Improved biocatalysts from a synthetic circular permutation library of the flavin-dependent oxidoreductase old yellow enzyme

Article abstract of DOI:10.1021/ja4074886

Members of the old yellow enzyme (OYE) family are widely used, effective biocatalysts for the stereoselective trans-hydrogenation of activated alkenes. To further expand their substrate scope and improve catalytic performance, we have applied a protein en

Full text of DOI:10.1021/ja4074886

Subscriber access provided by University | of Minnesota Libraries
Article
Improved biocatalysts from a synthetic circular permutation
library of the flavin-dependent oxidoreductase Old Yellow Enzyme.
Ashley B. Daugherty, Sridhar Govindarajan, and Stefan Lutz
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja4074886 • Publication Date (Web): 29 Aug 2013
Downloaded from http://pubs.acs.org on September 10, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Improved biocatalysts from a synthetic circular permutation library of the
flavin-dependent oxidoreductase Old Yellow Enzyme.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ashley B. Daugherty§, Sridhar Govindarajan¶, Stefan Lutz§*.
§ Department of Chemistry, Emory University, 1515 Dickey Drive, Atlanta, GA 30084, USA; ¶ DNA2.0,
1140A O’Brian Drive, Menlo Park, CA 94025, USA.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 36
1
2
3
4
ABSTRACT
5
6
7
8
9
Members of the the Old Yellow Enzyme (OYE) family are widelyꢀused, effective biocatalysts for the
stereoselective transꢀhydrogenation of activated alkenes. To further expand their substrate scope and
improve catalytic performance, we have applied a protein engineering strategy called circular permutation
(CP) to enhance the function of OYE1 from Saccharomyces pastorianus. CP can influence a biocatalyst’s
function by altering protein backbone flexibility and active site accessibility, both critical performance
features as the catalytic cycle for OYE1 is thought to involve rateꢀlimiting conformational changes. To
explore the impact of CP throughout the OYE1 protein sequence, we implemented a highly efficient
approach for cellꢀfree cpOYE library preparation by combining wholeꢀgene synthesis with in vitro
transcription/translation. The versatility of such an ex vivo system was further demonstrated by the rapid
and reliable functional evaluation of library members under variable environmental conditions with three
reference substrates ketoisophorone, cinnamaldehyde and (S)ꢀcarvone. Library analysis identified over 70
functional OYE1 variants with several biocatalysts exhibiting over an order of magnitude improved
catalytic activity. Although catalytic gains of individual cpOYE library members vary by substrate, the
locations of new protein termini in functional variants for all tested substates fall within the same four
distinct loop/lid regions near the active site. Our findings demonstrate the importance of these structural
elements in enzyme function and support the hypothesis of conformational flexibility as a limiting factor
for catalysis in wild type OYE.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 3 of 36
Journal of the American Chemical Society
1
2
3
4
INTRODUCTION
5
6
7
8
9
The reduction of alkenes is a widely used and effective strategy for the asymmetric synthesis of
chiral building blocks. While traditional catalysts for these reactions include chiral transition metal
complexes1ꢀ3 and organocatalysts4, enzymes offer a highly selective, evolvable and sustainable
alternative.5ꢀ7 More specifically, eneꢀreductases of the Old Yellow Enzyme (OYE) family [EC 1.3.1.31]
are known to catalyze the highly stereoselective transꢀhydrogenation of α,βꢀunsaturated aldehydes,
ketones, carboxylates, nitriles and nitroalkenes.8ꢀ14 The redox chemistry of OYEs is facilitated by a nonꢀ
covalently bound flavin mononucleotide (FMN) cofactor. Following reduction of the alkene substrate, the
FMN is regenerated via hydride transfer from NADPH (Scheme 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The popularity of OYEs in biotechnological and pharmaceutical applications has resulted in the
isolation and characterization of family members from a variety of organisms, providing a rich source of
oxidoreductases for the conversion of alkenes.10,12,13,15ꢀ19 At the same time, these biocatalysts represent
attractive targets for protein engineers to further improve and customize their functional performances. A
number of structureꢀguided siteꢀdirected and siteꢀsaturation mutagenesis studies, as well as directed
evolution experiments have yielded OYE variants with improved catalytic properties including enhanced
turnover rates, reversed enantioselectivity and increased stability.20ꢀ25 Unfortunately, the reported gains in
catalytic activity were generally modest (two to fourꢀfold) although potential benefits of amino acid
substitution in OYE family members are possible as demonstrated by the complete reversal of
enantioselectivity as a result of a single amino acid change.23 A review of past OYE engineering studies
also identified a major experimental challenge. Following standard heterologous expression of OYEs in
Escherichia coli (E. coli), the functional characterization of the exogenous oxidoreductase is complicated
by the host’s endogenous reductase activity, hence requiring additional purification steps prior to
functional evaluation.20 Consequently, library analysis has typically been limited to either small numbers
of variants or to substrates with low background from host reductases.22,23 Future efforts involving larger,
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 36
1
2
3
4
5
6
more comprehensive protein libraries for tailoring OYE family members would clearly benefit from a
more effective library preparation and analysis protocol.
7
8
9
Our OYE engineering efforts were motivated by a number of biochemical studies and
crystallographic structure analyses that suggested substantial conformational changes as part of the
enzyme’s catalytic cycle. Spectroscopic evidence for such conformational changes in kinetic
measurements and structural differences in crystals of apoꢀform and substrateꢀbound OYEs, as well as
more recent data from an engineering study of Zymomonas mobilis NCR enoate reductase all support the
hypothesis that structural rearrangements of loops and domains play an important and possibly rateꢀ
limiting role in the catalytic function of these enzymes.25ꢀ31 We therefore decided to test the functional
role of various flexible regions in OYE1 from Saccharomyces pastorianus (formerly S. carlsbergensis),
the most extensively studied member of the OYE family, by a protein engineering strategy called circular
permutation (CP). During CP, a protein’s original amino and carboxylꢀtermini are covalently linked by a
peptide linker and new termini are introduced elsewhere in the protein structure through the breakage of a
peptide bond.32 While such termini relocation leaves the amino acid composition of the protein
unchanged, the sequence reorganization has been shown to affect a protein’s local conformational
flexibility. For new termini positioned near an enzyme active site, the altered protein dynamics can
significantly impact catalytic performance by translating into greater active site accessibility and
modifying rateꢀdetermining structural changes.33ꢀ38
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Here, we report on the identification and characterization of catalytically improved OYE1
variants via CP. To address the aforementioned practical limitations for working with combinatorial
libraries of OYE, we have opted for a wholeꢀgene synthesis approach which gives access to an idealized
gene library with maximal (theoretical) diversity at minimal size. Subsequent synthesis of the
corresponding cpOYE variants is accomplished via in vitro transcription/translation (IVTT) using the
chemically defined PURE system which almost completely eliminates background reductase activity.39,40
This ex vivo protein engineering strategy has proved highly effective for the parallel synthesis of hundreds
of OYE library members whose catalytic activity can be assessed without further need for purification.
ACS Paragon Plus Environment

Page 5 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
The evaluation of our new protocol for OYE1 library screening on three reference substrates is described.
The approach has helped to identify a number of candidates with significantly (>10ꢀfold) enhanced
catalytic performance. Evidence in support of the critical role of loop and domains near the active site are
found in the systematic analysis of protein variants with multiple substrates and detailed characterization
of selected cpOYE variants.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MATERIAL AND METHODS
Materials. Wholeꢀgene synthesis was performed by DNA2.0 (Menlo Park, CA). PURExpress kits were
purchased from New England Biolabs (Ipswich, MA). All reagents, substrates, and reference materials
were obtained from SigmaꢀAldrich (St. Louis, MO) unless otherwise indicated.
cpOYE1 library synthesis. The gene encoding OYE1 from S. pastorianus (formerly S. carlsbergensis;
NCBI access number: X53597.1) served as starting sequence for our experiments. The CP library was
prepared by PCR amplification of gene sequences encoding the individual variants using an oye1 tandem
repeat as template (Fig. 1). The tandem repeat was created by wholeꢀgene synthesis, linking two copies of
oye1 headꢀtoꢀtail via an oligonucleotide sequence encoding a three amino acid linker (ꢀGlyꢀSerꢀThrꢀ). The
Start and Stop codons of the 399ꢀamino acid enzyme were eliminated in the tandem repeat. Subsequently,
gene sequences corresponding to individual circular permuted enzymes were obtained by PCR
amplification using the tandem repeat as template in combination with sequenceꢀspecific oligonucleotide
primers. In the 5’ꢀregions of the forward primers, a ATG start codon was included while reverse primers
carried a TAA termination codon. The PCR products were subcloned into pETꢀ14b (Novagen) via
flanking typeꢀIIs restriction endonuclease cleavage sites (BsaI) to enable ligation of PCR products into the
vector’s NcoI/XhoI sites independent of variations in the oye1 library sequences. Following
transformation of individual ligation reactions into E. coli, plasmid DNA from selected colonies was
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 36
1
2
3
4
5
6
extracted and gene sequences were verified by DNA sequencing. Aliquots of confirmed plasmids carrying
specific library members were stored separately in 96ꢀwell microtiter plates.
7
8
9
Primary library screening. All members of the cpOYE library were initially evaluated for catalytic
function by in vitro transcription/translation (IVTT) followed by an in situ eneꢀreductase activity assay.
Overall, the IVTT reactions were assembled using the PURExpress in vitro protein synthesis kit (PURE)
following manufactures protocol with a few adjustments (see below) optimized for our application.
Reactions were assembled in a 96ꢀwell microtiter plate on a 10ꢀꢁL scale containing PURExpress solution
A, PURExpress solution B, 100 ꢁM FMN, 10 units RNase Inhibitor (NEB), 100 ng DNA template, and
nucleaseꢀfree H2O. All reactions from the cpOYE library screening were ran simultaneously with wild
type OYE1 (positive control) and dihydrofolate reductase (DHFR; negative control; provided with the
PURExpress kit) under the same experimental conditions. The reactions were incubated at 37˚C for 2.5 h
to allow for adequate protein synthesis prior to the activity assay. The reactions were cooled to 4˚C to
stop the reaction. Subsequently, eneꢀreductase activity was assessed under anaerobic conditions (Coy
Laboratory Products, Grass Lake, MI) at ambient temperature utilizing glucose dehydrogenase (GDH)
from Thermoplasma acidophilum for NADPH regeneration. The IVTT reaction was mixed with substrate
(200 ꢁM ketoisophorone (1), 200 ꢁM cinnamaldehyde (3), or 1 mM Sꢀcarvone (5)), 200 ꢁM NADP+, 2 U
GDH, and 100 mM glucose. The 30ꢀꢁL assay ran for 2.5 h (due to high turnover, reaction time for 3 was
reduced to 30 min) and was quenched by mixing with an equal amount of ethyl acetate containing 1 mM
cyclohexanone as internal standard. A sample of the organic phase was injected on an Agilent
technologies 6850 GC machine equipped with a chiral CycloSilꢀB column (30 m x 0,32 mm / 0.25 ꢁm,
Agilent, Santa Clara, CA) using hydrogen as the carrier gas (flow rate 1.8 mL/min) and an FID detector
(detector temperature 200°C, split ratio 25:1). The temperature program for 1 ꢀ 4: 150°C hold 10 min.
(Retention time: 1 = 3.54 min; 2 = 3.86 min (Rꢀisomer); 3.95 min (Sꢀisomer), 3 = 6.1 min, 4 = 3.64 min.
For 5 and 6: 90°C hold 5 min, then 1°C/min to 120°C (Retention times: 5 = 27.54 min; 6 = 23.1 min
(1R/4Sꢀisomer); 22.92 min (1S/4Sꢀisomer). The percent conversions and enantiomeric excess was
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 7 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
calculated from substrate and product integration areas and were quantified using standard curves
generated using known amounts of the substrate and product.
7
8
9
Protein expression and purification. Selected members of the cpOYE library were chosen for a more
detailed in vitro characterization in order to confirm and further characterize interesting variants from the
primary library screening analysis. Individual plasmid DNA encoding the corresponding library members
was initially transformed in E. coli BL21(DE3)pLysS for expression. Colonies were cultured in 2 mL LB
medium containing chloramphenicol (34 ꢁg/ml) and ampicillin (100 ꢁg/ml) overnight at 37˚C. The
overnight culture was used to inoculate 2YT medium containing the same antibiotics and cultures were
grown at 37˚C until the OD(600) reached 0.5 to 0.7. Overexpression was induced by addition of IPTG to
a final concentration of 0.4 mM and cultures were overexpressed for 18 h at 20˚C. Subsequently, cultures
were centrifuged for 20 min at 4˚C and 4000g and cell pellets were stored at ꢀ20˚C until purification. The
purification of wild type OYE1 and variants followed the same purification procedure. Cell pellets from a
250 mL culture was resuspended in 6 mL of buffer A (40 mM TrisꢀHCl (pH 8.0), 20 mM NaCl). To the
mixture, 75 ꢁL of protease inhibitor cocktail (Sigma), 7.5 ꢁL of benzonase (Novagen), and 10 ꢁM PMSF
were added and stored on ice for 30 min. Cells were lyzed using sonication (8x with 10 sec pulses and 20
sec pauses). After centrifugation for 30 min at 4˚C and 10,000g, the clear lysate was further purified via
anionꢀexchange chromatography (HiTrap Q HPꢀ5mL column preꢀequilibrated with buffer A). The column
was washed with 2 column volumes (CV) of buffer A, followed by a linear gradient to 100% buffer B (40
mM TrisꢀHCl (pH 8.0), 1 M NaCl) over 10 CV. Product fractions were combined and concentrated to ~1
mL using a Millipore filter unit (MWCO: 10 kDa). In a final polishing step, proteins were purified by size
exclusion chromatography (Superdex 200, 10/300 GL column equilibrated with buffer C (40 mM Trisꢀ
HCl (pH 8.0), 300 mM NaCl); flow rate: 0.5 mL/min). Elution of protein was monitored by UVꢀdetection
at 280 nm and 460 nm and product fractions were combined. SDSꢀPAGE analysis of the final product
showed >95% purity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 36
1
2
3
4
5
6
7
8
9
Spectral properties of OYE1 variants. All spectral analysis of OYE1 and permutants was completed
using a Varian Cary 100 spectrophotometer. The extinction coefficient for each variant was determined
by recording the absorbance spectrum of the proteinꢀbound FMN solution in 10 mM TrisꢀHCl (pH 7.5)
buffer. A 25 ꢁL aliquot of 10% SDS was added to denature the protein and spectra were recorded until
changes were no longer observed. The extinction coefficient for each enzymeꢀbound FMN was calculated
using the following equation: ε = εfree(FMN) (12,500 Mꢀ1cmꢀ1) x (absorbance (at λmax) of FMNbound (prior to
SDS treatment) / absorbance (at λmax) of FMNbound (after SDS)).41
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Phenolate inhibitor binding assays were completed at room temperature in 100 mM potassium
phosphate buffer pH 7 using a cuvette with a 1ꢀcm path length (Fig. S1) The initial enzyme concentration
(10ꢀ30 ꢁM) in each experiment was calculated by the corresponding FMN extinction coefficient. Aliquots
of pꢀhydroxybenzaldehyde were added to the protein solution and spectra were recorded after 3 min of
equilibration until apparent saturation was reached. The long wavelength charge transfer maximum values
for each data set were determined by first derivative analysis and extinction coefficients were calculated
based on the absorbance values (Table 2). Separately, dissociation constants were calculated by plotting
[ligand]free versus Y (fractional saturation of OYE) and analysis by nonꢀlinear curve fit using GraphPad
Prism6 (GraphPad, La Jolla, CA) (Fig.S1 inserts).
Activity assays. The enzyme concentration for each cpOYE variant was determined based on the
extinction coefficient of proteinꢀbound FMN. Purified enzyme was assayed in an anaerobic chamber (Coy
LabProducts) by incubating 0.1ꢀ1 ꢁM enzyme with 1 mM 1, 200 ꢁM NADP+, 5 U GDH, 100 mM
glucose in 500ꢀꢁL total reaction volume. The reaction was monitored over 45 min (timed to achieve less
than 30% product formation) by taking aliquots and mixing with an equal amount of ethyl acetate and
analyzed via GC (see above). The product was quantified using a standard curve generated using known
amounts of the product (R)ꢀlevodione (2) and 1 mM cyclohexanone as internal standard. Alternatively,
turnover was monitored spectrophotometrically at ambient temperatures with the Bioteck EPOCH plate
reader housed in the anaerobic chamber. Purified enzyme was assayed following the decrease in
ACS Paragon Plus Environment

Page 9 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
absorbance of NADPH at 340 nm (molar extinction coefficient 6,220 Mꢀ1cmꢀ1). The 300ꢀꢁL reaction
contained 100 mM sodium phosphate buffer (pH 7.0), 1 mM of the appropriate substrate (from 20 mM
stock in ethanol), 100 ꢁM NADPH and 0.1ꢀ1 ꢁM enzyme. For 3, where substrate saturation was not a
problem, kinetic constants were determined by fitting initial rates to the MichaelisꢀMenten equation
through a nonꢀlinear curve fit using Origin7 (OriginLab, Northampton, MA) (Table S1). The
stereoselectivity displayed by the permutants for reducing (S)-5 was determined by incubating 25 ꢁg
enzyme with 5 mM substrate, 200 ꢁM NADP+, 5 U GDH, and 100 mM glucose in 500ꢀꢁL total reaction
volume. The reaction was carried out for 24 h at room temperature. After quenching, the reaction was
extracted with an equal amount of ethyl acetate and the organic layer analyzed by chiral GC.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Stopped flow experiments. Measurements were made on an OLIS RSMꢀ1000 (Olis Inc., Bogart, GA)
stopped flow spectrophotometer under anaerobic conditions in 50 mM TrisꢀHCl (pH 7.5) at 25 °C. For
the reductive half reaction, OYE1 or cpOYE303 (10 ꢁM) were mixed in the spectrophotometer with
NADPH (final concentration: 50 ꢀ 500 ꢁM). The oxidative half reaction was studied with 1 as substrate.
Enzyme (13 ꢁM) was preincubated with 20 ꢁM NADPH to completely reduce FMN, followed by mixing
with 1 (0.25 mM and 0.5 mM for OYE1; 0.5 mM and 1 mM for cpOYE303). Spectra were recorded with
delays of 10 sec and 100 sec after mixing. All data analysis was performed in GraphPad Prism 6
(GraphPad, La Jolla, CA).
RESULTS AND DISCUSSION
The structurally and functionally wellꢀcharacterized OYE1 from S. pastorianus was chosen to test
our hypothesis that CP of an OYE family member could yield variants with improved catalytic
performance. Initially, we analyzed representative structures of OYE1 (PDB access code: 1OYA30 and
1K0326) to determine a suitable linker sequence for connecting the native amino and carboxyl termini.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 36
1
2
3
4
5
6
7
8
Trimming the Cꢀterminus by three residues shortened the termini distance from 13Å to 6Å without
impacting the catalytic performance of the wild type enzyme. The remaining (shorter) gap was bridged by
a flexible, hydrophilic threeꢀamino acid residue linker (ꢀGlyꢀThrꢀSerꢀ).
9
Taking advantage of wholeꢀgene synthesis, we abandoned the previously employed random
circular permutation protocol 33,42 to instead prepare a fully synthetic DNA library of circular permuted
OYE1 (cpOYE) variants. A tandemꢀOYE1 gene sequence was prepared by chemical DNA synthesis that
served as template for PCR amplification of individual cpOYE variants (Fig. 1). This strategy
dramatically reduces the size of the circular permutation library by eliminating outꢀofꢀframe and inversely
cloned representatives in the sequence pool. Furthermore, parallel cloning of individual genes into an
appropriate DNA vector for protein expression, followed by verification of the correct DNA sequence for
each library member creates a chemically defined collection of enzyme variants. Such upꢀfront efforts
during library generation is advantageous as it dramatically simplifies library analysis by eliminating the
need for oversampling. For practical reasons, we created a focused cpOYE library by synthesizing only
every other possible variant, starting with evenꢀnumbered positions at the Nꢀterminus of OYE1. In
addition, variants within ± 6 amino acids to the native termini were left out as they are predicted to at best
show wild typeꢀlike activity. Following the initial functional evaluation of this primary library, oddꢀ
numbered variants in regions of interest were quickly prepared by the same PCRꢀbased method. Over the
course of this entire study, we synthesized and screened a total of 228 cpOYE variants, identifying
roughly 70 members (~30%) in our circular permutation library with equal or better than wild type
activity for reduction of ketoisophorone (1), a widely used reference substrate for the functional
evaluation of OYE family members (see below).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Building on the idea of a synthetic protein engineering approach, we employed the PURE system
for functional analysis of the cpOYE library. PURE distinguishes itself from other IVTT systems in that
all of its components are isolated and purified individually, followed by reconstitution of a chemically
fully defined functional protein synthesis machinery.39,40 As such, PURE dramatically lowers the
background signal caused by contaminating endogenous reductases in cellꢀbased in vitro and in vivo
ACS Paragon Plus Environment

Page 11 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
expression systems and allows for direct assaying of enzyme activity (Fig. S2). Furthermore, PURE is
highly scalable and allows for parallel synthesis of library members in 96ꢀwell microtiterplate format with
relatively consistent protein yields (±50% based on SDSꢀPAGE analysis; data not shown). In subsequent
activity assays, a 10ꢀꢁL PURE reaction produced enough enzyme (~ 0.1 ꢁM) for detecting reduction of 1
by OYE1 variants with as little as 10% of wild type activity. Activity assays performed under anaerobic
conditions maximize the signalꢀtoꢀnoise ratio, presumably eliminating side reactions and enzyme
inactivation by reactive oxygen species generated in the presence of reduced FMN.20 Finally, expression
of catalytically competent OYE required supplementation of the IVTT reaction mixture with FMN.
Maximum enzyme activity was reached at [FMN] >1 ꢁM which is approximately 100ꢀfold above the
reported dissociation constant for FMN in OYE1 (Fig. S3).43 We further raised the [FMN] to 100 ꢁM to
ensure saturation of enzyme with cofactor, even for engineered OYEs with potentially lower cofactor
binding affinity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Our initial functional evaluation of the cpOYE library focused on the stereoselective reduction of
1 to Rꢀ2, one of the most prominent industrial application for eneꢀreductases and a widely used standard
for the functional evaluation of OYE family members.16,17,44ꢀ47 Preliminary screening with PURE quickly
identified ~70 cpOYE variants with catalytic activity equal or better than wild type enzyme. The location
of the new termini in these active variants fell into four distinct sectors within the primary sequence (Fig.
2A). Sector I covers most of the exterior helical subdomain (amino acid residues 125ꢀ160; numbering
based on OYE1) while sectors II and III include loop/helix regions 5 (residues 250ꢀ265) and 6 (residues
290ꢀ310), respectively (Fig. 3). Sector IV represents a short loop (residues 375ꢀ380) near the native Cꢀ
terminus. The activity assay suggested that permutants with new termini in sector III were particularly
beneficial for effective reduction of 1, showing >400% of wild type activity for multiple cpOYE variants.
Analysis of the reaction mixtures by chiral GC confirmed unchanged Rꢀenantioselectivity for all active
cpOYE variants with nativeꢀlike eeRꢀvalues of >98%. From a practical perspective, we were interested in
whether the observed activity differences in PURE were reliable predictors of catalytic activity measured
with purified enzyme. To assess the predictability of activity in PURE, we selected nine cpOYE variants
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 36
1
2
3
4
5
6
7
8
9
(cpOYE146, 154, 160, 257, 260, 291, 292, 303, 306 and 378; the number indicates the location of a
variant’s new Nꢀterminal based on the numbering of residues in wild type enzyme) which showed 1 to 4ꢀ
fold activity gains over OYE1 in the PURE reactions with 1. Enzyme samples of these nine CP variants
and OYE1 were obtained through (traditional) heterologous expression and purification, followed by
kinetic measurements of individual conversion rates for 1 (Table 1). Determination of the Michaelisꢀ
Menten parameters was not possible as enzyme saturation could not be reached due to limited solubility
of 1. The kinetic data showed rate increases by up to 19ꢀfold (for cpOYE303) over wild type enzyme. The
data also indicate that PURE systematically underestimates the rates compared to purified enzyme.
Nevertheless, the observed changes in the IVTT system were overall proportional ꢀ the top performers in
the PURE system also showed the highest catalytic gains with purified enzyme. These results suggest that
the IVTT results are semiꢀquantitative and can be used as reliable predictors of enzyme activity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The experimental data for reduction of 1 led to three questions concerning the structureꢀfunction
relationship in these engineered OYE variants: a.) will the same four sectors be identified upon screening
with other substrates, b.) will sector III always be the preferred site for termini relocation, and c.) will
native Rꢀenantioselectivity be preserved for other substrates? We explored the first question by screening
the cpOYE library for improvements in eneꢀreductase activity with cinnamaldehyde (3) and Sꢀcarvone (5)
as substrates (Fig. 2B/C). Both compounds are important intermediates in the fragrance and flavor
industry. Library screening with these two substrates identified functional variants in the same four
sectors of the protein sequence. For conversion of 3, the catalytic improvements upon CP appeared
moderate, an observation that was confirmed in activity assays with purified enzyme which detected up to
2ꢀfold rate increases in catalytic efficiency for cpOYE260 (Table 1). We hypothesize that the already high
catalytic efficiency of OYE1 for reduction of 3 makes it more likely for additional functional gains to be
countered by undesirable structural perturbations upon CP. In contrast, the screening of our cpOYE
library with 5 showed significant catalytic improvements for variants in all four sectors. These functional
gains were independently verified in subsequent activity assays with purified enzyme, showing 3 to 13ꢀ
fold increases in the rate of conversion (Table 1). As for 1, the limited substrate solubility of 5 prevents
ACS Paragon Plus Environment

Page 13 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
the determination of detailed kinetic parameters and no changes in the native enzyme’s Rꢀselectivity for
the conversion of Sꢀ5 to cisꢀ1R/4Sꢀ6 were detected. Summarizing the screening results for 1, 3 and 5, all
active candidates among the OYE1 variants had their new protein termini located in the same four
sectors. The exclusivity of functional variants in these sectors might at least in part associated with
protein foldability upon termini relocation. Nevertheless, their proximity to the active site and in some
cases significant improvement in catalytic activity over the native enzyme suggests additional benefits
arising from changes in conformational flexibility and active site accessibility. The same data also showed
that relative catalytic performance of cpOYE variants differs with individual substrates. Although variants
in sector III are clearly the best biocatalysts for 1, sectors I and IV offer slight activity gains for 3 while
variants from all four sectors show substantive functional benefits for reduction of 5. The observed
preference of the three substrates for protein termini in certain sectors and those sectors’ location relative
to the substrate binding site could provide clues to the underlying structureꢀfunction relationship. Future
inꢀdepth structure studies on selected cpOYE variants via xꢀray crystallography will help to address these
questions in more detail. In summary, the relocation of protein termini in OYE1 introduced structural and
conformational changes in distinct portions of the active site, in turn affecting substrate binding affinity,
orientation and catalytic turnover.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To further investigate the nature of the catalytic rate enhancements, we conducted a series of
rapid reaction kinetic experiments with wild type OYE1 and cpOYE303, the variant with the most
significant functional improvements (Fig. 4). The enzyme’s catalytic cycle can be split into two steps; a
reductive half reaction representing the NADPHꢀdriven reduction of the oxidized flavoprotein and an
oxidative half reaction involving reoxidation of the reduced flavin cofactor upon substrate conversion
(Scheme 1). Based on studies by Massey and coworkers, the reductive half reaction can be further divided
into three subꢀsteps: i) initial binding of NADPH to enzyme (E)•FMNox followed by ii) conformational
changes and repositioning of the nicotinamide cofactor from E•FMNox•NADPH to E*•FMNox•NADPH
and finally iii) hydride transfer.29 For wild type OYE1, formation of E•FMNox•NADPH is observed in
stoppedꢀflow experiments as the adduct generates a shortꢀlived chargeꢀtransfer complex with a
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 36
1
2
3
4
5
6
7
8
9
characteristic redꢀshifted FMN absorption peak near 460 nm. Subsequent reorientation of the flavin
(E*•FMNox•NADPH) and reduction of the flavin cofactor results in the disappearance of the 460ꢀnm
band. In agreement with Massey’s data, we detected the intermediate in our reaction catalyzed by wild
type OYE1, yet the global fit analysis of spectral data for the engineered variant cpOYE303 did not
indicate accumulation of an observable chargeꢀtransfer complex prior to reduction (Fig. 4B). At the same
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
time, the overall rates of the reductive half reaction for OYE1 (kred = 5.1 ± 0.2 sꢀ1) and cpOYE303 (kred
=
5.4 ± 0.2 sꢀ1) remained largely unchanged (Fig. 4C). Given these observations, we rationalized the
absence of an intermediate in cpOYE303 with an increased dissociation rate for NADPH from
E•FMNox•NADPH. Such an explanation is also consistent with previous reports from structural studies
that emphasized the importance of conformational changes in loop region 5 (amino acid positions 290ꢀ
310) as part of the reductive half reaction.31 The location of the new protein termini in cpOYE303 in this
region is likely responsible for the lower NADPH binding affinity, reflecting the increased local
conformational flexibility upon cleavage of the polypeptide sequence which has been observed in
numerous structure studies with circularly permutated proteins.35,48ꢀ51 In summary, protein sequence
reorganization by CP does impact the NADPH binding step of the reductive half reaction but at least in
the case of cpOYE303 leaves the overall rate of the reductive half reaction unchanged.
In contrast, the observed rate for the oxidative half reaction increases significantly upon CP and
largely accounts for the gains in catalytic activity measured by steadyꢀstate kinetics. Following preꢀ
incubation of the two OYEs with NADPH to fully reduce the enzymeꢀbound FMN, the rates of cofactor
reoxidation (kox) were measured after mixing with 1 (Fig. 4D). For both enzymes, the data could be fitted
to a singleꢀexponential function, yielding initial rates of conversion for cpOYE303 and OYE1 of 0.63 ±
0.09 sꢀ1 and 0.06 ± 0.01 sꢀ1, respectively. The data indicate that FMN reoxidation with 1 is significantly
slower than cofactor reduction in both OYEs and likely represents the rateꢀlimiting step in the reaction
cycle. More importantly, we observe an elevenꢀfold rate increase for cpOYE303 over OYE1 which
accounts for most of the gains in catalytic activity observed for the OYE variant in our steadyꢀstate
kinetic measurements (Table 1). Although these initial studies do not reveal detailed information
ACS Paragon Plus Environment

Page 15 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
regarding the effects of CP on the individual steps along the reaction coordinate, the results clearly show
that rate enhancement in the OYEꢀcatalyzed reduction of 1 by cpOYE303 is due to changes in the
oxidative half reaction. However, it seems unlikely that the observed functional effect is general. Previous
studies have shown that the nature of the rateꢀlimiting step in OYE can change as a function of
substrate.28,29 While we believe that the catalytic gains measured for the reduction of Sꢀ(5) originate from
similar beneficial effects on the oxidative half reaction, the lack of change for 3 could be explained by
differences in the rateꢀdetermining step.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the absence of detailed crystallographic information, we investigated perturbations of the
active site environment as a result of CP via changes in the spectral properties of the flavin cofactor. Upon
binding of pꢀhydroxybenzaldehyde to OYE1, Massey and coworkers observed distinct long wavelength
chargeꢀtransfer bands ranging from 560 to 590 nm with λmax values depending on hydrogenꢀbonding
interactions of the phenolate and the FMN redox potential.43,52 The spectral properties of cpOYE variants
can therefore serve as sensitive probes for the active site environment. Our analysis of representative
variants from all four sectors with pꢀhydroxybenzaldehyde did indicate subtle but significant active site
changes (Table 2 & Fig. S1). While small, twoꢀfold differences in dissociation constants for the phenolate
are consistent with overall integrity of the active site binding pocket, the shifts in λmax values of the
chargeꢀtransfer bands from 564 nm to 582 nm (relative to 578 nm for wild type OYE1) reflect changes in
the FMN redox potential of ~20 mV which suggest small conformational and environmental perturbations
near the cofactor. Nevertheless, the change in redox potentials for the cpOYE variants did not show a
clear correlation with catalytic activity for our three substrates.
CONCLUSIONS
We have identified a series of circularly permuted OYE variants with enhanced eneꢀreductase
activity for two representative enones and an enal. For all three substrates, the new protein termini of
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 36
1
2
3
4
5
6
7
8
9
catalytically active and improved enzymes fall in the same four structural regions including three loops
and a helical subdomain near the active site. The functional benefits of new termini in these regions
varied with substrate, presumably reflecting subtle differences in protein structure and interaction with the
substrates. The native enzyme’s high enantioselectivity was preserved in all cpOYE variants and catalytic
activity for 1 and Sꢀ5 was improved by >10ꢀfold. The rapid reaction kinetics analysis of the oxidative and
reductive half reactions in wild type OYE1 and cpOYE303 suggests that the functional gains can almost
exclusively be attributed to improvements of the rateꢀlimiting oxidative half reaction with 1. Further inꢀ
depth studies of the individual steps in the reaction cycle will be required to explore whether the observed
rate enhancements are more specifically due to changes in binding of 1, release of (R)ꢀ2, or the catalytic
step. Initial evidence in support of beneficial changes to substrate/product affinity upon CP was found in
the absence of the characteristic chargeꢀtransfer complex (E•FMNox•NADPH) in the reductive half
reaction of cpOYE303. The accumulation of the complex in wild type OYE1 has been associated with a
required conformational change involving loop 5; the same loop where the new protein termini in
cpOYE303 are located. The increased conformational flexibility of a protein’s termini region is well
established in the literature and could explain the improved catalytic activity in our engineered OYE
variants. Crystallographic and inꢀdepth biophysical studies will help to further rationalize the observed
differences in catalytic performance. Beyond the exploration of fundamental aspects of enzyme catalysis,
our results demonstrate that CP offers an effective strategy for improving the catalytic performance of
OYE1. Although a direct comparison with other mutagenesisꢀbased engineering approaches is
complicated by differences in reaction conditions and substrates, the rate enhancements for our cpOYE
variants are of similar magnitude or better than for previously reported OYE variants.22,53 For future
experiments, the structural and functional similarities among enzymes of the OYE family suggests that
CP could result in similar functional gains in other family members. Furthermore, selected cpOYE
variants do not necessarily mark the end point of a protein engineering project but can serve as novel
templates for subsequent mutagenesis experiments, allowing for even greater diversity in tailoring these
biocatalysts to specific conditions and substrates. On the technical side, the use of a chemically defined
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 17 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
cpOYE gene library in combination with the PURE system dramatically accelerated the process of
identifying catalytically improved cpOYE variants for any substrate of choice. As technological advances
continue to raise efficacy and lower costs, we envision such synthetic protein engineering approaches to
become a popular new tool for synthetic biologists. For engineering OYE and cofactorꢀdependent
enzymes in general, the use of IVTT systems also offers unprecedented opportunities for combinatorial
approaches to exploring the functional impact of cofactor substitutions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 36
1
2
3
4
ASSOCIATED CONTENT
5
6
Supporting Information. Additional details on materials, methods, and data. This material is available
7
8
free of charge via the Internet at http://pubs.acs.org.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
* Email: sal2@emory.edu
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
The authors like to thank Drs. Roberto Orru and Dale Edmondson for their assistance with the stopped
flow experiments and members of the Atlanta Flavin Group and the Lutz lab for helpful comments and
suggestions. Support with PURExpress by Corinna Tuckey from New England Biolabs (Ipswich, MA) is
acknowledged. This work was in part funded by the US National Science Foundation (CBET 0730312 &
1159434).
ACS Paragon Plus Environment

Page 19 of 36
Journal of the American Chemical Society
1
2
3
4
REFERENCES
5
6
7
8
9
(1)
(2)
(3)
(4)
(5)
Lu, S. M.; Bolm, C. Chemistry 2008, 14, 7513.
Taylor, C. J. S. N.; Jaekel, C. Adv. Synth. Catal. 2008, 350, 2708.
Martin, N. J. A.; List, B. J. Am. Chem. Soc. 2006, 128, 13368.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Tuttle, J. B.; Ouellet, S. G.; MacMillan, D. W. C. J. Am. Chem. Soc. 2006, 128, 12662.
Bornscheuer, U. T.; Huisman, G. W.; Kazlauskas, R. J.; Lutz, S.; Moore, J. C.; Robins, K. Nature
2012, 485, 185.
(6)
(7)
(8)
(9)
Nestl, B. M.; Nebel, B. A.; Hauer, B. Curr. Opin. Chem. Biol. 2011, 15, 187.
Wohlgemuth, R. Curr. Opin. Biotechnol. 2010, 21, 713.
Kawai, Y.; Inaba, Y.; Tokitoh, N. Tetrahedron-Asymmetr. 2001, 12, 309.
Kosjek, B.; Fleitz, F. J.; Dormer, P. G.; Kuethe, J. T.; Devine, P. N. Tetrahedron-Asymmetr.
2008, 19, 1403.
(10)
Hall, M.; Stueckler, C.; Hauer, B.; Stuermer, R.; Friedrich, T.; Breuer, M.; Kroutil, W.; Faber, K.
Eur. J. Org. Chem. 2008, 9, 1511.
(11)
(12)
Swiderska, M. A.; Stewart, J. D. J. Mol. Catal. B-Enzym 2006, 42, 52.
Adalbjornsson, B. V.; Toogood, H. S.; Fryszkowska, A.; Pudney, C. R.; Jowitt, T. A.; Leys, D.;
Scrutton, N. S. Chembiochem 2010, 11, 197.
(13)
Toogood, H. S.; Fryszkowska, A.; Hare, V.; Fisher, K.; Roujeinikova, A.; Leys, D.; Gardiner, J.
M.; Stephens, G. M.; Scrutton, N. S. Adv. Synth. Catal. 2008, 350, 2789.
Toogood, H. S.; Gardiner, J. M.; Scrutton, N. S. Chemcatchem 2010, 2, 892.
Stueckler, C.; Mueller, N. J.; Winkler, C. K.; Glueck, S. M.; Gruber, K.; Steinkellner, G.; Faber,
K. Dalton T. 2010, 39, 8472.
(14)
(15)
(16)
(17)
Hall, M.; Stueckler, C.; Ehammer, H.; Pointner, E.; Oberdorfer, G.; Gruber, K.; Hauer, B.;
Stuermer, R.; Kroutil, W.; Macheroux, P.; Faber, K. Adv. Synth. Catal. 2008, 350, 411.
Winkler, C. K.; Tasnadi, G.; Clay, D.; Hall, M.; Faber, K. J. Biotechnol. 2012, 162, 381.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 36
1
2
3
4
5
6
7
8
(18)
Schittmayer, M.; Glieder, A.; Uhl, M. K.; Winkler, A.; Zach, S.; Schrittwieser, J. H.; Kroutil, W.;
Macheroux, P.; Gruber, K.; Kambourakis, S.; Rozzell, J. D.; Winkler, M. Adv. Synth. Catal.
2011, 353, 268.
9
(19)
(20)
Pompeu, Y. A.; Sullivan, B.; Walton, A. Z.; Stewart, J. D. Adv. Synth. Catal. 2012, 354, 1949.
Hulley, M. E.; Toogood, H. S.; Fryszkowska, A.; Mansell, D.; Stephens, G. M.; Gardiner, J. M.;
Scrutton, N. S. Chembiochem 2010, 11, 2433.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(21)
Toogood, H. S.; Fryszkowska, A.; Hulley, M.; Sakuma, M.; Mansell, D.; Stephens, G. M.;
Gardiner, J. M.; Scrutton, N. S. Chembiochem 2011, 12, 738.
(22)
(23)
(24)
(25)
(26)
Bougioukou, D. J.; Kille, S.; Taglieber, A.; Reetz, M. T. Adv. Synth. Catal. 2009, 351, 3287.
Padhi, S. K.; Bougioukou, D. J.; Stewart, J. D. J. Am. Chem. Soc. 2009, 131, 3271.
Hall, M.; Bommarius, A. S. Chem. Rev. 2011, 111, 4088.
Reich, S.; Hoeffken, H. W.; Rosche, B.; Nestl, B. M.; Hauer, B. Chembiochem 2012, 13, 2400.
Brown, B. J.; Hyun, J. W.; Duvvuri, S.; Karplus, P. A.; Massey, V. J. Biol. Chem. 2002, 277,
2138.
(27)
(28)
(29)
(30)
(31)
(32)
(33)
(34)
(35)
(36)
(37)
(38)
Vaz, A. D. N.; Chakraborty, S.; Massey, V. Biochemistry 1995, 34, 4246.
Niino, Y. S.; Chakraborty, S.; Brown, B. J.; Massey, V. J. Biol. Chem. 1995, 270, 1983.
Massey, V.; Schopfer, L. M. J. Biol. Chem. 1986, 261, 1215.
Fox, K. M.; Karplus, P. A. Structure 1994, 2, 1089.
Karplus, P. A.; Fox, K. M.; Massey, V. FASEB J. 1995, 9, 1518.
Yu, Y.; Lutz, S. Trends Biotechnol. 2011, 29, 18.
Qian, Z.; Lutz, S. J. Am. Chem. Soc. 2005, 127, 13466.
Qian, Z.; Fields, C. J.; Lutz, S. Chembiochem 2007, 8, 1989.
Qian, Z.; Horton, J. R.; Cheng, X.; Lutz, S. J. Mol. Biol. 2009, 393, 191.
Whitehead, T. A.; Bergeron, L. M.; Clark, D. S. Protein Eng. Des. Sel. 2009, 22, 607.
Stephen, P.; Tseng, K. L.; Liu, Y. N.; Lyu, P. C. Chem. Commun. (Camb). 2012, 48, 2612.
Guntas, G.; Kanwar, M.; Ostermeier, M. PLoS ONE 2012, 7, e35998.
ACS Paragon Plus Environment

Page 21 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(39)
(40)
Ueda, T.; Shimizu, Y.; Inoue, A.; Tomari, Y.; Suzuki, T.; Yokogawa, T.; Nishikawa, K. Nature
Biotechnol. 2001, 19, 751.
Protein Generation Using a Reconstituted System in Protein Engineering Handbook; Ying, B.
W.; Ueda, T., Eds.; WileyꢀVCH: Weinheim, 2009; Vol. 2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(41)
(42)
Brown, B. J.; Deng, Z.; Karplus, P. A.; Massey, V. J. Biol. Chem. 1998, 273, 32753.
Reitinger, S.; Yu, Y.; Wicki, J.; Ludwiczek, M.; D'Angelo, I.; Baturin, S.; Okon, M.; Strynadka,
N. C.; Lutz, S.; Withers, S. G.; McIntosh, L. P. Biochemistry 2010, 49, 2464.
Abramovitz, A. S.; Massey, V. J. Biol. Chem. 1976, 251, 5327.
Stuermer, R.; Hauer, B.; Hall, M.; Faber, K. Curr. Opin. Chem. Biol. 2007, 11, 203.
Wada, M.; Yoshizumi, A.; Noda, Y.; Kataoka, M.; Shimizu, S.; Takagi, H.; Nakamori, S. Appl.
Environ. Microbiol. 2003, 69, 933.
(43)
(44)
(45)
(46)
(47)
(48)
Kataoka, M.; Kotaka, A.; Hasegawa, A.; Wada, M.; Yoshizumi, A.; Nakamori, S.; Shimizu, S.
Biosci. Biotechnol. Biochem. 2002, 66, 2651.
Fryszkowska, A.; Toogood, H.; Sakuma, M.; Gardiner, J. M.; Stephens, G. M.; Scrutton, N. S.
Adv. Synth. Catal. 2009, 351, 2976.
Ay, J.; Hahn, M.; Decanniere, K.; Piotukh, K.; Borriss, R.; Heinemann, U. Proteins 1998, 30,
155.
(49)
(50)
Chu, V.; Freitag, S.; Le Trong, I.; Stenkamp, R. E.; Stayton, P. S. Protein Sci. 1998, 7, 848.
Manjasetty, B. A.; Hennecke, J.; Glockshuber, R.; Heinemann, U. Acta Crystallogr. D Biol.
Crystallogr. 2004, 60, 304.
(51)
(52)
(53)
Pieper, U.; Hayakawa, K.; Li, Z.; Herzberg, O. Biochemistry 1997, 36, 8767.
Matthews, R. G.; Massey, V.; Sweeley, C. C. J. Biol. Chem. 1975, 250, 9294.
van den Heuvel, R. H.; van den Berg, W. A.; Rovida, S.; van Berkel, W. J. J. Biol. Chem. 2004,
279, 33492.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 36
1
2
3
4
FIGURE LEGENDS
5
6
7
8
9
Figure 1. Assembly of cpOYE library via wholeꢀgene synthesis. Two copies of truncated oye1 (marked in
red and blue) were prepared that lack the three native Cꢀterminal amino acids and the Start/Stop codons.
The two genes were assembled in tandem and connected by a 9ꢀnucleotide sequence encoding the ꢀGlyꢀ
ThrꢀSerꢀ linker. PCR amplification with siteꢀspecific primers and the tandem repeat as template yielded
individual members of the CP library and reintroduced Start/Stop codons, as well as flanking restriction
sites for the subsequent cloning into pETꢀ14b. Correct gene constructs (cpOYE#) were confirmed by
DNA sequencing, sorted and stored as purified plasmids in microtiterplates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Primary screening data of cpOYE library for eneꢀreductase activity on A.) ketoisophorone (1)
to levodione (Rꢀ2), B.) cinnamaldehyde (3) to dihydrocinnamaldehyde (4), and C.) carvone (Sꢀ5) to
dihydrocarvone (cisꢀ(1R,4S)ꢀ6). Wholeꢀgene synthesis created a native cpOYE library with perfect
distribution (red lines in inner circle). Catalytic activity of library members for each substrate was
measured by semiꢀquantitative assay and is reflected in the length of the green lines. Wildꢀtype activity is
indicated by the greyꢀshaded area. The outerꢀmost circle marks the secondary structure elements (shades
of blue) and active site residues (yellow stars) of OYE1 (399 amino acid residues). For all three
substrates, four sectors (I ꢀ IV) in the protein sequence with activity equal or better than wildꢀtype were
identified.
Figure 3. Schematic of OYE1 structure (based on PDB access code 1K03 26). The FMN cofactor is
shown in blue sticks while the substrateꢀbinding site is occupied by pꢀhydroxybenzaldehyde (dotted
cloud). The location of permutation sites which resulted in catalytic improvements are highlighted in red
(sectors I), orange (sectors II), yellow (sectors III) and green (sectors IV), respectively.
Figure 4. Rapid reaction kinetics for wild type OYE1 and cpOYE303. A.) Complete reaction cycle for
OYE with rate constants for the reductive and oxidative half reactions of OYE1 (green) and cpOYE303
ACS Paragon Plus Environment

Page 23 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(red). B.) Changes in flavin spectra for wild type and cp variant during reductive half reaction. The
characteristic absorption band of FMNox near 460 nm (blue diamonds) disappears upon conversion to
FMNred (orange circles). For OYE1, global fit analysis indicates formation of a shortꢀlived intermediate
with a slightly redꢀshifted maximum near 460 nm (black squares) corresponding to the chargeꢀtransfer
complex in E•FMNox•NADPH. For cpOYE303, no such intermediate was detectable. C.) Observed rate
for the reductive half reaction as a function of [NADPH] based on the disappearance of the FMNox
absorption band. Given experimental restraints to [NADPH] >50 ꢁM, data analysis was limited to
determination of maximum rates (kred). D.) Flavin reoxidation (kox) upon rapid mixing of E•FMNred with 1
(0.5 mM) under anaerobic conditions followed singleꢀexponential kinetics. Experimental data are shown
for the overall and initial (insert) oxidative half reaction catalyzed by wild type OYE1 (grey) and
cpOYE303 (black). Curve fits are displayed as red lines.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 36
1
2
3
4
5
6
SCHEME 1
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 25 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
TABLE 1
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Comparison of biocatalytic rates for reduction of ketoisophorone (1), cinnamaldehyde (3), and Sꢀ
carvone (5).
EA(1)[b]
(minꢀ1)
EA(3)[c]
(minꢀ1)
EA(5)[c]
(minꢀ1)
Variant
IVT(1)[a]
OYE1
1x
6.5 (1x)
19 (3x)
53 (1x)
60 (1.1x)
83 (1.5x)
85 (1.6)
44 (0.8x)
113 (2x)
21 (0.4x)
24 (0.4x)
42 (0.8x)
57 (1x)
1.9 (1x)
1.4 (0.8x)
6 (3x)
cpOYE146
cpOYE154
cpOYE160
cpOYE257
cpOYE260
cpOYE291
cpOYE303
cpOYE306
cpOYE378
1.3x
0.8x
1.6x
1x
12.2 (2x)
10.1 (1.6x)
13 (2x)
9.5 (5x)
5.7 (3x)
2x
25.4 (4x)
36 (6x)
7.7 (4x)
3.2x
5.3x
4.2x
0.7x
3.4 (1.8x)
22.5 (12x)
23.6 (13x)
5.9 (3x)
122 (19x)
47.2 (7x)
nd
[a] Fold activity change in based on GC analysis. Conversion rate (foldꢀchange) of regular enzyme activity
(EA) measurements via [b] GC or [c] UV analysis. Standard error for IVT data ±100%; EA data ±10%. nd =
not determined.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 36
1
2
3
4
TABLE 2
5
6
7
8
Table 2. Spectral property changes upon inhibitor binding.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
E˚’
λmax(CT; nm) (ε λmax(FMN; nm)
Variant
KD(ꢁM)[a]
= Mꢀ1cmꢀ1)
(ε = Mꢀ1cmꢀ1)
(mV)[b]
OYE1
1.6 ± 0.4
1.0 ± 0.3
0.6 ± 0.2
1.9 ± 0.7
1.1 ± 0.4
4.0 ± 1.0
1.9 ± 0.2
2.9 ± 0.9
578 (4335)
582 (3753)
572 (2971)
566 (2485)
580 (2973)
564 (2956)
569 (2685)
565 (2756)
460 (10600)
457 (12784)
462 (11913)
463 (11341)
460 (11979)
464 (12260)
463 (13147)
458 (10774)
ꢀ207
ꢀ196
ꢀ210
ꢀ214
ꢀ200
ꢀ216
ꢀ212
ꢀ215
cpOYE154
cpOYE160
cpOYE257
cpOYE260
cpOYE291
cpOYE303
cpOYE306
cpOYE378
2.3 ± 0.9
578 (3702)
458 (12224)
ꢀ207
[a] Dissociation constant for pꢀhydroxybenzaldehyde. [b] E˚’ꢀvalues were estimated based on correlation of
λmax(CT) band with FMN E˚’ꢀvalues.43 Standard error: ±5 mV.
ACS Paragon Plus Environment

Page 27 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
FIGURE 1
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 36
1
2
3
4
FIGURE 2
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 29 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
FIGURE 3
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment