Engineered Citrobacter freundii methionine γ-lyase effectively produces antimicrobial thiosulfinates

Article abstract of DOI:10.1016/j.biochi.2016.07.007

Antimicrobial activity of thiosulfinates in situ produced by mixtures of Citrobacter freundii methionine γ-lyase (MGL) with new substrates, L-methionine and S-(alkyl/allyl)-L-cysteine sulfoxides has been recently demonstrated (Anufrieva et?al., 2015). This opens a way to the rational design of a new biotechnologically relevant antimicrobial drug producer. To increase the efficiency of the enzyme toward sulfoxides, the mutant forms of MGL, with the replacements of active site cysteine 115 with alanine (C115A MGL) and histidine (C115H MGL) were obtained. The replacement of cysteine 115 by histidine results in the loss of activity of the mutant enzyme in the γ-elimination reaction of physiological substrate, whereas the activity in the β-elimination reaction of characteristic substrates persists. However, the catalytic efficiency of C115H MGL in the β-elimination reaction of S-substituted L-cysteine sulfoxides is increased by about an order of magnitude compared to the wild type MGL. The antibacterial activity of C115H MGL mixtures with a number of sulfoxides was assessed against Gram-positive and Gram-negative bacteria. The bacteriostatic effect was more pronounced against Gram-positive than against Gram-negative bacteria, while antibacterial potential proved to be quite similar. Thus, the mutant enzyme C115H MGL is an effective catalyst, in particular, for decomposition of sulfoxides and the pharmacological couples of the mutant form with sulfoxides might be new antimicrobial agents.

Full text of DOI:10.1016/j.biochi.2016.07.007

Biochimie 128-129 (2016) 92e98
Contents lists available at ScienceDirect
Biochimie
journal homepage: www.elsevier.com/locate/biochi
Research paper
Engineered Citrobacter freundii methionine
produces antimicrobial thiosulfinates
g-lyase effectively
Elena A. Morozova, Vitalia V. Kulikova, Alexei N. Rodionov, Svetlana V. Revtovich,
Natalya V. Anufrieva, Tatyana V. Demidkina*
Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Vavilova Str., 32, Moscow, 119991, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
Antimicrobial activity of thiosulfinates in situ produced by mixtures of Citrobacter freundii methionine g-
Received 25 April 2016
Accepted 13 July 2016
Available online 16 July 2016
lyase (MGL) with new substrates, L-methionine and S-(alkyl/allyl)-L-cysteine sulfoxides has been recently
demonstrated (Anufrieva et al., 2015). This opens a way to the rational design of a new biotechnologically
relevant antimicrobial drug producer. To increase the efficiency of the enzyme toward sulfoxides, the
mutant forms of MGL, with the replacements of active site cysteine 115 with alanine (C115A MGL) and
histidine (C115H MGL) were obtained. The replacement of cysteine 115 by histidine results in the loss of
Keywords:
Methionine g-lyase
Citrobacter freundii
Site-directed mutagenesis
Cysteine sulfoxides
activity of the mutant enzyme in the
activity in the -elimination reaction of characteristic substrates persists. However, the catalytic effi-
ciency of C115H MGL in the -elimination reaction of S-substituted -cysteine sulfoxides is increased by
g-elimination reaction of physiological substrate, whereas the
b
b
L
about an order of magnitude compared to the wild type MGL. The antibacterial activity of C115H MGL
mixtures with a number of sulfoxides was assessed against Gram-positive and Gram-negative bacteria.
The bacteriostatic effect was more pronounced against Gram-positive than against Gram-negative bac-
teria, while antibacterial potential proved to be quite similar. Thus, the mutant enzyme C115H MGL is an
effective catalyst, in particular, for decomposition of sulfoxides and the pharmacological couples of the
mutant form with sulfoxides might be new antimicrobial agents.
Thiosulfinates
Design of pharmacological couple
Antibacterial activity
© 2016 Elsevier B.V. and Société Française de Biochimie et Biologie Moléculaire (SFBBM). All rights
reserved.
1. Introduction
of its degradation [5]. Therefore it seems that thiosulfinates should
not generate a resistance.
Antibiotic resistance requires a search for new antimicrobial
agents. In this regard garlic-derived thiosulfinates, in contrast to
conventional antibiotics, hold promise. The exact mechanism of
antibacterial action of 2-propenethiosulfinate (allicin) and other
thiosulfinates is not well understood although these compounds
are known for a long time since the work of Cavalitto et al. [1,2].
Allicin and a number of its congeners were shown to be active
against a broader spectrum of both Gram-positive and Gram-
negative bacteria than most conventional antibiotics [1e3]. Be-
sides, the antibacterial activity of allicin is attributed mainly to the
oxidation of SH-groups of proteins [4] and possibly to the products
Pyridoxal 5’-phosphate (PLP)-dependent methionine
(MGL; EC 4.4.1.11) catalyzes the -elimination reaction of
onine and its analogues, as well as the -elimination reaction of
cysteine and S-substituted -cysteine to yield ammonia, -keto
acids and volatile thiols [6]. The enzyme was found in many mi-
croorganisms, including pathogens, fungi and plants [7e11].
Recently, we have found that Citrobacter freundii MGL catalyzes the
g
-lyase
g
L-methi-
b
L-
L
a
b-elimination reaction of alliin to form allicin [12]. The application
of allicin as an antimicrobial agent is hindered by its instability in
the blood. To overcome this problem, it was proposed to conjugate
PLP-dependent alliinase with alliin for the production of allicin in
situ [13]. Our preliminary studies showed that mixtures of MGL and
S-substituted analogues of L-methionine or L-cysteine sulfoxides,
which in situ generate thiosulfinates, inhibit the growth of some
Gram-positive and Gram-negative bacteria [14]. Three cysteine
residues of C. freundii MGL were demonstrated to be oxidized by
Abbreviations: MGL, methionine
g
-lyase; alliin, ( )-S-allyl-
L
-cysteine sulfoxide;
methiin,
(
)-S-methyl-
L
-cysteine sulfoxide; ethiin, )-S-ethyl-L
(
-homocysteine
sulfoxide; PLP, pyridoxal 50-phosphate.
* Corresponding author.
allicin that is produced by the b-elimination reaction of alliin. This
E-mail address: tvd@eimb.ru (T.V. Demidkina).
http://dx.doi.org/10.1016/j.biochi.2016.07.007
0300-9084/© 2016 Elsevier B.V. and Société Française de Biochimie et Biologie Moléculaire (SFBBM). All rights reserved.

E.A. Morozova et al. / Biochimie 128-129 (2016) 92e98
93
leads to 31% inhibition of the enzyme in the
b
-elimination reaction,
2.3. Site-directed mutagenesis
which is attributed to the oxidation of the active site Cys115 [12].
This is a serious obstacle to the use of mixtures of MGL and amino
acid sulfoxides for the design of new antimicrobial agents. Mean-
while, it was found that site-directed mutagenesis of the active site
cysteine residue conserved in MGLs from different sources [15,16]
The Cys115/His mutant form was prepared as follows: the
plasmid mglBlue was obtained by cloning the mgl gene into the
Bluescript II vector SK( ); the mutagenesis was performed by PCR.
The fragment containing T/C (411) and G/A (412) replacements was
amplified using mglBlue as the template by PCR with the primers
F1, F2 and R1:
made it possible to increase the
b-eliminating activity of the
enzyme from Trichomonas vaginalis [17] and Pseudomonas putida
[16]. In order to prepare an efficient catalyst for the production of
thiosulfinates, we constructed and characterized two mutant forms
of C. freundii MGL, in which Cys115 was replaced by Ala (C115A
MGL) or His (C115H MGL).
(F1) GATATCCATGGCTGACTGTCGTAC
(R1) TTCGGCATGCTGTGCGACAAAAACGCGTG
(F2) TACGGCCACACCCACGCGTTTTTGTCGCA
The product was incubated with NcoI and SphI and ligated via
their sites into mglBlue. The fragment containing the Cys115/His
replacement was recloned from mglBlue to pET28. The Cys115/His
replacement was confirmed by two-directional sequencing.
2. Experimental
2.1. Materials
2.4. Expression and purification of recombinant MGLs
L-Methionine, S-ethyl-L-cysteine, L-cysteine hydrochloride hy-
drate, 5,5’-dithiobis-2-nitrobenzoic acid (DTNB), nicotinamide
adenine dinucleotide reduced form (NADH), lactate dehydrogenase
(LDH) from rabbit muscle, Tris-HCl and ( )-L-alliin were purchased
from Sigma-Aldrich; pyridoxal 5’-phosphate (PLP) and D,L-
dithiothreitol (DTT) were from Serva; kanamycin and ampicillin
were domestic products (OAO Biokhimik, OAO Sintez); DEAE-
sepharose was from Amersham. Restriction enzymes, Taq DNA
polymerase, calf-intestine alkaline phosphatase, dNTP mixture, T4
DNA polymerase, pBluescript II SK ( ), rapid DNA ligation kit and
GenJet plasmid miniprep kit were purchased from Fermentas. 2-
Nitro-5-thiobenzoate (NTB) was prepared according to [18]. The
plasmid with the gene of D-2-hydroxyisocaproate dehydrogenase
(HO-Hxo-DH) is a kind gift of K. Muratore.
E. coli BL21(DE3) cells containing the plasmids with the genes of
C115A MGL (obtained in the work [12]) and C115H MGL were
grown and the enzymes were purified as described previously [15];
the differences are that a DEAE-sepharose column was used and the
gel filtration step was omitted. The concentrations of homogeneous
proteins were determined using the absorbance at 278 nm, the
extinction coefficients (A1%278) being 0.8 [21].
The rates of the
100 mM potassium phosphate buffer, pH 8.0, supplemented with
0.2 mM PLP,1 mM DTTand either 30 mM S-methyl- -cysteine (the
elimination reaction) or 30 mM -methionine (the -elimination re-
action) at 30 ꢀC. In the case of the
-elimination reaction, a coupled
b- and g-elimination reactions were measured in
L
b-
L
g
b
assay with NADH and LDH was used to follow pyruvic acid accumu-
lation [22]. The assay with HO-Hxo-DH was utilized to determine
The C. freundii strain ATCC 21434 from the American Type Culture
Collection was kindly provided by R.S. Phillips. The Staphylococcus
aureus strain 015 and Bacillus megaterium strain 168 were kindly
provided by Yu. F. Belyi and G. B. Zavilgelsky. The Escherichia coli strain
BL21 (DE3) F- ompT hsdSB gal dcm (DE3) was obtained from Novagen.
accumulationof
One activity unit is defined as the amount of the enzyme that cata-
lyzes the transformation of 1 mol of a substrate per min. The specific
activities of C115A MGL and C115H MGL were 2.3 and 38.5 U/mg for
the -elimination reaction of S-methyl- -cysteine and 1.2 and 0.006
U/mg for the -elimination reaction of -methionine.
a-ketobutyricacidintheg-eliminationreaction[23].
m
b
L
2.2. Synthesis of sulfoxides of S-alkyl-substituted cysteine
g
L
( )-S-ethyl-
L
-cysteine sulfoxide was prepared according to
2.5. Steady-state kinetics
20
[19,20] with the yield 80%. M.p. 162e163 ꢀC dec., [
a
]
¼ ꢁ18ꢀ
D
(c ¼ 1, H2O). 1H NMR (400 MHz, D2O, ppm)
d: 1.29 (t, 3H, CH3,
Kinetic parameters for the b- and g-elimination reactions were
determined using reaction mixtures containing 100 mM potassium
phosphate buffer, pH 8.0, 0.2 mM PLP, 1 mM EDTA, 10 U LDH or
J ¼ 7.5); 2.79e2.94 (m, 1H, CH2), 3.10e3.25 (m, 1H, CH2); 2.95e3.01
(m, 2H, CH2); 3.59e3.64 (m, 1 H, CH). ESI-MS (m/z) ¼ 166 [MþH]þ.
( )-S-methyl-
L-cysteine sulfoxide was synthesized as follows.
L-
70
amounts of a substrate. The reactions were initiated by the addition
of the enzyme (C115A MGL or C115H MGL). The pyruvate or
mg HO-Hxo-DH, 0.2 mM NADH, 1 mM DTT and a different
Cysteine hydrochloride hydrate (21.08 g, 0.12 mol) was suspended
in absolute alcohol (350 ml), and cuts of sodium (11 g, 0.48 mol)
were added in 15-min intervals to the solution. After all the sodium
had dissolved, methyl iodide (8.3 ml, 0.13 mol) was added and the
mixture was stirred at room temperature for 15 min. Water was
added to dissolve precipitate, the solution was acidified by 48%
aqueous hydroiodic acid to pH 5.0. Diethyl ether (600 ml) was
added to solubilize a precipitate and the solution was cooled in a
refrigerator (8 h). Crystalline precipitate was collected by filtration,
washed with cold absolute ethanol, then with diethyl ether, and
a
-
ketobutyric acid accumulation was determined as described above.
The kinetic parameters were determined by fitting the data to the
Michaelis-Menten equation using the Enzfitter program. The mo-
lecular mass of the subunit was taken as 43 kDa.
2.6. Spectral studies
Absorption spectra of the holoenzymes and their complexes
dried. The yield of S-methyl-
L
-cysteine was 88%. M.p. 219-220 ꢀC
with L-methionine and S-ethyl-L-cysteine were recorded with a
dec., [a]2D5 ¼ ꢁ32ꢀ (c 0.8, H2O). 1H NMR (400 MHz, D2O, ppm)
d: 2.13
Cary-50 spectrophotometer (Varian, USA) at 25 ꢀC in 100 mM po-
tassium phosphate buffer, pH 8.0, supplemented with 1 mM DTT
and 1 mM EDTA at a protein concentration of 1 mg/ml and a sub-
strate concentration of 30 mM.
(s, 3H, CH3); 2.97 (dd, 1H, CH2, J ¼ 14.9, 7.7 Hz); 3.08 (dd, 1H, CH2,
J ¼ 14.9, 4.3 Hz); 3.92 (dd, 1H, CH, J ¼ 7.7, 4.3 Hz). ESI-MS (m/
z) ¼ 136 [MþH]þ.
( )-S-methyl-
tion of S-methyl-
[19]. The yield was 87%, m.p.166 ꢀC. 13C NMR (D2O,
L-cysteine sulfoxide was prepared by the oxida-
L-cysteine with hydrogen peroxide according to
2.7. Quantification of thiosulfinates formed in the b-elimination
reaction of cysteine sulfoxides
d
, ppm): 37.2,
38.1 (CH3); 50.1, 50.9 (CH); 52.1, 52.9 (CH2); 171.3, 171.4 (CO2H). ESI-
MS (m/z) ¼ 152 [MþH]þ.
The ( )-S-(alkyl/alkenyl)cysteine sulfoxides (2.5 mg/ml) were

94
E.A. Morozova et al. / Biochimie 128-129 (2016) 92e98
dissolved in the sterile potassium phosphate buffer (100 mM, pH
8,0, 1 mM DTT, 0.1 mM PLP) mixed with wild type or C115H MGL
(4 mg/ml) and incubated for one hour at room temperature. The
NTB assay [17] was used to quantify thiosulfinates.
reaction of S-methyl-L-cysteine catalyzed by the mutant form of
P. putida MGL in which homologous Cys116 was replaced by His,
was found to be ten times higher compared to the wild type MGL,
as evidenced by an about 30-fold increase in the kcat value [16]. This
probably suggests that, in spite of high sequence [24] and structural
[25] homology between MGL from C. freundii and P. putida, the
active sites of the enzymes differ in fine structural details.
2.8. Antibacterial activity assay
A pronounced increase in the catalytic efficiency of C115H MGL
was observed in the b-elimination reaction of ( )-S-(alkyl/alkenyl)
cysteine sulfoxides. The kcat/Km ratio in the reaction is about an
order of magnitude higher compared to the wild type enzyme
mainly due to a decrease in the Km values. The most visible effect
Strains of different cultures were cultured on an agar plate (1.5%
agar in LB-medium) overnight at 37 ꢀC. The pre-culture was pre-
pared by adding one colony from a night agar plate to fresh LB-
medium followed by the culturing with shaking at 37 ꢀC to OD600
of 0.2e0.3. Then 200
ml of each bacterial strain pre-culture was
was observed in the
b-elimination reaction of alliin, where kcat
evenly distributed onto the surface of agar plates, 50
ml of freshly
proved to be twice higher and Km was found to be ten times lower
than those of the wild type enzyme.
prepared mixtures of the wild type enzyme and C115H MGL with
sulfoxides were put on a sterile paper disk (12 mm in diameter),
and the disk was placed on a bacteria-inoculated plate. An effect of
wild type MGL, the mutant form and separately that of sulfoxides
on the growth of bacteria was verified. After incubation at 37 ꢀC for
24 h, the zone size was measured. The experiments were made in
triplicates. The samples to be tested were prepared as described in
Section 2.7. The data were processed in semi-logarithmic
coordinates.
The C115A MGL exhibited lower catalytic efficiency in the
elimination reaction of alliin and ethiin as a result of the lower kcat
values compared to wild type MGL. The catalytic efficiency in the
b-
b
-
elimination reaction of methiin is 3-fold higher due to an increase
in its affinity to this substrate. In general, C115H MGL was shown to
be more active in the b-elimination reaction of sulfoxides than the
wild type and C115A MGLs. The X-ray diffraction study is in prog-
ress to understand the features of the reaction of C115H MGL with
sulfoxides.
3. Results and discussion
3.1. Steady-state kinetic parameters
3.2. Spectral properties of the mutant enzymes
The catalytic parameters for the
g- and
b
-elimination reactions
The postulated chemical mechanisms of the b- and g-elimina-
catalyzed by the C115A and C115H enzymes and the parameters of
the wild type enzyme are presented in Table 1.
tion reactions are presented in Scheme 1 [26e28]. The absorption
spectra of the holoenzymes of the mutant forms (Figs. 1 and 2) are
identical to the spectrum of the wild type holoenzyme [29]. The
absorption spectra show the major band with a maximum at
420e425 nm belonging to internal aldimines (Scheme 1, Structure
1). Apparently, the replacement of Cys115 with Ala or His did not
lead to significant changes in the active sites of the mutant MGL.
The replacement of cysteine with alanine resulted in a decrease
of the catalytic efficiency in the
g-elimination reaction of the
physiological substrate by more than an order of magnitude, while
the replacement by histidine led to practically undetectable activ-
ity. The catalytic efficiency of C115H MGL and C115A MGL in the
elimination of the good leaving group of O-acetyl-
L
-homoserine
The spectrum of the complex of wild type MGL with L-methionine
increases by an order of magnitude and by 38%, respectively. This
confirms the hypothesis [12] that the interactions of the Cys115/
Tyr113/Arg60 triad in the active site of C. freundii MGL provide
shows two overlapping bands with maxima in the 450e480 nm
region [22]. For PLP-dependent enzymes, the absorption in the
region of 440e480 nm is conventionally assigned to
crotonates or -aminoacrylates [29,30] (Scheme 1, Structures 4, 10).
It was concluded that the deamination of aminocrotonate could be
the rate-limiting step of the -elimination reaction [21].
The spectrum of the complex of C115A MGL with -methionine
is similar to the spectrum of the wild type MGL complex with
a-amino-
effective catalysis of
In the -elimination reaction, C115H MGL showed almost the
same catalytic efficiency with S-alkyl-substituted cysteines and O-
g-substituent elimination.
a
b
g
acetyl-
in the
L
-serine as the wild type enzyme did. However, the kcat value
L
b
-elimination reaction of S-methyl- -cysteine proved to be
L
L-
increased fivefold, whereas the Km value was 4-fold higher. It is
methionine. It shows two overlapping bands at 450e480 nm and a
broad band in the region of 320e340 nm (Fig. 1). The region of
noteworthy that the catalytic efficiency in the
b-elimination
Table 1
Kinetic parameters for the
g- and
b-elimination reactions.
Substrate
Wild type MGL
C115H MGL
C115A MGL
kcat (sꢁ1
)
Km (mM)
kcat/Km (Mꢁ1 sꢁ1
)
kcat (sꢁ1
)
Km (mM)
kcat/Km (Mꢁ1 sꢁ1
)
kcat (sꢁ1
)
Km (mM)
kcat/Km (Mꢁ1 sꢁ1
)
L-Met
6.20a
2.10b
4.60a
3.48
0.70
2.91
0.71
4.14
0.17
2.46
4.28
4.70
8.85 ꢂ 103
7.21 ꢂ 102
6.48 ꢂ 103
8.4 ꢂ 102
2.96 ꢂ 104
2.7 ꢂ 103
4.98 ꢂ 102
1.26 ꢂ 103
ND
ND
ND
0.22
6.23
1.38
1.63
0.57
1.25
1.35
0.41
0.301
6.00
0.80
0.63
0.255
5.48
2.88
1.75
6.90 ꢂ 102
1.04 ꢂ 103
1.70 ꢂ 103
2.60 ꢂ 103
2.24 ꢂ 103
2.30 ꢂ 102
4.60 ꢂ 102
2.30 ꢂ 102
O-Ac-
L
-Hse
18.29
23.08
5.14
6.50
5.59
2.93
11.31
2.46
2.62
0.57
0.93
0.47
3.68
0.44
7.43 ꢂ 103
8.81 ꢂ 103
9.02 ꢂ 103
6.99 ꢂ 103
1.20 ꢂ 104
8.00 ꢂ 102
2.57 ꢂ 104
S-Met-L-Cys
Methiin
S-Et-L-Cys
Ethiin
O-Ac-L-Ser
Alliin
5.03a
6.71
2.13c
5.90d
The mean squared error of the experiment in the determination of the kinetic parameters did not exceed 10%.
ND e not detectable.
a
Data from Ref. [15].
Data from Ref. [21].
Data from Ref. [34].
Data from Ref. [12].
b
c
d

E.A. Morozova et al. / Biochimie 128-129 (2016) 92e98
95
Scheme 1. Mechanism of the
b- and
g-elimination reactions of sulfur-containing amino acids catalyzed by the MGL. Adopted from Refs. [31e33].
Fig. 2. The absorption spectra of C115H MGL. Holoenzyme (solid line), the complex
with L-methionine (dotted line), the complex with S-ethyl-L-cysteine (dash-dotted
line).
Fig. 1. The absorption spectra of C115A MGL. Holoenzyme (solid line), the complex
with L-methionine (dotted line), the complex with S-ethyl-L-cysteine (dash-dotted
line).
decrease in the band intensity at 420 nm,
a shoulder at
320e340 nm cannot be interpreted as the
product, -ketobutyric acid, intermediates of the
reaction with the tetrahedral C4’-atom of PLP, and enolic tautomers
of -aminocrotonate, external and internal aldimines exhibit bands
in it. The presence of the bands belonging to -aminocrotonate
suggests that, as in the case of the wild type enzyme, the deami-
nation of -aminocrotonate might be the rate-limiting step of the
-elimination reaction for C115A MGL.
In the spectrum of C115A MGL with S-ethyl-
g
-elimination reaction
438e440 nm, а very low-intensity peak at 495e497 nm and a broad
band at 320e340 nm are observed. The intermediate with a
maximum at 420 nm should belong to the external aldimine with
the substrate. The long-wavelength band can be assigned to a
quinonoid intermediate [31]. The absorption at 438e440 nm most
a
g
-elimination
a
a
likely belongs to
a-aminoacrylate.
a
The spectrum of the complex of C115H MGL (Fig. 2) with L-
methionine differs from the spectrum of C115A MGL. Thus, there
are no overlapping bands in the 450e480 nm region. The spectrum
g
L-cysteine (Fig. 1), a

96
E.A. Morozova et al. / Biochimie 128-129 (2016) 92e98
shows broad band at 320e340 nm and a band at 428 nm. In
addition, there is very low-intensity peak at 495 nm. Bearing in
similar peaks are present in the spectrum of C115A MGL with S-
ethyl- -cysteine, the -elimination reaction catalyzed by the
L
b
mind that no detectable activity was observed in the
g
-elimination
mutant enzymes apparently proceeds via the E1cB mechanism
[32], involving the step of a quinonoid intermediate formation. This
reaction of -methionine, one may conclude that the equilibrium
L
mixture contains the ketoenamine and enolimine tautomers of the
external aldimine, and/or intermediates with a tetrahedral C4’-
mechanism is generally observed for PLP-dependent
lyases, except for O-acetylserine sulfhydrylase that catalyzes the
elimination of acetate from O-acetyl- -serine by the E2 mechanism
[33]. The question as to whether wild type MGL catalyzes the b-
elimination reaction by the E1cB mechanism remains to be
elucidated.
b-eliminating
atom of the cofactor but not
a
-aminocrotonate. Presumably, the
g-
L
substituent elimination may be one of the rate-limiting steps for
the mutant enzyme.
The spectrum of C115H MGL with S-ethyl-L-cysteine, in contrast
to the spectrum of C115A MGL, shows intense peaks at 496 nm and
468 nm and a shoulder at 440 nm. Two long-wavelength peaks
undoubtedly belong to a quinonoid intermediate [31]. The ab-
3.3. Antibacterial potential
sorption at 440 nm may be assigned to
the vibrational structure of a quinonoid [31]. The mechanism of the
MGL-catalyzed -elimination reaction is poorly investigated and
the reliable assignment of the absorption bands requires compre-
hensive kinetic and spectral studies. Taking into account that the
a
-aminoacrylate and/or to
We compared the antibacterial activities of wild type MGL and
C115H MGL mixtures with alliin, methiin and ethiin against Gram-
negative and Gram-positive bacteria (Table 2). The antibacterial
activities of C115H MGL mixtures were higher, which is consistent
b
with a higher catalytic efficiency of the mutant enzyme in the b-
Table 2
Antibacterial activity of the mixtures of MGLs (10 mg/ml) with sulfoxides (2.5 mg/ml).
Sulfoxide
Zone of inhibition, mm2
Gram-negative
C. freundii
Gram-positive
E. coli BL21
B. megaterium
S. aureus
Wild type MGL
C115H MGL
Wild type MGL
C115H MGL
Wild type MGL
C115H MGL
Wild type MGL
C115H MGL
Alliin
Methiin
Ethiin
387 5.6
215 3.1
225 2.3
412 6.0
312 23.1
225 2.3
412 6.0
312 23.1
256 3.7
497 7.6
349 3.5
312 23.1
529 5.3
387 5.6
256 3.7
625 6.3
576 5.8
387 5.6
745 10.9
497 7.6
225 2.3
961 9.6
576 5.8
225 2.3
Fig. 3. Correlation between the concentration of antibiotics and their zones of inhibition: A) for C. freundii. The correlation coefficients are r2 ¼ 0.9864 for allicin, r2 ¼ 0.9765 for
kanamycin, r2 ¼ 0.9888 for dimethyl thiosulfinate, r2 ¼ 0.9924 for ampicillin, r2 ¼ 0.983 for diethyl thiosulfinate; B) for S. aureus. The correlation coefficients are r2 ¼ 0.9939 for
allicin, r2 ¼ 0.9998 for kanamycin, r2 ¼ 0.9805 for dimethyl thiosulfinate, r2 ¼ 0.9988 for ampicillin, r2 ¼ 0.9474 for diethyl thiosulfinate; C) for B. megaterium. The correlation
coefficients are r2 ¼ 0.9763 for allicin, r2 ¼ 0.9841 for kanamycin, r2 ¼ 0.9676 for dimethyl thiosulfinate, r2 ¼ 0.9942 for ampicillin, r2 ¼ 0.9413 for diethyl thiosulfinate; D) for E. coli.
The correlation coefficients are r2 ¼ 0.9851 for allicin, r2 ¼ 0.9403 for kanamycin, r2 ¼ 0.9693 for dimethyl thiosulfinate, r2 ¼ 0.9715 for ampicillin, r2 ¼ 0.9662 for diethyl thio-
sulfinate. The mixtures of C115H MGL (0.09 mM) and sulfoxides (2.5 mg/ml) were prepared as described in 2.7.

E.A. Morozova et al. / Biochimie 128-129 (2016) 92e98
97
elimination reaction of these substrates (Table 1). The antibacterial
References
potential of the mixtures of C115H MGL with sulfoxides was
assessed and compared with that of kanamycin and ampicillin. The
concentrations well correlate with the activities, and they give
straight lines when concentrations are plotted against the zones of
inhibition on a semi-logarithmic scale (Fig. 3).
As can be seen from these lines, the anti-staphylococcal activity
of allicin is about 2-fold higher than those of dimethyl thiosulfinate
and diethyl thiosulfinate and higher even than those of kanamycin
and ampicillin at concentrations of about 1 mg/ml (Fig. 3). The
potential of kanamycin is about 20-fold higher compared to the
investigated thiosulfinates. The anti-staphylococcal potential of
ampicillin was about two orders of magnitude higher than that of
thiosulfinates (Fig. 3).
[1] C.J. Cavallito, J.H. Bailey, Allicin, the antibacterial principle of Allium sativum. I.
Isolation, physical properties and antibacterial action, J. Am. Chem. Soc. 66
(1944) 1950e1951.
[2] L.V. Small, J.H. Bailey, C.J. Cavallito, Alkyl thiosulfinates, J. Am. Chem. Soc. 69
(1947) 1710e1713.
[3] G.R. Fenwick, A.B. Hanley, The genus Allium-Part 1, CRC Crit. Rev. Food Sci.
Nutr. 22 (1985) 199e271.
[4] A. Rabinkov, T. Miron, L. Konstantinovski, M. Wilchek, D. Mirelman, L. Weiner,
The mode of action of allicin: trapping of radicals and interaction with thiol
containing proteins, Biochim. Biophys. Acta 1379 (1998) 233e244.
[5] E. Block, S. Ahmad, J.L. Catalfamo, M.K. Jain, R. Apitz-Castro, Antithrombotic
organosulfur compounds from garlic: structural, mechanistic, and synthetic
studies, J. Am. Chem. Soc. 108 (22) (1986) 7045e7055.
[6] H. Tanaka, N. Esaki, K. Soda, A versatile bacterial enzyme: L-methionine
g
-
lyase, Enzyme Microb. Technol 7 (1985) 530e537.
[7] W. Kreis, C. Hession, Isolation and purification of L-methionine-a-deamino-
g
-
mercaptomethane-lyase (L-methioninase) from Clostridium sporogenes, Can-
cer Res. 33 (1973) 1862e1865.
[8] M. Yoshimura, Y. Nakano, Y. Yamashita, T. Oho, T. Saito, T. Koga, Formation of
methyl mercaptan from L-methionine by Porphyromonas gingivalis, Infect.
Immun. 68 (2000) 6912e6916.
In the case of B. megaterium, the inhibition zones for diethyl
thiosulfinate are larger compared with allicin and dimethyl thio-
sulfinate at concentrations of higher than 0.6 mg/ml, but the anti-
microbial potential of diethyl thiosulfinate is weaker compared
with allicin and dimethyl thiosulfinate (Fig. 3). The anti-bacillus
potential of kanamycin is about 10e20-fold higher than that of
allicin and dimethyl thiosulfinate, and about 90-fold higher than
that of diethyl thiosulfinate. The potential of ampicillin is about two
orders of magnitude higher compared to thiosulfinates (Fig. 3).
Thiosulfinates exhibit almost identical activity and potential
against Gram-negative bacteria (Fig. 3). The anti-escherichia po-
tential of thiosulfinates is about 50-fold weaker than that of
ampicillin and about 10-fold weaker than that of kanamycin (Fig. 3).
Compared to thiosulfinates, the anti-citrobacter potential is about
10-fold stronger for kanamycin and 7e11-fold stronger for ampi-
cillin (Fig. 3).
For all of the tested thiosulfinates at concentrations of about
1 mg/ml, the inhibition zones were 2e3-fold larger for S. aureus and
B. megaterium than those for E. coli and C. freundii. Therefore, the
antibacterial activities of thiosulfinates are more pronounced for
Gram-positive than for Gram-negative bacteria in the specified
concentration range. The antibacterial potentials of thiosulfinates
proved to be quite similar for all bacteria, except for more pro-
nounced effects of allicin and dimethyl thiosulfinate against
B. megaterium (Fig. 3).
[9] S.V. Revtovich, E.A. Morozova, N.V. Anufrieva, M.I. Kotlov, Y.F. Belyi,
T.V. Demidkina, Identification of methionine
g-lyase in genomes of some
pathogenic bacteria, Dokl. Biochem. Biophys. 445 (2012) 187e193.
[10] A.S. El-Sayed, Microbial L-methioninase: production, molecular characteriza-
tion, and therapeutic applications, Appl. Microbiol. Biotechnol. 86 (2010)
445e467.
[11] A. Goyer, E. Collakova, Y. Shachar-Hill, A.D. Hanson, Functional characteriza-
tion of a methionine g-lyase in Arabidopsis and its implication in an alter-
native to the reverse trans-sulfuration pathway, Plant Cell Physiol. 48 (2007)
232e242.
[12] E.A. Morozova, S.V. Revtovich, N.V. Anufrieva, V.V. Kulikova, A.D. Nikulin,
T.V. Demidkina, Alliin is a suicide substrate of Citrobacter freundii methionine
g-lyase: structural bases of inactivation of the enzyme, Acta Cryst. D. 70
(2014) 3034e3042.
[13] E. Appel, A. Vallon-Eberhard, A. Rabinkov, O. Brenner, I. Shin, K. Sasson,
Y. Shadkchan, N. Osherov, S. Jung, D. Mirelman, Therapy of murine pulmonary
aspergillosis with antibody-alliinase conjugates and alliin, Antimicrob. Agents
Chemother. 54 (2010) 898e906.
[14] N.V. Anufrieva, E.A. Morozova, V.V. Kulikova, N.P. Bazhulina, I.V. Manukhov,
D.I. Degtev, E.Y. Gnuchikh, A.N. Rodionov, G.B. Zavilgelsky, T.V. Demidkina,
Sulfoxides, analogues of L-methionine and L-cysteine as pro-drugs against
Gram-Positive and Gram-Negative bacteria, Acta Naturae 7 (2015) 128e135.
[15] I.V. Manukhov, D.V. Mamaeva, E.A. Morozova, S.M. Rastorguev, N.G. Faleev,
T.V. Demidkina, G.B. Zavilgelsky, L-Methionine
g-lyase from Citrobacter
freundii: cloning of the gene and kinetic parameters of the enzyme, Biochem.
Mosc. 71 (2006) 361e369.
[16] D. Kudou, S. Misaki, M. Yamashita, T. Tamura, N. Esaki, K. Inagaki, The role of
cysteine 116 in the active site of the antitumor enzyme L-methionine g-lyase
from Pseudomonas putida, Biosci. Biotechnol. Biochem. 72 (2008) 1722e1730.
[17] A.E. McKie, T. Edlind, J. Walker, J.C. Mottram, G.H. Coombs, The primitive
4. Conclusion
protozoon Trichomonas vaginalis contains two methionine
g-lyase genes that
encode members of the
g
-family of pyridoxal 50-phosphate-dependent en-
The replacement of cysteine 115 of C. freundii MGL with histi-
zymes, J. Biol. Chem. 273 (1998) 5549e5556.
[18] T. Miron, A. Rabinkov, D. Mirelman, L. Weiner, M. Wilchek,
A spectrophotometric assay for allicin and alliinase (alliin lyase) activity: re-
action of 2-nitro-5-thiobenzoate with thiosulfinates, Anal. Biochem. 265
(1998) 317e325.
[19] M. Frankel, D. Gertner, H. Jacobson, A. Zilkha, Synthesis of poly-S-alkyl-L-
cysteines, J. Chem. Soc. (1960) 1390e1393.
[20] W.H. Briggs, H. Xiao, K.L. Parkin, C. Shen, I.L. Goldman, Differential inhibition
of human platelet aggregation by selected Allium thiosulfinates, J. Agric. Food
Chem. 48 (2000) 5731e5735.
[21] E.A. Morozova, N.P. Bazhulina, N.V. Anufrieva, D.V. Mamaeva, Y.V. Tkachev,
S.A. Streltsov, V.P. Timofeev, N.G. Faleev, T.V. Demidkina, Kinetic and spectral
dine results in the loss of activity of the mutant form in the
elimination reaction of the physiological substrate. The catalytic
efficiency of C115H MGL in the -elimination reaction of charac-
teristic substrates is compared to that of the wild type enzyme.
The mutant form catalyzes the -elimination reaction of S-
substituted cysteine sulfoxides with a catalytic efficiency an order
of magnitude higher compared the wild type MGL. Thus, the
replacement of Cys115 with His allowed us to obtain an effective
catalyst for the production of thiosulfinates. The catalytic effi-
g-
b
b
parameters of interaction of Citrobacter freundii methionine
g-lyase with
ciency of C115A MGL in the
substituted cysteine sulfoxides proved to be significantly lower
compared the C115H enzyme.
b-elimination reaction of S-
amino acids, Biochem. Mosc. 75 (2010) 1272e1280.
[22] Y. Morino, E.E. Snell, Tryptophanase (E. coli B), Methods Enzymol. 17A (1970)
439e446.
[23] S.M. Aitken, D.H. Kim, J.F. Kirsch, Escherichia coli cystathionine
g-synthase
The C115H MGL mixtures with a number of sulfoxides were
shown to have antibacterial activities against some Gram-positive
and Gram-negative bacteria. The pharmacological couples of the
mutant enzyme and sulfoxides might be considered as potential
new antimicrobial agents.
does not obey ping-pong kinetics. Novel continuous assays for the elimination
and substitution reactions, Biochemistry 42 (2003) 11297e11306.
[24] I.V. Manukhov, D.V. Mamaeva, S.M. Rastorguev, N.G. Faleev, E.A. Morozova,
T.V. Demidkina, G.B. Zavilgelsky, A gene encoding L-methionine
g-lyase is
present in Enterobacteriaceae family genomes: identification and character-
ization of Citrobacter freundii L-methionine
3889e3893.
g-lyase, J. Bacteriol. 187 (2005)
[25] D. Mamaeva, E. Morozova, A. Nikulin, S. Revtovich, S. Nikonov, M. Garber,
Acknowledgments
T. Demidkina, Structure of Citrobacter freundii L-methionine
Cryst. F61 (2005) 546e549.
g-lyase, Acta
[26] P. Brzovic, E.L. Holbrook, R.C. Greene, M.F. Dunn, Reaction mechanism of
Escherichia coli cystathionine -synthase: direct evidence for a pyridoxamine
This research was supported by the Russian Science Foundation
(project No. 15-14-00009).
g

98
E.A. Morozova et al. / Biochimie 128-129 (2016) 92e98
derivative of vinylglyoxylate as a key intermediate in pyridoxal phosphate
dependent -elimination and -replacement reactions, Biochemistry 29
(1990) 442e451.
[30] L. Davis, D. Metzler, in: third ed., in: P.D. Boyer (Ed.), The Enzymes, 7, Aca-
g
g
demic Press, N. Y, 1972, pp. 33e74.
[31] C.M. Metzler, A.G. Harris, D.E. Metzler, Spectroscopic studies of quinonoid
species from pyridoxal 50-phosphate, Biochemistry 27 (1988) 4923e4933.
[32] W.H. Saunders, Distinguishing between concerted and nonconcerted elimi-
nations, Acc. Chem. Res. 9 (1976) 19e25.
[27] T. Clausen, R. Huber, B. Laber, H.D. Pohlenz, A. Messerschmidt, Crystal struc-
ture of the pyridoxal-50-phosphate dependent cystathionine
b-lyase from
Escherichia coli at 1.83 Å, J. Mol. Biol. 262 (2) (1996) 202e224.
[28] T. Clausen, R. Huber, L. Prade, M.C. Wahl, A. Messerschmidt, Crystal structure
[33] C.H. Tai, P.F. Cook, Pyridoxal 50-phosphate-dependent
a,b-elimination re-
of Escherichia coli cystathionine
(1998) 6827e6838.
g-synthase at 1.5 Å resolution, EMBO J. 23
actions: mechanism of O-acetylserine sulfhydrylase, Acc. Chem. Res. 34 (2001)
49e59.
[29] YuV. Morozov, V.O. Chekhov, N.P. Bazhulina, On the possibility to predict
spectral properties of pyridoxal-50-phosphate aldimines with amino acids and
some of their derivatives arising in the course of pyridoxal-depending enzy-
matic reactions, Biophys. Rus. 42 (2) (1997) 301e312.
[34] N.V. Anufrieva, N.G. Faleev, E.A. Morozova, N.P. Bazhulina, S.V. Revtovich,
V.P. Timofeev, Y.V. Tkachev, A.D. Nikulin, T.V. Demidkina, The role of active
site tyrosine 58 in Citrobacter freundii methionine
Acta 1854 (2015) 1220e1228.
g-lyase, Biochim. Biophys.