Tài liệu Báo cáo khoa học: Substrate specificity and inhibition of brassinin hydrolases, detoxifying enzymes from the plant pathogens Leptosphaeria maculans and Alternaria brassicicola ppt

Substrate specificity and inhibition of brassinin
hydrolases, detoxifying enzymes from the plant pathogens
Leptosphaeria maculans and Alternaria brassicicola
M. Soledade C. Pedras, Zoran Minic and Vijay K. Sarma-Mamillapalle
Department of Chemistry, University of Saskatchewan, Saskatoon, Canada
Introduction
Crucifers (family Brassicaceae, syn. Cruciferae) include
a wide variety of crops cultivated around the world,
including the oilseeds rapeseed and canola (Bras-
sica napus and Brassica rapa) and vegetables such ase
cabbage (Brassica oleraceae var. capitata), cauliflower
(B. oleraceae var. botrytis) and broccoli (B. oleraceae
Keywords
brassinin; cyclobrassinin; detoxification;
dithiocarbamate hydrolase; phytoalexin
Correspondence
M. S. C. Pedras, Department of Chemistry,
University of Saskatchewan, 110 Science
Place, Saskatoon, SK, Canada S7N 5C9
Fax: +1 306 966 4730
Tel: +1 306 966 4772
E-mail:
(Received 6 May 2009, revised
26 September 2009, accepted 22 October
2009)
doi:10.1111/j.1742-4658.2009.07457.x
Blackleg (Leptosphaeria maculans and Leptosphaeria biglobosa) and black
spot (Alternaria brassicicola) fungi are devastating plant pathogens known
to detoxify the plant defence metabolite, brassinin. The significant roles of
brassinin as a crucifer phytoalexin and as a biosynthetic precursor of sev-
eral other plant defences make it important in plant fitness. Brassinin

detoxifying enzymes produced by L. maculans and A. brassicicola catalyse
the detoxification of brassinin by hydrolysis of its dithiocarbamate group
to indolyl-3-methanamine. The purification and characterization of brassi-
nin hydrolases produced by L. maculans (BHLmL2) and A. brassicicola
(BHAb) were accomplished: native BHLmL2 was found to be a tetrameric
protein with a molecular mass of 220 kDa, whereas native BHAb was
found to be a dimeric protein of 120 kDa. Protein characterization using
LC-MS ⁄ MS and sequence alignment analyses suggested that both enzymes
belong to the family of amidases with the catalytic Ser ⁄ Ser ⁄ Lys triad. Fur-
thermore, chemical modification of BHLmL2 and BHAb with selective
reagents suggested that the amino acid serine was involved in the catalytic
activity of both enzymes. The overall results indicated that BHs have new
substrate specificities with a new catalytic activity that can be designated as
dithiocarbamate hydrolase. Investigation of the effect of various phytoal-
exins on the activities of BHLmL2 and BHAb indicated that cyclobrassinin
was a competitive inhibitor of both enzymes. On the basis of pH depen-
dence, sequence analyses, chemical modifications of amino acid residues
and identification of headspace volatiles, a chemical mechanism for hydro-
lysis of the dithiocarbamate group of brassinin catalysed by BHLmL2 and
BHAb is proposed. The current information should facilitate the design of
specific synthetic inhibitors of these enzymes for plant treatments against
blackleg and black spot fungal infections.
Abbreviations
BGT, brassinin glucosyl transferase; BH, brassinin hydrolase; BHAb, brassinin hydrolase from A. brassicicola; BHLmL2, brassinin hydrolase
from L. maculans; BO, brassinin oxidase; HRMS, high-resolution mass spectrometry; L2 ⁄ M2, Laird 2 and Mayfair 2; LC-ESI-MS ⁄ MS, liquid
chromatography-electrospray-tandem mass spectrometry.
7412 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
var. italica) [1]. Cruciferous oilseeds are the third larg-
est source of edible oil, after oil palm (Elaeis guineen-
sis) and soya bean (Glycine max). Both wild and

cultivated crucifers are known to have positive impact
on human health; a high intake of crucifers has been
convincingly associated with a reduced risk of cancer
[2–5]. The phytoalexin brassinin (1) is produced by
crucifers, including economically significant oilseed
crops within the genus Brassica [6,7]. Phytoalexins are
inducible secondary metabolites with antimicrobial
activity and are produced de novo by plants in
response to stress, including pathogen attack [8,9].
Depending on the type of stress, crucifers biosynthesize
different blends of phytoalexins that appear to have
multiple roles, including microbial growth inhibition
and inhibition of certain fungal enzymes [7,10]. The
antifungal activity of brassinin (1) is partly a result of
its dithiocarbamate group, known to be a potent
toxophore present in synthetic agrochemicals used to
control fungi and weeds [11].
It has been shown that some economically important
fungal plant pathogens can detoxify brassinin (1), a pro-
cess that facilitates the microbial colonization of plants
[12]. Such a depletion of brassinin (1) in plant tissues is
an ongoing concern because brassinin (1) is a biosyn-
thetic precursor of several other phytoalexins. Hence, a
decrease in the concentration of brassinin can make
plants more vulnerable to pathogen attack, while higher
concentrations of brassinin and derived phytoalexins
are expected to contribute to higher plant resistance to
disease. Consequently, technologies that prevent brassi-
nin degradation by pathogens could increase the
concentrations of plant defences and decrease the need

to apply fungicides at the onset of disease.
Recently, it was shown that a group of isolates of
the phytopathogenic ‘blackleg’ fungus [Leptosphae-
ria maculans (Desm.) Ces. et de Not., asexual stage
Phoma lingam (Tode ex Fr.) Desm.], virulent to
canola, detoxified brassinin (1) to 3-indolecarboxalde-
hyde (2) using brassinin oxidase (BO) [13]. Also,
another group of isolates of L. maculans (Laird 2 and
Mayfair 2, hereon called L2 ⁄ M2), virulent on brown
mustard (Brassica juncea), was shown to detoxify
brassinin (1) via hydrolysis to indolyl-3-methanamine
(3) [14]. Assays using cell-free homogenates incubated
with brassinin (1) demonstrated that the putative
hydrolase was induced by brassinin (1), N¢-methyl-
brassinin (a synthetic derivative of compound 1) and
camalexin (a phytoalexin of wild crucifers). Similarly,
the black spot fungus Alternaria brassicicola (Schwein.)
Wiltshire, also a pathogen of crucifers, detoxified
brassinin (1) via hydrolysis to indolyl-3-methanamine
(3) [15]. The summary of brassinin detoxification reac-
tions carried out by different fungal phytopathogens is
shown in Fig. 1.
A reasonable approach to control cruciferous phyto-
pathogens, such as L. maculans and A. brassicicola,
could utilize plant treatments with designer compounds
that we coined paldoxins [6,10]. Paldoxins are envi-
sioned as nontoxic and environmentally sustainable
plant treatments containing a combination of specific
synthetic inhibitors of phytoalexin-detoxifying
enzymes. However, to design paldoxins with such char-

acteristics, a significant understanding of the enzymes
involved in the detoxification of phytoalexins, includ-
ing their substrate specificity as well as the molecular
mechanisms of detoxification, is crucial. To this end,
we isolated, characterized and determined the substrate
specificities of brassinin hydrolase (BH) produced by
L. maculans L2 (BHLmL2) and BH produced by
A. brassicicola (BHAb) and report hereon the results
of this work.
Results and Discussion
Induction of BH activity
Previous studies have shown that some crucifer phyto-
alexins (e.g. brassinin and camalexin) [14] and other
chemicals (3-phenylindole) [13] could induce the bio-
synthesis of phytoalexin-detoxifying enzymes in plant
pathogenic fungi. Hence, the activity of BH was exam-
ined in mycelia obtained from cultures of L. maculans
and A. brassicicola incubated with various concentra-
tions (0.012–0.25 mm) of camalexin and 3-phenylin-
dole. After incubation, the cultures were filtered, the
mycelia were extracted and centrifuged, and the result-
ing cell-free extracts were analyzed for BH activity
using brassinin as the substrate (Fig. 2). BH activity
N
H
H
N
S
SCH
3

N
H
CHO
A
BorC
3
1
N
H
NH
2
2
Fig. 1. Transformation of brassinin (1) to indole-3-carboxaldehyde
(2) or to indolyl-3-methanamine (3) in fungal cultures. (A) Leptosp-
haeria maculans isolates virulent to canola; (B) L. maculans isolates
virulent to brown mustard; (C) Alternaria brassicicola.
M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7413
was only observed in cell-free extracts of cultures incu-
bated with camalexin or 3-phenylindole. 3-Phenylin-
dole induced the highest amount of BH activity at the
highest concentration tested in mycelial cultures of
both L. maculans and A. brassicicola. The substantially
higher induction of BH activity caused by 3-phenylin-
dole was particularly noticeable in cultures of L. macu-
lans (twofold higher than that of camalexin). For this
reason, 3-phenylindole (0.2 mm) was used to induce
the biosynthesis of BHs to facilitate their purification.
Purification of BH from L. maculans and
A. brassicicola

Enzymes with BH activity were purified from crude pro-
tein extracts of mycelia cultures, using brassinin as the
substrate, to monitor the enzymatic activity. The puri-
fied enzymes were designated as BHLmL2 for the BH
from L. maculans L2 ⁄ M2, and as BHAb for the BH
from A. brassicicola. Table 1 summarizes the purifica-
tion procedure used for BHLmL2 and indicates the
degree of purification and yield achieved for each step.
The purification protocol involved four column chroma-
tography separations: anion exchange chromatography,
hydroxyapatite chromatography, gel filtration chroma-
tography on Superdex 75 and gel filtration chromato-
graphy on Superdex 200. Table 2 summarizes the
purification procedure of BHAb and indicates the
degree of purification and yield achieved for each step.
The purification protocol involved three column chro-
matography separations: hydrophobic chromatography,
hydroxyapatite chromatography and gel filtration chro-
matography on Superdex 200. It is important to note
that a loss of BH activity obtained from both fungi was
observed during purification in the absence of Triton
X-100 and glycerol. Thus, to prevent enzymatic inacti-
vation, Triton X-100 (0.015%) and glycerol (1–3%)
were added to all buffers except for the extraction
buffer. Under these conditions and storage at )20 °C,
the activities of both BHs were stable for approximately
two weeks. Fractions with BH activity obtained after
the final chromatography step were pooled, concen-
trated and used for biochemical analysis.
SDS/PAGE of purified BHs

The purity of BHLmL2 obtained from L. maculans
was examined by denaturing SDS ⁄ PAGE, which, upon
staining with Coomassie Brilliant Blue R-250, revealed
two bands with apparent molecular mass values of 58
and 220 kDa (Fig. 3A). In addition, Superdex 200
chromatography of the purified BHLmL2 suggested
that the native protein was a tetramer because it was
eluted at a position corresponding to a molecular mass
of 220 kDa (data not shown), comparable to that
obtained under denaturing conditions with
SDS ⁄ PAGE. Likewise, purified BHAb obtained from
A. brassicicola revealed two bands on SDS ⁄ PAGE
with apparent molecular mass values of 60 kDa and
120 kDa (Fig. 3B). Similarly, the purified protein
BHAb was eluted from Superdex 200 (data not shown)
Fig. 2. Effect of camalexin and 3-phenylindole on the activity of BH
from mycelial cultures of Leptosphaeria maculans (A) and Alternaria
brassicicola (B). The results are expressed as means and standard
deviations of three independent experiments.
Brassinin hydrolases M. S. C. Pedras et al.
7414 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
at a position corresponding to a molecular mass of
about 120 kDa. Thus, these data suggest that native
BHAb is a dimer.
Identification and chemical modification of the
purified enzymes
To identify BHLmL2 and BHAb, the bands obtained
from SDS ⁄ PAGE were digested with trypsin and then
analyzed by LC-MS ⁄ MS using mascot algorithms. In
total, 11 peptides were deduced from the LC-MS ⁄ MS

spectral data (Table 3) for BHLmL2 and 9 peptides
were deduced for BHAb (Table 4). The sequence simi-
larity of the identified peptides was analyzed using the
NCBI blast algorithm. Peptide sequences obtained from
BHLmL2 and BHAb were aligned using SIM-Align-
ment Tools. This analysis indicated that several peptide
sequences of BHLmL2 and BHAb showed similarity
but none showed 100% identity to each other. The pep-
tide R.LSASEAMEASLAATRR obtained from a
BHLmL2 digest (Table 3) had 100% identity with a
putative amidase from Sinorhizobium medicae WSM419
(accession no. YP_001314042). Similarly, the peptide
K.TSVPQTLMVCETVNNIIGR obtained from a
BHAb digest showed 100% identity with different puta-
tive amidases from fungi such as Aspergillus oryzae
RIB40 (accession no. XP_001825134), Aspergillus
nidulans FGSC A4 (accession no. XP_682046), Emeri-
cella rugulosa (accession no. AAK29061) and Emericella
unguis (accession no. AAK29062). In addition, analyses
of the complete sequences of these putative fungal amid-
ases using Compute pi ⁄ mw tool ( />tools/pi_tool.html) software predicted that their molecu-
lar masses were about 60 kDa, which is in agreement
with the molecular mass obtained on SDS ⁄ PAGE for
BHAb. Moreover, other peptides (listed in Tables 3 and
4) showed similarity to putative amidases, suggesting
once more that BHs from L. maculans and A. brassici-
cola belong to the amidase superfamily. Additionally,
sequence alignment analyses of fungal putative
amidases that have 100% identity with the peptide
K.TSVPQTLMVCETVNNIIGR from BHAb and

malonamidase E2 from Bradyrhizobium japonicum,
showed that these amidases have a similar Ser ⁄ Ser ⁄ Lys
catalytic triad [16–18] (Supplementary Fig. S1). Overall,
these results suggest that BHs from L. maculans and
A. brassicicola belong to the amidase-signature super-
family of proteins that contain the Ser ⁄ Ser ⁄ Lys triad
active site.
Proteins of the amidase-signature superfamily from
various sources exist in multimeric forms, for example,
N-acetylmuramyl-l-alanine amidase from Bacillus sub-
tilis strain 168 (dimeric form) [19] and N-acetylmura-
moyl-l-alanine amidase from human serum [20].
Tetrameric forms of amidases were identified for the
polyamidase from Nocardia farcinica [21] and for
Table 1. Enzyme yields and purification factors for BHLmL2.
Purification step
Yield
Specific activity (nmol.min
)1
.mg
)1
) Recovery (%)
a
Purification factor (fold)
a
Protein (mg) Activity (nmol.min
)1
)
Crude homogenate
b

26.9 1070 39.8 100 1.0
HiTrapDEAE FF 4.16 525 126.2 49 3.1
Hydroxyapatite 1.86 466 250.3 44 6.3
Superdex 75 0.077 37 484.7 3.5 12
Superdex 200 0.019 16 833.7 1.5 21
a
Recoveries are expressed as percentage of initial activity and purification factors are calculated on the basis of specific activities.
b
Mycelia
from 1 L cultures yielded approximately 50 mg of protein.
Table 2. Enzyme yields and purification factors for BHAb.
Purification step
Yield
Specific activity (nmol.min
)1
.mg
)1
) Recovery (%)
a
Purification factor (fold)
a
Protein (mg) Activity (nmolÆmin
)1
)
Crude homogenate
b
29.4 561 19.1 100 1
Phenyl Sepharose 2.060 165 80.3 30 4.2
Hydroxyapatite 0.148 76 514.7 14 27
Superdex 200 0.012 8.7 728.0 1.5 38

a
Recoveries are expressed as percentage of initial activity and purification factors are calculated on the basis of specific activities.
b
Mycelia
from 1-L cultures yielded approximately 60 mg protein.
M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7415
numerous microbial amidohydrolases that belong to a
family of cyclic amidases [22]. In addition, some
enzymes, such as carbamate hydrolases from bacteria,
also exist in multimeric forms [23]. Therefore, it is not
surprising to find that purified BH from L. maculans is
a tetramer and that from A. brassicicola is a dimer
with apparent molecular mass values of 220 and
120 kDa, respectively.
To obtain information on the nature of the amino
acid residues occurring in the active site of BHs, pro-
tein-modifying reagents were used. The chemical modi-
fication of BH with selective reagents was carried out
by incubating the enzyme with a large excess of
reagent. The reaction conditions for the modification
of Asp, Glu, Cys and Ser are shown in Table 5. No
significant inactivation of BHs from either L. maculans
or A. brassicicola was observed upon treatment with
reagents specific for Asp and Glu [Woodward’s reagent
K and N-(3-dimethylaminopropyl)-N¢-ethylcarbodii-
mide] or for Cys (iodoacetamide). However, treatment
of BHs with a reagent specific for Ser (phen-
ylmethanesulfonyl fluoride) resulted in 55% and 51%
loss of the initial activities of BHLmL2 and BHAb,

respectively. These results strongly suggest that Ser is
involved in the catalytic activity of BHLmL2 and
BHAb and are consistent with our sequence analyses
indicating their high similarity to amidases containing
the Ser⁄ Ser⁄ Lys catalytic triad. The amidase-signature
domain is approximately 130 residues in length and
includes the conserved motif with the active-site Ser ⁄
Ser ⁄ Lys residues in which Ser is the nucleophilic
residue. Amidase-signature enzymes represent a large
family of nonclassical serine hydrolases that are wide-
spread in nature, exhibit very diverse biological
functions and use amides as substrates [16–18,24,25].
Kinetic properties, effects of metal ions, pH
optima and temperature
The substrate saturation curves of both BH enzymes
were determined in the presence of increasing concen-
trations of brassinin (Fig. 4), and the corresponding
kinetic parameters, calculated on the basis of the Hill
equation, are summarized in Table 6 (S
0.5
0.27 ±
0.02 mm, Hill coefficient was 1.6 ± 0.2 for BHLmL2,
Fig. 4A; S
0.5
0.24 ± 0.01 mm, Hill coefficient 1.4 ±
0.1, for BHAb, Fig. 4B). As shown in Table 6 and
Fig. 4, the Hill equation provided an excellent fit, with
coefficients of determination of 0.992 for BHLmL2
and 0.997 for BHAb. In addition, the fits of the
V = f(S) curves were sigmoidal and V ⁄ S = f(S) was

not a straight line (Fig. 4), suggesting kinetic charac-
teristics of allosteric enzymes, rather than Michaelis–
Menten kinetics. Overall, the kinetic parameters of the
BH enzymes were similar and exhibited slightly
positive cooperativity for the substrate brassinin.
The effect of metal ions, such as Mn
2+
,Ca
2+
,
Co
2+
,Ni
2+
and Zn
2+
, on the enzyme activity of BHs
was examined. No changes in enzyme activity were
found in the presence of these metal ions for either
BHLmL2 or BHAb. In addition, no effect of EDTA
kDa
A
B
225
150
100
75
50
35
25

15
225
150
100
75
50
35
25
Mr 1 2 3 4 5
220 kDa
120 kDa
60 kDa
58 kDa
kDa Mr 1 2 3 4
Fig. 3. SDS ⁄ PAGE of fractions and of purified BH enzymes. (A)
From L. maculans: lane M, marker proteins (molecular mass values
are indicated); lane 1, crude homogenate (40 lg); lane 2, HiTrap-
DEAE-FF pooled fractions (20 lg); lane 3, hydroxyapatite chroma-
tography (20 lg); lane 4, Superdex 75-pooled fractions (2 lg); lane
5, fraction (BHLmL2) after chromatography on Superdex 200
(1.5 lg). (B) From A. brassicicola: lane M, marker proteins (molecu-
lar mass values are indicated); lane 1, crude homogenate (30 lg);
lane 2, Phenyl Sepharose pooled fractions (20 lg); lane 3, hydroxy-
apatite chromatography (3 lg); lane 4, fraction (BHAb) after
chromatography on Superdex 200 (1 lg).
Brassinin hydrolases M. S. C. Pedras et al.
7416 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
was found on enzyme activity. Thus, these results sug-
gest the absence of divalent metal cations in the active
site of BHs. Both enzymes were inhibited by 1.0 mm

of 2-mercaptoethanol and dithiothreitol. For example,
1mm dithiothreitol caused 72% and 85% inhibition of
BHLmL2 and BHAb, respectively, while 1 mm 2-mer-
captoethanol caused 85% and 97% inhibition of
BHLmL2 and BHAb, respectively.
The influence of pH on the activities of the BH
enzymes was investigated in the pH range 6–11. The
pH optima were determined to be in the basic range
(pH 8.0–10.0) for BHLmL2 and BHAb (Fig. 5A,B).
Overall, the kinetic properties and pH optimum pro-
files of BHLmL2 and BHAb were comparable to those
reported for amidases and carbamate hydrolases
[23,24,26]. In fact, a general base lysine residue is used
in amidases that contain a Ser ⁄ Ser ⁄ Lys triad active site
such as FAAH [24]. The pH rate profiles of FAAH
indicated an increase in activity from pH 5–9, reveal-
ing a titratable group with a pK
a
of approximately 7.9,
similar to those profiles observed for BHLmL2 and
BHAb.
The temperature dependence of the activities of the
BHs was tested in the range 5–56 °C; the apparent
optimum temperatures of BHLmL2 were 25–30 °C
Table 3. Masses and scores of tryptic peptides obtained from purified BHLmL2. Observed = mass ⁄ charge of observed peptide; Mr
(expt) = observed mass of peptide; Mr (calc) = calculated mass of matched peptide, Delta = difference (error) between the experimental
and calculated masses; Score = ions score. The peptide shown in bold has 100% identity with a putative amidase from Sinorhizobium medi-
cae WSM419 (accession no. YP_001314042).
Observed Mr (expt) Mr (calc) Delta Score Peptide
400.9585 1199.8536 1199.5604 0.2932 20 SNYASIMTSAR

401.0218 1200.0435 1199.5856 0.4579 15 LYGSSGISSMAK
430.2589 858.5032 858.5287 )0.0255 25 VISGLSKR
537.2736 1072.5326 1072.5513 )0.0186 29 ELATQAGDLR
584.8255 1167.6364 1167.6798 )0.0434 42 MVPPLTINKR
608.8271 1215.6397 1215.6897 )0.0500 32 VIEMLIQDK
645.3196 1288.6246 1288.7099 )0.0853 33 NRSTVKEGSALK
667.3418 1332.6691 1332.7513 )0.0822 51 FLLEATKERAR
832.4039 1662.7932 1662.8359 )0.0427 40 R.LSASEAMEASLAATRR
844.4123 1686.8100 1686.8974 )0.0874 18 NLSLAELSLLAEQMR
846.4120 1690.8095 1690.8559 )0.0464 28 QSKTAATMAAEAAELAK
Table 4. Masses and scores of tryptic peptides obtained from purified BHAb. Observed = mass ⁄ charge of observed peptide; Mr
(expt) = observed mass of peptide; Mr (calc) = calculated mass of matched peptide, Delta = difference (error) between the experimental
and calculated masses; Score = ions score. The peptide shown in bold has 100% identity with putative amidases from Aspergillus oryzae
RIB40 (accession no. XP_001825134), Aspergillus nidulans FGSC A4 (accession no. XP_682046), Emericella rugulosa (accession no.
AAK29061) and Emericella unguis (accession no. AAK29062).
Observed Mr (expt) Mr (calc) Delta Score Peptide
400.7445 799.4744 799.4552 0.0192 28 GTIGLSPR
416.8871 1247.6395 1247.6986 )0.0591 19 YEALKARLER
428.7643 855.5139 855.4814 0.0326 35 IAAELSPR
443.5553 1327.6440 1327.6844 )0.0404 28 AERASIDGLGPSR
480.7742 959.5338 959.5287 0.0051 30 LDAGVASSLK
550.7971 1099.5797 1099.6032 )0.0235 65 RRPNMAAIR.R
567.2875 1132.5604 1132.6604 )0.1000 47 KLFGLESALR
658.8793 1315.7440 1315.6844 0.0596 42 DATGDKGLRIDR
716.7034 2147.0883 2147.0715 0.0168 32 K.TSVPQTLMVCETVNNIIGR
Table 5. Effect of modifying reagents on relative specific activities
of BHs. EDC, N-(3-dimethylaminopropyl)-N¢-ethylcarbodiimide; IAA,
iodoacetamide; PMSF, phenylmethanesulfonyl fluoride; WRK,
Woodward’s reagent K.
Modifying

reagent
Possible amino acid
residues modified
Specific relative
activity (100%)
BHLmL2 BHAb
WRK (10 m
M) Asp, Glu 95 ± 5 93 ± 4
EDC (10 m
M) Asp, Glu 133 ± 8 115 ± 5
IAA (5 m
M) Cys 100 ± 3 95 ± 3
PMSF (5 m
M) Ser 55 ± 2 51 ± 2
M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7417
(Fig. 6A) and of BHAb were 20–27 °C (Fig. 6B). The
activation energies of BHs were calculated using the
Arrhenius equation after determining enzyme activities
at 5, 10, 15 and 22 °C. The activation energies were 12
and 13 kJÆmol
)1
for BHLmL2 and BHAb, respectively.
These results indicated that BHs do not show substan-
tial differences with respect to the apparent optimal
temperatures and the activation energies.
Fig. 4. Substrate saturation curves of BHs. (A) Brassinin saturation
curve for BHLmL2 and (B) brassinin saturation curve for BHAb. The
mixture was incubated at 23 °C for 45 min in the presence of
increasing concentrations of brassinin (0–1 m

M). The curves
obtained were fitted to the Hill equation using KaleidaGraph. Inset,
corresponding Eadie plots.
Table 6. Kinetic parameters of the BHLmL2 and BHAb (kinetic
parameters were obtained from the saturation curves presented in
Fig. 4 fitted to the Hill equation. Standard deviation values were
obtained from this fit).
Source of BH (brassinin
hydrolase)
Kinetic parameters
V
max
(nmolÆmin
)1
) S
0.5
(mM) n
H
BHLmL2
Leptosphaeria maculans
0.49 ± 0.02 0.27 ± 0.02 1.6 ± 0.2
BHAb
Alternaria brassicicola
0.42 ± 0.01 0.24 ± 0.01 1.4 ± 0.1
Fig. 5. pH dependence of BHLmL2 (A) and BHAb (B) activities.
The enzyme activities were measured using protein extracts
obtained after the second purification step.
Brassinin hydrolases M. S. C. Pedras et al.
7418 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
Substrate specificities of purified enzymes

The substrate specificity of the purified enzymes was
tested using various synthetic compounds containing a
dithiocarbamate group or isosteres located at C-3 of
indolyl-3-methyl or naphthyl-1 or 2-methyl moieties.
As summarized in Table 7, BHLmL2 and BHAb
showed hydrolytic activity towards brassinin (1),
1-methylbrassinin (4), methyl tryptaminedithiocarba-
mate (8) and methyl tryptopholdithiocarbonate (16 ).
Brassinin (1) was the best substrate for both BHLmL2
and BHAb; however, BHAb exhibited relatively higher
activity (of about twofold) towards 1-methylbrassinin
(4) than towards BHLmL2. By contrast, the rates of
hydrolysis of methyl tryptopholdithiocarbonate (16)
and methyl tryptaminedithiocarbamate (9) catalysed
by BHLmL2 were substantially higher than those
catalysed by BHAb. In addition, no catalytic activities
were observed with thiolcarbamate (10), carbamate
(11), urea (12), thiourea (13) or amide (15), indicating
that both BHs are functional group specific. Moreover,
because naphthyldithiocarbamates (18 and 20) were
not transformed, it appeared that the indole group is
also important for catalysis. Finally, the ethyl dithio-
carbamate (14) (a homologue of brassinin containing
only an additional CH
2
) was also not transformed,
which indicates that the hydrophobic pocket binding
the methylthiol group of brassinin is rather small and
hence could not accommodate the additional CH
2

group. It is worthy of note that, similarly to BHLmL2,
L. maculans L2 ⁄ M2 was able to transform methyl
tryptaminedithiocarbamate (8) to tryptamine (9) and
methyl tryptopholdithiocarbonate (16) to tryptophol
(17). Moreover, the rates of transformation of both
compounds 8 and 16 in fungal cultures were substan-
tially slower than the rates observed for transformation
of brassinin (1) [14]. Altogether, these direct correla-
tions between BH activities and corresponding
biotransformations in fungal cultures appear to suggest
that cells of L. maculans L2 ⁄ M2 produce only one type
of dithiocarbamate hydrolase activity. Obviously, this
information is of great importance to the design of
potential paldoxins.
Overall, both BHs showed new substrate specifici-
ties, because they were only able to hydrolyse the
dithiocarbamate functional group (–HN–C(=S)–
SCH
3
) of brassinin and 1-methylbrassinin, but none of
the BHs exhibited activity towards the brassinin ana-
logue methylated at the side-chain (–CH
3
N–C(=S)–
Fig. 6. Temperature dependence of BHLmL2 (A) and BHAb (B)
activities. The enzyme activities were measured using protein
extracts obtained after the second purification step.
Table 7. Relative specific activities of BHs towards the phytoalexin
brassinin (1, natural substrate) and synthetic substrates 1-methyl-
brassinin (4), methyl tryptaminedithiocarbamate (8) and methyl tryp-

topholdithiocarbonate (16).
Substrate ⁄ compound
name (number)
Relative
specific
activity
a
(%)
of BHLmL2
Relative
specific
activity
a
(%)
of BHAb
Brassinin (1) 100 100
1-Methylbrassinin (4) 24±2 50±4
Methyl tryptaminedithiocarbamate (8)15±2 2±1
Methyl tryptopholdithiocarbonate (16)19±2 5±2
a
Activities are expressed as the percentage of activity compared
to the substrate activity obtained with brassinin (1.0 m
M; 100% of
activity for BHLmL2 is equivalent to 0.15 nmolÆmin
)1
and for BHAb
is equivalent to 0.10 nmolÆmin
)1
). The results are expressed as
means and standard deviations of four independent experiments.

M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7419
SCH
3
, dithiocarbamate compound 6) or isosteric
groups (e.g. thiolcarbamate and carbamate) except for
the dithiocarbonate 16 (–O–C(=S)–SCH
3
). Amidases,
which act on carbon–nitrogen bonds (EC 3.5.), and
esterases, which act on carbon–oxygen bonds (EC
3.1.), are enzymes with hydrolytic activities similar to
those of BHs, but to the best of our knowledge no
dithiocarbamate hydrolases have been reported to
date. BHLmL2 and BHAb are thus the first members
of this new group of enzymes within the amidase
superfamily (EC 3.5).
Identification of the volatile products and
chemical mechanism of BH catalysed reactions
Our previous studies of the biotransformation of brass-
inin (1) in liquid cultures of L. maculans (isolates avir-
ulent on canola, now reclassified as L. biglobosa [27])
revealed the presence of carbonyl sulphide (COS) and
methanethiol (CH
3
SH) in headspace volatiles, suggest-
ing that both products originated from dithiocarba-
mate hydrolysis. Consequently, it was suspected that,
in addition to amine (3), these volatiles were products
of the enzymatic transformation of brassinin (1)by

BHs. To identify these volatile reaction products, BHs
were incubated with brassinin in tightly closed vials, as
described in the Experimental procedures. A gas-tight
syringe was used to collect the headspace volatiles in
the vial and to inject them into a GC ⁄ high-resolution
mass spectrometry (GC ⁄ HRMS) instrument. Two
peaks with retention times of 6.0 and 12.5 min were
identified as carbonyl sulphide (O=C=S) and metha-
nethiol (CH
3
SH), respectively. Thus, these analyses
indicated that brassinin was enzymatically transformed
into 3-indolylmethanamine (3), carbonyl sulphide and
methanethiol.
The chemical mechanism of dithiocarbamate hydro-
lysis catalysed by BHs is expected to be similar to the
hydrolyses of amides and esters catalysed by amidases
and ⁄ or esterases, as depicted in Fig. 7. First, the sub-
strate binds covalently to the active site of hydrolase
via the hydroxyl group of Ser, yielding a first tetrahe-
dral intermediate stabilized by other amino acid resi-
due(s). Next, the free amine is released and a
dithiocarbonate–enzyme complex is formed. Finally,
nucleophilic attack of water on the thiocarbonyl
carbon of the enzyme complex forms a second tetrahe-
dral intermediate, which then releases the products
carbonyl sulphide and methanethiol, and regenerates
the free enzyme.
Effect of phytoalexins on BH activities
To identify potential inhibitors of BHs, inhibition

experiments were carried out using the purified
enzyme, brassinin (1) (0.10 mm final concentration)
and the phytoalexins brassicanal A, erucalexin, rutalex-
in, brassilexin, camalexin (chemical structures in
Fig. S2) and cyclobrassinin [7]. Interestingly, only
cyclobrassinin (Fig. 8A) exhibited an inhibitory effect
on both BHs. The type of inhibition caused by cyclo-
brassinin was determined from the kinetics of inhibi-
tion of BHs, shown in the form of Lineweaver–Burk
double reciprocal plots (Fig. 8B,C). These results
showed that the intersection points of all curves for
both BHs were on the 1 ⁄ V axis, strongly suggesting
that cyclobrassinin competitively inhibited the BH
activities of both enzymes. The K
i
values were deter-
N
H
H
N
S
SCH
3
3
1
Active site
N
H
H
N

S
SCH
3
O
Active site
O
N
H
NH
2
S
H
3
CS
+
Active site
OH
+H
2
O
C
O
S
1st tetrahedral intermediate
H
Ser
:B
Ser
HB
O

Active site
Ser
:B
:B
Ser
S
H
3
CS
O
Active site
Ser
:B
O
CH
3
SH
2nd tetrahedral intermediate
H
H
+
+
Fig. 7. Proposed chemical mechanism of
catalytic hydrolysis of brassinin (1)by
BHLmL2 and BHAb.
Brassinin hydrolases M. S. C. Pedras et al.
7420 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
mined to be 0.14 ± 0.02 mm for BHLmL2 and
0.41 ± 0.08 mm for BHAb.
Previously we found that both phytoalexins (cyclo-

brassinin and camalexin) competitively inhibited brass-
inin oxidase [13]. This inhibitory effect was thought to
be a result of the structural resemblance of each phyto-
alexin to two different intermediates in the oxidative
transformation mediated by brassinin oxidase. In this
work, the inhibitory effect of cyclobrassinin on the
activity of both BHs is probably caused by its struc-
tural resemblance to the substrate brassinin (1). Fur-
thermore, based on the mechanism proposed for the
hydrolysis of brassinin, it was not surprising to find
that camalexin had no inhibitory effect.
Conclusion and prospects for paldoxin application
In this work, we have purified and characterized
BHLmL2 and BHAb, two brassinin detoxifying
enzymes that exhibit BH activity. BHs are enzymes
produced by L. maculans and A. brassicicola, and
which require induction with specific compounds such
as 3-phenylindole and camalexin. Importantly, it was
demonstrated that both BHs were inhibited by the
phytoalexin cyclobrassinin. This discovery lends fur-
ther support to the hypothesis that phytoalexins have
multiple physiological roles in plant protection, which
include inhibition of microbial growth and detoxifying
enzymes produced by fungal plant pathogens [13].
Cyclobrassinin is biosynthetically derived from brassi-
nin, and both phytoalexins co-occur in various culti-
vated Brassica species [7]. To date, two other brassinin
detoxifying enzymes have been reported: BO, isolated
from L. maculans isolates virulent on canola [13]; and
brassinin glucosyl transferase (BGT), produced by the

fungal phytopathogen Sclerotinia sclerotiorum
(SsBGT1) [28]. Similarly to BHs, the activities of both
BO and BGT were inducible by 3-phenylindole and
camalexin. BO was isolated from the wild-type fungus
and found to be competitively inhibited by the crucif-
erous phytoalexins cyclobrassinin and camalexin.
Furthermore, recent results have shown that 5-meth-
oxycamalexin, a synthetic compound, was the most
effective inhibitor of BO [29]. Recombinant SsBGT1
was isolated from Saccharomyces cerevisiae after the
corresponding gene of S. sclerotiorum was cloned. The
relatively low expression levels of cloned BGT did not
allow inhibition studies to be carried out [28]; however,
BGT activity in crude cell-free homogenates of S. scle-
rotiorum was strongly inhibited by 3-phenylindole and
6-fluoro-3-phenylindole [30].
Knowledge that the Ser ⁄ Ser ⁄ Lys catalytic triad is
probably involved in the catalytic activity of both
BHLmL2 and BHAb will greatly facilitate the design
of inhibitors for both enzymes. In particular, the devel-
opment of mechanism-based inhibitors is anticipated
because inactivation of the hydroxyl group of Ser, the
probable nucleophile, has ample precedents. For
N
H
S
N
SCH
3
Cyclobrassinin

A
B
C
Fig. 8. Lineweaver–Burk plots of BH activities for (A) BHLmL2 and
(B) BHAb in the presence of the phytoalexin cyclobrassinin.
Enzyme activity was determined as described in the Experimental
procedures.
M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7421
instance, mechanism-based inhibitors of the amidase-
signature enzyme FAAH, and of other pharma-
ceutically important drugs, have been successfully
developed [31–33]. In conclusion, taken together, the
current findings suggest that inhibitors of brassinin
detoxifying enzymes, such as BHs, can be designed
using cyclobrassinin as the lead structure. It is hoped
that such inhibitors will offer sustainable alternatives
for crop protection against specific fungal phytopatho-
gens. The availability of BHs from two important
plant pathogens will greatly enhance our ability to
screen for cyclobrassinin-based paldoxins that could
protect plants against blackleg and black spot fungi.
Experimental procedures
General procedures and HPLC analysis
Brassinin (1), indolyl-3-methanamine (3), 1-methyl-
brassinin (4), N¢-methylbrassinin (6), N¢-methylindolyl-
3-methanamine (7), methyl tryptaminedithiocarbamate
(8), tryptamine (9), brassitin (10), methyl N¢-(3-indo-
lylmethyl)carbamate (11), N ¢-(3-indolylmethyl)-N¢¢-me-
thylurea (12), N ¢-(3-indolylmethyl)-N¢¢-methylthiourea

(13), ethyl N¢-(3-indolylmethyl)dithiocarbamate (14),
N¢-acetylindolyl-3-methanamine (15), methyl tryptop-
holdithiocarbonate (16), methyl 1-naphthylmethyl-
dithiocarbamate (17), 1-naphthylmethanamine (18),
methyl 2-naphthylmethyldithiocarbamate ( 19) and
2-naphthylmethanamine (20) were synthesized as previ-
ously reported [7,34].
HPLC analysis was carried out using an HPLC sys-
tem equipped with a quaternary pump, an automatic
injector and a diode array detector (wavelength range
190–600 nm), and a degasser and a column equipped
with an in-line filter. HPLC method A was used for
substrate and product analysis, excluding amines, as
follows: a Hypersil ODS column (5-lm particle-size sil-
ica, 200 · 4.6 mm internal diameter), a linear gradient
of H
2
O-CH
3
CN (75 : 25) to H
2
O-CH
3
CN (0 : 100) for
35 min and a flow rate of 1.0 mLÆmin
)1
. Previously,
the product of the enzymatic reaction of brassinin, ind-
olyl-3-methanamine (3), was determined indirectly by
reverse-phase HPLC after acetylation of brassinin with

acetic anhydride [14,15]. However, that indirect detec-
tion of 3-indolylmethanamine (as the resultant acetyl
amine) was rather time consuming and more prone to
error, because quantitative acetylation of the amine
occurred only in the absence of water. Hence, a direct
HPLC analytical method using a cyano column [35]
and elution under normal phase conditions (method B)
was developed for direct detection and quantification
of all amines [indolyl-3-methanamine (3), N¢-methy-
lindolyl-3-methanamine (7), tryptamine (9), 1-naph-
thylmethanamine (18) and 2-naphthylmethanamine
(20)]. HPLC analysis of amines, method B, was per-
formed as follows: a Zorbax XDB-CN (5-lm particle-
size silica, 150 · 4.6 mm internal diameter) column,
and isocratic elution with hexane-methanol-CH
2
Cl
2
(30 : 10 : 60), containing 0.2% n-propylamine, for
10 min at a flow rate of 1.0 mLÆmin
)1
. Calibration
curves were built using both methods A and B at vari-
ous wavelengths. The retention times of substrates and
potential reaction products, and minimum detection
limits, are shown in Table 8.
Fungal cultures
Fungal spores from the L. maculans virulent isolate
Laird 2 were grown on V8 agar under continuous light
at 23 ± 1 ° C; after 15 days, spores were collected and

stored at )20 °C [36]. Liquid cultures were initiated by
inoculating 250-mL Erlenmeyer flasks containing mini-
mal media (100 mL) with fungal spores at a concentra-
tion of 10
6
spores per mL; the culture-containing flasks
were then incubated on a shaker under constant light
at 23 ± 1 °C. After 74 h of incubation, liquid cultures
were induced with 3-phenylindole (0.20 mm). The cul-
tures were incubated for an additional 24 h and then
gravity filtered to separate the mycelia from the culture
broth.
A. brassicicola isolate ATCC 96866 was subcultured
in V8-agar plates under constant light at room temper-
ature for 10 days [15]. Cultures of A. brassicicola were
initiated by inoculating spores (10
6
spores per flask) in
250-mL Erlenmeyer flasks containing 100 mL of mini-
mal medium plus thiamine and incubated on a shaker
at 120 rpm, at 23 ± 1 °C. After 72 h, liquid cultures
were induced with 3-phenylindole (0.20 mm), and the
cultures were incubated for an additional 24 h and
then gravity filtered to separate the mycelia from the
broth. For analysis of BH induction, 72-h-old liquid
cultures (20 mL) were co-incubated with camalexin or
3-phenylindole (the final concentrations in culture were
0.012, 0.025, 0.050, 0.125 and 0.250 mm) and after fur-
ther incubation for 24 h, the mycelia were separated
from the culture broth by filtration and used for BH

activity assays.
GC–MS analysis
GC-MS analysis was carried out using a GC 8060 mass
spectrometer coupled to a VG-70SE magnetic sector
mass spectrometer operated at a mass resolution of
5000, using electron-impact ionization performed at an
electron energy of 70 eV. Chromatographic separation
Brassinin hydrolases M. S. C. Pedras et al.
7422 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
Table 8. Chemical structures of compounds tested as substrates of BHLmL2 and BHAb and potential products, references for synthesis,
and HPLC retention times using methods A and B described in the Experimental procedures (n.d. = not detected). Only boxed reactions are
catalyzed by BHLmL2 and BHAb.
Substrate ⁄ compound name (#) [reference
for synthesis] (HPLC method, retention
time) chemical structure
Potential product chemical structure,
name (#) [reference for synthesis]
(HPLC method, retention time)
Brassinin (1) [7]
(A, 18.6 min; B, 2.0 min)
N
H
H
N
S
SCH
3
N
H
NH

2
indolyl-3-methanamine (3) [7]
(A, n.d.; B, 5.7 min)
1-Methylbrassinin (4) [34]
(A, 24.6 min; B, 2.0 min)
N
H
N
S
SCH
3
CH
3
N
NH
2
CH
3
1-methylindolyl-3-methanamine (5) [7]
(A, n.d.; B, 3.3 min)
N¢-Methylbrassinin (6) [34]
(A, 24.3; B, 2.0 min)
N
H
N
S
SCH
3
CH
3

N
H
NH
CH
3
N¢-methylindolyl-3-methanamine (7) [7]
(A, n.d.; B, 4.1 min)
Methyl tryptaminedithiocarbamate (8) [7]
(A, 20.1; B, 2.0 min)
N
H
NH
S
SCH
3
N
H
NH
2
tryptamine (9) (A, n.d.; B, 5.0 min)
Brassitin (10) [34] (A, 11.3 min; B, 2.0 min)
N
H
NH
O
SCH
3
N
H
NH

2
indolyl-3-methanamine (3) [7]
(A, n.d.; B, 5.7 min)
Methyl N¢-(3-indolylmethyl)carbamate (11) [34]
(A, 11.5 min; B, 2.0 min)
N
H
NH
O
OCH
3
N
H
NH
2
indolyl-3-methanamine (3) [7]
(A, n.d.; B, 5.7 min)
N¢-(3-Indolylmethyl)-N¢¢-methylurea (12) [34]
(A, 5.3; B, 2.0 min)
N
H
NH
O
NHCH
3
N
H
NH
2
indolyl-3-methanamine (3) [7]

(A, n.d.; B, 5.7 min)
N¢-(3-Indolylmethyl)-N¢¢-methylthiourea (13) [34]
(A, 8.5 min; B, 2.0 min)
N
H
NH
S
NHCH
3
N
H
NH
2
indolyl-3-methanamine (3) [7]
(A, n.d.; B, 5.7 min)
Ethyl N¢-(3-indolylmethyl)dithiocarbamate (14) [34]
(A, 20.3 min; B, 2.0 min)
N
H
NH
S
SCH
2
CH
3
N
H
NH
2
indolyl-3-methanamine (3) [7]

(A, n.d.; B, 5.7 min)
N¢-acetylindolyl-3-methanamine (15)
(A, 6.0 min; B, 2.0 min)
N
H
H
N
CH
3
O
N
H
NH
2
indolyl-3-methanamine (3) [7]
(A, n.d.; B, 5.7 min)
Methyl tryptopholdithiocarbonate (16) [34]
(A, 28.0 min; B, 2.0 min)
N
H
O
S
SCH
3
N
H
OH
tryptophol (17) (A, 6.7 min)
Methyl 1-naphthylmethyldithiocarbamate (18) [34]
(A, 23.9 min; B, 2.0 min)

H
N
SCH
3
S
NH
2
1-naphthylmethanamine (19) [34]
(A, n.d.; B, 2.3 min)
Methyl 2-naphthylmethyldithiocarbamate (20) [34]
(A, 24.2 min; B, 2.0 min)
NH
SCH
3
S
NH
2
2-naphthylmethanamine (21) [34]
(A, n.d.; B, 2.4 min)
M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7423
was achieved in a GS-Q capillary column (30 m ·
0.546 mm), with oven temperature at 40 °C for 5 min,
then increased to 200 °C for 10 min and kept at 200 °C
for 5 min; the injection volume was 1.0 mL, the carrier
gas was helium, and the flow rate was 1 mLÆmin
)1
.
For GC ⁄ HRMS analysis of the headspace volatiles of
enzymatic reactions, samples were prepared as follows:

1 mL of the reaction mixture in a vial tightly closed with
a screw cap and silicon septum containing brassinin
(1.0 mm in 5 lL of dimethylsulfoxide) and 100 lLof
protein extract was incubated at 23 °C for 24 h. Head-
space volatiles were withdrawn from the vial using a
gas-tight syringe and injected into the GC-MS injector.
Preparation of protein extracts
Frozen mycelia (8–10 g) obtained from L. maculans (1
l) or A. brassicicola (0.8 l) were suspended in ice-cold
extraction buffer (10 mL) and ground (using mortar
and pestle) for 5 min. The extraction buffer consisted
of 25 mm diethanolamine (DEA, pH 8.3) 5% (v ⁄ v)
glycerol, 1 mm EDTA and 1 : 200 (v ⁄ v) protease
inhibitor cocktail (P-8215; Sigma, Mississauga,
Canada). The suspension was centrifuged for 60 min
at 58 000 g . The resulting supernatant (10 mL) was
used for chromatographic analyses.
Chromatographic purification of enzyme
exhibiting BH activity from L. maculans isolate
L2
⁄
M2
An A
¨
kta FPLC (GE Healthcare, Baie d¢Urfe
´
, Canada),
equipped with a P-920 pump, UPC-900 UV absorbance
monitor and a Frac950 fraction collector, and stored at
5 °C, was used for all protein separations.

Step 1: the soluble protein extract from mycelia
(10 mL) was equilibrated by dialysis against 20 mm
DEA buffer (pH 8.3), 0.015% (v ⁄ v) Triton X-100 and
2% (v ⁄ v) glycerol and then loaded onto a HiTrap
DEAE FF (GE Healthcare) anion-exchange column
(5 mL). Proteins were eluted with 20 mL of the same
buffer, first alone and then with a 60-mL 0.0–0.45 m
NaCl gradient. Fractions of 1.8 mL were collected at a
flow rate of 0.5 mLÆmin
)1
, and 100 lL was assayed for
BH activity. Fractions (15–17) showing BH activity were
pooled and used in the second step of purification.
Step 2: the protein extract (5.0 mL) obtained from
step 1 was equilibrated by dialysis against 5 mm potas-
sium phosphate buffer (pH 6.8), 0.015% (v ⁄ v) Triton
X-100 and 2% (v ⁄ v) glycerol. The protein extract was
loaded onto a hydroxyapatite column (1.6 · 6.0 cm)
(Bio-Rad). Proteins were eluted using a linear gradient
of sodium phosphate (0.005–0.30 m) in the same
buffer. Fractions of 2 mL were collected at a flow rate
of 0.5 mL Æ min
)1
and 100 lL was assayed for BH activ-
ity. Fractions 26–28 were pooled and used in the third
step of purification.
Step 3: pooled fractions showing BH activity after
step 2 were concentrated to 1.5 mL (Amicon Ultra 15,
10 kDa; Millipore, Etobicoke, Canada) and fraction-
ated on a Superdex 75 HR16 ⁄ 60 column. Equilibration

and elution were performed at 8 °C with 20 mm DEA
(pH 8.3), 0.015% (v ⁄ v) Triton X-100, 1% (v ⁄ v)
glycerol and 0.15 m NaCl. Fractions of 1.5 mL were
collected at a flow rate of 0.6 mLÆmin
)1
, and 50 lL
of each fraction was assayed for BH activity. A peak
of BH activity was observed for fractions 30–32.
Step 4: pooled fractions showing BH activity after
step 3 were concentrated to 500 lL (Amicon, Bedford,
MA, USA) and fractionated using a Superdex 200
HR10 ⁄ 30 column, precalibrated with the following
markers of known molecular mass: Blue dextran
(2000 kDa), apoferritin (443 kDa), b-amylase
(200 kDa), alcohol dehydrogenase (150 kDa), albumin
(67 kDa), ovalbumin (43 kDa) and carbonic anhydrase
(29 kDa). Equilibration and elution were performed at
8 °C using 20 mm DEA (pH 8.3), 0.015% (v ⁄ v) Triton
X-100, 1% (v ⁄ v) glycerol and 0.15 m NaCl. Fractions of
0.4 mL were collected at a flow rate of 0.4 mLÆmin
)1
,
and 20 lL of each fraction was assayed for BH activity.
Fractions 24–26 were pooled and concentrated (Nano-
sep 10K Omega filters; Pall Corporation, New York,
NY, USA) to 120 lL and then used for biochemical
analysis.
Chromatographic purification of enzyme
exhibiting BH activity from Alternaria brassicicola
isolate ATCC 96866

The A
¨
kta FPLC described above was used for all
protein separations.
Step 1: the soluble protein extract from mycelia
(7 mL) was equilibrated by dialysis against 1.5 m
ammonium sulphate and 20 mm Tris ⁄ HCl buffer (pH
8.0) and was then loaded onto a phenyl-Sepharose
CL-4B column (1.6 · 8 cm). Proteins were eluted with
55 mL of the same buffer, first alone and then with a
200 mL of a decreasing gradient of ammonium sul-
phate in 20 mm Tris⁄ HCl buffer (pH 8.0) with a flow
of 0.8 mLÆmin
)1
. Fractions of 2.5 mL were collected
and 100 lL was assayed for BH activity. Fractions
65–67, showing BH activity, were pooled and used in
the second step of purification.
Step 2: the protein extract (6.7 mL) obtained from
step 1 was equilibrated by dialysis against 5 mm
potassium phosphate buffer (pH 6.8) and 0.015%
Brassinin hydrolases M. S. C. Pedras et al.
7424 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
(v ⁄ v) Triton X-100. The protein extract was loaded
onto a hydroxyapatite column (1.6 · 6.0 cm) (Bio-
Rad). Proteins were eluted using a linear gradient of
sodium phosphate (0.005–0.30 m) in the same buffer
with a flow rate of 0.5 mLÆmin
)1
. Fractions of 3.5 mL

were collected and assayed (100 lL) for BH activity.
Fractions 12–14 were pooled, concentrated (Amicon)
and used in the third step of purification.
Step 3: pooled fractions (500 lL) showing BH activ-
ity after step 2 were fractionated on a Superdex 200
HR10 ⁄ 30 column. Equilibration and elution were per-
formed at 8 °Cin20mm DEA (pH 8.3), 0.015% (v ⁄ v)
Triton X-100, 1% (v ⁄ v) glycerol and 0.15 m NaCl at a
flow rate of 0.4 mLÆmin
)1
. Fractions of 0.4 mL were
collected, and 20 lL of each fraction was assayed for
BH activity. Fractions 24–26 were pooled and concen-
trated (Nanosep) to 100 lL, then used for biochemical
analysis.
SDS
⁄
PAGE
Protein-denaturing SDS ⁄ PAGE was carried out using
10% polyacrylamide gels. Standard markers (molecular
mass range 10–225 kDa; V8491; Promega, Madison,
WI, USA) were used to determine the approximate
molecular mass values of purified proteins in gels
stained with Coomassie Brilliant Blue R-250.
Identification of tryptic peptides of BH by liquid
chromatography-electrospray-tandem mass
spectrometry
Analyses were carried out at the Plant Biotechnology
Institute, National Research Council of Canada (Saska-
toon, SK). A protein gel slice was manually excised

from a Coomassie Brilliant Blue-stained gel and placed
in a 96-well microtitre plate. The protein was then auto-
matically destained, reduced with dithiothreitol, alkyl-
ated with iodoacetamide, digested with porcine trypsin
and Q-TOF liquid chromatography-tandem mass
spectrometry (LC-MS ⁄ MS) analysis was carried out
as described previously [13]. LC-MS ⁄ MS data were
processed using mascot distiller (version 2.1.1.0; Matrix
Science, London, UK; available at rix-
science.co.uk). The main search parameters were methi-
onine oxidation as differential modification and trypsin
as enzyme. Protein homology analysis was carried out
using peptide sequences obtained by automated inter-
pretation of the MS ⁄ MS by NCBI blast (http://ca.
expasy.org/tools/blast/). Molecular weight calculations
were carried out using pi ⁄ mw tool software (http://ca.
expasy.org/tools/pi_tool.html). Sequence multiple align-
ments were generated at />services/malign_reduced.html. Peptide sequence align-
ments were carried out using SIM-Alignment Tools
( />Chemical modification of BHs
Chemical modifications of purified BHLmL2 and
BHAb were carried out in 100 mm phosphate buffer
(pH 7.5) and 10 lLofN-ethyl-5-phenylisoxazolium-3¢-
sulphonate (Woodward’s reagent K; 10 mm), N-(3-
dimethylaminopropyl)-N¢-ethylcarbodiimide (10 mm),
iodoacetamide (5 mm) and phenylmethanesulfonyl
fluoride (5 mm). The enzyme activity was measured
and expressed as the percentage relative activity calcu-
lated as the ratio of the specific activity of BH with
chemical modification reagents to that without chemi-

cal modification reagents.
BH activity assays
The reaction mixture contained 20 mm DEA (pH 8.3),
0.1% Triton X-100, 1.0 mm brassinin (in dimethylsulf-
oxide, 5 lL) and 50–100 lL of protein extract in a
total volume of 1 mL. The reaction was carried out at
23 °C for 30–45 min. A control reaction was stopped
by the addition of 0.1 mL of solution containing 20%
(w ⁄ v) ammonium sulphate and 15% (w ⁄ v) ammonium
hydroxide at t = 0. The reaction assays were extracted
with 2 mL of chloroform ⁄ methanol (95 : 5, v ⁄ v) and
concentrated to dryness in a rotary evaporator.
Extracts were dissolved in methanol (200 lL) and ana-
lyzed by HPLC using both methods A and B; quantifi-
cation was carried out using integration of peak areas
of 3-indolylmethanamine and a calibration curve of
indolyl-3-methanamine (3).
Kinetic analysis and protein measurements
Kinetic parameters of the purified enzyme were deter-
mined for the substrate brassinin (1) in a concentration
range of 0.05–1.0 m
m. The standard deviation values
for the assays were < 3%. Kinetic data were pro-
cessed using the kaleidagraph program (Synergy
Software, Reading, PA, USA) based on Hill enzyme
kinetics. Protein concentrations were determined by
the Bradford method [37] using the Sigma prepared
reagent and BSA as the standard.
Effects of divalent metal ions and thiols on BH
activities

The effect of different metal cations on the BH activity
was determined by incubating an appropriate dilution
M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7425
of the enzyme with metal ions (Mn
2+
,Ca
2+
,Co
2+
,
Ni
2+
and Zn
2+
, all used as a chloride salt, 1 mm) for
5 min. The reaction was started by the addition of
brassinin (1 mm), and the reaction mixtures were then
incubated for an additional 40 min at 23 °C. In addi-
tion, the effects of EDTA (1 mm), 2-mercaptoethanol
(0.1 mm) and dithiothreitol (0.1 mm) on the enzyme
activity, using brassinin as the substrate, were investi-
gated.
pH profiles and temperature
To determine the pH optimum, the temperature was
set to 23 °C and the pH was varied from 3.0 to 11.0 in
50 mm diethanolamine, 50 mm N-ethylmorpholine and
100 mm morpholine ethanesulphonic acid buffer.
The determination of temperature dependence of
BH activity was carried out at pH 8.3, as described

above for the BH assay, except that the temperature
ranged from 5 to 56 °C. The activation energy (E
a
)
was determined from the slope ()E
a
⁄ R) of Arrhenius
plots of log V (nmol of product per minute) versus
1 ⁄ T.
Substrate specificity
The substrate specificity of BHs was determined using
the synthetic compounds shown in Table 8 (1, 4, 6, 8,
10–16, 18, 20). Stock solutions (100 mm) of each com-
pound were prepared in dimethylsulfoxide. BH activity
was assayed by mixing 0.1 lg of the purified enzyme
with 1 mm of substrate under experimental conditions
identical to those described for the BH activity assay.
After 45 min of incubation at 23 °C, the reaction was
stopped and extracted as described above. The extracts
were dissolved in methanol (200 lL) and analyzed by
HPLC using both methods A and B, as described in
the section on HPLC analysis. Peaks corresponding to
the product of each enzymatic reaction were integrated
and quantified using calibration curves built for each
product shown in Table 8 (3, 5, 7, 9, 17, 19, 21).
Effect of phytoalexins on BH activities
To determine potential inhibitors of BHs, inhibition
experiments were carried out using the purified
enzyme, brassinin (0.10 mm final concentration) and
the known phytoalexins brassicanal A, erucalexin,

rutalexin, brassilexin, cyclobrassinin and camalexin
(0.10 and 0.30 mm final concentrations). The phyto-
alexin that displayed an inhibitory effect was further
examined in order to determine the type of inhibition
and the K
i
using brassinin (0.10–0.40 mm) as a sub-
strate in the presence of inhibitor (0.10, 0.20 and
0.40 mm). The kinetics of inhibition was transformed
into Lineweaver–Burk double reciprocal plots (1 ⁄ S
versus 1 ⁄ V). To determine the K
i
, the secondary plot
was constructed using the apparent S
0.5
values
obtained in the presence and absence of inhibitor ver-
sus inhibitor concentration. The linear regression was
used to determine the x-axis intercept value that is
equal to the negative K
i
.
Acknowledgements
Financial support from the Natural Sciences and Engi-
neering Research Council of Canada (Discovery Grant
to M.S.C.P.), Canada Foundation for Innovation
(Infrastructure Fund to M.S.C.P.), Canada Research
Chairs program (Research Grant to M.S.C.P.) and the
University of Saskatchewan (Teaching Assistantship to
V.K.S.M.) is gratefully acknowledged. We thank M.

Jha for the synthesis of compounds 4, 6, 8, 12–15, 18
and 20. We acknowledge the technical assistance of P.
B. Chumala, Department of Chemistry, for develop-
ment of the HPLC method B used for amine detection
and K. Thoms, Department of Chemistry, for
GC ⁄ HRMS method development and data acquisition.
References
1 Schranz ME, Song BH, Windsor AJ & Mitchell-Olds T
(2007) Comparative genomics in the Brassicaceae: a
family-wide perspective. Curr Opin Plant Biol 10,
168–175.
2 Talalay P & Fahey JW (2001) Phytochemicals from cru-
ciferous plants protect against cancer by modulating
carcinogen metabolism. J Nutr 131, 3027S–3033S.
3 Shukla Y & Pal SK (2004) Dietary cancer chemopre-
vention: an overview. Int J Hum Genet 4, 265–276.
4 Jeffery EH & Araya M (2009) Physiological effects of
broccoli consumption. Phytochemistry Rev 8, 283–298.
5 Traka M & Mithen R (2009) Glucosinolates, isothiocy-
anates and human health. Phytochemistry Rev 8, 269–
282.
6 Pedras MSC, Jha M & Ahiahonu PWK (2003) The syn-
thesis and biosynthesis of phytoalexins produced by
cruciferous plants. Curr Org Chem 7, 1635–1647.
7 Pedras MSC, Zheng QA & Sarma-Mamillapalle VK
(2007) The phytoalexins from Brassicaceae: Structure,
biological activity, synthesis, and biosynthesis. Nat Prod
Commun 2, 319–330.
8 Bailey JA & Mansfield JW (1982) Phytoalexins. Blackie
& Son, Glasgow, UK, 1–334.

9 Essenberg M (2001) Prospects for strengthening plant
defenses through phytoalexin engineering. Physiol Mol
Plant Pathol 59, 71–81.
Brassinin hydrolases M. S. C. Pedras et al.
7426 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS
10 Pedras MSC (2008) The chemical ecology of crucifers
and their fungal pathogens: boosting plant defenses
and inhibiting pathogen invasion. Chem Rec 8, 109–
115.
11 Szolar OHJ (2007) Environmental and pharmaceutical
analysis of dithiocarbamates. Anal Chim Acta 582, 191–
200.
12 Pedras MSC & Ahiahonu PWK (2005) Metabolism and
detoxification of phytoalexins and analogs by phyto-
pathogenic fungi. Phytochemistry 66, 391–411.
13 Pedras MSC, Minic Z & Jha M (2008) Brassinin
oxidase, a fungal detoxifying enzyme to overcome to
overcome a plant defense: purification, characterization
and inhibition. FEBS J 275, 3691–3705.
14 Pedras MSC, Gadagi RS, Jha M & Sarma-Mamillapalle
VK (2007) Detoxification of the phytoalexin brassinin
by isolates of Leptosphaeria maculans pathogenic on
brown mustard involves an inducible hydrolase. Phyto-
chemistry 68, 1572–1578.
15 Pedras MSC, Chumala PB, Jin W, Islam MS & Hauck
DW (2009) The phytopathogenic fungus Alternaria
brassicicola: phytotoxin production and phytoalexin
elicitation. Phytochemistry 70, 394–402.
16 Shin S, Lee T-H, Ha N-C, Koo HM, Kim S-Y, Lee
H-S, Kim YS & Oh B-H (2002) Structure of malano-

amidase E2 reveals a novel Ser-cisSer-Lys catalytic triad
in a new serine hydrolase fold that is prevalent in nat-
ure. EMBO J 21, 2509–2516.
17 Neumann S, Granzin J, Kula M-R & Labahn J (2002)
Crystallization and preliminary X-ray data of the
recombinant peptide amidase from Stenotrophomonas
maltophilia. Acta Crystallogr Sect D 58, 333–335.
18 Bracey MH, Hanson MA, Masuda KR, Stevens RC &
Cravatt BF (2002) Structural adaptations in a mem-
brane enzyme that terminates endocanabinoid signaling.
Science 298, 1793–1796.
19 Rogers HJ, Taylor C, Rayter S & Ward JB (1984) Puri-
fication and properties of autolytic endo-beta-N-acetyl-
glucosaminidase and the N-acetylmuramyl-L-alanine
amidase from Bacillus subtilis strain 168. J Gen Micro-
biol 130, 2395–2402.
20 Vanderwinkel E, de Vlieghere M, de Pauw P, Cattalini
N, Ledoux V, Gigot D & ten Have JP (1990) Purifica-
tion and characterization of N-acetylmuramoyl-L-ala-
nine amidase from human serum. Biochim Biophys Acta
1039, 331–338.
21 Heumann S, Eberl A, Fischer-Colbrie G, Pobeheim H,
Kaufmann F, Ribitsch D, Cavaco-Paulo A & Guebitz
GM (2009) A novel aryl acylamidase from Nocardia
farcinica hydrolyses polyamide. Biotechnol Bioeng 102,
1003–1011.
22 Yoon J, Oh B, Kim K, Park JE, Wang J, Kim HS &
Kim Y (2003) Modifying the oligomeric state of cyclic
amidase and its effect on enzymatic catalysis. Biochem
Biophys Res Commun 310, 651–659.

23 Chaudhry GR, Mateen A, Kaskar B, Bloda M &
Riazuddin S (2002) Purification and biochemical
characterization of the carbamate hydrolase from
Pseudomonas sp. 50432. Biotechnol Appl Biochem 36,
63–70.
24 Ekici OD, Paetzel M & Dalbey RE (2008) Unconven-
tional serine proteases: variations on the catalytic Ser ⁄
His ⁄ Asp triad configuration. Protein Sci
17, 2023–2037.
25 Holmquist M (2000) Alpha ⁄ beta-hydrolase fold
enzymes: structures, functions and mechanisms. Curr
Protein Pept Sci 1, 209–235.
26 Suzuki Y & Ohta H (2006) Identification of a thermo-
stable and enantioselective amidase from the thermo-
acidophilic archaeon Sulfolobus tokodaii strain 7.
Protein Expr Purif 45, 368–373.
27 Pedras MSC & Taylor JL (1993) Metabolism of the
phytoalexin brassinin by the blackleg fungus. J Nat
Products 56, 731–738.
28 Sexton AC, Minic Z, Cozijnsen AJ, Pedras MS & How-
lett BJ (2009) Cloning, purification and characterisation
of brassinin glucosyltransferase, a phytoalexin-detoxify-
ing enzyme from the plant pathogen Sclerotinia sclero-
tiorum. Fungal Genet Biol 46, 201–209.
29 Pedras MSC, Minic Z & Sarma-Mamillapalle VK
(2009) Synthetic inhibitors of the fungal detoxifying
enzyme brassinin oxidase based on the phytoalexin
camalexin scaffold. J Agric Food Chem 57, 2429–2435.
30 Pedras MSC & Hossain M (2007) Design, synthesis,
and evaluation of potential inhibitors of brassinin

glucosyltransferase, a phytoalexin detoxifying enzyme
from Sclerotinia sclerotiorum. Bioorg Med Chem 15,
5981–5996.
31 Seierstad M & Breitenbucher JG (2008) Discovery and
development of fatty acid amide hydrolase (FAAH)
inhibitors. J Med Chem 51, 7327–7343.
32 Mileni M, Johnson DS, Wang Z, Everdeen DS, Liimat-
ta M, Pabst B, Bhattacharya K, Nugent RA, Kamtekar
S, Cravatt BF et al. (2008) Structure-guided inhibitor
design for human FAAH by interspecies active site con-
version. Proc Natl Acad Sci USA 105, 12820–12824.
33 Ahn K, Johnson DS, Mileni M, Beidler D, Long JZ,
McKinney MK, Weerapana E, Sadagopan N, Liimatta
M, Smith SE et al. (2009) Discovery and characteriza-
tion of a highly selective FAAH Inhibitor that reduces
inflammatory pain. Chem Biol 16, 411–420.
34 Pedras MSC & Jha M (2006) Toward the control of
Leptosphaeria maculans: Design, synthesis, biological
activity, and metabolism of potential detoxification
inhibitors of the crucifer phytoalexin brassinin. Bioorg
Med Chem 14, 4958–4979.
35 Kagan M, Chlenov M, Melnikov S, Greenfield A,
Gross J & Bernotas RC (2008) Optimization of normal-
phase chromatographic separation of compounds with
primary, secondary and tertiary amino groups. J Chro-
matogr A 1194, 80–89.
M. S. C. Pedras et al. Brassinin hydrolases
FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS 7427
36 Pedras MSC & Khan AQ (1996) Biotransformation of
the brassica phytoalexin brassicanal A by the blackleg

fungus. J Agric Food Chem 44, 3403–3407.
37 Bradford MM (1976) A rapid and sensitive method for
quantitation of microgram quantities of protein utilizing
the principle of protein-dye-binding. Anal Biochem 72,
248–254.
Supporting information
The following supplementary material is available:
Fig. S1. Sequence alignments of malonamidase E2
(MAE2) from Bradyrhizobium japonicum and enzymes
from various sources.
Fig. S2. Chemical structures of the phytoalexins tested
for inhibitory activity against BHs from Leptosphae-
ria maculans and Alternaria brassicicola.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
Brassinin hydrolases M. S. C. Pedras et al.
7428 FEBS Journal 276 (2009) 7412–7428 ª 2009 The Authors Journal compilation ª 2009 FEBS