Xylella fastidiosa esterase rather than hydroxynitrile lyase

Article abstract of DOI:10.1002/cbic.201402685

In 2009, we reported that the product of the gene SCJ21.16 (XFa0032) from Xylella fastidiosa, a xylem-restricted plant pathogen that causes a range of diseases in several important crops, encodes a protein (XfHNL) with putative hydroxynitrile lyase activi

Full text of DOI:10.1002/cbic.201402685

DOI: 10.1002/cbic.201402685
Full Papers
Xylella fastidiosa Esterase Rather than Hydroxynitrile Lyase
Guzman Torrelo,*[a] Fayene Zeferino Ribeiro de Souza,[a, b] Emanuel Carrilho,[b] and
Ulf Hanefeld[a]
In 2009, we reported that the product of the gene SCJ21.16
(XFa0032) from Xylella fastidiosa, a xylem-restricted plant
pathogen that causes a range of diseases in several important
crops, encodes a protein (XfHNL) with putative hydroxynitrile
lyase activity. Sequence analysis and activity tests indicated
that XfHNL exhibits an a/b-hydrolase fold and could be classi-
fied as a member of the family of FAD-independent HNLs.
Here we provide a more detailed sequence analysis and new
experimental data. Using pure heterologously expressed XfHNL
we show that this enzyme cannot catalyse the cleavage/syn-
thesis of mandelonitrile and that this protein is in fact a non-
enantioselective esterase. Homology modelling and ligand
docking simulations were used to study the active site and
support these results. This finding could help elucidate the
common ancestor of esterases and hydroxynitrile lyases with
an a/b -hydrolase fold.
Introduction
Sequence alignment and database search tools have been
used in various life science fields to find alternate or missing
enzymes in metabolic pathways, as well as in drug discovery
and in toxicity studies.[1] Additionally, they can shed light on
the evolutionary aspects of enzyme mechanisms. These tools
have improved over time,[2] and currently they constitute a
powerful set of methods for analysing relationships among se-
quences in protein (super)families and developing hypotheses
about structure–function relationships in families and super-
families.
terase/lipase superfamily: the S-selective enzymes from Hevea
brasiliensis (HbHNL), Manihot esculenta (MeHNL), Sorghum bicol-
or (SbHNL) and Baliospermum montanum (BmHNL) and the R-
selective enzyme from Arabidopsis thaliana (AtHNL).[7] Most
HNLs have been discovered in plants, and several have been
cloned successfully and expressed in Escherichia coli or yeasts
such as Pichia pastoris.[8]
The amino acid sequence of XfHNL was compared with
HNLs of the esterase/lipase superfamily, and the sequence
alignment showed an overall identity of about 30% between
all sequences, with greater similarity to the R-selective AtHNL,
although in the work of Caruso et al., the enantioselectivity of
XFa0032 (XfHNL) was not described and the crystal structure
was not available.[4]
In 2000, the genome of Xylella fastidiosa, the first plant phy-
topathogen to be completely sequenced, was unravelled.[3] In
a search to elucidate the plant–pathogen interaction, a se-
quence similarity search assigned the gene product XFa0032 (a
251 amino acids protein) as an FAD-independent hydroxynitrile
lyase (HNL) with an a/b-hydrolase fold.[4]
XfHNL is the first example of a bacterial HNL, although re-
cently HNLs showing sequence similarity to proteins of the
cupin superfamily have been described in various bacteria.[9]
Our findings show that XfHNL is able to hydrolyse esterase
substrates and, despite the high sequence homology and the
results of a previous report,[4] no hydroxynitrile lyase activity
could be detected.
HNLs form an interesting group of enzymes, and are used
for the formation of enantiopure cyanohydrins. Over the last
two decades, the enzymes have been investigated extensively,
especially regarding their potential as catalysts in organic
chemistry,[5] as cyanohydrins constitute important building
blocks for fine chemicals.[6] Within this group of enzymes,
which belong to different protein families with no significant
sequence similarity, some of the best characterised HNLs are Results and Discussion
those having an a/b-hydrolase fold. They are related to the es-
HNL activity of XfHNL
As mentioned above, a sequence similarity search can help to
[a] Dr. G. Torrelo, Dr. F. Z. Ribeiro de Souza, Prof. U. Hanefeld
Gebouw voor Scheikunde, Biokatalyse Afdeling Biotechnologie
Technische Universiteit Delft
determine the activity of an unknown protein, though this is
a first step. Subsequently, the protein has to be purified and
an activity test has to be carried out with natural substrates for
each class of enzyme (oxidoreductase, hydrolase, lyase, etc.) in
order to identify the enzymatic activity. The cleavage of man-
delonitrile is the common reaction to test for HNL activity
(Scheme 1).[10] This substrate has the disadvantage that it is not
stable above pH 5; it rapidly decomposes to benzaldehyde, so
spontaneous decomposition has to be subtracted from the
Julianalaan 136, 2628 BL Delft (The Netherlands)
E-mail: G.TorreloVilla@tudelft.nl
[b] Dr. F. Z. Ribeiro de Souza, Prof. E. Carrilho
Instituto de Quꢀmica de S¼o Carlos, Universidade de S¼o Paulo
Av. Trabalhador S¼o-carlense, 400, CEP 13560-970
CP 780, S¼o Carlos-SP (Brazil)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201402685.
ChemBioChem 2015, 16, 625 – 630
625
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 1. Cleavage of mandelonitrile catalysed by HNLs.
enzymatic reaction in an activity test. Another difficulty is that
a change of 18C in the cuvette of the UV spectrophotometer
results in a measuring error of approximately 10%. Finally, the
mandelonitrile purity should be high enough to give an initial
absorption of 0.4–0.8 (at 280 nm) to obtain reliable results.
HNL activity of 42.9 Umgꢀ1 at pH 7 was reported for XfHNL
by Caruso et al.[4] The authors mentioned that the enzymatic
reaction at pH 8.0 (where XfHNL is more stable) was not evalu-
ated, because of the significant degradation of mandelonitrile.
In our experience, even at pH 7 spontaneous degradation of
mandelonitrile is quite fast. This makes it difficult to detect an
enzymatic reaction. Caruso et al. also reported 5.1 Umgꢀ1 at
pH 5; this is a low but acceptable activity.
Figure 1. Absorbance (280 nm) comparison between the reaction activity
tests with enzyme (a) and the control reactions without enzyme (c) at
pH 5, 5.5, 6, 6.5 and 7.
shares about 25% amino acid identity with the HNLs and that
Ser87 is one of the catalytically active residue. The small amino
acids flanking the serine allow the formation of the sharp bend
that is typical for the “nucleophile elbow” of a/b-hydrolase-
fold proteins.[12] The two other residues of the catalytic triad
are also found at conserved positions in XfHNL (Asp200 and
His226).
Here, in order to elucidate the enantioselectivity of this HNL,
we synthesised the gene based on gene accession number
NP_061688 according to the method of Caruso et al.[4] (Sup-
porting Information). The synthetic gene was cloned into the
expression vector pET28a. E. coli TOP10 competent cells were
used for plasmid propagation, then the plasmid was trans-
formed into E. coli BL21(DE3) for expression and purification of
the full-length protein. All experiments were also performed
with the originally published clone of the enzyme,[4]
with as expected identical results.
Interestingly, members of the a/b-hydrolase fold show both
R and S selectivity.[7,13] For the S-selective HNLs (HbHNL,
MeHNL, and BmHNL), the enzymatic mechanism involves the
Ser-His-Asp catalytic triad as a general base to deprotonate the
cyanohydrin hydroxyl group. In this triad, the histidine residue
Surprisingly, during the preliminary activity assays,
no activity was observed at pH 5. Additionally, at
pH 5.5, 6, 6.5 and 7, the same results were obtained
for the controls as for the reactions with enzyme
(Figure 1). It is important to note that as the pH in-
creases, so does the rate of decomposition of man-
delonitrile. Control reactions for background subtrac-
tions must be carried out carefully. Additionally,
mandelonitrile synthesis reactions at pH 5 and 6.5
were carried out in order to confirm the total ab-
sence of HNL activity in both directions of the reac-
tion. A biphasic system using an excess of HCN in
MTBE was used (see Experimental Section). As in the
case of the cleavage of mandelonitrile, no activity
was observed.
Amino acid sequence analysis of XfHNL
In view of these results, a second analysis of the
sequence alignment of XfHNL was performed with
the HNLs of the a/b-hydrolase fold family for which
X-ray data are available (MeHNL, HbHNL, AtHNL, and
BmHNL; Figure 2). SbHNL was not included because
this enzyme has a completely different active-site ar-
chitecture: the conventional classic triad is not pres-
ent.[11] The sequence analysis showed that XfHNL
Figure 2. Multiple alignment of XfHNL (NP_061688.1) with the plant hydroxynitrile lyases
HbHNL (7YAS_A), MeHNL (1DWP_A), BmHNL (3WWP_A) and AtHNL (3DQZ_A). Identical
and similar amino acids are indicated by dark and pale grey backgrounds, respectively
(the following amino acids were considered similar: A, S, T; D, E; F, Y, W; I, L, M, V; N, Q;
R, K). The residues of the catalytic triad are marked #. The crucial amino acids involved in
the enzymatic mechanism and enantioselectivity of HNLs are boxed.
ChemBioChem 2015, 16, 625 – 630
www.chembiochem.org
626
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
acts as a base, the serine hydroxyl group acts as a mediator,
and a lysine residue (Lys236) gives the positive charge that sta-
bilises the negative charge evolving at the cyano group during
the reaction together with Asp.[14] The Ser-His-Asp catalytic
triad is conserved in the R-selective AtHNL and is crucial for
catalysis,[15] whereas HbHNL and AtHNL differ at several other
residues surrounding the active site. Specifically, the lysine resi-
due crucial for stabilising the negative charge on the cyanide
(Lys236 in HbHNL and BmHNL, Lys237 in MeHNL) is replaced
by methionine. In a recently proposed catalytic mechanism for
AtHNL,[7] His236 from the catalytic triad acts as a general base
and the emerging negative charge on the cyano group is sta-
bilised by main-chain amide groups (Ala13 and Phe82) and an
a-helix dipole very similar to a/b-hydrolases. In both cases,
deprotonation of the cyanohydrin is facilitated by hydrogen-
bond interactions between the hydroxyl group and Asn12
(AtHNL) or Thr11 (HbHNL).
Later, the Schwab group provided clear evidence for the
conversion of the BgEstC into an active hydroxynitrile lyase by
the exchange of only few amino acids (conference poster, re-
sults not published).[18b] They exchanged Ser276 (correspond-
ing to the Val237 in XfHNL) in BgEstC with the essential lysine.
They also found two other positions that had an impact on
the desired HNL activity. Gly24 (corresponding to Gly31 in
XfHNL, at the active site in the homology model) was ex-
changed to threonine, and His111 (His86 in XfHNL), close to
the catalytic serine, was exchanged to glutamic acid (which is
present in S-HNLs). Activity assays based on the cleavage reac-
tion provided evidence for successful conversion into an HNL.
The activity was also confirmed for these mutants by using
benzaldehyde as substrate. In 2010, the Kazlauskas group also
reported the conversion of a plant esterase into an HNL with
just two amino acid substitutions.[17] This analysis explains the
absence of HNL activity for XfHNL. At the same time, it strongly
suggests an esterase activity.
The analysis of the sequence alignment of the XfHNL
showed that these crucial residues involved in the enzymatic
mechanism of HNLs (both for S and R enantioselectivity; boxed
in Figure 2) are missing. Other amino acids are present. Thr11
and Lys236 involved in the catalytic mechanism of S-HNLs are
different in XfHNL. In place of Thr11 there is Gly31, an amino
acid with quite different properties. Furthermore, there is
Val237 instead of the charged Lys236 of S-HNLs. Valine is not
a charged amino acid (it is a hydrophobic amino acid), there-
fore it cannot stabilise the negative charge on the cyanide
group.
Esterase activity
Purified XfHNL protein was examined with different assays for
esterase and lipase activity (Scheme 2, Table 1). As expected,
Regarding the amino acids involved in the catalytic mecha-
nism of the R-enantioselective AtHNL (Asn12, Ala13 and Phe82)
there are fewer differences. Ala13 in the AtHNL corresponds to
Ala32 in XfHNL, according to the homology model (see Homol-
ogy model and docking simulations, below). Phe82 corre-
sponds to Tyr88, which has similar properties. The difference
lies in the important Asn12, which is a Gly31 in XfHNL. Glycine
is shorter than asparagine, and the backbone NH group, neces-
sary to facilitate the deprotonation of the cyanohydrin, is in
a different orientation and not in close proximity to the active
site pocket. The positions of all these amino acids were con-
firmed in the homology model of XfHNL.
Screening databases with the Jackhmmer server indicated
that the amino acid sequence of XfHNL has high homology
(61% identity) with an esterase from Ralstonia solanacearum
UW551 (UniProt accession: A3RYV8), as well as hydrolases from
Gallaecimonas xiamenensis 3-C-1 (K2JM70), Pseudomonas bras-
sicacearum (W8PX61) and fluorescens (U7DGE0), and a putative
hydrolase from Pseudomonas sp. GM60 (J3B303). Unfortunately,
these enzymes have not been studied, although, interestingly,
there is an esterase that has been well characterised and
shows significant homology to HbHNL and MeHNL: EstC from
Burkholderia gladioli (BgEstC; accession number AAF66687.1).[16]
XfHNL shares about 41% sequence similarity with the BgEstC,
and a sequence alignment showed that both enzymes lack the
crucial amino acids for HNL activity previously mentioned. Fur-
thermore, the two enzymes share almost exactly the lipase
motif around the catalytic serine (VVLVGHSXGG) and the His-
Gly motif (VLVHGAXX) described by Reiter et al.[16]
Scheme 2. Reactions used for determining esterase and lipase activity.
Table 1. Catalytic activity of XfHNL at different pH values.
Catalytic activity
[mmol per min per mg of enzyme]
pH 7
pH 8
p-nitrophenyl acetate
p-nitrophenyl butyrate
tributyrin
10.9
3.1
0
66.9
4.2
0
1-phenyl ethyl acetate
butyl 2-phenyl propanoate
0
0
0
0
ChemBioChem 2015, 16, 625 – 630
www.chembiochem.org
627
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
XfHNL displayed good enzymatic activity in the hydrolysis of
nitrophenyl esters, particularly for p-nitrophenyl acetate, which
showed higher activity at pH 8 than at pH 7. With p-nitrophen-
yl butyrate, the pH difference was lower. Lipase activity was
tested using tributyrin and two racemic substrates, 1-phenyl
acetate and butyl 2-phenyl propanoate (substrates that are
commonly used to check the enantioselectivity of these en-
zymes).[18] No significant lipase activity was detected with
these substrates, as is the case for BgEstC.[16]
Homology model and docking simulations
The X-ray structures of all HNLs in the sequence alignment are
known (with the exception of XfHNL). For this reason, a homol-
ogy model of XfHNL was made using the software YASARA.
This allowed us to analyse the positions of the residues in the
active site and carry out some docking simulations. The best
result showed a protein (dimer) with an overall structure slight-
ly similar to that of MeHNL and AtHNL, with root mean square
deviation (RMSD) values for Ca atoms of 17.137 ꢁ between
XfHNL and MeHNL, and 18.140 ꢁ between XfHNL and AtHNL. A
superposition of these structures showed differences in the
loop regions, and two b-sheets on the XfHNL surface are short-
er, probably because of the difference in length between these
proteins (Supporting Information). An analysis of the catalytic
triad (Ser-Asp-His) showed that these residues are in similar ori-
entations in the active sites of these enzymes, thus highlight-
ing the importance of other catalytic residues for HNLs (quite
different in XfHNL).
Figure 3. Stereo views (PyMOL) of the active-site structure of XfHNL (homol-
ogy model, YASARA). A) Docking simulation with 100 molecules of mandelo-
nitrile (orange). B) Docking simulation with 100 molecules of p-nitrophenyl
acetate (orange). Active-site residues (S87, D200, H226) and the correspond-
ing residues for the Lys236 and Thr11/Asn12 in S- and R-HNLs (G31, V227)
are shown as stick models.
The MeHNL and AtHNL crystal structures and the XfHNL ho-
mology model were used to carry out docking studies. (A simi-
lar investigation with BgEstC was not possible as no crystal
structure is available.) Mandelonitrile was chosen as the ligand,
and one monomer of the enzyme was used as the receptor.
The docking of mandelonitrile was assayed 100 times by using
the minimum-energy structures of the enzymes. In the case of
MeHNL and AtHNL, the 100 mandelonitrile molecules were
docked to the active site in the same way (same orientation
for each enzyme, but differences between them because of
the different enantioselectivity). The 100 mandelonitrile mole-
cules in the active site of XfHNL had six different orientations,
thus indicating unspecific binding (Figure 3A). As explained
above, the lack of some of the HNL catalytic residues can ex-
plain this unspecific binding, and hence the absence of HNL
activity observed in the experiments.
show that all the p-nitrophenyl acetate molecules were placed
in the active site in the same orientation, as would be expect-
ed for a substrate for which the enzyme showed activity
(Table 1).
Conclusions
In this study, it was demonstrated that the product of the
gene SCJ21.16 (XFa0032) from X. fastidiosa encoded a protein
with esterase activity. Despite the sequence homology, no hy-
droxynitrile lyases activity could be detected. Sequence align-
ment, docking simulations and activity assays support this con-
clusion. We think that our finding should support the renam-
ing of the protein and could contribute to elucidating the
common ancestor of esterases and hydroxynitrile lyases with
an a/b-hydrolase fold.[20]
Some authors have pointed out the importance of analysis
of the hydrophobic residues surrounding the active site.[14b,19]
These can play an important role in recognising various sub-
strates and introducing the cyanide ion into the active site. An
analysis of the XfHNL residues corresponding to the nine hy-
drophobic residues in the active site of HNLs showed that only
five are hydrophobic, and only four surround the active site.
The rest are polar or are on the surface of the protein, far from
the active site.
Experimental Section
CAUTION! All procedures involving hydrogen cyanide were per-
formed in a well-ventilated fume hood equipped with a HCN de-
tector. HCN-containing waste was neutralised by using commercial
bleach and stored independently for disposal with a large excess
of bleach.
In order to validate the homology model and the docking
results, a docking simulation using p-nitrophenyl acetate as
ligand was carried out. The results (100 molecules; Figure 3B)
Preparation and purification of XfHNL: The synthetic gene (se-
quence deposited in the database under accession number NP_
061688, without codon optimisation) cloned into the expression
ChemBioChem 2015, 16, 625 – 630
www.chembiochem.org
628
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
vector pET28a was ordered from BaseClear (Leiden, The Nether-
lands). E. coli TOP10 competent cells were used for the plasmid
propagation. The plasmid was transformed into E. coli BL21(DE3)
for expression and purification of the full-length protein. Heterolo-
gous expression and purification of recombinant protein (XfHNL)
was performed according to Caruso et al.[4] A BC assay was used
for protein quantification.[21]
p-nitrophenol was performed under the same experimental condi-
tions.
Lipase activity was assayed at pH 7 and 8 (buffers as above) by
a pH-stat method with tributyrin, 1-phenylethyl acetate and butyl
2-phenyl propanoate as substrates.[24] Small-volume reactions with
the last two substrates were also performed and checked by GC
and HPLC in order to reduce the detection limit and rule out any
activity. The reactions consisted of enzyme (0.5–10 mg) and 1-phe-
nylethyl acetate (10–25 mg, 0.06–0.15 mmol) or butyl 2-phenyl
propanoate (10–25 mg, 0.05–1.2 mmol) in citrate-phosphate buffer
(1 mL, 50 mm, pH 5), potassium phosphate buffer (1 mL, 50 mm
pH 7) or Tris·HCl buffer (50 mm, pH 8, 1 mL). The reaction mixture
was vigorously stirred in a Thermomixer comfort (Eppendorf) at
258C for 24 h. The hydrolysis of 1-phenylethyl acetate was moni-
tored by GC, and the hydrolysis of butyl 2-phenyl propanoate was
monitored by HPLC.
Chemicals: (ꢁ)-Mandelonitrile (Sigma–Aldrich) was purified by
column chromatography (PE/EtOAc 9:1!3:7) prior to use. Benz-
aldehyde (Acros Organics) was distilled prior to use and stored
under nitrogen at 48C. Tributyrin, p-nitrophenylacetate, p-nitrophe-
nylbutyrate, 2-phenylpropionic acid, n-butanol, 1-phenylethanol,
isopropanol and heptane (HPLC grade) were purchased from
Sigma–Aldrich. 1-phenylethyl acetate was purchased from Acros
Organics. Petroleum ether and ethyl acetate (technical grade) were
purchased from VWR. Methyl tert-butyl ether (MTBE, 99.9% extra
pure, Acros Organics) was used without further treatment unless
otherwise specified. Aqueous buffers were prepared from analytical
grade salts.
All enzymatic assays and control reactions were carried out in at
least duplicate.
GC method: Reaction mixtures were extracted with diethyl ether
(2ꢂ), and the organic phase was dried with anhydrous MgSO4. The
final clear solution (1 mL) was injected into a model GC 2010 chro-
matograph (Shimadzu) equipped with a CP-Chirasil-Dex CB column
(25 mꢂ0.32 mmꢂ0.25 mm; Agilent Technologies) with helium as
the carrier gas: injector 2508C, detector 2708C, split 50, flow rate:
1.59 mLminꢀ1, maximum: 2708C. The temperature program was
708C (0.1 min), increase to 1108C (308Cminꢀ1), hold (11 min), in-
crease to 2458C (308Cminꢀ1), hold (1 min). Retention times: (R)-1-
phenyl ethyl acetate, 9.1 min; (S)-1-phenyl ethyl acetate, 9.5 min:
(R)-1-phenyl ethanol, 10.2 min; (S)-1-phenyl ethanol, 10.7 min.
Synthesis of butyl-2-phenyl propanoate: 2-Phenyl propionic acid
(6 g, 40 mmol), n-butanol (5.92 g, 80 mmol), toluene (30 mL) and
sulphuric acid (1 mL) were placed in a 100 mL round-bottom flask
connected to a Dean–Stark-type trap. The mixture was heated to
1208C, and the reaction was stopped when no more water
formed. The mixture was washed three times with ice water
(40 mL), saturated Na2CO3 (40 mL) and water (40 mL). The toluene
layer was dried over MgSO4, then the solvent was evaporated
1
under vacuum to obtain the product (7.82 g, 95% yield). H NMR
(400 MHz, CDCl3) butyl-2-phenyl propanoate: d=7.27–7.33 (m, 5H,
5ꢂCH, Ph), 4.07 (t, J=6.8 Hz, 2H, OCH2), 3.71 (q, J=7.2 Hz, 1H,
CHMe), 1.55–1.60 (m, 2H, CH2), 1.51 (d, J=7.2 Hz, 3H, CHMe), 1.27–
1.32 (m, 2H, CH2), 0.88 ppm (t, J=7.5 Hz, 3H, Me); these data are
in accordance with ref. [22].
HPLC method: Reaction mixtures were extracted with ethyl ace-
tate and diethyl ether, and the combined organic phases were
dried with anhydrous MgSO4. The final clear solution (10 mL) was
injected into an HPLC device (Waters). Analyses were performed
on a Chiralpak AD-H column (4.6ꢂ250 mm, 5 mm; Daicel, Tokyo,
Japan) coupled to an SpH 99 column thermostat (Chrompack),
a 515 HPLC pump (Waters), a 717autosampler (Waters), and an
SPD-10A UV/Vis detector (Shimadzu). The column temperature was
maintained at 408C; mobile phase: heptane/isopropanol, 95:5
(0.1% TFA); flow rate: 1 mLminꢀ1; detection: 254 nm. Retention
times: butyl 2-phenyl propanoate, 3.8 min; (S)-2-phenylpropionic
acid, 6.3 min; (R)-2-phenylpropionic acid, 6.75 min,
Hydrogen cyanide 1.5–2m in MTBE: HCN solution was prepared
as described in ref. [21]; determination of HCN concentration was
as in ref. [23].
Enzyme assays: HNL activity (cleavage of mandelonitrile) used pu-
rified XfHNL and was according to reported procedures.[10a] Citrate-
phosphate buffer (50 mm) was used for pH 5, 5.5 and 6; potassium
phosphate buffer (50 mm) was used for pH 6.5 and 7, for both en-
zymatic and control reactions.
Sequence alignment and database search tools: The structure-
based multiple sequence alignment was constructed with
MUSCLE,[25] provide by The European Bioinformatics Institute.[26]
Homology searches were performed with BLAST[27] and Jackhmmer
(HMMER suite).[28]
The reaction with benzaldehyde as substrate was carried out at
two pH values. The enzyme sample (up to 50 mL) was added to cit-
rate-phosphate buffer (50 mm, 450 mL, pH 5) or potassium phos-
phate buffer (50 mm, 450 mL, pH 6.5). Then, HCN (1.7m in MTBE,
500 mL) with benzaldehyde (1 mmol) and 1,3,5-triisopropylbenzene
as the internal standard (0.01 mmol), previously mixed under a
nitrogen atmosphere, were added. The reaction was monitored by
chiral HPLC[21] over 24 h while the reaction flask was stirred vigo-
rously at room temperature (228C).
Homology modelling: The set of homology models was made
using YASARA (version 14.7.17, http://www.YASARA.org)[29] with the
FASTA sequence of XfHNL. The best was selected based on the
quality parameters of models estimated with ProSA.[30] The RMSD
values for Ca atoms between MeHNL, AtHNL and XfHNL were cal-
culated from structures previously overlaid by the MUSTANG algo-
rithm[31] in YASARA.
Esterase activity was determined using the generic esterase sub-
strates p-nitrophenyl-acetate (two carbons) and p-nitrophenyl-bu-
tyrate (four carbons). Working solutions of p-nitrophenyl acetate
and butyrate were prepared at 100 mm in acetonitrile. The reaction
mixture consisted of enzyme (0.2 mg) and p-nitrophenyl ester
(1 mm) in potassium phosphate buffer (50 mm, 1 mL, pH 7) or
Tris·HCl buffer (50 mm, pH 8). The release of p-nitrophenol was
monitored continuously at OD405 nm over 5 min at 258C in a Cary 60
spectrophotometer (Agilent Technologies). A calibration curve of
Docking simulations: The HNL crystal structures (PDB IDs: 1DWP
and 3DQZ) and the XfHNL homology model were used. Initially, all
the hydrogen atoms were shown, and the minimum energy struc-
tures were calculated. Active-site amino acids were identified
based on the active sites in the crystal structures. After assignment
of the substrate-binding site, mandelonitrile was docked to the
structures by using AutoDock 4.2.3[32] (initial position, orientation
ChemBioChem 2015, 16, 625 – 630
www.chembiochem.org
629
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[10] a) U. Hanefeld, A. J. J. Straathof, J. J. Heijnen, Biochim. Biophys. Acta Pro-
tein Struct. Mol. Enzymol. 1999, 1432, 185–193; b) M. Bauer, R. Geyer, M.
Boy, H. Griengl, W. Steiner, J. Mol. Catal. B 1998, 5, 343–347.
[11] H. Lauble, B. Miehlich, S. Fçrster, H. Wajant, F. Effenberger, Biochemistry
2002, 41, 12043–12050.
[12] D. L. Ollis, E. Cheah, M. Cygler, B. Dijkstra, F. Frolow, S. M. Franken, M.
Harel, S. J. Remington, I. Silman, J. Schrag, J. L. Sussman, K. H. G. Ver-
schueren, A. Goldman, Protein Eng. 1992, 5, 197–211.
[13] J.-K. Guterl, J. N. Andexer, T. Sehl, J. von Langermann, I. Frindi-Wosch, T.
Rosenkranz, J. Fitter, K. Gruber, U. Kragl, T. Eggert, M. Pohl, J. Biotechnol.
2009, 141, 166–173.
[14] a) K. Gruber, G. Gartler, B. Krammer, H. Schwab, C. Kratky, J. Biol. Chem.
2004, 279, 20501–20510; b) S. Nakano, M. Dadashipour, Y. Asano, Bio-
chim. Biophys. Acta Proteins Proteomics 2014, 1844, 2059–2067.
[15] J. Andexer, J. von Langermann, A. Mell, M. Bocola, U. Kragl, T. Eggert, M.
Pohl, Angew. Chem. Int. Ed. 2007, 46, 8679–8681; Angew. Chem. 2007,
119, 8833–8835.
and torsions of the ligand molecules were set randomly, 100 runs)
implemented in YASARA. Interactions at the active site were visual-
ised with PyMol Molecular Graphics System (Version 1.4.1 Schrç-
dinger). The same procedure was performed for the docking simu-
lation of XfHNL and p-nitrophenyl acetate.
Acknowledgements
G.T. thanks the Fundaciꢁn Alfonso Martꢀn Escudero (Spain) for
a postdoctoral fellowship. F.Z.R.S thanks the Coordenażo de
AperfeiÅoamento de Pessoal de Nꢀvel Superior (CAPES-PDEE, Proc-
esso BEX: 0333/11-5) for the fellowship, and E.C. thanks the IN-
CTBio (Brazil). The authors thank Dr. Dmitry Suplatov and
Prof. Dr. Vytas Svedas for fruitful discussions.
[16] B. Reiter, A. Glieder, D. Talker, H. Schwab, Appl. Microbiol. Biotechnol.
2000, 54, 778–785.
[17] S. K. Padhi, R. Fujii, G. A. Legatt, S. L. Fossum, R. Berchtold, R. J. Kazlaus-
kas, Chem. Biol. 2010, 17, 863–871.
Keywords: docking simulation · enzyme models · esterases ·
lyases · protein structures · Xylella fastidiosa
[18] a) D. Stetca, I. W. C. E. Arends, U. Hanefeld, J. Chem. Educ. 2002, 79,
1351; b) M. Paravidino, P. Bçhm, H. Grçger, U. Hanefeld in Enzyme Catal-
ysis in Organic Synthesis (Eds.: K. Drauz, H. Grçger, O. May), 3rd. ed.,
Wiley-VCH, Weinheim, 2012, pp. 249–362; c) U. T. Bornscheuer, R. J. Kaz-
lauskas in Hydrolases in Organic Synthesis, Wiley-VCH, Weinheim, 2006,
pp. 61–140.
[19] A. Schmidt, K. Gruber, C. Kratky, V. S. Lamzin, J. Biol. Chem. 2008, 283,
21827–21836.
[20] D. M. Nedrud, H. Lin, G. Lopez, S. K. Padhi, G. A. Legatt, R. J. Kazlauskas,
Chem. Sci. 2014, 5, 4265–4277.
[21] G. Torrelo, N. van Midden, R. Stloukal, U. Hanefeld, ChemCatChem 2014,
6, 1096–1102.
[22] E. Coulbeck, J. Eames, Synlett 2008, 333–338.
[23] L. M. van Langen, F. van Rantwijk, R. A. Sheldon, Org. Process Res. Dev.
2003, 7, 828–831.
[24] L. Veum, M. Kuster, S. Telalovic, U. Hanefeld, T. Maschmeyer, Eur. J. Org.
Chem. 2002, 1516–1522.
[1] a) S. F. Altschul, W. Gish in Methods in Enzymology, Vol. 266: Computer
Methods for Macromolecular Sequence Analysis (Ed.: R. F. Doolittle), Aca-
demic Press, San Diego, 1996, pp. 460–480; b) S. F. Altschul, M. S. Bo-
guski, W. Gish, J. C. Wootton, Nat. Genet. 1994, 6, 119–129; c) S. McGin-
nis, T. L. Madden, Nucleic Acids Res. 2004, 32, W20–W25; d) J. Ye, S.
McGinnis, T. L. Madden, Nucleic Acids Res. 2006, 34, W6–W9; e) M. John-
son, I. Zaretskaya, Y. Raytselis, Y. Merezhuk, S. McGinnis, T. L. Madden,
Nucleic Acids Res. 2008, 36, W5–W9.
[2] S. R. Eddy, PLoS Comput. Biol. 2011, 7, e1002195.
[3] A. J. G. Simpson, F. C. Reinach, P. Arruda, F. A. Abreu, M. Acencio, R. Al-
varenga, L. M. C. Alves, J. E. Araya, G. S. Baia, C. S. Baptista, M. H. Barros,
E. D. Bonaccorsi, S. Bordin, J. M. Bovꢃ, M. R. S. Briones, M. R. P. Bueno,
A. A. Camargo, L. E. A. Camargo, D. M. Carraro, H. Carrer, et al., Nature
2000, 406, 151–157.
[4] C. S. Caruso, R. de Fꢄtima Travensolo, R. de Campus Bicudo, E. G. de Ma-
cedo Lemos, A. P. Ulian de Araffljo, E. Carrilho, Microb. Pathog. 2009, 47,
118–127.
[25] R. C. Edgar, BMC Bioinf. 2004, 5, 113.
[26] H. McWilliam, W. Li, M. Uludag, S. Squizzato, Y. M. Park, N. Buso, A. P.
Cowley, R. Lopez, Nucleic Acids Res. 2013, 41, W597–W600.
[27] S. F. Altschul, W. Gish, W. Miller, E. W. Myers, D. J. Lipman, J. Mol. Biol.
1990, 215, 403–410.
[28] R. D. Finn, J. Clements, S. R. Eddy, Nucleic Acids Res. 2011, 39, W29–
W37.
[5] M. Sharma, N. N. Sharma, T. C. Bhalla, Enzyme Microb. Technol. 2005, 37,
279–294.
[6] a) J. Holt, U. Hanefeld, Curr. Org. Synth. 2009, 6, 15–37; b) A. Ranoux, U.
Hanefeld in Stereoselective Synthesis of Drugs and Natural Products (Eds.:
V. Andrushko, N. Andrushko), Wiley, New York, 2013, Chapter 28.
[7] J. N. Andexer, N. Staunig, T. Eggert, C. Kratky, M. Pohl, K. Gruber, Chem-
BioChem 2012, 13, 1932–1939.
[29] E. Krieger, G. Koraimann, G. Vriend, Proteins Struct. Funct. Bioinf. 2002,
47, 393–402.
[8] a) M. Dadashipour, Y. Asano, ACS Catal. 2011, 1, 1121–1149; b) K. Stein-
er, A. Glieder, M. Gruber-Khadjawi in Biocatalysis in Organic Synthesis 2
(Eds.: K. Faber, W.-D. Fessner, N. J. Turner), Thieme, Stuttgart, 2015,
Chapter 2.1.1..
[9] a) Z. Hussain, R. Wiedner, K. Steiner, T. Hajek, M. Avi, B. Hecher, A. Ses-
sitsch, H. Schwab, Appl. Environ. Microbiol. 2012, 78, 2053–2055; b) I.
Hajnal, A. Łyskowski, U. Hanefeld, K. Gruber, H. Schwab, K. Steiner, FEBS
J. 2013, 280, 5815–5828; c) R. Wiedner, M. Gruber-Khadjawi, H. Schwab,
K. Steiner, Comput. Struct. Biotechnol. J. 2014, 10, 58–62.
[30] M. Wiederstein, M. J. Sippl, Nucleic Acids Res. 2007, 35, W407–W410.
[31] A. S. Konagurthu, J. C. Whisstock, P. J. Stuckey, A. M. Lesk, Proteins
Struct. Funct. Bioinf. 2006, 64, 559–574.
[32] G. M. Morris, R. Huey, W. Lindstrom, M. F. Sanner, R. K. Belew, D. S. Good-
sell, A. J. Olson, J. Comput. Chem. 2009, 30, 2785–2791.
Received: December 1, 2014
Published online on February 12, 2015
ChemBioChem 2015, 16, 625 – 630
www.chembiochem.org
630
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim