Probing the role of highly conserved residues in triosephosphate isomerase - Analysis of site specific mutants at positions 64 and 75 in the Plasmodial enzyme

Article abstract of DOI:10.1111/febs.13384

Highly conserved residues in enzymes are often found to be clustered close to active sites, suggesting that functional constraints dictate the nature of amino acid residues accommodated at these sites. Using the Plasmodium falciparum triosephosphate isomerase (PfTIM) enzyme (EC 5.3.1.1) as a template, we have examined the effects of mutations at positions 64 and 75, which are not directly involved in the proton transfer cycle. Thr (T) occurring at position 75 is completely conserved, whereas only Gln (Q) and Glu (E) are accommodated at position 64. Biophysical and kinetic data are reported for four T75 (T75S/V/C/N) and two Q64 (Q64N/E) mutants. The dimeric structure is weakened in the Q64E and Q64N mutants, whereas dimer integrity is unimpaired in all four T75 mutants. Measurement of the concentration dependence of enzyme activity permits an estimate of K_d values for dimer dissociation (Q64N = 73.7 ± 9.2 nm and Q64E = 44.6 ± 8.4 nm). The T75S/V/C mutants have activities comparable to the wild-type enzyme, whereas a fourfold drop is observed for T75N. All four T75 mutants show a dramatic fall in activity between 35 C and 45 C. Crystal structure determination of the T75S/V/N mutants provides insights into the variations in local interactions, with the T75N mutant showing the largest changes. Hydrogen-bond interactions determine dimer stability restricting the choice of residues at position 64 to Gln (Q) and Glu (E). At position 75, the overwhelming preference for Thr (T) may be dictated by the imperative of maintaining temperature stability of enzyme activity.

Full text of DOI:10.1111/febs.13384

EDITOR’S CHOICE
Probing the role of highly conserved residues in
triosephosphate isomerase – analysis of site specific
mutants at positions 64 and 75 in the Plasmodial enzyme
Debarati Bandyopadhyay1, Mathur R. N. Murthy1, Hemalatha Balaram2 and Padmanabhan
Balaram1
1 Molecular Biophysics Unit, Indian Institute of Science, Bangalore, India
2 Molecular Biology and Genetics Unit, Jawaharlal Nehru Centre for Advanced Scientific Research, Bangalore, India
Keywords
Highly conserved residues in enzymes are often found to be clustered close to
active sites, suggesting that functional constraints dictate the nature of amino
amino acid conservation; conserved
residues in enzymes; enzyme structure and
acid residues accommodated at these sites. Using the Plasmodium falciparum
triosephosphate isomerase (PfTIM) enzyme (EC 5.3.1.1) as a template, we
function; mutational effects on
triosephosphate isomerase; triosephosphate
have examined the effects of mutations at positions 64 and 75, which are not
directly involved in the proton transfer cycle. Thr (T) occurring at position
isomerase
75 is completely conserved, whereas only Gln (Q) and Glu (E) are accommo-
dated at position 64. Biophysical and kinetic data are reported for four T75
(T75S/V/C/N) and two Q64 (Q64N/E) mutants. The dimeric structure is
weakened in the Q64E and Q64N mutants, whereas dimer integrity is unim-
paired in all four T75 mutants. Measurement of the concentration depen-
dence of enzyme activity permits an estimate of Kd values for dimer
dissociation (Q64N = 73.7 ꢀ 9.2 nM and Q64E = 44.6 ꢀ 8.4 nM). The
T75S/V/C mutants have activities comparable to the wild-type enzyme,
whereas a fourfold drop is observed for T75N. All four T75 mutants show a
dramatic fall in activity between 35 °C and 45 °C. Crystal structure determi-
nation of the T75S/V/N mutants provides insights into the variations in local
interactions, with the T75N mutant showing the largest changes. Hydrogen-
bond interactions determine dimer stability restricting the choice of residues
at position 64 to Gln (Q) and Glu (E). At position 75, the overwhelming
preference for Thr (T) may be dictated by the imperative of maintaining tem-
perature stability of enzyme activity.
Correspondence
P. Balaram, Molecular Biophysics Unit,
Indian Institute of Science,
Bangalore 560012, India
Fax: +9180 2360 0535
Tel: +9180 2293 3000
E-mail: pb@mbu.iisc.ernet.in
(Received 12 June 2015, revised 8 July
2015, accepted 16 July 2015)
doi:10.1111/febs.13384
Database
Structural data have been deposited in the Protein Data Bank under accession numbers 4ZZ9,
5BMW, 5BMX, 5BNK and 5BRB.
Introduction
Protein sequences bear the imprint of the selective
pressures of evolution. Examination of sequences of
enzymes, derived from a large set of diverse organisms,
reveals that relatively few residues are completely con-
served across all species. Fully conserved residues are
likely to be critically important for maintaining opti-
Abbreviations
ANS, 1-anilino-8-naphthalene sulfonate; DHAP, dihydroxyacetone phosphate; GAP, glyceraldehyde-3-phosphate; LC, liquid chromatography;
Lm, Leishmania mexicana; PDB, Protein Data Bank; Pf, Plasmodium falciparum; TIM, triosephosphate isomerase; TWT, PfTIM wild-type.
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3863

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
mal catalytic activity, either by contributing directly to
the chemistry of catalysis or, more indirectly, by being
indispensable for folding, stability and oligomerization
of enzymes [1]. Sequence comparisons over large data-
sets can also reveal specific sites at which minimal
amino acid replacements are observed, suggesting that
functional selection imposes severe constraints on the
nature of the amino acid side chains that can be
accommodated at these positions [2]. The role of con-
sensus mutations at significantly conserved position on
protein thermal stability has been explored in an
extensive study of Saccharomyces cerevisiae triosephos-
phate isomerase (EC 5.3.1.1) [3,4]. Using the malarial
parasite Plasmodium falciparum (Pf) triosephosphate
isomerase (TIM) as a model system, we have been
exploring the role of highly conserved residues on
structure and function [5,6]. TIM is arguably one of
the best studied enzymes in the literature of biochem-
istry [7–12], earning the distinction of being labelled as
the ‘perfect enzyme’ [13]. Its central role in glycolysis
ensures that it is ubiquitous in nature. Recent studies
also highlight the nonglycolytic roles of glycolytic
enzymes, including TIM [14,15]. The growing database
of protein sequences derived from genomic data has
been used to generate a dataset of 3398 TIM sequences
from bacterial sources [16]. Ten residues (K12, T75,
H95, E97, C126, E165, P166, G209, G210 and G228)
are fully conserved (the numbering scheme used corre-
sponds to the sequences of the yeast and Pf enzymes).
Of these, the roles of K12, H95 and E165 have been
extensively explored in the seminal work of Knowles
and coworkers [17–24], which laid the foundation for
the current understanding of the mechanistic details of
the isomerization reaction [25]. In previous work, we
have explored the role of C126 and E97, using specific
mutational replacements. Although the former appears
to indirectly determine the stability of dimeric enzyme
[5], a direct role in the catalytic proton transfer cycle
has been suggested for the latter [6].
The integrity of the enzyme active site, which
lies close to the dimer interface, is determined by
inter-subunit interactions, making TIM an obligatory
dimer. Attempts to engineer monomeric TIMs yielded
well folded b8a8 TIM barrels with highly impaired
activity, with the catalytic rates dropping by three
orders of magnitude [27]. In the present study, we
assess the effects of mutations at residues 75 and 64
on TIM dimers using the Pf enzyme as a template.
Results
In the present study, four mutants at position 75
(T75S/T75V/T75C and T75N) and two mutants
at position 64 (Q64E and Q64N) were examined.
Figure 1 shows the interactions involving the side
chains of Thr75 and Gln64. The side-chain -OH group
of Thr75 makes two important inter-subunit contacts
with the Asn10 -NH2 and Glu97 carboxyl group,
whereas the backbone -NH and backbone -CO are
connected to the Glu97 carboxyl and Arg98 guani-
dinium group, respectively. The Gln64 side-chain -CO
forms an inter-subunit hydrogen bond to the Gly76
backbone -NH, while the side-chain -NH2 group forms
two hydrogen bonds with the backbone -CO groups of
Lys12 and Phe42. Inspections of these interactions sug-
gest that mutations at these sites might perturb dimer
interface stability and may also influence activity, in
view of the proximity of these residues to the active
site residues Lys12, His95 and Glu97 [5]. All mutant
proteins were characterized by ESI-MS (data not
shown) and their corresponding DNA sequences were
determined, to establish their identity.
Biophysical studies
Figure 2 shows the gel filtration profiles for both the
sets of mutant proteins. All four Thr75 mutants elute at
the same position as that of the wild-type enzyme, con-
firming that stable dimers are formed under the condi-
tions used. In contrast, both Q64E and Q64N mutants
elute more slowly from the column, suggesting that
dimer dissociation and equilibration on the column do
occur in these cases. CD spectra of the four Thr75
mutants showed that both near and far UV-CD bands
were very similar to those observed for the wild-type
protein (TWT) (data not shown). Figure 3 compares
the spectra for both the wild-type and T75V mutant. In
contrast, the Q64E and Q64N mutants possess distinc-
tively different spectral characteristics. In both cases,
there is almost complete abolition of the near UV-CD
band, centered at approximately 280 nm. Although it is
not possible to unambiguously ascribe this band to con-
In the present study, we address the role of the fully
conserved T75 residue, which is involved in key hydro-
gen-bond interactions across the dimer interface. We
also examine the consequences of residue replacement
at position 64, a site at which Q is present in 3011
sequences, whereas E is present in 383 sequences, in
the dataset examined. Q64 is involved in hydrogen-
bond interactions across the dimer interface and in
intra-subunit interactions with residues proximal to the
active site K12 residue. Interestingly, a study by
Williams et al. [26] on the E65Q mutant of the enzyme
from Leishmania mexicana (LmTIM; E65 in LmTIM is
structurally equivalent to Q64 in PfTIM) reveals a
large increase in thermal stability.
3864
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
Fig. 1. Environment of Gln64 and Thr75 in
PfTIM (PDB code: 1O5X). The important
network of hydrogen bonds in the dimer
interface; Gly72 to Glu77 (in cyan) is from
the other subunit. Inter-subunit hydrogen-
bonding interactions are shown in blue
and intra-subunit hydrogen-bonding
interactions are shown in red.
Fig. 2. Elution profiles for all six mutants (T75S/V/C/N; Q64N/E in
inset) and PfTIM wild-type (TWT). The protein concentration was
20 lM.
Fig. 3. CD spectra of TWT and the three mutants (Q64N/E; T75V).
(A) Far UV-CD (protein concentration: 5 lM) and (B) near UV-CD
(inset; protein concentration 30 lM).
tributions from a specific aromatic residue, it may be
noted that Trp (W) 11 is very close to the site of muta-
tion. Interestingly, the aromatic CD band is also lost in
the case of the wild-type protein at pH 3, a feature
noted in earlier studies [28]. The spectral differences
between TWT and Q64E/N mutants are less pro-
nounced in the far UV region, although there is a dis-
tinct reduction in the ellipticities at 220 nm. This region
contains contributions from both backbone amide and
side-chain aromatic group electronic transitions. The
Q64E and Q64N mutants also show significant binding
of the hydrophobic fluorescent probe 1-anilino-8-naph-
thalene sulfonate (ANS) at neutral pH (Fig. 4). The
wild-type enzyme and the Thr75 mutants do not cause
an enhancement of ANS fluorescence intensity. ANS
binding and loss of tertiary CD bands as a result of aro-
matic residues have been correlated in literature to mol-
ten globule state of proteins [29,30]. TIM has been
shown to bind ANS at low pH values of approximately
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3865

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
tion of pH, a dramatic enhancement between pH 3 and
4 is observed for the wild-type protein. A significantly
greater enhancement was observed for both mutants
between pH 4 and 4.5, suggesting that mutations at this
site destabilize the protein with respect to acid-induced
unfolding.
PfTIM undergoes irreversible thermal unfolding and
aggregation resulting in precipitation of the protein at
temperatures > 50 °C. The four T75 mutants and
Q64E behave like the wild-type enzyme in precipitating
at temperatures > 50 °C. The Q64N mutant exhibits a
thermal precipitation temperature that is significantly
lower, at approximately 42 °C. The thermal unfolding
and precipitation of these proteins are preceded by a
structural transition in the pre-melting region, which
results in distinct changes in CD spectra at tempera-
tures below the onset of precipitation. As noted earlier,
far UV-CD spectra of the wild-type protein and the
four Thr75 mutants are largely similar at ambient tem-
perature (23 °C). Small changes have been noted in
the case of Q64E and N mutants (Fig. 3A). Figure 5
compares CD spectra for wild-type protein and the
mutants at position 64 and 75 at temperatures, in the
temperature range 40–45 °C. In the case of T75N, a
blue shift of the long wavelength band (219 nm) is
observable, while the spectra for the wild-type and
T75C mutant are largely indistinguishable at 40 °C
(Fig. 5A). T75S and T75V mutants behave in a man-
ner similar to T75C and T75N, respectively. Upon
increasing the temperature to 45 °C, differences are
clearly observable for all the T75 mutants as compared
Fig. 4. Fluorescence emission spectra of bound ANS to TWT and
the Q64N/E mutants. The protein concentration was 5 lM and the
ANS concentration was 2 lM (kexcitation = 370 nm).
3, where an acid expanded state of the protein is
obtained [28]. The magnitude of the enhancement at
acidic pH is significantly more than that shown in
Fig. 4 (data not shown). The limited enhancement
observed in the case of Gln (Q) 64 mutants may be a
consequence of ANS binding at a relatively polar site at
the dimer interface, which becomes accessible upon per-
turbation of inter-subunit interactions. Further evi-
dence of ANS binding to the Q64E/N mutants is also
obtained by excitation of the intrinsic fluorescence of
the protein at 280 nm. Although no Trp-ANS energy
transfer is observed in the case of the wild-type protein,
significant transfer is observed for both mutants (data
not shown). When ANS binding is monitored as a func-
Fig. 5. Thermal melting effect on the
mutants. (A, B) On T75 mutants and (C, D)
on Q64 mutants at 40 and 45 °C. The
protein concentrations used were 15 lM.
Pathlength = 1 mm.
3866
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
to the wild-type enzyme (Fig. 5B). Figure 5C,D shows
a comparison of the CD spectra of TWT, Q64E and
Q64N. At 40 °C, a dramatic reduction of ellipticity in
the far UV-CD spectrum of Q64N is observed and
even more substantial loss of CD bands is noticeable
at 45 °C. The onset of precipitation in the case of
Q64N is at a temperature almost 10 °C lower than
wild-type protein. In the case of Q64E, small but sig-
nificant changes are observed at 40 °C, whereas, at
45 °C, there is a distinct change in the band shape pre-
sumably corresponding to a pre-melting structural
transition. Mutations at positions 64 and 75 destabilize
the three-dimensional structure of the enzyme with
respect to thermal perturbation. The Q64N mutation
appears to be the least thermally stable. The b8a8 bar-
rel fold in TIM has been previously shown to be
robust in the presence of high concentrations of urea
up to approximately 6 M [31]. All six mutants exam-
ined in the present study, unfold in the presence of
urea with a midpoint of the structural transition at
approximately 3.5–4 M (Fig. 6A) as monitored by shift
of the fluorescence emission maxima from approxi-
mately 330 nm (folded) to 348 nm (unfolded) and loss
of CD ellipticity at 222 nm (Fig. 6B).
The biophysical evidence presented thus far suggests
that mutations at positions 64 and 75 destabilize the
protein when exposed to thermal perturbation or dena-
turing agents, such as urea, at ambient temperatures
and near neutral pH. The four Thr75 mutants behave
in a manner similar to the wild-type enzyme, as indi-
cated by gel filtration studies and spectroscopic prop-
erties. Destabilization of subunit interactions appears
to result in the case of mutants at position 64.
Fig. 6. Urea denaturation study with Q64N/E and T75 mutants at
pH 8. (A) Monitoring fluorescence emission maxima and (B)
monitoring changes in ellipticity at 222 nm. Protein (5 lM) was
incubated with different concentrations of urea for 45 min. The
data monitored by CD spectra (B) were normalized by taking the
ellipticity at 222 nm in the absence of urea as 100%.
specific activity with respect to the wild-type enzyme.
The T75N mutant shows a fourfold reduction in activ-
ity. The Q64E and Q64N mutants were significantly
less active even at high enzyme concentrations
(150 nM), suggesting that both dimer destabilization
and conformational alteration at the active site may
contribute to the observed five- to 10-fold decrease in
specific activity. Figure 7 compares the concentration
dependence of specific activity for the Thr75 and
Gln64 mutants. Although no dependence of concentra-
tion is observed for the wild-type and Thr75 mutants
even in the range 0.1–5 nM, a relatively sharp depen-
dence is observed for Q64E and Q64N mutants in the
range 5–150 nM. These results suggest that the TIM
dimer is unaffected by the mutations at position 75,
whereas mutational changes at position 64 led to a
considerable loss of dimer integrity.
Enzyme kinetics
Measurements of enzyme activity were carried out
over a wide range of protein (0.15–150 nM) and sub-
strate (0.25–4.0 mM) concentrations. Saturation of the
measured specific activity was achieved at substrate
concentrations > 3 mM for the wild-type enzyme and
four Thr75 mutants (data not shown), permitting kcat
and Km values to be derived (Table 1). By contrast,
saturation of the measured specific activity could not
be achieved for the Q64E and Q64N mutants even at
D-glyceraldehyde-3-phosphate (D-GAP) concentrations
> 4 mM. The Gln64 mutants also had significantly
lower activities, necessitating the use of higher enzyme
concentrations in the assays (data not shown). Table 1
summarizes the measured activity and kinetic parame-
ters for the wild-type protein, Thr75 mutants and
Gln64 mutants. The T75S/V/C mutants were substan-
tially active and exhibited only marginal decreases in
Making the assumption that the monomeric species
is inactive and the measured activity may be ascribed
entirely to dimeric species, an estimate of the dimer
dissociation constant (Kd) for Gln64 mutants may be
made, as previously reported by Pray et al. [32] for a
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3867

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
Table 1. Kinetic parameters for TWT and all six mutants.
Specific activity (lmolꢃmgꢂ1ꢃminꢂ1
)
ꢂ1
Enzyme
kcat 9 (102 sꢂ1
)
Km (mM)
kcat/Km 9 (102 sꢂ1ꢃmM
)
[enzyme concentration]a
TWT
4.0 ꢀ 0.1
3.8 ꢀ 0.1
3.8 ꢀ 0.2
3.4 ꢀ 0.2
1.3 ꢀ 0.1
–
0.62 ꢀ 0.07
0.80 ꢀ 0.08
0.91 ꢀ 0.12
1.15 ꢀ 0.15
0.86 ꢀ 0.08
–
6.45
4.75
4.18
2.96
1.51
–
679.6 [5 nM]
T75S
T75V
T75C
T75N
Q64Nb
Q64Eb
573.3 [5 nM]
560.7 [5 nM]
502.4 [5 nM]
178.7 [5 nM]
134.1 (301.4) [150 nM]
63.3 (126.2) [150 nM]
–
–
–
a
Substrate concentration was 3 mM; assays performed at 23 °C at pH 7.6. Limiting specific activity values for Q64 mutants are given in
b
parenthesis. kcat and Km values could not be obtained since saturation of specific activity could not be achieved at a substrate concentra-
tion as high as 4 mM.
neered monomeric TIM exhibits a 1000-fold drop in
measured activity.
To separate the effects of mutations on the local
structure of the active site from the effects on dimer
stability, it is necessary to estimate the limiting specific
activity for the Gln64 mutants. This is readily achieved
by fitting the experimental data to the following equa-
tions [32] using PRISM, version 5 (GraphPad Software
Inc., San Diego, CA, USA) (Fig. 7A).
½Pr2ꢁ
Aspec ¼ Amax
½Prꢁ þ ½Pr2ꢁ
2 þ ½Prꢁ ꢂ E ¼ 0
½Prꢁ
2
tot
Kd
where Amax is the maximum specific activity, Aspec is
the measured specific activity in the assay, and [Pr]
and [Pr2] are the concentration of free monomer and
dimer in the solution. Etot represents the total enzyme
concentration used in each assay ([Pr] + 2[Pr2]); [Pr]
was derived in terms of Kd by using the second equa-
tion and eventually Amax and Kd were solved from the
first equation. These yield limiting values of
126.2 lmolꢃmgꢂ1ꢃminꢂ1 and 301.4 lmolꢃmgꢂ1ꢃminꢂ1
for Q64E and Q64N mutants, respectively. Thus, an
approximately 2.5-fold drop in activity is observed for
the Q64N mutant, whereas a significant greater drop
of fivefold was observed for the Q64E mutant.
Fig. 7. Dependence of specific activity on enzyme concentrations.
(A) For Q64 mutants and (B) for wild-type (TWT) and T75 mutants.
All assays were performed at pH 7.6; enzymes at different
concentrations were preincubated for 45 min in assay buffer prior
to initiation of the assay with 3 mM substrate (D-GAP).
Q64 mutants: pH effects
The replacement of Q at position 64 by E raises the pos-
sibility that the ionization state of the glutamic acid
side-chain influences the structure of the active site. In
the unionized form, the E (Glu) side chain might be
expected to minimally perturb the local network of
hydrogen bonds, whereas ionization to the carboxylate
viral protease. The Kd values thus obtained are
73.7 ꢀ 9.2 nM for Q64N and 44.6 ꢀ 8.4 nM for Q64E
mutants. These values are several orders higher than
the estimated value (0.01–0.0001 nM) for the wild-type
TIM [33,34]. The assumption of inactive monomers is
supported by Borchert et al. [27] reporting that engi-
3868
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
Structural analysis of Q64 mutants
Obtaining crystals diffracting to high resolution
ꢀ
(< 3.0 A) for the Q64E mutant was difficult. However,
ꢀ
single crystal, diffracting to 2.51 A, could be
a
obtained from which data of limited quality could be
collected. Structure solution and refinement provided a
view of the site of mutation (data not shown). The car-
boxylic acid side chain of E64 makes potential hydro-
gen bonds, with the backbone -NH of Gly76 of the
other subunit and with backbone carbonyl (-CO)
groups of Lys12 and Phe42 of the same subunit. These
are comparable to the hydrogen bonds observed in the
cases where Gln is in the position 64 (Fig. 9). The
Q64N mutant shows a significant fall in activity at pH
values > 8 with a transition that probably corresponds
to changes in the ionization state of the catalytic
Lys12 residue. Attempts to obtain diffracting quality
crystals of Q64N mutant proved futile. To assess the
possible structural consequences of the Q to N muta-
tion, the Asn side chain was modelled at position 64 in
the structure of the wild-type enzyme (Fig. 10). Of the
seven Asn side-chain rotamers in the Coot library
[35,36]; three rotamers showed unfavourable steric
contacts with neighbouring atoms (rotamers 2, 3 and
6). The modelled structures of the remaining four rota-
mers (rotamers 1, 4, 5 and 7) are shown in Fig. 10. In
all cases, the residue 64-Lys12 side chain-backbone
hydrogen bond is no longer present. Of the three key
hydrogen-bond contacts made by the Gln residue at
position 64 in the wild-type enzyme, only one, the side
chain-backbone hydrogen bond to the -CO group of
Phe 42, is maintained in three of the rotamers
(rotamer 1, 5 and 7). Replacement of Gln by Asn at
position 64 is thus likely to weaken the dimer interface
because of the absence of the hydrogen bond to Gly76
-NH of the neighboring subunit. Further, the loss of
the hydrogen bond to the Lys12 backbone -CO may
have an influence on activity. It may be noted that, in
all native TIM structures, Lys12 is held in an
Fig. 8. Specific activities of (A) TWT and (B) Q64N/E mutants at
different pH values; 5 nM TWT, 150 nM Q64 mutants and 2 mM
substrate (D-GAP) were used in the assay.
may be expected to have appreciable effects on the local
structure. Figure 8 compares the pH dependence of the
specific activity of the Q64E/N mutants. At pH values
below 7, both mutants have comparable activities. The
Q64E mutant shows a sharp fall in activity between pH
7 and pH 8, presumably due to titration of the E64 car-
boxylic acid group. The observed high pKa of approxi-
mately 7.5 may be a consequence of the stabilizing
effect of the hydrogen-bonded network on the proto-
nated form of glutamic acid (E) at position 64.
Fig. 9. Comparison of the hydrogen-
bonding interactions near the residue 64 in
(A) PfTIM wild-type (PDB code: 1O5X) and
(B) Q64E mutant (PDB code: 5BRB).
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3869

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
Fig. 10. Possible rotamers of modelled
Q64N on PfTIM wild-type structure (PDB
code: 1O5X).
3870
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
unusual backbone conformation (φ = 54.3 ꢀ 5.5 and
w = –144.1 ꢀ 7.0) [37]. This strained conformation is
stabilized by hydrogen bonds formed by the flanking
peptide units with proximal side chains. Our experi-
mental studies, detailed above, suggest a significantly
weakened dimer interface in the Q64N mutant result-
ing in concentration dependence of enzyme activity
and dissociation under gel filtration conditions. The
Q64N mutation has a significantly greater destabilizing
effect on the dimer than the Q64E mutation. A com-
parison of the limiting activity obtained at high pro-
tein concentration suggests that the Q64N mutant is
approximately twice as active as the Q64E mutant.
Both mutants exhibit similar activities at low pH val-
ues, suggesting that ionization of the E64 side chain in
the Q64E mutant probably disrupts the local confor-
mation of the active site residues as a result of the loss
of the hydrogen-bond interactions described above.
-NH2 of Asn75 and backbone -CO of Gln64 is
observed. In the T75C mutant, the electron density
near the site of mutation was poor, preventing the
building of a good structural model. These structural
results suggest that mutations at position 75 cause rela-
tively small perturbations confined to local structure.
Despite the proximity of Thr75 to the active site resi-
dues Lys12 and Glu97, the relative orientations of these
residues remain unchanged in all the mutants. The bio-
physical studies described earlier established that the
dimer integrity is undiminished upon replacement of
Thr (T) 75 by S, V, C and N. The T75S, T75V and
T75C mutants exhibit enzyme activities very close to
that obtained for the wild-type enzyme. Only the T75N
mutant shows
a
fourfold reduction in activity
(Table 1). A comparison was made between the struc-
tures of T75N and the wild-type enzyme, T75S and
T75V, around the site of mutation. This analysis pro-
vided an explanation for the diminished activity of the
T75N mutant. In the TWT structure, the oxygen atom
(-Oc) of the Thr side chain makes a close contact with
the -CcH2 of the active site K12 side chain (O—C;
Structural analysis of Thr75 mutants
Diffraction quality crystals were obtained for the four
mutants: T75S, T75V, T75C and T75N. Structures were
determined at a resolution of 1.80, 1.86, 1.80 and
ꢀ
3.7 A). In the T75S mutant, the corresponding distance
c
c
ꢀ
between Ser -O and Lys -C H2 is 3.6 A, whereas, in
ꢀ
1.80 A, respectively. Details of the diffraction data,
the T75V mutant, the contact distance between Val
-CcH3 and Lys -CcH2 is 3.4 A. In contrast, in the
T75N mutant, the shortest distances between the atoms
ꢀ
structure solution and refinement statistics for T75S/V/
N/C mutants are provided in Table 2. In all cases, unli-
ganded dimeric structures were obtained with the active
site loop 6 occurring in the open form. In all four T75
mutant structures, the active site was occupied by a
molecule of ethylene glycol, which was used as a cry-
oprotectant. The active site geometries for all four
mutants were almost identical (Fig. 11). Thr75 lies at
the dimer interface making hydrogen-bond contacts
with residues from the neighboring subunit. An exami-
nation of B-factor plots for the wild-type enzyme and
the Thr75 mutants did not reveal any significant differ-
ences for the loop residues in the vicinity of the site of
mutation. Figure 12 compares the environment of resi-
due 75 in the structure of the wild-type enzyme and
three mutants: T75V, T75S and T75N. In the serine
mutant, the primary hydroxyl group of the side-chain
forms hydrogen bond to the side-chain -CO of Asn10
and side-chain carboxylic group of Glu97 in a manner
similar to that observed with Thr at this position. In
the T75V mutant, both these inter-subunit hydrogen
bonds are lost and are replaced by van der Waals con-
tacts to the side chains of Asn10 and Glu97. The valine
side chain takes up an orientation almost identical to
that observed for Thr. The largest changes in dimer
interface interactions are observed for the T75N
mutant. In this case, the hydrogen bond to Glu 97 is
lost but a new hydrogen bond between the side-chain
ꢀ
of N75 side chain and Lys12 are 4.6 A (N75 side-chain
c
ꢀ
-CO and K12 side-chain -C H2) and 4.7 A (N75
-CbH2 and K12 side-chain -CcH2) (Fig. 13). The poorer
packing of the K12 side chain with the proximal N75
residue may contribute to greater conformational flexi-
bility, resulting in marginally diminished activity. The
accommodation of as many as four diverse residues at
position 75 with relatively limited effects on structure
and activity raises important questions. Why is Thr (T)
75 completely conserved in the very large dataset of
bacterial TIM sequences and, indeed, in all of the avail-
able archaeal and eukaryotic sequences? What is the
phenotype that imposes a strong constraint on the resi-
due selected at this site? In order to further understand
the possible reasons for conservation of Thr (T) 75 resi-
due, we aimed to determine the temperature depen-
dence of enzyme activity, as discussed below.
Temperature dependence of enzyme activity in
Thr (T) 75 mutants
Figure 14A shows the change in specific activity as a
function of temperature for the wild-type enzyme and
the mutants T75S and T75N. The temperature depen-
dence determined for the T75V and T75C mutants are
very similar to that observed for T75S (data not
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3871

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
Table 2. Crystallographic data collection and refinement statistics. Values given in parenthesis are for the highest resolution shell.
Mutant proteins
T75S
T75V
T75N
T75C
Q64E
PDB code
Space group
Unit cell
4ZZ9
P21
5BMW
P21
5BMX
5BNK
P21
5BRB
P21
P212121
ꢀ
a, b, c (A)
39.23,76.54, 74.94 39.21,76.02, 74.36 50.64,106.72, 179.62
39.36,76.88, 74.48
97.79
39.02,75.35,74.13
97.72
b(°)
Resolution (A)
98.27
97.47
90
ꢀ
38.82–1.80
(1.90–1.80)
174 650 (23 676)
38.88–1.86
(1.96–1.86)
203 849 (27 042)
30.58–1.80 (1.90–1.80) 33.26–1.80 (1.90–1.80) 73.46–2.51
(2.65–2.51)
Number of observed
reflections
668 294 (80 908)
167 317 (24 014)
45 770 (5215)
Number of unique
reflections
40 317 (5839)
36 392 (5211)
88 535 (12379)
40 771 (5902)
14 204 (1734)
Redundancy
4.3 (4.1)
5.6 (5.2)
7.5 (6.5)
4.1 (4.1)
3.2 (3.0)
Completeness (%)
Overall Rmerge (%)
Mean (I/rI)
99.9 (99.5)
0.100 (0.480)
10.1(2.8)
99.8 (98.4)
0.056 (0.156)
22.2(9.8)
97.2 (94.6)
0.116 (0.297)
11.3(4.8)
100.0 (100.0)
0.058 (0.593)
12.6(2.2)
97.2 (81.4)
0.153 0.(441)
5.6(2.3)
R
work/Rfree
0.175/0.237
0.131/0.182
0.167/0.216
0.216/0.258
0.229/293
Model quality
Number of atoms
Number of water
molecules
4382
525
4411
483
8651
931
3827
273
3732
62
Wilson B-factor
ꢀ
Average B-factor (A )
17.7
15.6
13.6
24.9
0.018
1.76
13.1
11.6
9.8
10.3
15.7
13.2
26.3
0.02
2.01
24.8
29.8
30.7
43.5
0.018
1.84
40
2
23.6
24.8
18.1
0.007
1.07
Protein
Water
23.6
0.019
1.8
ꢀ
rmsd bond lengths (A)
rmsd bond angles (°)
Ramachandran map (%)
Most favored (%)
94.4
94.9
4.6
0.4
0
94.9
4.8
0.3
0
95.2
4.8
0
93.4
6.4
0.2
0
Additionally allowed (%) 5.1
Generously allowed (%) 0.4
Outliers (%)
0
0
shown). In the case of TIM wild-type (TWT), there is a
steady increase in the specific activity in the range 25–
45 °C, with a levelling off observed at 50 °C. At higher
temperatures, the enzyme begins to precipitate. This
temperature-dependent increase in activity is generally
observed for enzymes and has been the subject of recent
discussion [38]. In contrast to the wild-type enzyme, the
T75S mutant shows a very small increase in activity
over the range of 25–40 °C and a sharp fall in activity
at higher temperatures, with > 50% loss of activity at
temperatures > 45 °C. The T75N mutant exhibits a
completely different temperature dependence, with
> 50% reduction of activity even at 30 °C. In all the
cases, the assays were carried out by incubating a stock
protein solution at a given temperature for 4 min, fol-
lowed by the addition of an aliquot of protein to a pre-
equilibrated assay mixture to initiate the reaction. Pre-
cipitation of the protein is minimized under these condi-
tions.
tures, followed by centrifugation and removal of
precipitated protein and subsequent addition of an
aliquot from the supernatant to the assay mixture main-
tained at 25 °C. A plot of the activity as a function of
temperature following this protocol is provided in
Fig. 14B. In this case, the mutants T75V/N and T75S
show a dramatic fall in activity at temperatures between
35 °C and 40 °C, whereas the fall in activity for the
T75C mutant is pronounced at temperatures close to
45 °C. The wild-type enzyme exhibits substantial activ-
ity even at 45 °C with precipitation and loss of activity
being observed only at temperatures > 45 °C. These
results demonstrate that while the T75 mutants show lit-
tle or no loss of activity compared to the wild-type
enzyme at 25 °C, there is a pronounced diminution of
activity at temperatures in the range 30–45 °C.
Discussion
In a separate set of experiments, a stock enzyme solu-
tion was incubated for 45 min at the desired tempera-
A relatively small number of residues are completely
conserved in enzyme sequences when large datasets
3872
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
TWT (1O5X)
T75S (4ZZ9)
T75V (5BMW)
T75C (5BNK)
T75N (5BMX)
Fig. 11. Electron density maps (2Fo ꢂ Fc, contoured at 1r) at the active site residues in PfTIM wild-type (1O5X), T75S (4ZZ9), T75V
(5BMW), T75N (5BMX) and T75C (5BNK) structures.
from diverse organisms are examined. In the case of
TIM, as few as ten residues are completely conserved
in sequences that have a length distribution between
220 and 270 residues. Of the ten conserved residues,
four Lys (K) 12, His (H) 95, Glu (E) 97 and Glu (E)
165 are positioned in close proximity to the substrate
and have been implicated in the proton transfer reac-
tion [6,10–12]. The remaining six conserved residues
T75, C126, P166, G209, G210 and G228 do not have
a direct role in the catalytic process, although they are
undoubtedly critical for maintaining the structural
integrity of the active site and in permitting key
dynamic processes that may be important for achiev-
ing key catalytic efficiency. C126 lies proximal to the
active site residue E165. Although this fully conserved
residue is not involved in inter-subunit interactions, it
is mainly important for stability and proper folding of
the dimer [5,39,40]. P166 is a critical residue in the N-
terminal hinge that controls the opening and closing
movement of active site loop 6, which is crucial for
substrate sequestration and subsequent product release
[41–44]. G209 and G210 are part of the important
highly conserved 208–211 segment that undergoes a
peptide flip on going from unliganded to liganded
structure. This helps in the movement of catalytic
loop 6 allowing catalytic base E165 to attain a
conformation necessary for TIM activity. The flip of
209–210 and 210–211 also helps S211 to form a crucial
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3873

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
Fig. 12. Comparison near the site of mutation (position 75) in TWT (1O5X) and the three T75 mutants [T75S (4ZZ9), T75V (5BMW) and
T75N (5BMX)]. The PDB entries are given in parenthesis. Inter-subunit hydrogen-bonding interactions are shown in blue and intra-subunit H-
bonding interactions are shown in red.
interaction with the phosphate moiety of the substrate
in the liganded state [analysis performed with the help
of yeast TIM structures, Protein Data Bank (PDB)
code: 1NEY and 1YPI]. G228 helps in appropriate
positioning of the 208–209 segment of the aforemen-
tioned 208–211 segment by a backbone-backbone
hydrogen bond (G228 -CO and Y208 -NH). The back-
bone torsion angles adopted by this residue are stereo-
chemically feasible only for a Gly residue (φ = 147.1,
w = –161.9, in 1NEY) [45].
In the present study, we have examined the conse-
quences of mutation of the T75 residue. In addition to
complete conserved position in enzymatic sequences,
there are a limited number of sites at which very few
residues appear to be accommodated in natural
sequences. Position 64 in TIM (i.e. the numbering
scheme used corresponds to that for the yeast and the
Plasmodial enzymes) is one such site where, of the
3398 sequences examined, an overwhelming majority
have Gln (Q) (3011/3398), whereas a smaller but sig-
nificant number have Glu (E) (383/3398), a conserva-
tive substitution, at this position. These observations
clearly suggest that a strong selective pressure operates
to limit amino acid variability at position 64. Inspec-
tion of the three-dimensional structures of TIM from
different sources establishes that both residues lie close
to the active site and participate in a network of intra
and inter-subunit hydrogen-bond interactions. Thr75
lies near the TIM active site at the dimer interface
making key contacts with N10 and E97 of the other
subunit. N10, in turn, makes an important hydrogen
bond to the K12 backbone -NH, which is presumably
3874
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
Fig. 13. Comparative view of the van der Waals interactions in PfTIM wild-type (PDB code: 1O5X), T75S (PDB code: 4ZZ9), T75V (PDB
code: 5BMW) and T75N (PDB code: 5BMX) mutants. Similar interactions present in TWT and the mutants are shown in blue; new van der
Waals interactions as a result of the mutations are shown in red; interactions that are lost as a result of the mutations are shown in black.
essential for maintaining the unusual conformation of
this key residue [37]. E97 interacts directly with the
side-chain amino group of K12 and may contribute to
the proton transfer process [6]. We therefore antici-
pated that replacement of T75 may result in both
destabilization of the TIM dimer and/or impaired
activity. The Q64 side chain makes an important
hydrogen bond to the backbone -CO of K12 and an
inter-subunit hydrogen bond to the backbone -NH of
G76, which flanks the Thr75 residue. In this case also,
mutational replacement might be expected to affect
dimer stability and/or enzyme activity. The results
reported in the present study establish that mutations
at position 75 (T75S, T75V, T75C and T75N) do not
appreciably affect dimer stability. Spectroscopic studies
establish that these substitutions have little effect on
the overall structure of the enzyme, a conclusion that
is supported by the observation of intact dimers in the
crystal structures of T75S, T75V and T75N. Indeed,
the active site geometries in the mutant structures are
almost identical to that observed in the wild-type
enzyme. Elution profiles of these mutants in gel filtra-
tion studies are similar to TIM wild-type (TWT), while
measured enzyme activities are also very similar to
those obtained for TWT, with a marginal fall observed
for the T75N mutant. Why then is T75 conserved?
Examination of the temperature dependences of
enzyme activity in the range 25–50 °C reveals that the
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3875

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
73.7 ꢀ 9.2 nM, respectively. These are significantly
higher than the estimated values for the TIM wild-type
enzyme from different organisms. The value of
0.00001 nM was estimated by Lolis et al. [33] on the
basis of the buried surface area in the crystallographi-
cally determined dimer structure of yeast TIM; the
value of 0.01 nM was estimated by Rietveld et al. [34]
by calculating the Gibbs free energy change of dissoci-
ation as a result of GdnHCl-induced denaturation of
rabbit muscle TIM [34]. Using these lower
(0.00001 nM) and upper (0.01 nM) limits of Kd for
dimerization of wild-type enzyme from the above stud-
ies, the Gibbs free energy changes upon mutation
(DDG) of 5.2–9.2 kcalꢃmolꢂ1 are estimated for Q64N,
while the corresponding values for the Q64E mutant
are 4.9–8.9 kcalꢃmolꢂ1. These DDG values are broadly
compatible with the loss of hydrogen-bond interactions
upon mutational replacements together with concomi-
tant changes in local packing (van der Waals) interac-
tions. The Q64E mutant displays a pronounced pH
dependence of activity, with a dramatic fall in the
range pH 7 to pH 8. This is distinct from the beha-
viour of the wild-type enzyme, which shows a broad
range of unimpaired activity between pH 7 and pH 9.
The titration of the buried glutamate residue in the
Q64E mutant must contribute to the structural alter-
ations that influence the orientation of catalytic resi-
dues. This indirect effect on the observed pH
dependence of enzyme activity has long been an
important factor that prevents meaningful interpreta-
tion of the pKa of groups directly involved in catalytic
reactions in enzymes [46]. The influence of this buried
glutamic acid on the stability of the TIM from
L. mexicana has been demonstrated previously. In that
case, the wild-type enzyme contains a Glu (E) residue
at position 65 (equivalent to position 64 in the yeast
and Plasmodial sequences). As noted earlier, the E65Q
mutation in this case results in a dramatic enhance-
ment of the thermal stability of the enzyme [26,47],
while, in the present study, the thermal stability of the
mutant structure has diminished as a result of muta-
tion at the sites 64 and 75, as monitored by CD spec-
troscopy.
Fig. 14. (A) Temperature dependence of specific activity for T75
mutants. (B) Residual activity at different temperatures. In each
case, a protein concentration of 1 nM was used in the presence of
a substrate concentration of 3 mM.
wild-type enzyme shows a steady increase in activity
with temperature up to 45 °C, whereas all four
mutants show a dramatic fall in activity in the range
35–45 °C. All the T75 mutants were significantly less
stable in the presence of high concentration of urea.
While TIM wild-type remains unaffected by concentra-
tions of urea as high as 6 M, all the T75 mutants
unfold at urea concentrations in the range 3.5–4 M.
These results point to the possibility that temperature
stability in this range may be the phenotype being
selected, rendering position 75 resistant to evolutionary
drift.
The Q64E and Q64N mutants display a significant
destabilization of the TIM dimer, as evidenced by the
gel filtration elution profile and concentration depen-
dence of enzyme activity in the range 5–150 nM. In
addition, both mutants show a loss of near UV-CD
bands and enhanced binding of the fluorescent probe
ANS, establishing significant perturbation of the ter-
tiary structure in the vicinity of the site of mutation.
Both mutants also unfold at relatively low concentra-
tions of urea (3.5–4 M) in contrast to the wild-type
enzyme, which remains stable even at significantly
higher urea concentrations. The concentration depen-
dence of enzyme activity has been used to estimate the
dissociation constant for the Q64E and Q64N mutant
dimers, yielding values of 44.6 ꢀ 8.4 nM and
The small number of fully conserved and highly
conserved residues in a large dataset of TIM sequences
suggests that the mutational substitutions are well tol-
erated at most positions in the polypeptide sequence.
The inference that there are multiple sequence solu-
tions to the problems of folding and construction of
protein–protein interfaces appears to be reasonable. In
contrast, the constraints of catalytic chemistry impose
a unique sequence solution or a restricted set of solu-
tions for residues directly involved in the catalytic pro-
3876
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
(0.01 mM), DTT (2 mM) and 10% glycerol and disrupted
by sonication. After centrifugation (19 320 g for 30 min at
4 °C) and removal of the cell debris, the supernatant was
fractionated with ammonium sulfate. The protein fractions,
containing T75 mutants, were in the supernatant solution
of 60% ammonium sulfate saturation. The supernatant was
thoroughly dialyzed against 20 mM Tris-HCl (pH 8.0) buf-
fer solution. The dialyzed solution was centrifuged
(19 320 g for 45 min at 4 °C) and the soluble fraction was
further purified by anion exchange (Q-Sepharose; GE
Healthcare UK Ltd, Little Chalfont, UK) chromatography,
with a linear gradient of NaCl (0–1 M) salt [5]. On the
other hand, the almost pure Q64 mutants remained in the
supernatant of 80% ammonium sulfate saturation (first
step: 40% and second step: 80% ammonium sulfate precip-
itation to the supernatant of 40% precipitation). The puri-
fied protein obtained was then extensively dialyzed against
20 mM Tris-HCl (pH 8.0) at 4 °C for desalting. Protein
purity was checked by 12% SDS/PAGE. The masses of the
mutant proteins were confirmed by LC-ESI-MS with a
Bruker Q-TOF mass spectrometer (Bruker Instruments,
Inc., Bellerica, MA, USA), Massobs (Masscalc): 27831.3 Da
(27831.5 Da) TWT; 27818.2 Da (27817.8 Da) T75S;
27828.7 Da (27829.6 Da) T75V; 27832.6 Da (27833.6 Da)
T75C; 27844.4 Da (27844.6 Da) T75N; 27817.8 Da
(27817.5 Da) Q64N and 27833.3 Da (27832.5 Da) Q64E
(data not shown). Protein concentrations were determined
by Bradford’s method using BSA as the standard [51] and
are expressed (lM or nM) using the monomer molecular
mass.
cess and others that form the immediate scaffolding.
In the case of TIM, all the fully conserved residues are
proximal to the active site and highly conserved resi-
dues, such as Q64, interact directly with the active site
residues. Selective pressures undoubtedly shape TIM
sequences, optimizing specific features such as temper-
ature dependence of activity and the dissociation con-
stants of the dimer, parameters that assume
considerable importance as the integrity of the active
site is inextricably linked to the integrity of the struc-
ture of the protein dimer.
Experimental procedures
Cloning expression and purification
The TIM gene from the organism Pf has been cloned into
the expression vector pTrc99A, known as pARC1008 [48].
The protein was overexpressed into a host TIM null
mutant Escherichia coli strain AA200 [49]. To probe the
role of residue 64 and 75, mutagenesis was carried out in
the Pf TIM gene by site-directed mutagenesis using single
primer method as described previously [50]. The absence of
an appropriate restriction site necessitated a two-step pro-
cedure to generate mutants at the two sites. In the first
step, two intermediate clones at each position were gener-
ated (Q64int and T75int mutants). This was carried out
using the wild-type gene as template and using the single
primers with EcoRV restriction sites (primers shown in
Table S1). The generated clones were confirmed by restric-
tion digestion. In the second step, mutant clones Q64N,
Q64E and T75S, T75V, T75C and T75N were constructed,
with the subsequent removal of the restriction site of
EcoRV and the introduction of the desired mutations,
using the Q64int and T75int clone gene as the template for
PCR.
Size exclusion chromatography
Size exclusion chromatography was performed using a
Superdex-200 column (length 300 mm, internal diameter
10 mm) (GE Healthcare) for Q64 mutants and a Superdex-
75 column (length 300 mm, internal diameter 10 mm) for
T75 mutants, attached to an AKTA Basic FPLC system at
a flow rate of 0.5 mLꢃminꢂ1. The solvent system was
20 mM Tris-HCl (pH 8.0). Protein elution was monitored
at a wavelength of 280 nm. The column was calibrated
with BSA (66 kDa) and carbonic anhydrase (29 kDa). All
the chromatographic runs were performed at 25 °C.
For both PCRs, the thermostable proofreading poly-
merase enzyme Phusion (Thermo Scientific, Waltham, MA,
USA) was used. The PCR cycling conditions involved an
initial step at 95 °C for 5 min followed by 40 cycles consist-
ing of 1 min at 95 °C, annealing at 50 °C for 1 min and
extension at 72 °C for 10 min. Final incubation at 72 °C
for 20 min was applied. The presence of mutations was
confirmed by restriction digestion and DNA sequencing.
The primers used to generate the two Q64 and four T75
mutants are shown in Table S1. The mutant TIM genes,
cloned into the expression vector pTrc99A, were expressed
in E. coli AA200 (devoid of inherent bacterial TIM gene)
strain. The cells were grown in Terrific broth (containing
100 lgꢃmLꢂ1 antibiotic ampicillin) at 37 °C. The cells were
induced with 400 lM IPTG at OD600 of 0.6–0.8, followed
by harvesting using centrifugation (7245 g for 20 min at
4 °C). Later, the cells were resuspended in lysis buffer con-
taining 20 mM Tris-HCl (pH 8.0), EDTA (1 mM), PMSF
CD
Far UV and near UV-CD measurements were carried out
on a Jasco-815 spectropolarimeter (Jasco Inc., Easton,
MD, USA) connected to a thermostatted cell holder, which
was temperature controlled inside the cuvette with a Peltier
device. Denaturation studies were performed by incubating
5 lM protein with various concentrations of urea for
45 min and spectra were recorded in the range 250–200 nm
and averaged over three scans. For thermal melting studies,
far UV-CD spectra were recorded over the range 20–75 °C
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3877

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
at intervals of 5 °C. At each temperature, the protein was
incubated for 10 min to follow the unfolding transition. A
cuvette of path length of 1 mm and a protein concentration
of 15 lM were used. The spectra were averaged over three
scans at a scanning speed of 10 nmꢃminꢂ1. The change of
ellipticity was measured as a function of temperature for
thermal melting. Individual spectra (far UV-CD: 250–
200 nm; near UV-CD: 300–250 nm) were averaged over
four scans with the same scan speed. All experiments were
carried out in 20 mM Tris-HCl (pH 8.0).
decrease in absorbance of NADH at 340 nm. Theꢂe1xtinc-
tion coefficient of NADH was taken to be 6220 M ꢃcmꢂ1
at 340 nm [53]. In the range studied, the linear dependence
of the initial rate on the enzyme concentration ensures the
validity of the assay [52]. The values for the kinetic param-
eters of T75 mutants (Km, kcat) were determined by fitting
the initial velocity data to the Michaelis–Menten equation
using PRISM, version 5. For this experiment, 14 ng of
PfTIM wild-type and T75S/V/C mutants and 28 ng of
T75N mutant were used. The kinetic parameters for Q64
could not be determined because of the absence of satura-
tion at higher concentrations of substrate. As a result, the
specific activities at a particular substrate concentration
were compared. Activities of all the mutants and of wild-
type were measured as a function of protein concentrations
at ambient temperature (23 °C). Concentrations varied in
the range 0.15–5 nM in presence of 3 mM substrate for the
T75 mutants and wild-type TIM, whereas, for the less
active Q64 mutants, the concentrations varied in the range
5–200 nM. The effect of temperature on the activity of T75
mutants was also measured by monitoring the reaction at
different temperatures, as well as by incubating the protein
at 28 lM at different temperatures and then diluting it to
1 nM at room temperature for measuring residual activity.
A concentration of 100 mM of mixed Tris HCl-MES-
acetate-glycine buffer of different pH values (ionic
strength = 0.1) was used for pH dependence of the specific
activity of Q64 mutants.
Fluorescence spectroscopy
Emission spectra were recorded on a spectrofluorimeter
(Hitachi-F-2500; Hitachi, Tokyo, Japan). For denaturation
studies, the protein samples were excited at 295 nm and the
emission spectra were recorded inthe range 300–400 nm
(kexcitation = 295 nm) after incubating 2 lM of each protein
with different concentrations of urea for 45 min. Excitation
and emission band passes were kept at 5 nm and 10 nm,
respectively for the Q64 mutants, whereas they were both
set at 5 nm for the T75 mutants. For the ANS binding
study, 5 lM of TIM wild-type and both the Q64 mutants
were incubated with different concentrations of ANS and
excited either at 280 nm or at 370 nm and emission spectra
were recorded in the range 300–600 nm and 400–600 nm,
respectively.
Enzyme activity
Crystallization and X-ray diffraction data
collection of PfTIM T75 and Q64 mutants
The kinetics of the conversion of GAP to dihydroxyacetone
phosphate (DHAP) by the enzyme TIM (Pf wild-type and
mutants) was monitored in the presence of coupling
enzyme, a-glycerol phosphate dehydrogenase, as described
by Plaut and Knowles [52]. The final 500-lL reaction mix-
ture contained 100 mM TEA (pH 7.6), 0.1 mM NADH,
5 mM EDTA, 10 lg of the a-glycerol phosphate dehydroge-
nase and varying concentrations (0.25–4.5 mM) of the sub-
strate (GAP) to which TIM was added to initiate the
reaction. In case of the Q64 mutants the aforementioned
assay technique gave a two-phase reaction profile (data not
shown). The two-phase assay profiles for the Q64 mutants
occurred probably as a result of perturbation in dimer–
monomer equilibrium as a result of sudden dilution of the
stock solution. Knowles’ protocol has been slightly modi-
fied in the present study, which is the first to achieve the
dimer–monomer equilibrium and then monitor the enzyme
kinetics. For this, the Pf wild-type and Q64 TIM mutants
were incubated for 45 min in the assay mixture and then
the reaction was monitored by adding the substrate GAP.
PfTIM wild-type, TWT (14 ng), Q64N (120 ng) and Q64E
(300 ng) were used for substrate titration in accordance
with the Knowles protocol, whereas, for the concentration-
and pH-dependent experiments, the alternate protocol was
used. The progress of the reaction was monitored by the
The Q64 and T75 mutants were crystallized by the hanging
drop method at 23 °C [54]. The concentrations of the two
mutant proteins were 1.5 and 10 mgꢃmLꢂ1, respectively. The
unliganded T75S crystal was obtained under the conditions:
22% poly(ethylene glycol)-1450, 100 mM HEPES buffer (pH
7.0), 10 mM calcium chloride, 0.5 mM EDTA; the unli-
ganded T75V crystal was obtained under the conditions:
20% poly(ethylene glycol)-1450, 100 mM MES buffer (pH
6.5), 10 mM calcium chloride, 0.5 mM EDTA; the unli-
ganded T75N crystal was obtained under the conditions:
28% poly(ethylene glycol)-1450, 100 mM Tris-HCl buffer
(pH 8.0), 10 mM lithium sulfate, 0.5 mM EDTA; and the
unliganded T75C crystal was obtained under the conditions:
16% poly(ethylene glycol)-1450, 100 mM HEPES buffer (pH
7.5), 10 mM CaCl2, 0.5 mM EDTA. The hanging drop con-
tained 3 lL of protein and 3 lL of crystallization cocktail
equilibrated against 400 lL of crystallization cocktail as the
reservoir buffer. The crystals appeared within 4 days and
grew to required sizes within 10–15 days. The Q64N and
Q64E mutants crystallized under a number of conditions.
However, these crystals did not diffract X-rays to useful res-
olution. The crystal obtained with Q64E mutant diffracted
ꢀ
X-rays to a resolution of 2.51 A only under one of the con-
3878
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
ditions (28% poly(ethylene glycol)-1450, 100 mM HEPES,
pH 7.5, 10 mM CaCl2).
not shown) for the mutants, confirmed the presence of the
mutations. The data collection and refinement statistics for
the mutant structures are shown in Table 2. Density for the
ligand was not observed in any of the mutant crystal struc-
tures, obtained by cocrystallization with the ligand. Images
were generated using PYMOL (http://www.pymol.org) [63].
The experimental structure factors and the atomic coordi-
nates have been deposited in the PDB.
Diffraction quality crystals obtained under the above
mentioned conditions were soaked in 10% (v/v) ethylene
glycol solution mixed with the reservoir buffer before flash
freezing in liquid nitrogen. X-ray diffraction data for
Q64E, T75S and T75V mutants were collected using a RU
200 rotating anode X-ray generator (Rigaku, Tokyo,
ꢀ
Japan) (k = 1.54 A). Datasets for T75N and T75C mutants
ꢀ
were collected using synchrotron radiation (k = 0.954 A) at
Acknowledgements
the European Synchrotron Radiation Facility (ESRF),
Grenoble and on a CCD MARmosaic 225 detector. The
unliganded T75S, T75C and T75N crystals diffracted to
ꢀ
1.80 A, whereas, the unliganded T75V crystal diffracted to
ꢀ
1.86 A. All datasets were processed using MOSFLM [55,56]
We thank N. V. Joshi for analysis of the TIM
sequences and helpful discussions. We also thank the
support staff at MBU X-ray facility and beamline sci-
entists at ESRF, Grenoble, for help during data collec-
tion. DB was a senior research fellow of the Council
for Scientific and Industrial Research, Government of
India. Research on TIM was supported under an insti-
tutional grant from the Department of Biotechnology
(India). X-ray diffraction and MS facilities were sup-
ported by program grants from the Department of
Science and Technology and Department of Biotech-
nology (India).
and scaled with SCALA [57] of the CCP4 (Collaborative Com-
putational Project, Number 4, 1994) software suite [58].
Structure determination and refinement
Structures of all the five mutants were determined with the
molecular replacement software PHASER of the CCP4 package
[59]. The native PfTIM crystal structure (PDB code: lO5X;
ꢀ
1.1 A) was used as the template for structure determination
of all the mutants and their complexes, Q64E-unlig, T75S-
unlig, T75V-unlig, T75C-unlig and T75N-unlig. The coordi-
nates of 1O5X were modified by removing loop 6 residues,
residues near the site of mutations and the regions that
interact with it, ligand, water molecules and alternate con-
formations, before initiating molecular replacement calcula-
tions. Refinements of all the structures were carried out
using REFMAC5 [60,61], with an initial 20 cycles of rigid
body refinement, followed by 100 cycles of restrained
refinement. On the basis of the electron density maps,
2Fo ꢂ Fc and Fo ꢂ Fc, contoured at 1r and 3r, respec-
tively, the loop 6 residues, the residues that were absent in
the model, ligand and water molecules, were built. Model
building was carried out using COOT [35,36]. Two subunits
were present in one asymmetric unit, in the case of the
Q64E-unlig, T75S-unlig, T75V-unlig and T75C-unlig struc-
tures, whereas, for T75N-unlig, which belonged to a differ-
ent space group, four subunits were observed. The presence
of the Q64E, T75S, T75V and T75N mutations were con-
firmed from the difference Fourier maps. The T75C mutant
structure was not well-defined near the site of mutation as
a result of static disorder. Water molecules were automati-
cally located by COOT and accepted if the peak observed
was above 1.5r in a 2Fo ꢂ Fc map and 3r in an Fo ꢂ Fc
map. Alternate conformations of residues were included
wherever necessary and the B-factors of all the atoms were
refined. All the structures were refined to reasonable Rwork
and Rfree values and good geometry. The quality of the
Author contributions
DB carried out all of the experimental work. DB,
MRNM, HB and PB were involved in data analysis.
DB, MRNM, HB and PB have approved the final ver-
sion of the manuscript submitted for publication.
References
1 Harms MJ & Thornton JW (2013) Evolutionary
biochemistry: revealing the historical and physical
causes of protein properties. Nat Rev Genet 14, 559–
571.
2 Guo HH, Choe J & Loeb LA (2004) Protein tolerance
to random amino acid change. Proc Natl Acad Sci USA
101, 9205–9210.
3 Sullivan BJ, Durani V & Magliery TJ (2011)
Triosephosphate isomerase by consensus design:
dramatic differences in physical properties and activity
of related variants. J Mol Biol 413, 195–208.
4 Sullivan BJ, Nguyen T, Durani V, Mathur D, Rojas S,
Thomas M, Syu T & Magliery TJ (2012) Stabilizing
proteins from sequence statistics: the interplay of
conservation and correlation in triosephosphate
isomerase stability. J Mol Biol 420, 384–399.
5 Samanta M, Banerjee M, Murthy MRN, Balaram H &
Balaram P (2011) Probing the role of the fully
conserved Cys126 in triosephosphate isomerase by site-
specific mutagenesis – distal effects on dimer stability.
FEBS J 278, 1932–1943.
structures was evaluated by
CCP4 suite
[62]. The electron density maps (2Fo
PROCHECKꢂoFfc,thceontoured at
1.0r) surrounding the residues at position 64 and 75 (data
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3879

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
6 Samanta M, Murthy MRN, Balaram H & Balaram P
(2011) Revisiting the mechanism of the triosephosphate
isomerase reaction: the role of the fully conserved
glutamic acid 97 residue. ChemBioChem 12, 1886–1896.
7 Wierenga R (2001) The TIM-barrel fold: a versatile
framework for efficient enzymes. FEBS Lett 492, 193–
198.
Cold Spring Harb Symp Quant Biol pp. 603–613. Cold
Spring Harbor Laboratory Press.
22 Knowles JR (1991) Enzyme catalysis: not different, just
better. Nature 350, 603–613.
23 Knowles JR (1991) To build an enzyme. Philos Trans R
Soc Lond B Biol Sci 332, 115–121.
24 Lodi PJ, Chang LC, Knowles JR & Komives EA
(1994) Triosephosphate isomerase requires a positively
charged active site: the role of lysine-12. Biochemistry
33, 2809–2814.
25 Rieder SV & Rose IA (1959) The mechanism of the
triosephosphate isomerase reaction. J Biol Chem 234,
1007–1010.
26 Williams JC, Zeelen JP, Neubauer G, Vriend G,
Backmann J, Michels PA, Lambeir A-M & Wierenga
RK (1999) Structural and mutagenesis studies of
leishmania triosephosphate isomerase: a point mutation
can convert a mesophilic enzyme into a superstable
enzyme without losing catalytic power. Protein Eng 12,
243–250.
27 Borchert TV, Abagyan R, Jaenicke R & Wierenga RK
(1994) Design, creation, and characterization of a
stable, monomeric triosephosphate isomerase. Proc Natl
Acad Sci USA 91, 1515–1518.
28 Ray SS, Singh SK & Balaram P (2001) An electrospray
ionization mass spectrometry investigation of 1-anilino-
8-naphthalene-sulfonate (ANS) binding to proteins.
J Am Soc Mass Spectrom 12, 428–438.
29 Bhattacharjya S & Balaram P (1997) Hexafluoroacetone
hydrate as a structure modifier in proteins:
8 Cui Q & Karplus M (2003) Catalysis and specificity in
enzymes: a study of triosephosphate isomerase and
comparison with methyl glyoxal synthase. Adv Protein
Chem 66, 315–372.
9 Malabanan MM, Amyes TL & Richard JP (2010) A
role for flexible loops in enzyme catalysis. Curr Opin
Struct Biol 20, 702–710.
10 Richard JP (2012) A paradigm for enzyme-catalyzed
proton transfer at carbon: triosephosphate isomerase.
Biochemistry 51, 2652–2661.
11 Richard JP, Amyes TL, Goryanova B & Zhai X (2014)
Enzyme architecture: on the importance of being in a
protein cage. Curr Opin Chem Biol 21, 1–10.
12 Richard JP, Zhai X & Malabanan MM (2014)
Reflections on the catalytic power of a TIM-barrel.
Bioorg Chem 57, 206–212.
13 Albery WJ & Knowles JR (1976) Evolution of enzyme
function and the development of catalytic efficiency.
Biochemistry 15, 5631–5640.
ꢁ
ꢁ
14 Orosz F, Olah J & Ovadi J (2009) Triosephosphate
isomerase deficiency: new insights into an enigmatic
disease. Biochim Biophys Acta 1792, 1168–1174.
15 Lincet H & Icard P (2015) How do glycolytic
enzymes favour cancer cell proliferation by
nonmetabolic functions & quest. Oncogene 34,
3751–3759.
characterization of a molten globule state of hen egg-
white lysozyme. Protein Sci 6, 1065–1073.
30 Hameed M, Ahmad B, Fazili KM, Andrabi K &
Khan RH (2007) Different molten globule-like folding
intermediates of hen egg white lysozyme induced by
high pH and tertiary butanol. J Biochem 141, 573–
583.
16 Consortium U (2011) Ongoing and future developments
at the Universal Protein Resource. Nucleic Acids Res
39, D214–D219.
17 Coulson A & Knowles J (1970) Uniquely labelled active
site sequence in chicken muscle triose phosphate
isomerase. Nature 227, 180–181.
18 Webb MR & Knowles JR (1974) The existence of an
electrophilic component in the reaction catalysed by
triose phosphate isomerase. Biochem J 141, 589.
19 Belasco JG & Knowles JR (1980) Direct observation of
substrate distortion by triosephosphate isomerase using
Fourier transform infrared spectroscopy. Biochemistry
19, 472–477.
31 Gokhale RS, Ray SS, Balaram H & Balaram P (1999)
Unfolding of Plasmodium falciparum triosephosphate
isomerase in urea and guanidinium chloride: evidence
for a novel disulfide exchange reaction in a covalently
cross-linked mutant. Biochemistry 38, 423–431.
32 Pray TR, Reiling KK, Demirjian BG & Craik CS
(2002) Conformational change coupling the
dimerization and activation of KSHV protease.
Biochemistry 41, 1474–1482.
20 Straus D, Raines R, Kawashima E, Knowles JR &
Gilbert W (1985) Active site of triosephosphate
isomerase: in vitro mutagenesis and characterization of
an altered enzyme. Proc Natl Acad Sci USA 82, 2272–
2276.
33 Lolis E, Alber T, Davenport RC, Rose D, Hartman FC
& Petsko GA (1990) Structure of yeast triosephosphate
ꢀ
isomerase at 1.9 A resolution. Biochemistry 29, 6609–
6618.
34 Rietveld AW & Ferreira ST (1998) Kinetics and
energetics of subunit dissociation/unfolding of TIM: the
importance of oligomerization for conformational
persistence and chemical stability of proteins.
Biochemistry 37, 933–937.
21 Alber T, Davenport R, Giammona D, Lolis E, Petsko
G & Ringe D (1987) Crystallography and site-directed
mutagenesis of yeast triosephosphate isomerase: what
can we learn about catalysis from a “simple” enzyme?
3880
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS

D. Bandyopadhyay et al.
Conserved residues in triosephosphate isomerase
35 Emsley P & Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallographica
Section D 60, 2126–2132.
36 Emsley P, Lohkamp B, Scott WG & Cowtan K (2010)
Features and development of Coot. Acta
48 Ranie J, Kumar VP & Balaram H (1993) Cloning of
the triosephosphate isomerase gene of Plasmodium
falciparum and expression in Escherichia coli. Mol
Biochem Parasitol 61, 159–169.
49 Anderson A & Cooper R (1970) Genetic mapping of a
locus for triosephosphate isomerase on the genome of
Escherichia coli K 12. J Gen Microbiol 62, 329–334.
50 Shenoy AR & Visweswariah SS (2003) Site-directed
mutagenesis using a single mutagenic oligonucleotide
and DpnI digestion of template DNA. Anal Biochem
319, 335–336.
51 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein
utilizing the principle of protein-dye binding. Anal
Biochem 72, 248–254.
52 Plaut B & Knowles J (1972) pH-dependence of the
triose phosphate isomerase reaction. Biochem J 129,
311–320.
Crystallographica Section D 66, 486–501.
37 Banerjee M, Balaram H & Balaram P (2009) Structural
effects of a dimer interface mutation on catalytic
activity of triosephosphate isomerase. FEBS J 276,
4169–4183.
38 Elias M, Wieczorek G, Rosenne S & Tawfik DS (2014)
The universality of enzymatic rate–temperature
dependency. Trends Biochem Sci 39, 1–7.
ꢁ
ꢁ
39 Gonzalez-Mondragon E, Zubillaga RA, Saavedra E,
ꢁ
ꢁ
ꢁ
ꢁ
Chanez-Cardenas ME, Perez-Montfort R & Hernandez-
Arana A (2004) Conserved cysteine 126 in
triosephosphate isomerase is required not for enzymatic
activity but for proper folding and stability.
Biochemistry 43, 3255–3263.
53 Horecker BL & Kornberg A (1948) The extinction
coefficients of the reduced band of pyridine nucleotides.
J Biol Chem 175, 385–390.
ꢁ
ꢁ
40 Hernandez-Santoyo A, Domınguez-Ramırez L, Reyes-
ꢁ
ꢁ
Lopez CA, Gonzalez-Mondragon E, Hernandez-Arana
ꢁ
ꢁ
ꢁ
ꢁ
A & Rodrıguez-Romero A (2012) Effects of a buried
54 Velanker SS, Ray SS, Gokhale RS, Suma S, Balaram
H, Balaram P & Murthy MRN (1997) Triosephosphate
isomerase from Plasmodium falciparum: the crystal
structure provides insights into antimalarial drug
design. Structure 5, 751–761.
55 Leslie AG & Powell HR (2007) Processing diffraction
data with Mosflm. In Evolving methods for
macromolecular crystallography, pp. 41–51. Springer.
56 Battye TGG, Kontogiannis L, Johnson O, Powell HR
& Leslie AG (2011) iMOSFLM: a new graphical
interface for diffraction-image processing with
MOSFLM. Acta Crystallographica Section D 67, 271–
281.
cysteine-to-serine mutation on yeast triosephosphate
isomerase structure and stability. Int J Mol Sci 13,
10010–10021.
41 Brown FK & Kollman PA (1987) Molecular dynamics
simulations of “loop closing” in the enzyme triose
phosphate isomerase. J Mol Biol 198, 533–546.
42 Joseph D, Petsko GA & Karplus M (1990) Anatomy of
a conformational change: hinged” lid” motion of the
triosephosphate isomerase loop. Science 249, 1425–
1428.
43 Xiang J, Sun J & Sampson NS (2001) The importance
of hinge sequence for loop function and catalytic
activity in the reaction catalyzed by triosephosphate
isomerase. J Mol Biol 307, 1103–1112.
44 Casteleijn MG, Alahuhta M, Groebel K, El-Sayed I,
Augustyns K, Lambeir AM, Neubauer P & Wierenga
RK (2006) Functional role of the conserved active site
proline of triosephosphate isomerase. Biochemistry 45,
15483–15494.
57 Evans P (2006) Scaling and assessment of data quality.
Acta Crystallographica Section D 62, 72–82.
58 Winn MD, Ballard CC, Cowtan KD, Dodson EJ,
Emsley P, Evans PR, Keegan RM, Krissinel EB, Leslie
AG & McCoy A (2011) Overview of the CCP4 suite
and current developments. Acta Crystallographica
Section D 67, 235–242.
45 Samanta M (2011) Exploring the role of highly
conserved residues in triosephosphate isomerase. PhD
Thesis, Indian Institute of Science, Bangalore.
46 Knowles JR & Jencks WP (1976) The intrinsic PkA-
values of functional groups in enzymes: improper
deductions from the Ph-dependence of steady-state
parameter. Crit Rev Biochem Mol Biol 4, 165–173.
47 Lambeir AM, Backmann J, Ruiz-Sanz J, Filimonov V,
Nielsen JE, Kursula I, Norledge BV & Wierenga RK
(2000) The ionization of a buried glutamic acid is
thermodynamically linked to the stability of Leishmania
mexicana triose phosphate isomerase. Eur J Biochem
267, 2516–2524.
59 McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn
MD, Storoni LC & Read RJ (2007) Phaser
crystallographic software. J Appl Crystallogr 40, 658–
674.
60 Murshudov GN, Vagin AA & Dodson EJ (1997)
Refinement of macromolecular structures by the
maximum-likelihood method. Acta Crystallographica
Section D 53, 240–255.
ꢁ
61 Murshudov GN, Skubak P, Lebedev AA, Pannu NS,
Steiner RA, Nicholls RA, Winn MD, Long F & Vagin
AA (2011) REFMAC5 for the refinement of
macromolecular crystal structures. Acta
Crystallographica Section D 67, 355–367.
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS
3881

Conserved residues in triosephosphate isomerase
D. Bandyopadhyay et al.
62 Laskowski RA, MacArthur MW, Moss DS &
Thornton JM (1993) PROCHECK: a program to check
the stereochemical quality of protein structures. J Appl
Crystallogr 26, 283–291.
Podcast
This article is accompanied by a podcast, listen now.
Or listen in iTunes.
63 DeLano WL (2002) The PyMol Molecular Graphics
System. DeLano Scientific, San Carlos, CA.
Supporting information
Additional supporting information may be found in the
online version of this article at the publisher’s web site:
Table S1. List of mutagenic primers used to generate
the mutants.
3882
FEBS Journal 282 (2015) 3863–3882 ª 2015 FEBS