G. Tjallinks et al.
Archives of Biochemistry and Biophysics 704 (2021) 108888
found to be most potent and was the only AOX variant used for further
studies.
Table 1
Conversions[a] and kinetic parameters[b] for the oxidation of secondary alcohols
by AOX. [c]
substrate
.
Conversions with all identified secondary alcohols were performed
for 24 h and it was observed that AOX selectively oxidizes the (S)-al-
cohols to the corresponding ketones. Fig. 1 illustrates the enantiose-
lective conversion of the (S)-enantiomer when AOX is incubated with
racemic 1-phenylethanol. The (R)-alcohols were left almost untouched.
As a consequence, high E values of >90 were obtained for 1-indanol, 1-
phenylethanol and 2-pentanol (Table 1). Only for 3-butyn-2-ol a some-
what lower enantioselectivity was observed (E value = 9). AOX was
found to be most active towards 2-pentanol (Table 1). 2-Pentanol is a
relatively small compound with a flexible aliphatic chain which could
allow for facile entry into the active site and for obtaining an optimal
conformation for subsequent alcohol oxidation. The reaction rate of 1-
phenylethanol and 3-butyn-2-ol were lower but in the same range as
2-pentanol. Overall, the reaction rates were relatively low, however
almost full conversions could be obtained for the secondary alcohols
within 24 h.
Conversion[d] (%) e.e.[e] (%) E[f]
kobs (sꢀ 1
)
1-indanol (1)
50
>99
>200 n.d.[g]
1-phenylethanol (2)
43
72
93
0.13 ± 0.01
2-pentanol (3)
50
42
>99
>200 0.38 ± 0.02
15 0.12 ± 0.01
3-butyn-2-ol (4)
57
a
Values obtained using 20 mM substrate and 20 μM AOX in 100 mM potas-
sium phosphate pH 7.5, 24 h at 35 ◦C. Details can be found in the Supporting
Information.
b
Values obtained using 200 mM substrate and 0.2 μM AOX in 100 mM po-
tassium phosphate pH 7.5.
The underlying reason for the observed (S)-enantioselectivity of AOX
was examined by inspecting the crystal structure and performing in silico
substrate docking. The selectivity can be explained by investigating the
role of active-site residues and residues that form the substrate binding
pocket. Just as for other members of the GMC-oxidoreductase super-
family, AOX contains a strictly conserved catalytic base His561 [18–21].
In addition, it holds a spatially conserved Asn604 residue that, as
postulated previously, most likely functions as a hydrogen bond donor
during catalysis. Both residues structurally occupy the Re-face of the
flavin cofactor, hence the substrate binding domain is also located on the
c
Commercial enantiopure substrates were used as standards to establish the
absolute configuration.
d
Conversion is based on the amount of formed ketone, except for 3-butyn-2-ol
which was based on racemic substrate depletion.
e
f
Enantiomeric excess of substrate.
g
Value could not be obtained due to insolubility of the substrate at high
concentrations, this was indicated as n.d. (not determined).
holo enzymes. No addition of FAD during purification was required
which indicates a tight binding of the flavin cofactor (Fig. S1). A typical
UV–Vis absorption spectrum was obtained for the purified AOX variants
with an absorption maximum at 455 nm, suggesting that the flavin is in
the oxidized state. The substrate acceptance profile and enantiose-
lectivity of wild-type AOX for secondary alcohols was analyzed using
chiral GC analysis and a set of 30 different aromatic and aliphatic sec-
ondary alcohols (see Table S2 for the complete overview of screened
substrates). This revealed that wild-type AOX was able to convert four
secondary alcohols: 1-indanol, 1-phenylethanol, 2-pentanol, and 3-
butyn-2-ol. Activity of AOX on these apolar aromatic and aliphatic
secondary alcohols had not yet been reported before (Table 1). Three
AOX mutants (Thr315Ser, Leu317Phe and Trp560Phe AOX), which are
expected to have an altered substrate binding pocket, were tested for
their conversion of 1-indanol, 1-phenylethanol, and 2-pentanol. How-
ever, all AOX mutants displayed low or no significant activity and/or
lower enantioselectivity (Fig. S4-S6). Ultimately, wild-type AOX was
Re-face. In the oxidation reaction, proton abstraction of the
group by the histidine catalytic base His561 with a simultaneous or
stepwise with hydride transfer from the substrate C atom to flavin N5
α-hydroxyl
α
atom occurs [19,21–23]. Hence, the specific orientation of the substrate
towards the FAD cofactor, His561 and Asn604 is decisive in the enan-
tioselective outcome of the reaction. Docking of 1-phenylethanol
revealed that residues Phe101, Phe399, Phe422, Tyr424 and Trp560
create an apolar cavity in which the substrate can comfortably undergo
van der Waals and pi-stacking interactions (Fig. 2). Consequently, only
(S)-1-phenylethanol can bind in the active site in a productive manner:
the hydride can only be transferred to the Re-face of the flavin cofactor.
The aromatic ring of 1-phenylethanol prevents the formation of any
other conformation necessary to convert the (R)-enantiomer. Docking of
the other three substrates resulted in analogous optimal binding poses,
in which the hydrophobic environment drives the observed (S)-enan-
tioselectivity (Figs. S10-S12).
The enantioselectivity of wild-type AOX was then compared to the
Fig. 1. GC chromatogram after 24 h conversion of 20 mM racemic 1-phenylethanol (2) by 20 μM AOX in 100 mM potassium phosphate buffer pH 7.5.
2