectrophotometer (Agilent Technologies, California, USA), coupled with Peltier temperature controller. CdGSTM1 (0.1 mg mL -1 ) at different pH (1.0-10.0) in a 10-mm path-length cuvette was excited at 295 nm (bandwidth, 5 nm each) for recording total and tryptophan fluorescence emission spectra, respectively. For aggregation studies of CdGSTM1at different pH, the Rayleigh light scattering (RLS) method was used. CdGSTM1 concentration at different pH was maintained at 0.1 mg mL -1 . The CdGSTM1 sample was excited at 500 nm, and emission was collected from 450-550 nm. The bandwidth of excitation and emission slit was 1.5 nm. Fluorescence intensity at 500 nm was plotted against different pH. The change in surface hydrophobicity with respect to pH was measured using 8-anilinonaphthalene-1-sulfonic acid (ANS) extrinsic fluorophore. Fifty-molar excess of ANS was added in the CdGSTM1 samples at different pH and incubated at room temperature for 30 min. Subsequently, samples were excited at 380 nm (2.5 nm bandwidth), and emission of ANS-bound proteins was measured at 400-600 nm (5-nm bandwidth).

Acrylamide quenching
Stepwise two μl of acrylamide quencher was added from the stock of 5.0 M into incubated protein (0.10 mg mL -1 ) samples. The quencher was added to the protein samples till saturation was achieved (0-0.12 M). The Trp residues of protein sample were excited at 295 nm, and the emission spectrum was recorded at 300-400 nm. Reduction in fluorescence intensity at wavelength maximum was analyzed using the Stern-Volmer equation.
where Fo and F are the fluorescence intensities at maximum wavelength in the absence and presence of the quencher, respectively, Ksv is the Stern-Volmer quenching constant, and [Q] is the concentration of the quencher.

Thermal shift assay
Thermal stability of CdGSTM1 at different pH was analysed by continuously heating samples at 1˚C min -1 from 30-90˚C. The actual temperature of the samples was monitored using an internal temperature probe. The samples were excited at 295 nm (10-nm bandwidth) and data were collected at 330, 340, and 350 nm (10-nm bandwidth).

Circular dichroism spectroscopy
Far-UV CD measurements were obtained using a Chirascan Plus spectropolarimeter (Applied Photophysics Ltd, UK), coupled with a Peltier temperature controller. The instrument was calibrated with (1S)-(+)-10-camphorsulphonic acid. The far-UVCD spectra were measured at a CdGSTM1 concentration of 0.2 mg mL -1 with a 1-mm path length cell at 25˚C. Three spectra of each sample were obtained at 200-260 nm. Air and buffer blank were subtracted from each spectrum. Percent secondary structure content was calculated using CDNN software. Dynamic multi-mode spectroscopy of CdGSTM1 at pH 2.0 and 7.0 was performed with a Chirascan-Plus spectrophotometer. Protein samples (0.2 mg mL -1 ) equilibrated overnight at pH 2.0 and 7.0 were gradually heated at 1˚C min -1 , and far-UV CD spectra were recorded at 200-250 nm. The actual sample temperature was monitored using an internal temperature probe. Thermal melting curves were analysed using Chirascan's Global 3 software, Applied Photophysics Ltd, UK.

Glutaraldehyde cross-linking
Glutaraldehyde cross-linking was performed in accordance with a previously reported method [69] with slight modification. Briefly, 0.2 mL of CdGSTM1 (0.5 mg mL -1 ) at pH 2.0, 4.0, and 7.0 were treated with 0.25% glutaraldehyde at room temperature for 15 min. Subsequently, cross-linking was quenched by addition of 0.2 M sodium borohydride and incubated for 20 min at room temperature. Thereafter, 0.005 mL of 10% sodium deoxycholate was added. The cross-linked protein was precipitated by decreasing the pH to 2.0-2.5 via addition of orthophosphoric acid. The precipitated protein was separated via centrifugation at 13,000 rpm for 10 min at 4˚C. The pellet was dissolved in SDS-loading dye.