Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application to prevent human fibrinogen degradation and bacteria growth in vitro

Article abstract of DOI:10.1016/j.biochi.2014.05.006

SufA, which belongs to the subtilisin-like serine protease family, contains a non-canonical Asp-His-Ser catalytic triad. Under in vitro conditions, SufA is capable of human fibrinogen hydrolysis leading to inhibition of fibrin network formation, thus suggesting its important role in the development and progression of Finegoldia magna infections. In addition, it has been demonstrated that SufA can hydrolyze antibacterial peptides such as LL-37 and the chemokine MIG/CXCL 9, hence evading host defence mechanisms. Although the SufA protease from F. magna was discovered several years ago, its optimal substrate preference has not yet been identified. Considering the role of SufA, we have focused on the profiling of its substrate sequence preference spanning S1-S3 binding pockets using the FRET (fluorescence resonance energy transfer) approach. Next, based on the structure of the P1 residue of the developed substrate, we narrowed the inhibitor screening to the phosphonic analogues of amino acids containing an arginine-like side chain. Among all the compounds tested, only Cbz-6-AmNphth^P(OPh)₂ showed any inhibitory activity against SufA displaying k₂/K_i value of 10 800 M ^-1 s^-1. In addition, it prevented SufA-mediated human fibrinogen hydrolysis in vitro and exhibited potent antibacterial activity against F. magna, Staphylococcus aureus and Escherichia coli. Herein, we report on the substrate specificity, synthesis and kinetic evaluation of phosphonic inhibitors of SufA protease from F. magna which could help to establish its function in pathogenesis development and may lead to the elaboration of new antibacterial drugs.

Full text of DOI:10.1016/j.biochi.2014.05.006

Biochimie xxx (2014) 1e7
Contents lists available at ScienceDirect
Biochimie
journal homepage: www.elsevier.com/locate/biochi
Research paper
Substrate profiling of Finegoldia magna SufA protease, inhibitor
screening and application to prevent human fibrinogen degradation
and bacteria growth in vitro
a
,
*
,
1
a
,
1
b
c
Ewa Burchacka
, Marcin Sie nꢀ czyk , Inga-Maria Frick , Magdalena Wysocka ,
c
a
Adam Lesner , J oꢀ zef Oleksyszyn
Division of Medicinal Chemistry and Microbiology, Faculty of Chemistry, Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370
a
Wroclaw, Poland
b
Division of Infection Medicine, Department of Clinical Sciences, Lund University, SE-221 84 Lund, Sweden
Faculty of Chemistry, Gdansk University, Wita Stwosza 63, 80-952 Gdansk, Poland
c
a r t i c l e i n f o
a b s t r a c t
Article history:
SufA, which belongs to the subtilisin-like serine protease family, contains a non-canonical Asp-His-Ser
catalytic triad. Under in vitro conditions, SufA is capable of human fibrinogen hydrolysis leading to in-
hibition of fibrin network formation, thus suggesting its important role in the development and pro-
gression of Finegoldia magna infections. In addition, it has been demonstrated that SufA can hydrolyze
antibacterial peptides such as LL-37 and the chemokine MIG/CXCL 9, hence evading host defence
mechanisms. Although the SufA protease from F. magna was discovered several years ago, its optimal
substrate preference has not yet been identified. Considering the role of SufA, we have focused on the
profiling of its substrate sequence preference spanning S1eS3 binding pockets using the FRET (fluo-
rescence resonance energy transfer) approach. Next, based on the structure of the P1 residue of the
Received 30 December 2013
Accepted 12 May 2014
Available online xxx
Keywords:
Finegoldia magna
SufA protease
Substrate mapping
Phosphonic inhibitors
developed substrate, we narrowed the inhibitor screening to the phosphonic analogues of amino acids
P
containing an arginine-like side chain. Among all the compounds tested, only Cbz-6-AmNphth (OPh)
2
ꢀ
1
ꢀ1
showed any inhibitory activity against SufA displaying k
2
/K
i
value of 10 800 M
s . In addition, it
prevented SufA-mediated human fibrinogen hydrolysis in vitro and exhibited potent antibacterial activity
against F. magna, Staphylococcus aureus and Escherichia coli.
Herein, we report on the substrate specificity, synthesis and kinetic evaluation of phosphonic in-
hibitors of SufA protease from F. magna which could help to establish its function in pathogenesis
development and may lead to the elaboration of new antibacterial drugs.
©
2014 Elsevier Masson SAS. All rights reserved.
1
. Introduction
Staphylococcus aureus, since both contribute to skin dysfunction [9].
The combined action of these bacterial species results in difficulties
in the treatment of skin infections, especially when increasing
resistance to antibiotics is taken into consideration. Although
F. magna is a predominant bacterial species found in clinical in-
fections caused by anaerobic cocci, its role in pathogenesis is not
fully understood. F. magna produces various virulence factors, such
as collagenases, albumin-binding proteins, protein L, adhesion
factors and the SufA protease [2]. SufA is a member of the subtilisin-
like serine protease family with a non-canonical Asp-His-Ser cat-
alytic triad. It has been demonstrated that SufA protease cleaves
human fibrinogen and thus inhibits the fibrin network formation
Gram-positive anaerobic coccus Finegoldia magna is a member
of the human commensal skin flora and is also present in the
oropharynx as well as in gastrointestinal and urogenital tracts [1,2].
In addition, F. magna is a clinically important opportunistic path-
ogen associated with dangerous and onerous human diseases
[1,3e5]. Most F. magna strains have been isolated from diabetic
wounds and skin, bone, chest, obstetrical and gynaecological in-
fections [6e8]. F. magna displays similar pathogenic features to
[
10]. This suggests that SufA may play an important role in the
*
Corresponding author. Tel.: þ48 71 320 24 39; fax: þ48 71 40 64.
E-mail address: ewa.burchacka@pwr.wroc.pl (E. Burchacka).
These authors contributed equally to this work.
development and progression of F. magna infections. Additionally, it
has been shown that SufA is able to hydrolyze antibacterial
1
http://dx.doi.org/10.1016/j.biochi.2014.05.006
300-9084/© 2014 Elsevier Masson SAS. All rights reserved.
0
Please cite this article in press as: E. Burchacka, et al., Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application
to prevent human fibrinogen degradation and bacteria growth in vitro, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.006

2
E. Burchacka et al. / Biochimie xxx (2014) 1e7
peptides, such as LL-37 [11]. Moreover, chemokine MIG/CXCL 9
produced by human keratinocytes in response to inflammatory
stimuli is also degraded by SufA-expressing F. magna [12]. In this
respect, SufA represents an important target for the development of
specific and selective inhibitors and could thus help to establish the
role of this protease during an infection. Additionally, optimization
of SufA inhibitors could lead to the development of new antibac-
terial agents against F. magna species.
2. Materials and methods
2.1. Substrate library synthesis
2.1.1. Peptide synthesis
All peptides were synthesized manually using the Fmoc solid-
phase chemistry, as described previously [25]. TentaGel S RAM
(substitution 0.25 meq/g; RAPP Polymere, Tübingen, Germany) was
used as a solid support. The first step of the synthesis was the
conjugation of 5-amino-2-nitrobenzoic acid (ANB) to the resin [26].
Shortly, ANB (3 eq.) was dissolved in dimethylformamide (DMF)
(5 ml) and O-(benzotriazol-1-yl)-N,N,N ,N -tetramethyluronium
tetrafluoroborate (TBTU) (3 eq.) coupling agent was added followed
by an addition of 4-dimethylaminopyridine (DMAP) (2 eq.). The
solution was then added to the resin and after 30 s N,N-diisopro-
pylethylamine (DIPEA) (6 eq.) was added. The reaction was per-
formed at room temperature for 3 h. The solution was filtered and
the resin was washed with DMF. The procedure was repeated three
times. Next, the first Fmoc-protected amino acid was conjugated to
Considering the role of SufA protease and the lack of its
complete substrate preference data, we started with a mapping
of protease binding pockets and the development of selective
and specific substrates. In general, profiling of SufA protease,
especially the preference towards P1 residue of the substrate, is
highly useful for the design of potent inhibitors. The technique
of fluorescence resonance energy transfer (FRET) is commonly
used for the preparation of fluorogenic peptide substrates for
biochemical characterization of proteases in order to determine
their specificity of action. The main concept of such a method is
based on the presence of two “opposite” chemical functions
located within the peptide chain, where one serves as a donor of
fluorescence and the second as a quencher. Protease-mediated
hydrolysis of any peptide bond of the substrate results in its
separation which leads to an increase in fluorescence which
directly corresponds to the protease activity. One interesting
fluorophore-quencher pair is aminobenzoic acid (ABZ) and 5-
amino-2-nitrobenzoic acid (ANB) which display dual, chromo-
genic (ANB) and fluorogenic (ABZ) properties, allowing the
measurement of protease activity in an easy readout [13,14].
For the screening of potential SufA inhibitors, we focused our
0
0
3
ANB using a POCl /pyridine system. Further, Fmoc-deprotection
steps were performed using 20% piperidine in DMF/NMP and
Fmoc amino acids were conjugated using the DIPCI/HOBt method.
0
Briefly, the mixture of N-protected amino acid derivative, N,N -
diisopropylcarbodiimide (DIPCI) and 1-hydroxybenzotriazole
(HOBt) (molar ratio, 1:1:1) was dissolved in DMF:NMP (N-
methyl-2-pyrrolidone) solution (1:1, v/v) and was added into the
resin. A three-fold excess to the resin active sites was used. The Boc-
protected 2-aminobenzoic acid (ABZ) was introduced as the N-
terminal group. After completing the synthesis, the peptides were
attention on
a
-aminoalkylphosphonate diaryl esters which are
2
cleaved from the resin using a TFA/phenol/triisopropylsilane/H O
known selective and potent inhibitors reacting with the active site
of serine proteases in an irreversible manner. They are highly
specific towards serine proteases and do not react with acetyl-
cholinesterase, cysteine, aspartyl, and metalloproteinases
mixture (88:5:2:5, v/v) [27]. The purity of peptides was examined
using a reversed phase high pressure liquid chromatography on a
Varian Pro Star HPLC system (Varian, Australia) equipped with a
Kromasil 100 C
8
column (8 ꢁ 250 mm) (Knauer, Berlin, Germany)
[
15e17]. The overall design of the phosphonic inhibitors of serine
and a UVeVIS detector. A linear gradient from 10 to 90% B was
applied within 40 min (A: 0.1% TFA; B: 80% acetonitrile in A). The
analyzed peptides were monitored at 226 nm. Mass spectra of
synthesized peptides were recorded on a Biflex III MALDI TOF mass
proteases is based on the replacement of the C-terminal amino
acid residue of the substrate of best fit with a structurally corre-
sponding
phosphonic
analogue
[17,18].
a-Amino-
alkylphosphonate diaryl esters have been successfully used as
inhibitors of different human serine proteases, including
cathepsin G, neutrophil elastase, DPP IV, chymase, uPA and
granzymes [19]. Recently, phosphonic inhibitors of bacterial pro-
teases have been reported: the phosphonic analogue of glutamic
acid and its peptidyl derivatives displayed the second-order in-
spectrometer (Bruker, Bremen, Germany) using
hydroxycinnamic acid (CHCA) as a matrix.
a-cyano-4-
2.1.2. Peptide library preparation
The peptide library was synthesized via the split-and-mix
method [28,29]. Initially, 5 g of the solid support (substitution
0.25 meq/g; RAPP Polymere, Tübingen, Germany) was used. The
residue in position P1 was attached to the solid support, as
described above, followed by several peptide elongation steps for
P2eP4 positions: coupling of the appropriate amino acid derivative
to the resin-attached amino acid, washing, mixing and splitting,
and removal of the group protecting the N-terminal. A three-fold
molar excess of amino acid was used for the coupling reaction.
The Boc-protected 2-aminobenzoic acid was introduced as the N-
terminal group. Other synthetic methods employed were as
described above. The library thus obtained was used without pu-
rification. In order to prove that the synthesized library contained
the intended peptides pattern, MS analyses were performed for
randomly chosen sublibraries at each stage of the synthesis. Mass
spectra were recorded using a Bruker Biflex III MALDI TOF mass
spectrometer (Bruker, Bremen, Germany) (data not shown).
3
ꢀ1
ꢀ1
and
hibition rate constant values of 5.3
ꢁ
10
and Boc-Phe-Leu-Glu-
towards S. aureus V8 protease, respectively [20,21]. A
phosphonic peptide containing a C-terminal analogue of gluta-
M
s
3
ꢀ1 ꢀ1
P
8
(
.5 ꢁ 10
M
s
6 5 2
for Ac-Glu (OC H )
P
6 5 2
OC H )
P
mine (Cbz-Glu-Leu-Gln (OC
6
H
4
)
2
) showed an inhibitory activity
/K value of
ꢀ1 ꢀ1
s , whereas phosphonic analogues of phenylalanine
towards S. aureus SplB protease with
a
k
2
i
2
5
ꢁ 10 M
and leucine were active against subtilisin and S. aureus SplA
protease [22,23].
In this study, we present the substrate specificity profiling
(
P4eP1 residues, according to the Schechter and Berger
nomenclature [24]) of the SufA protease using a synthetic pep-
tide library and preliminary results of phosphonic-type inhibitor
screening. One of the phosphonic inhibitors tested (5) was able
to completely block SufA-mediated human fibrinogen degrada-
tion in vitro. Moreover, compound 5 showed a high antibacterial
effect in cell culture. To the best of our knowledge, this is the
first report describing SufA substrate profiling using the
combinatorial approach, its effective inhibition as well as pre-
venting fibrinogen degradation in vitro and potent antibacterial
activity.
2.2. Enzymatic studies
2.2.1. Initial screening
For substrate library evaluation, SufA protease (0.7
used. The assay buffer was 50 mM Tris HCl (pH 8.0) supplemented
mg/ml) was
Please cite this article in press as: E. Burchacka, et al., Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application
to prevent human fibrinogen degradation and bacteria growth in vitro, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.006

E. Burchacka et al. / Biochimie xxx (2014) 1e7
3
with 20 mM KCl. Iterative deconvolution of the sublibraries was
performed using the fluorescent method [27]. The stock solutions
of tested peptides were prepared in DMSO (3 mg/ml). All mea-
surements were performed using an Omega plate reader (BMG,
under Swern conditions gave the corresponding aldehyde. The
resulting 6-formyl-2-naphthonitrile was used for the amidoalky-
lation reaction of triphenyl phosphite with benzyl carbamate
P
leading to Cbz-(6-CN)Nphth (OPh)
2
[31]. The nitrile group was next
Ortenberg, Germany). The fluorescent intensity was measured for
converted into amidine via ethyl imidate formation.
ꢂ
6
0 min at 37 C at 325 nm (ABZ excitation) and 400 nm (ABZ
emission). The final enzyme concentration used in this study was
2.4. SufA inhibition studies
ꢀ10
ꢀ11
M (position X2
1.2 ꢁ 10
M (position X4 and X3) and 5.2 ꢁ 10
ꢀ7
and X1). The concentration of each library was 2.76 ꢁ 10 M.
The enzyme inhibition assay was performed using 96-well mi-
crotiter plates (COSTAR, Corning Inc., Gdansk, Poland) at 37 C in a
ꢂ
2.2.2. Determination of Michaelis constants (K
M
), catalytic
total volume of 200 ml. The assay buffer was 0.1 M TriseHCl (pH 8.0)
constants (kcat), and specificity constants (kcat/K
M
)
supplemented with 0.01 M KCl. The substrate used was Abz-Ile-Ser-
Lys-ANB (Ex. 325 nm, Em. 400 nm). To calculate the apparent
Enzymatic hydrolysis of selected optimal substrates was per-
ꢂ
formed in 0.1 M TriseHCl (pH 8.0) at 37 C. The velocity of the
second-order rate constant values (k
inhibitors in a range of 5 Me200
concentration of the substrate (25
The hyperbolic model of inhibition was used to calculate k
The K and k /K values were calculated from the slope of the linear
range of the plot using the following equation: kobs/[I] ¼ k /(K þ[I]),
where K is the dissociation constant of the enzymeeinhibitor
complex, [I] is the concentration of inhibitor and k is the velocity of
the enzymeeinhibitor complex formation. The final enzyme con-
centration was 9.75 g/mL.
2
m
/K
M were tested at a constant
M) under [S] << K conditions.
values.
i
), various concentrations of
chromophore release was measured within 5 min at 405 nm using
a Cary 3E spectrophotometer (Varian, Mulgrave, Australia) at an
m
m
m
ꢀ
9
enzyme concentration of 4.12 ꢁ 10 M. The substrates concen-
2
ꢀ7
ꢀ5
tration ranged from 1 ꢁ 10 to 4 ꢁ 10 M. Three to five mea-
surements were performed for each compound (systematic error
expressed as a standard deviation did not exceed 20%). The calcu-
lated initial hydrolysis rates were used as a measure of the sub-
strate activity of the investigated peptides. All details of kinetic
i
2
i
2
i
i
2
studies and methods for kinetic parameters calculation (K
and kcat/K ) were performed as described previously [30].
M
, kcat
,
m
M
2.5. Inhibition of SufA-mediated degradation of human fibrinogen
2
2
.3. Inhibitor synthesis
Ten microlitres of 5% citrate-treated human plasma in PBS
(SigmaeAldrich, Stockholm, Sweden) or 10 l of human fibrinogen
1 mg/ml in PBS; SigmaeAldrich, Stockholm, Sweden) were incu-
bated with SufA (250 ng) or buffer for 60 min at 37 C. The cleavage
of fibrinogen was blocked with the specific SufA inhibitor at a final
concentration of 2.5 mM (in DMSO/PBS 1:1, v/v). Next, the samples
were resolved via SDS PAGE followed by staining with 0.25% Coo-
massie R-250 Brilliant blue (SigmaeAldrich, Stockholm, Sweden) in
.3.1. General procedure
All reagents and solvents used in this study were purchased
m
(
ꢂ
from SigmaeAldrich and were used without purification. A detailed
description of the synthetic procedures, as well as spectroscopic
data for all inhibitors used in this study, can be found in the
Supplementary material [31e33]. In general, an amidoalkylation
reaction of triaryl phosphite with benzyl carbamate and an
appropriate aldehyde in acetic acid was applied [34]. For acid-
sensitive aldehydes, a modified amidoalkylation reaction under
mild conditions was used [35]. A representative example of com-
pound 5 preparation is described below. The synthesis of 5 started
with the preparation of 6-formyl-2-naphthonitrile (Scheme 1).
Briefly, dimethyl 2,6-naphthalenedicarboxylate was first selectively
hydrolyzed into 6-(methoxycarbonyl)-2-naphthoic acid [36]. Next,
the carboxylic group was transferred into amide and subsequent
dehydratation led to nitrile formation. The ester group was sub-
sequently reduced into alcohol, the further oxidation of which
40% EtOH and 7% acetic acid.
2.6. Antibacterial activity of compound 5
F. magna strain 505 and Escherichia coli strain B1351 were clin-
ical isolates collected at the Department of Clinical Microbiology,
Lund University Hospital (SUS), Lund, Sweden. F. magna was grown
ꢂ
under strict anaerobic conditions at 37 C (Whitley A35 Anaerobic
Workstation, VWR International AB, Lund, Sweden) in
ToddeHewitt broth (TH, Becton Dickinson AB, Nordic Biolabs, T a€ by,
ꢂ
Scheme 1. General synthesis of compound 5. Reagents and conditions: (a) KOH/MeOH, 1,4-dioxane, 80e90 C; (b) SOCl
2
, ClCH
2
CH
2
Cl, reflux; (c) NH
3
/MeOH, CH
2
Cl
2
, r.t.; (d)
ꢂ
N, ꢀ60 ꢂC; (g) triphenyl phosphite, benzyl carbamate, AcOH, 80e90 C; (h) HCl, CHCl
ꢂ
pyridine, 1,4-dioxane, TFAA, 0 C; (e) LiBH
4
, THF, r.t.; (f) oxalyl chloride, CH
2
Cl
2
, DMSO, Et
3
3
,
ꢂ
EtOH, 0 C; (i) NH
3
/MeOH, r.t.
Please cite this article in press as: E. Burchacka, et al., Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application
to prevent human fibrinogen degradation and bacteria growth in vitro, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.006

4
E. Burchacka et al. / Biochimie xxx (2014) 1e7
Sweden) supplemented with 0.5% Tween-80. E. coli was grown at
final 200 ml volume of MuellereHinton Broth (MHB, Merck, War-
ꢂ
37
C and 5% CO
2
in TH broth. Stationary phase F. magna bacteria
saw, Poland), pH 7.5. Inhibitor 5 or gentamicin was dissolved in
DMSO and diluted with MHB just before addition to the plate.
S. aureus inoculum was prepared from an agar plate, cultured for
(
7
strain 505) were washed with 10 mM TriseHCl, 5 mM glucose, pH
6
ꢀ1
.5 and adjusted to a concentration of 1% (2 ꢁ 10 c.f.u. ml ). Fifty
5
ꢂ
microlitres of bacterial suspension (approximately 10 c.f.u. of
24 h at 37 C and diluted with medium to 0.5 of the McFarland
8
ꢀ1
bacteria) were incubated with various concentrations of inhibitor 5
turbidity standard (1 ꢁ 10 c.f.u. ml ). Then, serially diluted in-
ꢂ
for 1 h at 37 C under strict anaerobic conditions. Serial dilutions of
hibitor 5 or gentamicin was added into the wells containing 190
ml
ꢂ
the mixtures were then plated on TH agar in duplicates and incu-
of S. aureus suspension and incubated at 37 C for 24 h. The number
of survivors were measured (OD600) and MIC values were calcu-
lated. All measurements were performed in duplicate and repre-
sent the average of at least four independent experiments. The
results are presented as the mean ± SEM.
ꢂ
bated at 37 C for three days under strict anaerobic conditions. The
number of c.f.u. were counted. All dilutions were performed in
10 mM TriseHCl, 5 mM glucose, pH 7.5. For analysis of growth in
the presence of gentamicin, F. magna inoculate was prepared from
stationary phase bacteria (strain 505), diluted to a final concen-
6
tration of 2 ꢁ 10 cfu/ml in TH and 2.5 ml was added to 2.5 ml TH
2.7. SufA isolation and purification
containing increasing concentrations of gentamicin (final bacterial
6
ꢂ
concentration 1 ꢁ10 cfu/ml). The bacteria were cultivated at 37 C
SufA used in this study was isolated and purified as described
before [11]. Briefly, F. magna strain ALB8 was treated with papain
and, following digestion, the papain was inactivated by the addition
of E64. The supernatant was collected by centrifugation and papain
was removed by dialysis. The proteins in the supernatant were
separated using ion-exchange chromatography (Mono Q 5/50 GL).
Fractions containing SufA were combined and further purified via
gel filtration (Superose 12 10/300 GL). The enzymatic activity was
analyzed using gelatin zymography and purity of SufA was exam-
ined via SDS PAGE and Western blot analyses.
for 48 h and the number of survivors were measured (OD620) after
2
4 h and 48 h. E. coli strain B1351 was grown to mid-log phase
(
2
OD620 approximately 0.4) in TH broth, washed and diluted to
ꢁ 10 c.f.u. ml in 10 mM TriseHCl, 5 mM glucose, pH 7.5. Fifty
5
6
ꢀ1
microlitres of bacterial suspension (approximately 10 c.f.u. of
bacteria) were incubated with various concentrations of inhibitor 5
ꢂ
for 1 h at 37 C. Serial dilutions of the mixtures were then plated on
ꢂ
TH agar in duplicates and incubated overnight at 37 C. The number
of c.f.u. were then counted. All dilutions were performed in 10 mM
TriseHCl, 5 mM glucose, pH 7.5. For analysis of growth in the
presence of gentamicin, E. coli inoculate was prepared from sta-
tionary phase bacteria (strain B1351) diluted in THY to a final
3. Result and discussion
6
concentration of 1 ꢁ 10 cfu/ml in TH and 2.5 ml was added to
3.1. Substrate profiling
2
.5 ml TH containing increasing concentrations of gentamicin. This
6
gave a bacterial concentration of 0.5 ꢁ10 cfu/ml. The bacteria were
Based on the ability of SufA protease to cleave human fibrinogen
and MIG/CXCL 9 peptide after lysine or arginine [11,12], we fixed
these residues at P1 position of the designed libraries. Iterative
deconvolution of synthesized sublibraries showed the highest
fluorescence intensity for Lys and Thr, incorporated as the P4
ꢂ
cultivated at 37 C and the number of survivors were measured
(
OD620) after 24 h. S. aureus 6538P was purchased from the Polish
Collection of Microorganisms in Wrocław. The assay was performed
using 96-well microtiter plates (Sarstedt, Stare Babice, Poland) in a
Fig. 1. Substrate specificities of SufA protease at P1eP4 positions.
Please cite this article in press as: E. Burchacka, et al., Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application
to prevent human fibrinogen degradation and bacteria growth in vitro, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.006

E. Burchacka et al. / Biochimie xxx (2014) 1e7
5
Table 1
Physicochemical and kinetic characterization of synthesized optimal SufA substrates.
cat [s 1] ꢁ 10ꢀ1
ꢀ
K
[m
M]
kcat/K
M
ꢀ1
[M ꢁ sꢀ1] ꢁ 103
Substrate
MW calc/found
t
R
[min]
k
M
ABZ-Lys-Ile-Ser-Arg-ANB-NH
2
(Su1)
(Su2)
781.5/782.4
757.3/758.5
12.34
11.22
1.3 ± 0.3
4.1 ± 0.5
5.7 ± 0.8
19.7 ± 2.5
226 ± 12
208 ± 15
ABZ-Lys-Ile-Ser-Lys-ANB-NH
2
residues; thus, they were chosen for further screening studies
Fig. 1). The analysis also showed that Ser and Gln residues can be
potential displayed by compound 5 cannot be fully explained.
Currently, study of SufA protease crystallization is underway.
(
relatively efficiently accommodated into the enzyme S4 subside. In
contrast, for libraries containing Trp, Pro or Asp at this position no
visible fluorescence was observed. At position P3 the highest in-
crease of fluorescence was observed for the sublibrary containing
isoleucine. This residue significantly dominated over all the other
amino acid residues tested. In addition, negatively charged amino
acid residues such as Glu or Asp were not recognized by the SufA
protease. A similar pattern of recognition in terms of a near abso-
lute preference towards particular amino acid residues was
observed for position P2, where the highest proteolytic suscepti-
bility revealed a sublibrary with Ser residue. Slightly lower activity
was displayed by libraries with Thr, Tyr and Asn residues at P2
position, whereas no activity was recorded for libraries with Pro
and Glu. Further analysis showed that SufA protease exclusively
recognized positively charged side chains of lysine and arginine at
the S1 binding pocket. However, the SufA preference towards the
Arg residue at P1 position was higher than Lys, since for ABZ-Lys-
3
.3. Inhibition of SufA-mediated fibrinogen degradation in vitro
In order to evaluate the potential of compound 5 to prevent
human fibrinogen degradation by SufA, we used purified fibrinogen
as well as whole human plasma. After pre-incubation of SufA with
compound 5, followed by the addition of fibrinogen or plasma, the
sample was resolved using SDS PAGE under reducing conditions
(Fig. 2). Human fibrinogen is composed of three pairs of chains e
A
a
, B and e that are linked by disulfide bonds into a complex of
b
g
Table 2
Inhibition of SufA protease from Finegoldia magna by phosphonic analogues of
arginine.
Ile-Ser-Arg-ANB-NH
approximately twice as high as for ABZ-Lys-Ile-Ser-Lys-ANB-NH
Su2).
For both of the obtained optimal substrates, the kinetic pa-
rameters were determined (Table 1). Surprisingly, both substrates
2
(Su1) the increase of fluorescence was
2
(
displayed similar specificity constant values (kcat/K
M
)
of
5
ꢀ1 ꢀ1
5
ꢀ1 ꢀ1
s
2
.26 ꢁ 10 M
s
(Su1) and 2.08 ꢁ 10 M
(Su2); however,
Compd
R
1
R
2
K
i
[m
M]
k
2
/K
i
[M 1
ꢀ
s
ꢀ1
]
for substrate containing lysine, the kcat value was approximately 3-
times higher when compared to its arginine counterpart S1, thus
making ABZ-Lys-Ile-Ser-Lys-ANB-NH
assay.
2
more suitable for fluorogenic
1
2
3
4
5
eH
>200
<10
<10
<10
<10
eOeCH
eH
3
>200
3.2. SufA inhibition studies
>200
In parallel to the profiling of SufA protease, we screened a library
of various phosphonic-type compounds for potential inhibitors.
Among all the tested simple -aminoalkylphosphonate diaryl ester
eH
>200
a
e analogues of natural and unnatural amino acids (Val, nVal, Phg,
Phe, hPhe, Leu, Ile, Orn, Ala, Abu) as well as their peptidyl de-
rivatives, we were unable to select any compound showing any
inhibitory activity towards SufA protease. Considering the revealed
SufA substrate specificity, especially its preference towards Arg/Lys
residues at P1 position, we screened the library of structurally
diverse phosphonic analogues of arginine. To our surprise, with
only one exception, none of the tested phosphonic arginine-like
compounds inhibited the proteolytic activity of SufA (Table 2).
4
eH
44.03 ± 1.5
1.08ꢁ10
<10
6
7
8
9
eH
>200
>200
>200
eOeCH
3
<10
The only active derivative that showed notable inhibitory potency
P
was Cbz-6-AmNpth (OPh)
2 2 i
(5), displaying a k /K value of
eCH
3
<10
ꢀ1
ꢀ1
1
0 800 M
very weak inhibitor of human cathepsin G (5% inhibition at 50
32]. The lack of inhibition observed for phosphonic lysine and
s
. Compound 5 had previously been reported as a
mM)
[
eH
eH
>200
>200
<10
<10
arginine analogues may be the result of the high flexibility of their
aliphatic side chains which limits the optimal occupation of the S1
binding pocket of SufA protease. Binding of more rigid naphthyl
rings of compound 5 may lead to the formation of a close bond
network at the bottom of the S1 pocket between an amidino group
of inhibitor and acidic residue(s) of SufA. Unfortunately, since the
crystal structure of SufA protease is unsolved, the inhibitory
1
1
0
1
eH
>200
<10
Please cite this article in press as: E. Burchacka, et al., Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application
to prevent human fibrinogen degradation and bacteria growth in vitro, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.006

6
E. Burchacka et al. / Biochimie xxx (2014) 1e7
4. Conclusions
Application of the FRET-based combinatorial approach for the
profiling of SufA protease revealed an optimal substrate sequence.
The highest preference of SufA protease towards P1 residue was
narrowed to positively charged side chains of arginine and lysine.
SufA also showed a relatively high preference towards amino acid
residues at P2 and P3 positions which were Ser and Ile, whereas it
displayed a broad tolerance towards the residue occupying the S4
pocket accepting side chains of Lys > Thr > Gln > Ser > Ala > Gly.
The optimal substrates of SufA protease were ABZ-Lys-Ile-Ser-Arg-
ANB-NH
2
(Su1) and ABZ-Lys-Ile-Ser-Lys-ANB-NH
2
(Su2) displaying
5
ꢀ1 ꢀ1
5
ꢀ1 ꢀ1
k
cat/K
M
values of 2.26 ꢁ 10
M
s
and 2.08 ꢁ 10
M
s ,
respectively, but since substrate Su2 showed a higher kcat value
ꢀ
1
(0.41 s ) than S1, it was used for further SufA inhibition assay.
Among all synthesized phosphonic analogues of arginine and
P
lysine, only Cbz-6-AmNpth (OPh)
2
(5) showed inhibitory activity
4
ꢀ1 ꢀ1
against SufA protease displaying a k
2
/K
i
value of 1.08 ꢁ 10 M
s .
Fig. 2. Inhibition of SufA-mediated degradation of human fibrinogen by compound 5.
Further analysis demonstrated that compound 5 completely
inhibited degradation of fibrinogen, either purified or human
plasma-derived. Moreover, inhibitor 5 showed an antibacterial ef-
fect on F. magna, S. aureus and E. coli species with the MIC values of
The arrowhead points at the fibrinogen Aa chain and arrows point at degradation
products of the A -chain.
a
1
.4, 14, and 2.8 mg/l, respectively. In addition, since no significant
3
40 kDa [37]. Under reducing conditions, these chains migrate as
cytotoxic effect of compound 5 on normal cell lines (BALB/3T3 and
HLMEC) was observed, this provides the possibility for potential
in vivo experiments.
approximately 70 kDa, 60 kDa and 55 kDa bands, respectively. The
results clearly demonstrated that compound 5 completely pre-
vented degradation of fibrinogen, either purified or human plasma-
derived. In the absence of the inhibitor, control SufA samples
In summary, we have presented complete data on SufA protease
profiling towards residues occupying its S1eS4 binding pockets. In
addition, we have selected a lead compound that not only showed
inhibitory activity towards target protease in an enzymatic assay,
but also was able to prevent the degradation of human fibrinogen
and displayed an antibacterial effect in vitro similar to gentamicin.
Considering the involvement of SufA protease in pathogenesis of
F. magna infection, the developed substrate and inhibitor could help
to establish its role in vivo.
showed a complete disappearance of the fibrinogen A
~70 kDa) into degradation products (32 kDa and 37 kDa).
a-band
(
3.4. Antibacterial effect of compound 5 on bacteria growth
Previous studies have demonstrated that SufA degrades anti-
bacterial peptides, such as LL-37, leading to bacterial survival [11].
Pre-incubation of bacteria with compound 5 resulted in a partial
inhibition of LL-37 cleavage and bacteria death (data not shown).
Interestingly, we found that compound 5 had a direct bactericidal
effect on F. magna (strain 505) in a dose dependent manner. In the
next steps, we investigated if the observed antibacterial effect is
more general and therefore the growth of two other bacteria spe-
cies, S. aureus and E. coli, were included in the study. The MIC values
measured for all bacterial species tested here are presented in
Table 3. Although the highest antibacterial activity of compound 5
was observed for F. magna (MIC ¼ 1.4 mg/l), it showed the potential
to inhibit the growth of E. coli (MIC ¼ 2.8 mg/l) and S. aureus
Conflicts of interest
The authors state that there is nothing to declare.
Acknowledgements
The work was supported by the Swedish Research Council
(project 7480). The authors would like to thank Wroclaw University
of Technology for financial support (Statute Funds S30134/Z0313).
The authors would like to thank Mateusz Psurski from the NeoLek
Laboratory co-financed by The European Regional Development
Fund POIG.02.01.00-02-073/09 for performing the human cell line
cytotoxicity assay.
(
MIC ¼ 14 mg/l). The antibacterial activity of 5 against F. magna and
E. coli was in a range of gentamicin action. In order to further
examine the cytotoxic effect of compound 5 on eukaryotic cells, we
used two normal cell lines: BALB/3T3 (mouse) and HLMEC (hu-
man). Compound 5 did not show any significant cytotoxic effect
displaying IC50 values of 69.7
T3 and HLMEC, respectively. Under the conditions of our assay,
cisplatin, which was used as the control compound, showed the
IC50 values of 4.1 g/mL and 0.21 g/mL towards BALB/3T3 and
m
g/mL and 97.7
m
g/mL towards BALB/
Appendix A. Supplementary material
3
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.biochi.2014.05.006.
m
m
HLMEC, respectively.
References
Table 3
Antibacterial activity of inhibitor 5 and gentamicin on S. aureus, E. coli, F. magna
growth in vitro.
[1] D.A. Murdoch, Gram-positive anaerobic cocci, Clin. Microbiol. Rev. 11 (1)
(1998) 81e120.
[
2] E.C. Murphy, I.M. Frick, Gram-positive anaerobic cocciecommensals and
opportunistic pathogens, FEMS Microbiol. Rev. 37 (4) (2013) 520e553.
3] P.G. Bowler, B.I. Duerden, D.G. Armstrong, Wound microbiology and associ-
ated approaches to wound management, Clin. Microbiol. Rev. 14 (2) (2001)
Bacterial species
MIC [mg/l]
[
Compound 5
Gentamicin
2
44e269.
S. aureus ATTC 6638P
E. coli B1351
F. magna 505
14
2.8
1.4
1.0
16
8.0
[
4] C. Hansson, et al., The microbial flora in venous leg ulcers without clinical
signs of infection. Repeated culture using a validated standardised microbio-
logical technique, Acta Derm. Venereol. 75 (1) (1995) 24e30.
Please cite this article in press as: E. Burchacka, et al., Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application
to prevent human fibrinogen degradation and bacteria growth in vitro, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.006

E. Burchacka et al. / Biochimie xxx (2014) 1e7
7
[
5] P. Stephens, et al., Anaerobic cocci populating the deep tissues of chronic
[22] E. Burchacka, et al., The development of first Staphylococcus aureus SplB
protease inhibitors: phosphonic analogues of glutamine, Bioorg. Med. Chem.
Lett. 22 (17) (2012) 5574e5578.
wounds impair cellular wound healing responses in vitro, Br. J. Dermatol. 148
(
3) (2003) 456e466.
[
[
[
[
6] C. Neut, et al., Analysis of gram-positive anaerobic cocci in oral, fecal and
[23] E. Pietrusewicz, M. Sienczyk, J. Oleksyszyn, Novel diphenyl esters of peptidyl
alpha-aminoalkylphosphonates as inhibitors of chymotrypsin and subtilisin,
J. Enzyme Inhib. Med. Chem. 24 (6) (2009) 1229e1236.
[24] I. Schechter, A. Berger, On the active site of proteases. 3. Mapping the active
site of papain; specific peptide inhibitors of papain, Biochem. Biophys. Res.
Commun. 32 (5) (1968) 898e902.
[25] M. Wysocka, et al., Selection of new chromogenic substrates of serine pro-
teinases using combinatorial chemistry methods, Comb. Chem. High
Throughput Screen. 10 (3) (2007) 171e180.
[26] K. Hojo, et al., Amino acids and peptides. XXXV. Facile preparation of p-
nitroanilide analogs by the solid-phase method, Chem. Pharm. Bull. 48 (11)
(2000) 1740e1744.
vaginal flora, Eur. J. Clin. Microbiol. 4 (4) (1985) 435e437.
7] S. Higaki, M. Morohashi, Characteristics of anaerobes from skin specimens,
Drugs Exp. Clin. Res. 29 (4) (2003) 153e155.
8] I. Brook, Peptostreptococcal infection in children, Scand. J. Infect. Dis. 26 (5)
(
1994) 503e510.
9] S. Higaki, et al., Anaerobes isolated from infectious skin diseases, Anaerobe 5
6) (1999) 583e587.
(
[
[
[
10] C. Karlsson, et al., SufA e a bacterial enzyme that cleaves fibrinogen and
blocks fibrin network formation, Microbiology-Sgm 155 (2009) 238e248.
11] C. Karlsson, et al., SufA e a novel subtilisin-like serine proteinase of Finegoldia
magna, Microbiology 153 (Pt 12) (2007) 4208e4218.
12] C. Karlsson, et al., SufA of the opportunistic pathogen Finegoldia magna
modulates actions of the antibacterial chemokine MIG/CXCL9, promoting
bacterial survival during epithelial inflammation, J. Biol. Chem. 284 (43)
[27] N.A. Sole, G. Barany, Optimization of solid-phase synthesis of [Ala8]-dynor-
phin-A, J. Org. Chem. 57 (20) (1992) 5399e5403.
[28] A. Furka, et al., General-method for rapid synthesis of multicomponent pep-
tide mixtures, Int. J. Pept. Protein Res. 37 (6) (1991) 487e493.
[29] R.A. Houghten, et al., Generation and use of synthetic peptide combinatorial
libraries for basic research and drug discovery, Nature 354 (6348) (1991)
84e86.
[30] A. Lesner, G. Kupryszewski, K. Rolka, Chromogenic substrates of bovine beta-
trypsin: the influence of an amino acid residue in P1 position on their
interaction with the enzyme, Biochem. Biophys. Res. Commun. 285 (5) (2001)
1350e1353.
[31] M. Sienczyk, J. Oleksyszyn, Inhibition of trypsin and urokinase by Cbz-
amino(4-guanidinophenyl)methanephosphonate aromatic ester derivatives:
the influence of the ester group on their biological activity, Bioorg. Med.
Chem. Lett. 16 (11) (2006) 2886e2890.
[32] M. Sienczyk, et al., New potent cathepsin G phosphonate inhibitors, Bioorg.
Med. Chem. 16 (19) (2008) 8863e8867.
(
2009) 29499e29508.
[
13] K. Paschalidou, et al., Highly sensitive intramolecularly quenched fluorogenic
substrates for renin based on the combination of L-2-amino-3-(7-methoxy-4-
coumaryl)propionic acid with 2,4-dinitrophenyl groups at various positions,
Biochem. J. 382 (2004) 1031e1038.
[
[
[
[
[
14] M. Szabelski, M. Rogiewicz, W. Wiczk, Fluorogenic peptide substrates con-
taining benzoxazol-5-yl-alanine derivatives for kinetic assay of cysteine
proteases, Anal. Biochem. 342 (1) (2005) 20e27.
15] J. Oleksyszyn, J.C. Powers, Irreversible inhibition of serine proteases by pep-
tidyl derivatives of alpha-aminoalkylphosphonate diphenyl esters, Biochem.
Biophys. Res. Commun. 161 (1) (1989) 143e149.
16] J. Oleksyszyn, J.C. Powers, Irreversible inhibition of serine proteases by pep-
tide derivatives of (alpha-aminoalkyl)phosphonate diphenyl esters,
Biochemistry 30 (2) (1991) 485e493.
17] J. Oleksyszyn, J.C. Powers, Amino acid and peptide phosphonate derivatives as
specific inhibitors of serine peptidases, Methods Enzymol. 244 (1994)
[33] K.K. Sivaraman, et al., Synthesis and structure-activity relationships of
phosphonic arginine mimetics as inhibitors of the M1 and M17 amino-
peptidases from plasmodium falciparum, J. Med. Chem. 56 (12) (2013)
5213e5217.
[34] J. Oleksyszyn, L. Subotkowska, P. Mastalerz, Diphenyl 1-amino-
alkanephosphonates, Synthesis (1979) 985e986.
423e441.
18] M. Sienczyk, J. Oleksyszyn, Irreversible inhibition of serine proteases - design
and in vivo activity of diaryl alpha-aminophosphonate derivatives, Curr. Med.
Chem. 16 (13) (2009) 1673e1687.
[
[
19] R. Grzywa, M. Sienczyk, Phosphonic esters and their application of protease
control, Curr. Pharm. Des. 19 (6) (2013) 1154e1178.
20] R. Hamilton, B. Walker, B.J. Walker, Synthesis and proteinase inhibitory
properties of diphenyl phosphonate analogues of aspartic and glutamic acids,
Bioorg. Med. Chem. Lett. 8 (13) (1998) 1655e1660.
[35] J. Joossens, et al., Development of irreversible diphenyl phosphonate in-
hibitors for urokinase plasminogen activator, J. Med. Chem. 47 (10) (2004)
2411e2413.
[36] M.D. Wendt, et al., Identification of novel binding interactions in the devel-
opment of potent, selective 2-naphthamidine inhibitors of urokinase. Syn-
thesis, structural analysis, and SAR of N-phenyl amide 6-substitution, J. Med.
Chem. 47 (2) (2004) 303e324.
[
21] E. Burchacka, et al., Phosphonic analogues of glutamic acid as irreversible
inhibitors of Staphylococcus aureus endoproteinase GluC: an efficient syn-
thesis and inhibition of the human IgG degradation, Bioorg. Med. Chem. Lett.
[37] M.W. Mosesson, Fibrinogen and fibrin structure and functions, J. Thromb.
Haemost. 3 (8) (2005) 1894e1904.
23 (5) (2013) 1412e1415.
Please cite this article in press as: E. Burchacka, et al., Substrate profiling of Finegoldia magna SufA protease, inhibitor screening and application
to prevent human fibrinogen degradation and bacteria growth in vitro, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.006