Probing the role of the fully conserved Cys126 in triosephosphate isomerase by site-specific mutagenesis – distal effects on dimer stability


P. Balaram, Molecular Biophysics Unit, Indian Institute of Science, Bangalore-560012, India
Fax: +91 80 2360 0535
Tel: +91 80 2293 3000


Cys126 is a completely conserved residue in triosephosphate isomerase that is proximal to the active site but has been ascribed no specific role in catalysis. A previous study of the C126S and C126A mutants of yeast TIM reported substantial catalytic activity for the mutant enzymes, leading to the suggestion that this residue is implicated in folding and stability [Gonzalez-Mondragon E et al. (2004) Biochemistry43, 3255–3263]. We re-examined the role of Cys126 with the Plasmodium falciparum enzyme as a model. Five mutants, C126S, C126A, C126V, C126M, and C126T, were characterized. Crystal structures of the 3-phosphoglycolate-bound C126S mutant and the unliganded forms of the C126S and C126A mutants were determined at a resolution of 1.7–2.1 Å. Kinetic studies revealed an approximately five-fold drop in kcat for the C126S and C126A mutants, whereas an approximately 10-fold drop was observed for the other three mutants. At ambient temperature, the wild-type enzyme and all five mutants showed no concentration dependence of activity. At higher temperatures (> 40 °C), the mutants showed a significant concentration dependence, with a dramatic loss in activity below 15 μm. The mutants also had diminished thermal stability at low concentration, as monitored by far-UV CD. These results suggest that Cys126 contributes to the stability of the dimer interface through a network of interactions involving His95, Glu97, and Arg98, which form direct contacts across the dimer interface.

Structural data are available in the Protein Data Bank under the accession numbers 3PVF, 3PY2, and 3PWA.

Structured digital abstract 


glyceraldehyde 3-phosphate


dihydroxyacetone phosphate


Protein Data Bank


Plasmodium falciparum triosephosphate isomerase




triosephosphate isomerase


melting temperature


The conserved amino acids in enzymes are, most often, associated with the key steps of substrate recognition and catalysis. The availability of rapidly expanding databases of enzyme sequences may be effectively used to identify key residues. Triosephosphate isomerase (TIM) is an extremely well-studied enzyme [1–4], and provides a good model system for exploring the role of residues that are completely conserved or minimally replaced during evolution. Examination of a dataset of 503 sequences of TIM from different organisms reveals only nine fully conserved residues: Lys12, Thr75, His95, Glu97, Cys126, Glu165, Pro166, Gly209, and Gly228 [the numbering scheme used here corresponds to that for Plasmodium falciparum TIM (Pf TIM), and, for all of the fully conserved residues, this is identical to that of yeast TIM]. Of these, Lys12, His95, Glu97 and Glu165 surround the substrate, with the carboxylate of Glu165 acting as the base for abstraction of a proton from the C2 position of glyceraldehyde 3-phosphate (GAP) and dihydroxyacetone phosphate (DHAP) [5–8]. Lys12 and His95 are involved in substrate/transition state binding and proton transfer, respectively [6,9,10]. Pro166 is a hinge residue located in loop 6, which undergoes dynamic interconversion between open and closed states, with the latter corresponding to the catalytically competent form [11–15]. Gly209 is located near the active site in the highly conserved 208–212 segment. Gly228 adopts a backbone conformation accessible only for Gly residues, enabling appropriate positioning of the facing 208–209 segment by backbone–backbone hydrogen bonds. Thr75 is a critical residue at the dimer interface [16]; the side chain of this residue from one subunit makes key hydrogen bonding contacts with Asn10 and Glu97 of the other subunit, which are proximal to the active site. Cys126 is a completely conserved residue that is spatially proximal to the active site residue Glu165 (Fig. S1).

Interestingly, a preliminary analysis of a dataset of over 800 putative TIM sequences extracted from a dataset of bacterial sequences of marine origin [17] also revealed the occurrence of Cys at position 126. Inspection of several 3D structures of TIM from diverse organisms available in the Protein Data Bank (PDB) does not immediately suggest a structural explanation for the complete conservation of this residue. Indeed, an earlier investigation of the C126S and C126A mutants of yeast TIM revealed that their activity remained undiminished, with the mutants displaying a significantly lower degree of thermal stability. This study suggested that Cys126 may be required for efficient folding and stability rather than being involved in maintaining catalytic activity [18]. A recent treatise on enzymology highlights Cys126 in a discussion of TIM [19]. As part of a program directed towards understanding the role of conserved residues, we describe the characterization of five Cys126 mutants of Pf TIM. The mutants studied were C126S, C126A, C126V, C126M, and C126T. We describe crystal structures of unliganded forms of the C126S and C126A mutants, and the liganded form of the C126S mutant. Temperature-dependent activity measurements and spectroscopic studies suggest that Cys126 may be involved in maintaining the structural integrity of the active site in the temperature range 40–50 °C. Furthermore, the residue also contributes to the thermal stability of the dimer interface through an extended interaction network involving His95, Glu97 and Thr75 of the neighboring subunit, all of which are fully conserved residues.


Analysis of crystal structures

Diffraction-quality crystals were obtained for the C126S mutant complexed with phosphoglycolate (PGA) and the unliganded C126S mutant. For the C126A mutant, a structure could be determined only for the unliganded form. PGA was bound to the active site of the C126S mutant structure in a manner similar to that for wild-type Pf TIM, whereas the C126A mutant structure had no ligand bound to the active site after cocrystallization. The difference in electron density at the ligand position is shown in Fig. 1A. Figures were generated with pymol ( The active site loop 6 was in the ‘closed’ form in the structure of the C126S–PGA complex. In the unliganded forms of the C126S and C126A mutants, both of which contained a dimer in the asymmetric unit, the active site loop 6 was in the ‘open’ conformation. In the C126S mutant unliganded structure, the active site was occupied by an ethylene glycol molecule and a single water molecule in one subunit. In addition, a proximal sulfate ion, derived from the lithium sulfate in the crystallization medium, could also be identified near the active site. The other subunit in the C126S mutant and both subunits in the C126A mutant contained two water molecules in the active site, along with a distal sulfate ion. The electron density maps (2Fo − Fc, contoured at 1.0σ) surrounding the residues at position 126 for the mutants are shown in Fig. 1.

Figure 1.

 The electron density maps (2Fo − Fc contoured at 1.0σ) at position 126 and the active site residues in: (A) the Pf TIM C126S PGA-bound structure; (B) the Pf TIM C126S-unliganded structure with the molecule ethylene glycol (cryoprotectant) at the active site; and (C) the Pf TIM C126A-unliganded structure.

Figure S2 compares the relationships between the active site residues and Cys/Ser126 in the unliganded and liganded forms of the wild-type enzyme and the C126S mutant. The most notable difference is in the orientation of the Ser side chain, with the hydroxyl group forming a hydrogen bond with the carboxylate of Glu165 in the liganded form. A change of χ1 from −62.5° in the unliganded form to −170.8° in the liganded form is observed. In contrast, the Cys126 side chain remains unchanged in orientation upon ligand binding. Interestingly, both the unliganded forms contain two invariant water molecules, which form hydrogen bonds with one another. Water 512 in the wild-type enzyme (PDB ID: 1LYX) and water 349 in the C126S mutant also form hydrogen bonds with the fully conserved His95 and highly conserved Asn10 (Asn in 465 out of 470 sequences) side chains. Water 558 in wild-type TIM (PDB ID:1LYX) and water 36 in the C126S mutant form hydrogen bonds with the side chain of the active site Glu165 and the backbone CO of the fully conserved Gly209. These two invariant water molecules form a similar network of interactions in the C126A unliganded structure and also in the previously reported, unliganded yeast structure (PDB ID:1YPI) [20]. Ligand binding and loop closure result in the expulsion of these water molecules and a change in the backbone conformational angles for the highly conserved Gly209-Gly210-Ser211 segment. This results in a change in orientation of the Gly209 backbone CO group.

Kinetic parameters

The kinetic parameters determined for Pf TIM and the five mutants at position 126 are listed in Table 1. The parameters determined for the wild-type yeast enzyme and the C126S and C126A mutants by Gonzalez-Mondragon et al. [18] are also shown for comparison. In the earlier study of the yeast enzyme, the wild-type and the Cys126 mutant enzymes had comparable kinetic parameters, with a small reduction in kcat (approximately four-fold). Temperature-dependent activity measurements were not reported in that study. In the present study of Pf TIM, an approximately 5.8-fold drop in kcat was observed for both the C126S and C126A mutants. The other three mutants, C126V, C126M, and C126T, showed significantly lower kcat values, corresponding to a reduction of approximately 10-fold in catalytic activity. These results suggest that all five Cys126 mutants show a high degree of catalytic activity, despite the fact that a completely conserved residue, proximal to the active site Glu165 and His95 side chains, has been replaced by residues of varying size and hydrogen-bonding ability. Figure 2A compares the temperature dependence of the specific activity of wild-type Pf TIM and the five Cys126 mutants, at a protein concentration of 3.7 nm. For the wild-type enzyme, there was the expected increase in activity over the temperature range 25–40 °C, with a leveling off between 40 °C and 50 °C. In sharp contrast, all five mutants showed a dramatic reduction in activity in the temperature range 40–50 °C, with essentially complete absence of activity at 50 °C. The activities of the wild-type enzyme and the five Cys126 mutants were also measured as a function of protein concentration at 50 °C. The results summarized in Fig. 2B establish that all of the Cys126 mutants exhibited a pronounced fall in activity upon lowering of the protein concentration to below 20 μm. Indeed, a fall in activity of approximately 10–1000-fold was observed on the change from 30 μm to 1 μm. The pronounced concentration dependence of activity in the Cys126 mutants is suggestive of diminished stability of the dimeric protein at high temperature.

Table 1.   Kinetic parameters of Pf TIM and its five Cys126 mutants.
Enzymekcat (s−1)Km (mm)kcat/Km (mm−1·s−1)
  1. a P. falciparum (present study). b S. cerevisiae [18].

Wild typea(4.3 ± 0.3) × 1030.35 ± 0.051.2 × 104
C126Sa(7.5 ± 0.1) × 1021.4 ± 0.205.4 × 102
C126Aa(7.7 ± 0.2) × 1021.5 ± 0.205.2 × 102
C126Va(1.6 ± 0.8) × 1021.0 ± 0.101.6 × 102
C126Ma(1.9 ± 0.3) × 1021.5 ± 0.101.2 × 102
C126Ta(3.3 ± 0.2) × 1021.2 ± 0.202.8 × 102
Wild typeb(4.7 ± 0.7) × 1031.1 ± 0.44.3 × 103
C126Sb(1.1 ± 0.2) × 1030.3 ± 0.13.7 × 103
C126Ab(3.1 ± 0.2) × 1030.8 ± 0.33.9 × 103
Figure 2.

 (A) Temperature dependence of specific activity of wild-type Pf TIM (TWT) and the Cys126 mutants. (B) Concentration dependence of specific activity of the five Cys126 mutants at 50 °C.

Structural stability

Figure 3A,B show the far-UV CD and fluorescence emission spectra of Pf TIM and the five Cys126 mutants, determined at a protein concentration of 3 μm. The near identity of the observed spectra established that there were no dramatic structural consequences of the mutations at position 126. The far-UV CD spectra also remained unchanged over the concentration range 0.5–15 μm at 25 °C, suggesting the absence of any concentration-dependent structural effects at ambient temperature. Figure 3C shows a comparison of the thermal melting profiles for wild-type Pf TIM and the five mutants, obtained by monitoring the CD ellipticity at 222 nm as a function of temperature, at a protein concentration of 15 μm. The sharp reduction in CD ellipticity at temperatures greater than 60 °C corresponds to unfolding, aggregation, and precipitation. The wild-type and the mutant enzymes behaved in a very similar way under these conditions. These results suggest that replacement of Cys at position 126 does not significantly perturb the overall folded structure of the protein or its thermal stability, at this relatively high protein concentration. However, when the protein concentration was reduced to 0.5 μm, the melting curves determined using the fall in ellipticity at 222 nm (shown in Fig. 3D) were dramatically different for the wild-type enzyme and the mutants. The melting temperature (Tm) for wild-type Pf TIM was unaffected by lowering the concentration, whereas the mutants melted at a significantly lower temperature (midpoint of transition, 50 °C). This concentration dependence of protein thermal stability is consistent with the fall in enzyme activity of the mutants at low concentration and high temperature. The reversibility of the thermal unfolding transition was investigated by measurements of ellipticity at 222 nm upon cooling from a temperature of 55 °C for the mutants and 60 °C for the wild-type enzyme, at a protein concentration of 0.5 μm. Under these conditions, aggregation and irreversible precipitation of the thermally unfolded protein structure was minimized. Figure 4 summarizes the results obtained for the heating and cooling cycles for the wild-type enzyme and the five Cys126 mutants. Wild-type Pf TIM recovered almost 90% of the original ellipticity upon cooling to 20 °C. The observed hysteresis in the cooling cycle has also been previously noted for the wild-type enzyme from Saccharomyces cerevisiae [18,21]. In the case of all five mutants, only 60% of the CD ellipticity was recovered upon cooling. These results correspond well with those reported in the previous study of yeast TIM C126S and C126A mutants. Gonzalez-Mondragon et al. have noted that the reduction in Tm observed for the C126S and C126A mutants of the yeast enzyme ‘should be taken as an indication of diminished kinetic, rather than thermodynamic, stability of the native dimer’ [18]. They have also presented evidence for the dependence of refolding rates of the C126S yeast mutant at 30 °C and enzyme concentrations of 1.1 μm and 1.9 μm. More rapid refolding is observed at higher protein concentrations [18]. Our present study points to a greater tendency of the Cys126 mutants than of the wild-type enzyme to dissociate at low concentrations and high temperatures.

Figure 3.

 CD and fluorescence spectra of wild-type Pf TIM (TWT) and the five Cys126 mutants. (A) Far-UV CD, protein concentration 15 μm, 25 °C. (B) Fluorescence spectra, protein concentration 3 μm, 25 °C. (C) Thermal melting profile monitored at 222 nm, pathlength 1 mm, and protein concentration 15 μm. (D) Thermal melting profile monitored at 222 nm, pathlength 1 cm, and protein concentration 0.5 μm. All spectra were recorded in 20 mm Tris/HCl (pH 8.0).

Figure 4.

 Thermal unfolding and refolding study on wild-type Pf TIM and the five Cys126 mutants. A protein concentration of 0.5 μm was used in 20 mm Tris/HCl (pH 8.0). Ellipticity changes at 222 nm were monitored with heating and cooling rates of 0.5 °C·min−1. The cooling cycle was started immediately after completion of the denaturation transition. Black line: unfolding. Gray line: refolding.

The relative stability of the wild-type enzyme and the five mutants with respect to guanidinium chloride-induced and urea-induced perturbation was probed by measuring the position of fluorescence maxima. Unfolding results in a shift in the emission maximum from 328 to 355 nm. It is evident from the data in Fig. 5 that all five mutants were significantly less stable to urea-induced and guanidinium chloride-induced denaturation. The observed Cm values (midpoint of transition) for guanidinium chloride-induced denaturation were 1.7 m for wild-type Pf TIM and ∼ 1.0 m for all five Cys126 mutants; in the case of urea-induced denaturation, the Cm for wild-type Pf TIM was > 8 m, and that for all five Cys126 mutants was ∼ 4 m. The precise nature of the side chain at position 126 did not appear to have a significant influence, with all of the mutants exhibiting very similar unfolding transitions, suggesting that the Cys side chain is unique in imparting local stability.

Figure 5.

 Unfolding study on wild-type Pf TIM (TWT) and the five Cys126 mutants in the presence of urea and guanidinium chloride, by fluorescence. Protein at a concentration of 3 μm was incubated with different concentrations of urea and guanidinium chloride in 20 mm Tris/HCl (pH 8.0) for 45 min. The data for unfolding were normalized by taking the spectroscopic parameter to be 100% in the absence of any denaturant.


We began this study with the intention of establishing the role of the completely conserved Cys126 in the structure and function of TIM. In a previously reported study of S. cerevisiae TIM, Gonzalez-Mondragon et al. had concluded that Cys126 ‘is required not for enzymatic activity but for folding and stability’ [18] Their studies of the C126S and C126A mutants of the yeast enzyme established that these mutations had little effect on enzymatic activity, but resulted in greater susceptibility to thermal denaturation. In addition, the mutations slowed down the folding rate by a factor of 10. We have now re-examined the C126S and C126A mutants of Pf TIM, and determined their 3D structures by X-ray diffraction, in order to gain further insights into the structural consequences of mutations at position 126. We have also compared the kinetic and biophysical properties of three additional mutants: C126V, C126M, and C126T. The C126S and C126A mutants show a five-fold drop in kcat, whereas the other three mutants show a 10-fold drop. The observation of significantly high catalytic rates in all five mutants suggests that the conservation of Cys126 cannot be directly attributed to the imperatives of catalysis.

Our results clearly establish that the temperature dependence of enzyme activity is strongly concentration-dependent. At a temperature of 50 °C, the measured activity of all of the mutants show a concentration dependence over the range 1–20 μm. At low concentrations (3.7 nm), whereas the wild-type enzyme does not show marked temperature dependence over the range 40–50 °C, all of the mutants show a sharp loss in activity beyond 40 °C. Biophysical studies also confirm a concentration dependence of thermal stability, as probed with CD ellipticities at 222 nm. The mutants are significantly less stable with respect to thermal unfolding at low protein concentrations. Furthermore, the mutants are also much more structurally labile at appreciably lower concentrations of the denaturants urea and guanidinium chloride than the wild-type enzyme. These results lead to the conclusion that mutation at position 126 must cause a destabilization of subunit interactions, despite the apparent noninvolvement of this residue in any direct contacts across the dimer interface. We therefore turned to a re-examination of the structures of wild-type Pf TIM and the C126S and C126A mutants and the yeast enzyme DHAP complex reported by McDermott et al. (PDB ID: 1NEY) [22].

From Fig. S2, it can be seen that Cys126 closely approaches two active site residues, His95 and Glu165. The shortest contact distances lie between 4.0 and 4.5 Å in the case of the Pf TIM–PGA complex. In the ligand-bound C126S mutant structure, the serine OH group swings away from His95, in order to form a hydrogen bond with the carboxylate of Glu165. Figure 6 provides a view of the environment of Cys126, illustrating a network of interactions that connect this site to key residues at the subunit interface. The Cys126 backbone CO and NH groups are held by a pair of hydrogen bonds to the Arg99 guanidine side chain and the backbone CO of Ile93, respectively. The CO group of the fully conserved Gly94 is also held by a second guanidine group on the side chain of Arg99. The Cβ methylene group of Cys126 is in close proximity to Gly94 (3.71 Å). Arg 99 is also a very highly conserved residue, and is found in as many as 464 of 470 bacterial and eukaryotic sequences. Crucial hydrogen bond interactions across the subunit interface are made between the carboxylate of the fully conserved Glu97 and the Cβ hydroxyl of the fully conserved Thr75 from the other subunit. The guanidine group of Arg98 of one subunit also forms hydrogen bonds with the backbone CO of Thr75 and the side chain carboxylate of Glu77. The residues at positions 98 and 77 are also strongly conserved. Arg98 occurs in 441 of 470 sequences in our dataset, whereas, at position 77, Glu is observed in 409 examples and Asp in 51 examples from 470 sequences.

Figure 6.

 Environment of Cys126 in Pf TIM (PDB ID: 1O5X), showing the important network of hydrogen bond interactions involving subunit interface residues. Thr75 and Glu77 are from the other subunit.

Figure 7 shows a view of the environment of the Cys126 side chain. The thiol group of Cys126 does not appear to be involved in any significant hydrogen-bonding interaction. The closest potential hydrogen bond acceptors are the backbone carbonyl oxygen atoms of Ile93 (S–O=C: 4.12 Å) and Ile124 (S–O=C: 4.39 Å). A similar observation has been made in the atomic resolution structure of Leishmania mexicana TIM (0.83 Å), where the distances are as follows: 3.91 Å for S(Cys126)–O=C(Leu93); and 4.17 Å for S(Cys 126)–O=C(Ile124) [23]. No evidence for the involvement of the Cys126 thiol group in strongly directional hydrogen bond interactions is obtained from the crystal structures of TIMs from diverse organisms. The three proximal side chains are those of Glu165, His95, and Ile92. The closest distances of approach involving the thiol sulfur atom are 3.85 Å for S(Cys126)–OOC(Glu165), 4.21 Å for S(Cys126)–Cδ2(His95), (Fig. S1) and 4.10 Å for S(Cys126)–Cγ2H3(Ile 92) (Fig. 7). The corresponding residues are shown in the same orientation in the C126S–PGA complex structure. It is evident that the only difference is with respect to the orientation of the Ser126 hydroxyl group. The absence of any significant change in the relative orientations of His95 and Glu165 is consistent with the relatively high kcat values determined for the mutants at ambient temperature. However, creation of a cavity at position 126 in the case of the mutants (as shown in Fig. 7) may be expected to result in enhanced flexibility of the fully conserved Gly94-His95 segment, with the possibility of greater variability of the His95 side chain conformations upon heating.

Figure 7.

 View of the Cys/Ser/Ala126 side chain with 92, 94, 95 and 165 residues. (A) Wild-type Pf TIM PGA-bound structure (PDB ID: 1LYX). (B) C126S PGA-bound structure. (C) C126S-unliganded structure. (D) C126A-unliganded structure.

The structural data provide a possible explanation for the observed instability of the dimeric structure in the Cys126 mutants at elevated temperature. Perturbation of dimer interface contacts may be mediated by altered interactions between His95 and Glu97, and also through the Arg98-Arg99 segment (Fig. 6). The space-filling interactions involving the side chain of Cys126 (Fig. 7) appear to be critical in maintaining the observed network of hydrogen-bonding interactions, which must contribute to the stability of both active site residue orientation and subunit interface structure. Complete conservation of Cys126 suggests that selective pressures for optimal dimer stability at low concentrations and physiological temperatures may have been operative during the evolution of TIM sequences.

Experimental procedures


The Pf TIM gene was cloned into the pTrc99A vector pARC1008 [24]. The protein was overexpressed in Escherichia coli strain AA200, which has a null mutation for the host TIM gene [25]. For the present study, the five single mutants at position 126 were constructed by site-directed mutagenesis with the single primer method [26]. A single primer was sufficient to generate mutant ssDNA, which was subsequently transformed into E. coli DH5α cells to finally obtain the plasmid DNA with the desired mutation. As only one primer was used to achieve the mutation, the mutation site lies in the middle of a stretch of oligonucleotides, with sufficient flanking residues to obtain a high Tm, close to 78 °C. A primer length of 35-mer to 40-mer was successfully used to obtain the required mutations. The thermostable proofreading polymerase enzyme Pfu was used. The PCR mixture contained, in a total volume of 25 μL: template DNA, 150 ng; mutagenic primer, 20 pmol; thermostable polymerase buffer (× 10), 2.5 μL; dNTPs, 6 μL of a solution containing 2.5 mm each dNTP; and polymerase, 2.5 U. The cycling conditions for the PCR were as follows. The PCR tube was initially taken to 95 °C for 5 min, and then 40 cycles consisting of 1 min at 95 °C, annealing at 45 °C for 1 min and extension at 72 °C for 10 min were applied. Following this, a final extension at 72 °C for 20 min was applied. One microliter of DpnI (equivalent to 10 U) was directly added to the reaction mixture and incubated for 6–8 h at 37 °C, to digest the methylated template (parent) DNA. Ten microliters of the reaction mix was directly transformed into chemically competent DH5α cells, after which the presence of mutations was confirmed by restriction digestion and sequencing. In this study, five mutations were constructed at the same position. Because of the absence of a restriction site at the desired mutation position, a two-step process was followed: step 1, generating an intermediate clone, C126int, with the introduction of EcoRV restriction site at the desired mutation position; and step 2, taking C126int as the template and generating the mutant clones C126S, C126A, C126V, C126M, and C126T, with the subsequent removal of the EcoRV restriction site at the desired mutation position. The primer used for generating the C126int clone, with the introduction of the EcoRV restriction site, was 5′-TAATTTAAAAGCCGTGATATCTTTTGGTGAATCTT-3′, and the primers used for generating the five mutants were: C126S, 5′-TAATTTAAAAGCCGTTGTATCCTTTGGTGAATCTT-3′; C126A, 5′-TAATTTAAAAGCCGTTGTAGCTTTTGGTGAATCTT-3′; C126V, 5′-TAATTTAAAAGCCGTTGTAGTTTTTGGTGAATCTT-3′; C126M, 5′-TAATTTAAAAGCCGTTGTAATGTTTGGTGAATCTT-5′; and C126T, 5′-TAATTTAAAAGCCGTTGTAACTTTTGG TGAATCTT-3′.

Protein expression and purification

The TIM gene carrying the mutation was expressed in E. coli AA200 (a null mutant for the inherent TIM gene) cells carrying the pTrc99A recombinant vector. Cells were grown at 37 °C in Terrific broth, containing 100 μg·mL−1 ampicillin. Cells were induced with 300 μm isopropyl thio-β-d-galactoside at a D600 nm of 0.6–0.8, harvested by centrifugation at 4 °C, resuspended in lysis buffer containing 20 mm Tris/HCl (pH 8.0), 1 mm EDTA, 0.01 mm phenylmethanesulfonyl fluoride, 2 mm dithiothreitol, and 10% glycerol, and disrupted by sonication. After centrifugation (7245 g, 15 min, 4 °C) and removal of cell debris, the supernatant was fractionated with ammonium sulfate. The protein fraction containing TIM was precipitated between 60% and 80% ammonium sulfate saturation. The precipitate was obtained by centrifugation (19 320 g, 45 mins, 4 °C), and after resuspension in buffer A (20 mm Tris/HCl (pH 8.0), 2mm dithiothreitol, and 10 % glycerol), the following steps were followed. Firstly, it was subjected to gel filtration chromatography (Sephacryl-200), equilibrated with the same buffer A. The fractions containing the protein were pooled and further purified by anion exchange (Q-Sepharose) chromatography, with a linear gradient of 0-1 m NaCl. The purified protein obtained was then extensively dialyzed overnight against buffer A at 4 °C. Protein purity was checked by 12% SDS/PAGE. Mutations were confirmed by ESI MS: mobs (mcalc): wild-type TIM, 27 831 Da (27 831 Da); C126S, 27 815.7 Da (27 815 Da); C126A, 27 799.8 Da (27 799 Da); C126V, 27 827.2 Da (27 827 Da); C126M, 27 859.6 Da (27 859 Da); and C126T, 27 829 Da (27 829 Da) (Fig. S3). The protein concentration was determined with the Bradford method [27], using BSA as a standard.

Enzyme activity

Enzyme activity was measured by a coupled assay method. The conversion of GAP to DHAP by TIM was monitored in the presence of the coupling enzyme, α-glycerol phosphate dehydrogenase [28]. Enzymes were freshly prepared in 100 mm triethanolamine-HCl (pH 7.6). The reaction mixture contained (final volume, 1 mL) 100 mm triethanolamine-HCl, 5 mm EDTA, 0.5 mm NADH and 20 μg·mL−1α-glycerol phosphate dehydrogenase and GAP, to which TIM was added to initiate the reaction. In the case of the wild-type enzyme, the assay was started by addition of 10 ng of protein, and in the case of the Cys126 mutants 100 ng was used. Substrate concentrations varied from 0.25 mm to 4.0 mm. The progress of the reaction was monitored by the decrease in absorbance of NADH at 340 nm. The extinction coefficient of NADH was taken to be 6220 m−1·cm−1 at 340 nm [29]. The initial rates showed a linear dependence on the enzyme concentration in the range studied. This ensures the validity of the assay [28]. The values for the kinetic parameters (Km, kcat) were determined by fitting to the Michaelis–Menten equation with graphpad prism (Version 5 for windows; graphpad Software, San Diego, CA, USA;

Fluorescence spectroscopy

Fluorescence emission spectra were recorded on a HITACHI-250 spectroflorimeter. The protein samples were excited at 295 nm, and the emission spectra were recorded from 300 nm to 400 nm. Excitation and emission bandpasses were kept as 5 nm and 10 nm, respectively. Denaturation studies were performed by incubating 3 μm protein with different concentrations of urea and guanidinium chloride for 45 min. Spectra were acquired from 300 nm to 400 nm, after excitation at 295 nm.


Far-UV CD measurements were carried out on a JASCO-715 spectropolarimeter equipped with a thermostatted cell holder. The temperature of the sample solution in the cuvette was controlled with a Peltier device. For thermal melting studies, ellipticity changes at 222 nm were monitored. The temperature was varied at a rate of 0.5 °C·min−1 to follow the unfolding and refolding transitions. Spectra were averaged over four scans at a scanning speed of 10 nm·min−1. The change of ellipticity was measured as a function of temperature for thermal melting. Individual spectra (250–200 nm) were averaged over four scans.

Crystallization of Pf TIM Cys126 mutants

The Cys126 mutants were purified as described, and concentrated to approximately 10 mg·mL−1. Crystals were allowed to grow by the hanging drop method, at 23 °C [30]. The C126S–PGA crystal was obtained under the following conditions: 20% poly(ethylene glycol), 1 m Hepes buffer (pH 7.5), and 10 mm lithium sulfate. The unliganded C126S crystal was obtained under the following conditions: 24% poly(ethylene glycol), 1 m Hepes buffer (pH 7.0), and 10 mm lithium sulfate. The unliganded C126A crystal was obtained under the following conditions: 24% poly(ethylene glycol), 1 m Hepes buffer (pH 7.0), and 10 mm lithium sulfate. The crystals appeared within 2 days, and grew to the required sizes within 4–5 days.

Data collection and processing

Ethylene glycol (20%) was used as the cryoprotectant before flash-freezing of the crystals. X-ray diffraction data were collected with a Rigaku rotating anode generator and a MAR Research image plate detector system. The data were processed with mosflm and scala [31] of the ccp4 suite of programs [32]. The details of the datasets collected and the data collection statistics are shown in Table 2.

Table 2.   Data collection statistics.
  1. a Values in parentheses correspond to the last resolution shell.

Space groupC121P21212P21212
Unit cell
 a (Å)87.550.851.0
 b (Å)63.1173.8175.4
 c (Å)53.253.554.2
α (°)909090
β (°)117.239090
γ (°)909090
Resolution range (Å)26.2–1.743.4–1.951.8–2.0
No. of reflections53 371208 875297 253
No. of unique reflections25 34131 24329 675
Completion (%)a94.2 (79.7)85.6 (78.0)93.2 (75.3)
Overall R merge(%)a3.3 (22.5)8.2 (25.2)8.2 (26)
Multiplicitya2.1 (2.1)6.67(6.67)10 (10.3)
<I>/<σI>a20.2 (5.3)17.1 (8.3)20.2 (8.9)
Average mosaicity0.440.330.47

Structure solution and refinement

The mutant structures were solved with the molecular replacement program phaser of the ccp4 package [33]. The native Pf TIM crystal structure (PDB ID: lLYX) was used as the starting model for structure determination for the datasets of C126S-liganded. The structure with the PDB ID of 1O5X was used as the starting model in the case of the datasets for C126S-unliganded and C126A. The coordinates of 1LYX and of 1O5X were modified by removing the loop 6 residues, ligand, water molecules, and alternative conformations. Refinements of all the structures were carried out with refmac [34], with an initial 20 cycles of rigid body refinement followed by 50 cycles of restrained refinement. The loop 6 residues, ligand and water molecules were added on the basis of 2Fo − Fc and Fo − Fc maps contoured at 1σ and 3σ, respectively. Model building was performed with coot [35]. One subunit in the case of the C126S-liganded structure and two subunits in the case of the C126S-unliganded and C126A structures were present in the asymmetric unit. The existence of the C126S and C126A mutations was confirmed from difference Fourier maps. Water molecules were first located automatically by coot, and validated if a peak was observed above 3σ on a difference map and above 1.5σ on a double difference map. The B-factors of all atoms were also refined, and alternative conformations were included wherever necessary. All of the structures were refined to reasonable Rwork and Rfree values and good geometry, and then validated with procheck [36] in the ccp4 package. The electron density maps (2Fo − Fc contoured at 1.0σ) surrounding the residues at position 126 for the mutants are shown in Fig. 1. The refinement statistics for the mutant structures are shown in Table 3.

Table 3.   Refinement statistics.
Resolution range (Å)26.2–1.743.4–1.951.8–2.0
Number of subunits/asymmetric unit122
Number of used reflections25 34131 29729 723
% observed94.485.893.1
Rwork (%)
Rfree (%)
Model quality
 Number of atoms234945374401
 Number of water molecules343616491
 Number of ligand atoms (PGA)9
Average B-factor (Å2)
 Ligand (PGA)9.5
rmsd from ideal
 Bond length (Å)0.0260.0060.006
 Bond angle (°)2.00.8970.908
Ramachandran statistics
 Most allowed region (%)94.294.594.9
 Allowed region (%)
 Generously allowed region (%)000
 Disallowed region (%)000


One of us (P. Balaram) is deeply indebted to N. V. Joshi for his analysis of TIM sequences and helpful discussions. M. Samanta was supported by a Senior Research Fellowship from the Council of Scientific and Industrial Research (India). X-ray diffraction and MS facilities are supported by program grants from the Department of Biotechnology (India).