M. Rabuffetti et al.
Bioorganic Chemistry 108 (2021) 104644
the hydroxyl moiety of Tyr150 to the substrate to form the alcohol
product. This step is favored by the H-bond network that links the ribose
of the co-factor with Lys154 and Tyr150, which ultimately lowers the
electronegativity of Tyr150 hydroxyl oxygen, facilitating the proton loss
and assisting in the stabilization of the resulting negative charge. The
enzyme is eventually regenerated by water molecules (Fig. 4).
KRED1-Pglu was experimentally reviewed using different symmetric
and asymmetric benzil derivatives (e.g., diaryl 1,2-diketones). High
enantioselectivity was always observed, whereas chemoselectivity to-
wards asymmetric benzil was mostly dependent on the substitution on
the aromatic rings. The structural rationale for the stereoselective
monoreduction of diketones by KRED1-Pglu was revealed by the crystal
structure, combined with QM molecular dynamics simulations.
2
.5. In silico prediction of substrates reactivity
3
. Materials and methods
We evaluated in silico the reactivity of selected dicarbonyls, which
3
.1. Chemicals, reagents and enzymes
were experimentally reduced by KRED1-Pglu with different rates and
conversions. The correct positioning of the substrate for the reaction to
occur was evaluated by assessing the presence of a non-covalent inter-
action between NADPH hydride and the carbonyl; the appropriate pose
determines which stereoface is attacked by the hydride released by
NADPH. QM/MM geometry optimizations was employed to determine
how the positioning of substrates within KRED1-Pglu binding pocket
could affect the outcome of the reaction. Indeed, the relatively tight
shape of the enzyme binding pocket makes the enzyme specific and
enantioselective. When the pose of the substrate was close enough to
have a productive interaction, Fukui function indices (f+) were calcu-
lated to predict the electrophilicity of the C of each carbonyl group,
depending on the orientation of the substrate (si or re face) within
KRED1-Pglu active site. For instance, the positioning of benzil (1a) to
produce (R)-benzoin is geometrically incompatible with the nucleo-
philic attack by NADPH hydride, and only f+ corresponding to the
electrophilicity of the C placed for producing the (S)-benzoin could be
simulated; the same situation is observed in the reduction of ethyl 2-oxo-
All reagents were purchased from Sigma-Aldrich (Milan, Italy) and/
or from VWR International (Milan, Italy) and were used without further
purification. All the solvents were of HPLC grade. Analytical Thin Layer
Chromatography TLC was performed on silica gel 60 F254 precoated
aluminum sheets (0.2 mm layer; Merck, Darmstadt, Germany); compo-
nents were detected either under an UV lamp (λ 254 nm) or with a
KMnO
4
solution (1.5 g of KMnO
4
, 10 g of K
2
CO
3
and 1.25 mL of 10%
◦
NaOH in 200 mL of H
2
O), followed by heating at about 150 C. Sub-
strates 1a, 1b, 1h, 1i, 1j, 1k, 1r and 1s were purchased from Sigma-
Aldrich (Milan, Italy) and/or from VWR International (Milan, Italy).
Substrates 1c, 1d, 1e, 1f and 1g were synthesized by oxidation of their
respective benzoins (see SI for details) [25]. Substrates 1l, 1m, 1n, 1o,
1
p and 1q were synthesized as reported in Supporting Information.
3
.2. Enzymes
The gene sequence of KRED1-Pglu (GenBank: AKP95857) was pre-
2
-phenylacetate 1h, where only ethyl (S)-2-hydroxy-2-phenylacetate 2k
viously identified from the genome of the non-conventional yeast Pichia
glucozyma (subsequently reclassified as Ogataea glucozyma) [11].
KRED1-Pglu and GDH from Bacillus megaterium DSM509 expression in
E. coli BL21(DE3) strains and their production were carried out as pre-
viously described [11].
is formed.
1
-Phenylpropane-1,2-dione 1b is the most reactive substrate and it is
reduced by KRED1-Pglu with high enantioselectivity and high regiose-
lectivity, since only the carbonyl in position 2 (adjacent to the methyl
group) is reduced. In our model, the position of the substrate is
compatible with the hydride attack on both carbonyls; moreover, hy-
dride delivery is possible on both si- and re-face of C1, whereas only the
attack to the si face is geometrically viable when C2 is attacked.
Calculation of f+ did not allow for understanding the discrimination
between the two carbonyls and activation energy of the reaction was
calculated as the difference between the energies of the initial optimized
geometries and the highest energy intermediate of the coordinate scan.
This calculation suggests that totally selective reduction of the si face of
carbonyl at C2 is due mostly to a much lower activation energy, thus
explaining the experimentally observed high regio- and enantiose-
lectivity. The simple 1,3-dione 1k (1-phenylbutane-1,3-dione) is poorly
reactive and this is confirmed by the little favored pose of the substrate
resulting in a low activation observed only for the carbonyl adjacent to
the methyl group.
3
.3. KRED1-Pglu activity test
Activity measurements were performed spectrophotometrically at
◦
3
40 nm by determining the consumption of NADPH at 25 C in a half-
microcuvette (total volume
Spectrometer spectrophotometer). One unit (U) of activity corresponds
to the amount of enzyme which catalyzes the consumption of 1 mol of
1 mL) for 5 min (Eppendorf Bio-
μ
ꢀ 1
ꢀ 1
NADPH (
ε = 6220 M cm ) per minute under reference conditions,
namely with 0.25 mM NADPH and 0.47 mM substrate (added as DMSO
solution, final DMSO concentration in cuvette amounts to 0.1%), in 50
mM Tris⋅HCl buffer (pH 8.0).
3.4. GDH activity test
A full transition state optimization was also performed to assess that
these intermediates were indeed valid transition states; the structure and
a clip of the transition state corresponding to the reduction of benzil is
reported in SI.
Activity measurements were performed spectrophotometrically at
◦
340 nm by determining the formation of NADPH at 25 C in a half-
microcuvette (total volume
Spectrometer spectrophotometer). One unit (U) of activity corresponds
to the amount of enzyme which catalyzes the reduction of 1 mol of
1 mL) for 5 min (Eppendorf Bio-
In summary, the monoreduction of different 1,2 dicarbonyls giving
homochiral benzoins (1,2-diaryl-2-hydroxyethanones) catalyzed by
μ
Fig. 4. Proposed mechanism for the monoreduction of benzil catalyzed by KRED1-Pglu.
5
Rabuffetti, Marco
