Rhodococcus erythropolis Oleate Hydratase: a New Member in the Oleate Hydratase Family Tree—Biochemical and Structural Studies

Article abstract of DOI:10.1002/cctc.201701350

Recently, the enzyme family of oleate hydratases (OHs: EC 4.2.1.53) has gained increasing scientific and economic interest, as these FAD-binding bacterial enzymes do not require cofactor recycling and possess high thermal and pH stability. Their products, hydroxy fatty acids, are used in specialty chemical applications including surfactant and lubricant formulations. The “oleate hydratase engineering database”, established by Schmid et al. (2017), divides all OHs into 11 families (HFam1 to 11). To date, only two crystal structures of homodimeric OHs from the families HFam2 and HFam11 have been reported. In this study, we biophysically characterized an OH belonging to the HFam3 family, originating from the marine bacterium Rhodococcus erythropolis, for the first time. The crystal structure revealed that this new OH (OhyRe) surprisingly is a monomer in its active form. This particular feature provides new avenues for enzyme engineering and recycling through immobilization.

Full text of DOI:10.1002/cctc.201701350

Accepted Article
Title: Rhodococcus erythropolis Oleate Hydratase: a New Member in
the Oleate Hydratase Family Tree - Biochemical and Structural
Studies
Authors: Jan Lorenzen, Ronja Janke, Ayk Waldow, Farah Qoura,
Bernhard Loll, and Thomas Brück
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201701350
Link to VoR: http://dx.doi.org/10.1002/cctc.201701350
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201701350
Rhodococcus erythropolis Oleate Hydratase: a New Member in the Oleate Hydratase
Family Tree - Biochemical and Structural Studies
Jan Lorenzen[a,d], Ronja Janke[b,d], Ayk Waldow , Farah Qoura , Bernhard Loll
[b]
[a]
[b,c,e]
and Thomas
Brück[
a,e]
a
Professorship for Industrial Biocatalysis, Technical University Munich, Lichtenberg Str. 4, 85748
Garching, Germany
b
Institute of Chemistry and Biochemistry, Freie Universität Berlin, Takustr. 6, 14195 Berlin,
Germany
c
d
e
moloX GmbH, Takustr. 6, 14195 Berlin, Germany
These authors contributed equally to this work
To whom correspondence should be addressed:
Prof. Dr. Thomas Brück
Professorship for Industrial Biocatalysis
Department of Chemistry
Technical University Munich
Lichtenberg Str. 4, 85748 Garching, Germany
Tel.: +49-(0) 89 289-13253
E-mail: brueck@tum.de
Dr. Bernhard Loll
Institute of Chemistry and Biochemistry
Structural Biochemistry
Freie Universität Berlin
Takustr. 6,
14195 Berlin, Germany
Phone: +49 (0) 30 838-57348
E-mail: loll@chemie.fu-berlin.de
1
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
Abstract
Recently, the enzyme family of oleate hydratases (OHs: EC 4.2.1.53) have gained a rising
scientific and economic interest, as these FAD-binding bacterial enzymes do not require cofactor
recycling and feature high thermal and pH stability. Their products, hydroxy fatty acids, are used
in specialty chemical applications including surfactant and lubricant formulations. The “oleate
hydratase engineering database”, established by Schmid et al. (2017), divides all OHs into 11
families (HFam1 to 11). To date, only two crystal structures of homodimeric OHs from the families
HFam2 and HFam11 have been reported. In this study, we biophysically characterized an OH
belonging to the HFam3 family, originating from the marine bacterium Rhodococcus erythropolis,
for the first time. The crystal structure revealed that this new OH (OhyRe) surprisingly is a
monomer in its active form. This particular feature provides new avenues for enzyme engineering
and recycling via immobilization.
Introduction
Unsaturated fatty acids are toxic to many bacteria due to their surfactant properties,
destabilization of membranes and due to the unsaturated fatty acids inhibitory effect on the fatty
[1]
acid biosynthesis. Hence, mechanisms have evolved to counteract these effects when
organisms are exposed to unsaturated fatty acids, such as linoleic acid or oleic acid. In
consequence, unsaturated fatty acids are saturated by hydrogenation. However, an intermediate
[
2]
step in this hydration process is catalyzed by fatty acid hydratases. Hydration of unsaturated
fatty acids has so far only been identified in Gram-negative and Gram-positive bacteria [3] and is
mediated by hydratases.
[
4]
Hydratases are a group of lyases that catalyze hydration and dehydration of a substrate . Even
though many hydratases are known, only few OHs have been described.[2] OHs belong to the
group of fatty acid hydratases (EC 4.2.1.53) and convert oleic acid into (R)-10-hydroxystearic acid
10-HSA).[ To date all characterized OHs bind an FAD cofactor, albeit the role of FAD has
5]
[2]
(
remained enigmatic.
Figure 1: Oleate hydratase (OH) reaction scheme.
Merely two structures of OHs have been reported. One of the OH from Lactobacillus acidophilus
[
6]
(
LAH), crystalized in the apo state as well as bound to the preferred substrate linoleic acid. The
second structure of an OH (OhyA) has been determined from Elizabethkingia meningoseptica,
[
7]
bound to its cofactor FAD. Even though LAH and OhyA convert different substrates, both utilize
FAD as a cofactor. Interestingly the bound FAD was oxidized in both hydratases, but reduced
[
7]
FAD led to significantly higher activity in OhyA.
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
Most recently, Schmid et al. assembled a “hydratase engineering database” including a total of
[8]
2
046 protein sequences. Sequence comparison revealed 11 distinct families (HFam1 to 11) with
a sequence identity of 62 % within the respective family. The only two structurally characterized
OHs, LAH and OhyA fall into two different families, HFam2 and HFam11. For all other families
structural information has remained elusive. Remarkably, in regions attributed to decorate the
active site, amino acid sequences are not strictly conserved within these families.
Hydratases are in particular interesting for commercial applications since their products, hydroxyl
fatty acids, have a wide range of industrial and pharmaceutical applications. Hydroxyl fatty acids
can be utilized in applications for the food industry as precursors of lactones,[9] as well as an
emollient for cosmetics. Further applications are the employment as a surfactant or as additives
in lubricant formulation and as a green material in polymer science.[10]
In this study, we aimed at obtaining novel insights into the reaction mechanism of the OhyRe
enzyme of R. erythropolis, using a combined structural, biophysical and biochemical approach.
The OhyRe enzyme was biochemically characterized in terms of pH and temperature optimum
as well as substrate specificity. The kinetic constants were determined for the conversion reaction
of oleic acid to 10-HSA. Furthermore, we were able to solve the crystal structure of OhyRe at 2.64
Å resolution. The structure of OhyRe provides structural information for the OH HFam3 family of
OHs for the first time. We have compared the overall architecture as well as structural features of
the OhyRe active site to a set of reported OHs. As we identified two key residues, which differed
in the OhyRe structure, these were replaced by site directed mutagenesis to residues conserved
in the other family members. Most interestingly, these conservative amino acid replacements
resulted in a strong decrease of hydration activity, indicating that OhyRe may follow a distinct
reaction mechanism.
Results & Discussion
Isolation & characterization of OhyRe derived from R. erythropolis
R. erythropolis is a Gram-positive, non-motile representative of the nocardiaceae family.[11] R.
erythropolis is a wide spread soil bacterium which is mainly found in marine habitats, from coastal
[
12]
[13]
to deep sea sediments
and exhibits a broad range of biochemically interesting enzymes. As
the genome of R. erythropolis has recently been deciphered,[14] we have performed sequence
specific genome mining, targeting alternative OH activities in this extremophile organism.
To this end we have identified the 1.68 kB sequence related to known OHs albeit at a very low
sequence identity level of max. 38 %. The gene for the putative OH was termed ohyRe, according
[
5a]
to the nomenclature used by Bevers et al. for the Elizabethkingia meningoseptica OH ohyA.
The corresponding OhyRe protein is composed of 559 amino acids with a calculated molecular
weight of 66.3 kDa. The N-terminally His -tagged protein was successfully expressed in
6
Escherichia coli BL21(DE3) cells. Initial activity tests with the crude extract resulted in high
conversion rates (approx. 85%) from oleic acid to 10-HSA, confirming that OhyRe indeed
2+
displayed OH activity. To further characterize the enzyme, we have devised a Ni -affinity
purification protocol. After final His-Trap HP affinity chromatography, the purified enzyme was
colorless and lost its hydration activity (Fig. S1). This observation suggested a loss of the FAD
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
cofactor, which has been uniformly reported to be essential for the hydration activity in all known
OHs. Moreover, the only reported loss of FAD during purification was also associated with a
[
6]
strong decrease in hydration activity. Therefore, the addition of various FAD concentrations (5
µM to 100 µM) to the purified enzyme preparation has been evaluated to reconstitute the activity
of purified OhyRe. The addition of FAD to a final concentration of 20 µM lead to the recovery of
the enzymatic activity of OhyRe and was applied for all following experiments. After establishing
an efficient purification protocol, OhyRe was biochemically characterized.
Initial tests demonstrated that purified OhyRe has a temperature optimum for the hydration activity
at 28 °C (Fig 2). Interestingly, OhyRe exhibits a relatively wide temperature tolerance with high
activity in the low temperature region (72 % of max. at 20 °C) and still moderate activity in the
high temperature area (21 % of max. at 45 °C). This wide temperature tolerance might reflect the
diversity of the occurrence of R. erythropolis from costal to deep sea habitats.
100
90
80
70
60
50
40
30
20
10
0
15
20
25
30
35
40
45
Temperature [°C]
Figure 2: Determination of the temperature optimum for the oleate hydratase (OH) OhyRe from R. erythropolis in 20
mM Tris-HCl buffer at pH 7.2. Relative activity is given to the temperature optimum at 28 °C (100 %). All experiments
were carried out in biological duplicates and technical triplicates.
The determination of the pH optimum (Fig. 3) showed a high pH tolerance in a range from pH 5.0
to 8.0 for OhyRe, the highest activity was reached at pH of 7.2.
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
100
95
90
85
5
6
7
8
pH
Figure 3: Determination of the pH optimum for the oleate hydratase (OH) OhyRe from R. erythropolis at 28°C. Relative
activity is given to the optimum at pH 7.2 (100 %). All experiments were carried out in biological duplicates and technical
triplicates.
These results correlate with the recommended cultivation conditions for R. erythropolis published
by the German Collection of Microorganisms and Cell Culture (Leibniz Institute DMSZ).
The substrate spectrum of OhyRe was experimentally determined in reactions with the purified
enzyme and different free fatty acids (FFAs) (Table 1). All experiments were carried out in
biological duplicates and technical triplicates. The enzyme showed neither detectable activity for
any saturated fatty acid tested (C14:0 to C18:0), nor for myristoleic acid, vaccenic acid,
eicosatrienoic acid and eicosatetraenoic acid, respectively. Low conversion rates could be
observed for linoleic acid, α- and γ-linolenic acid. The highest affinity, under the given conditions,
was detected for oleic acid, followed by palmitoleic acid. Complex lipids, like the tested triolein or
microalgae derived oils were not converted by the enzyme. These results correspond to published
[
6]
data for other known OHs and confirm the preference of OHs for the 9-Z double bond in the
fatty acid carbon chain. The kinetic parameters were experimentally defined for OhyRe, with oleic
acid as the preferred substrate, and resulted in a K value of 0.49 ± 0.1 mM and a kcat value of 34
M
-
1
±
5 min .
Table 1: Tested substrates, identified products and corresponding specific activity of conversion reaction with OhyRe.
All experiments were carried out in biological duplicates and technical triplicates.
specific activity
substrate
product
-
1
-1
[
nmol mg-enzyme min ]
C16:1 ω7 palmitoleic Acid 10-hydroxypalmitic acid
1205 ± 70.6
1266 ± 30.3
156 ± 3.3
C18:1 ω9 oleic Acid
10-hydroxystearic acid
C18:2 ω6 linoleic Acid
10-hydroxy-12(Z)-octadecenoic acid
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
C18:3 ω3 α-linolenic Acid
C18:3 ω6 γ-linolenic Acid
10-hydroxy-12,15(Z,Z)-octadecadienoic acid 150 ± 1.7
10-hydroxy-6,12(Z,Z)-octadecadienoic acid 615 ± 36.0
OhyRe mutants
Engleder et al. [7] demonstrated by multiple sequence alignment of 10 OHs that the catalytically
crucial residues are highly conserved within the OH family. The alignment of OhyRe with 14
different OHs (Fig. S2) predominantly confirmed this hypothesis, but also depicted two
exceptions. Interestingly, two amino acid residues that are highly conserved in all other OHs differ
in the sequence of OhyRe, namely the residues M77 and V393. Further structural comparison of
OhyRe and OhyA (Fig. S3) showed that these residues correlate with the residues E122 and
T436 from OhyA, which are responsible for the activation of the water molecule or the binding of
[
7]
the carboxylate, respectively. A conservative substitution of the two identified residues to the
familial conserved residues (M77E, V393T) was executed by site directed mutagenesis and
resulted in the effect depicted in Fig. 4.
Figure 4: Relative conversion activity from oleic acid to 10-HSA from OhyRe wild type enzyme (OhyRe WT), set as
1
00 %, and its respective mutants (M77E, V393T). All experiments were carried out in biological duplicates and
technical triplicates.
Fig. 4 shows the relative hydration activity of the OhyRe wild type enzyme compared to the two
constructed mutants M77E and V393T with oleic acid as substrate. The conservative substitution
of both residues resulted in a strong decrease in the conversion efficiency from oleic acid to 10-
HSA of 81 % and 90 %, respectively. This strong decrease in hydration activity supports the
hypothesis that OHs from the HFam3 family might follow a distinct catalytic mechanism.
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
In comparison to the two other structurally characterized OHs LAH [6] and OhyA , OhyRe shows
a sequence identity of 35 % and 34 %, respectively. Beside the low sequence identity, both the
LAH and OhyA have large N- and C-terminal amino acid sequence extensions (Fig. S4 and 6A,B),
which are absent in OhyRe. To understand whether those extensions are common for all other
OHs, we performed a sequence alignment (Fig. S2). Interestingly, OH sequences can be roughly
divided in sequences with amino acid extensions at the N- and the C-termini present or absent.
Consequently, this raised the question, whether OhyRe might be structurally different compared
to the members of the HFam2 and HFam11 families. Hence we tried to crystalize the protein for
structure determination by X-ray crystallography.
[7]
Overall structure
We obtained crystals of OhyRe that belong to space group P6 22. The diffraction of the OhyRe
5
crystals was anisotropic, resulting in complete data set to a resolution of 2.64 Å. The structure of
OhyRe was determined by molecular replacement with two protein chains in the asymmetric unit
corresponding to a very high solvent content of about 75%. The structure was refined with good
stereochemistry to good R factors (Table S1, supporting information). Both polypeptide chains
are practically identical with a root mean square deviation (rmsd) of 0.1 Å for 522 pairs of Cα-
atoms. The structure is complete, except for the 5 N-terminal residues and a loop region ranging
from residue 42 to 67 due to missing electron density caused by its high flexibility (Fig. 5A). No
redox cofactor could be identified in the electron density maps. Hence our structure represents
the apo-state of OhyRe from R. erythropolis.
Figure 5: Overall structure of OhyRe in cartoon representation. (A) Domain organization of OhyRe: Domain I in orange,
domain II in green, domain III in pink and domain IV in yellow. The bound Mg2+ and coordinating waters are shown as
orange and red spheres, respectively. Magenta spheres indicate the boarders of a flexible linker, ranging from residue
6
1 to 86, which could not be modelled due to its flexibility. (B) Zoom onto domain IV with a superposition of OhyRe and
7
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
LAH, domain IV of LAH is shown in gray. (C) Superposition of OhyRe and OhyA, domain IV of OhyA is shown in gray.
Dashed line indicates an un-modelled loop region.
The closest related structures, as analyzed by the DALI software,[15] are the OHs LAH (PDB-id
4ia6) and OhyA (PDB-id 4uir). OhyRe superimposes with a rmsd of 1.8 Å for 500 pairs of Cα-
atoms and 1.6 Å for 501 pairs of Cα-atoms, respectively. In analogy to the structures of LAH and
OhyA, OhyRe is composed of three core domains (residues 1–540; Fig. 5A and 6) that are related
to other FAD-dependent enzymes. Domain I (residues 6–115, 223–300, 322–344, and 478–540)
is a mixed α/β domain composed of a parallel five-stranded β-sheet packed between two α-helices
on one side and a three-stranded antiparallel β-sheet on the other side (Fig. 5A), resembling a
variant of the Rossmann fold. Domain II (residues 116–131, 301–321, and 345–477) consists of
an antiparallel β-sheet (Fig. 5A) flanked by three α-helices defining the cofactor- and substrate-
binding site in conjunction with domain I. Domain III (residues 132-222) is exclusively α-helical
(
5
Fig. 5A) and its fold is structurally related to monoamine oxidases.[16] Domain IV (residues 541-
59) of OhyRe is composed of merely one single α-helix. In contrast to the structures of LAH and
OhyA the C-terminal domain is truncated by 32 or 48 residues, respectively. As a consequence
two α-helices are absent in domain IV of OhyRe (Fig. 5B and C). The structure of LAH has been
determined in apo and product-bound state. Upon substrate recognition and binding the two C-
terminal α-helices of domain IV undergo a large conformational change and thereby allow the
[
6]
substrate to enter a cavity that ends at a cleft between domain I and III.
For both the LAH and OhyA a homodimeric quaternary structure in solution as well as in crystallo
[
6-7]
[17]
[18]
has been reported.
M. caseolyticus
For other OHs such as the one from L. fusiformis, S. maltophilia
and
[20]
[
19]
dimers in solution were reported as well. As analyzed by the PISA server,
OhyRe is a monomer in crystallo. To analyze the situation in solution and to confirm our structural
analysis, we performed multi-angle light scattering (MALS) experiments. Our experiments clearly
revealed that OhyRe is monomeric in solution (Fig. S5). The latter finding is in agreement to our
observations for OhyRe assembly in the crystal and strengthens our hypothesis that OHs, lacking
N- and C-terminal extensions, are monomers. Additionally we examined whether the extensions
might have functional implications. As discussed above, in the structure of LAH, domain IV
reflects the C-terminal extension, suggested to be involved in substrate gating. Since the two
reported and our crystal structures cover nearly the complete amino acid sequence of the
respective protein, we superimposed the structures and inspected in particular the monomer-
monomer interfaces of dimeric LAH and OhyA. Most of the dimer interface is established by
residues that reside within the N- or C-terminal extension, including domain IV of LAH (Fig. 6A)
and OhyA (Fig. 6B). In summary, we propose that the presence or absence of N- and C-terminal
extensions within the amino acid sequence directly influences the oligomeric state of OHs. In case
of OhyRe the extensions are absent and hence the protein is monomeric in solution.
2+
Notably, the crystal structure revealed a coordinated Mg ion close to the putative active site of
2+
OhyRe (section S1). However, experimental data showed that the Mg ion has no influence on
the hydration activity of OhyRe.
8
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
Figure 6: Oligomeric state of oleate hydratases (OHs). (B) LAH monomer I is shown in dark gray. Monomer II in surface
and cartoon representation (light gray). The N-terminal domain (NTD) and domain IV in cyan. (C) The NTD OhyA of
each monomer meanders into the other monomer.
FAD binding
All characterized OHs utilize FAD as cofactor.[2] Different roles of the FAD cofactor have been
proposed either being directly involved in catalysis or indirectly having merely a structural role to
[
6]
stabilize the enzyme. In the crystal structure of LAH, no FAD could be located in its binding site.
In contrast, in the crystal structure of OhyA, FAD is bound, but only to one of the two polypeptide
[
7]
-6
[7]
chains. Furthermore a K of 1.8 x 10 M was determined by isothermal calorimetry. All indicate
d
a rather low binding affinity of FAD to studied OHs. For our structural studies, we attempted to
add FAD prior to the final purification step or to incubate OhyRe with FAD prior to the
crystallization experiments. Moreover, we tried to soak FAD in our apo OhyRe crystals. All our
attempts to gain structural information of OhyRe bound to FAD remained unsuccessful so far. A
primary sequence analysis reveals the presence of a characteristic recognition sequence
GXGXXGX21E/D for nucleotide binding, also reflected by the Rossmann fold of the protein. A
superposition of the structures of OhyA and OhyRe (Fig. 7A and B) clearly reveals that the
cofactor binding site is large enough to accommodate a FAD molecule. Furthermore in the crystal
structure of OhyA a PEG molecule was identified that might mimic a bound oleate molecule. For
latter substrate enough space would be in the active site of our OhyRe structure (Fig. 7A and B).
9
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
Figure 7: Superposition of OhyRe and OhyA. Same color coding as in Figure 5A. (A) OhyRe is colored as in Figure 1
and OhyA in gray. View on the active site of OhyA. Magenta spheres indicate the boarders of a flexible linker in OhyRe,
ranging from residue 42 to 67, which could not be modelled due its flexibility. In contrast upon FAD binding in OhyA
this loop region is getting structured, indicated by two gray arrows. Moreover a short loop fragment rearranges upon
cofactor/substrate binding (gray arrow-headed arc). In general, there would be enough space in OhyRe to
accommodate the cofactor as well as the substrate. Solvent exposed on the back of the active site, the Mg2+ binding
site located. (B) Same view in (A) zoomed on the FAD cofactor binding site of OhyA.
Putative active site with catalytically important residues
As mentioned above OHs can be, based on their amino acid sequences, sub-divided in 11
families.[ Based on the structure of OhyA, that belongs to the family HFam1, E122 was
8]
[8]
proposed to be important for cofactor binding and being involved in catalysis by activating a water
[
7]
molecule for the attack on the partially charged double bond of the substrate. Latter residue
resides within a conserved loop region 1 RGGREM , almost strictly conserved within all OHs.
In the sequence of the OhyRe, that falls into the family HFam3, the ultimate and penultimate
amino acids are different, but characteristic for members of the largest HFam1 as well HFam3
18
123
[
8]
73
78
family. In the respective loop region OhyRe has the sequence RGGRML . The exchange of
OhyA E122 to OhyRe M77, will drastically alter the chemistry and can certainly not activate a
water molecule. To investigate the influence of amino acid at position 77 in OhyRe, we generated
a point mutation M77E. This single mutation led to a drastic decrease in hydration activity of
OhyRe. OhyA Y241 was proposed to be as well important for catalysis, by protonation of the
[
7]
double bond and is conserved with OhyRe Y205. Hence, it is tempting to propose that OhyRe
might have a different catalytic mechanism as proposed for OhyA.
Interestingly, the fatty acid double-bond hydratase from Lactobacillus plantarum, LPH,[21] also
exhibits the HFam3 family specific motive RGGREM. In contrast to LPH, which solely converts
linoleic acid, OhyRe exhibits hydration activity for both, C16 and C18 unsaturated fatty acids. In
addition, LPH was shown to be a homotrimer in its active form, whereas OhyRe is a monomer.
The exact binding site of the substrates of OHs could not yet be resolved. In the crystal structure
of LAH, a cavity was observed running from the surface of the protein towards the interface formed
by domain I and III. This site is known to be a substrate binding site in other related enzyme
families. Volkov et al. identified a linoleic acid (LA) molecule bound in this cavity with its
carboxylate oriented towards the protein surface and the hydrophobic tail pointing to the protein
1
0
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
interior (Fig. 8). In contrast, a PEG molecule has been identified close to the FAD cofactor in the
structure of OhyA, possibly mimicking a bound substrate molecule (Fig. 8). Strictly this binding
site is fundamental different to the “interdomain”-binding site as described for LAH (Fig. 7B and
Fig. 8). To date, there is no structural information on an OH with a real substrate bound to the
enzyme. Therefore it is very speculative which amino acid residues are crucial for catalysis. A
definitive answer could only provide a structure with bound substrate. This would allow the
identification of key amino acids, important for catalysis. These residues could then be targeted
by structure-based mutagenesis. Moreover, the given resolution of 2.64 Å of our OhyRe structure,
does not allow the identification of water molecules within the active site. The location of the water
molecules represents the second bottle neck for the identification of catalytically relevant
residues, as the precise position of the water molecules might be as well crucial for catalysis.
Based on the structure of OhyA, the T436A mutant had been generated, showing a dramatic
decrease in activity. Latter observation was interpreted as the hydroxyl of T436 being important
for recognition and binding of the carboxylate function of the substrate. Taken together, both
structures not only display distinct substrate binding sites, but also differ in the orientation of the
carboxylate. In order to increase the enzymatic activity of OhyRe, we speculated that a mutation
V393T (equivalent to T436 in OhyA) could potentially improve recognition and binding of the
substrate. But unexpectedly, the activity was drastically reduced (Fig 4).
Figure 8: View on the potential nucleotide and substrate binding site. Superposition of LAH (magenta) and OhyA (cyan)
as ribbon. FAD and PEG bound to LAH are shown in stick representation. The linoleic acid substrate of LAH and 2-
methyl-2,4-pentanediol (MPD) molecules bound to LAH are shown in stick representation. Notably a 2-methyl-2,4-
pentanediol (MPD) molecule in the LAH structure superimposed partially with the bound PEG molecule in the OhyA
structure. The terminal carbon atom of linoleic acid is in distance of 21 Å to FAD, indicated by a dashed black line.
Conclusion
The enzyme family of OHs offers the sustainable conversion and valorization of biogenic fatty
acids into high value lubricant additives. In contrast to heme dependent P450 hydroxylases, the
conversion of oleic acid to 10-hdroxysteric acid by OHs can be accomplished without the need
for co-factors. In this study a new OH from R. erythropolis could be identified by targeted genome
mining. The biochemical characterization showed that the enzyme, OhyRe, exhibits a high pH-
tolerance as well as high hydration activity over a wide temperature range. The crystal structure
of OhyRe was solved to a resolution of 2.64 Å, surprisingly revealing a monomeric structure for
1
1
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
the enzyme, which contrasts all other reported members of the OH family. These results were
ultimately confirmed by MALS experiments of the protein in solution. To our knowledge, OhyRe
is the first strictly monomeric OH and the first member of the HFam3 family of OHs that was
characterized and crystallized so far. The HFam3 family of OHs exhibits high potential for further
biocatalytical applications, as the monomeric enzyme provides a variety of options for
immobilization strategies. The free N- and C-terminal ends are not involved in dimer association
and could e.g. be feasible targets for the fusion of anchor-proteins. Furthermore, two mutants of
OhyRe, M77E and V393T, were generated to clarify the importance of these two residues in the
catalytic mechanism of OhyRe. These conservative substitutions, to residues being highly
conserved within in the OH family, resulted in a drastic decrease of hydration activity of 80 to 90
%. This suggests that OhyRe might follow a distinct reaction mechanism, which has to be clarified
in future experiments. Data in part presented in this manuscript are included in a patent
application, regarding the technical process for 10-HSA production (date of submission to the
EPO: 29.09.2017).
Materials and Methods
Chemicals
All chemicals utilized in this work were purchased from Sigma-Aldrich (Munich, Germany) and
Carl Roth (Karlsruhe, Germany), at the highest purity grade available.
Bacterial strains & plasmids
The Escherichia coli strains XL-1 Blue and DH5α were used for cloning whereas the strains BL21
(
DE3) and Rosetta2 were employed for protein expression. All plasmids and E. coli strains were
obtained from Novagene/Merck Millipore.
Cloning for characterization experiments
The myosin-cross-reactive antigen coding gene O5Y_00450 from Rhodococcus erythropolis
CCM2595, was taken as a template for a codon-optimized gene-synthesis (Life Technologies,
Regensburg), for an E. coli host strain. The obtained synthetic gene was sub-cloned in a pET28a
expression plasmid and transformed into chemically competent F cells.
Cloning for crystallization experiments
For protein production in E. coli, the OhyRe ORF was amplified with the forward primer 5’-
TATACCATGGGAATGAGCAGCAATCTGAGCC-3’
and
reverse
primer
5’-
TATAAAGCTTTTAACGAAACATGGTAACTGCTGC-3’ containing HindIII and NcoI recognition
sites, respectively) and cloned into the pETM11 vector, resulting in a construct with N-terminal
hexa-histidine-tag that could be cleaved off by TEV protease.
OhyRe mutant construction
1
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
The mutants OhyRe-M77E and OhyRe-V393T were constructed by overlap-extension-PCR using
primer sets listed in Table S2 (supporting information).
Protein expression and protein purification for characterization experiments
The expression of the native OhyRe or respective mutants thereof was carried out in E. coli
BL21DE3 cells, grown in Laure Broth (LB) medium. Pre-cultures were inoculated from a cryo-
stock and grown over night in 100 ml LB in a 500 ml baffled flask at 120 rpm at 37°C. Main cultures
were grown up to an optical density, measured at 600 nm (OD600), of 0.6 – 0.8, before the
expression was induced by the addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final
concentration of 0.1 mM. After 16 h of incubation at 16°C the cells were harvested and the
resuspended (20 mM Tris-HCl buffer and 25 mM imidazole pH 7.2) cell pellets were disrupted by
high pressure homogenization (EmulsiFlex-B15, AVESTIN). A subsequent centrifugation step at
20.000 g for 40 min at 4°C (Beckmann coulter J-20 XP) was applied for the separation of the cell
debris from the soluble protein fraction. The soluble protein fraction was utilized for affinity
2
+
chromatography via a Ni -NTA His-trap column (HisTrap FF, GE Healthcare; flow rate 1 ml/min).
The purified protein solution was desalted using HiPrep 26/10 desalting column (GE Healthcare).
Protein amounts were quantified using 2-D quant kit (GE Healthcare) according to manufacturer’s
instructions.
Protein expression and protein purification for crystallization experiments
OhyRe fused to an N-terminal hexa-histidine-tag in pET-M11 vector was transformed in E. coli
Rosetta2. Protein was expressed in auto-induction media at 37 °C until an OD600 ~ 0.8 was
[22]
reached and subsequently cooled down to 20 °C. Cells grew over night and were harvested by
centrifugation (6 min, 6’000 rpm at 4 °C). For resuspension of the cell pellet, buffer A was used
(
50 mM Tris/HCl pH 7.5, 500 mM NaCl, 30 mM imidazole and 1 mM DTT). Cells were lysed by
homogenization at 4 °C and the lysate was cleared by centrifugation (1 h, 21’000 rpm at 4 °C).
2+
Ni -NTA beads (cv ~ 1 ml; GE Healthcare) were equilibrated with buffer A. OhyRe was loaded
on the column and washed with 10 cv of buffer A. OhyRe was eluted in a linear gradient to buffer
A supplemented to 500 mM imidazole. Pooled fractions were dialyzed over night against buffer B
(
20 mM Tris/HCl pH 7.5, 150 mM NaCl and 1 mM DTT) supplemented with TEV protease to
cleave off the N-terminal His-tag. Uncleaved protein was separated from cleaved protein via an
2+
additional Ni -NTA affinity chromatography. Size exclusion chromatography was performed with
a HighLoad Superdex S200 16/60 column (GE Healthcare), equilibrated with buffer B. Pooled
protein fractions were concentrated with Amicon-Ultra 30’000 to 19.2 mg/ml as measured by the
absorbance at 280 nm.
Enzyme characterization
The determination of the enzymatic properties of the OH from Rhodococcus erythropolis CM2595
was (unless otherwise stated) executed under the following standard reaction conditions. The
tests were carried out in a reaction volume of 200 µl in 20 mM Tris-HCl buffer (pH 7.2), containing
the purified enzyme (final conc. of 5 µM), FAD (final conc. 20 µM) and 720 µM OA as substrate
at 28°C for 15 min. Thermo-stability was tested in a temperature range from 20 – 45°C. pH
1
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
tolerance was monitored in a range from pH 5 – 8. To determine the substrate specificity of
the myosin-cross-reactive antigen from Rhodococcus erythropolis CM2595, the purified enzyme
was tested on myristic acid (C14:0), myristoleic acid (C14:1), palmitic acid (C16:0), palmitoleic
acid (C16:1), stearic acid (C18:0), vaccenic acid (C18:1 trans-11) oleic acid (C18:1, cis-9), linoleic
acid (C18:2), linolenic acid (C18:3), eicosatrienoic acid (C20:3), eicosatetraenoic acid (C20:4)
and triolein as potential substrates. For the determination of the kinetic parameters of OhyRe
towards oleic acid, different substrate concentrations (90 µM to 1.44 mM) and reaction times (1
to 15 min) were tested under standard conditions. The reactions were stopped by the addition of
an equal volume of ethyl acetate (EtOAc) and instant, intensive vortexing. The resulting solvent
phase was separated from the water phase and applied for silylation and subsequent analysis.
Lipid analysis
The preparation of the extracted lipid fractions for GC measurements was performed according
[
23]
to Volkov et al. Hydroxylated fatty acids were identified by GC-MS by injecting 1 µl into a
Thermo Scientific™ TRACE™ Ultra Gas Chromatograph coupled to a Thermo DSQ™ II mass
spectrometer and the Triplus™ Autosampler injector. Column: Stabilwax ® fused silica capillary
(
30 m × 0.25 mm, film thickness 0.25 μm) Restek. (Program: initial column temperature 150 °C,
increasing (4 °C/min) up to a final temperature of 250 °C. Carrier gas: hydrogen (flow rate 3.5
ml/min). Peaks were identified by comparison to fatty acid standards or by specific molecular
masses detected. Extracts resulting from kinetic experiments were analyzed with the Shimadzu™
GC-2025 system equipped with a flame ionization detector. Column: Zebron ZB-WAX (30 m x
0.32 mm, film thickness 0,25 µm) Phenomenex. Carrier gas: hydrogen (3.00 ml/min). Program:
initial column temperature 150 °C for 1 min; increasing 5°C/min to 240°C, hold for 6 min. Peaks
were identified by comparison to the respective standards.
Crystallization
Crystals were obtained by the sitting-drop vapor-diffusion method at 18 °C with a reservoir
solution composed of 100 mM Bis-Tris/HCl pH 5.5 to pH 6.0, 255 mM to 300 mM magnesium
formate and 5 % (v/v) glycerol. Crystals were cryo-protected with 25% (v/v) MPD supplemented
to the reservoir resolution and subsequently flash-cooled in liquid nitrogen.
Diffraction data collection, structure determination and refinement
Synchrotron diffraction data were collected at the beamline 14.1 of the MX Joint Berlin laboratory
at BESSY (Berlin, Germany) or beamline P14 of Petra III (Deutsches Elektronen Synchrotron,
Hamburg, Germany). X-ray data collection was performed at 100 K. Diffraction data were
[
24]
processed with XDS
(Table S1). The structure was solved via molecular replacement in
PHASER [ by using the structure of the hydratase from Lactobacillus acidophilus (PDB 4ia5) [6]
25]
5
as a search model. Crystals of OhyRe belong to the space group P6 22, with two molecules in
the asymmetric unit. For the refinement 2.8% of the reflections were set aside for the calculation
[
26]
of the Rfree. Model building and water picking was performed with COOT. The structure was
initially refined by applying a simulated annealing protocol and in later refinement cycles by
[
27]
maximum-likelihood restrained refinement using PHENIX.refine. Model quality was evaluated
1
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
with MolProbity [28] and the JCSG validation server.
[29]
Secondary structure elements were
[
30]
[31]
assigned with DSSP
and ALSCRIPT
was used for secondary structure based sequence
[
32]
alignments. Figures were prepared using PyMOL.
Multi-angle light scattering (MALS)
MALS experiment was performed at 18°C. OhyRe was loaded onto a Superdex 200 increase
0/300 column (GE Healthcare) that was coupled to a miniDAWN TREOS three-angle light
1
scattering detector (Wyatt Technology) in combination with a RefractoMax520 refractive index
detector. For calculation of the molecular mass, protein concentrations were determined from the
differential refractive index with a specific refractive index increment (dn/dc) of 0.185 ml/g. Data
were analyzed with the ASTRA 6.1.4.25 software (Wyatt Technology).
Accession numbers
The atomic coordinates and structure factor amplitudes have been deposited in the Protein Data
Bank under the accession code 5ODO.
Acknowledgement
We gratefully acknowledge the financial support by the German Federal Ministry of Education
and Research (Project: Advanced Biomass Value, 03SF0446C). We also want to thank Martina
Haack for her excellent technical support in the analytics and Tom Schuffenhauer for his excellent
experimental support. We are grateful to C. Alings for excellent technical support. We are grateful
to M. Wahl for continuous encouragement and support. R. Janke is supported by the Elsa-
Neumann stipend of the state of Berlin and the Nüsslein-Volhard foundation. We acknowledge
beamtime and support at beamline P14 of PETRA III (Deutsches Elektronen Synchrotron,
Hamburg, Germany). We accessed beamlines of the BESSY II (Berliner Elektronenspeicherring-
Gesellschaft für Synchrotronstrahlung II) storage ring (Berlin, Germany) via the Joint Berlin MX-
Laboratory sponsored by the Helmholtz Zentrum Berlin für Materialien und Energie, the Freie
Universität Berlin, the Humboldt-Universität zu Berlin, the Max-Delbrück Centrum, and the
Leibniz-Institut für Molekulare Pharmakologie.
Keywords: Lyases • Oleate Hydratase • Fatty Acids • Hydration • Protein Structure
1
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
References
[
1]
a) D. L. Greenway, K. G. Dyke, Journal of General Microbiology 1979, 115, 233-245;
b) C. J. Zheng, J. S. Yoo, T. G. Lee, H. Y. Cho, Y. H. Kim, W. G. Kim, FEBS Lett 2005,
79, 5157-5162.
5
[
[
[
[
2]
3]
4]
5]
A. Hiseni, I. W. C. E. Arends, L. G. Otten, ChemCatChem 2015, 7, 29-37.
D. L. Greenway, K. G. Dyke, J Gen Microbiol 1979, 115, 233-245.
J. Jin, U. Hanefeld, Chem Commun (Camb) 2011, 47, 2502-2510.
a) L. E. Bevers, M. W. Pinkse, P. D. Verhaert, W. R. Hagen, J Bacteriol 2009, 191,
5010-5012;
b) E.-Y. Jeon, J.-H. Lee, K.-M. Yang, Y.-C. Joo, D.-K. Oh, J.-B. Park, Process
Biochemistry 2012, 47, 941-947.
[
[
6]
7]
A. Volkov, S. Khoshnevis, P. Neumann, C. Herrfurth, D. Wohlwend, R. Ficner, I.
Feussner, Acta Crystallogr D Biol Crystallogr 2013, 69, 648-657.
M. Engleder, T. Pavkov-Keller, A. Emmerstorfer, A. Hromic, S. Schrempf, G.
Steinkellner, T. Wriessnegger, E. Leitner, G. A. Strohmeier, I. Kaluzna, D. Mink, M.
Schurmann, S. Wallner, P. Macheroux, K. Gruber, H. Pichler, Chembiochem 2015, 16,
1730-1734.
[
[
[
8]
J. Schmid, L. Steiner, S. Fademrecht, J. Pleiss, K. B. Otte, B. Hauer, Journal of
Molecular Catalysis B: Enzymatic 2017.
Y. Wache, M. Aguedo, A. Choquet, I. L. Gatfield, J. M. Nicaud, J. M. Belin, Appl Environ
Microbiol 2001, 67, 5700-5704.
9]
10]
a) C. T. Hou, N Biotechnol 2009, 26, 2-10;
b) J. W. Song, E. Y. Jeon, D. H. Song, H. Y. Jang, U. T. Bornscheuer, D. K. Oh, J. B.
Park, Angew Chem Int Ed Engl 2013, 52, 2534-2537.
[
[
11]
12]
M. Goodfellow, G. Alderson, Journal of General MIcrobiology 1977, 100, 25.
a) B. R. Langdhal, P. Bisp, K. Ingvorsen, Microbiology 1996, 142;
b) S. C. Heald, P. F. B. Brandao, R. Hardicre, A. T. Bull, Antonie van Leeuwenhoek
2001, 80.
[
[
13]
14]
C. C. de Carvalho, M. M. da Fonseca, Appl Microbiol Biotechnol 2005, 67, 715-726.
H. Strnad, M. Patek, J. Fousek, J. Szokol, P. Ulbrich, J. Nesvera, V. Paces, C. Vlcek,
Genome Announc 2014, 2.
[
[
15]
16]
L. Holm, P. Rosenstrom, Nucleic Acids Res 2010, 38, W545-549.
L. De Colibus, M. Li, C. Binda, A. Lustig, D. E. Edmondson, A. Mattevi, Proc Natl Acad
Sci U S A 2005, 102, 12684-12689.
[
[
[
17]
18]
19]
B. N. Kim, Y. C. Joo, Y. S. Kim, K. R. Kim, D. K. Oh, Appl Microbiol Biotechnol 2012, 95,
9
29-937.
Y. C. Joo, E. S. Seo, Y. S. Kim, K. R. Kim, J. B. Park, D. K. Oh, J Biotechnol 2012, 158,
7-23.
Y. C. Joo, K. W. Jeong, S. J. Yeom, Y. S. Kim, Y. Kim, D. K. Oh, Biochimie 2012, 94,
07-915.
1
9
[
[
[
[
20]
21]
22]
23]
E. Krissinel, K. Henrick, J Mol Biol 2007, 372, 774-797.
J. Ortega-Anaya, A. Hernandez-Santoyo, Biochim Biophys Acta 2015, 1848, 3166-3174.
F. W. Studier, Protein Expr Purif 2005, 41, 207-234.
A. Volkov, A. Liavonchanka, O. Kamneva, T. Fiedler, C. Goebel, B. Kreikemeyer, I.
Feussner, J Biol Chem 2010, 285, 10353-10361.
[
[
24]
25]
W. Kabsch, Acta Crystallogr D Biol Crystallogr 2010, 66, 125-132.
A. J. McCoy, R. W. Grosse-Kunstleve, P. D. Adams, M. D. Winn, L. C. Storoni, R. J.
Read, Journal of applied crystallography 2007, 40, 658-674.
[
26]
P. Emsley, B. Lohkamp, W. G. Scott, K. Cowtan, Acta Crystallogr D Biol Crystallogr
2010, 66, 486-501.
1
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201701350
[
27]
a) P. D. Adams, P. V. Afonine, G. Bunkoczi, V. B. Chen, I. W. Davis, N. Echols, J. J.
Headd, L. W. Hung, G. J. Kapral, R. W. Grosse-Kunstleve, A. J. McCoy, N. W. Moriarty,
R. Oeffner, R. J. Read, D. C. Richardson, J. S. Richardson, T. C. Terwilliger, P. H.
Zwart, Acta Crystallogr D Biol Crystallogr 2010, 66, 213-221;
b) P. V. Afonine, R. W. Grosse-Kunstleve, N. Echols, J. J. Headd, N. W. Moriarty, M.
Mustyakimov, T. C. Terwilliger, A. Urzhumtsev, P. H. Zwart, P. D. Adams, Acta
Crystallogr D Biol Crystallogr 2012, 68, 352-367.
[
[
28]
29]
V. B. Chen, W. B. Arendall, 3rd, J. J. Headd, D. A. Keedy, R. M. Immormino, G. J.
Kapral, L. W. Murray, J. S. Richardson, D. C. Richardson, Acta Crystallogr D Biol
Crystallogr 2010, 66, 12-21.
H. Yang, V. Guranovic, S. Dutta, Z. Feng, H. M. Berman, J. D. Westbrook, Acta
Crystallogr D Biol Crystallogr 2004, 60, 1833-1839.
[
[
[
30]
31]
32]
W. Kabsch, C. Sander, Biopolymers 1983, 22, 2577-2637.
G. J. Barton, Protein Eng 1993, 6, 37-40.
W. DeLano, The PyMOL Molecular Graphics System http://www.pymol.org . 2002.
1
7
This article is protected by copyright. All rights reserved.