5630
A. Patterson-Orazem et al. / Bioorg. Med. Chem. 22 (2014) 5628–5632
Enzyme-catalyzed alkene reductions were carried out at room
temp for 24 h at pH 7.0 in total volumes of 0.30 mL with 10 mM
substrate and 4.6 lM fusion protein. Conversions and stereoselec-
tivities (when appropriate) were assessed by GC after extracting
with EtOAc.
In contrast to b-substituted 2-cyclohexenones, a-substitution
does not significantly impair reduction rates by alkene reductases.
This was readily apparent in conversions of (R)- and (S)-carvone (8
and 9, respectively), which were reduced almost quantitatively by
all proteins examined (Fig. 5). In addition, no change in stereo-
chemical preference was observed for reductions of 8 and 9 when
Ile 113 was substituted by other amino acids. This contrasts shar-
ply with the behavior of analogous mutants of S. pastorianus OYE.19
We tested two exocyclic enones, (R)-pulegone 10 and n-butyl-
substituted 11 since these have proven to be challenging sub-
strates for alkene reductases. The presence of a tetrasubstituted
alkene in 10 makes this enone especially resistant to enzyme-cat-
alyzed reductions. As noted above, efficient reduction requires that
the ketone oxygen and alkene b-carbon occupy specific locations
and while it is not possible to place 10 and 11 exactly congruently
with the most preferred positions, the arrangements shown in Fig-
ure 6 are reasonable approximations. The wild-type and most
mutants reduced 10 with good efficiencies and with consistent ste-
reoselectivities that could be explained by the binding orientation
shown in Figure 6. When compared with the parent 2-cyclohexe-
none, most of the added steric bulk of (R)-pulegone 10 resides on
the ‘western’ side of the active site. Even without additional muta-
genic sculpting, this region of the OYE 2.6 active site is relatively
open, consistent with the observed behavior. By contrast, alkene
11 presents a more difficult challenge since the n-butyl chain
impinges on the locations of side-chains at positions 78 and 113
and the ring requires space on the ‘western’ side, near the
side-chain at position 247. Not unexpectedly, alkene 11 is a poor
substrate for wild-type OYE 2.6 (ca. 20% conversion after 24 h);
however, conversion was increased four-fold when the Tyr 78
Trp/Ile 113/Phe 247 His triple mutant was substituted. This is a
very significant improvement and clearly demonstrates the value
of combining mutations on both the ‘eastern’ and ‘western’ sides
of the active site.
2.2. Cyclohexyl substrates
In common with all conjugate additions, enzyme-catalyzed
alkene reductions are acutely sensitive to steric hindrance at the
b-carbon. This is one factor that makes alkenes 4–7 challenging
substrates. Beyond their intrinsically disfavored b,b-disubstitution
patterns, reduction rates can be further depressed by steric clash
between the exocyclic b-substituent and active site protein moie-
ties. Indeed, an earlier study using S. pastorianus OYE to reduce
2-cyclohexenones 4–7 found a very significant negative correlation
between b-substituent size and reduction rates and usable conver-
sions were only obtained for 4 and 5.18a A similar trend was
observed for wild-type P. stipitis OYE 2.6, where even the best sub-
strate (4) gave only 55% conversion after 24 h (Fig. 4). Replacing
Tyr 78 with Trp, either singly or with concomitant changes at posi-
tion 113 and/or 247 gave >98% conversion of 4 within the same
time period. This is a significant improvement. Similar trends were
observed for 5 and 6, although the optimal replacements for Ile 113
differed (Met and Leu, respectively). The Tyr 78 Trp/Ile 113 Met
double mutant reduced 7 to the greatest extent; however, the
low conversion for even this ‘best’ mutant lessens the practical
utility. In each case, better rates were obtained from enzymes with
smaller side-chains at position 113, consistent with a steric
explanation of the results. It might be possible to obtain even bet-
ter conversions by exploring additional, smaller Ile 113 replace-
ments. It is important to note that the mutations did not erode
stereoselectivity, and only the (S)-products were detectable by
GC in all cases.
8
9
100
80
100
10
80
60
60
4
11
40
40
5
6
20
20
0
7
0
Figure 4. Catalytic activities of OYE 2.6 and mutants for alkenes 4–7. Data are
grouped by substrate (4, green; 5, yellow; 6, orange; 7, red). In all cases,
(S)-products were obtained in P98% ee.
Figure 5. Catalytic activities of OYE 2.6 and mutants for alkenes 8–11. Data are
grouped by substrate (8, green; 9, yellow; 10, orange; 11, red). In all cases, products
were obtained in P98% de (8–10) or P98% ee (11).