Directed Evolution of Alcohol Dehydrogenase for Improved Stereoselective Redox Transformations of 1-Phenylethane-1,2-diol and Its Corresponding Acyloin

Article abstract of DOI:10.1021/acs.biochem.8b00055

Laboratory evolution of alcohol dehydrogenase produced enzyme variants with improved turnover numbers with a vicinal 1,2-diol and its corresponding hydroxyketone. Crystal structure and transient kinetics analysis aids in rationalizing the new functions of these variants.

Full text of DOI:10.1021/acs.biochem.8b00055

Communication
pubs.acs.org/biochemistry
Cite This: Biochemistry XXXX, XXX, XXX−XXX
Directed Evolution of Alcohol Dehydrogenase for Improved
Stereoselective Redox Transformations of 1‑Phenylethane-1,2-diol
and Its Corresponding Acyloin
Emil Hamnevik, Dirk Maurer, Thilak Reddy Enugala, Thao Chu, Robin Lofgren, Doreen Dobritzsch,*
̈
and Mikael Widersten*
Department of Chemistry-BMC, Uppsala University, Box 576, SE-751 23 Uppsala, Sweden
S
* Supporting Information
Table 1. Steady State Kinetic Parameters
ABSTRACT: Laboratory evolution of alcohol dehydro-
enzyme/substrate
kcat (s−1
)
KM (mM)
kcat/KM (s−1 mM−1
)
genase produced enzyme variants with improved turnover
numbers with a vicinal 1,2-diol and its corresponding
hydroxyketone. Crystal structure and transient kinetics
analysis aids in rationalizing the new functions of these
variants.
a
a
a
a
wild type/(S)-1
wild type/(R)-2
wild type/3
wild type/4
C1/(R)-2
80 20
0.63 0.05
130 30
a
a
a
a
0.73 0.01
17 0.6
0.044 0.0008
a
a
36 0.8
1.2 0.09
30
2
a
a
2.0 0.08
1.9 0.05
3.7 0.4
0.55 0.03
37
36
2
6
0.050 0.001
0.53 0.02
C1/4
19
5.5 0.4
8.8
2
C1B1/(R)-2
C1B1/4
120 20
20
0.050 0.003
0.44 0.04
nzymes can provide effective alternative routes as catalysts
E
in synthetic chemistry,1−4 and the “green-ness” of
biocatalysis can contribute to a more sustainable manufacturing
of chemicals. An important chemical transformation in
synthetic chemistry is the oxidation of secondary alcohols and
the reduction of the corresponding ketones. Partial oxidation of
1,2-substituted (vicinal) diols can produce the corresponding α-
hydroxy ketones (acyloins) that are attractive building blocks
for chiral auxiliaries, natural products, and pharmaceuticals.5−7
Enzyme-catalyzed production of acyloins has been demon-
strated to be feasible using monooxygenases or alcohol
dehydrogenases.8−10
1
4
a
Data from ref 11.
Scheme 1. Kinetic Mechanism for ADH-A-Catalyzed Alcohol
Oxidation
Table 2. Kinetic Rates
a
a
enzyme/substrate
k3 (s−1
)
k5 (s−1
)
kcat (s−1
)
b
c
b
Alcohol dehydrogenase A (ADH-A) from the bacterium
Rhodococcus ruber DSM 44541 is an interesting candidate
biocatalyst of redox reactions. ADH-A is unusually tolerant
toward organic solvents, is highly regio- and enantioselective,
and displays activity with a wide range of alcohols and ketones,
including aryl-substituted vicinal diols.11−14The catalyzed
oxidation of 1,2-diols by the wild-type enzyme is, however,
relatively inefficient; the kcat/KM for oxidation of (R)-2 (Chart
1) into the corresponding hydroxy ketone 4 is approximately
wild type/(S)-1
wild type/(R)-2
C1/(S)-1
630 40
51
51
6
6
80 20
c
b
42
8
0.73 0.01
−
76 20
76 20
93 10
93 10
110
1.9 0.05
120
5.5 0.4
7
C1/(R)-2
25
2
C1B1/(S)-1
C1B1/(R)-2
340 20
58 10
6
a
See the Supporting Information for a description of the estimation of
k3 and k5. Example tracks recorded during the pre-steady state or pre-
equilibrium phases and the model fittings to the observed transient
b
c
rates are shown in Figures S3 and S4. Data from ref 14. Data from
ref 23.
Chart 1. Substrate Alcohols and Ketones
to be independent of the alcohol substrate (or ketone product)
because this step involves only the binary enzyme−nucleotide
complex. ADH-A displays a kcat with (R)-2 of 0.73 s−1, which is
70-fold lower than the NADH release rate (Table 2). Assuming
that oxidation of (R)-2 still obeys an ordered mechanism, the
lower turnover rate is presumably due to nonproductive
substrate binding,15 where the alcohol is bound in the ternary
3000-fold lower than that for the oxidation of (S)-1 into
acetophenone (3) (Table 1). ADH-A follows an ordered
sequential bi-bi mechanism with a low degree of accumulation
of the ternary complex at the steady state.14 The rate-limiting
step for the turnover of preferred substrates such as (S)-1 is
NADH release (k5 in Scheme 1, Table 2). It is noteworthy that
in a strictly ordered mechanism the NADH off rate is expected
Received: January 16, 2018
Revised: January 31, 2018
Published: January 31, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.biochem.8b00055
Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Communication
complex, in a configuration(s) that does not facilitate its
oxidation.16
Table 3. Selection of ADH-A Variants Scored as Top Hits
during Library Screenings
In this work, we set out to improve the catalytic activity of
ADH-A with the aryl-substituted vicinal diol (R)-2. More
efficient enzyme variants can be directly employed in enzyme-
catalyzed reaction pathways such as those described in Scheme
2 in which racemic styrene oxide can be transformed into
a
enzyme
variant
relative activity (variant/wild
type)
substitution
F43H
comments
C1
4.9
3
eight
siblings
C1B1
F43H/Y54L
9.0
6
two siblings
a
Measured with 10 mM (R)-2 and 1.6 mM NAD+ at pH 8.0. Activities
Scheme 2. Enantioconvergent Hydrolysis of Styrene Oxide
were measured in crude bacterial lysates without normalization for
enzyme expression levels (n ≥ 3).
a
into (R)-2 Is Catalyzed by Epoxide Hydrolase20−22
its crystal structure was determined (Figure S1 and Table S2).
The increased activity with (R)-2 displayed by C1 can be
attributed to a 2.6-fold increase in kcat (Table 1). However,
because KM is also increased to the same degree, the overall
catalytic efficiency expressed as kcat/KM is unchanged. C1
displays similar behavior in the reduction of ketone 4, with a
10-fold increase in both turnover rate and KM (Table 1).
The parallel increases in kcat and KM suggest a lower degree
of nonproductive binding.15 Furthermore, neither the rate of
oxidation of (R)-2 into 4 nor the rate of NADH release is
significantly changed by the F43H substitution in C1 (Table 2).
Hence, the positive effect on kcat is likely due to destabilization
of nonproductive alcohol binding, thus allowing for a higher
frequency of formation of productive ternary complexes.
The F43H substitution results in a noteworthy structural
alteration. The imidazole side chain of H43 adopts a
conformation different from that of the F43 benzyl group
and is directed toward the C-2′ hydroxyl of the nicotinamide
nucleotide ribose. In this conformation, the imidazole side
chain flips away from the substrate-binding site, enlarging it. We
have observed the same F43H substitution when searching for
ADH-A variants with improved activities toward other
alcohols.23 Interestingly, this substitution “restores” a hydro-
gen-bonding chain that has been proposed to be part of the
catalytic machinery in the related horse liver ADH24−27 (Figure
2A). The significance of this change in enzyme−cofactor
interactions is unclear but does obviously facilitate a higher rate
of catalytic turnover in the case studied here.
a
(R)-2 can subsequently be oxidized into 4 by ADH-A.
acyloin 4. To achieve adequate flux through the coupled
reaction steps, however, the catalytic efficiency of ADH-A with
(R)-2 requires improvements. To optimize the turnover of
alcohol oxidation in this case presumably requires that a
relaxation of nonproductive binding be accomplished. We
approached this by applying CASTing17−19 involving saturation
mutagenesis of active-site residues predicted to interact with
bound alcohol substrates. The targeted residues were grouped
into three different sites, each comprising two residues: site A,
Y294 and W295; site B, Y54 and L119; site C, F43 and I271
(Figure 1). Gene libraries encoding full unbiased codon sets at
The C1 variant was subsequently employed as a parent
scaffold for additional rounds of saturation mutagenesis. First,
the A site was re-explored, but, again, without success.
Revisiting the B site, however, now resulted in several improved
variants (Table S1). The most active enzyme from the screen,
C1B1 [F43H, Y54L (Table 3)], was purified and further
characterized functionally and structurally. The apparent
increase in the activity of C1B1 observed in the screen with
(R)-2 as a substrate can be attributed to an additional 2.9-fold
increase in kcat by the added Y54L substitution (Table 1 and
Table S3). C1B1 also displays a parallel increase in KM, again
resulting in an unchanged kcat/KM. Hence, without a decrease in
the overall catalytic efficiency, the turnover number is increased
by >7-fold with (R)-2. As described previously, the value of kcat
is much lower than the rate of either (R)-2 oxidation or NADH
release (Table 2). This indicates a fractional increase in the
productively bound ternary enzyme−substrate complex as a
result of the combined F43H and Y54L substitution effects.
Because the positive effect of the Y54L mutation in the C1B1
variant did not penetrate without the prior introduction of the
F43H substitution, the successful isolation of improved
enzymes from this second-generation C1B library suggests
synergistic (epistatic) coupling between these mutations.
Figure 1. Active-site residues subjected to saturation mutagenesis. The
catalytic zinc ion is shown as a gray sphere and the bound coenzyme
(NAD+) as blue sticks. The carbon atoms of the targeted residues are
colored as follows: magenta for the A site, yellow for the B site, and
green for the C site. The image was created with PyMOL 1.8.729 using
the wild-type ADH-A coordinates (Protein Data Bank entry 3jv7).30
both positions were constructed (see the Supporting
Information for experimental details), and the expressed
proteins were challenged for activity with (R)-2. Screening
>1200 enzyme variants of the A or B libraries did not produce
any candidate hits. The C library, however, did present ADH-A
variants with apparently improved oxidation activity with (R)-2
(Table S1). The most active variant identified, dubbed “C1”
(F43H) (Table 3), was further characterized functionally, and
B
DOI: 10.1021/acs.biochem.8b00055
Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Communication
substrate access is normally limited by only solubility, and
possible inhibitory effects on the enzyme catalyst from high
reactant concentrations can be alleviated if the binding affinity
of the reactant is decreased, i.e., as reflected in a higher KM
value. Also, under a high substrate load, the catalytic efficiency
is more accurately measured by kcat than by kcat/KM because
[ES] approaches [E]tot.28
The parallel increases in kcat and KM for (R)-2, while kcat/KM
stays unchanged, indicate a larger number of accessible binding
modes in the active site, increasing the likelihood of turnover at
the cost of affinity. The degree of productive binding increases
through the evolutionary trajectory, from the wild-type enzyme,
via the F43H substitution in C1, and enhanced by the
additional Y54L substitution in C1B1. This enzyme variant also
shows excellent enantioselectivity in the oxidation and
reduction of (R)-2 and 4, respectively.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.bio-
chem.8b00055.
Experimental details; supplementary tables of kinetic
parameters, library screening outcomes, and crystallo-
graphic data and refinement statistics; and supplementary
figures of kinetic data and product ratios (PDF)
Accession Codes
Crystallographic models and data were deposited in the Protein
Data Bank as entries 6ffx for C1 and 6ffz for C1B1.
AUTHOR INFORMATION
■
Figure 2. (A) Hydrogen bonding network involving the inserted H43
in the C1 variant (cyan). The side chain of F43 in wild-type ADH-A
(pink) is oriented in a different direction. NAD+ and the Zn ligands
are shown as sticks with blue and white carbon atoms, respectively. (B)
The amino acid residue substitution in C1B1 frees the space occupied
by the side chains of F43 and Y54 in the wild-type enzyme (pink),
resulting in an expansion of the active site. The surface representation
shows the dimensions of the active-site cavity in C1B1 (green).
Corresponding Authors
*E-mail: doreen.dobritzsch@kemi.uu.se.
*E-mail: mikael.widersten@kemi.uu.se.
ORCID
Thilak Reddy Enugala: 0000-0001-5915-1514
Mikael Widersten: 0000-0002-3203-3793
Funding
The work was supported by Stiftelsen Olle Engkvist
As in C1, the side chain of H43 interacts with the 2′-hydroxyl
of the bound coenzyme, which enlarges the active site to the
same degree (Figure 2B). The inserted L54 is located
perpendicular to H43, and the Y54L substitution further
increases the active-site volume. The larger binding pocket may
allow for conformational sampling of the substrate (alcohol or
ketone) that increases the frequency of productive binding
modes, observed as increases in kcat. C1B1 displays high
enantioselectivity toward (R)-2 over (S)-2, with almost
undetectable activity with (S)-2. The stereoselectivity is also
reflected in the reduction of 4, producing exclusively (R)-2 as a
product (Figure S2). Hence, the improved catalytic trans-
formation of 4 also affords a more efficient production of
enantiopure (R)-2.
In conclusion, we have shown that it is possible to increase
the rate of turnover of ADH-A with the aryl-substituted vicinal
diol (R)-1-phenylethane-1,2-diol, and the acyloin 2-hydrox-
yacetophenone. However, the overall catalytic efficiencies with
either of these substrates, if expressed as kcat/KM, did not
increase due to parallel increases in KM. From an applied
perspective, using the C1B1 enzyme as a biocatalyst for this
transfomation, a higher value of kcat is favored over a low KM;
Byggmastare (183-358).
̈
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The support from COST action 1303 Systems Biocatalysis is
also gratefully acknowledged. The authors thank the Diamond
Light Source for beam time (proposal mx11171) and the staff
of beamline I04 for assistance with crystal testing and data
collection. Access was supported in part by EU FP7
infrastructure grant BIOSTRUCTX (Contract 283570).
REFERENCES
■
(1) Schmid, A., Dordick, J. S., Hauer, B., Kiener, A., Wubbolts, M.,
and Witholt, B. (2001) Industrial biocatalysis today and tomorrow.
Nature 409, 258−268.
(2) Schoemaker, H. E., Mink, D., and Wubbolts, M. G. (2003)
Dispelling the myths - biocatalysis in industrial synthesis. Science 299,
1694−1697.
(3) Anastas, P., and Eghbali, N. (2010) Green Chemistry: Principles
and Practice. Chem. Soc. Rev. 39, 301−312.
C
DOI: 10.1021/acs.biochem.8b00055
Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Communication
(4) Clouthier, C. M., and Pelletier, J. N. (2012) Expanding the
organic toolbox: a guide to integrating biocatalysis in synthesis. Chem.
Soc. Rev. 41, 1585−1605.
productive binding and increased rate of coenzyme release in an
alcohol dehydrogenase increases turnover with a nonpreferred alcohol
enantiomer. FEBS J. 284, 3895−3914.
(24) Eklund, H., Nordstrom, B., Zeppezauer, E., Soderlund, G.,
(5) Bhowmick, K. C., and Joshi, N. N. (2006) Syntheses and
applications of C2-symmetric chiral diols. Tetrahedron: Asymmetry 17,
1901−1929.
̈
̈
Ohlsson, I., Boiwe, T., and Branden
́
, C.-I. (1974) The structure of
̈
horse liver alcohol dehydrogenase. FEBS Lett. 44, 200−204.
(25) Kvassman, J., and Pettersson, G. (1980) Unified mechanism for
proton-transfer reactions affecting the catalytic activity of liver alcohol
dehydrogenase. Eur. J. Biochem. 103, 565−575.
́
(6) Hoyos, P., Sinisterra, J. V., Molinari, F., Alcantara, A. R., and
Domínguez de María, P. (2010) Biocatalytic strategies for the
asymmetric synthesis of α-hydroxy ketones. Acc. Chem. Res. 43,
288−299.
(26) LeBrun, L. A., and Plapp, B. V. (1999) Control of coenzyme
binding to horse liver alcohol dehydrogenase. Biochemistry 38, 12387−
12393.
(7) Palomo, C., Oiarbide, M., and García, J. M. (2012) α-Hydroxy
ketones as useful templates in asymmetric reactions. Chem. Soc. Rev.
41, 4150−4164.
(27) Plapp, B. V. (2010) Conformational changes and catalysis by
alcohol dehydrogenase. Arch. Biochem. Biophys. 493, 3−12.
(28) Eisenthal, R., Danson, M. J., and Hough, D. W. (2007) Catalytic
efficiency and kcat/KM: a useful comparator? Trends Biotechnol. 25,
247−249.
(8) Zhang, J., Xu, T., and Li, Z. (2013) Enantioselective biooxidation
of racemic trans-cyclic vicinal diols: one-pot synthesis of both
enantiopure (S,S)-cyclic vicinal diols and (R)-α-hydroxy ketones.
Adv. Synth. Catal. 355, 3147−3153.
(9) Agudo, R., Roiban, G.-D., Lonsdale, R., Ilie, A., and Reetz, M. T.
(2015) Biocatalytic route to chiral acyloins: P450-catalyzed regio- and
enantioselective α-hydroxylation of ketones. J. Org. Chem. 80, 950−
956.
(10) Zhang, J., Wu, S., Wu, J., and Li, Z. (2015) Enantioselective
cascade biocatalysis via epoxide hydrolysis and alcohol oxidation: one-
pot synthesis of (R)-α-hydroxy ketones from meso- or racemic
epoxides. ACS Catal. 5, 51−58.
(11) Stampfer, W., Kosjek, B., Moitzi, C., Kroutil, W., and Faber, K.
(2002) Biocatalytic asymmetric hydrogen transfer. Angew. Chem., Int.
Ed. 41, 1014−1017.
(12) Stampfer, W., Kosjek, B., Faber, K., and Kroutil, W. (2003)
Biocatalytic asymmetric hydrogen transfer employing Rhodococcus
ruber DSM 44541. J. Org. Chem. 68, 402−406.
(29) The PyMOL Molecular Graphics System, version 1.8,
Schrodinger, LLC, Portland, OR.
̈
(30) Karabec, M., Łyskowski, A., Tauber, K. C., Steinkellner, G.,
Kroutil, W., Grogan, G., and Gruber, K. (2010) Structural insights into
substrate specificity and solvent tolerance in alcohol dehydrogenase
ADH-‘A’ from Rhodococcus ruber DSM 44541. Chem. Commun. 46,
6314−6316.
(13) Kosjek, B., Stampfer, Pogorevc, W. M., Goessler, W., Faber, K.,
and Kroutil, W. (2004) Purification and characterization of a
chemotolerant alcohol dehydrogenase applicable to coupled redox
reactions. Biotechnol. Bioeng. 86, 55−62.
(14) Hamnevik, E., Blikstad, C., Norrehed, S., and Widersten, M.
(2014) Kinetic characterization of Rhodococcus ruber DSM 44541
alcohol dehydrogenase A. J. Mol. Catal. B: Enzym. 99, 68−78.
(15) Fersht, A. (1999) in Structure and Mechanism in Protein Science,
pp 114−116, W. H. Freeman, New York.
(16) Bar-Even, A., Milo, Noor, R. E., and Tawfik, D. S. (2015) The
moderately efficient enzyme: futile encounters and enzyme floppiness.
Biochemistry 54, 4969−4977.
(17) Reetz, M. T., Bocola, M., Carballeira, J. D., Zha, D., and Vogel,
A. (2005) Expanding the range of substrate acceptance of enzymes:
combinatorial active-site saturation test. Angew. Chem., Int. Ed. 44,
4192−4196.
(18) Reetz, M. T., and Carballeira, J. D. (2007) Iterative saturation
mutagenesis (ISM) for rapid directed evolution of functional enzymes.
Nat. Protoc. 2, 891−903.
(19) Reetz, M. T., Kahakeaw, D., and Sanchis, J. (2009) Shedding
light on the efficacy of laboratory evolution based on iterative
saturation. Mol. BioSyst. 5, 115−122.
(20) Monterde, M. I., Lombard, M., Archelas, A., Cronin, A., Arand,
M., and Furstoss, R. (2004) Enzymatic transformations. Part 58:
enantioconvergent biohydrolysis of styrene oxide derivatives catalysed
by the Solanum tuberosum epoxide hydrolase. Tetrahedron: Asymmetry
15, 2801−2805.
(21) Cao, L., Lee, J., Chen, W., and Wood, T. K. (2006)
Enantioconvergent production of (R)-1-phenyl-1,2-ethanediol from
styrene oxide by combining the Solanum tuberosum and an evolved
Agrobacterium radiobacter AD1 epoxide hydrolases. Biotechnol. Bioeng.
94, 522−529.
(22) Carlsson, Å. J., Bauer, P., Ma, H., and Widersten, M. (2012)
Obtaining optical purity for product diols in enzyme-catalyzed epoxide
hydrolysis: contributions from changes in both enantio- and
regioselectivity. Biochemistry 51, 7627−7637.
(23) Hamnevik, E., Enugala, T. R., Maurer, D., Ntuku, S., Oliveira, A.,
Dobritzsch, D., and Widersten, M. (2017) Relaxation of non-
D
DOI: 10.1021/acs.biochem.8b00055
Biochemistry XXXX, XXX, XXX−XXX