Reactivities of a,b-Unsaturated Carboxylic Esters
FULL PAPER
entries 2 and 6). Changing the electron-donating a-methyl
moiety to an electron-withdrawing a-chloro substituent
translated into greatly improved reactivity: all enzymes
showed enhanced conversion with methyl 2-chloroacrylate
(7a) compared with 6a, in particular OYE3, YqjM, OPR3,
and OPR1, which displayed a 1.4 to 3.8-fold improvement in
conversions (Table 1, entry 7). It is tempting to relate this
effect of electronic activation to the mechanism of OYEs,
because the reaction proceeds through a Michael-type addi-
tion through nucleophilic attack of a hydride onto the b-
carbon followed by proton addition onto the a-position.[51]
Thus, the additional electron-withdrawing a-chloro group
(in contrast to the methyl moiety) enhances the degree of
C=C polarization, thereby favoring the overall addition of
[2H]. However, comparing 7a with unsubstituted methyl ac-
rylate 2a, it appears that also steric factors play a role. Al-
though (electronically expected) higher reaction rates were
achieved with YqjM and OPR1, OYE3 and OPR3 showed
reduced conversion with nonactivated 7a. Moderate-to-high
that the a-cyano group would enhance the C=C bond polari-
zation (Figure 1B). Unfortunately, substrates 15a–17a un-
derwent decomposition and/or polymerization under screen-
ing conditions (data not shown) and were not further inves-
tigated.
As an alternative, we tested b-substituted a-haloacrylic
ester derivatives. To our delight, methyl 2-chloropentenoate
[(Z)-8a] was nicely accepted and yielded (S)-8b with excel-
lent stereoselectivity (conversion up to>99%, >99% ee) in
the presence of OYEs 1–3 and YqjM (Table 1, entry 8). This
supports the hypothesis that C=C bond activation can be
tuned by the presence of a-halo substituents, because the
corresponding a-unsubstituted derivative [(E)-12a] was not
converted at all. By using the glucose/glucose dehydrogen-
ase (NAD+/GDH) cofactor recycling system, 97% conver-
sion and 98% ee were obtained by using OYE2 (Table 1,
entry 9). Interestingly, NCR and OPR1 showed moderately
expressed opposite stereopreference [ee up to 72% for (R)-
8b]. The E-isomer was similarly well accepted, and the
change of isomer geometry had a strong influence on the
stereochemical outcome for some of the enzymes: OYE1,
OYE2, YqjM, and OPR3 showed a strong switch, that is,
(Z)-8a gave (S)-8b and (E)-8a led to (R)-8b. In contrast,
OYE3 always formed (S)-8b, and NCR and OPR1 were
always R-selective, regardless of the E/Z-configuration of
substrate 8a (Table 1, entry 10). This behavior has been pre-
viously observed with OYEs.[53] The cofactor recycling
system yielded noticeable variations in conversion and ee
values (Table 1, entry 11). Such an influence of the nicotin-
S-stereoACHTUNGTRENNUNGselectivity was obtained with OYEs 1–3 (40–83%
enantiomeric excess (ee)), whereas YqjM, NCR, and OPRs
showed excellent opposite stereopreference (86 up to
>99% ee for (R)-7b).
Next, we investigated the effects of b-substituents. Inter-
estingly, the addition of an alkyl substituent, regardless of its
size, totally deactivated the system (compounds 9a–12a,
<1% conversion, Figure 1A). Herein, the result can also be
rationalized by considering the reaction mechanism. Indeed,
the electron-donating group diminished the d+ at the b-
carbon; in addition, the b-position may be particularly sensi-
tive towards steric hindrance, because the hydride has to be
delivered at this point from N5 of the flavin cofactor. In
need of additional activation of unreactive b-alkyl- or b-
aryl-substituted acrylic esters, we looked for potential a-acti-
vating groups. Because methyl (Z)-2-ethoxy-3-(4-methoxy-
phenyl)propenoate with a moderately electron donating a-
alkoxy group was not converted by OYE3,[52] a-cyano-sub-
stituted acrylates, such as methyl 2-cyanoacrylate (15a),
methyl 2-cyano-3,3-dimethylacrylate (16a), and methyl 2-
cyano-3-phenylpropenoate [(E)-17a], were tested, expecting
AHCTUNGTREGaNNNU mide regeneration system on the bioreduction reaction has
been observed in numerous cases.[54–57] Although not fully
understood, it may result from differences in the (bi-bi Ping-
Pong) kinetics between ene-reductase and recycling enzyme,
leading to variations in the ratio oxidized/reduced cofactor.
Prompted by these results, a series of a-halogenated cin-
namic esters (18a–20a) were tested to overcome the insuffi-
cient degree of activation in (unreactive) methyl cinnamate
((E)-13a, <1% conversion, Figure 1A). Again, the intro-
duction of an a-halo substituent proved to be successful, be-
cause all substrates were converted with OYEs 1–3 (Table 2,
entries 1–8). Among them, OYE3 showed highest activities
(conversion 36–97%) and S-stereoselectivity (79% up to
>99% ee) with methyl a-chloro-, bromo-, and iodocinna-
mates. The ability of OYE3 to accept larger substrates is
likely due to an enlarged binding pocket caused by an ex-
change of phenylalanine F296 (in OYE1 and OYE2) by
serine S297 in OYE3.[58,59] The type of halogen had a strong
effect on the activity of OYE3: although a-fluoro-substitu-
tion led to an unreactive substrate ((Z)-14a, <1% conver-
sion, Figure 1A), the Cl-, Br-, and I-analogues (Z)-18a, (Z)-
19a, and (Z)-20a were successfully converted at different
rates, and the a-bromo derivative (Z)-19a was the most re-
active of the series (Table 2, entry 3). The S-stereoprefer-
ence of OYEs 1–3 was conserved throughout the series 18a–
20a, reduced stereoselectivity was observed with the steri-
cally most demanding iodo analogue (Z)-20a (ee 79–80%).
The E/Z-configuration of the brominated substrate had no
Figure 1. A) Nonsubstrates for ene-reductases (<1% conversion, data
not shown). B) Substrates undergoing polymerization or decomposition.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
3
&
ÞÞ
These are not the final page numbers!