Substrate scope and synthetic applications of the enantioselective reduction of α-alkyl-β-arylenones mediated by Old Yellow Enzymes

Article abstract of DOI:10.1039/c3ob40076j

The ene-reductases mediated bioreduction of a selection of open-chain α-alkyl-β-aryl enones afforded the corresponding saturated α-chiral ketones in high yield and optical purity in several cases. The stereo-electronic requirements of the reaction have been investigated, considering the nature and location of substituents on the aromatic ring as well as the steric hindrance at the α-position and adjacent to the carbonyl functionality. The general considerations drawn allow us to guide the design of α,β-unsaturated ketones to be employed as substrates of ene-reductases in future preparative applications. An interesting case of orthogonality between enzyme-based and substrate-based stereocontrol within the highly homologous ene-reductases from Saccharomyces species (OYE1-3) has been reported and rationalized with the help of computational docking studies. Furthermore, to demonstrate the synthetic versatility of the reaction, the key chiral precursors of biologically active compounds such as (2′R)- stenusines and (S)-iopanoic acid were obtained. The very robust protocol allowed us to run the reactions on preparative scale in quantitative yields, with a simple work-up and no chromatographic purification steps. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ob40076j

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Substrate scope and synthetic applications of the
enantioselective reduction of α-alkyl-β-arylenones
mediated by Old Yellow Enzymes†
Cite this: Org. Biomol. Chem., 2013, 11,
2988
Elisabetta Brenna,a Sara Lucia Cosi,b Erica Elisa Ferrandi,b Francesco G. Gatti,a
Daniela Monti,b Fabio Parmeggiani*a and Alessandro Sacchettia
The ene-reductases mediated bioreduction of a selection of open-chain α-alkyl-β-aryl enones afforded
the corresponding saturated α-chiral ketones in high yield and optical purity in several cases. The stereo-
electronic requirements of the reaction have been investigated, considering the nature and location of
substituents on the aromatic ring as well as the steric hindrance at the α-position and adjacent to the car-
bonyl functionality. The general considerations drawn allow us to guide the design of α,β-unsaturated
ketones to be employed as substrates of ene-reductases in future preparative applications. An interesting
case of orthogonality between enzyme-based and substrate-based stereocontrol within the highly homo-
logous ene-reductases from Saccharomyces species (OYE1–3) has been reported and rationalized with
the help of computational docking studies. Furthermore, to demonstrate the synthetic versatility of the
reaction, the key chiral precursors of biologically active compounds such as (2’R)-stenusines and (S)-iopa-
noic acid were obtained. The very robust protocol allowed us to run the reactions on preparative scale in
quantitative yields, with a simple work-up and no chromatographic purification steps.
Received 14th January 2013,
Accepted 12th March 2013
DOI: 10.1039/c3ob40076j
www.rsc.org/obc
microbiological reduction of α,β-unsaturated ketones mediated
by whole cells of various microorganisms (mainly Saccharo-
Introduction
Optically active α-substituted ketones are very interesting myces cerevisiae, known as baker’s yeast,6,7 but also other
chiral building blocks for the synthesis of many drugs and yeasts and fungi like Rhodotorula rubra8 and Beauveria bassi-
natural products, e.g., pheromones.1 These valuable com- ana9 or bacteria such as Pseudomonas fluorescens10). This
pounds are usually prepared by different chemical methods, system is very simple, the reaction proceeds in aqueous
such as the metal-catalyzed asymmetric hydrogenation of suit- medium under mild and environmentally friendly conditions
ably substituted α,β-unsaturated ketones,2 the asymmetric and it applies to a very wide range of substrates. Furthermore,
protonation of prochiral enols and equivalents,3 the in the case of the synthesis of flavours and fragrances, an
enantioselective α-alkylation with chiral auxiliaries4 or the additional advantage that boosted the biotechnological
hydroacylation of 1-alkenes.5 Each of these general strategies is research towards this aim is the possibility to employ the pro-
well-established, but shows distinctive disadvantages, e.g., the ducts as “natural” aromas for the food industry (e.g., raspberry
use of stoichiometric chiral reagents, the limited substrate ketone), since the transformation is performed by a biological
scope, the high cost of the catalysts and the requirement for catalyst.11
toxic heavy metals.
Notably, whole cells mediated processes usually suffer from
As opposed to the chemical methods listed above, biocataly- several drawbacks which hinder their large-scale application,
sis offers a highly enantioselective alternative which is the e.g., the scarce tolerance to high substrate loadings, the poor
chemoselectivity arising from the presence of competitive
enzymatic activities (for instance carbonyl reduction) and the
aDipartimento di Chimica, Materiali e Ingegneria Chimica “G. Natta”,
Politecnico di Milano, Via Mancinelli 7, 20131 Milano, Italy.
E-mail: fabio.parmeggiani@chem.polimi.it; Fax: +39 02 2399 3180;
Tel: +39 02 2399 3077
consequent need for chromatographic separation steps down-
stream to remove undesired by-products.
In the last few decades, the development of molecular
biology techniques allowed us to address these issues by
affording overexpressed and purified enzymes in sufficient
amounts and at a reasonable price. The enzymes responsible
for the reduction of CvC double bonds activated by an
bIstituto di Chimica del Riconoscimento Molecolare – CNR, Via Mario Bianco 9,
20131 Milano, Italy
†Electronic supplementary information (ESI)
available. See DOI:
10.1039/c3ob40076j
2988 | Org. Biomol. Chem., 2013, 11, 2988–2996
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
electron-withdrawing group are called ene-reductases,12 most
of which belong to the family of the Old Yellow Enzymes
(OYEs). These biocatalysts mediate the stereospecific anti
hydrogenation of activated olefins, based on the transfer of a
formal hydride from a flavin mononucleotide (FMNH2) pros-
thetic group (regenerated by the NAD(P)H cofactor) to the
β-carbon atom, followed by the delivery of a proton from the
hydroxyl group of a tyrosine residue to the α-carbon atom of
the resulting enolate.12a Since 1991, when the first member of
the family (OYE1) was cloned and overexpressed from Saccharo-
myces pastorianus (brewer’s yeast),13 many other homologs
quickly followed, such as OYE2 and OYE3 from baker’s yeast.14
As a consequence, nowadays more and more chemists are
given the opportunity to exploit the activity of these versatile
enzymes in the synthesis of chiral compounds in optically
pure form.15
Nevertheless, in spite of the large number of papers dealing
with whole cells bioreduction of α,β-unsaturated ketones avail-
able in the literature, up to now systematic investigations to
assess the behaviour of isolated ene-reductases on this class of
substrates have only been carried out on cyclic enones (e.g.,
alkylcyclopentenones and substituted cyclohexenones).12a,16,17
Herein we present the results of the bioreduction of several
open-chain α-alkyl-β-aryl enones mediated by OYE1–3, demon-
strating also the relevant role of the reaction in preparative
organic synthesis by obtaining key precursors of a natural alka-
loid and a radiocontrast agent.
Scheme 1 Synthesis and OYE1–3 mediated reduction of α-alkyl-β-aryl enones
1–11. i. H2SO4, AcOH; ii. NaOH, H2O–EtOH; iii. H2, Pd/C, EtOAc.
Stereo-electronic effects in the bioreduction
of β-arylenones
purified to homogeneity by affinity chromatography on a Ni-
At first, a set of representative substrates was chosen in order NTA stationary phase.
to investigate the influence of stereo-electronic factors in the
After 12 h, extraction of the reaction mixture afforded the
ene-reductase mediated reduction of enones (Scheme 1). In optically enriched saturated ketones 12–22 (and residual start-
particular, the selection of substrates 1–11 (Table 1) was meant ing material, if the conversion was not complete). The conver-
to illustrate the effects of increasing steric hindrance either at sion was determined by GC-MS and the enantiomeric excess
the α-position (R1) or on the opposite side of the carbonyl by GC on a chiral stationary phase (racemic samples of the pro-
group (R2), and the influence of the presence and the position ducts 12–22 were obtained by hydrogenation of the corres-
of an electron-donating substituent on the phenyl ring (R3).
ponding starting materials as shown in Scheme 1). The results
Enones 1–11 were easily synthesized by aldol condensation of the bioreduction are reported in Table 1. No evidence of
of an aryl aldehyde with the suitable aliphatic ketone. With double bond isomerisation under the reaction conditions
non-symmetric ketones the condensations were performed was apparent by GC-MS analysis of the unconverted starting
under acidic conditions (to afford the more substituted materials.
product), while in the case of symmetric ketones the conden-
The biotransformations of all the substrates which were
sations were carried out under basic conditions, which provide converted substantially on analytical scale were also carried
higher yields. In all cases only the (E)-stereoisomers were out on a larger scale (50 mg) with a suitable amount of the
obtained (diastereoisomeric excess >95% by 1H-NMR).
best biocatalyst, in order to achieve complete conversion of the
The substrates were then submitted to bioreduction substrate. Such a procedure allowed the isolation of the bio-
mediated by isolated OYE1–3 (Scheme 1). The regeneration of transformation product for characterization and measurement
the expensive NADPH cofactor was performed in situ with a of the specific optical rotatory power, without the need for a
recombinant glucose dehydrogenase (GDH) from Bacillus chromatographic purification.
megaterium, using glucose as a sacrificial cosubstrate. All the
The absolute configurations of the products were deter-
enzymes used (OYE1–3 and GDH) were overexpressed in mined by comparison of the specific optical rotatory power
Escherichia coli BL21 (DE3) cells in His-tagged form and values with those reported in the literature. For compounds
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2988–2996 | 2989

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 OYE1–3 mediated bioreduction of α,β-unsaturated-β-arylketones 1–11
OYE1
OYE2
OYE3
Substrate
Product
Conv.a (%)
eeb (%)
Conv.a (%)
eeb (%)
Conv.a (%)
eeb (%)
1
2
3
4
5
6
7
8
12
13
14
15
16
17
18
19
20
21
22
>99
>99
>99
92
67
3
53
41
<1
>99
16
56 (S)
97 (S)
95 (S)
42 (S)
59 (R)
n.d.
20 (R)
60 (R)
n.d.
>99
61
>99
64
9
<1
11
26
<1
52
3
61 (S)
95 (S)
97 (S)
59 (S)
76 (R)
n.d.
42 (R)
69 (R)
n.d.
>99
46
>99
45
14
<1
8
98 (S)
96 (S)
99 (S)
97 (S)
89 (S)
n.d.
95 (S)
81 (S)
n.d.
7
9
10
11
<1
>99
6
99 (S)
98 (S)
99 (S)
n.d.
99 (S)
n.d.
a Conversion values determined by GC-MS. b Enantiomeric excess values determined by GC on a chiral stationary phase.
18, 19 and 22 whose [α]D values are not available in the litera- (S)-enantiomer (81–95% ee) while the other two homologues
ture, the absolute configuration was assigned by chemical cor- provide the opposite absolute configuration, though with
relation with known compounds (see ESI†).
lower ee values (up to 76% ee).
Both the activity and the selectivity of the enzymes turned
The strong preference for (S)-enantiomers is consistent with
out to be very high in several instances, ensuring the possi- the stereochemical outcome of OYE mediated reductions of
bility to integrate such reactions as a chirality generation step many related non-cyclic α,β-unsaturated substrates12a (e.g.,
in more complex multistep syntheses, as will be exemplified in α-substituted cinnamaldehydes18,19), which originates by the
the last section. Concerning the substrate scope of the biore- binding mode of the substrate to the active site shown in
duction, the following general considerations could be made:
Fig. 1A. In the literature, this is referred to as “flipped”
(i) Increasing the steric hindrance at the α-position (R1) has binding mode.12a,18
a contrasting effect: switching from Me (1) to Et (10) to n-Bu
Nevertheless, in some cases, the preference for the (R)-
(11), the conversions decrease markedly while the enantio- enantiomers was observed. In principle, if a syn hydrogen
meric excess (ee) values increase. In the case of substrate 11 addition mechanism is excluded, the latter observation might
the conversion is too low to be of practical significance, be explained only by the exposure of the opposite stereohetero-
though the ee observed in the OYE1 catalyzed bioreduction is topic face of the α-carbon to the hydroxyl group of the tyrosine
still excellent. The same general trend has been recently residue, i.e., either by a 180° rotation of the whole substrate
observed also in the closely related family of α-alkylcinnamal- within the active site (Fig. 1B, known as the “classical” binding
dehydes,18 though with higher conversions and lower optical mode for historical reasons) or by the accommodation of the
purities.
s-cis rotamer instead of the more stable s-trans one (Fig. 1C),
(ii) All the enzymes tested are very sensitive to the steric hin- as proposed in a recent molecular modelling study on the
drance of the chain adjacent to the carbonyl functionality (R2) known cases of stereopreference inversion.20
since the conversions of ethyl ketones 5–8 are all considerably
Although an enzyme-based enantioselectivity switch has
lower than those of the corresponding methyl ketones 1–4 and already been reported in several instances,20,21 it has rarely
the isopropyl ketone 9 is not accepted at all. Indeed, OYE1 been observed for the three ene-reductases from Saccharomyces
seems more tolerant than the two homologues, providing the (OYE1–3) because of their high sequence homology (92% iden-
highest conversions.
tity between OYE1 and OYE2, 80% between OYE1 and OYE3,
(iii) The presence of electron-donating substituents such as 82% between OYE2 and OYE3). In particular, to the best of our
the methoxy group on the aromatic ring (R3) considerably knowledge, a stereoinversion between OYE1–2 and OYE3 has
decreases the conversion and the effect is more apparent when only been described for dimethyl (Z)-2-butyramidofumarate21b
the electron-donating effect is stronger (o- and p-substituted and for methyl (E)-2-chloro-2-pentenoate.21c
rings, 2, 4, 6, 8). Specifically, in the case of enone 6, the pres-
Even more interestingly, a closer inspection of the results
ence of a methoxy group in the ortho position together with an shown in Table 1 reveals that the stereopreference inversion is
ethyl chain next to the carbonyl functionality provided such a not only enzyme-based (e.g., substrate 5 gives (R)-16 with
low yield that the product could barely be detected.
OYE1–2 and (S)-16 with OYE3), but also substrate-based, since
(iv) Concerning the stereoselectivity, for almost all of the the same enzyme can afford products with opposite configur-
substrates the absolute configuration of the product is invari- ation starting from either methyl ketones or the corresponding
ably (S) with the three proteins, and the highest ee values are ethyl ketones (e.g., OYE1–2 give the (S)-enantiomer from 1 and
always provided by OYE3. Instead, in the bioreduction of ethyl the (R)-enantiomer from 5). This behaviour, which demon-
ketones 5, 7 and 8 only OYE3 retains the preference for the strates the remarkable effectiveness of “substrate engineering”
2990 | Org. Biomol. Chem., 2013, 11, 2988–2996
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 1 Binding modes of substrate 5 to the active site of OYEs and potential explanations of the stereopreference inversion observed in the bioreduction of enones
5–8. A: “flipped” binding mode; B: “classical” binding mode, resulting from a 180° rotation; C: binding of the s-cis rotamer (numbering of the residues according to
the pdb file 1OYB).
in the control of the stereochemical outcome, provides a rep- β-carbon atom of enone 1 were 3.48 Å, 3.71 Å and 3.85 Å for
resentative example of “orthogonality” between enzyme-based OYE1–3, respectively.
and substrate-based stereocontrol strategies.
The docking study of enone 5 revealed instead that the
same binding mode is retained only with OYE3 (Fig. 2F), while
the “classical” configuration shown in Fig. 1B is preferred with
OYE1 and OYE2 (Fig. 2D and E), keeping the carbonyl oxygen
and the β-carbon atom essentially fixed. This substrate flipping
necessarily leads to the opposite enantiomer of the product, as
shown above (Fig. 1B), accounting for the stereoinversion
observed. None of the results involved the binding of the s-cis
Rationalization of the enzyme-based and
substrate-based stereopreference inversion
by computational studies
In order to identify the most likely explanation of the stereo- rotamer shown in Fig. 1C, neither did any of the other higher
chemical outcome observed, a docking study of methyl ketone energy poses obtained from the docking simulation. In these
1 and ethyl ketone 5 to the active sites of the three ene- cases, the distances between the N(5) of FMNH2 and the
reductases was carried out. The structures of OYE2 and OYE3 β-carbon atom of enone 5 were 4.00 Å, 4.29 Å and 4.39 Å for
were modelled on the X-ray crystallographic data of OYE1 OYE1–3, respectively.
obtained from the PDB database (entry code: 1OYB). The
results are shown in Fig. 2.
Interestingly enough, the role of the residue 296 seems
determinant for the observed stereoselectivity control. In fact,
For enone 1 the “flipped” binding mode sketched in Fig. 1A in OYE3 the repositioning of the phenyl group due to the
can be observed with all the three biocatalysts (Fig. 2A–C), increased steric hindrance at the R2 position is well tolerated,
completely matching with the case of α-substituted cinnamal- thanks to the reduced size of Ser296 with respect to the
dehydes18,19 and reliably in agreement with the observed (S) phenylalanine residue. Therefore, the typical binding mode is
stereoselectivity. The carbonyl group is held in position by still productive, although with slightly reduced enantioselecti-
interaction with the hydrogen bond donor residues (His191 vity. Instead, in the case of OYE1–2, the substrate cannot fit in
and Asn194), and the phenyl ring is accommodated in a hydro- the active site with the same orientation without a clashing of
phobic pocket formed by the aromatic residues of Tyr82, its aromatic ring with the Phe296 residue, so the flipped
Phe123, Phe250, Phe296 and Tyr375, conserved in the three binding mode is adopted as preferential. Therefore, the
proteins (except for Phe296 which is replaced with Ser296 in docking study confirms and justifies the unusual stereochemi-
OYE3). The distances between the N(5) of FMNH2 and the cal outcome observed experimentally.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2988–2996 | 2991

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 2 Binding modes of methyl ketone 1 (shown in purple) and ethyl ketone 5 (shown in orange) to the active sites of OYE1–3 obtained by docking studies. A, B,
C: 1 to OYE1, OYE2, OYE3, respectively; D, E, F: 5 to OYE1, OYE2, OYE3, respectively.
(2′S,3R)/(2′R,3R)/(2′R,3S)-23, S. clavicornis almost only (2′S,3R)-
Exploitation of the bioreduction in the
synthesis of relevant chiral building blocks
23, while S. juno mainly the (2′R,3R)-isomer.25 The pygidial
fluid spread on the beetles’ bodies also acts as an antibacterial
agent and as a deterrent against ants and other predators.
Since stenusine 23 has been demonstrated to be the active
Even though in some instances the absence of a definite pre-
ference for one of the multiple binding modes available
ingredient in both these functions, to evaluate the differences
results in moderate or low ee values (e.g., 7), many OYE
in the biological activity of the four stereoisomers, several
mediated bioreductions provide quantitative yield and excel-
lent stereoselectivity (e.g., 3) and are very valuable for prepara-
tive applications. In order to demonstrate the practical
synthetic value of the reaction in the obtainment of highly
enantioenriched α-chiral ketones, two more open-chain
α-alkyl-β-arylenones were synthesized and reduced to afford
key chiral precursors of a natural product and an active
pharmaceutical ingredient.
attempts have been made to stereoselectively synthesize them
all in optically pure form.25a,26 The two (2′R)-isomers are the
most challenging, since the literature methods either start from
the commercially unavailable enantiomer of a natural chiral
reagent or do not allow strict stereocontrol. Very recently the
total synthesis of all four stereoisomers of 23 was successfully
carried out by Seifert et al.,26e the access to (2′R)-stereoisomers
being provided by the development of a chemoenzymatic syn-
thesis (Scheme 2). Specifically, the approach was based on the
baker’s yeast mediated bioreduction of the suitable β-pyridyl
A. Synthesis of the (2′R)-stenusines precursor
The rove beetles of the genus Stenus Latreille are well known enone 24, already reported in 20016c within the large number
for their outstanding capability of locomotion on the surface of papers on yeast mediated reduction of enones. Although the
of water, with up to 70 cm s−1 speed. From their pygidial original publication described a yield of 76% and a very high ee
glands they release a surface-active fluid which spreads very value (>95%) of the saturated ketone (S)-25, disappointingly the
rapidly on the water providing the thrust to jet over the Seifert group was not able to reproduce the same results,
surface.22 The main component of the fluid, responsible for neither with fresh nor with lyophilized baker’s yeast. After an
this property, is a piperidine alkaloid isolated and identified extensive optimization of the protocol, a similar yield but only
in the mid 1970s,23,24 called stenusine (23) after the name of a lower ee (92%) were obtained.26e
the genus.
Hence, we envisaged the opportunity to improve this syn-
The pygidial secretions of different Stenus species contain thetic strategy by exploiting isolated ene-reductases in the
stenusine with a very diverse stereoisomeric pattern: S. comma reduction of the β-(3-pyridyl) enone 24 according to our
produces a 43 : 40 : 13 : 4 mixture of the stereoisomers (2′S,3S)/ protocol.
2992 | Org. Biomol. Chem., 2013, 11, 2988–2996
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 2 Chemoenzymatic enantioselective synthesis of (2’R)-stenusines. i. H2SO4, AcOH; ii. baker’s yeast, glucose, H2O; iii. OYE/GDH, glucose, NADP+, phosphate
buffer pH 7.0; iv. H2, Pd/C, EtOAc; v. (CF3CO)2O, H2O2, DCM.
Table 2 Baker’s yeast and OYE1–3 mediated bioreduction of α,β-unsaturated
ketone 24
B. Synthesis of the (S)-iopanoic acid precursor
Iopanoic acid (27, also known as iodopanoic acid) is a radio-
opaque contrast agent largely employed in cholecystography
for X-ray diagnosis of gallstones. It is administered orally,
adsorbed very rapidly and excreted with bile.
It has also been subjected to clinical studies for the treat-
ment of hyperthyroidism and its mechanism of action has
been thoroughly studied.27 In particular, it has been found
Biocatalyst
Conv.a (%)
eeb (%)
Baker’s yeastc
Baker’s yeastd
OYE1
OYE2
OYE3
76
62
>99
>99
>99
95
92
95
94
98
a Conversion values determined by GC-MS. b Enantiomeric excess that the two enantiomers are converted into their respective
glucuronides at different rates.28 Accordingly, much effort has
values determined by HPLC on a chiral stationary phase of ester 26
obtained by the Baeyer–Villiger reaction. c Reported from ref. 6c.
d Reported from ref. 26e.
been directed not only to the obtainment of both the enantio-
mers of 27, through classical29 and enzymatic30 resolution,
but also to the stereoselective synthesis of the desired (S)-
enantiomer.2,6b
It is worth mentioning that in this particular case none of
In this respect, one of the approaches proposed was based
the available chiral stationary phases for GC or HPLC analysis on baker’s yeast mediated reduction of enone 28 (Scheme 3),
proved effective in separating the enantiomers of the product affording the almost optically pure saturated ketone (S)-29 in a
25. Also the attempt to discriminate between the two enantio- 56% yield.6b Unfortunately, also in the latter case the authors
1
mers by H NMR, adding the chiral shift reagent Eu(hfc)3, was reported a considerable decrease in yield upon scaling-up the
fruitless. Eventually, we chose to perform a Baeyer–Villiger standard whole cells protocol.
reaction on the samples, converting ketone 25 into the corres-
ponding acetate ester 26 (Scheme 2), whose enantiomers were isolated ene-reductases OYE1–3 in the reduction of enone 28
easily separated by HPLC analysis.
with the standard protocol on an analytical scale. In this case,
Therefore, we investigated the performances of the three
The results of the biotransformations are reported in racemic 29 to be used as an analytical reference could not be
Table 2, along with the best ones previously obtained with prepared by hydrogenation of the unsaturated precursor 28,
yeast whole cells. The data agree with the general trends due to the presence of the nitro group. It was then synthesized
already highlighted, a quantitative conversion being ensured by alkylation of ethyl 2-ethylacetoacetate 30 with m-nitrobenzyl
by the small methyl substituents, and a very high, though not bromide, followed by hydrolysis and decarboxylation of the
complete, optical purity.
intermediate 31 (Scheme 3).
In addition, the reaction was successfully carried out on a
The results of the biotransformations are reported in
50 mg scale with the best biocatalyst (i.e., OYE3), obtaining Table 3, compared to those previously obtained with yeast
completely reproducible values: a practically quantitative yield whole cells. Again, the data corroborate the general trends dis-
of (S)-25, with 97% ee. Thus, it is apparent how the use of iso- cussed in the previous sections. A small substituent (Me) next
lated enzymes can help not only to improve the performance to the carbonyl group and a relatively small group (Et) at the
of a whole cells process but also to address its possible repro- α-position, together with a strong electron-withdrawing group
ducibility issues, with minimal effort.
on the aromatic ring in the meta position, contribute to afford
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2988–2996 | 2993

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 3 Chemoenzymatic enantioselective synthesis of (S)-iopanoic acid. i. H2SO4, AcOH; ii. baker’s yeast, glucose, H2O; iii. OYE/GDH, glucose, NADP+, phosphate
buffer pH 7.0; iv. NaH, Et2O; v. m-nitrobenzyl bromide, Et2O; vi. NaOH, H2O–MeOH.
enzymes allows us to design more practical processes, over-
coming those hurdles in many instances.
Table 3 Baker’s yeast and OYE1–3 mediated bioreduction of α,β-unsaturated
ketone 28
Such a strategy has been exemplified in this work with the
bioreduction of a family of α,β-unsaturated α-alkyl-β-aryl-
ketones, yielding in most cases the corresponding α-chiral
saturated ketones in high yield and high optical purity.
The selection of a wide range of structural features enabled
us to assess the stereo-electronic requirements of the reaction.
To maximize the conversion the substituents in the α-position
and adjacent to the carbonyl group must both be small and
the substituents on the ring should be electron-withdrawing
(best in the meta position). To obtain high enantiomeric
excesses it is better to have a reasonably larger group in the
α-position, compatible with the conversion obtainable. In
addition, ortho and meta substituents on the ring more often
lead to a higher enantioselection than para substituents (most
likely for steric reasons).
During the substrate screening, an interesting case of enan-
tiodivergent behaviour of the ene-reductases from Saccharo-
myces species has been shown, allowing us to access both
enantiomers of suitable products simply by employing
different biocatalysts. This enzyme-based stereocontrol, quite
unexpected with OYE1–3 given their high degree of homology,
is driven by the increased bulkiness at the position adjacent to
the carbonyl moiety, and has been explained and rationalized
by computational docking studies.
Biocatalyst
Conv.a (%)
eeb (%)
Baker’s yeastc
OYE1
OYE2
56
>99
73
99
99
99
99
OYE3
88
a Conversion values determined by GC-MS. b Enantiomeric excess
values determined by GC on a chiral stationary phase. c Reported from
ref. 6b.
very high conversion, quantitative with OYE1. As previously
observed in the reduction of enone 10, an ethyl group in the
α-position seems to exert a positive effect on the enantioselec-
tion, since the formation of an almost enantiomerically pure
product was observed with all of the ene-reductases tested.
Scaling-up of the bioreduction catalyzed by OYE1 (which
gave complete conversion also in the analytical-scale screen-
ing) afforded (S)-29 in quantitative yield and in optically pure
form, further proving the practical potential of isolated
enzyme mediated reductions.
The saturated ketone (S)-29 can then be submitted to a
haloform reaction to provide the corresponding carboxylic acid
without any loss of optical purity. The latter is easily converted
into the desired enantiomer of iopanoic acid (S)-27, through
functional group interconversions and electrophilic aromatic
substitution with iodine monochloride.31
These considerations on the stereochemical outcome of the
bioreductions should proficiently guide the design of enones
to be employed as substrates of ene-reductases in future pre-
parative applications. Two examples of such a possibility have
been presented in this work, leading to the key precursors of
two valuable bioactive molecules (the piperidine alkaloid
stenusine and the radiocontrast agent iopanoic acid) in opti-
cally pure form, with great advantages in yield, optical purity
and ease of isolation.
Conclusions
The introduction of biocatalyzed transformations in multi-step
chemical processes provides evident advantages and is becom-
ing a well-established practice. The very high enantioselectivity
of ene-reductases in the generation of chiral synthones by
reduction of suitably substituted CvC bonds has mainly been
exploited in whole cell mediated processes, although this
approach shows significant drawbacks (e.g., presence of side-
Experimental section
General methods
activities, large amounts of biomass to process, low substrate All chemical reagents and solvents were purchased from sup-
loadings and complex work-up procedures). On the other pliers and used without further purification. All the enzymes
hand, the relatively recent access to overexpressed and purified employed were overexpressed in E. coli BL21 (DE3) strains
2994 | Org. Biomol. Chem., 2013, 11, 2988–2996
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
harbouring a specific plasmid prepared according to standard Computational studies
molecular biology techniques: pET30a-OYE1 from the original
The structures of OYE2 and OYE3 were obtained by homology
modelling on the Swiss-model server automated system.34
With respect to OYE1, OYE2 showed a sequence identity of
92% and OYE3 of 80%, so they were modelled on the known
crystal structure of OYE1 as obtained from the PDB database
(1OYB.pdb). This OYE1 complex contained both the oxidized
flavin cofactor (FMN) and the ligand HBA (p-hydroxybenzalde-
hyde), which were maintained in the modelling of OYE2. The
cofactor FMN was further converted to its reduced form
FMNH2 and the resulting complex was then optimized with
the “protein preparation wizard” tool as implemented in
Maestro 9.0 (Schrödinger),35 using the OPLS2001 force field
for the minimization. The receptor grid was generated with
the software Glide 5.5 (Schrödinger)36 around the HBA ligand.
The structures of the enone substrates were previously opti-
mized by an MC/MM conformational search with Spartan
‘08.37 Docking studies were performed with Glide 5.5 (Schrö-
dinger). The ligands were docked flexibly onto the receptor
grid using the extra precision mode. Up to ten poses for each
ligand were analyzed and docking scores were generated using
the standard functions as implemented in the software. The
poses were then viewed using PyMOL.38
plasmid provided by Neil C. Bruce,32 pET30a-OYE2 and
pET30a-OYE3 from S. cerevisiae BY4741 and pKTS-GDH from
B. megaterium DSM509.33
General procedure for OYE1–3 mediated biotransformations
(screening)
The enone (5 μmol) dissolved in DMSO (10 μL) was added to a
phosphate buffer solution (1.0 mL, 50 mM, pH 7.0) containing
glucose (20 μmol), NADP+ (0.1 mM), GDH (4 U) and the
required OYE (40 μg mL−1). The mixture was incubated for
24 h in an orbital shaker (160 rpm, 30 °C). The solution was
extracted with EtOAc (2 × 250 μL), centrifuging after each
extraction (15 000 g, 1.5 min), and the combined organic solu-
tions were dried over anhydrous Na2SO4 for further analysis.
General procedure for OYE1–3 mediated biotransformations
(50 mg scale)
For all the substrates which gave a reasonable conversion, a
similar protocol was followed on a larger scale, employing the
OYE which provided the best conversion and/or ee. The suit-
able enone (250 μmol, ab. 50 mg) dissolved in i-PrOH (1 mL)
was added to a phosphate buffer solution (25 mL, 50 mM, pH
7.0) containing the required OYE (40–120 μg mL−1), GDH
(100 U), glucose (1 mmol, 180 mg) and NADP+ (5 μmol, 3.7 mg). Acknowledgements
The reaction was monitored by GC until complete conversion.
Prof. Claudio Fuganti (Politecnico di Milano) is warmly
The mixture was then extracted with EtOAc (3 × 10 mL) and
processed as described for the small-scale protocol. Bioreduc-
tion of 1 with OYE3 gave (S)-12, 96% isolated yield, 97% ee by
chiral GC, [α]D = +38.6 (c 0.9, CHCl3) [lit.2 98% ee, [α]D = +40.7
(c 0.9, CHCl3)]. Bioreduction of 2 with OYE1 gave (S)-13, 93%
isolated yield, 97% ee by chiral GC, [α]D = +51.6 (c 1.0, CHCl3)
[lit.2 99% ee, [α]D = +57.3 (c 1.1, CHCl3)]. Bioreduction of 3
with OYE3 gave (S)-14, 91% isolated yield, 99% ee by chiral
GC, [α]D = +29.6 (c 1.2, CHCl3) [lit.2 99% ee, [α]D = +29.0 (c 1.1,
CHCl3)]. Bioreduction of 4 with OYE1 gave (S)-15, 98% isolated
yield, 44% ee by chiral GC, [α]D = +17.6 (c 0.9, CHCl3) [lit.2 98%
ee, [α]D = +42.1 (c 1.1, CHCl3)]. Bioreduction of 5 with
OYE1 gave (R)-16, 93% isolated yield, 65% ee by chiral GC,
[α]D = −41.4 (c 1.1, CHCl3) [lit.2 (S) 99% ee, [α]D = +70.9 (c 1.0,
CHCl3)]. Bioreduction of 7 with OYE1 gave (R)-18, 91% isolated
yield, 17% ee by chiral GC, [α]D = −15.8 (c 0.9, CHCl3). Bio-
reduction of 8 with OYE1 gave (R)-19, 92% isolated yield, 63%
ee by chiral GC, [α]D = −38.8 (c 1.2, CHCl3). Bioreduction of 10
with OYE3 gave (S)-21, 94% isolated yield, 99% ee by chiral
GC, [α]D = +35.3 (c 1.0, CHCl3) [lit.2 98% ee, [α]D = +35.8 (c 0.9,
CHCl3)]. Bioreduction of 11 with OYE1 gave (S)-22, 43% iso-
lated yield, 98% ee by chiral GC, [α]D = +31.8 (c 1.0, CHCl3).
Bioreduction of 24 with OYE3 gave (S)-25, 97% isolated yield,
97% ee by chiral HPLC (of ester (S)-26), [α]D = +9.0 (c 0.8,
CHCl3) [lit.26e 92% ee, [α]D = +8.5 (c 1.0, CHCl3)]. Bioreduction
of 28 with OYE1 gave (S)-29, 98% isolated yield, 99% ee by
chiral GC, [α]D = +9.4 (c 1.1, CHCl3) [lit.2 98% ee, [α]D = +9.7
(c 0.9, CHCl3)].
acknowledged for fruitful discussions. Prof. Neil C. Bruce (Uni-
versity of York) is kindly acknowledged for the gift of plasmid
pT7-OYE1. Prof. Sven Panke and Christian Femmer (ETH-Z,
Basel) are kindly acknowledged for the help provided in the
preparation of the plasmids and the overexpressing strains.
References
1 (a) R. G. Riley, R. M. Silverstein and J. C. Moser, Science,
1974, 183, 760–762; (b) K. Mori, S. Masuda and T. Suguro,
Tetrahedron, 1981, 37, 1329–1340; (c) K. Mori and
S. Harashima, Tetrahedron Lett., 1991, 32, 5995–5998;
(d) S. C. Sinha and E. Keinan, J. Am. Chem. Soc., 1995, 117,
3653–3654.
2 S.-M. Lu and C. Bolm, Angew. Chem., Int. Ed., 2008, 47,
8920–8923.
3 (a) H. Zimmerman, Acc. Chem. Res., 1987, 20, 263–268;
(b) M. Naodovic and H. Yamamoto, Chem. Rev., 2008, 108,
3132–3148.
4 A. Job, C. F. Janeck, W. Bettray, R. Peters and D. Enders,
Tetrahedron, 2002, 58, 2253–2329.
5 H. Heeres, A. J. Minnaard, B. Hessen and H. J. Heeres, Org.
Process Res. Dev., 2012, 16, 400–408.
6 (a) Y. Kawai, K. Saitou, K. Hida and A. Ohno, Tetrahedron:
Asymmetry, 1995, 6, 2143–2144; (b) Y. Kawai, K. Saitou,
K. Hida, D. H. Dao and A. Ohno, Bull. Chem. Soc. Jpn.,
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2988–2996 | 2995

View Article Online
Paper
1996, 69, 2633–2638; (c) Y. Kawai, M. Hayashi and
N. Tokitoh, Tetrahedron: Asymmetry, 2001, 12, 3007–3013.
7 (a) T. Sakai, S. Matsumoto, S. Hidaka, N. Imajo, S. Tsuboi
Organic & Biomolecular Chemistry
Catal., 2011, 353, 1169–1173; (c) G. Tasnádi, C. K. Winkler,
D. Clay, N. Sultana, W. M. F. Fabian, M. Hall, K. Ditrich
and K. Faber, Chem.–Eur. J., 2012, 18, 10362–10367.
and M. Utaka, Bull. Chem. Soc. Jpn., 1991, 64, 3473–3475; 22 J. W. M. Bush and D. L. Hu, Annu. Rev. Fluid Mech., 2006,
(b) E. P. Siqueira Filho, J. A. R. Rodrigues and
P. J. S. Moran, Tetrahedron: Asymmetry, 2001, 12, 847–852.
8 M. Zagozda and J. Plenkiewicz, Tetrahedron: Asymmetry,
2006, 17, 1958–1962.
9 C. Fuganti, J. Minut, G. Pedrocchi Fantoni and S. Servi,
J. Mol. Catal. B: Enzym., 1998, 4, 47–52.
38, 339–369.
23 H. Schildknecht, D. Krauß, J. Connert, H. Essenbreis and
N. Orfanides, Angew. Chem., Int. Ed. Engl., 1975, 14, 427.
24 H. Schildknecht, D. Berger, D. Krauß, J. Connert,
J. Gehlhaus and H. Essenbreis, J. Chem. Ecol., 1976, 2,
1–11.
10 M. De Mancilha, R. De Conti, P. J. S. Moran and 25 (a) D. Enders, J. Tiebes, N. De Kimpe, M. Keppens,
J. A. R. Rodrigues, ARKIVOC, 2001, 85–93.
11 (a) E. Brenna, G. Fronza, C. Fuganti, F. G. Gatti and
S. Serra, in Innovation in Food Engineering – New Techniques
C. Stevens, G. Smagghe and O. Betz, J. Org. Chem., 1993,
58, 4881–4884; (b) I. Lusebrink, K. Dettner, A. Mosandl and
K. Seifert, Naturwissenschaften, 2008, 95, 751–755.
and Products, ed. M. L. Passos and C. P. Ribeiro, CRC Press, 26 (a) C. Stevens and N. De Kimpe, J. Org. Chem., 1993, 58,
2010, pp. 81–106; (b) G. Fronza, C. Fuganti, M. Mendozza,
R. S. Rallo, G. Ottolina and D. Joulain, Tetrahedron, 1996,
52, 4041–4052.
132–134; (b) L. Micouin, T. Varea, C. Riche, A. Chiaroni,
J.-C. Quirion and H.-P. Husson, Tetrahedron Lett., 1994, 35,
2529–2532; (c) E. Poupon, N. Kunesch and H.-P. Husson,
Angew. Chem., Int. Ed., 2000, 39, 1493–1495; (d) A. Castro-
C., J. Juárez-P., D. Gnecco, J. L. Terán, A. Galindo, S. Bernès
and R. G. Enríquez, Tetrahedron: Asymmetry, 2005, 16,
949–952; (e) T. Müller, K. Dettner and K. Seifert, Eur. J. Org.
Chem., 2011, 6032–6038.
12 (a) H. Toogood, J. M. Gardiner and N. S. Scrutton, Chem-
CatChem, 2010, 2, 892–914; (b) M. Hall and
A. S. Bommarius, Chem. Rev., 2011, 111, 4088–4110.
13 K. Saito, D. J. Thiele, M. Davio, O. Lockridge and V. Massey,
J. Biol. Chem., 1991, 266, 20720–20724.
14 (a) K. Stott, K. Saito, D. J. Thiele and V. Massey, J. Biol. 27 (a) P. R. Larsen, J. E. Silva and M. M. Kaplan, Endocr. Rev.,
Chem., 1993, 268, 6097–6106; (b) Y. S. Niino,
S. Chakraborty, B. J. Brown and V. Massey, J. Biol. Chem.,
1995, 270, 1983–1995.
1981, 2, 87–102; (b) L. V. Oberkotter and L. K. Racaniello,
Life Sci., 1990, 46, 1–7; (c) K. Renko, C. S. Hoefig, F. Hiller,
L. Schomburg and J. Kohrle, Endocrinology, 2012, 153,
2506–2513.
15 (a) E. Brenna, F. G. Gatti, A. Manfredi, D. Monti and
F. Parmeggiani, Eur. J. Org. Chem., 2011, 4015–4022; 28 (a) W. J. Cooke and L. M. Cooke, J. Pharmacol. Exp. Ther.,
(b) M. Hall, C. Stueckler, B. Hauer, R. Stuermer,
T. Friedrich, M. Breuer, W. Kroutil and K. Faber, Eur. J. Org.
1983, 225, 85–93; (b) W. J. Cooke and L. M. Cooke, Invest.
Radiol., 1983, 18, 285–292.
Chem., 2008, 1511–1516; (c) M. A. Swiderska and 29 D. Pitré and S. Boveri, J. Med. Chem., 1968, 11, 406.
J. D. Stewart, Org. Lett., 2006, 8, 6131–6133; (d) E. Brenna, 30 M. Colombo, M. De Amici, C. De Micheli, D. Pitré,
F. G. Gatti, D. Monti, F. Parmeggiani and A. Sacchetti,
Chem. Commun., 2012, 48, 79–81; (e) A. Müller, B. Hauer
G. Carrea and S. Riva, Tetrahedron: Asymmetry, 1991, 2,
1021–1030.
and B. Rosche, Biotechnol. Bioeng., 2007, 98, 22–29; 31 T. R. Lewis and S. Archer, J. Am. Chem. Soc., 1949, 71,
(f) E. Brenna, F. G. Gatti, A. Manfredi, D. Monti and
F. Parmeggiani, Org. Process Res. Dev., 2012, 16, 262–268.
16 M. A. Swiderska and J. D. Stewart, J. Mol. Catal. B: Enzym.,
2006, 42, 52–54.
17 C. K. Winkler, C. Stueckler, N. J. Mueller, D. Pressnitz and
K. Faber, Eur. J. Org. Chem., 2010, 6354–6358.
3753–3755.
32 R. E. Williams, D. A. Rathborne, N. S. Scrutton and
N. C. Bruce, Appl. Environ. Microbiol., 2004, 70, 3566–3574.
33 M. Bechtold, E. Brenna, C. Femmer, F. G. Gatti, S. Panke,
F. Parmeggiani and A. Sacchetti, Org. Process Res. Dev.,
2012, 16, 269–276.
18 E. Brenna, F. G. Gatti, D. Monti, F. Parmeggiani and 34 (a) L. Bordoli, F. Kiefer, K. Arnold, P. Benkert, J. Battey and
A. Sacchetti, ChemCatChem, 2012, 4, 653–659.
19 C. Stueckler, N. J. Mueller, C. K. Winkler, S. M. Glueck,
K. Gruber, G. Steinkellner and K. Faber, Dalton Trans.,
2010, 39, 8472–8476.
20 G. Oberdorfer, K. Gruber, K. Faber and M. Hall, Synlett,
2012, 1857–1864.
T. Schwede, Nat. Protoc., 2008, 4, 1–13; (b) K. Arnold,
L. Bordoli, J. Kopp and T. Schwede, Bioinformatics, 2006,
22, 195–201; (c) F. Kiefer, K. Arnold, M. Künzli, L. Bordoli
and T. Schwede, Nucleic Acids Res., 2009, 37, D387–D392;
(d) M. C. Peitsch, Bio/Technology, 1995, 13, 658–660.
35 Maestro, version 9.0, Schrödinger, LLC, New York, NY, 2011.
21 (a) M. Hall, C. Stueckler, W. Kroutil, P. Macheroux and 36 Glide, version 5.5, Schrödinger, LLC, New York, NY, 2011.
K. Faber, Angew. Chem., Int. Ed., 2007, 46, 3934–3937; 37 Spartan ‘08, Wavefunction, Inc., Irvine, CA.
(b) C. Stueckler, C. K. Winkler, M. Hall, B. Hauer, 38 The PyMOL Molecular Graphic System, Schrödinger, LLC,
M. Bonnekessel, K. Zangger and K. Faber, Adv. Synth.
2010.
2996 | Org. Biomol. Chem., 2013, 11, 2988–2996
This journal is © The Royal Society of Chemistry 2013