N.N. Vorobjeva, et al.
B
B
A-GeneralSubjects1864(2020)129601
digesting the protein with concentrated HNO3 at 200–250 °C.
Hsp70, which form tight complexes with o-KduI [8].
Prompted by the above structural analysis, sugar isomerization ac-
tivity of recombinant o-KduI was tested against a panel of D-sugars and
their derivatives: glucose-6-phosphate (G6P), fructose-6-phosphate
(F6P), glucose-1-phosphate, mannose-6-phosphate, arabinose, xylose,
fructose, glucose, galactose, glucuronate, galacturonate, and tagatur-
onate. Hexuronates were used at 10 mM concentration, and all other
sugars at 40 mM. The potential substrates were incubated with 0.2 mg/
ml o-KduI at pH 7.5 for 5 min (G6P and F6P) or 2.5 h (other sugars),
and ketose concentration in the assay mixture was measured with the
resorcine method. Significant changes in absorbance were only ob-
served with glucose-6-phosphate and glucuronate (ketose content in-
creased) and fructose-6-phosphate (ketose content decreased). The
calculated activity values for the three substrates were 2.9, 0.12 and
4.7 IU/mg, respectively. These findings indicated that o-KduI exhibits a
glucose-6-phosphate isomerase activity and a smaller glucuronate iso-
merase activity. Notably, the value of glucose-6-phosphate isomerase
activity of E. coli o-KduI is similar to that of P. furiosus cPGI [19,20].
The ability of o-KduI to catalyze G6P isomerization was further
corroborated by measuring equilibrium concentrations of G6P and F6P
after their interconversion. These sugars were separately incubated
with o-KduI, and the concentration of ketose (i.e. F6P) in the reaction
mixture was monitored for 3 h (Fig. 3A). In both cases, a constant level
change after addition of fresh enzyme, indicating that a true equili-
brium was attained. The equilibrium constant for the isomerization
reaction, [F6P]eq/[G6P]eq, can be calculated from these data as 0.32,
which agrees well with the values reported by others [11].
2.4. Sedimentation
Analytical ultracentrifugation was performed at 20 °C in a Spinco E
instrument (Beckman Instruments) equipped with a computerized data
collection unit, with scanning at 280 nm. Samples contained 0.4 mg/ml
o-KduI, 0.1 M Tris/HCl, pH 7.5, and 10 mM NaCl. The sedimentation
velocity was measured in triplicate at 60,000 rpm, sedimentation
coefficients (s20,w), which characterize the ability of macromolecules to
sediment under the action of centrifugal acceleration, and molecular
masses were calculated with the program SedFit [www.
3. Results
3.1. Comparison of the crystal structures of o-KduI and P. furiosus
phosphoglucose isomerase
Protein Data Bank contains two very similar crystal structures of o-
KduI determined at a resolution of 1.94 and 2.3 Å (PDB ID: 1XRU [17]
is composed of two β-barrel domains connected by a short linker. The
domains are structurally similar, but only the C-terminal domain con-
tains a competent metal binding pocket (Fig. 2A). This pocket contains
a metal ion tentatively identified as zinc [17]. Three dimers form a
homohexamer.
Comparison with the available three-dimensional structures of other
cupin proteins using PDBeFOLD [18] has revealed close similarity be-
hyperthermophylic archaeon P. furiosus (cPGI, PDB ID: 2GC2 [19]),
bound metal ion, tentatively identified as Zn2+, and fructose-6-phos-
phate in both subunits, whereas the o-KduI structure contains pre-
sumable Zn2+ in one of the two cupin domains. In both structures, the
metal ion is four-coordinated via three His and one Glu residues typical
of cupins. cPGI subunit is formed by only one cupin domain, therefore
the structures of the two-domain subunit of o-KduI and cPGI homo-
dimer were superimposed for comparison. The superposition by Cα
atoms of two-domain o-KduI and two-subunit cPGI yields RMSD of
1.29 Å (48 pairs of matching atoms). A clear similarity of the overall
structures was evident (Fig. 2B) despite differences in domain sizes.
Furthermore, at least two important substrate-binding residues of cPGI
(Thr71 and Tyr99) and the key catalytic general base (Glu97) have
clear structural matches in the C-terminal metal-bound domain of o-
KduI (Fig. 2C). The o-KduI cavity containing these residues could easily
accommodate a fructose-6-phosphate molecule with a similar co-
ordination. This structural analysis pinpointed thus the C-terminal do-
main cavity as the active site in o-KduI. In contrast, the corresponding
cavity in the metal-free domain is markedly distorted, disallowing
fructose-6-phosphate binding to it. Moreover, the metal-binding re-
sidues are absent in this domain, explaining its inability to bind metal
ion.
3.3. Sugar isomerization kinetics
The rates of the glucose-6-phosphate ⇆ fructose-6-phosphate iso-
merization reaction in both directions exhibited sigmoidal de-
pendencies on initial substrate concentration (Fig. 3B), indicating po-
sitive kinetic cooperativity. The value of the Hill coefficient (h), which
characterizes the degree of the cooperativity, was close to 2.0 in both
cases (Table 1). The use of the Hill equation allowed also estimation of
the maximal specific activity (Amax) and the concentrations of the
substrate leading to half-maximal activity (K0.5) in both directions
(Table 1). Both parameters demonstrated only small dependence on the
reaction direction, consistent with the high reversibility of the iso-
merization rection.
The pH dependence of the rate of the o-KduI-catalyzed isomeriza-
tion reaction measured at a saturating substrate concentration (i.e., kcat
)
indicated that the reaction in both directions requires a basic group
with a pKa value of 6.2 0.1 (Fig. 3C). In addition, F6P conversion is
The limiting residual activity of F6P conversion in the alkaline medium
was estimated to be 55
10% of the maximal activity.
3.4. Oligomeric structure
Sedimentation velocity data (Fig. 4) and the parameters derived
partially dissociated into dimers after incubations with EDTA and, more
readily, after incubation with substrate, G6P, as indicated by appear-
ance of a species with a lower sedimentation coefficient, s20,w, in Fig. 4B
strate not only stimulated dissociation of o-KduI, but also made its
structure more compact, as indicated by a shift of the peaks in Fig. 4C to
lower s20,w values.
3.2. o-KduI production and substrate specificity
E. coli o-KduI gene was expressed in the host cells from an en-
gineered plasmid. The produced protein without any tag was purified
by hydrophobic and ion exchange chromatography. The final yield of o-
KduI in terms of protein content was 10–15 mg per 1 l culture. SDS-
PAGE analysis of the final preparation showed one major ~30 kDa band
and several minor higher-mass bands (Fig. A.1). The major and five
minor bands were analyzed by an MALDI-TOF mass spectrometry. The
major and two minor bands contained only o-KduI, and three other
minor bands were identified as known o-KduI partners, fructose-1,6-
bisphosphate aldolase, L-glutamate decarboxylase, and chaperone
The specific activities of o-KduI pre-equilibrated at different con-
centrations were measured to further define the conditions that would
3