Whole-Cell-Catalyzed Multiple Regio- and Stereoselective Functionalizations in Cascade Reactions Enabled by Directed Evolution

Article abstract of DOI:10.1002/anie.201605990

Biocatalytic cascade reactions using isolated stereoselective enzymes or whole cells in one-pot processes lead to value-added chiral products in a single workup. The concept has been restricted mainly to starting materials and intermediate products that are accepted by the respective wild-type enzymes. In the present study, we exploited directed evolution as a means to create E. coli whole cells for regio- and stereoselective cascade sequences that are not possible using man-made catalysts. The approach is illustrated using P450-BM3 in combination with appropriate alcohol dehydrogenases as catalysts in either two-, three-, or four-step cascade reactions starting from cyclohexane, cyclohexanol, or cyclohexanone, respectively, leading to either (R,R)-, (S,S)-, or meso-cyclohexane-1,2-diol. The one-pot conversion of cyclohexane into (R)- or (S)-2-hydroxycyclohexanone in the absence of ADH is also described.

Full text of DOI:10.1002/anie.201605990

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201605990
German Edition: DOI: 10.1002/ange.201605990
Cascade Reactions
Whole-Cell-Catalyzed Multiple Regio- and Stereoselective
Functionalizations in Cascade Reactions Enabled by Directed
Evolution
Aitao Li, Adriana Ilie, Zhoutong Sun, Richard Lonsdale, Jian-He Xu, and Manfred T. Reetz*
Abstract: Biocatalytic cascade reactions using isolated stereo-
selective enzymes or whole cells in one-pot processes lead to
value-added chiral products in a single workup. The concept
has been restricted mainly to starting materials and intermedi-
ate products that are accepted by the respective wild-type
enzymes. In the present study, we exploited directed evolution
as a means to create E. coli whole cells for regio- and
stereoselective cascade sequences that are not possible using
man-made catalysts. The approach is illustrated using P450-
BM3 in combination with appropriate alcohol dehydrogenases
as catalysts in either two-, three-, or four-step cascade reactions
starting from cyclohexane, cyclohexanol, or cyclohexanone,
respectively, leading to either (R,R)-, (S,S)-, or meso-cyclo-
hexane-1,2-diol. The one-pot conversion of cyclohexane into
(R)- or (S)-2-hydroxycyclohexanone in the absence of ADH is
also described.
oxidative functionalization. Our approach utilizes an inex-
pensive starting material derived from petro-chemical sources
and is therefore different from using renewable feedstocks
such as glucose, which would require metabolic pathway
engineering. Four different whole cells were envisioned, each
leading to one of the three possible stereoisomeric forms of
1,2-cyclohexanediol, namely (R,R)-5, (S,S)-5, (S,R)-5 (meso),
or (R,S)-5 (also meso), in a multi-step sequence (Figure 1). To
construct R,R-selective whole cells, the plan was to use
a single mutant of an appropriate cytochrome P450 mono-
oxygenase (CYP)[5] to regio- and enantioselectively access
acyloin (R)-4 as the primary product followed by reduction
with R-selective alcohol dehydrogenase (ADH) to form
(R,R)-5. The advantages of a single CYP mutant for the
ꢀ
first three C H activating steps include simplicity and
minimal cell stress. Enantiomeric (S,S)-5 as well as the meso
stereoisomer of 5 could be targeted analogously.
R
etrosynthetic analysis based on man-made reagents, cata-
We first focused on cyclohexanone 3 as the starting
substrate and designed whole cells for its conversion into the
final stereoisomeric diols 5. Thereafter, the plan was to extend
the cascade sequence in a stepwise fashion, specifically by
starting from cyclohexanol 2 and finally from cyclohexane 1.
This bottom-up approach was thought to constitute a useful
guide in the construction of the final four-step sequence 1!
2!3!4!5. The self-sufficient P450-BM3 monooxygenase
from Bacillus megaterium[5,6] was chosen as the catalyst. WT
P450-BM3 with its large binding pocket shows very poor
activity towards cyclohexane 1,[6b] probably owing to its small
size. A variant was recently generated by protein engineering
to accept this substrate with formation of 2, which was further
oxidized to 3 with an ADH.[7] Owing to the aforementioned
reasons, this alternative route to 3 is not suitable for our goal,
lysts, and synthons is a routine technique used by organic
chemists when planning the preparation of structurally simple
or complex compounds.[1] Biocatalytic retrosynthesis[2]
exploiting enzyme mixtures, whole cells, or enzymes in
combination with transition-metal catalysts has not reached
a similar state of maturity.[3] The design of such artificial
reaction sequences is different from metabolic engineering as
this approach is independent of naturally occurring pathways.
The former generally involve wild-type (WT) enzymes,[3]
which restricts their scope. Fortunately, advanced methods
in directed evolution provide a means to lift the traditional
limitations of enzyme catalysis, such as insufficient stereo-
and regioselectivity.[4] Herein, we applied directed evolution
to achieve cascade reactions that are not readily amenable to
state-of-the-art synthetic transition-metal catalysis or organo-
catalysis.
We chose cyclohexane 1 as the model starting material,
a compound that traditionally requires harsh conditions for
[*] Dr. A. T. Li, Dr. A. Ilie, Dr. Z. Sun, Dr. R. Lonsdale, Prof. Dr. M. T. Reetz
Max-Planck-Institut fꢀr Kohlenforschung
Kaiser-Wilhelm-Platz 1, 45470 Mꢀlheim an der Ruhr (Germany)
and
Fachbereich Chemie der Philipps-Universitꢁt
Hans-Meerwein-Strasse, 35032 Marburg (Germany)
E-mail: reetz@mpi-muelheim.mpg.de
Prof. Dr. J. H. Xu
State Key Laboratory of Bioreactor Engineering
East China University of Science and Technology
Shanghai 200237 (China)
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/anie.201605990.
Figure 1. Construction of E. coli whole cells for producing either (R,R)-
5, (S,S)-5, or meso-5, respectively.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
in addition to the anticipated problems with the final ADH-
catalyzed steps.
11:89 in favor of (S)-4). Further hits of comparable quality are
shown in the Supporting Information (see Table S3). The
genes of the best variants from library A were then used as
templates for N-I-F based saturation mutagenesis at site B
(ISM scheme A!B), and the analogous procedure based on
pathway B!A was also explored (see Table S4). The best
results, with e.r. = 95:5 (or 5:95) for both R and S selectivity,
originate from ISM scheme A!B (Figure 2C). This ratio
corresponds to DDG# = 7.3 kJmolꢀ1 relative to a 50:50 ratio.
Kinetic characterization of the two best variants and coupling
efficiencies are shown in the Supporting Information,
Table S5. We also performed docking computations to gain
insight into the origin of the regio- and stereoselectivity
(Figure S1).
To extend the cascade sequence from 3 to the stereoiso-
meric cyclohexane-1,2-diols 5 (Figure 1), we considered 2,3-
butanediol dehydrogenases from different sources (Table S6)
because they catalyze the structurally related oxidation of 2,3-
butanediol with acyloin formation or the reverse reaction.[12]
We initially tested two stereocomplementary enzymes,
BDHA and BUDC, as catalysts in the oxidative kinetic
resolution of rac-5. Gratifyingly, they proved to be effective,
leading to selectivity factors of E > 100 in favor of (R)- and
(S)-4, respectively, a sign that genetic optimization is not
necessary. With meso-5 as the substrate, BDHA provided (S)-
4 with e.r. = 3:97 whereas BUDC favored the formation of
(R)-4 (e.r. = 97:3). We also tested l-2,3-butanediol dehydro-
genase (LBDHA), but it showed no appreciable enantiose-
lectivity in the oxidative kinetic resolution of rac-5 (formation
of almost racemic 4). However, when meso-5 was subjected to
oxidation, it proved to be selective for (R)-4 (e.r. = 95:5).
Thus directed evolution was not necessary.
As the gene mutagenesis method for evolving P450-BM3
mutants for the hydroxylation of cyclohexanone 3, we chose
structure-guided saturation mutagenesis at sites lining the
enzyme binding pocket.[4c,8] To minimize screening, we tested
triple code saturation mutagenesis (TCSM) utilizing three
amino acids as the reduced amino acid alphabet (in addition
to WT) at medium-sized randomization sites, also in combi-
nation with iterative saturation mutagenesis (ISM).[4c,9] WT
P450-BM3 does not accept 3, nor do any of the variants
evolved earlier for a-hydroxylation of structurally different
ketones.[10]
Based on the crystal structure of P450-BM3,[6a] at least two
dozen residues lining the binding pocket can be chosen for
saturation mutagenesis as shown by previous studies.[11]
Following these guidelines and exploratory mutagenesis
experiments, which included NNK-based randomization at
selected amino acid positions (see the Supporting Informa-
tion for details), three amino acids were chosen for TCSM,
followed by ISM. Accordingly, we first docked substrate 3
into the crystal structure of P450-BM3.[6a] Figure 2A features
the arrangement in which the substrate is positioned closest to
=
the catalytically active high-spin species heme-Fe O (com-
ꢀ
pound I). None of the C H entities of 3 are close enough for
smooth hydroxylation, which explains the inertness of this
substrate. Eight residues lining the binding pocket were
chosen as potential randomization candidates, which were
grouped into two four-residue sites, A (L75/L181/I263/A264)
and B (V78/A82/A328/T438). We analyzed all exploratory
data (see Tables S1 and S2) and designed a triple code
consisting of asparagine, isoleucine, and phenylalanine (N-I-
F) for combinatorial randomization at sites A and B (Fig-
ure 2A). This entailed the screening of only 552 and 736
transformants, respectively, for 95% library coverage. The
best variants from library A proved to be L181F (e.r. = 80:20
in favor of (R)-4) and L181F/I263N (e.r. = 22:78 in favor of
(S)-4; Figure 2B). Library B was more productive (Fig-
ure 2B), the best variants being V78I/A82I/A328N (e.r. =
89:11 in favor of (R)-4) and V78I/A82F/A328F/T438I (e.r. =
With these results in hand, we first designed whole cells
for cascade reactions starting from 3 with final formation of
cyclohexane-1,2-diols 5. The E. coli host cells were trans-
formed with one plasmid carrying both the P450-BM3 mutant
and alcohol dehydrogenase (see Figure S2A), which resulted
in six different whole cells based on different combinations of
two P450s (P450ATC06 and P450ATD04) and three ADHs
(BDHA, BUDC, and LBDHA). For comparison, E. coli cells
=
Figure 2. A) Binding pocket of P450-BM3 harboring cyclohexanone 3 closest to heme-Fe O, featuring two four-residue randomization sites A
(green) and B (blue). B) Typical P450-BM3 variants from libraries A (green arrows) and B (blue arrows) showing activity as well as regio- and
enantioselectivity in the reaction of 3 into 4. C) Best ISM pathways for evolving R- and S-selective variants. Green arrow: variants originating from
initial library A; purple arrow: improved R-or S-selective variants resulting from ISM at site B.
2
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
with two plasmid systems showed very poor protein expres-
Surprisingly, whereas meso-5 was formed in the case of
sion for P450 (see Figures S2B and S2C). The designed whole
cells with one plasmid system were then tested in the
conversion of 3 into 5. Under these conditions, the cofactor
is provided by the E. coli host cells using glucose as an energy
source. As shown in Table 1, all three stereoisomeric diols 5
were obtained with high enantio- and diastereoselectivity on
an optional basis. Thus the E. coli (P450ATC06 + BDHA)
cells that combine the R-selective P450-BM3 mutant and R,R-
selective BDHA provided (R,R)-5 (Table 1, entry 1) whereas
E. coli (P450ATD04 + BDHA) or E. coli (P450ATC06 +
BUDC) containing the S/R-selective P450-BM3 mutant and
the R,R/S,S-selective ADH favored meso-5, the former being
more selective (Table 1, entries 2 and 3). The exclusive
formation of (S,S)-5 was achieved by E. coli (P450ATD04 +
BUDC) co-expressing the S-selective P450-BM3 mutant and
the S,S-selective BUDC (Table 1, entry 4). (S,S)-5 was also
obtained using the E. coli (P450ATD04 + LBDHA) strain
that harbors the S-selective P450-BM3 mutant and LBDHA
(Table 1, entry 6). This ADH is known to accept (S)-2-
hydroxybutanone with formation of (S,S)-2,3-butanediol.[12c]
The remaining combination of P450ATC06 with LBDHA
failed to show acceptable levels of enantio- and diastereose-
lectivity. In control experiments, it was shown that both
BUDC and LBDHA catalyze the reduction of 3, whereas
BDHA does not. Therefore, a trace amount of 2 was detected
in the whole-cell systems that contained BUDC or LBDHA.
At this point, mixtures of two E. coli cells, one harboring
a P450-BM3 mutant and the other one an ADH, or two cell
lysates, were tested, but all led to inacceptable results (20–
50% conversion; see the Supporting Information for details).
Next, we tested whether the two best P450-BM3 mutants
ATC06 and ATD04 were capable of catalyzing the sequence
2!3!4 (see Table S7), which proved to be the case.
mutant ATC06, it was not detected when 3 was used as the
substrate in the respective control experiment. This indicates
that the ATC06 variant is also able to catalyze the direct
hydroxylation of 2 into meso-5. The best whole cells were then
tested in the three-step cascade sequence 2!3!4!5. Using
the best cell systems, it was possible to obtain all three
stereoisomeric cyclohexane-1,2-diols 5 as the major products
with high enantio- and diastereoselectivity (Table 2). In
contrast, E. coli (P450ATC06 + BDHA) cells gave a mixture
of trans- and meso-5 with reduced selectivity (85:15) owing to
the competing direct oxidation of 2 into meso-5.
Finally, the utility of the two best mutants P450ATC06 and
P450ATD04 was tested in the cascade sequence of 1!2!3!
4. Unfortunately, only low conversions into 3 were observed
(Table S8). This may come as a surprise but we postulated
that these P450-BM3 variants, although highly regio- and
stereoselective, are not active enough to smoothly catalyze all
three reactions 1!2!3!4. Indeed, the E. coli whole cells
(Table 1) failed to function well in the longer cascade
sequence 1!2!3!4!5. We therefore turned our attention
to the mutants generated in the initial library (Table S1), and
discovered that the two enantiocomplementary P450 mutants
A82F and A82F/A328F catalyze the sequence 1!2!3!4
with good selectivity (Table S8). These variants are consid-
erably more active than ATC06 and ATD04 in the hydrox-
ylation of 3, although a little less selective. Consequently, four
additional whole cells were engineered according to the best
combinations. These were then used in the cascade sequences
starting from 1 with final formation of cyclohexane-1,2-diols 5
(Table 3). Good results were obtained in the case of the
enantiomeric diols (R,R)- and (S,S)-5, these being the major
products with enantiomeric ratios ranging between 96:4 and
> 99:1 and diastereoselectivities of 85–93%. When aiming for
Table 1: Conversion of 3 into stereoisomeric diols 5 using E. coli cells co-expressing P450-BM3 mutants and ADHs.
Entry
Catalyst
Conv. [%][a]
trans/meso
e.r. [%][c]
Favored stereoisomer
Product distribution[b] [%]
2
4
5
1
2
3
4
5
6
E. coli (P450ATC06+BDHA)
E. coli (P450ATD04+BDHA)
E. coli (P450ATC06+BUDC)
E. coli (P450ATD04+BUDC)
E. coli (P450ATC06+LBDHA)
E. coli (P450ATD04+LBDHA)
90
98
84
98
91
97
98:2
5:95
8:92
95:5
51:49
96:4
>99:1
(R,R)
meso
meso
(S,S)
N.A.
(S,S)
0
0
17
3
29
8
1
1
1
0
2
1
99
99
82
97
69
91
[d]
–
–
[d]
>99:1
[d]
–
98:2
[a] Conditions: 10 mm substrate, 308C, 200 rpm, 5 h. The conversion was determined by GC analysis and is based on the amount of converted
substrate. For detailed conditions see the Supporting Information. [b] Relative amounts based on the peak areas in the GC chromatograms.
[c] Determined for trans-cyclohexane-1,2-diols 5. [d] Values not determined. N.A.=not available.
Table 2: Conversion of 2 into stereoisomeric diols 5 using E. coli cells co-expressing P450-BM3 mutants and ADHs.
Entry
Catalyst
Conv. [%][a]
trans/meso
e.r. [%][c]
Favored stereoisomer
Product distribution [%][b]
3
4
5
1
2
3
4
5
E. coli (P450ATC06+BDHA)
E. coli (P450ATD04+BDHA)
E. coli (P450ATC06+BUDC)
E. coli (P450ATD04+BUDC)
E. coli (P450ATD04+LBDHA)
96
20
97
99
99
85:15
4:96
19:81
95:5
>99:1
(R,R)
meso
meso
(S,S)
(S,S)
4
35
3
2
2
1
0
1
1
0
95
65
96
97
98
[d]
–
–
[d]
>99:1
97:3
96:4
[a] Conditions: 5 mm substrate, 308C, 200 rpm, 5 h. The conversion was determined by GC analysis and is based on the amount of converted
substrate. For detailed conditions see the Supporting Information. [b] Relative amounts based on the peak areas in the GC chromatograms.
[c] Determined for trans-cyclohexane-1,2-diols 5. [d] Values not determined.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
Table 3: Conversion of 1 into stereoisomeric diols 5 using E. coli cells co-expressing P450-BM3 mutants and ADHs.
Entry
Catalyst
Conv. [%][a]
trans/meso
e.r. [%][c]
Favored stereoisomer
Product distribution [%][b]
2
3
4
5
1
2
3
4
E. coli (P450A82F+BDHA)
72
77
83
87
88:12
20:80
85:15
93:7
>99:1
(R,R)
meso
(S,S)
(S,S)
5
16
7
15
19
23
15
3
0
0
0
77
65
70
82
[d]
E. coli (P450A82F/A328F+BDHA)
E. coli (P450A82F/A328F+BUDC)
E. coli (P450A82F/A328F+LBDHA)
–
>99:1
96:4
3
[a] Conditions: 5 mm substrate, 258C, 200 rpm, 8 h. The conversion was determined by GC analysis and is based on the amount of converted
substrate. For detailed conditions see the Supporting Information. [b] Relative amounts based on the peak areas in the GC chromatograms. [c] Values
determined for trans-cyclohexane-1,2-diols 5. [d] Values not determined.
Dickerhoff, A. Nobili, H. J. Joosten, K. Weisz, M. Hohne, U. T.
Bornscheuer, ChemBioChem 2015, 16, 1041 – 1045; f) S. Riva,
meso-5, a diastereoselectivity of 80% was achieved. Several
reactions were scaled up; the optimal workup procedure
involved the use of n-butanol in the extraction step (see the
Supporting Information for details), which led to yields of
isolated products of 29–66% (Table S15).
W.-D. Fessner, Cascade Biocatalysis, Wiley-VCH, Weinheim,
2014; g) S. K. Wu, Y. Z. Chen, Y. Xu, A. T. Li, Q. S. Xu, A.
Glieder, Z. Li, ACS Catal. 2014, 4, 409 – 420; h) H. Grçger, W.
Hummel, Curr. Opin. Chem. Biol. 2014, 19, 171 – 179; i) R.
Agudo, M. T. Reetz, Chem. Commun. 2013, 49, 10914 – 10916;
j) E. Ricca, B. Brucher, J. H. Schrittwieser, Adv. Synth. Catal.
2011, 353, 2239 – 2262; k) S. K. Wu, Y. Zhou, T. W. Wang, H. P.
Too, D. I. C. Wang, Z. Li, Nat. Commun. 2016, 7, 11917.
[4] a) A. S. Bommarius, Annu. Rev. Chem. Biomol. 2015, 6, 319 –
345; b) C. A. Denard, H. Q. Ren, H. M. Zhao, Curr. Opin.
Chem. Biol. 2015, 25, 55 – 64; c) M. T. Reetz, Angew. Chem. Int.
Ed. 2011, 50, 138 – 174; Angew. Chem. 2011, 123, 144 – 182;
d) E. M. Brustad, F. H. Arnold, Curr. Opin. Chem. Biol. 2011, 15,
201 – 210; e) C. Jꢁckel, D. Hilvert, Curr. Opin. Biotechnol. 2010,
21, 753 – 759; f) N. J. Turner, Nat. Chem. Biol. 2009, 5, 568 – 574;
g) S. Lutz, U. T. Bornscheuer, Protein Engineering Handbook,
Wiley-VCH, Weinheim, 2009.
[5] a) H. M. Girvan, A. W. Munro, Curr. Opin. Chem. Biol. 2016, 31,
136 – 145; b) P. R. Ortiz de Montellano, Chem. Rev. 2010, 110,
932 – 948.
[6] a) D. C. Haines, D. R. Tomchick, M. Machius, J. A. Peterson,
Biochemistry 2001, 40, 13456 – 13465; b) A. Glieder, E. T.
Farinas, F. H. Arnold, Nat. Biotechnol. 2002, 20, 1135 – 1139.
[7] S. Staudt, E. Burda, C. Giese, C. A. Muller, J. Marienhagen, U.
Schwaneberg, W. Hummel, K. Drauz, H. Grçger, Angew. Chem.
Int. Ed. 2013, 52, 2359 – 2363; Angew. Chem. 2013, 125, 2415 –
2419.
In conclusion, we have shown that the use of directed
evolution enables the creation of E. coli whole cells for
synthetically challenging cascade reactions that are not
possible with wild-type enzymes or man-made catalysts. In
terms of method development in directed evolution, it also
shows that recently reported triple code saturation muta-
genesis (TCSM)[9] is effective for mechanistically and struc-
turally more complex enzymes such as P450 monooxygenases.
Starting from the achiral substrates cyclohexane, cyclohex-
anol, or cyclohexanone, cyclohexane-1,2-diols as functional-
ized products were shown to be accessible in three different
stereoisomeric forms. The cascade sequence can also be
terminated at the stage of (R)- or (S)-4, acyloins that are
likewise value-added compounds.[13] Numerous other enzyme
types await consideration when designing bacterial or yeast
whole cells for synthetically useful stereoselective cascade
sequences requiring only a single workup.
Acknowledgements
[8] Z. Sun, Y. Wikmark, J. E. Bꢁckvall, M. T. Reetz, Chem. Eur. J.
2016, 22, 5046 – 5054.
[9] a) Z. Sun, R. Lonsdale, L. Wu, G. Li, A. T. Li, J. Wang, J. Zhou,
M. T. Reetz, ACS Catal. 2016, 6, 1590 – 1597; b) Z. Sun, R.
Lonsdale, A. Ilie, G. Li, J. Zhou, M. T. Reetz, ACS Catal. 2016, 6,
1598 – 1605.
Support from the Max-Planck-Society and the LOEWE
Research cluster SynChemBio is gratefully acknowledged.
ꢀ
Keywords: cascade reactions · C H activation ·
[10] R. Agudo, G. D. Roiban, R. Lonsdale, A. Ilie, M. T. Reetz, J.
Org. Chem. 2015, 80, 950 – 956.
directed evolution · enantioselectivity · P450 monooxygenase
[11] a) S. Kille, F. E. Zilly, J. P. Acevedo, M. T. Reetz, Nat. Chem.
2011, 3, 738 – 743; b) C. J. C. Whitehouse, S. G. Bell, L. L. Wong,
Chem. Soc. Rev. 2012, 41, 1218 – 1260; c) R. Bernhardt, V. B.
Urlacher, Appl. Microbiol. 2014, 98, 6185 – 6203; d) R. Fasan,
ACS Catal. 2012, 2, 647 – 666.
[12] a) Y. J. Yan, C. C. Lee, J. C. Liao, Org. Biomol. Chem. 2009, 7,
3914 – 3917; b) J. D. Zhang, T. T. Xu, Z. Li, Adv. Synth. Catal.
2013, 355, 3147 – 3153; c) S. Ui, Y. Takusagawa, T. Sato, T.
Ohtsuki, A. Mimura, M. Ohkuma, T. Kudo, Lett. Appl. Micro-
biol. 2004, 39, 533 – 537.
[1] a) E. J. Corey, The Logic of Chemical Synthesis, Wiley, New
York, 1995; b) C. Grondal, M. Jeanty, D. Enders, Nat. Chem.
2010, 2, 167 – 178; c) K. C. Nicolaou, J. S. Chen, Chem. Soc. Rev.
2009, 38, 2993 – 3009; d) Catalytic Cascade Reactions (Eds.: P.-F.
Xu, W. Wang), Wiley, New York, 2014.
[2] N. J. Turner, E. OꢀReilly, Nat. Chem. Biol. 2013, 9, 285 – 288.
[3] a) Enzyme Catalysis in Organic Chemistry (Eds.: K. Drauz, H.
Grçger, O. May), Wiley-VCH, Weinheim, 2012; b) P. Both, H.
Busch, P. P. Kelly, F. G. Mutti, N. J. Turner, S. L. Flitsch, Angew.
Chem. Int. Ed. 2016, 55, 1511 – 1513; Angew. Chem. 2016, 128,
1533 – 1536; c) T. Bayer, S. Milker, T. Wiesinger, F. Rudroff,
M. D. Mihovilovic, Adv. Synth. Catal. 2015, 357, 1587 – 1618;
d) F. G. Mutti, T. Knaus, N. S. Scrutton, M. Breuer, N. J. Turner,
Science 2015, 349, 1525 – 1529; e) L. Skalden, C. Peters, J.
[13] P. Hoyos, J. V. Sinisterra, F. Molinari, A. R. Alcantara, P. D.
De Maria, Acc. Chem. Res. 2010, 43, 288 – 299.
Received: June 22, 2016
Revised: July 18, 2016
Published online: && &&, &&&&
4
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Cascade Reactions
Four steps in one pot: E. coli cells har-
boring P450-BM3 mutants and appropri-
ate alcohol dehydrogenases were
obtained by directed evolution. Such cells
enable one-pot cascade reactions of cy-
clohexane with selective formation of all
three stereoisomeric cyclohexane-1,2-
diols.
A. T. Li, A. Ilie, Z. Sun, R. Lonsdale,
J. H. Xu, M. T. Reetz*
&&&& — &&&&
Whole-Cell-Catalyzed Multiple Regio- and
Stereoselective Functionalizations in
Cascade Reactions Enabled by Directed
Evolution
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!