.
Angewandte
Communications
corresponding aldehyde, thus yielding furan-2,5-dicarbalde-
hyde (4), followed by additional oxidations to 5 and 1.
Because HMFO was shown to be active towards both 3 and 4,
both routes would be feasible for HMFO.
varied. In addition, FAD was added to prevent the inactiva-
tion of HMFO because of cofactor dissociation. HMFO relies
on a noncovalently bound FAD cofactor for catalysis and it
can dissociate from the enzyme, thus rendering it inactive.
The effect of additional FAD is marginal but clear, thus
increasing the yield of 1 from 5% to 6.1% (Table 1, entries 2
and 3). High yields of 1 were obtained using a biocatalyst/
substrate ratio of 1:200. A reaction with 20 mm HMFO and
4 mm 2 produces a yield of 95% of 1 after 24 hours at ambient
temperature and pressure (Table 1, entry 4). It is worth noting
that the residual activity of HMFO was close to 100% after
24 hours of incubation. This data shows that the biocatalyst is
rather robust and suitable for prolonged incubations.
To gain more insight into the catalytic properties of
HMFO, the conversion of 2 and its derivatives was monitored
over time. Reactions were started upon addition of the
enzyme and samples were taken during conversion. The
reaction was terminated by heating the sample for five
minutes at 708C to denature the enzyme. All samples were
analyzed by HPLC. Using 2 as a substrate, the oxidation
products 4, 5, and 1 were detected during the conversion. The
compound 4 was found to be the main intermediate in the first
hours of the reaction, whereas increasing amounts of 5 and
The oxidation route (Scheme 1) raises the question as to
why HMFO is an efficient catalysis for oxidation of aldehydes
like 4 and 8, but leaves the aldehyde groups of 2 and its
phenylic analogue 7 unaffected. Alcohol oxidation by FAD-
dependent oxidases is typically initiated by the abstraction of
a proton from the alcohol group by an active-site base, and is
followed by a hydride transfer from the a-carbon atom of the
1
appeared from the 4 to 15 hour time point (Table 1 and see
Table 1: Oxidation of furans to 1 at 258C in a 100 mm potassium
phosphate buffer of pH 7.0. The reaction time was 15 h, unless stated
otherwise.
Entry
Sub-
strate
[mm]
HMFO
[mm]
FAD
[mm]
Conv.
[%]
Yield [%]
FDCA
[
11]
[
a]
alcohol to the FAD cofactor.
With aldehydes, such
a mechanism is not possible and may be the reason why
most FAD-dependent oxidases are not capable of oxidizing
1
2
3
4
5
6
7
6
2
2
2
4
3
5
2
2
2
4
2
2
2
1
1
1
20
1
–
–
10
20
–
100
100
100
100
100
99
4.4Æ0.2
5.0Æ0.1
6.1Æ0.5
[12]
aldehydes.
[
b]
However, several oxidases have been reported to perform
aldehyde oxidation reactions. In all these cases, the aldehyde
itself is not oxidized, but the hydrated gem-diol form is
95Æ1.7
5.9Æ0.1
6.1Æ0.6
7.4Æ0.1
1
1
–
–
[
13–15]
7.4
converted by the enzyme .
To investigate the mechanism
1
8
of oxidation by HMFO, O-labeled water was used during
catalysis (Scheme 2). When the hydrated aldehyde is oxidized
[a] 100% conversion means the substrate is fully converted. [b] The
reaction time was 24 h.
Tables S1 and S2 in the Supporting Information). The
exclusive formation of 4 shows that route B is the preferred
oxidation pathway and route A is not used by HMFO
(
Scheme 1).
To determine whether the preference for route B is valid
for other substrates as well, we expanded the analysis to the
phenylic analogues of the examined furans. By using the
phenylic analogue of 2, 4-(hydroxymethyl)benzaldehyde (7),
as the substrate, the conversion was monitored over time.
Similar to the experiments on the furanic compounds,
terephthaldehyde (8), 4-formylbenzoic acid (9), and tereph-
thalic acid (10) were formed. The phenylic analogue of 3, 4-
Scheme 2. Hydration and oxidation of an aldehyde. The formed gem-
diol is oxidized by the enzyme to form the carboxylic acid. The
1
8
presence of H2 O in the reaction will lead to the formation of
a product with m/z values of M+2 and M+4.
(
hydroxymethyl)benzoic acid (11), was again not formed as
an intermediate (see Table S6, S7, and S12 in the Supporting
Information).
to a carboxylic acid by HMFO, one water-derived oxygen
1
8
These findings imply that HMFO oxidizes the alcohol
group of 2 and 7 first, rather than the aldehyde substituent.
This forces the enzyme to perform an aldehyde oxidation as
the next step, as 4 and 8 do not contain an alcohol group. The
good activity towards the furanic and phenylic dialdehydes (4
and 8) is surprising because HMFO does not react with the
aldehyde groups of 2 and 7. The subsequent oxidation of
aldehydes 5 and 9 into the final product is much less efficient
as evidenced by the presence of these oxidation intermediates
as main products when relatively little enzyme was used
atom ( O) will be incorporated. In contrast, if the oxidation is
1
6
spontaneous, O from molecular oxygen will be incorporated.
To distinguish between these reactions, the oxidation product
[14]
of the readily hydrated 4-nitrobenzylaldehyde, formed by
HMFO, was analyzed by LC-MS. In unlabeled water, the
formed product has the expected mass of 166 (MS in negative
mode), thus corresponding to 4-nitrobenzoic acid. A similar
1
8
experiment, performed in 30% H2 O, gives additional mass
peaks of M+2 and M+4 (Figure 1). This shows that the
inserted oxygen originates from the hydration of the aldehyde
and HMFO subsequently oxidizes the hydration product to
the carboxylic product.
(
Table 1, entries 1–3 and 5–7). To optimize the yield of 1,
the pH value, temperature, and enzyme concentration were
6
516
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 6515 –6518