Stabilization of Hydroxynitrile Lyases from Two Variants of Passion Fruit, Passiflora edulis Sims and Passiflora edulis Forma flavicarpa, by C-Terminal Truncation

Article abstract of DOI:10.1002/cbic.201900468

Because the synthesis of chiral compounds generally requires a broad range of substrate specificity and stable enzymes, screening for better enzymes and/or improvement of enzyme properties through molecular approaches is necessary for sustainable industri

Full text of DOI:10.1002/cbic.201900468

Accepted Article
Title: Stabilization of hydroxynitrile lyases from two variants of passion
fruits, Passiflora edulis Sims and Passiflora edulis forma
flavicarpa by C-terminal truncation
Authors: Aem Nuylert, Fumihiro Motojima, Chartchai Khanongnuch,
Tipparat Hongpattarakere, and Yasuhisa Asano
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: ChemBioChem 10.1002/cbic.201900468
Link to VoR: http://dx.doi.org/10.1002/cbic.201900468
A Journal of
www.chembiochem.org

ChemBioChem
10.1002/cbic.201900468
1
Stabilization of hydroxynitrile lyases from two variants of passion
fruits, Passiflora edulis Sims and Passiflora edulis forma
flavicarpa by C-terminal truncation
[
a, b]
[a, b]
[c]
[d]
Aem Nuylert,
Fumihiro Motojima,
Chartchai Khanongnuch, Tipparat Hongpattarakere, and
[
a, b]
Yasuhisa Asano*
Abstract: As the synthesis of chiral compounds generally requires a
broad range of substrate specificity and stable enzymes, screening
for better enzymes and/or improvement of enzyme properties using
molecular approaches is necessary for the sustainable industrial
development. Here, we report the discovery of unique hydroxynitrile
lyases (HNLs) from two species of passion fruits, Passiflora
edulis forma flavicarpa (yellow passion fruit, PeHNL-Ny) and
Passiflora edulis Sims (purple passion fruit, PeHNL-Np) isolated and
purified from passion fruit leaves. These are the smallest HNLs
Np, we speculated that protein engineering around the only one
glycosylation site, Asn105 located at the C-terminal region of
PeHNL-Ny might contribute to stabilization of PeHNL. Therefore, we
focused on the improvement of stability of the non-glycosylated
PeHNL by truncating its C-terminal region. The C-terminal truncated
PeHNLΔ107 was obtained by truncating 15 amino acids from its C-
terminal and was expressed in Escherichia coli. PeHNLΔ107
expressed in E. coli was not glycosylated, and showed improved
thermostability, solvent stability, and reusability similar to the wild-
type glycosylated form of PeHNL expressed in Pichia pastoris.
These data reveal that the lack of the high flexibility region at the C-
terminal of PeHNL might be the possible reason for improving the
stability of PeHNL.
(comprising 121 amino acids). Amino acid sequences of both
enzymes are 99% identical; there is a difference of one amino acid
in a consensus sequence. Purple passion fruit PeHNL-Np has Ala
residue at position 107 and is non-glycosylated at Asn105. As it was
confirmed that natural and glycosylated PeHNL-Ny showed superior
thermostability, pH stability, and organic tolerance than the PeHNL-
and characterized from cyanogenic plants, and they include the (R)-
[
4]
Introduction
selective ones derived from Prunus amygdalus (PaHNL), Prunus
[
5]
[6]
serotina Ehrh. (PsHNL),
Prunus mume (PmHNL),
Prunus
[
7]
[8]
armeniaca L. (ParsHNL), Prunus communis (PcHNL), Linum
Hydroxynitrile lyases (HNLs, EC 4.1.2.10, 4.1.2.11, 4.1.2.46, and
[
9]
[10]
usitatissimum (LuHNL),
Phlebodium aureum (PhaHNL),
4
.1.2.47) act during the final step of the cyanohydrin biodegradation
[
1]
[11]
[12]
pathway. In cyanogenic plants, HNL plays a role in the defense
mechanism of plants by acting decomposition of stored cyanogenic
glycosides into aldehydes and hydrogen cyanide for protection
against microbes and herbivores. Correspondingly, it was
discovered that an arthropod (millipede) synthesizes mandelonitrile
as a chemical for defense in order to produce hydrogen cyanide by
Arabidopsis thaliana (AtHNL),
Eriobotrya japonica (EjHNL),
[
13]
Passiflora edulis f. flavicarpa (PeHNL),
Davallia tyermannii
[
14]
[15]
(DtHNL),
and Nandina domestica (NdHNL),
and the (S)-
[
16]
selective ones derived from Hevea brasiliensis (HbHNL), Manihot
[
17]
[18]
esculenta (MeHNL),
Baliospermum montanum (BmHNL).
Sorghum bicolor (SbHNL), and
Among the HNLs of plant
[
19]
[
18]
[12]
[13c]
[20]
[8]
[
2]
origin, SbHNL,
EjHNL,
PeHNL,
PaHNL,
PcHNL, and
catalysis of a HNL. As the reversible reaction of HNLs, the enzyme
also catalyzes the synthesis of optically pure cyanohydrins which are
important synthetic intermediates for fine chemical and
[
21]
PmHNL
are glycosylated proteins, whereas, LuHNL, BmHNL,
AtHNL, NdHNL, HbHNL, MeHNL, and PhaHNL are not glycosylated.
Furthermore, EjHNL, PaHNL PcHNL, and PmHNL contain flavin
adenine dinucleotide (FAD), in contrast to the LuHNL, PhaHNL,
AtHNL, PeHNL, BmHNL, HbHNL, MeHNL, SbHNL, and NdHNL
which do not contain FAD. Moreover, the cupin fold-containing HNLs
derived have also been discovered from bacteria, Granulicella
[
3]
pharmaceutical industries. Mainly, HNLs have been discovered
[
a]
A. Nuylert, Dr. F. Motojima, Prof. Dr. Y. Asano
Biotechnology Research Center
[
22]
[23]
tundricola (GtHNL) and Acidobacterium capsulatum (AcHNL).
Deparment of Biotechnology, Toyama Prefectural University
5
180 Kurokawa, Imizu, Toyama 939-0398 (Japan)
Two major passion fruit sub-species, Passiflora edulis Sims,
which yields purple fruits and Passiflora edulis f. flavicarpa, which
gives yellow ones are used in commercial passion fruit products
such as concentrated juice, jam, nectar, etc. Whether the yellow
passion fruit is a mutant of the purple variety P. edulis or its hybrid
E-mail: asano@pu-toyama.ac.jp
A. Nuylert, Dr. F. Motojima, Prof. Dr. Y. Asano
Asano Active Enzyme Molecule Project, ERATO, JST
[b]
[c]
[d]
5180 Kurokawa, Imizu, Toyama 939-0398 (Japan)
Asst. Prof. Dr. C. Khanongnuch
[
24]
Division of Biotechnology, School of Agro-Industry, Faculty of Agro-
Industry, Chiang Mai University, Chiang Mai, 50100, Thailand
Assoc. Prof. Dr. T. Hongpattarakere
Department of Industrial Biotechnology, Faculty of Agro-Industry,
Prince of Songkla University, Hat Yai, Songkhla, 90112, Thailand
with other species is not elucidated. Both species contain large
quantities of the cyanogenic compound, especially prunasin
[
25]
representing 80 % of cyanogenic glycosides. First, the HNL from
Passiflora edulis f. flavicarpa (PeHNL-Ny) was isolated and purified
[
13a, 13b]
in our laboratory,
and its cDNA was successfully cloned into E.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
2
[
13c]
coli and P. pastoris.
Glycosylated PeHNL-Ny showed high
Np levels (2.42 U/g leaf sample) detected in leaves were
approximately 3.3-fold less than in P. edulis f. flavicarpa (8.13 U/ g
leaf sample. PeHNL-Np was also purified to near homogeneity using
several rounds of column chromatography as summarized in Table
S1. The specific activity for the synthesis of (R)-mandelonitrile from
benzaldehyde and potassium cyanide (KCN) was 120 U/mg, which
was marginally lower than that from PeHNL-Ny (136 U/mg). PeHNL-
Np has a molecular mass of 14 kDa, same as PeHNL-Ny, as
revealed by SDS-PAGE; moreover, they have 99% identity. At
position 107, PeHNL-Np from purple passion fruit has an Ala residue,
whereas that from yellow passion fruit has a Thr residue (Figure S2).
Therefore, the purple passion fruit PeHNL cannot be glycosylated on
the nitrogen atom of the Asn residue of nascent polypeptide in the
stability at low pH and in organic solvents that are always required
for cyanohydrin synthesis for industrial use. Therefore, we expected
that P. edulis Sims might also be a valuable resource for discovering
new HNLs.
We recently reported the X-ray crystallographic analysis of
PeHNL-Ny, and found that PeHNL belongs to the dimeric α+β barrel
(
DABB) superfamily consisting of a central β-barrel in the middle of
the dimer, and the high flexibility region was observed in the rich
hydrophobic residues at a residue 107 which lies beyond the C-
[
26]
terminal.
We also found that the wild-type glycosylated form of
natural PeHNL and the one produced in P. pastoris PeHNL both of
which have one glycosylation site at Asn105 located near the C-
terminus, showed significantly higher stability than the enzyme
without glycosylation produced in E. coli which lack the glycosylation
[
28]
recognition tripeptide sequence Asn-X-Ser/Thr.
Periodic acid
Schiff (PAS) staining of the purified PeHNL-Np confirmed that it is
not a glycoprotein (Figure S1). Using various approaches such as
the use of N-glycosylation and N-glycan-processing inhibitors and
site-directed mutagenesis of N-glycosylation site, glycans in plant
glycoproteins have long been shown to play a major role in the
folding, processing, and secretion of proteins from the endoplasmic
[
13c]
system.
Because of the drawback of the time-consuming process
of production of glycosylated PeHNL in P. pastoris, we aimed to
improve the stability of PeHNL expressed in E. coli using a targeted
mutagenesis approach. The flexible regions are potential targets and
thus have been selected for molecular engineering to improve
[
27]
[29]
stability of the enzyme.
This region can be evaluated with the
reticulum (ER) and the Golgi apparatus. For example, treatment of
highest B-factor, value determined on the basis of atomic
a
suspension-cultured carrot (Daucus carota) cells with tunicamycin,
displacement obtained from X-ray data.
an antibiotic which inhibits Asn-linked glycosylation, yielded
In this study, we looked for a novel PeHNL from purple P.
edulis cultivated in Thailand, purified the enzyme and cloned its
cDNA for expression of PeHNL-Np, and discovered one mutation
among the C-terminal region which affects the stability of PeHNLs
upon comparison of its structure and activity. Taken together, the
finding of the natural diversity of PeHNLs among the sub-species of
passion fruits, and our previous results obtained using molecular
engineering of PeHNL-Ny, we speculated that the presence of the
glycan at the C-terminal might significantly contribute to the stability
of PeHNLs. We improved the activity of PeHNL by reducing the
length of the enzyme at the C-terminus using site-directed
mutagenesis. The engineered enzymes were purified, and their
improved properties were characterized and compared with those of
the wild-type enzymes. The substrate specificity of the synthesized
enzymes and the enantiomeric excess of the cyanohydrins
synthesized in biphasic systems were determined.
unglycosylated
carrot
cell-wall
β-fructosidase,
and
the
unglycosylated enzyme was easily degraded during the last stages
in the secretory pathway or immediately after its arrival in the cell
[
30]
wall.
The N-glycan is also known to protect the protein from
proteolytic degradation and is responsible for thermostability,
solubility, and biological activity of concanavalin A (Con A), the lectin
[
31]
from jack bean seeds.
The amount of glycosylated PeHNL-Ny in the whole leaves of
the yellow passion fruit was 3.3-fold higher per mg of the total leaf
protein than that of non-glycosylated PeHNL-Np in the leaves of
purple passion fruits. It seems likely that the presence of glycan
affected the transport of PeHNL in the yellow passion fruit cells. The
yellow passion fruit grows faster and has a greater resistance to soil
[
24]
fungi, whereas the purple one is more resistant to cold injury.
Thus, yellow and purple passion fruits may have been adopted to
their environment and evolved to control the amounts of HNLs by
glycosylation, according to their survival strategies.
Results and Discussion
Biochemical properties of two PeHNLs
To compare enzyme kinetics of two natural PeHNLs, the rate of
synthesis (R)-mandelonitrile from benzaldehyde and KCN was
measured using HPLC-chiral column analyses. The Vmax, kcat, and
PeHNL is present in two species of passion fruits
The HNL from P. edulis f. flavicarpa (yellow passion fruit, PeHNL-
Ny) has been discovered, characterized, and gene coding for it was
successfully cloned and expressed to produce the recombinant
-
1
-1
k
m
cat/K values of glycosylated PeHNL-Ny were 220 µmol min mg ,
-
1
-1
-1
5
0.4 s , and 3.70 mM s , respectively, and these values were
-
1
[
13]
higher than those of non-glycosylated PeHNL-Np (150 µmol min
PeHNL-Ny.
Recently, we solved the PeHNL-Ny structure and
-
1
-1
-1 -1
[
26]
mg , 35.7 s , and 1.86 mM s , respectively,) (Table 1.). The K
PeHNL-Ny (13.7 mM) was slightly lower than that of PeHNL-Np
19.2 mM), which corresponds with our previous report that the
glycan on PeHNL enhances the catalytic efficiency without reducing
m
of
proposed its catalytic mechanism. Here, we report the discovery
of HNL from another commercial passion fruit species, P. edulis SIM
(
(purple passion fruit, PeHNL-Np), which is cultivated on a large scale
in Thailand. A large amount of P. edulis SIM leaves (5000 g) was
collected at a plantation in Chaing-Mai, Thailand. The HNL activity
from the leaf extract was determined by (R)-mandelonitrile ((R)-
MAN) synthesis using HPLC equipped with a chiral column. PeHNL-
[
13c]
the substrate affinity for benzaldehyde and KCN.
Also, we proved
that those kinetic parameters of natural PeHNL were caused by the
presence of glycan.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
3
Table 1. Kinetic parameters of native and C-truncated mutant PeHNLs.
Enzyme
Spec. act.
V
max
-1
K
m
k
cat
-1
kcat/K
m
-
1
-1
-1 -1
(
U mg )
136
120
110
122
134
167
118
112
60
(µmol min mg )
220±0.50
150±3.20
145±3.27
168±2.37
228±0.67
270±0.80
167±0.37
158±0.40
103±0.20
n.d.
(mM)
(s )
(mM s )
3.70±0.01
1.86±0.01
1.33±0.01
1.40±0.32
2.30±0.52
2.53±0.30
1.50±0.05
1.40±0.06
0.73±0.01
n.d.
PeHNL-Ny
PeHNL-Np
PeHNL121p
PeHNL121y
PeHNLΔ117
PeHNLΔ112
PeHNLΔ109
PeHNLΔ107
PeHNLΔ105
PeHNLΔ102
13.7±0.01
19.2±0.03
25.2±0.46
26.3±0.14
23.4±0.14
25.0±0.16
26.1±0.11
26.6±0.13
33.0±0.12
n.d.
50.4±0.01
35.7±0.04
33.6±0.50
36.4±0.50
53.4±0.16
63.1±0.20
39.0±0.10
37.2±0.10
24.1±0.04
n.d.
n.d.
The purified PeHNLs were prepared from various sources; Ny and Np were purified from natural yellow and purple passion fruit leaves, respectively; the subscript
number 121p and 121y represented to the whole amino acid sequences of purple and yellow passion fruit, respectively, which were expressed and purified from
recombinant in E. coli; the C-truncated mutant PeHNLs (PeHNLΔ) with a subscript number corresponds to the total PeHNL length after truncated at C-terminal. The
kinetic parameters were obtained by monitoring synthesis of (R)-MAN by an HPLC at 254 nm with conducting experiments in triplicates. n.d. = not determined
We determined the kinetic parameters using both purified
product by the two natural PeHNLs were more than 95% at low
temperatures between 5°C to 20°C, and low ee products were
obtained when the temperature was increased, due to the
acceleration of the non-enzymatic reaction. In addition, the activity of
PeHNL-Ny remained above 80% after 1 h incubation at 60°C (Figure.
1D), while PeHNL-Np retained only 40% of activity under the same
condition. Obviously, the thermostability was significantly improved
by glycosylation in PeHNL-Ny. Furthermore, glycosylated PeHNL-Ny
was more stable in the presence of an organic solvent in a biphasic
system than non-glycosylated PeHNL-Np toward all the organic
solvents tested (Figure. S3). This evidence supports that the
glycosylated PeHNL is more stable than non-glycosylated
recombinant PeHNLs lacking glycosylation expressed in E. coli. The
m m
K , Vmax, kcat, and kcat/K values of PeHNL121y were estimated to be
-
1
-1
-1
-1
-1
2
6.3 mM, 168 µmol min mg , 36.4 s , and 1.85 mM s ,
respectively, and these values were similar to those of PeHNL121p
-
1
-1
-1
-1
-1
(
25.2 mM, 145 µmol min mg , 33.6 s , and 1.33 mM s ,
respectively) (Table 1). Thus, the difference of one amino acid at
position 107 of PeHNL appears not to affect the affinity and catalytic
efficiency of PeHNL.
The effect of pH and temperature on the synthesis (R)-
mandelonitrile was studied using the purified natural PeHNLs.
PeHNL-Ny and PeHNL-Np exhibited the highest (R)-MAN synthetic
activity at pH 4.0 with >95% ee (Figure. 1A). The ee of 99% (R)-
MAN was obtained at pH 2.5-3.5 for both PeHNLs, but it obviously
decreased at pH higher than 4.5 due to a rapidly increasing the
spontaneous non-enzymatic reaction yielding the racemic of both
enantiomers. The pH stability analyses revealed that both natural
PeHNLs were broadly stable over the range of 3.5 to 10 after
preincubation at 30°C for 1 h without substrate (Figure. 1C).
The optimum temperature range of both PeHNLs for the (R)-
MAN synthetic activity was 20 to 25 °C (Figure. 1B). The ee of the
[
13c]
PeHNL.
Recently, we solved the structure of PeHNL and
discussed that a glycosylation site on Asn105 of PeHNL might
stabilize the high flexibility region at C-terminus, which could improve
[
26]
PeHNL stability. Thus, we hypothesized that reduction in the high
flexibility region of PeHNL by the deletion of part of C-terminal region
increases the stability of non-glycosylated PeHNL.
Figure 1. The comparison of biochemical properties of two PeHNLs; PeHNL-Ny (yellow line) and PeHNL-Np (purple line) and rac-mandelonitrile formed by non-enzymatic
reaction (black line). A) Effect of pH and B) temperature on initial velocity (solid line) and enantiomeric excess (dashed line). C) pH stability and D) temperature stability
after incubation of enzymes at various pH or temperatures for 1 h. Remaining activity was estimated by monitoring the synthesis of (R)-mandelonitrile from benzaldehyde
and potassium cyanide using HPLC with a chiral column. Data are presented as mean ± SD (n = 3)
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
4
further study the roles of the C-terminal loop of PeHNL, truncation
mutants were constructed lacking the chosen amino acid residues
from this region. Five to 18 residues (4.1 to 14% of the total PeHNL
length) were deleted from the C-terminus. Up to 17 amino acid
residues of PeHNL could be removed (PeHNLΔ105
) without
abolishing its activity. A 6xHis tag was attached to the N-terminus of
truncation mutants and they were heterologously expressed in E.
coli BL21 (DE3) cells. The yields of all C-terminal truncation mutants
were similar, except PeHNLΔ105, which showed the lowest PeHNL
production. The truncation mutants were purified using Ni-sepharose
6
Fast Flow and a Mono Q 5/50 column. The purified PeHNLΔ117,
PeHNLΔ112, PeHNLΔ109, PeHNLΔ107, and PeHNLΔ105 had molecular
masses of approximately 14.9, 14.4, 14.3, 14.2, and 14.2 kDa,
respectively, in SDS-PAGE analyses (Figure S4). The molecular
masses of these mutants as determined by high-performance gel
chromatography on Superdex 200 10/30 were 27.6, 26.8, 26.8, 26.6,
and 26.4 kDa, respectively, indicating that the native enzyme was
active as a homodimer.
Analysis of specific activity and kinetic parameters of C-
terminal truncation mutants
Figure 2. Flexibility in PeHNL. A) A B-factor putty drawing of wild-type PeHNL
(
PDB ID:5XZQ) structure in the nucleocapsid-like particle. Regions colored red
The C-terminal truncation mutants affected the mandelonitrile
synthesis activity of PeHNL as shown in Table 1. The specific
also have a large diameter, which corresponds to higher B factors and are
associated with higher structural flexibility. A consensus sequence (N105) for N-
glycosylation site on PeHNL is shown as cyan spheres. B) Root mean square
deviation (RMSD) between the crystal structure of wild-type and PeHNLΔ107
-
1
activity of PeHNLΔ117 and PeHNLΔ112 were 134 U mg and 167 U
-1
mg , respectively, which were marginally higher than that of wild-
(
PDB ID: 5XZT).
type PeHNL121. In contrast, the specific activity of PeHNLΔ105 was 60
-
1
U mg , significantly (2-fold) lower than that of wild-type PeHNL121
;
PeHNLΔ102 activity could not be detected. These results indicated
that the 15 amino acid deletion from C-terminus did not affect
PeHNL activity. Next, we determined kinetic parameters of the C-
terminal truncation mutants for the synthesis of mandelonitrile from
Identification of flexible regions in PeHNL
The secondary structure of wild-type PeHNL (PDB ID:5XZQ)
consists mainly of four antiparallel β-sheets (β1, 6-14 aa; β2, 44-47
aa; β3, 62-68 aa; β4, 94-103 aa), three α-helices (α1, 20-36 aa; α2,
m
benzaldehyde and KCN. The Vmax, kcat, and kcat/K of PeHNLΔ117 and
7
1-79 aa; α3, 81-93 aa), and a special loop at the C-terminal that
mainly contains hydrophobic amino acid residues
104INETFPYTWALPNKYVVT121). During structural analysis, we
-
1
-1
PeHNLΔ112 were 228±0.67 and 270±0.80 µmol min mg , 53.4±0.16
-
1
-1 -1
[
26]
and 63.1±0.20 s , 2.30±0.52 and 2.53±0.30 mM s , respectively,
(
missed the C-terminus region (residues 110-121) because of the
poor electron density, indicating that this region has high flexibility
[
26]
and does not form a stable structure.
Generally, the B-factor,
which can evaluate the flexibility of an individual residue in a protein,
is used for explaining the behavior of atomic electron densities from
[
27a]
X-ray data.
The available C-terminal region of wild-type PeHNL
crystal structure showed remarkably high B-factor values at the
residues E106, T107, F108, and P109, which were 90, 70.5, 72.6,
and 82.5, respectively (Figure. 2A). In agreement with the root mean
squared deviation (RMSD) at the C-terminal region of the similarity
between two superimposed atomic coordinates of PeHNL wild-type
[26]
and PeHNLΔ107 (PDB ID: 5XZT)
structure also showed a high
RMSD, with an average value above 1.281Å (Figure. 2B). The
RMSD value of this region is much higher than that of other regions,
indicating that the C-terminus region is highly flexible, which might
be responsible for the instability of PeHNL.
Expression and purification of C-terminal truncation
mutants
Figure 3. The stability profile of wild-type and C-terminal truncated PeHNL variants. A)
Stability at low pH. After incubation in reaction mixtures with varying pH values for 1 h, the
remaining activity was measured as described in Figure 1. B) Solvent stability, the remaining
activities of PeHNL mutants were measured after incubation in each biphasic system in a
citrate buffer (200 μL)/organic solvent (various solvent phase content) at 10 °C for 12 h using
The deletion of unnecessary amino acid or domains at the N- or C-
terminus of some enzymes have been used for improving the
enzymatic properties and/or enzyme stability, such as endo-β-1,4-
HPLC with
a chiral column. Dashed bar: PeHNL121y. Open bar: PeHNLΔ117. Gray bar:
PeHNLΔ112. Cross bar: PeHNLΔ109. Solid bar: PeHNLΔ107. Dot bar: PeHNLΔ105. Data are
presented as means ± SD (n = 3)
[
32]
[33]
[34]
[35]
glucanase, amylopullulanase, keratinase, and amylase. To
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
5
while the Vmax, kcat, and kcat/K
m
of other C-terminal truncation mutants
Previously, we found that diisopropyl ether (DIPE) is a suitable
organic solvent for cyanohydrin synthesis by non-glycosylated
PeHNL expressed in E. coli when used in a biphasic system;
however, the wild-type non-glycosylated PeHNL expressed in E. coli
were similar to those of the wild-type, except for PeHNLΔ105, which
had the lowest values. The K values of wild-type, PeHNLΔ117
PeHNLΔ112, PeHNLΔ109, and PeHNLΔ107 were 26.3±0.14, 23.4±0.14,
5.0±0.16, 26.1±0.11, and 26.6±0.13 mM, respectively (Table 1).
m
,
[
13c]
2
has low stability in an organic solvent.
In this study, we compared
Thus, truncation of 5-15 amino acid residues from C terminal did not
the solvent stability of the wild-type and C-terminus truncated
mutants of PeHNL in a biphasic system of buffer and different
concentration of DIPE. The stabilities of the variants were monitored
after incubation at 10 °C for 12 h with shaking at 1,500 rpm and the
residual activity of enzyme were compared with the initial activity in
the aqueous phase saturated with organic solvents before incubation.
The highest solvent stability was shown by PeHNLΔ107 in all
concentrations of DIPE (20−80%), with more than 90% residual
affect the affinity of PeHNL for benzaldehyde.
C-terminal truncation mutant PeHNLΔ107 improved PeHNL
stability
To prove that the flexible region at C-terminus of PeHNL contributes
to enzyme stability, we further investigated the resistance of the wild-
type enzyme and truncation mutants to low pH, temperature, and
organic solvent. A low pH is necessary for the enantioselective
synthesis of cyanohydrins because of the spontaneous unselective
addition of HCN to the carbonyl group occurs as a background
activity. On the other hand, activity of the wild-type, PeHNLΔ117
,
PeHNLΔ112, and PeHNLΔ105 enzymes reduced to 90% to 50% as the
DIPE concentration increased from 20% to 50% v/v, and these
enzymes were almost inactivate at a high DIPE concentration of
70% to 80% v/v (Figure 4B). These results demonstrated that
deleting 15 amino acid from the C-terminus (PeHNLΔ107) improved
the thermostability and solvent stability of PeHNL.
[
6, 36]
reaction at elevated pH.
We analyzed the stability of C-terminal
truncated mutants of PeHNL at low pH (3.0 to 5.0). All PeHNLs were
stable over the pH range tested (Figure 3A). Although stability under
most pH condition was comparable, a marginal difference in the
stability was observed among the PeHNLs. Wild-type, PeHNLΔ117,
and PeHNLΔ107 showed > 95% remaining activity at low pH, which
was slightly higher than the remaining activities of PeHNLΔ112, and
PeHNLΔ109 (90%) under same pH conditions for 1 h incubation. It
appears that the deletion of 5-17 amino acid residues from C-
terminus does not increase the stability of PeHNL under the pH
range tested.
PeHNLΔ107 catalyzes the synthesis of cyanohydrins
We tried to improve the stability of non-glycosylated PeHNL
expressed in E. coli by truncation of its C-terminus, it showed
potential for cyanohydrin synthesis in a biphasic system with an
organic solvent. It has been reported that biphasic systems could
improve the enantiomeric excess of the cyanohydrin product and
minimized the non-enzymatic cyanation reaction upon extraction and
separation of an aldehyde into the organic phase in a biphasic
Thermostability analyses revealed that the partial deletion of 5
to 13 amino acid (PeHNLΔ117, PeHNLΔ112, and PeHNLΔ109) from C-
terminus decreased the thermostability at 45°C compared with that
of wild-type for 1 h preincubation. Surprisingly, only PeHNLΔ107
exhibited >80% and >60% remaining activity at 50°C and 55°C,
respectively, which was highest for 1 h preincubation (Figure 4A).
PeHNLΔ107 also showed >40% remaining activity after 2 h incubation
at 50°C, in contrast to the wild-type and other C-terminal truncated
variants that were rapidly inactivated (Figure 4B). These results
suggest that the deletion of high flexibility region of 15 amino acid at
C-terminal likely contributes to the thermostability of PeHNL.
[
13b, 13c, 37]
system.
In this study, we synthesized cyanohydrins in the
presence of either the wild-type form PeHNL or PeHNLΔ107 (10 U) in
biphasic systems containing several aldehydes at high concentration
(250 mM) in DIPE (0.75 mL) and an equal volume of KCN (300 mM)
in citrate buffer (400 mM, pH 3.5) used as a substrate. PeHNL
showed a preference for several aromatic aldehydes other than
bulky aromatic aldehydes such as 3-methoxybenzaldehyde (1f), 1-
naphthaldehyde (1i) and 2-
Figure 4. A) Thermostability at different temperatures, B) thermostability at 50 °C of C-terminal truncated variants of PeHNL121y (◊), PeHNLΔ117 (□), PeHNLΔ112 (Δ), PeHNLΔ109
×), PeHNLΔ107 (ж) and PeHNLΔ105 (○). Activity of the enzyme was detected by incubating the enzyme for 1 h at different temperatures and then the remaining activity was
(
measured as described in Figure 1. Data are presented as means ± SD (n = 3)
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
6
Table 2. Conversion and enantioselectivity of purified wild type PeHNL121 and C-truncated
PeHNLΔ107 toward different aldehydes.
naphthaldehyde (1j) which proved to be poor
substrates for PeHNL. The maximum
conversion and more than 99% ee were
obtained in less than 30 min for the synthesis
of (2S)-thiophen-2-yl-hydroxyacetonitrile (2g).
Compared to the wild-type PeHNL121y
,
1a:Ph; 1b:2-Me-Ph; 1c:3-Me-Ph; 1d:2-Cl-Ph; 1e:3-Cl-Ph; 1f:3-MeO-Ph; 1g:2-thiophene;
1h:2-furan; 1i:1-naphthalene; 1j:2-naphthalene
PeHNLΔ107 showed good ee (>90%) and the
over 90% conversion after 3 h, which are
higher than that of the wild-type PeHNL121y
for the synthesis of (R)-mandelonitrile (2a),
Sub.
1[R=]
t=0.5 h
Conv. [%]
t=3 h
Conv. [%]
t=24 h
Conv. [%]
ee [%]
95.5
97.1
92.0
93.5
53.3
67.5
76.6
95.1
2.5
99.8
22.0
30.5
95.5
99.1
90.3
93.6
16.4
18.3
n.d.
ee [%]
90.7
93.4
82.7
91.0
30.1
58.1
60.0
90.7
2.5
99.4
9.10
26.0
90.7
96.4
70.0
78.0
4.8
ee [%]
60.5
75
1a
1b
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
PeHNL121
PeHNLΔ107
72.2
76.4
48.3
55.5
28.0
29.5
60.8
69.1
46.5
54.2
30.7
32.6
82.2
86.4
52.4
54.0
30.3
30.6
n.d.
87.7
92.2
81.2
91.1
46.0
59.0
85.6
93.7
50.4
89.0
72.5
72.2
97.7
99.2
56.5
58.0
34.4
34.2
n.d.
90.5
94.1
95.6
96.5
87.4
90.3
99.5
99.5
97.2
97.4
96.6
96.6
n.t.
(
R)-2-methylmandelonitrile
(2b),
and
(R)-2-
(R)-3-
chloromandolonitrile (2d)
65.4
78.8
15.6
38.5
45.5
90.5
7.23
98.9
5.70
16.8
n.t.
chloromandolonitrile (2e). The substrate
specificity toward aldehyde or substituted
benzaldehydes for PeHNL is the major
intermediate cyanohydrins for the preparation
of bioactive drugs such as duloxetine (2g,
used for treatment of depression and general
1c
1d
1e
[
38]
anxiety
disorder),
semi-synthetic
1f
[
39]
cephalosporins (2a, antibacterial agents),
[
40]
and clopidogrel (2d, anticoagulant).
Thus,
1
g
h
PeHNL is a good potential biocatalyst that
n.t.
n.t.
can be used for industrial application.
1
64.1
66.8
60.1
60.1
n.d.
26.0
17.0
1.6
3.4
n.d.
n.d.
Reusability of PeHNLΔ107
1i
1j
The reusability of PeHNLΔ107 has been
investigated for the production of (R)-MAN
synthesis in biphasic batch reaction systems.
Wild-type PeHNL and PeHNLΔ107 showed no
significant differences in the production of
13.6
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d. = not determined; n.t. = not tested
(R)-MAN in the range of 50-55 µM for the first
cycle of reaction. After four cycles in the reusability experiment, the Conclusion
production of (R)-MAN of wild-type PeHNL was significantly
decreased in the DIPE biphasic system, whereas the production of
PeHNLΔ107 was almost the same as that in the first cycle of the
reaction (Figure 5A, 5B). These results are in agreement with the
high solvent stability of PeHNLΔ107 in contrast to that of the wild-type
form which lost more than 50 % activity upon incubation in a
biphasic system for 6 h (Figure S5). Thus, the C-terminal truncated
PeHNLΔ107 without glycosylation which was expressed in E. coli
could improve the stability and reusability similar to that shown by
The natural and glycosylated form of PeHNL from the yellow passion
fruit showed much higher thermostability, pH stability, and the
tolerance to organic solvents than PeHNL from the purple passion
fruit, indicating the important role of the glycosylation. The glycan
present at one glycosylation site located on a high flexibility C-
terminal region might stabilize the structure and thus reduce the
structural flexibility of PeHNL. This hypothesis has been proved by
generation of a C-terminal truncated form of PeHNL lacking 15
amino acid residues, and then expressing it in E. coli resulting in an
enzyme lacking in glycosylation. The non-glycosylated and C-
[
13c]
the glycosylated wild-type PeHNL expressed in P. pastoris.
Figure 5. Reusability of A) PeHNL121, B) PeHNLΔ107 during the production of (R)-mandelonitrile from benzaldehyde and potassium
cyanide. ○: Amount of (R)-mandelonitrile. ▲: Enantiomeric excess [%].
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
7
[
42]
terminal truncated mutant PeHNLΔ107 could improve the stability and
reusability, as effectively as the glycosylated natural PeHNL and
expressed in P. pastoris.
Periodic acid-Schiff (PAS)
staining using a GelCode Glycoprotein
Staining Kit (Thermo Fisher Scientific) according to the manufacturer`s
instructions.
Cloning of PeHNL-Np cDNA
Experimental Section
Total RNA was isolated from a young leaf of P. edulis Sims after 3
months of cultivation and reverse transcribed. As purified PeHNL-Np
showed a molecular mass of 14000 Da which is almost the same as
PeHNL-Ny, we first tried to identify the PeHNL-Np gene from P. edulis
Sims using degenerate and gene-specific primers as previously
reported. . DNA sequences were determined using a 3500 Genetic
Analyzer (Thermo Fisher Scientific) and assembled and analyzed using
ATGC and Genetyx Ver. 12 (Genetyx, Tokyo, Japan), respectively.
Materials and chemicals
Leaves of Passiflora edulis Sims and Passiflora edulis forma flavicarpa
were collected from the Thung Roeng Royal Project Development Center
in Chiang Mai, Thailand, and the Botanic Garden of Toyama, Japan,
respectively, and stored at -20 ºC. Benzaldehyde (redistilled, 99.5 %)
and (rac)-mandelonitrile were purchased from Sigma-Aldrich. Other
chemicals used in the experiments were purchased from chemical
sources and used without further purification.
[13c]
B-factor and root mean squared deviation analysis
The B-factor of wild-type PeHNL (PDB ID:5XZQ) was extracted from the
Assays to determine the synthetic activity of HNL
[
27a]
pdb structure using the B-FITTER software.
The root mean squared
deviation (RMSD) of the similarity between the structures of the two
PeHNLs, wild-type PeHNL and PeHNLΔ107 (PDB ID: 5XZT) have been
calculated by the Molecular Operating Environment (MOE, Chemical
Computing Group Inc Canada).
Enzymatic activity was quantified by determining (R)-mandelonitrile
synthesis using benzaldehyde and KCN as substrates following our
[
13a]
previous study.
The reaction was initiated by adding KCN (100 mM) in
citrate buffer (pH 4.0, 400 mM) into the reaction mixture (0.5 mL)
containing benzaldehyde (50 mM; 20 µL of 1250 mM benzaldehyde
dissolved in dimethyl sulfoxide (DMSO)) and enzyme sample (0.5-5 U
Bacterial strain and construction of the C-terminal truncated PeHNL
-
1
mL ) which was incubated at 25 ºC for 5 min. The reaction was stopped
by transferring aliquots (100 µL) of the reaction mixture to 900 µL of n-
hexane/propan-2-ol mixture (85:15), mixed vigorously, and centrifuged at
E. coli JM 109 (Takara Bio) and BL21 (DE3) (Thermo Fisher Scientific)
cells were used as hosts for amplification of plasmid DNAs and
recombinant protein expression, respectively. PeHNL from P.
edulis forma flavicarpa was cloned and expressed using the pColdI
15000 × g for 5 min at 4 ºC. The organic layers containing benzaldehyde,
(
R)- and (S)-mandelonitrile were analyzed using chiral HPLC as
[
13c]
[
13c]
expression vector (pColdI-PeHNL) as previously described.
Wild-type
described in our previous study.
One unit of HNL activity was defined
pColdI-PeHNL was truncated at the C-terminus of each variant using
QuikChange® Lightning Mutagenesis Kit (Agilent Technologies, USA)
and the truncation primers shown in Table S2. PCR products were
as the amount of enzyme that produced 1 µmol of (R)-mandelonitrile
from benzaldehyde and KCN per minute.
treated with 10
U of DpnI at 37ꢀ°C for 1 h, and then used for
Purification of PeHNLs from leaves
transformation of E. coli JM 109 using the heat shock method.
Sequences were confirmed using the 3500 Genetic Analyzer.
The PeHNL from P. edulis forma flavicarpa was purified as described
[
13c]
previously . Here we describe a procedure to purify PeHNL-Np. Briefly,
P. edulis Sims leaves were frozen using liquid nitrogen and ground with
mortar and pestle. Extraction and protein fractionation with ammonium
Gene expression, purification, and determination of molecular mass of
wild-type and C-terminal truncated PeHNL
[
13c]
sulfate was performed in the same manner as previously reported.
E. coli BL21(DE3) cells harboring recombinant plasmids were cultured in
After dialysis of the active ammonium sulfate fraction, proteins from the
extract were loaded onto the Toyopearl DEAE -650M column (Tosoh, 50
mm i.d., column volume 150 mL) and eluted with a linear gradient of
sodium chloride from 0 to 500 mM in potassium phosphate buffer (KPB,
pH 6.0, 20 mM). The active fractions were added ammonium sulfate
−
1
Luria–Bertani (LB, 5 mL) broth containing ampicillin (50 μg ml ) and
incubated overnight at 37 °C with shaking at 300 rpm. The pre-culture
starter was transferred into LB broth (500 mL) containing ampicillin
−
1
(
50 μg ml ) in an Erlenmeyer flask (2 L) and cultured at 37 °C with
shaking at 150 rpm until A600 reached 0.6, followed by a cold-shock at
6 °C. After 2 h, isopropyl-β-D-thiogalactopyranoside (IPTG) was added
(30 % saturated) and loaded onto a Butyl-Toyopearl column (Tosoh, 25
1
mm i.d., column volume 50 mL). Bound proteins were eluted stepwise
with ammonium sulfate (1330, 860, 420, and 0 mM) in KPB (pH 6.0, 20
to a final concentration of 1.0 mM, and cells were cultured at 16 °C for 24
h with the same rate of shaking. Protein extraction and purification were
mM). Active fractions were combined, dialyzed thrice against KPB (pH
[
13c]
TM
nd
performed exactly as described previously.
Superdex 10/300 GL
6
.0, 20 mM) and then loaded onto the 2 Toyopearl DEAE-650M column
(GE Healthcare) was used to calculate the molecular mass of the purified
(25 mm i.d., column volume 20 mL). Bound protein was pre-washed with
recombinant wild-type PeHNL and its truncated form based on their
relative mobility as compared with values for standard marker proteins as
follows: glutamate dehydrogenase (290,000 Da), lactate dehydrogenase
5
0 mM sodium chloride in KPB (pH 6.0, 20 mM) and eluted with a linear
gradient of sodium chloride from 50 to 100 mM in the same buffer. Active
fractions were pooled, concentrated, desalted by using a centrifugal
filtration device (Amicon Ultra-15, 10000 NMWL, EMD Millipore) and then
loaded onto a Mono Q 5/50 GL column (1 mL, GE Healthcare). Proteins
were then eluted with a linear gradient of sodium chloride from 0 to 100
mM in KPB (pH 6.0, 20 mM) at a flow rate of 0.8 mL.min . Active
fractions were pooled, concentrated and further purified using the
Superdex 200 10/300 GL column (GE Healthcare). Purified PeHNL-Np
was monitored using sodium dodecyl sulfate-polyacrylamide gel
(142,000 Da), enolase (67,000 Da), adenylatekinase (32,000 Da), and
cytochrome c (12,400 Da).
-
1
Measurement of kinetic parameters
Kinetic parameters of purified wild-type and C-terminal truncation
mutants of PeHNL were assayed by monitoring synthesis of (R)-
mandelonitrile in sodium citrate buffer (pH 4.0, 400 mM), according to the
[
41]
electrophoresis (SDS-PAGE)
and glycosylation was detected using
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
8
method described above, at various concentrations (0.5-100 mM) of
benzaldehyde (followed by mixing with total 20 µL of 12.5-2500 mM
benzaldehyde dissolved in DMSO). The kinetic parameters of the
PeHNL was then reused for the next batch of reactions under the same
conditions. The amount of (R)-mandelonitrile produced was estimated as
described above.
m
enzyme (kcat and K ) were calculated by non-liner least-squares curve
fitting against plots of the initial velocity vs substrate concentrations using
the Michaelis-Menten equation.
Acknowledgements
Effect of pH on enzyme activity and stability
The financial support given to A. N. from Ministry of Education,
Culture, Sport, Science and Technology (MEXT) of Japan is
deeply appreciated. This work was supported by ERATO
(Exploratory Research for Advanced Technology Program),
Asano Active Enzyme Molecule Project of Japan Science and
Technology Agency (Grant No. JPMJER1102). This research
was also supported in part by a grant-in-aid for Scientific
Research (S) from The Japan Society for Promotion of Sciences
Two natural purified PeHNLs were used to determine optimum pH for
(
R)-mandelonitrile synthesis. The reaction was performed at 25 °C
between pH 2.5 to pH 6.0 (400 mM) for 5 min. To determine enzyme
stability, each PeHNL was preincubated at 30 °C between the range of
pH 3.5 to pH 10.0 (40 mM) for 1 h. The remaining activity of the enzyme
was measured as described above.
Effect of temperature on enzyme activity and stability
(
Grant No. 17H06169) to Y. Asano. We are thankful to Prof.
Aran H-Kittikun, Prince of Songkla University, for his kind
discussion upon the occurrence of HNL in Passion fruits, and for
his help in the collaboration between Thailand and Japan. We
thank Prof. Sujinda Sriwattana, the Dean of Faculty of Agro-
Industry, Chiang Mai University, for letting us use their
laboratories and facilities. We would like to thank Dr. Masashi
Nakata, the chairman of the Botanic Gardens of Toyama:
Botanic Gardens of Toyama, Toyama, Japan for providing us
leaves of P. edulis used in some part of this study.
Optimum temperature required for enzyme activity was examined by
incubating the reaction mixture at temperatures ranging from 5 °C to
45 °C in citrate buffer (pH 4.0, 400 mM) for 5 min. For determination of
thermostability, the enzyme was preincubated in potassium phosphate
buffer (pH 6.0, 20 mM) over the range of 30 to 80 °C for 1 h. The
remaining activity of the enzyme was measured as described above.
Effect of organic solvent on stability and cyanohydrin synthesis in a
biphasic system
Organic solvent stability of two natural PeHNLs were compared using
[
13c]
different organic tests as described in detail by Nuylert et al.
In the
study on cyanohydrin synthesis of C-terminal truncation PeHNL mutants,
diisopropyl ether (DIPE) was selected for synthesis of cyanohydrins in a
biphasic system, organic solvent phases containing various aldehyde
Keywords: enzyme catalysis • hydroxynitrile lyase • Passiflora
edulis • C-terminal truncation• enzyme stability
(
250 mM) and mixed with citrate buffer (pH 3.5, 400 mM) containing each
purified C-terminal truncated PeHNL mutants (10 U) in total volumes of
.5 mL in 2.0 mL microcentrifuge tubes. Reactions were initiated by
adding potassium cyanide (300 mM), and the mixtures were incubated at
[
[
[
1]
2]
3]
a) M. Zagrobelny, S. Bak, B. L. Moller, Phytochemistry 2008, 69, 1457-
468; b) T. Yamaguchi, Y. Kuwahara, Y. Asano, FEBS Open Bio 2017,
7, 335-347.
a) M. Dadashipour, Y. Ishida, K. Yamamoto, Y. Asano, Proc Natl Acad
Sci U S A 2015, 112, 10605-10610; b) T. Yamaguchi, A. Nuylert, A. Ina,
T. Tanabe, Y. Asano, Sci Rep 2018, 8, 3051.
M. Dadashipour, Y. Asano, ACS Catal 2011, 1, 1121-1149.
I. Jansen, R. Woker, M.-R. Kula, Biotechnol. Appl. Biochem. 1992, 15,
1
1
1
0
0 °C with shaking at 1500 rpm in an incubator shaker (BioShaker M-BR-
22UP, Taitec Corporation, Tokyo, Japan). Aliquots of samples (10 µL)
were collected from the organic phase at various time-points and
analyzed using HPLC at condition as described in detail by Dadashipour
[4]
[
2a]
9
0-99.
R. S. Yemm, J. E. Poulton, Arch. Biochem. Biophys. 1986, 247, 440-
45.
et al.
Alternatively, for chiral high performance liquid chromatography
[
5]
(
HPLC) analysis of the conversion of 2-chlorobenzaldehyde, 3-
4
chlorobenzaldehyde and 4-chlorobenzaldehyde were performed with a
CHIRALCEL OD-H column (Diacel, Osaka, Japan) under the following
condition: mobile phase n-hexane/propan-2-ol/trifluoroacetic acid with a
[
[7]
[8]
6]
S. Nanda, Y. Kato, Y. Asano, Tetrahedron 2005, 61, 10908-10916.
M. Asif, T. C. Bhalla, Catalysis Letters 2016, 146, 1118-1127.
Y. C. Zheng, J. H. Xu, H. Wang, G. Q. Lin, R. Hong, H. L. Yu, Adv.
Synth. Catal. 2017, 359, 1185-1193.
-
1
volume ratio of 96:4:0.2, flow rate 1 mL min , absorbance 228 nm,
column oven temperature 25 °C. ee values were determined according to
relative concentration (mM) of the two enantiomers using Equation (1):
[
9]
J. Albrecht, I. Jansen, M. R. Kula, Biotechnol. Appl. Biochem. 1993, 17,
191-203.
[10] H. Wajant, S. Forster, D. Selmar, F. Effenberger, K. Pfizenmaier, Plant
physiol. 1995, 109, 1231-1238.
[
[
푅]−[푆]
푅]+[푆]
[11] J. Andexer, J. von Langermann, A. Mell, M. Bocola, U. Kragl, T. Eggert,
M. Pohl, Angew. Chem. Int. Ed. 2007, 46, 8679-8681.
푒푒 [%] =
× 100
(1)
[
12] T. Ueatrongchit, A. Kayo, H. Komeda, Y. Asano, A. H-Kittikun, Biosci.
Biotechnol. Biochem. 2008, 72, 1513-1522.
Reusability
[
13] a) Y. Asano, K. i. Tamura, N. Doi, T. Ueatrongchit, A. H-Kittikun, T.
Ohmiya, Biosci. Biotechnol. Biochem. 2005, 69, 2349-2357; b) T.
Ueatrongchit, K. i. Tamura, T. Ohmiya, A. H-Kittikun, Y. Asano, Enzyme
Microb. Technol. 2010, 46, 456-465; c) A. Nuylert, Y. Ishida, Y. Asano,
ChemBioChem 2017, 18, 257-265.
14] E. Lanfranchi, T. Pavkov-Keller, E.-M. Koehler, M. Diepold, K. Steiner,
B. Darnhofer, J. Hartler, T. Van Den Bergh, H.-J. Joosten, M. Gruber-
Khadjawi, Sci Rep 2017, 7, 46738.
15] K. Isobe, A. Kitagawa, K. Kanamori, N. Kashiwagi, D. Matsui, T.
Yamaguchi, K.-i. Fuhshuku, H. Semba, Y. Asano, Biosci. Biotechnol.
Biochem. 2018, 82, 1760-1769.
16] U. Hanefeld, A. J. Straathof, J. J. Heijnen, Biochim. Biophys. Acta,
Protein Struct. Mol. Enzymol. 1999, 1432, 185-193.
The reusability of recombinant purified PeHNL, wild-type (PeHNL121y),
and C-terminal truncated PeHNLΔ107 expressed in E. coli was compared
in the cyanohydrin synthesis reaction batch containing benzaldehyde
[
[
[
(250 mM) and KCN in biphasic systems of citrate buffer (pH 3.5, 400 mM,
0.75 mL) and DIPE (0.75 mL) in a microcentrifuge tube (2 mL). A
reaction was also started by addition of each PeHNL (5 U) the same as
described above. After 3 h of incubation, the buffer phase containing the
enzyme was recovered and dialyzed against citrate buffer (pH 3.5, 400
mM) using a centrifugal filtration device. The exchanged buffer containing
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
9
[
17] J. Hughes, J. Decarvalho, M. A. Hughes, Arch. Biochem. Biophys.
994, 311, 496-502.
[32] Y. Wang, H. Yuan, J. Wang, Z. Yu, Bioresour. Technol. 2009, 100, 345-
349.
1
[
[
18] H. Wajant, K.-W. Mundry, Plant Sci. 1993, 89, 127-133.
19] M. Dadashipour, M. Yamazaki, K. Momonoi, K. Tamura, K. Fuhshuku,
Y. Kanase, E. Uchimura, G. Kaiyun, Y. Asano, J. Biotechnol. 2011, 153,
[33] M. Nisha, T. Satyanarayana, Applied Microbiology and Biotechnology
2015, 99, 5461-5474.
[34] Z. Fang, J. Zhang, G. Du, J. Chen, Sci Rep 2016, 6, 27953.
[35] Z. Lu, Q. Wang, S. Jiang, G. Zhang, Y. Ma, Sci Rep 2016, 6, 22465.
[36] a) M. Bauer, H. Griengl, W. Steiner, Enzyme Microb. Technol. 1999, 24,
514-522; b) Z. Liu, B. Pscheidt, M. Avi, R. Gaisberger, F. S. Hartner, C.
Schuster, W. Skranc, K. Gruber, A. Glieder, ChemBioChem 2008, 9,
58-61; c) B. Pscheidt, Z. Liu, R. Gaisberger, M. Avi, W. Skranc, K.
Gruber, H. Griengl, A. Glieder, Adv. Synth. Catal. 2008, 350, 1943-
1948.
[37] a) R. Wiedner, B. Kothbauer, T. Pavkov‐Keller, M. Gruber‐Khadjawi, K.
Gruber, H. Schwab, K. Steiner, ChemCatChem 2015, 7, 325-332; b) B.
Pscheidt, M. Avi, R. Gaisberger, F. S. Hartner, W. Skranc, A. Glieder, J.
Mol. Catal., B Enzym. 2008, 52-53, 183-188.
1
00-110.
20] I. Dreveny, K. Gruber, A. Glieder, A. Thompson, C. Kratky, Structure
001, 9, 803-815.
[
[
[
[
[
2
21] Y. Fukuta, S. Nanda, Y. Kato, H. Yurimoto, Y. Sakai, H. Komeda, Y.
Asano, Biosci Biotechnol Biochem 2011, 75, 214-220.
22] I. Hajnal, A. Lyskowski, U. Hanefeld, K. Gruber, H. Schwab, K. Steiner,
FEBS J 2013, 280, 5815-5828.
23] R. Wiedner, M. Gruber-Khadjawi, H. Schwab, K. Steiner, Comput Struct
Biotechnol J 2014, 10, 58-62.
24] D. B. Rodriguez-Amaya, in Encyclopedia of Food Sciences and
Nutrition (Second Edition) (Ed.: B. Caballero), Academic Press, Oxford,
2
003, pp. 4368-4373.
[38] R. K. Rej, T. Das, S. Hazra, S. Nanda, Tetrahedron Asymmetry 2013,
24, 913-918.
[39] M. Terreni, G. Pagani, D. Ubiali, R. Fernandez-Lafuente, C. Mateo, J.
M. Guisán, Bioorg. Med. Chem. Lett. 2001, 11, 2429-2432.
[40] T. Ema, S. Ide, N. Okita, T. Sakai, Adv. Synth. Catal. 2008, 350, 2039-
2044.
[
25] D. Chassagne, J. C. Crouzet, C. L. Bayonove, R. L. Baumes, J. Agric.
Food Chem. 1996, 44, 3817-3820.
26] F. Motojima, A. Nuylert, Y. Asano, FEBS J. 2018, 285, 313-324.
[
[
27] a) M. T. Reetz, J. D. Carballeira, A. Vogel, Angew. Chem. Int. Ed. Engl.
2
2
006, 45, 7745-7751; b) H. Yu, Y. Yan, C. Zhang, P. A. Dalby, Sci Rep
017, 7, 41212.
[41] U. K. Laemmli, Nature 1970, 227, 680-685.
[
28] Y. Nagashima, A. von Schaewen, H. Koiwa, Plant Sci. 2018, 274, 70-
[42] R. M. Zacharius, Anal. Biochem. 1969, 30, 148-152.
79.
[
[
[
29] C. Rayon, P. Lerouge, L. Faye, J. Exp. Bot. 1998, 49, 1463-1472.
30] L. Faye, M. J. Chrispeels, Plant physiol. 1989, 89, 845-851.
31] L. Faye, M. J. Chrispeels, Planta 1987, 170, 217-224.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

ChemBioChem
10.1002/cbic.201900468
1
0
Entry for the Table of Contents (Please
choose one layout)
Layout 1:
FULL PAPER
PeHNL of the yellow passion fruit
showed much better thermostability,
pH stability, and organic tolerance
than the PeHNL purple passion fruit.
The non-glycosylated and C-truncated
mutant PeHNLΔ107 could improve the
stability and reusability properties as
well as the glycosylated natural
A. Nuylert, F. Motojima, C. Khanongnuch,
T. Hongpattarakere, and Y. Asano*
Page No. – Page No.
Stabilization of hydroxynitrile lyases
from two variants of passion fruits,
Passiflora edulis Sims and Passiflora
edulis forma flavicarpa by C-terminal
truncation
PeHNL and expressed in P. pastoris.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.