Exploring castellaniella defragrans linalool (De)hydratase-isomerase for enzymatic hydration of alkenes

Article abstract of DOI:10.3390/molecules24112092

Acyclic monoterpenes constitute a large and highly abundant class of secondary plant metabolites and are, therefore, attractive low-cost raw materials for the chemical industry. To date, numerous biocatalysts for their transformation are known, giving access to highly sought-after monoterpenoids. In view of the high selectivity associated with many of these reactions, the demand for enzymes generating commercially important target molecules is unabated. Here, linalool (de)hydratase-isomerase (Ldi, EC 4.2.1.127) from Castellaniella defragrans was examined for the regio- and stereoselective hydration of the acyclic monoterpene β-myrcene to (S)-(+)-linalool. Expression of the native enzyme in Escherichia coli allowed for identification of bottlenecks limiting enzyme activity, which were investigated by mutating selected residues implied in enzyme assembly and function. Combining these analyses with the recently published 3D structures of Ldi highlighted the precisely coordinated reduction–oxidation state of two cysteine pairs in correct oligomeric assembly and the catalytic mechanism, respectively. Subcellular targeting studies upon fusion of Ldi to different signal sequences revealed the significance of periplasmic localization of the mature enzyme in the heterologous expression host. This study provides biochemical and mechanistic insight into the hydration of β-myrcene, a nonfunctionalized terpene, and emphasizes its potential for access to scarcely available but commercially interesting tertiary alcohols.

Full text of DOI:10.3390/molecules24112092

molecules
Article
Exploring Castellaniella defragrans Linalool
De)hydratase-Isomerase for Enzymatic Hydration
of Alkenes
(
1
,2
3
3
3
3
Matthias Engleder
, Monika Müller , Iwona Kaluzna , Daniel Mink , Martin Schürmann ,
Erich Leitner 1
,4
, Harald Pichler
1,2
and Anita Emmerstorfer-Augustin
1,2,
*
1
Austrian Centre of Industrial Biotechnology, Petersgasse 14, 8010 Graz, Austria;
matthias.engleder@acib.at (M.E.); erich.leitner@tugraz.at (E.L.); harald.pichler@tugraz.at (H.P.)
Institute of Molecular Biotechnology, Graz University of Technology, NAWI Graz, BioTechMed Graz,
Petersgasse 14, 8010 Graz, Austria
InnoSyn B.V., Urmonderbaan 22, 6167 RD Geleen, The Netherlands; monika.muller@innosyn.com (M.M.);
iwona.kaluzna@innosyn.com (I.K.); daniel.mink@innosyn.com (D.M.);
martin.schuermann@innosyn.com (M.S.)
2
3
4
Institute of Analytical Chemistry and Food Chemistry, Graz University of Technology, NAWI Graz,
Stremayrgasse 9, 8010 Graz, Austria
*
Correspondence: emmerstorfer-augustin@tugraz.at; Tel.: +43-316-873-4082
ꢀ
ꢁꢂꢀꢃꢄꢅꢆꢇ
Academic Editor: Paola Vitale
Received: 28 April 2019; Accepted: 31 May 2019; Published: 1 June 2019
ꢀ
ꢁꢂꢃꢄꢅꢆ
Abstract: Acyclic monoterpenes constitute a large and highly abundant class of secondary plant
metabolites and are, therefore, attractive low-cost raw materials for the chemical industry. To date,
numerous biocatalysts for their transformation are known, giving access to highly sought-after
monoterpenoids. In view of the high selectivity associated with many of these reactions, the
demand for enzymes generating commercially important target molecules is unabated. Here, linalool
(
de)hydratase-isomerase (Ldi, EC 4.2.1.127) from Castellaniella defragrans was examined for the regio-
and stereoselective hydration of the acyclic monoterpene -myrcene to (S)-(+)-linalool. Expression
β
of the native enzyme in Escherichia coli allowed for identification of bottlenecks limiting enzyme
activity, which were investigated by mutating selected residues implied in enzyme assembly and
function. Combining these analyses with the recently published 3D structures of Ldi highlighted the
precisely coordinated reduction–oxidation state of two cysteine pairs in correct oligomeric assembly
and the catalytic mechanism, respectively. Subcellular targeting studies upon fusion of Ldi to different
signal sequences revealed the significance of periplasmic localization of the mature enzyme in the
heterologous expression host. This study provides biochemical and mechanistic insight into the
hydration of β-myrcene, a nonfunctionalized terpene, and emphasizes its potential for access to
scarcely available but commercially interesting tertiary alcohols.
Keywords: hydratase; dehydratase; isomerase; monoterpene hydration; Castellaniella defragrans;
myrcene; linalool; geraniol; tertiary alcohol
1
. Introduction
Asymmetric addition of water to monoterpenes is of huge interest, since it enables access to
high-value compounds from renewable, inexpensive starting materials [
dehydratase-isomerase (Ldi, EC 4.2.1.127) offers access to the desired tertiary monoterpene alcohol
S)-(+)-linalool by reversible (de)hydration of -myrcene (Scheme 1). (S)-(+)-linalool is isomerized to
1]. The bifunctional linalool
(
β
geraniol in a follow-up reaction catalyzed by the same enzyme [2]. The formation of (S)-(+)-linalool by
Molecules 2019, 24, 2092; doi:10.3390/molecules24112092
www.mdpi.com/journal/molecules

Molecules 2019, 24, 2092
2 of 14
Ldi is highly selective (e.e.
grant facilitated access to biotechnologically produced (S)-(+)-linalool [3,4].
≥
95%), and due to its poor commercial availability, the Ldi reaction may
Scheme 1. Ldi-catalyzed hydration of β-myrcene to (S)-(+)-linalool and its subsequent isomerization
to geraniol.
Ldi catalyzes the initial steps of monoterpene mineralization in the facultative anaerobic
β-proteobacterium Castellaniella defragrans strain 65Phen when grown under anaerobic conditions with
β
-myrcene as the sole carbon source [2]. In a proposed pathway, geraniol is then further metabolized
+
to geranial and geranic acid by NAD -dependent dehydrogenases, and introduced into
Since the thermodynamic equilibrium of the reactions favor isomerization of geraniol and dehydration
of (S)-(+)-linalool, Ldi may additionally confer detoxification of monoterpene alcohols in vivo ].
This was illustrated by 100- to even 1000-fold higher reaction rates for geraniol isomerization (Vmax of
β-oxidation.
[
2
−
1
−1
−1
−1
approx. 25
µ
mol min mg ) and (S)-(+)-linalool dehydration (Vmax of approx. 9
µ
mol min mg )
−1
−1
compared to the reverse reactions (Vmax of approx. 8 nmol min mg ). Optimal conditions for the
Ldi reaction were reported at pH 9.0 and 35 C, and enzyme activity was inhibited when concentrations
◦
of NaCl, KCl, or MgCl exceeded 0.2 M. To date, the linalool isomerase from Thauera linaloolentis 47
2
Lol is the only sequence with notable similarity to Ldi [
enzyme in the protein sequence space.
5], emphasizing the unique attributes of the
Mature Ldi is a periplasmic protein. It is translocated to the periplasm via SEC-dependent
membrane transport of the unfolded peptide mediated by an N-terminal signal sequence. Ldi is
sensitive towards molecular oxygen and requires a mild reducing agent such as DTT for full activity
in vitro. This suggests that the reduction–oxidation state of the four cysteines in the Ldi sequence is
important for enzyme function [
cytoplasm reduced which impairs disulfide formation, the oxidative environment in the bacterial
periplasm allows for assembly of disulfide bonds [ ]. In earlier work, Ldi was actively expressed in
E. coli with the N-terminal secretion signal [ ]. The crystal structure of Ldi from C. defragrans was
independently solved in the groups of Harder as well as Grogan and Hauer. In both cases, the enzyme
crystallized as toroidal pentamer, with each monomer showing an ( )6 barrel fold. The subunits,
either obtained in complex with -myrcene [ ] or geraniol [ ], are tightly joined by large interfacial
areas. The active site of Ldi is located at the interface of two subunits, which is unique among ( )6
2,6]. While both eukaryotic and prokaryotic organisms keep their
7
6,8
α,α
β
6
6,8
α,α
barrel proteins. The importance of cysteines for Ldi activity was highlighted by an essential disulfide
bond capping the substrate channel and contribution of two additional cysteines in the putative
reaction mechanisms (Figure 1) [6,8].
Both independent studies reported on acid/base catalysis for the (de)hydration and isomerization
reactions of Ldi, in which a carbocation intermediate is formed from (S)-(+)-linalool via
cysteine-catalyzed protonation, followed by rehydration for geraniol or deprotonation for
β
-myrcene [6,8]. Nestl and colleagues suggested a second mechanism that requires formation
of a covalent thioterpene intermediate from (S)-(+)-linalool by attack of the terminal alkene by one
of the active-site cysteines [ ]. Investigation of the proposed mechanisms with combined quantum
mechanics and molecular mechanics approaches revealed high relative Gibbs free energy barriers for
the reaction pathways suggested by Nestl et al., which rendered them rather unfavorable [ ]. In contrast,
8
9
the calculations indicated that the catalytic reaction may instead be initiated by aspartic-acid-mediated
0
protonation of C3 of
β-myrcene, giving an unstable carbocation intermediate that is subsequently
hydrated by activated water to form (S)-(+)-linalool. This stable intermediate may then isomerize
to geraniol by undergoing a concerted dehydration–hydration reaction conferred by an active-site
cysteine and activated water, respectively [9].

Molecules 2019, 24, 2092
3 of 14
Figure 1. Overall structure and a close-up view of the active site of Ldi (PDB code: 5G1U [6]). Bound
geraniol (purple) and the essential cysteines are highlighted in sticks. Cysteines 48 and 101 form a
disulfide bond that caps off the active-site cavity. Reduced cysteines 170 and 179 in close vicinity to a
monoterpenoid substrate participate in the catalytic mechanism.
In order to examine the biocatalytic potential of Ldi for hydration of monoterpenes, we compared
several vectors and E. coli strains for functional expression of the enzyme and used both whole cells
and isolated Ldi to stereoselectively convert β-myrcene to (S)-(+)-linalool. We investigated the role
of periplasmic translocation by determining the assembly of subunits in our expression system and
evaluated the function of the cysteines by site-directed mutagenesis. We found that a functional Ldi
oligomer was formed only in the periplasm, while cytosolic expression resulted in a distinctly different
assembly and complete loss of activity. Despite an unimpaired overall structure, this also applied for
variants harboring single amino acid exchanges of essential cysteines, thus confirming the essential
role of the thiols for Ldi structure and function.
2
. Results and Discussion
2
.1. Activity of Recombinant Ldi in E. coli
We initiated our investigations by evaluating different E. coli strains for recombinant expression of
Ldi. We used an Ldi nucleotide sequence harboring the N-terminal, native signal sequence in codon
optimized form, and compared different vector systems. The levels of recombinant protein in total cell
lysates (TCLs) and cell-free extracts (CFEs) were assessed via SDS-PAGE and immunoblot analysis
(Figure S2). Expression of Ldi varied notably among the tested strains and vectors. The amount of
soluble Ldi obtained with E. coli BL21star (DE3), E. coli Origami, and E. coli LEMO21(DE3) harboring a
pMS470 vector was too low for further use independent of the cultivation conditions, as the enzyme
was mostly present in inclusion bodies. In consequence, we decided to evaluate several modified E. coli
strains for potentially improved soluble expression of the enzyme (Table S7). Jointly, we exchanged
the vector system from pMS470 to pET26b(+). This strategy led to vastly improved amounts of Ldi
in CFEs (Figure S3). Due to the good yield of soluble enzyme as well as the relatively few additives
needed for cultivation, we selected E. coli BL21-CodonPlus(DE3)-RP in combination with the pET26b(+)
vector as the expression system for further work. In this setup, Ldi was expressed with its native
N-terminal signal sequence and was fused to a C-terminal His -tag. Most probably, the beneficial
10
effect of this strategy for Ldi expression was a combination of the rather tightly regulated induction
at a low concentration of 0.02 mM IPTG and the presence of extra copies of the argU and proL tRNA
genes. This may favor translation of heterologous proteins from organisms with a GC-rich genome,
such as C. defragrans (GC-content of 69%) [10,11]. Even though a codon-optimized version of Ldi was
used in our studies, overall GC content was still slightly increased (57%).
Focusing on the hydration reaction, functional expression of Ldi was confirmed by whole-cell
biotransformations of β-myrcene (Figure 2). For unambiguous identification of reaction products

Molecules 2019, 24, 2092
4 of 14
from β-myrcene by GC–MS, the retention times and mass fragmentation patterns obtained from
biotransformations were compared with an authentic standard of racemic linalool (Figure 2A). Only
the conversions with E. coli overexpressing Ldi resulted in additional peaks absent from the control
reactions with the empty vector control (EVC). Due to the formation of (S)-(+)-linalool and geraniol,
Ldi-catalyzed hydration and isomerization of monoterpenoids as well as expression of the enzyme
in its active form were confirmed. As reported previously [
2,4], biotransformation of β-myrcene
to (S)-(+)-linalool was highly stereoselective, since (R)-( )-linalool was not detected by chiral GC
−
−1
(
Figure 2B). In addition to the formation of approx. 600 ng
µL
of (S)-(+)-linalool in the organic
β-myrcene phase, small amounts of geraniol were generated during conversion (Figure 2C). This was
also in accordance with the earlier observation that geraniol is formed only after accumulation of a
certain concentration of (S)-(+)-linalool, as well as the supposed thermodynamic equilibrium of the
Ldi reaction on the side of geraniol isomerization and (S)-(+)-linalool dehydration, respectively [2].
A8.0E+06
1
1.52 min
13.57 min
7
6
5
4
3
2
1
0
.0E+06
.0E+06
.0E+06
.0E+06
.0E+06
.0E+06
.0E+06
.0E+00
11.52 min
7
1.1
93.1
11.52 min
69.1
55.1
13.57 min
4
1.1
8
0.1
121.1
93.1
5
3.1
123.1
3
5
65 95 125 155
m/z
35 65 95 125 155
m/z
1
1
11.5
12
12.5
13
13.5
14
14.5
15
Time / min
B3.0E+02
C
(
(
R)-(-)-linalool
S)-(+)-linalool
geraniol
7
6
5
.0E+02
2
2
1
1
5
0
.5E+02
.0E+02
.5E+02
.0E+02
.0E+01
.0E+00
.0E+02
.0E+02
2
5.94 min
4.0E+02
3.0E+02
2
9.85 min
2
1
0
.0E+02
.0E+02
.0E+00
2
5.75 min 25.94 min
2
9.85 min
E. coli E. coli LDI
EVC
2
5
25.5 26 26.5 27 27.5 28 28.5 29 29.5 30
Time / min
E. coli EVC E. coli Ldi
Figure 2. Whole-cell biotransformation of
activity. ( ) GC-MS of an authentic standard of racemic linalool standard (black) and a representative
sample from the biotransformation (green). ( ) Chiral GC-FID of authentic racemic linalool (black) and
geraniol (red) standards and a representative sample from the biotransformation (green). (R)-( )-linalool
and (S)-(+)-linalool were identified on basis of earlier publications [ ] which showed that only the
S-enantiomer of the tertiary alcohol is formed by Ldi. ( ) Quantification of -myrcene conversion to
S)-(+)-linalool and geraniol by an E. coli empty vector control (EVC) and E. coli cells expressing the
β-myrcene with E. coli expressing Ldi confirms enzyme
A
B
−
2,3
C
β
(
enzyme (E. coli Ldi).
2
.2. In Vitro Activity of Purified Ldi
After achieving proper expression of soluble Ldi in E. coli and verification of activity, C-terminally
His -tagged Ldi was purified from E. coli lysate with self-packed Ni-NTA affinity chromatography
10
columns (Figure 3A). Highly pure Ldi with an apparent molecular weight of approx. 45 kDa for the

Molecules 2019, 24, 2092
5 of 14
monomer was obtained via affinity tag purification. This matched well with the expected size of the
mature protein without the native periplasmic targeting sequence (41.8 kDa). Ldi was most likely
purified as a cofactor-free enzyme as indicated by the virtually colorless solution and the absence of
any chromophores absorbing between 300 and 800 nm in the UV-Vis absorption spectrum (Figure S4).
Directly after isolation, Ldi was subjected to activity tests with β-myrcene as the substrate (Figure 3B).
Assays with purified Ldi under anaerobic conditions led to stereoselective formation of (S)-(+)-linalool.
A clear difference between whole-cell and in vitro assays was the absence of detectable levels of
geraniol with purified Ldi, which may have resulted from slightly lower (S)-(+)-linalool formation
being prohibitive for the isomerization reaction [2].
B
A
1
2
3
4
5
6
7
8
9
10 11 12 13 14
(
(
R)-(-)-linalool
S)-(+)-linalool
7
6
5
4
3
2
1
00
00
00
00
00
00
00
0
Geraniol
5
0 kDa
0 kDa
4
Figure 3. Purified Ldi catalyzes the enantioselective formation of (S)-(+)-linalool from
) Purification of Ldi from E. coli lysate via Ni-NTA affinity chromatography. Four L of TCL and
CFE (lanes 2 and 3), 10 L of flow through (lane 4), washing fractions with 10 mM imidazole (lanes 5
β-myrcene.
(
A
µ
µ
to 10) and washing fractions with 50 mM imidazole (lanes 11 to 13), as well as 1 µL of purified and
TM
concentrated Ldi were loaded onto the gel. Novex
as the molecular weight marker (lane 1). (
purified Ldi.
Sharp Prestained Protein Standard was used
−
1
B) In vitro conversions of β-myrcene with 1 mg mL of
2
.3. Impact of Cysteine Exchange on Ldi Activity
Next, we focused on the role of the four cysteine residues in enzyme activity. We assessed whether
mutating them to alanines and serines causes inactivation and/or disrupts the overall structure of
the enzyme (Figure 1). We generated eight different single amino acid exchange variants of Ldi by
site-directed mutagenesis and transformed E. coli BL21-CodonPlus(DE3)-RP with pET26b(+) plasmids
harboring Ldi variants. Each variant was purified in parallel with Ldi wild type after confirming
expression by immunoblot analysis (Figure S5A). We did not observe obvious differences between
secondary structure elements of purified Ldi wild type and variant enzymes by CD spectroscopy, which
indicated that none of the exchanges markedly affected the overall structure (Figure S5B). Consistent
with previous data [6,7], site-directed mutagenesis of cysteines to alanines led to complete inactivation
of the enzyme, underlining function of the thiols in active-site capping and catalysis. Earlier studies did
not, however, investigate whether loss of Ldi activity was in fact a direct consequence of obstructing
disulfide formation by cysteines 48 and 101 and acid/base catalysis aided by cysteines 170 and 179 [6,8],
or rather caused by a disruption of the overall structure. Our results render the latter explanation highly
unlikely and emphasize the involvement of oxidized and reduced thiols in active-site architecture and
catalysis. Cysteine-supported acid/base catalysis is not unprecedented. For instance, catalytic cysteines
have been described in maleate isomerase and glutamate racemase reactions [12,13]. We also created
single amino acid exchange variants harboring serines instead of cysteines, and also ended up with
properly folded but inactive enzyme (Table S8). For Ldi C48S and C101S, this was most likely due to
elimination of the essential disulfide bond. In the case of Ldi C170S and C179S, we hypothesize that, in
contrast to the thiols of the cysteines, the alcohol side chains of the serines will not be ionized at or near

Molecules 2019, 24, 2092
6 of 14
neutral pH. This would prevent formation of a covalently bound terpenoid intermediate resulting from
nucleophilic attack on the substrate, and would object the proposed (de)hydration mechanism [8].
2
.4. Periplasmic and Cytosolic Targeting of Ldi
While our initial studies mainly addressed the expression and isolation of active Ldi in E. coli,
we were still confronted with an overall yield of less than 1% of (S)-(+)-linalool from
9 h of both in vitro and whole-cell reactions. Brodkorb and colleagues reported in their initial study
on detection and isolation of Ldi from C. defragrans that the thermodynamic equilibrium of the reaction
favors the formation of -myrcene from (S)-(+)-linalool, which would strictly limit the hydration
reaction [ ]. Also, poor solubility of -myrcene limits overall contact of enzyme and substrate,
whereas solubility of (S)-(+)-linalool is better, which favors the dehydration reaction. However,
several additional bottlenecks could limit -myrcene hydration by Ldi. Supposing that either enzyme
β-myrcene after
1
β
2
β
β
localization caused by periplasmic targeting, the limited expression level, or a combination of both
were the most critical factors impacting the Ldi reaction, we expressed Ldi constructs with different
N-terminal sequences in E. coli to investigate these assumptions in more detail. In addition to the
C-terminally His -tagged Ldi with its native signal sequence (Ldi), we designed an Ldi-His -tag
10
10
construct where the native N-terminal signal sequence was replaced by E. coli OmpA signal peptide [14
OmpA-Ldi, Table S2) and tested whether this setup positively influenced the level of active protein
in the periplasm. We also generated constructs lacking the N-terminal secretion signal but carrying
a His -tag either N- or C-terminally (His -nosig-Ldi and nosig-Ldi-His , Tables S3 and S4) to
]
(
10
10
10
promote cytosolic Ldi localization and, thereby, avoid periplasmic translocation as a potential limitation
for expression. The four different sequences were cloned into pET26b(+) and expressed in E. coli
BL21-CodonPlus(DE3)-RP under the conditions described in the Materials and Methods section.
Applying both N-terminal signal peptides, Ldi was predominantly translocated to the periplasm
(Figure S6). Removing the signal sequence markedly affected protein secretion, since nosig-Ldi was
mainly found in the cytosol. Comparing the hydration activity of different Ldi constructs in whole-cell
biotransformations demonstrated a remarkable discrepancy between activity and expression level
(Figure 4A,B).
Only Ldi constructs harboring an N-terminal signal peptide for periplasmic translocation were
able to generate notable levels of (S)-(+)-linalool and small amounts of geraniol regardless of the
low concentration of soluble protein (Figure 4B). OmpA-Ldi was expressed slightly better than Ldi,
which correlated with a tendency for higher product yield. Ldi activity was severely affected when
the enzyme was expressed without a signal peptide. C-terminally His -tagged nosig-Ldi was not
10
expressed at all and, consequently, no conversion of
β-myrcene was observed. We were able to
express nosig-Ldi by fusing an N-terminal His -tag to the enzyme at an exceptionally high yield,
10
but to our surprise, only ended up with traces of (S)-(+)-linalool. These results suggest several
direct consequences of the Ldi signal sequence and its removal from the protein. By comparing
His -nosig-Ldi with nosig-Ldi-His we concluded an overall stabilizing effect of the N-terminal
10
10
His -tag for the protein. Equal conclusions were drawn in earlier studies for other proteins [15,16].
10
Since fusion of an N-terminal tag to Ldi or OmpA-Ldi would have resulted in masking the secretion
signal for periplasmic localization, we refrained from expressing the corresponding Ldi constructs
in E. coli. The vast disproportion between Ldi and His -nosig-Ldi expression and activity indicated
10
that the amount of soluble protein per se was not pivotal for activity. Instead, based on our data,
we propose that the transport of the nascent polypeptide into the periplasm is much more decisive
for Ldi function. The increased activity of the enzyme in the periplasm may be mainly linked to two
factors. Firstly, Sec pathway-dependent translocation of Ldi [2] implies that the protein is translocated
in an unfolded state and adopts the native overall fold only in the periplasm [17]. Considering that Ldi
is a homopentamer in solution, this would be of particular importance for correct functional assembly
of the oligomeric protein [
6,8]. Secondly, as already discussed above, the Ldi 3D structure revealed that
the reduction–oxidation state of its four cysteine residues needs to be strictly controlled for enzyme

Molecules 2019, 24, 2092
7 of 14
assembly and activity [
6,
8]. Oxidized cysteines 48 and 101 trigger formation of an intrasubunit disulfide
bond for capping of the active-site cavity, while cysteines 170 and 179 must be kept in a reduced
state for all Ldi reactions. Essentially, the formation and isomerization of disulfides in E. coli are only
catalyzed in the periplasm by the Dsb protein family, which likely further accounts for the enhanced
enzyme activity after periplasmic translocation [18,19].
A
(S)-(+)-linalool
geraniol
6.0E+03
5.0E+03
4.0E+03
3.0E+03
2.0E+03
1.0E+03
0.0E+00
C-terminal His- C-terminal His- N-terminal His- C-terminal His-
EVC
L tda ig
Omp At a- gL di His -n toa sg ig-Ldi nosig- Lt ad gi -His
10 10
E. coli EVC
E. coli LDI
E. coli OmpA-
LDI
E.coli His10-
nosig-LDI
E. coli nosig-
LDI-His10
B
EVC
Ldi
OmpA-Ldi His -nosig-Ldi nosig-Ldi-His
10 10
TCL CFE
TCL CFE
TCL CFE TCL CFE
TCL CFE
Figure 4. Subcellular targeting impacts the expression and activity of recombinant Ldi in E. coli.
The enzyme was located to the periplasm by expressing Ldi constructs harboring either the native
N-terminal signal sequence (Ldi) or an N-terminal E. coli OmpA signal sequence (OmpA-Ldi). Cytosolic
enzyme localization was targeted with constructs lacking an N-terminal leader peptide (nosig-Ldi;
either N- or C-terminally His -tagged). (
A
) Whole-cell biotransformations of
β-myrcene with E. coli
10
expressing Ldi. (
B) Expression analysis of Ldi in total cell lysates (TCL) and cell-free extracts (CFE) of
an E. coli empty vector control (EVC), as well as in strains expressing Ldi, OmpA-Ldi, His10-nosig-Ldi,
and nosig-Ldi-His10 by immunoblotting.
2
.5. Impact of Cellular Localization on Ldi Assembly and Activity
To examine the presumed central role of Ldi translocation on its correct functional assembly,
we detected the oligomeric state of the different purified Ldi constructs via gel filtration. For Ldi
and OmpA-Ldi, three major enzyme forms were obtained (Figure 5A,B). Elution times were largely
independent of the Ldi construct investigated, except for one peak of the nosig-Ldi sample.

Molecules 2019, 24, 2092
8 of 14
Figure 5. An N-terminal periplasmic signal sequence provokes formation of distinct Ldi oligomers in
solution. Different oligomeric assemblies of His -tag purified Ldi ( ), OmpA-Ldi ( ), and nosig-Ldi
) in solution were separated by gel filtration. Relative molecular masses of major fractions were
A
B
1
0
(
C
determined on basis of a calibration with standard proteins.
The peak with a maximum at approx. 43 mL was corresponding to an Ldi multimer, whereas
the peaks with maxima at approx. 63 mL and 78 mL apparently corresponded to the elution of Ldi
dimers and monomers, respectively. For both Ldi and OmpA-Ldi, the dimer was the most abundant
form, followed by the monomer and the multimer. At approx. 54 mL, we expected to observe the peak
maximum of an Ldi homopentamer—the native form of the protein in solution [6,8]. Instead, only a
slight shoulder appeared at a roughly similar elution volume, indicating that the Ldi homopentamer
was not one of the predominant protein forms in our expression system. Apart from the Ldi multimer,
the elution profile of nosig-Ldi was characteristically different from the ones seen for Ldi and OmpA-Ldi
(
Figure 5C). Instead of Ldi dimers and monomers, we only encountered one additional distinct protein
form with a peak maximum at 87.2 mL, and observed only a small shoulder instead of the dimer at
approx. 64 mL. On the basis of our gel filtration calibration (Figure S1), we were not able to propose a
conclusive enzyme assembly to these two fractions.
−1
Upon activity tests with 1 mg mL of separated fractions, Ldi and OmpA-Ldi dimers were the
most active forms in solution (Table 1). The monomers showed a vastly reduced -myrcene hydration
β
activity (28% and 25% relative activity compared to Ldi and OmpA-Ldi dimers), and multimers were
only barely active in the case of Ldi, and not active at all in the case of OmpA-Ldi. We attribute the
apparent residual activity of the Ldi monomer fraction to an incomplete separation of the monomer
and dimer in gel filtration. In consequence, this implies that the activity detected for the monomer
was originating from residual Ldi dimer, whereas the monomer, in line with previous structural and
biochemical investigations, was unable to catalyze the reaction. In vitro, all tested nosig-Ldi fractions
were inactive. This was overall in good accordance with whole-cell biotransformations, where only
a slight residual activity of nosig-Ldi was detected in spite of the notably higher expression level
compared with Ldi and OmpA-Ldi (Figure 4).

Molecules 2019, 24, 2092
9 of 14
Table 1. The oligomeric assembly of Ldi subunits in solution strongly affects enzyme activity. Ldi gel
filtration fractions corresponding to the different oligomers in solution were pooled and applied in
-myrcene conversions for 19 h under anaerobic conditions. For each Ldi construct, the activities of
different assemblies were set relative to the oligomeric form with the highest -myrcene conversion.
Pooled enzyme fractions without any hydration activity are denoted as ‘ia’ (inactive).
β
β
Relative Molecular
Assembly in
Solution
Relative Activity for
β-myrcene Hydration [%]
Ldi Construct
Pooled Fraction
Mass/kDa
Multimer
>8 subunits)
Oligomer
2 to 4 subunints)
Monomer
3
7–46 mL
930–430
4
(
Ldi
57–70 mL
2–84 mL
168–55
47–17
100
28
(
7
Multimer
>8 subunits)
38–47 mL
854–395
ia
(
OmpA-Ldi
nosig-Ldi
Oligomer
2 to 4 subunints)
Monomer
5
7
8–70 mL
3–83 mL
154–55
43–18
100
25
(
Multimer
>8 subunits)
Inconclusive
Inconclusive
4
0–48 mL
720–363
ia
(
6
7
0–70 mL
4–83 mL
130–55
39–18
ia
ia
Based on the fact that we detected enzyme activity in our assays even without obtaining the
native pentameric structure, we propose that the minimum functional Ldi unit is a dimer. This is
supported by the 3D structure, which suggests that two subunits would, in principle, be sufficient
for (de)hydration and isomerization reactions at the interface of two monomers [6,8]. These results
back our earlier hypothesis that oligomerization in the periplasm is eminent for the correct functional
assembly of Ldi.
3
. Conclusions
The highly enantioselective hydration of
of a tertiary monoterpene alcohol that is only hardly accessible with other enzymes or organic
biotransformations [ ]. Characterization of this intriguing bifunctional (de)hydratase-isomerase has
β
-myrcene to (S)-(+)-linalool by Ldi permits production
3,4
opened up routes towards its improvement, but also revealed challenges for its biocatalytic application.
These can be primarily attributed to the low expression of functional Ldi and the low specific activity.
In this study, we tested functional expression of different Ldi constructs and amino acid exchange
variants in E. coli and determined activities in whole-cell biotransformations and in vitro with isolated
enzymes. Targeting studies demonstrated that a functional Ldi oligomer is assembled in the periplasm,
and while the native enzyme in solution is a homopentamer, the minimal functional unit is a dimer.
In addition to the appropriate functional assembly, periplasmic translocation is of utmost importance
for formation of the correct reduction–oxidation state of each of the four essential cysteines. Periplasmic
translocation is a common method for functional heterologous expression of disulfide-containing
proteins in E. coli, but it is often connected with low yields due to yet poorly understood reasons [20].
In Ldi, this is aggravated by the heterogeneity of oligomers, which further reduces the efficiency of
the E. coli expression system. The 1000-fold lower specific activity of Ldi for hydration of
compared to the dehydration of (S)-(+)-linalool additionally hinders the commercial production of the
more valuable monoterpene alcohol [ 21]. Further examination of whether Ldi is indeed limited by
β-myrcene
2
,
the thermodynamic equilibrium as hydration biocatalyst or can be developed by protein engineering
or by modifying the reaction conditions to become a viable biocatalyst for the sustainable valorization
of organic feedstock will follow.

Molecules 2019, 24, 2092
10 of 14
4
. Materials and Methods
4
.1. Chemicals and Media
Unless otherwise stated, all chemicals, solvents, and reference compounds were obtained from
®
Sigma-Aldrich (Steinheim, Germany) or Carl Roth GmbH & Co. KG (Karlsruhe, Germany) with the
highest purity available.
β
-myrcene (purity
≥
95%), racemic linalool (purity
≥
97%), (R)-(
−
)-linalool
®
(
purity 99%), and geraniol (purity
≥
≥
99%) were purchased from Sigma-Aldrich (Steinheim,
Germany). pEHISTEV [22] and pMS470 [23] vectors were employed for expression. Restriction
enzymes were acquired from Thermo Scientific (St. Leon-Rot, Germany). Sterile water was purchased
0
0
q
from Fresenius Kabi (Graz, Austria). For cloning and plasmid replication, E. coli Top10 F (F [lacI
R
Tn10(tet )] mcrA
∆
(mrr-hsdRMS-mcrBC)
φ
80lacZ
−
) from Life technologies (Vienna, Austria) was used. Recombinant Ldi
∆M15 ∆lacX74 deoR nupG recA1 araD139 ∆(ara-leu)7697
R
galU galK rpsL(Str ) endA1
λ
−
−
+
r
was expressed in E. coli BL21-CodonPlus(DE3)-RP (F-ompT hsdS(rB mB ) dcm Tet gal endA Hte [argU
r
proL Cam ]) (Life technologies, Vienna, Austria).
4
.2. Cloning and Site-Directed Mutagenesis
Standard molecular cloning technologies were used according to established procedures [24].
Correct integration of the insert was confirmed by sequencing (LGC Genomics, Berlin, Germany).
®
For gene amplification, Phusion High Fidelity DNA polymerase (Thermo Fisher Scientific
Inc., St. Leon-Rot, Germany) was utilized in accordance with the recommended PCR protocol.
Codon-optimized gene variants of Ldi (GenBank: CBW30776.1) harboring the native N-terminal signal
sequence (Ldi, Table S1) or, instead, an N-terminal E. coli OmpA signal sequence (OmpA-Ldi, Table S2)
were obtained from DSM Ahead R & D (Geleen, The Netherlands). N- and C-terminally His -tagged
10
Ldi constructs without an N-terminal signal sequence (nosig-Ldi, Tables S3 and S4) were generated by
PCR from an Ldi template using suitable primers with NdeI and HindIII restriction sites for cloning
into pET-26b(+). Synthetic His -tagged genes were amplified using primers with NdeI and HindIII
10
restriction sites (Table S5) and were cloned into a pET-26b(+) expression vector (Merck, Germany).
The resulting constructs were used to transform E. coli BL21-CodonPlus(DE3)-RP.
Single amino acid exchanges of cysteines to alanines or serines were introduced into the Ldi
coding sequence by site-directed mutagenesis (Table S6). The Stratagene QuikChange site-directed
mutagenesis protocol was slightly modified by preparing two separate PCR reactions containing
forward and reverse primers, respectively. After five cycling steps, PCR reactions were combined and
PCR was continued for 20 additional cycles according to the Stratagene manual. Mutated plasmids
were verified by DNA sequencing of the coding regions of the constructs.
4
.3. Expression of Recombinant Ldi in E. coli
Ldi constructs were expressed in E. coli BL21-CodonPlus(DE3)-RP. First, a preculture was inoculated
−
1
−1
in 10 mL LB supplemented with 15
µ
g mL of kanamycin and 20
µ
g mL of chloramphenicol and
◦
was cultivated at 25 C and 130 rpm overnight. Main cultures were inoculated in LB to an OD600 of 0.1
and protein expression was induced with 0.02 mM IPTG at an OD6 of 0.6–0.8. Recombinant Ldi was
00
◦
expressed for 14 h at 25 C and 130 rpm. Cells were harvested by centrifugation for 15 min at 4400
and 22 C and were used instantly for whole-cell biotransformations or were stored at
×
g
◦
◦
−
20 C until
protein purification.
4
.4. Whole-Cell Biotransformation of β-myrcene with E. coli
Bioconversions of -myrcene with E. coli were initiated immediately after protein expression.
β
Fifty OD600 units, which correspond to a cell dry weight (CDW) of 50 mg, were harvested and
®
resuspended in 1 mL of 20 mM sodium phosphate, pH 7.4, in Pyrex glass culture tubes. To initiate
the reactions, 100
µL of β-myrcene were added to the cell suspensions. Biotransformations were
◦
◦
incubated in technical triplicates for 19 h at 35 C and 150 rpm with the reaction tubes fixed in a 55

Molecules 2019, 24, 2092
11 of 14
angle relative to horizontal axis. Conversions were quenched by addition of 500
µL of ethyl acetate and
monoterpenoids were extracted for 30 min at 1500 rpm on a Vibrax VXR basic (IKA, Germany). After
◦
phase separation by spinning for 5 min at 2900
×
g and 22 C, terpenoids in 10
µ
L of organic phase were
verified by GC-MS and were quantified by solid-phase microextraction (SPME) coupled to GC-FID.
4
.5. Cell Lysis and Purification of Ldi
In order to purify recombinant Ldi from E. coli, cell pellets were resuspended in 50 mM sodium
phosphate, pH 8.0, containing 10 mM imidazole and disrupted by ultrasonication for 4 min with a
®
Sonifier 250 (Branson, Danbury, CT) at 80% duty cycle and output control 8. Cell-free extract (CFE)
◦
was separated from the total cell lysate (TCL) by centrifugation for 35 min at 48,300
was filtered through 0.22
×
g and 4 C and
µ
m filters (Millipore, Bedford, MA) prior to loading onto a pre-equilibrated
self-packed Ni-NTA affinity chromatography column (GE Healthcare, Little Chalfont, United Kingdom).
Nonspecifically bound protein was removed by washing the column with 7 column volumes of 50 mM
sodium phosphate, pH 8.0, containing 10 mM imidazole and 5 column volumes of the same buffer
containing 50 mM imidazole. His -tagged Ldi was eluted with 7 column volumes of the same
10
buffer containing 250 mM imidazole. The eluate was concentrated and buffer was exchanged to
0 mM Tris-HCl, pH 9.0, using disposable PD-10 desalting columns (GE Healthcare, United Kingdom)
8
according to the recommended protocol.
Different oligomeric forms of Ldi were separated by gel filtration on a Superdex 200 HiLoad 16/60
column (GE Healthcare, United Kingdom) using 80 mM Tris-HCl, pH 9.0, for elution. The apparent
relative molecular mass of native Ldi was determined by comparison of the size of eluted protein with
a calibration curve (Figure S1) using
44 kDa), myoglobin (17 kDa), and Vitamin B12 (1.35 kDa) from a commercially obtained gel filtration
standard set (Bio-Rad, Hercules, CA).
γ-globulin (158 kDa), bovine serum albumin (66 kDa), ovalbumin
(
4
.6. In Vitro β-myrcene Conversion
In vitro bioconversions were performed with purified Ldi. Before reactions, oxygen was removed
from protein solutions by incubation for 1 h in a glove box under anaerobic conditions. Degassed Ldi
was treated with either a 50-fold molar excess of DTT for 30 min or the same volume of ddH O in
2
−
1
®
controls. For conversions, 200
µ
L of
β
-myrcene were added to 400
µ
L of 1 mg mL Ldi in Pyrex glass
culture tubes. In vitro reactions were performed in technical triplicates for 19 h at 35 C and 300 rpm
in reaction tubes fixed in a 55 angle in a glove box for maintaining anaerobic reaction conditions.
◦
◦
Conversions were stopped, extracted, and prepared for SPME coupled to GC-FID as described for the
whole-cell biotransformations.
4
.7. Analysis of Monoterpenoids by GC-MS and GC-FID
Monoterpenoids in organic solvent were identified by comparison of retention times and mass
fragmentation patterns of analytes to authentic standards in gas chromatography–mass spectrometry
GC-MS). An HP-5 column (crosslinked 5% Ph-Me Siloxane; 30 m length, 0.25 mm in diameter, and
(
0
.25
µ
m film thickness) on a Hewlett-Packard 6890 Series II GC equipped with a 5973 mass selective
◦
detector was used. Sample aliquots of 1
temperature and 280 C interface temperature with He as carrier at a flow rate set to 0.93 mL min
µ
L were injected in split mode (split ratio 5:1) at 230 C injector
◦
−1
◦
(
35 cm/s linear velocity) in constant flow mode. The oven temperature program was as follows: 50 C
◦
◦
−1
◦
◦
−1
for 1 min, a ramp to 160 C at 10 C min , followed by a ramp to 250 C at 20 C min (1.5 min).
The total run time was 18 min. The mass selective detector was operated in a mass range of 35–300
amu at an electron multiplier voltage of 1671 V with a scan rate of 3.6 scans per second.
For separation of linalool enantiomers and quantification of monoterpenoids, a GC-FID
method preceded by SPME for extraction of volatiles was developed. Therefore, a chiral BGB-174
column (30% 2,3-diacetyl-6-tert-butyLdimethylsilyl-beta-cyclodextrin dissolved in BGB-1701 (14%
cyanopropylphenyl-, 86% methylpolysiloxane)) on a Hewlett-Packard 6890 GC equipped with a

Molecules 2019, 24, 2092
12 of 14
flame ionization detector (FID) was used. An automated SPME system was used for extraction and
enrichment of the analytes on a divinylbenzene-carboxen-polydimethylsiloxane fiber (50/30 m, 2 cm
µ
stable flex, Supelco, Bellfonte, PA, USA). A CTC Combi PAL sampler (CTC Analytics, Switzerland)
equipped with a single magnetic mixer (Chromtech, Germany) was used for sampling. Samples were
◦
equilibrated in the autosampler for 5 min at 60 C under stirring with a glass-coated magnetic stirrer.
◦
Extraction onto the SPME fiber was performed for 10 min at 60 C. Then, the fiber was transferred to
the GC injector for application of the sample onto the GC column. The oven program was as follows:
◦
◦
−1
◦
5
0 C for 1 min and 6 C min ramp to 240 C. The total run time was 36 min.
4
.8. Protein Analysis and Immunoblotting
Routinely, Ldi expression levels in E. coli cell lysate and the levels of purified Ldi were assessed by
SDS-PAGE or Western blot. SDS-PAGE was performed with the NuPAGE® SDS-PAGE Gel System
(
Life Technologies, Austria) according to the manufacturer’s recommendations. For analysis of proteins
via immunoblot, a primary anti-His antibody as well as an HRP-conjugated secondary antibody and
TM
enhanced chemiluminescent signal detection (SuperSignal , Pierce Biotechnology, Rockford, IL, USA)
were used for visualization of immunoreactive bands.
4
.9. Isolation of Recombinant Ldi from E. coli Periplasm and Cytosol
E. coli periplasm containing recombinant Ldi was isolated by preparing spheroplasts as described
previously [25]. Briefly, after protein expression, cells were harvested for 4 min at 10,000
were resuspended in an equal volume of 10 mM Tris-HCl, pH 8.0. After centrifugation for 4 min
at 10,000 g, cells were resuspended in an equal volume of spheroplasting buffer containing 20%
sucrose and 10% of 0.1 M EDTA in 30 mM Tris-HCl, pH 8.0. A freshly prepared lysozyme solution
×
g and
×
−
1
◦
was added to a concentration of 0.1 mg mL . The suspension was incubated for 30 min at 0 C
◦
before separation of periplasm from spheroplasts by spinning for 10 min at 10,000× g and 4 C. After
withdrawing the supernatant containing periplasmic Ldi, the cytosolic Ldi fraction was recovered by
TM
disrupting spheroplasts with the B-PER Bacterial Protein Extraction Reagent (Thermo Scientific,
Germany) according to the recommended protocol. Periplasmic and cytosolic Ldi was analyzed by
immunoblot analysis.
4
.10. Circular Dichroism Spectroscopy
Circular dichroism (CD) spectral measurements of Ldi wild-type enzyme and single amino acid
exchange variants were carried out with a Jasco J-810 spectropolarimeter (Jasco, MD). Far-UV CD
−
1
spectra were collected between 190 nm and 260 nm with a scan speed of 50 nm min in quartz cells of
◦
5
mm path length at 22 C. Protein concentrations were ranging from 6.3
µ
M to 13.4
µM in 80 mM
Tris-HCl, pH 8.0. Obtained values were normalized by subtracting the baseline recorded for the buffer.
2
−1
.
Data were expressed in mean residue ellipticity [Θ] in deg cm dmol
Supplementary Materials: The following are available online, Figure S1: Gel filtration calibration curve generated
with standard proteins listed in the Materials and Methods section; Figure S2: Initial expression studies of Ldi in
different E. coli strains indicated poor expression of soluble protein. Ldi was expressed from a pMS470 vector with
and without co-expression of DsbC. The level of Ldi in E. coli lysate was analyzed by immunoblotting; Figure S3:
Expression of Ldi in different E. coli strains increases the yield of soluble protein in specific set ups. Nine different
E. coli strains were tested for expression of Ldi, and after harvesting the cells, His -tagged Ldi was detected in
1
0
total cell lysates and cell free extracts by immunoblotting; Figure S4: UV-Vis absorption spectrum of purified
Ldi. The enzyme was His10-tag purified from cell free extract after expression in E. coli BL21-CodonPlus(DE3)-RP.
The lack of any noticeable absorbance in the range between 300 and 800 nm indicates that Ldi was purified as a
cofactor-free enzyme; Figure S5: An N-terminal signal sequence favors localization of Ldi in the E. coli periplasm.
After expression, E. coli periplasm (P) and cytosol (C) were separated as described in the Materials and Methods
section. Recombinant Ldi in isolated fractions was detected via immunoblotting. Controls were treated with
ddH O instead of the solutions inducing formation of spheroplasts; Figure S6: Comparison of circular dichroism
2
(
CD) spectra of Ldi wild type enzyme and variants confirms the overall integrity of single amino acid exchange
variants. Ldi wild type enzyme and variants were purified in parallel. After dilution to an A280 of approx. 0.8,
far-UV CD spectra were collected between 190 and 260 nm. Data are expressed in mean residue ellipticity [ ] in
Θ

Molecules 2019, 24, 2092
13 of 14
2
−1
deg cm dmol ; Table S1: Codon-optimized DNA sequence for expression of C-terminally His -tagged Ldi
1
0
in E. coli; Table S2: Codon-optimized DNA sequence for expression of C-terminally His -tagged OmpA-Ldi
1
0
in E. coli; Table S3: Codon-optimized DNA sequence for expression of N-terminally His -tagged nosig-Ldi
1
0
in E. coli; Table S4: Codon-optimized DNA sequence for expression of C-terminally His -tagged nosig-Ldi in
1
0
E. coli; Table S5: Primers used for amplification of Ldi constructs for cloning into pET26b(+) vector. Underlined
nucleotides are marking restriction sites; Table S6: Primers used for site-directed mutagenesis of Ldi cysteine
residues; Table S7: E. coli strains tested for expression of Ldi.
Author Contributions: Conceptualization, M.E., M.M., I.K., D.M., M.S., E.L., H.P., and A.E.-A.; Methodology,
M.E., M.M., I.K., D.M., M.S., E.L., H.P., and A.E.-A.; Validation, M.E., E.L., and A.E.-A.; Formal Analysis, E.L. and
A.E.-A.; Investigation, M.E., and A.E.-A.; Data Curation, M.E., and A.E.-A.; Writing—Original Draft Preparation,
M.E. and A.E.-A.; Writing—Review & Editing, M.E., M.M., I.K., D.M., M.S., E.L., H.P., and A.E.-A.; Visualization,
M.E., and A.E.-A.; Project Administration, H.P., D.M., M.S., and A.E.-A.; Funding Acquisition, D.M., M.S., and
H.P.; Supervision, H.P., M.S., and A.E.-A.
Funding: This research was funded by the Austrian Science Fund (FWF) project J3787 (A.E.-A.). M.E., H.P.,
and A.E.-A. acknowledge the support of this work by the Federal Ministry of Science, Research and Economy
(BMWFW), the Federal Ministry of Traffic, Innovation and Technology (bmvit), the Styrian Business Promotion
Agency SFG, the Standortargentur Tirol, the Government of Lower Austria and Business Agency Vienna through
the COMET-Funding Program managed by the Austrian Research Promotion Agency FFG. The APC was funded
by the Austrian Science Fund (FWF).
Acknowledgments: We thank Thomas Netscher for helpful discussions and careful revision of this paper
manuscript, as well as Eva Baumhackl for excellent technical assistance.
Conflicts of Interest: The authors declare to have no conflict of interest. Martin Schürmann, Daniel Mink, Iwona
Kaluzna, and Monika Müller have company affiliations (InnoSyn B.V., Geleen, The Netherlands). This does not
alter the authors’ adherence to all Molecules policies on sharing data and materials.
References
1
.
Bicas, J.L.; Dionísio, A.P.; Pastore, G.M. Bio-oxidation of terpenes: An approach for the flavor Industry.
Chem. Rev. 2009, 109, 4518–4531. [CrossRef] [PubMed]
2
.
Brodkorb, D.; Gottschall, M.; Marmulla, R.; Lüddeke, F.; Harder, J. Linalool dehydratase-isomerase, a
bifunctional enzyme in the anaerobic degradation of monoterpenes. J. Biol. Chem. 2010, 285, 30436–30442.
[CrossRef]
3.
4.
5.
Demming, R.M.; Fischer, M.-P.; Schmid, J.; Hauer, B. (De)hydratases-recent developments and future
perspectives. Curr. Opin. Chem. Biol. 2018, 43, 43–50. [CrossRef] [PubMed]
Lüddeke, F.; Harder, J. Enantiospecific (S)-(+)-Linalool Formation from
Dehydratase-Isomerase. Z. Naturforsch. C 2011, 66, 409–412.
β-Myrcene by Linalool
Marmulla, R.; Šafari c´ , B.; Markert, S.; Schweder, T.; Harder, J. Linalool isomerase, a membrane-anchored
enzyme in the anaerobic monoterpene degradation in Thauera linaloolentis 47Lol. BMC Biochem. 2016, 17,
1
–11. [CrossRef]
6.
7.
8.
Weidenweber, S.; Marmulla, R.; Ermler, U.; Harder, J. X-ray structure of linalool dehydratase/isomerase from
Castellaniella defragrans reveals enzymatic alkene synthesis. FEBS Lett. 2016, 590, 1375–1383. [CrossRef]
De Marco, A. Strategies for successful recombinant expression of disulfide bond-dependent proteins in
Escherichia coli. Microb. Cell Fact. 2009, 8, 1–18. [CrossRef]
Nestl, B.M.; Geinitz, C.; Popa, S.; Rizek, S.; Haselbeck, R.J.; Stephen, R.; Noble, M.A.; Fischer, M.P.; Ralph, E.C.;
Hau, H.T.; et al. Structural and functional insights into asymmetric enzymatic dehydration of alkenols.
Nat. Chem. Biol. 2017, 13, 275–281. [CrossRef] [PubMed]
9
.
Ling, B.; Wang, X.; Su, H.; Liu, R.; Liu, Y. Protonation state and fine structure of the active site
determine the reactivity of dehydratase: Hydration and isomerization of
dehydratase/isomerase from Castellaniella defragrans. Phys. Chem. Chem. Phys. 2018, 20, 17342–17352.
CrossRef]
0. Oliver, J.L.; Mar
16–223. [CrossRef]
1. Petasch, J.; Disch, E.M.; Markert, S.; Becher, D.; Schweder, T.; Hüttel, B.; Reinhardt, R.; Harder, J.
The oxygen-independent metabolism of cyclic monoterpenes in Castellaniella defragrans 65Phen.
BMC Microbiol. 2014, 14, 1–13. [CrossRef] [PubMed]
β-myrcene catalyzed by linalool
[
1
1
ín, A. A relationship between GC content and coding-sequence length. J. Mol. Evol. 1996, 43,
2

Molecules 2019, 24, 2092
14 of 14
1
2. Fisch, F.; Fleites, C.M.; Delenne, M.; Baudendistel, N.; Hauer, B.; Turkenburg, J.P.; Hart, S.; Bruce, N.C.;
Grogan, G. A covalent succinylcysteine-like intermediate in the enzyme-catalyzed transformation of maleate
to fumarate by maleate isomerase. J. Am. Chem. Soc. 2010, 132, 11455–11457. [CrossRef] [PubMed]
3. Glavas, S.; Tanner, M.E. Active site residues of glutamate racemase. Biochemistry 2001, 40, 6199–6204.
1
1
1
1
[CrossRef]
4. Movva, N.R.; Nakamura, K.; Inouye, M. Amino acid sequence of the signal peptide of ompA protein, a major
outer membrane protein of Escherichia coli. J. Biol. Chem. 1980, 255, 27–29.
5. Gräslund, S.; Nordlund, P.; Weigelt, J.; Hallberg, B.M.; Bray, J.; Gileadi, O.; Knapp, S.; Oppermann, U.;
Arrowsmith, C.; Hui, R.; et al. Protein production and purification. Nat. Methods 2008, 5, 135–146.
6. Booth, W.T.; Schlachter, C.R.; Pote, S.; Ussin, N.; Mank, N.J.; Klapper, V.; Offermann, L.R.; Tang, C.;
Hurlburt, B.K.; Chruszcz, M. Impact of an N-terminal Polyhistidine Tag on Protein Thermal Stability.
ACS Omega 2018, 3, 760–768. [CrossRef]
1
1
1
2
7. Tsirigotaki, A.; De Geyter, J.; Šoštari c´ , N.; Economou, A.; Karamanou, S. Protein export through the bacterial
Sec pathway. Nat. Rev. Microbiol. 2017, 15, 21–36. [CrossRef] [PubMed]
8. Nakamoto, H.; Bardwell, J.C.A. Catalysis of disulfide bond formation and isomerization in the Escherichia
coli periplasm. Biochim. Biophys. Acta Mol. Cell Res. 2004, 1694, 111–119. [CrossRef]
9. Mamathambika, B.S.; Bardwell, J.C. Disulfide-Linked Protein Folding Pathways. Annu. Rev. Cell Dev. Biol.
2
008, 24, 211–235. [CrossRef]
0. Schlegel, S.; Rujas, E.; Ytterberg, A.J.; Zubarev, R.A.; Luirink, J.; de Gier, J.W. Optimizing heterologous protein
production in the periplasm of E. coli by regulating gene expression levels. Microb. Cell. Fact. 2013, 12, 1–11.
[
CrossRef]
1. Engleder, M.; Pichler, H. On the current role of hydratases in biocatalysis. Appl. Microbiol. Biotechnol. 2018
02, 5841–5858. [CrossRef] [PubMed]
2. Liu, H.; Naismith, J.H. A simple and efficient expression and purification system using two newly constructed
vectors. Protein Expr. Purif. 2009, 63, 102–111. [CrossRef] [PubMed]
3. Balzer, D.; Ziegelin, G.; Pansegrau, W.; Kruft, V.; Lanka, E. KorB protein of promiscuous plasmid RP4
recognizes inverted sequence repetitions in regions essential for conjugative plasmid transfer. Nucleic
Acids Res. 1992, 20, 1851–1858. [CrossRef] [PubMed]
2
2
2
,
1
2
4. Ausubel, M.; Brent, R.; Kingston, R.E.; Moore, D.D.; Seidman, J.G.; Smith, J.A.; Struhl, K. Current Protocols in
Molecular Biology; Wiley-Liss, Inc.: New York, NY, USA, 1987.
2
5. Ueguchi, C.; Ito, K. Escherichia coli sec mutants accumulate a processed immature form of maltose-binding
protein (MBP), a late-phase intermediate in MBP export. J. Bacteriol. 1990, 172, 5643–5649. [CrossRef]
[PubMed]
Sample Availability: Samples of the compounds are not available from the authors.
©
2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).