BBA - General Subjects 1862 (2018) 1401–1409

Contents lists available at ScienceDirect

BBA - General Subjects

journal homepage: www.elsevier.com/locate/bbagen

Differential effects on enzyme stability and kinetic parameters of mutants T

related to human triosephosphate isomerase deficiency
Nallely Cabreraa, Alfredo Torres-Lariosa, Itzhel García-Torresb, Sergio Enríquez-Floresb,
⁎
Ruy Perez-Montforta,
a
Departamento de Bioquímica y Biología Estructural, Instituto de Fisiología Celular, Universidad Nacional Autónoma de México, Av. Universidad 3000, Coyoacán, 04510,
Ciudad de México, Mexico
b
Laboratorio de Bioquímica-Genética, Instituto Nacional de Pediatría, Insurgentes Sur 3700-C, Col. Insurgentes Cuicuilco, Coyoacán, 04530, Ciudad de México, Mexico

A R T I C LE I N FO A B S T R A C T

Keywords: Human triosephosphate isomerase (TIM) deficiency is a very rare disease, but there are several mutations re-
Triosephosphate isomerase deficiency ported to be causing the illness. In this work, we produced nine recombinant human triosephosphate isomerases
Site directed mutagenesis which have the mutations reported to produce TIM deficiency. These enzymes were characterized biophysically
Recombinant enzyme and biochemically to determine their kinetic and stability parameters, and also to substitute TIM activity in
Activity
supporting the growth of an Escherichia coli strain lacking the tim gene. Our results allowed us to rate the
Stability
Crystal structure
deleteriousness of the human TIM mutants based on the type and severity of the alterations observed, to classify
four “unknown severity mutants” with altered residues in positions 62, 72, 122 and 154 and to explain in
structural terms the mutation V231M, the most affected mutant from the kinetic point of view and the only
homozygous mutation reported besides E104D.

1. Introduction
The fifth enzyme of the glycolytic pathway, triosephosphate iso-
merase (TIM, E.C. 5.3.1.1), catalyzes the reversible conversion between
dihydroxyactetone 3 phosphate (DHAP) and glyceraldehyde 3 phos-
phate (GAP). After this enzymatic reaction, GAP can be further meta-
bolized by subsequent enzymes in the pathway, continuing the break-
down of glucose, which produces two ATPs per molecule of the
carbohydrate [1,2].
Almost all organisms have this enzyme, which, in most cases, is a
homodimer with two identical subunits with a molecular mass of
27 kDa composed of approximately 250 amino acids. Both monomers
have all catalytic residues but TIM is only active as a dimer. In some
archaea TIM has an active and stable homotetrameric quaternary
structure, but, in all cases, monomers of this enzyme are inactive and
unstable, emphasizing that the enzyme acts as either a dimer or a tet-
ramer [3].
Human triosephosphate isomerase (HsTIM) is formed by two sub-
units, which have 248 amino acids, that associate to form a 54 kDa
homodimer. The enzyme with the wild type (WT) sequence of HsTIM
has kinetic parameters that make it a nearly perfect enzyme, because
the reaction is a diffusion-limited process. It also has a high thermal
resistance and its monomers interact with high affinity making it a very
stable dimer [4,5].
The sequence of TIM has high homology for many species. A few
key residues are completely conserved among all known sequences, and
in some other positions have relatively little variation. In the sequence
of HsTIM, very few mutations occur naturally that allow an individual
to survive. The extremely small number of persons who have their
HsTIM with an altered sequence suffer from HsTIM deficiency. This
condition represents the most severe glycolytic enzyme defect in hu-
mans which is almost always lethal in early childhood. Although this
disease, and the mutations in HsTIM that produce it, have been ex-
tensively reviewed [6–8], biochemical characterization using purified
recombinant mutant enzymes has not been investigated. There have
been several attempts to predict in silico the effect of mutations in
HsTIM as they might reflect in the patients with TIM deficiency.
Schneider [6] and Oliver and Timson [9] proposed that mutant proteins
associated with pathological phenotypes are less stable than those as-
sociated with a less severe disease. Schneider [6], based on structural
information, also predicted that catalytic abnormalities might be asso-
ciated with the severity of the disease. Oliver and Timson [9] could not
include kinetic parameters of enzyme function into their bioinformatics
investigations. So, both these analyses lack the direct study of the

⁎
Corresponding author at: Departamento de Bioquímica y Biología Estructural, Instituto de Fisiología Celular, Universidad Nacional Autónoma de México, Circuito Exterior S/N,
Delegación Coyoacán, 04510, Ciudad de México, Mexico.
E-mail address: ruy@ifc.unam.mx (R. Perez-Montfort).

https://doi.org/10.1016/j.bbagen.2018.03.019
Received 8 February 2018; Received in revised form 14 March 2018; Accepted 19 March 2018
Available online 20 March 2018
0304-4165/ © 2018 Elsevier B.V. All rights reserved.
N. Cabrera et al. BBA - General Subjects 1862 (2018) 1401–1409

function of the mutant HsTIMs having the mutated amino acids de- Table 1
scribed in patients with HsTIM deficiency and the characterization of Enzymatic kinetic parameters for WT HsTIM and the mutant enzymes. The mean of three
independent experiments is shown. For simplicity, a common color code is followed in all
these enzymes in vitro. In this study, we offer such a comparison of
tables and figures to describe the behavior of the enzymes.
several mutants related to HsTIM deficiency. We rate these mutants in
terms of their deleterious effects according to kinetic, thermal and di- Enzyme Km (mM) kcat kcat/ Km kcat/ Km
(105 min-1) (106 M- 1 s-1) relative
merization stability assays, and also bacterial growth inhibition. In
addition, using X-ray crystallography, we analyze the molecular basis of WT 0.46 + 0.07 2.73 9.8 1
the deficiency produced by mutant HsTIMV231M, the only other C41Y 0.62 + 0.03 2.16 5.8 0.59
A62D nd nd nd nd
known homozygous mutant reported besides HsTIME104D.
G72A 3.0 + 0.36 4.11 2.3 0.2
E104D 0.57 + 0.08 3.31 9.6 0.97
2. Materials and methods G122R 0.57 + 0.08 3.57 10.4 1.06
V154M 0.49 + 0.04 2.7 9.0 0.91
2.1. Cloning and purification I170V 0.03 + 0.004 0.11 5.8 0.59
V231M 4.02 + 0.99 2.67 1.1 0.1
F240L 0.83 + 0.06 3.32 6.6 0.6
The DNA sequence X69723.1 (NCBI database) for HsTIM was used
to make the single mutants. All mutants were constructed on a modified
plasmid pET-3a [5] using the QuikChange protocol (Agilent Technol- corresponding mutant enzymes (Table 1). To calculate the kinetic
ogies, CA). The plasmids containing the different mutants were trans- parameters, GAP concentration was varied between 0.05 and 3 mM.
formed into the Escherichia coli BL21 Codon Plus (DE3)-RIL strain The data were adjusted to the Michaelis-Menten model and the values
(Agilent Technologies, CA). Cells were grown at 37 °C in Luria-Bertani of Km and Vmax were calculated using non-linear regression. Activity
medium containing ampicillin and chloramphenicol until an A600 of 0.8 was measured in a Cary 60 spectrophotometer (Agilent Technology,
was obtained. At that time, they were induced with 1 mM isopropyl-D-1- CA) with a multi-cell attachment. All assays were performed in tripli-
thiogalactopyranoside following the conditions described in Supple- cate.
mentary Table 1, which were obtained after testing two periods of time
for induction (3 and 18 h), and three different temperatures (18, 30 and
37 °C). 2.4. Thermal shift assay (differential scanning fluorimetry)
The cell pellet from 1 L of culture was suspended in 20 mL of buffer
A containing 50 mM sodium phosphate pH 8.0, 10 mM imidazole and We followed the protocol described by Niesen and collaborators
different concentrations of NaCl (Supplementary Table 1). Cells were [12]. Briefly, the assay system had 0.2 μg of each mutant protein, 8 μl of
lysed by sonication and centrifuged at 20,000 ×g for 20 min. The su- 100 mM TEA pH 7.4,10 mM EDTA and a 1:100 dilution of SYPRO Or-
pernatant was loaded on a 5 mL HisTrap (GE Healthcare) column. A ange dye (Invitrogen, CA), in a final volume of 10 μl. The dye was ex-
1 mL column was used, for the HsTIMA62D mutant. The proteins were cited at 490 nm and the emitted light intensity was recorded at 575 nm.
eluted with a linear gradient of buffer A containing 500 mM imidazole Data were collected at 1 °C intervals from 25 to 99 °C on a StepOnePlus
and dialyzed for 2 h against a buffer containing 50 mM Tris pH 8.0, real time PCR system using a 96-well reaction plate (Applied Biosys-
0.5 mM EDTA and 1 mM dithiothreitol (DTT). The proteins were then tems 4346907, MA) and analyzed with the Protein Thermal Shift
cleaved using purified recombinant His tagged tobacco etch virus (TEV) Software v1.3 from Applied Biosystems to define the thermal melting
protease expressed using vector pRK508 [10]. The protease was added temperature (Tm). All assays were performed in triplicate.
in a proportion of 1:20 (w/w) and incubated at 30 °C for 18 h. Subse-
quently, the His-tagged TEV protease was removed using a HisTrap
column equilibrated with buffer A. The enzymes were precipitated with 2.5. Dimer stability
ammonium sulfate at 75% saturation and maintained at 4 °C until their
use. All enzymes we analyzed were incubated at the following con-
centrations: 0.01, 0.05, 0.1, 0.5, 1, 2 and 5 μg of protein/mL for 2 h at
2.2. Growth curves of E. coli Δtim- BL21-gold (DE3) cells complemented 36 °C. At that time, the specific activities of the samples were de-
with WT HsTIM and different mutants termined. The amount of enzyme used to measure the specific activity
was different for each mutant (Supplementary Table 1). All assays were
We used a genetically manipulated strain BL21-Gold (DE3) without performed in triplicate. The percentage of residual catalytic activity
the tim gene (E. coli Δtim-BL21-Gold (DE3)) [Saab et al., unpublished was plotted against the logarithm of protein concentration and the data
data]. Cells were grown at 37 °C in M9 minimal medium supplemented were adjusted to a non-linear-fit-three-parameter equation, included in
with glucose (0.2%), casamino acids (0.006%) and ampicillin (100 μg/ the GraphPad Prism 7.0 software.
mL). The assay was performed in triplicate in 96-well, clear, flat-
bottom, plates (Corning Costar 3697), in a total volume of 80 μl in-
oculated either with a single colony of WT bacteria, mutants or cells 2.6. Crystallization and data collection of HsTIMV231M
containing the plasmid without the gene, using a Synergy MX equip-
ment (BioTek, VT). The growth was followed for 16 h and the results HsTIMV231M was crystallized via vapor diffusion using the sitting
were monitored with the software of the equipment (Gen 1.11). drop method. One microliter of a solution at 35 mg/ of protein mL was
mixed with 1 μL of reservoir solution. Crystals were obtained after three
2.3. Activity assay weeks of incubation at 20 °C in condition A6 of the Crystal Screen HT
kit from Hampton Research (200 mM MgCl2, 100 mM Tris pH 8.5, 30%
Enzyme activity was measured at 25 °C following the conversion of polyethylene glycol 4000). The crystals were cryoprotected in a solu-
glyceraldehyde 3 phosphate (GAP) to dihydroxyacetone phosphate tion prepared with the mother liquor supplemented with 20% glycerol;
using α-glycerolphosphate dehydrogenase (α-GDH) as a coupling en- they were immediately frozen in liquid nitrogen. Diffraction data were
zyme, as described by [11]. NADH oxidation was monitored at 340 nm. collected at 100 K using a wavelength of 0.9785 Å at the Life Sciences
The assay system (1 mL) had 100 mM triethanolamine (TEA) pH 7.4, Collaborative Access Team (LS-CAT) 21-ID-G beamline at the Advanced
10 mM EDTA,1 mM GAP, 0.2 mM NADH and 20 μg/mL of α-GDH. The Photon Source (Argonne National Laboratory, IL). The data were pro-
reaction was initiated by the addition of different quantities of the cessed with iMOSFLM [13] and reduced with Aimless [14].

1402
N. Cabrera et al. BBA - General Subjects 1862 (2018) 1401–1409

2.7. Structure determination and refinement

The structure was solved by the molecular replacement method

with the program PHASER [15] using the coordinates of WT HsTIM, at
a resolution of 1.7 Å (Protein Data Bank code 2JK2), as the search
model. The existence of the V231 M mutation was initially confirmed
by difference Fourier maps calculated using the structure of the wild
type enzyme. Refinement was alternated with manual building/re-
finement in COOT [16], PHENIX [17] and the PDB_REDO server [18].
At the start, non-crystallographic symmetry restraints were used for the
refinement, using one group that included the four monomers of the
asymmetric unit. The restraints were released during the last cycles of
refinement, before the addition of water molecules. The electron den-
sity of monomer A is very well defined. For monomers B and C, residues
170–177 (in loop 6 or the flexible loop) are less clear in the electron
density map. These regions are normally poorly defined in all apo-TIMs.
In addition, one of the monomers (D) is somewhat more disordered
than the other three (residues 130–240). The regions that comprise the
active sites were well defined for all monomers. Five percent of the data
were used to validate the refinement. Data collection and refinement
statistics are given in Supplementary Table 2. Figures were generated
with PyMOL (the Pymol Molecular Graphics System, Version 1.7.4,
Schrödinger, LCC).

3. Results

3.1. Cloning and purification

All the mutant enzymes and WT HsTIM were cloned and expressed
in Escherichia coli strain BL21-CodonPlus (DE3)-RIL. To optimize pro-
tein expression, the time and temperature of induction was fine tuned
for each mutant, as well as the amount of NaCl used in the lysis buffer
(Supplementary Table 1). The enzymes were expressed with a His-tag at
the N-terminal position, which was cleaved with TEV protease, as part
of the purification process. In general, the yield of protein was between
20 and 200 mg/L of culture. The purity and homogeneity of all the
proteins (which was more than 95%) was analyzed by SDS-PAGE and
size exclusion chromatography coupled to a multi-angle light scattering
instrument, which showed a dimeric behavior for all the enzymes at a
concentration of 2 mg/mL (data not shown).
HsTIMA62D had a very low yield after purification (5 mg/L of
culture) and had a very strong tendency to precipitate. This behavior
was due to an inherent low stability of the enzyme (Fig. 1, and growth Fig. 2. The enzymatic kinetic constants are affected differently for each mutant. A. Km is
curves described below). Thus, most of the described assays could not increased for mutants G72A (6.5 fold) and V231 M (8.7 fold). It decreases 15-fold for
mutant I170V. B. kcat is affected for mutant I170V (25 fold). C. The catalytic efficiency
be performed with this mutant, and will be shown as not determined
(kcat/Km) is only strongly affected for mutants G72A (4.2 fold) and V231 M (9 fold). The
(nd). assays were performed in triplicate.

3.2. Growth of E. coli Δtim-BL21 gold (DE3) cells complemented with the
different enzymes

In order to test for major survival impairments due to the mutations

of HsTIM, we monitored the growth of bacterial cells lacking their
endogenous tim gene and transformed them with the plasmids over-
expressing WT HsTIM or the mutant enzymes (Fig. 1). Notably, there is
a clear difference in the growth rate of mutant A62D, when compared
to all the other mutants and the WT enzyme. No growth at all is seen for
HsTIMA62D, when the assay is performed at 42 °C (Supplementary
Fig. 1).

Fig. 1. Mutant A62D impairs the growth of a Δtim bacterial strain. Growth curves of
Escherichia coli strain Δtim BL21-Gold (DE3) transformed with the overexpression plasmid
3.3. Enzyme activity
pET-3a (Novagen) encoding WT HsTIM or the mutant enzymes or the empty plasmid. The
logarithm of OD600 is plotted against time of incubation in M9 minimal medium sup- Steady state kinetics assays for all enzymes were determined in the
plemented with 0.006% casamino acids and 0.2% glucose at 37 °C. The assay was per- direction of glyceraldehyde 3-phosphate to dyhydroxyacetone-phos-
formed in triplicate on a Synergy MX microplate reader on 96-well plates. No growth is phate (Table 1 and Fig. 2). All enzymes exhibited a Michaelis-Menten
seen for the A62D mutants when the assay is performed at 42 °C (Supplementary Fig. 1).
behavior. No activity was detected with HsTIMA62D.

1403
N. Cabrera et al.
3.3.1. Km
Regarding this parameter, we found mainly two groups of mutant
enzymes (depicted in Fig. 2A): 1) those that had a “WT-like” behavior,
including HsTIMC41Y, HsTIME104D, HsTIMG122R, HsTIMV154M and
HsTIMF240L and 2) severely affected enzymes, including HsTIMG72A
and HsTIMV231M, which have a Km 6 and 9 times higher than the WT
enzyme, respectively. Interestingly, the Km value of HsTIMI170V is 15
times lower than that of WT HsTIM.
3.3.2. kcat
As shown in Fig. 2B, the only mutant that is strongly affected in this
parameter is HsTIMI170V. Its kcat value is 25 times lower as compared
to that of the WT enzyme. The rest of the mutants have a similar be-
havior to the WT enzyme, with a kcat that is 1.5 times higher than the
WT enzyme for HsTIMG72A and 1.2 times lower for HsTIMC41Y.
3.3.3. Catalytic efficiency (kcat/Km)
Fig. 2C shows that HsTIMG72A and HsTIMV231M are strongly af-
fected in this global parameter, with a respective four- and nine-fold
decrease when compared to the WT enzyme. The catalytic efficiency
values of HsTIMC41Y, HsTIMI170V and HsTIMF240L decreases by half
with respect to the WT enzyme.
3.4. Thermostability
We used thermal shift assays to measure the thermal stability of the
mutants (Fig. 3 and Supplementary Fig. 2). HsTIMG122R,
HsTIMV154M and HsTIMI170V display a melting temperature (Tm)
similar to the WT enzyme (60 to 63 °C). HsTIMV231M and HsTIMF240L
have a Tm of around 53 °C. HsTIMC41Y, HsTIMG72A and HsTIME104D
show a difference of more than 10 °C compared to the WT enzyme, with
Tm values ranging between 49.7 and 51.5 °C.
3.5. Dimer stability
Stability to dilution experiments measure the reduced activity of the
TIM enzymes related to monomerization events [19–21]. We incubated
the HsTIM mutants at different decreasing concentrations for 2 h and
measured the remaining activity (Fig. 4 and Supplementary Fig. 3). At
the lowest enzyme concentration assayed (1 ng/mL), we noticed that
the residual activity showed an important gap for the enzymes that are
still active (between 31 and 57%) and another group of enzymes:
HsTIMC41Y, HsTIMG72A and HsTIME104D, which are strongly af-
fected by the dilution effect and had between 2.6 and 7.7% activity of
their original activity.
Remarkably, we found a strong correlation for some of the mutants
between thermostability data and the results of the dimer stability
1404
BBA - General Subjects 1862 (2018) 1401–1409
assays (Fig. 5). This is the case for the most affected enzymes in these
parameters: HsTIMG72A, HsTIME104D and HsTIMC41Y. There is an-
other group of enzymes that are only slightly affected: HsTIMG122R,
HsTIMV154M and HsTIMI170V. HsTIMV231M has a residual activity
similar to WT HsTIM, but its Tm value is affected. Together with
HsTIMF240L, both mutants represent the borderline between the “WT-
like” group and the most affected mutants.
3.6. Overall comparison and rating of the deleteriousness of the mutant
enzymes
In order to provide an overall and comparative view of this study,
the strength of the effects seen in the different measured parameters
described in Figs. 1-4, compared with the WT enzyme, is summarized in
Table 2. According to these results, we propose a rating of the dele-
teriousness of the mutants, based on their effects on growth (g), kinetics
(k, represented by Km, kcat or kcat/Km) or stability (s, represented by
the thermal or dilution stability) (Fig. 6). The critical mutations are
those whose three parameters (g,k,s) are affected. This is the case for
HsTIMA62D, which was extremely difficult to purify and whose pre-
cipitation behavior precluded the measurement of any parameter. This
mutant cannot support the growth of the Δtim strain of E. coli.
HsTIMG72A is a severe mutant for which two parameters (k and s) are
strongly affected. Other important mutations are those that show a
strong effect in one parameter. This is the case for HsTIMC41Y,
HsTIME104D (with an effect on s) and HsTIMV231M (with an effect on
k). HsTIMI170V and HsTIMF240L have moderate effects, since their
kinetic or stability parameters are only somewhat affected. Finally,
HsTIMG122R and HsTIMV154M do not have any effect on these
parameters of enzyme function.
3.7. Structural basis for the kinetic effects of mutation HsTIMV231M
HsTIMV231M is the most affected mutant from the kinetic point of
view, considering the catalytic efficiency parameter. In addition, this
enzyme is, together with E104D, the only homozygous mutation re-
ported to date. We therefore solved the crystal structure of this mutant
in the apo conformation at 2.3 Å resolution (Fig. 7, Supplementary
Table 2 and Supplementary Fig. 8). The inspection of the four molecules
that constitute the asymmetric unit revealed the structural basis of TIM
deficiency for this mutant (Fig. 7). Comparing it with the available
crystal structures of the WT enzyme, both in the apo conformation (PDB
ID: 2JK2) and in complex with the substrate 2-phosphoglycerate (2PG
or PGA), allowed us to define the conformational changes that occur as
a consequence of the V231 M mutation (Figs. 7A, B). The most im-
portant changes are found just around residue 231 (Fig. 7C), which is
located very close to the active site (Figs. 7A, B). These changes are
Fig. 3. Mutants C41Y, G72A and E104D are
strongly affected in thermal stability. The
Thermal Shift Assay (TSA) was used to measure
the thermal stability of the WT and the mutant
enzymes. Each assay had 0.2 μg of protein in
100 mM TEA pH 7.4 and SYPRO Orange dye in a
final volume of 10 μl. The experiment was per-
formed using a 96-well plate in a real time PCR
system. The assay was performed in triplicate.
The thermal melting temperature was calculated
using the peak of the first derivative plot shown
in Supplementary Fig. 2.
N. Cabrera et al. BBA - General Subjects 1862 (2018) 1401–1409

Fig. 4. Mutants C41Y, G72A and E104D are

strongly affected in dimeric stability. The en-
zymes were incubated at different protein con-
centrations (see Supplementary Fig. 3) in
100 mM TEA pH 7.4 and 10 mM EDTA for 2 h at
36 °C. After that time, their residual catalytic
activity was determined. This residual catalytic
activity was plotted against the logarithm of
protein concentration. The apparent dissociation
constants (Kd app) were determined for the WT
enzyme and each mutant. The values of each
enzyme were adjusted using the Three-parameter
Logistic equation of Prism 7.

Fig. 5. Correlation between thermal and dimeric stability of mutants C41Y, G72A, E104D
and F240L. Double plot of the residual activity at 1 ng/ml (depicted in Fig. 4) and the Tm
temperatures (depicted in Fig. 3). The order of the proteins in the x-axis follows the
residual activity values, from highest to lowest. In general, a mild effect is seen for both
parameters in mutants G122R, V154 M and I170V, and a strong effect in mutants G72A,
E104D and C41Y. Mutants V231 M and F240 L represent a borderline between these two
extremes.
Fig. 6. Deleteriousness rating of the HsTIM mutants. The rating is based upon the severity
of the effects on kinetic parameters (k), stability (s) and growth (g), also shown in Table 2.
Critical: implicates that all these parameters are affected. Severe: both k and s are af-
fected. Important: either k or s are affected. Moderate: a mild effect on k or s.
substrate. This effect is correlated with the change in the kinetic
parameter Km, described in this work.

represented by movements in the main chain around residues 212–219 4. Discussion

and 231–234, which show a maximal displacement of 1.1 and 2.1 Å,
respectively, when compared to both conformations of the WT enzyme
(apo and 2PG, Fig. 7). In particular, the movement of residues 231–234
would approach the main chain of the protein to 2PGA by 0.5 Å
(Fig. 7C), which is now too close, and hinders the binding of the
In this work, we performed an experimental analysis of most of the
reported mutants related to human triosephosphate isomerase defi-
ciency. Other attempts have been made to evaluate the range of effects
on mutants of TIM [9] but this report defines more precisely the extent

Table 2
Summary of the effects seen in different mutants, compared to the wild-type enzyme. The
table summarizes the results shown in Figs. 1-4.

Mutant Kinetic Stability Growth

Km kcat k kcat/Km Tm Dilution

C41Y X XX XX
A62D ND ND ND ND XX
G72A XX XX XX XX
E104D XX XX
G122R
V154M
I170V XX XX X
V231M XX XX X
F240L X X X

XX = strong, X = moderate, blank = no effect, ND = not determined.

1405
N. Cabrera et al. BBA - General Subjects 1862 (2018) 1401–1409

of the active site; the remaining seven mutations, which do have an

effect, are located on the same side as the active site of the enzyme
(Supplementary Fig. 4).
In the following sections of the discussion, we will enlist, following
the numbering in the sequence of HsTIM, the effects reported in the
disease and the main findings observed in this study with corresponding
recombinant mutant protein. In addition, we summarize the molecular
contexts and the effects seen with each mutant (Supplementary Figs. 4
to 9) in order to help in the understanding of the molecular basis of TIM
deficiency and to aid in the prediction of the severity of the mutations
in a heterozygous situation.

4.1. C41Y

4.1.1. Occurrence
Three cases have been reported in compound heterozygotes to-
gether with the E104D mutation [25,27].

4.1.2. Severity in patients

Chronic axonal neuropathy. The damage became evident at the age
of 2 years [27].

4.1.3. Previous experimental characterization

In a yeast model with the deleted endogenous gene tim, this mutant
showed the same activity as the WT enzyme [28]. In addition, the di-
merization properties of this mutant were monitored using a yeast two-
hybrid system. It had an altered dimerization behavior, with increased
interaction activity.

4.1.4. Findings in this work

Thermal and dimer stability strongly affected (13.5 °C difference in
the value of Tm and residual activity at 1 ng/mL, 32% lower than that
of the WT enzyme).

4.1.5. Molecular context and predictions

Residue 41 is in the vicinity of positions 62, 231 and 240
(Supplementary Fig. 7), which have critical, strong and moderate ef-
fects on the enzyme, respectively. A non-conservative change in this
residue might have a strong effect on the stability of the enzyme, as was
indeed seen for mutation C41Y. In combination with the E104D mu-
tation, we predict that the heterodimer would have completely dis-
ruptive effects on the stability of the enzyme. The patient might have
Fig. 7. Structural basis of human TIM deficiency produced by mutation V231M. Mutation limited survival in the presence of the homodimers having each mu-
V231 M perturbs the active site region. A. Cα RMSD plot calculated using apo WT HsTIM tation.
(PDB ID: 2JK2) and the model of the crystal structure of mutant V231 M, solved at 2.3 Å
resolution (PDB ID: 6C2G). The overall RMSD is 0.311 and 0.562 Å for the superpositions
4.2. A62D
of the mutant V231 M with WT HsTIM apo and complexed with 2PG, respectively. The
regions that are displaced are indicated by the residue numbers or structural feature that
differ the most. Loop 6 (residues 169–173) is usually flexible in all TIMs. B. Superposition 4.2.1. Occurrence
of the model of the crystal structure of mutant V231 M with WT HsTIM complexed with One case in a compound heterozygote, with E104D as the second
2PGA, showing the regions of conformational movements depicted in panel A. The re- mutation [22].
gions correspond to residues 134–141, loop 6, 212–219 and 231–234. The substrate 2PGA
is shown as spheres. C. Perturbations at the active site. Regions 212–219 and 231–234 are
4.2.2. Severity in patients
located very close to the active site. The movement of these residues, with its peak at
residue 233, which is displaced by 2.1 Å with respect to the WT structure. This correlates
Unknown.
with the presence of the mutation and hinders the binding of the substrate.
4.2.3. Previous experimental characterization
None.
and cause of the enzymatic deficiency at the biochemical level, in an all
versus all comparative scenario. Although most of the mutants are
4.2.4. Findings in this work
present in humans as heterozygous proteins, as a first approach we
The mutation impairs growth when expressed in bacterial strain E.
studied single mutants in order to investigate their individual con-
coli Δtim.
tribution to the kinetics and stability of human TIM.
We were also able to classify the “unknown severity” mutants
4.2.5. Molecular context and predictions
HsTIMA62D, HsTIMG72A, HsTIMG122R and HsTIMV154M [22–24].
Residue 62 is located in a hydrophobic patch formed by residues
Interestingly, HsTIMG122R and HsTIMV154M do not seem to have any
V40, A42, A62 and V92 (Supplementary Fig. 5). It is possible that a
major biochemical alteration, and only show a very modest effect on
conservative mutation in this position could be tolerated well, but a
thermal stability. Residues 122 and 154 are located on the opposite side
mutation that involves a charged residue, like A62D, completely

1406
N. Cabrera et al.
disrupts this hydrophobic region, which in addition is close to residue
41, and compromises the survival of the organism. In combination with
the E104D mutation, we predict that the heterodimer would have to-
tally disruptive effects on the stability of the enzyme. The patient might
have limited survival in the presence of only one mutant type of
homodimer having mutation E104D.
4.3. G72A
4.3.1. Occurrence
Found in a phenotypic heterozygote, for which TIM activity was
reduced, together with the WT enzyme [24].
4.3.2. Severity in patients
No patients. Study in a population of healthy heterozygotes.
4.3.3. Previous experimental characterization
None.
4.3.4. Findings in this work
Effects upon both kinetic (fourfold decrease in the catalytic effi-
ciency) and stability parameters (11.7 °C difference in the Tm value and
the residual activity at 1 ng/mL, 28% lower compared with the WT
enzyme).
4.3.5. Molecular context and predictions
Residue 72 is located in loop 3, which is very close to the active site
and also forms part of the dimer interface. (Supplementary Fig. 6). Any
change in this position would be severe for both the kinetic and stability
parameters, as is shown by mutation G72A, which would now be too
close to N15 of the dimeric partner (Supplementary Fig. 6). However,
since both residues in positions 72 in both monomers are far apart from
each other in the dimeric enzyme, if a dimer is formed in combination
with one WT monomer, we predict that the heterodimer will have en-
ough stability. The patient might survive in the presence of all types of
homo and heterodimers involving this mutation.
4.4. E104D
4.4.1. Occurrence
At least 15 cases from multiple unrelated families, although most of
the patients are thought to have a common ancestor. HsTIME104D
together with HsTIMV231M, are the only homozygote mutants reported
to date.
4.4.2. Severity in patients
Hemolytic anemia, neuromuscular impairment and increased sus-
ceptibility to infection. No patient has survived beyond the age of six.
4.4.3. Previous experimental characterization
The molecular basis of the deficiency has been reported elsewhere
[5,28] and is caused by alterations in dimer stability, rather than on the
effects in catalytic activity.
4.4.4. Findings in this work
Thermal and dimeric stability are strongly affected (13.5 °C differ-
ence in Tm and the residual activity at 1 ng/mL is 32% lower compared
to the WT enzyme).
4.4.5. Molecular context and predictions
Residue 104 is located very close to the interface and forms a net-
work of interactions with other highly conserved residues
(Supplementary Fig. 7). Any greater change in this position would
compromise the stability of the enzyme, as is implied by the highly
conservative mutation E104D. This is worsened by the fact that the
counterpart of this region in the other monomer is in close proximity.
1407
BBA - General Subjects 1862 (2018) 1401–1409
(Supplementary Fig. 7). A patient has limited survival having
HsTIME104D homodimers.
4.5. G122R
4.5.1. Occurrence
One individual [29].
4.5.2. Severity in patients
No patients. The study was performed with a population of healthy
heterozygotes.
4.5.3. Previous experimental characterization
Reported as a thermolabile enzyme [23]. In a yeast model, with a
deleted endogenous gene for TIM, this mutant had the same activity
compared to the WT enzyme [28]. In addition, the dimerization prop-
erties of this mutant were monitored with a yeast two-hybrid system,
showing no effect.
4.5.4. Findings in this work
The enzyme is not affected. Interestingly, the residual activity at
1 ng/mL is 21.6% higher with respect to the WT enzyme.
4.5.5. Molecular context and predictions
Residue 122 is located on the opposite side of the active site, at the
end of a loop and facing the solvent (Supplementary Fig. 9). Any change
for a hydrophilic residue would have no effect on the enzyme, as is
shown by non-conservative mutation G122R. Patients should show no
symptoms or disease.
4.6. V154M
4.6.1. Occurrence
Found in a phenotypic heterozygote, for which TIM activity was
reduced, together with the WT enzyme [24].
4.6.2. Severity in patients
No patients. The study was performed in a population of healthy
heterozygotes.
4.6.3. Previous experimental characterization
None.
4.6.4. Findings in this work
The enzyme is not affected.
4.6.5. Molecular context and predictions
Residue 154 is located on the opposite side of the active site, in a
hydrophobic patch formed by residues L177, V154, W157 and V160
(Supplementary Fig. 9). Most changes for a hydrophobic residue would
have no effect on the enzyme, as is shown by conservative mutation
V231M. Patients should not show any symptom or disease.
4.7. I170V
4.7.1. Occurrence
Found in one case in a compound heterozygote together with the
E104D mutation [25] [26].
4.7.2. Severity in patients
One patient with hemolytic anemia, but no muscular impairment
[26]. A model in Drosophila has shown behavioral abnormalities [30].
4.7.3. Previous experimental characterization
This mutant had very decreased activity in the patient (10% or less)
when compared to the wild type enzyme, and decreased activity
N. Cabrera et al.
(around 30%) in a yeast model with a deleted endogenous TIM gene. In
addition, the dimerization properties of this mutant were examined
using a two-hybrid system in yeast, with a slight enhancement of the
stability [28]. The mutation lowered Km and kcat values 20 and 30-
fold, respectively, and increased enzyme stability (Tm = 59.4 °C, com-
pared to 46.5 °C of the WT enzyme) [30].
4.7.4. Findings in this work
Enzyme kinetics are moderately affected. The Km and the kcat va-
lues decrease 15 and 25 times, respectively. However, the overall effi-
ciency of the enzyme is not strongly affected (less than twofold). The
stability of the enzyme is unchanged.
4.7.5. Molecular context and predictions
Residue 170 is very close to the active site, in loop 6, which is im-
portant for catalysis (Supplementary Fig. 8). A non-conservative change
in this residue might have a strong effect on the catalysis of the enzyme.
The effect is indeed moderate for mutation I170V, because the enzyme
also shows an increased substrate affinity. The structural basis of the
deficiency caused by this mutant has been analyzed in [30]. In com-
bination with the E104D mutation, we predict that the heterodimer
should be functional, but with severe effects on its kinetic parameters
and stability. Yet the patient with the hetero dimers does survive at
least until late childhood and, patients having homo dimers would
probably not be severely affected.
4.8. V231M
4.8.1. Occurrence
Homozygous mutation linked to clinical deficiency in two cases
[31] [32]
4.8.2. Severity in patients
Hemolytic anemia and neurological dysfunction.
4.8.3. Previous experimental characterization
None.
4.8.4. Findings in this work
Enzyme kinetics are strongly affected. The Km increases nine-fold
and the catalytic efficiency decreases accordingly. The crystal structure
of the mutant shows that the active site is perturbed and hinders the
binding of the substrate (Fig. 7).
4.8.5. Molecular context and predictions
Patients might have limited survival (and see above).
4.9. F240L
4.9.1. Occurrence
Found in two brothers who are compound heterozygotes (E145Stop)
[33,34]; reviewed in [8]. A mutation in the same position (F240S), is
found in a compound heterozygote together with mutation E104D [35].
4.9.2. Severity in patients
Neurological symptoms for only one of the brothers with the F240 L
mutation and also for the patient with the F240S mutation.
4.9.3. Previous experimental characterization
Decreased catalytic activity [36]. In a yeast model with a deleted
TIM endogenous gene, this mutant showed the same activity as com-
pared to the wild type enzyme [28]. In addition, the dimerization
properties of this mutant were measured using a yeast two-hybrid
system, and no effect was detected.
1408
BBA - General Subjects 1862 (2018) 1401–1409
4.9.4. Findings in this work
The mutation has moderate effects upon the kinetics (less than a
twofold decrease in the catalytic efficiency) and stability parameters
(10 °C difference in the Tm value, but a relatively high residual activity
value of 22.6% compared to 35.6% of the WT enzyme).
4.9.5. Molecular context and predictions
Residue 240 is close to positions 41 and 231 (Supplementary Fig. 8).
A conservative change in this residue might have an effect on the sta-
bility and perhaps the catalytic performance of the enzyme because of
its vicinity with V231. The effect is indeed moderate for mutation
F240L and perhaps it would be more severe for mutation F240S in
combination with mutation E104D. We predict that the heterodimer
would be functional, but severely affected in its stability. The patient
might have limited survival in the presence of the heterodimers, with
little contribution to the deficiency because of the F240S homodimers.
The combination of F240L with E145Stop is probably not functional as
a heterodimer, but the F240L homodimer should be functional.
Overall, it is clear that the perturbation of the stability of the qua-
ternary structure may define some TIM deficiencies (mutations C41Y,
E104D) as well as disturbances mainly in kinetic parameters (mutations
I170V, V231M), effects in kinetics and stability (mutation G72A) or
completely aberrant enzymes (mutation A62D) for others. Notably, this
study is in general agreement with experimental studies dealing with
the in vitro characterization of one or several mutants [28]; [5]; [30].
Further studies will be needed to define the molecular effects of some
mutations, notably C41Y, as well as the effects seen in heterodimeric
constructions, which can be expected from the compound heterozygous
mutations found in most cases of this disease.
Data accessibility
The atomic coordinates and structure factors of mutant V231M of
human triosephosphate isomerase have been deposited in the Protein
Data Bank (PDB) with the accession code 6C2G.
Conflict of interests
The authors declare they have no conflict of interest.
Authors' contributions
NC, ATL and RPM conceived and designed the experiments; NC,
ATL, IGT and SEF performed the experiments; NC, ATL, IGT, SEF and
RPM analyzed the data; NC, ATL and RPM wrote the manuscript. All
authors gave final approval for publication.
Transparency document
The Transparency document associated this article can be found, in
the online version.
Acknowledgements
We thank Gloria Saab-Rincón, PhD and Leticia Olvera-Rodríguez,
BSc (Instituto de Biotecnología-UNAM, México) for providing the E. coli
Δtim-BL21-Gold (DE3) strain. We also thank Gabriel Del Río, PhD, who
kindly allowed the use of the Synergy MX equipment, María-Teresa
Lara-Ortiz, Jessica Díaz-Salazar, Beatriz Aguirre, Manuel Ortinez
Benavides, Aurey Galván Lobato and Unidad de Biología Molecular at
Instituto de Fisiología Celular-UNAM (Dr. Laura Ongay and Guadalupe
Codiz) for technical support.
Use of the Advanced Photon Source, an Office of Science User
Facility operated for the U.S. Department of Energy (DOE) Office of
Science by Argonne National Laboratory, was supported by the U.S.
DOE under Contract No. DE-AC02-06CH11357. Use of the LS-CAT
N. Cabrera et al.
Sector 21 was supported by the Michigan Economic Development
Corporation and the Michigan Technology Tri-Corridor, Grant
085P1000817. This work was supported by grant 254694 from Consejo
Nacional de Ciencia y Tecnología, México, to RPM.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bbagen.2018.03.019.
References
[1] S.V. Rieder, I.A. Rose, The mechanism of the triosephosphate isomerase reaction, J.
Biol. Chem. 234 (1959) 1007–1010.
[2] J.R. Knowles, Enzyme catalysis: not different, just better, Nature 350 (1991)
121–124.
[3] A.R. Katebi, R.L. Jernigan, The critical role of the loops of triosephosphate iso-
merase for its oligomerization, dynamics, and functionality, Protein Sci. 23 (2014)
213–228.
[4] S.C. Mande, V. Mainfroid, K.H. Kalk, K. Goraj, J.A. Martial, W.G. Hol, Crystal
structure of recombinant human triosephosphate isomerase at 2.8 A resolution.
Triosephosphate isomerase-related human genetic disorders and comparison with
the trypanosomal enzyme, Protein Sci. 3 (1994) 810–821.
[5] C. Rodriguez-Almazan, R. Arreola, D. Rodriguez-Larrea, B. Aguirre-Lopez, M.T. de
Gomez-Puyou, R. Perez-Montfort, M. Costas, A. Gomez-Puyou, A. Torres-Larios,
Structural basis of human triosephosphate isomerase deficiency: mutation E104D is
related to alterations of a conserved water network at the dimer interface, J. Biol.
Chem. 283 (2008) 23254–23263.
[6] A.S. Schneider, Triosephosphate isomerase deficiency: historical perspectives and
molecular aspects, Baillieres Best Pract. Res. Clin. Haematol. 13 (2000) 119–140.
[7] F. Orosz, J. Olah, J. Ovadi, Triosephosphate isomerase deficiency: facts and doubts,
IUBMB Life 58 (2006) 703–715.
[8] F. Orosz, J. Olah, J. Ovadi, Triosephosphate isomerase deficiency: new insights into
an enigmatic disease, Biochim. Biophys. Acta 1792 (2009) 1168–1174.
[9] C. Oliver, D.J. Timson, In silico prediction of the effects of mutations in the human
triose phosphate isomerase gene: towards a predictive framework for TPI defi-
ciency, Eur. J. Med. Genet. 60 (2017) 289–298.
[10] R.B. Kapust, D.S. Waugh, Escherichia coli maltose-binding protein is uncommonly
effective at promoting the solubility of polypeptides to which it is fused, Protein Sci.
8 (1999) 1668–1674.
[11] B. Plaut, J.R. Knowles, pH-dependence of the triose phosphate isomerase reaction,
Biochem. J. 129 (1972) 311–320.
[12] F.H. Niesen, H. Berglund, M. Vedadi, The use of differential scanning fluorimetry to
detect ligand interactions that promote protein stability, Nat. Protoc. 2 (2007)
2212–2221.
[13] T.G. Battye, L. Kontogiannis, O. Johnson, H.R. Powell, A.G. Leslie, iMOSFLM: a new
graphical interface for diffraction-image processing with MOSFLM, Acta
Crystallogr. D Biol. Crystallogr. 67 (2011) 271–281.
[14] P.R. Evans, An introduction to data reduction: space-group determination, scaling
and intensity statistics, Acta Crystallogr. D Biol. Crystallogr. 67 (2011) 282–292.
[15] A.J. McCoy, Solving structures of protein complexes by molecular replacement with
Phaser, Acta Crystallogr. D Biol. Crystallogr. 63 (2007) 32–41.
[16] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta
Crystallogr. D Biol. Crystallogr. 60 (2004) 2126–2132.
[17] P.D. Adams, P.V. Afonine, G. Bunkoczi, V.B. Chen, I.W. Davis, N. Echols, J.J. Headd,
L.W. Hung, G.J. Kapral, R.W. Grosse-Kunstleve, A.J. McCoy, N.W. Moriarty,
R. Oeffner, R.J. Read, D.C. Richardson, J.S. Richardson, T.C. Terwilliger,
P.H. Zwart, PHENIX: a comprehensive python-based system for macromolecular
structure solution, Acta Crystallogr. D Biol. Crystallogr. 66 (2010) 213–221.
1409
BBA - General Subjects 1862 (2018) 1401–1409
[18] R.P. Joosten, F. Long, G.N. Murshudov, A. Perrakis, The PDB_REDO server for
macromolecular structure model optimization, IUCrJ. 1 (2014) 213–220.
[19] T.V. Borchert, R. Abagyan, K.V. Kishan, J.P. Zeelen, R.K. Wierenga, The crystal
structure of an engineered monomeric triosephosphate isomerase, monoTIM: the
correct modelling of an eight-residue loop, Structure 1 (1993) 205–213.
[20] T.V. Borchert, J.P. Zeelen, W. Schliebs, M. Callens, W. Minke, R. Jaenicke,
R.K. Wierenga, An interface point-mutation variant of triosephosphate isomerase is
compactly folded and monomeric at low protein concentrations, FEBS Lett. 367
(1995) 315–318.
[21] V. Mainfroid, S.C. Mande, W.G. Hol, J.A. Martial, K. Goraj, Stabilization of human
triosephosphate isomerase by improvement of the stability of individual alpha-he-
lices in dimeric as well as monomeric forms of the protein, Biochemistry 35 (1996)
4110–4117.
[22] L. Manco, M.L. Ribeiro, Novel human pathological mutations. Gene symbol: TPI1.
Disease: triosephosphate isomerase deficiency, Hum. Genet. 121 (2007) 650.
[23] B.A. Perry, H.W. Mohrenweiser, Human triosephosphate isomerase: substitution of
Arg for Gly at position 122 in a thermolabile electromorph variant, TPI-Manchester,
Hum. Genet. 88 (1992) 634–638.
[24] M. Watanabe, B.C. Zingg, H.W. Mohrenweiser, Molecular analysis of a series of
alleles in humans with reduced activity at the triosephosphate isomerase locus, Am.
J. Hum. Genet. 58 (1996) 308–316.
[25] R. Arya, M.R. Lalloz, A.J. Bellingham, D.M. Layton, Evidence for founder effect of
the Glu104Asp substitution and identification of new mutations in triosephosphate
isomerase deficiency, Hum. Mutat. 10 (1997) 290–294.
[26] A. Ationu, A. Humphries, M.R.A. Lalloz, R. Arya, B. Wild, J. Warrilow, J. Morgan,
A.J. Bellingham, D.M. Layton, Reversal of metabolic block in glycolysis by enzyme
replacement in triosephosphate isomerase-deficient cells, Blood 94 (1999)
3193–3198.
[27] J.M. Wilmshurst, G.A. Wise, J.D. Pollard, R.A. Ouvrier, Chronic axonal neuropathy
with triosephosphate isomerase deficiency, Pediatr. Neurol. 30 (2004) 146–148.
[28] M. Ralser, G. Heeren, M. Breitenbach, H. Lehrach, S. Krobitsch, Triose phosphate
isomerase deficiency is caused by altered dimerization–not catalytic inactivity–of
the mutant enzymes, PLoS One 1 (2006) e30.
[29] J. Asakawa, H.W. Mohrenweiser, Characterization of two new electrophoretic
variants of human triosephosphate isomerase: stability, kinetic, and immunological
properties, Biochem. Genet. 20 (1982) 59–76.
[30] B.P. Roland, C.G. Amrich, C.J. Kammerer, K.A. Stuchul, S.B. Larsen, S. Rode,
A.A. Aslam, A. Heroux, R. Wetzel, A.P. VanDemark, M.J. Palladino,
Triosephosphate isomerase I170V alters catalytic site, enhances stability and in-
duces pathology in a Drosophila model of TPI deficiency, Biochim. Biophys. Acta
1852 (2015) 61–69.
[31] B.A. Neubauer, A. Pekrun, S.W. Eber, M. Lakomek, W. Schröter, Relation between
genetic defect, altered protein structure, and enzyme function in triose-phosphate
isomerase (TPI) deficiency, Eur. J. Pediatr. 151 (1992) 232.
[32] G. Serdaroglu, Y. Aydinok, S. Yilmaz, L. Manco, E. Ozer, Triosephosphate isomerase
deficiency: a patient with Val231Met mutation, Pediatr. Neurol. 44 (2011)
139–142.
[33] M.L. Chang, P.J. Artymiuk, X. Wu, S. Hollan, A. Lammi, L.E. Maquat, Human
triosephosphate isomerase deficiency resulting from mutation of Phe-240, Am. J.
Hum. Genet. 52 (1993) 1260–1269.
[34] S. Hollan, H. Fujii, A. Hirono, K. Hirono, H. Karro, S. Miwa, V. Harsanyi, E. Gyodi,
M. Inselt-Kovacs, Hereditary triosephosphate isomerase (TPI) deficiency: two se-
verely affected brothers one with and one without neurological symptoms, Hum.
Genet. 92 (1993) 486–490.
[35] E. Fermo, P. Bianchi, C. Vercellati, D.C. Rees, A.P. Marcello, W. Barcellini,
A. Zanella, Triose phosphate isomerase deficiency associated with two novel mu-
tations in TPI gene, Eur. J. Haematol. 85 (2010) 170–173.
[36] F. Orosz, J. Olah, M. Alvarez, G.M. Keseru, B. Szabo, G. Wagner, Z. Kovari,
M. Horanyi, K. Baroti, J.A. Martial, S. Hollan, J. Ovadi, Distinct behavior of mutant
triosephosphate isomerase in hemolysate and in isolated form: molecular basis of
enzyme deficiency, Blood 98 (2001) 3106–3112.