NMR solution structure of the precursor for carnobacteriocin B2,Implications of the a-helical leader section for export and inhibition of type IIa bacteriocin activity Tara Sprules1, Kar
Trang 1NMR solution structure of the precursor for carnobacteriocin B2,
Implications of the a-helical leader section for export and inhibition of type IIa bacteriocin activity
Tara Sprules1, Karen E Kawulka1, Alan C Gibbs2, David S Wishart2and John C Vederas1
1
Department of Chemistry and2Faculty of Pharmacy, University of Alberta, Edmonton, AB, Canada
Type IIa bacteriocins, which are isolated from lactic acid
bacteria that are useful for food preservation, are potent
antimicrobial peptides with considerable potential as
therapeutic agents for gastrointestinal infections in
mam-mals They are ribosomally synthesized as precursors with
an N-terminal leader, typically 18–24 amino acid residues
in length, which is cleaved during export from the
produ-cing cell We have chemically synthesized the full precursor
of carnobacteriocin B2, precarnobacteriocin (preCbnB2),
which has a C-terminal amide rather than a carboxyl, and
also produced preCbnB2(1–64), which is missing two
amino acid residues at the C-terminus (Arg65 and Pro66),
via expression in Escherichia coli as a maltose-binding
protein fusion that is then cut with Factor Xa
PreC-bnB2(1–64) is readily labeled with15N and13C for NMR
studies using the latter approach Multidimensional NMR
analysis of preCbnB2(1–64) shows that, like the parent
bacteriocin, it exists as a random coil in water but assumes
a defined conformation in water/trifluoroethanol mixtures
In 70 : 30 trifluoroethanol/water, the 3D structure of the
preCbnB2 section corresponding to the mature bacteriocin
is essentially the same as reported previously by us for carnobacteriocin B2 (CbnB2) This structure maintains the highly conserved a-helix corresponding to residues 20–38
of CbnB2 that is believed to be responsible for interaction with a target receptor in sensitive cells, including Listeria monocytogenes PreCbnB2 also has a second a-helix from residues 3–13 (i.e )15 to )5 relative to CbnB2) in the leader section of the peptide This helix appears to be conserved in related type IIa bacteriocin precursors based
on sequence analysis It is likely to be a key recognition element for export and processing, and is probably responsible for the considerably reduced antimicrobial activity of preCbnB2 The latter effect may assist the pro-ducing cell in avoiding the toxic effects of the bacteriocin This is the first 3D structure determined for a prebacte-riocin from lactic acid bacteria
Keywords: antibacterials; bacteriocin; NMR structure; pep-tide synthesis; precarnobacteriocin B2
Bacteriocins are potent antimicrobial peptides secreted by
bacteria Those produced by lactic acid bacteria are the
focus of extensive studies because of their potential
application as nontoxic food preservatives, as well as their
possible therapeutic uses in both human and veterinary
medicine [1–5] Nisin is approved in over 80 countries as
a food additive [6] Most bacteriocins from lactic acid
bacteria are synthesized as prepeptides which undergo a
variety of post-translational modifications, ranging from
extensive formation of dehydro residues and lanthionine
bridges in the case of lantibiotics (e.g nisin A) to simple
cleavage of a leader peptide and export across the cell membrane These antimicrobial compounds are divided into classes according to their structural characteristics Type IIa bacteriocins are single peptides characterized by
a conserved YGNGVXC motif in the N-terminus, with the cysteine involved in a disulfide bridge, and are otherwise unmodified except for cleavage of the leader from their precursor (Table 1) They show potent activity against a number of potential Gram-positive food spoilage and pathogenic bacteria, e.g Listeria monocyto-genes, but display no toxicity toward humans or other eukaryotes We reported purification and the primary structure of the first member of this class, leucocin A [7], and in the meantime over 20 such compounds have been identified These heat-stable, cationic peptides typically have 37–48 amino acid residues The solution structures
of three type IIa bacteriocins have been determined by NMR methods: leucocin A [8], carnobacteriocin B2 (CbnB2) [9], and, very recently, sakacin P [10] Interest-ingly, the high sequence homology of the N-terminal portion does not lead to conserved 3D structures for this section of the molecules because of polarity differences of the few varying residues [9] However, in contrast, the much more variable sequences of the C-terminal halves of
Correspondence to J C Vederas, Department of Chemistry,
University of Alberta, Edmonton, AB, T6G 2G2, Canada.
Fax: + 1 780 492 8231, Tel.: + 1 780 492 5475,
E-mail: john.vederas@ualberta.ca
Abbreviations: ABC, ATP-binding cassette; BHI, brain heart infusion;
CbnB2, carnobacteriocin B2; DMF, N,N-dimethylformamide;
HSQC, heteronuclear single quantum coherence; MBP,
maltose-binding protein; preCbnB2, precarnobacteriocin B2.
(Received 7 January 2004, revised 24 February 2004,
accepted 11 March 2004)
Trang 2these bacteriocins maintain a highly conserved a-helical
structure which determines antimicrobial specificity [9,11]
and is likely to be responsible for recognition of a cell
surface protein receptor in the target bacterial cells
Disruptions of this helix through mutations that change
amino acid polarity generally abolish antimicrobial
activ-ity [12–14]
The leader peptides of type IIa bacteriocins are also
homologous [5,15] (Table 1) They range in length from
18 to 24 amino acids, terminating in two glycine residues
Hydrophobic residues are found at positions)4, )7, )12
and)15, and hydrophilic residues are found at positions
)5, )6, )8, )9, )11, )13 and )14 Along with genes that
encode bacteriocin and immunity proteins, which protect
the producer strain from attack by its own antimicrobial
peptides, genes that encode ATP-binding cassette (ABC)
transporter proteins are usually found in these bacteriocin
operons [5] ABC transporters are a family of large
transmembrane proteins responsible for the ATP-driven
transport of a variety of compounds ranging from ions to
oligosaccharides and proteins The ABC proteins
associ-ated with type IIa bacteriocin production have an extra
N-terminal cysteine protease domain, which is responsible
for cleavage of the leader peptide after the double glycine
motif [16,17] This cleavage occurs on the cytoplasmic side
of the membrane during export of the bioactive peptide
[18] The leader peptide is undoubtedly recognized by the
ABC protein responsible for export and processing In the
case of carnobacteriocin B2 (CbnB2), a cationic
thermo-stable type IIa bacteriocin produced by Carnobacterium
piscicolaLV17B [19], the additional 18 amino acids in the
leader peptide of precarnobacteriocin B2 (preCbnB2) also
render it 125 times less active than the mature peptide
[14] To help determine the structural basis of the
inhibition of the antimicrobial activity of CbnB2 by the
leader and to assist future analysis of the interaction of
preCbnB2 with its ABC transporter protease, it is
essential to establish the preferred geometry of the
precursor We now report the chemical synthesis,
bio-chemical production, and solution structure of preCbnB2,
and compare it with the structure of the mature
bacteriocin, CbnB2
Materials and methods
Chemical synthesis of preCbnB2 with a C-terminal amide Stepwise synthesis of preCbnB2 (but with a C-terminal amide instead of carboxyl) was achieved manually on a 0.3 mmol scale of Rink amide resin using standard Fmoc solid-phase peptide chemistry Fmoc-protected L-amino acids (Sigma) were used with the following orthogonal protection: Arg(Pmc), Asn(Trt), Asp(tBu), Cys(Acm), Cys(Trt), Glu(tBu), His(Trt), Lys(Boc), Ser(tBu), Thr(tBu), Trp(Boc), Tyr(tBu) Initially, Fmoc-Pro was coupled to Rink resin using N,N-dicyclohexyl-carbodiimide as the activating agent in the presence of a catalytic amount
of N-hydroxybenzotriazole in N,N-dimethylformamide (DMF) The peptide chain was assembled in a stepwise manner with deprotection, activation and coupling cycles All steps were followed by washing sequentially with DMF, dichloromethane, propan-2-ol, dichloromethane and DMF,
to remove excess reagents and protecting groups Fmoc deprotection was catalyzed by 20% piperdine in DMF Amino acid activation was achieved with O-benzotriazole-N,N,N¢,N¢-tetramethyluronium hexafluorophosphate in DMF, and a fourfold excess of the Fmoc-protected amino acid was used to maximize yield of the N-hydroxybenzo-triazole-amino acid ester The extent of peptide elongation,
or coupling, was monitored by using a ninhydrin assay to check for residual free a-amine To facilitate coupling of difficult residues, combinations of the following condi-tions were used: elevated temperatures of up to 50C, O-7-azabenzotriazole-1-yl-N,N,N¢,N¢-tetramethyluronium hexafluorophosphate as activation agent, and N-methyl-pyrrolidinone as solvent Test cleavages were performed after every five-residue coupling, and the desired product was confirmed by MALDI-TOF MS Peptide was cleaved from the resin with a mixture of 87.5% trifluoroacetic acid, 5% phenol, 5% water, 2.5% dithiothreitol, and 2.5% anisole for 90 min at 20 C The filtrate from the cleavage reactions was collected, combined with trifluoroacetic acid washes (3· 2 min, 1 mL), and concentrated in vacuo Cold diethyl ether ( 15 mL) was added to precipitate the crude cleaved peptide Disulfide bond formation was achieved by
Table 1 Comparison of the amino acid sequence of selected type IIa bacteriocins Conserved hydrophobic residues within the leader peptide are italicized, and hydrophilic residues are underlined The YGNGVXC motif is highlighted in bold text.
Bacteriocin Leader peptide Mature peptide Reference PreCbnB2 MNS V KE L NVKE M KQ L HGG VN YGNGVSC SKTKCSVNWGQAFQERYTAGINSFVSGVASGAGS
IGRRP
[19] Sakacin G MKN A KS L TIQE M KS I TGG KY YGNGVSC NSHGCSVNWGQAWTCGVNHLANGGHGVC [20] Plantaricin 423 MMKK I EK L TEKE M AN I IGG KY YGNGVTC GKHSCSVNWGQAFSCSVSHLANFGHGKC [21] Piscicolin 126 MKT VKELS V KEMQLT TGG KY YGNGVSC NKNGCTVDWSKAIGIIGNNAAANLTTGGAAGWNKG [22] CbnBM1 MKS VKELN K KEMQQI NGG AIS YGNGVYC NKEKCWVNKAENKQAITGIVIGGWASSLAGMGH [19] Leucocin A MMNMKPTES YEQLD N SALEQV VGG KY YGNGVHC TKSGCSVNWGEAFSAGVHRLANGGNGFW [7] Pediocin MKK I EK L TEKE M AN I IGG KY YGNGVTC GKHSCSVDWGKATTCIINNGAMAWATGGHQGNHKC [23,24] Mesentericin
Y105
MTNMKSVEA Y QQ L DNQN L KK V VGG KY YGNGVHC TKSGCSVNWGEAASAGIHRLANGGNGFW [25] Sakacin A MNN V KE L SMTE L QT I TGG ARS YGNGVYC NNKKCWVNRGEATQSIIGGMISGWASGLAGM [26] Divercin V41 MKNLKEGS Y TA V NTDE L KS I NGG TKY YGNGVYC NSKKCWVDWGQASGCIGQTVVGGWLGGAIPGKC [27] Sakacin P MEK F IE L SLKE V TA I TGG KY YGNGVHC GKHSCTVDWGTAIGNIGNNAAANWATGGNAGWNK [28]
Trang 3stirring the reaction mixture with elemental iodine (0.1Min
methanol) under argon for 2 h The oxidation was
quenched by addition of aqueous ascorbic acid (1M), and
the crude product was purified by RP-HPLC The fraction
showing the correct mass spectrum (mass 6990.8) was
collected and lyophilized to give 10 mg peptide with >90%
purity
Expression of labeled MBP-PreCbnB2 fusion protein,
cleavage and purification of PreCbnB2(1–64)
Escherichia coli BL21 (DE3) cells transformed with the
plasmid pLQP, expressing preCbnB2 as a maltose-binding
protein (MBP) fusion, was grown with shaking at 37C
in M9 minimal medium as described previously [14]
For labeling experiments, [15NH4]2SO4, or alternatively
[15NH4]2SO4 and D-[U-13C]glucose (99% isotopic purity;
Cambridge Isotope Laboratories, Woburn, MA, USA),
were used as sole nitrogen and carbon sources
Recom-binant protein production was induced with 0.3 mM
isopropyl thio-b-D-galactoside when A600of the cell culture
reached 0.5 The culture was incubated for a further 3 h at
37C, and the cells were harvested by centrifugation The
cell pellet was resuspended in column buffer (20 mMTris/
HCl, 200 mMNaCl, 1 mMEDTA, 1 mMNaN3and 1 mM
dithiothreitol; 20 mLÆL)1 cell culture), and lysozyme
(0.1 mgÆmL)1) was added The cells were subjected to three
freeze-thaw cycles and sonicated for 2 min The cells were
then centrifuged, applied to a column of amylose resin (New
England Biolabs), washed overnight with column buffer,
and eluted with 10 mMmaltose, according to the
manufac-turer’s protocol The fractions were analyzed by
MALDI-TOF MS and also by SDS/PAGE (12%) to determine if
they contained the expected MBP-PreCbnB2 fusion protein
The appropriate fractions were combined and dialyzed
extensively against distilled, deionized water and
lyophi-lized Typical yield was 70–100 mg fusion protein per litre of
culture The target peptide was cleaved from MBP with
Factor Xa (Borean Biologics Aps, Aarhus, Denmark) in
20 mMTris/HCl/100 mMNaCl/1 mMCaCl2[0.01 mgÆ(mg
fusion protein))1] overnight at room temperature The
resulting peptide was separated from MBP on a C18
column (Waters PrepPak), with a 20 minute gradient of
20–60% acetonitrile in water with 0.1% trifluoroacetic acid
The target peptide was eluted at 28% acetonitrile Fractions
containing this peptide were combined and lyophilized
Typical yield is 4–6 mg per litre of initial fermentation MS
analysis (see below) of the MBP fusion protein was in
agreement with that expected (mass 48 300 Da) for
unlabe-led protein However, the MS of the target peptide showed
that the two C-terminal amino acid residues (Arg65 and
Pro66) were absent [observed 6739.2 ± 0.8 for unlabeled
peptide; calculated 6738.5 for preCbnB2(1–64) missing the
two C-terminal amino acids; calculated 6991.8 for
preC-bnB2 having all 66 residues] The mass spectra of
15N-labeled preCbnB2(1–64) derived from [15NH4]2SO4
showed a predominant mass peak at 6825.8 ± 0.8,
corres-ponding to complete 15N-labeling of all 87 nitrogens
(backbone and side chain) PreCbnB2(1–64) showed CD
spectra and biological activity that were indistinguishable
from mature preCbnB2 As the peptide samples were
susceptible to degradation when stored in solid form for a
prolonged period, they were dissolved and stored in 70 : 30 trifluoroethanol-d3 (99.9%)/H2O or 70 : 30 trifluoroetha-nol-d3(99.9%)/D2O (100%) at a concentration of 1 mM Mass spectrometry
MS analysis used MALDI-TOF on an Applied Biosystems Voyager Elite instrument in the positive ion mode with an acceleration voltage of 20 kV using a nitrogen laser (k¼ 337 nm) Samples were prepared using a-cyano-4-hydroxycinnamic acid (Aldrich) or sinapinic acid (Aldrich)
as a matrix, and fixed to a gold or stainless-steel target before analysis The instrument was calibrated daily before each experiment using apomyoglobin [MH+¼ 16 952.56] and trypsinogen [MH+¼ 23 981.9] as standards for MBP fusion proteins and insulin [MH+¼ 5734.59] and insulin chain B [MH+¼ 3496.96] for peptides
CD spectroscopy All CD measurements were performed by R Luty (Depart-ment of Biochemistry, University of Alberta) on a Jasco J-720 spectrophotometer equipped with JASCO J700 soft-ware A thermally controlled quartz cell with a 0.02 cm path length over 180–250 nm was used CD spectra of preCbnB2
at different concentrations and varying the solvent from 0%
to 70% trifluoroethanol in water were collected at 25C Data were collected every 0.05 nm and were the average of eight scans The bandwidth was set at 1.0 nm and the sensitivity at 50 mdegrees, and the response time was 0.25 s
In all cases, baseline scans of aqueous buffer were subtracted from the experimental readings Results are expressed in units of molar ellipticity per residue (degreesÆcm2Ædmol)1) and plotted against wavelength
Antibacterial activity Antibacterial activity was determined by the spot-on-lawn test essentially as previously reported [14] The indicator organism, L monocytogenes LI0502, was grown overnight
in 7.5 mL tryptic soy broth with yeast extract or brain heart infusion (BHI) broth without shaking at 30C Serial twofold dilutions of preCbnB2 (in Luria–Bertani broth) were spotted (20 mL) on to a BHI hard agar plate, allowed
to dry, and overlaid with 7.5 mL of a 1% solution of the indicator organism in BHI soft agar Zones of inhibition were measured after 16 h incubation at 37C
NMR spectroscopy NMR experiments were recorded on a Varian INOVA-600 spectrometer Unless otherwise stated, all experiments were performed on labeled preCbnB2(1–64), the biological activity and CD spectra of which were indistinguishable from the parent 66-amino acid peptide, preCbnB2 15N HSQC [29],15N HSQC-TOCSY [30],15N HSQC-NOESY [30], 13C HSQC, 13C HSQC-NOESY [31] and a 2D NOESY spectrum were recorded at 35C 15N HSQC,
15N HSQC-NOESY, HNHA [32–34], 13C HSQC and
13C HSQC-NOESY spectra were recorded at 20C 15N HSQC-NOESY mixing times were 200 ms; the13C HSQC-NOESY experiments were recorded with 150 ms mixing
Trang 4times The TOCSY experiment was recorded with a 60-ms
spinlock Chemical shifts were referenced to an internal
standard of 2,2-dimethyl-2-silapentane-5-sulfonic acid [35]
Data were processed with NMRpipe [36], and data analysis
was performed with NMRView [37] Data were multiplied
by a 90-shifted sine-bell-squared function in all dimensions
Indirect dimensions were doubled by linear prediction
and zero-filled to the nearest power of 2, before Fourier
transformation
Structure calculations
A total of 649 NOE restraints were obtained from15N and
13C edited HSQC-NOESY experiments on labeled
preC-bnB2(1–64), and classified as strong, medium and weak,
corresponding to distance restraints of 1.8–2.8 A˚, 1.8–3.4 A˚
and 1.8–5.0 A˚, respectively Thirty-three backbone / angles
were obtained from analysis of the diagonal-peak to
cross-peak intensity ratio in the HNHA experiment, with torsion
angles calculated from the Karplus equation [34] and
assigned a variance of ± 15 Fifty-eight backbone w angle
restraints were obtained from analysis of dNa/daN ratios
[38] The w angle restraint was set to)30 ± 110 for dNa/
daN ratios less than 1, and to 120 ± 100 for dNa/daN
ratios greater than 1 Initially 100 structures were calculated
with CNS 1.1 [39] using 327 intraresidue, 201 sequential, 115
medium-range and six long-range NOEs The resulting
structures were subjected to a second round of simulated
annealing with the addition of 22 hydrogen bonds and 91
dihedral angles The 20 lowest energy accepted structures
(no NOE violations >0.5 A˚, no dihedral angle violations
>5) with no residues in the disallowed region of the
Ramachandran plot (PROCHECK[40] analysis) were chosen
to represent the structure The co-ordinates of preCbnB2(1–
64) have been deposited in the Protein Data Bank as 1RY3
Results and discussion
Production of target peptides
Like other bacteriocins, CbnB2 is ribosomally synthesized
as a prepeptide, PreCbnB2, by C piscicola [19] PreCbnB2
consists of the mature CbnB2 sequence (48 amino acids)
preceded at the N-terminus by an 18-amino acid leader
peptide, which is cleaved at the Gly–Gly site during the
maturation process to liberate the active peptide The goal
of this study was determination of the 3D solution structure
of preCbnB2 to assist in obtaining a molecular level
understanding of the reduced activity of preCbnB2 (relative
to the mature bacteriocin) and the recognition events
required for export and cleavage by the ABC transporter
Initially it appeared that the structure of preCbnB2 could be
obtained by modern protein NMR techniques using the
compound without isotopic labels Hence, we chemically
synthesized this 66-amino acid peptide as its C-terminal
amide using standard solid-phase methods with Fmoc
chemistry and acid-labile protecting groups where necessary
on the side chains This provided sufficient material (10 mg)
suitable for a variety of biological (e.g antimicrobial
activity) and spectroscopic (e.g MS, CD) studies However,
it soon became evident that preCbnB2 has limited solubility
as well as a strong tendency to self-associate and precipitate
as insoluble aggregates from aqueous solutions at concen-trations required for effective NMR analysis PreCbnB2 also displays a tendency to decompose or denature in the absence of solvent Thus, the enhanced sensitivity in NMR studies afforded by isotopic labeling with 15N and13C of more dilute samples proved essential
A practical route to isotopic labeling involves the previously reported [14] expression of a fusion of MBP and preCbnB2 in E coli This system allows use
of (15NH4)2SO4, or alternatively (15NH4)2SO4 and
D-[U-13C]glucose, as sole nitrogen and carbon sources Purification of substantial quantities (70–100 mg per litre of fermentation) of fusion protein is easy, but large-scale proteolytic cleavage of the MBP portion by Factor Xa also results in removal of two C-terminal amino acids of preCbnB2, namely Arg65 and Pro66, to yield the truncated preCbnB2(1–64) as the major product However, this material displays CD spectra and antimicrobial properties indistinguishable from the complete preCbnB2 The two missing hydrophilic residues are at the C-terminus of the peptide, which is unstructured in the mature CbnB2 [9], and distant from the N-terminal leader portion The C-terminal section is also random coil in preCbnB2(1–64) (see below) Significantly, substitution of Arg46, corresponding to Arg64
in preCbnB2, with glycine in CbnB2 has no effect on its activity [14] Hence no significant conformational differ-ences are expected between preCbnB2 and preCbnB2(1–64) Structure of preCbnB2
The structures of all mature type IIa bacteriocins examined thus far are primarily random coil in water, but in lipophilic solvents, such as trifluoroethanol or dodecylphosphocholine micelles, assume a highly conserved amphipathic a-helix which begins at about residue 18 and continues for several turns toward the C-terminus [8–10] As these bacteriocins are well known to interact with membrane lipids and recognize a chiral membrane-bound receptor molecule [2,11], the a-helical portion of the structure of mature type IIa bacteriocins is critical for their biological activity [8–10]
CD experiments (data not shown) demonstrate that, like the mature bacteriocin, preCbnB2 is unstructured in pure water but contains one or more a-helical sections in more lipophilic trifluoroethanol/water mixtures To directly compare the structure of PreCbnB2(1–64) with that reported by us for mature CbnB2 [9], the prepeptide was initially dissolved in a
90 : 10 trifluoroethanol/water However, preCbnB2 is very sparingly soluble in this solvent system, possibly because of the increase in the number of charged and hydrophilic residues in the leader peptide region Increasing the propor-tion of water to 30% results in the best combinapropor-tion of peptide solubility and spectral dispersion of the amide protons, which indicates the maintenance of organized structural elements (Fig 1) As described below, the use of a higher proportion of water does not significantly alter the 3D structure of the section corresponding to the mature peptide Complete proton, nitrogen and carbon assignments were obtained for PreCbnB2(1–64) using a combination of
15N HSQC-TOCSY,15N HSQC-NOESY and13C HSQC spectra Sequential assignments were made by following the pattern of dN,N±1 NOEs Analysis of 3JHNHa coupling constants, Ha, Ca and Cb chemical shifts, and NOEs
Trang 5indicated the presence of two a-helices: one running from
residues)15 to )5 in the leader peptide, and a second from
residues 20–38 in the C-terminal portion of the molecule No
other regular secondary-structure elements were identified
The solution structure of PreCbnB2(1–64) was calculated
based on 649 NOEs, 91 dihedral angles, and 22 hydrogen
bonds The two a-helices are separated by a stretch of
relatively unstructured peptide (Fig 2) The rmsd for helix
1, from residues)15 to )5 is 0.6 ± 0.2 A˚, and for helix 2
0.8 ± 0.3 A˚ In contrast with mature CbnB2, where no
NOEs were observed indicative of a disulfide bridge [9], at
35C Ha-Ha and Ha-Hb NOEs are seen across the
disulfide bridge, although at lower temperature (20C) line
broadening precludes their observation Cysteine b carbon
chemical shifts have been found to be quite sensitive to
oxidation state [41] The Cb chemical shifts for oxidized
cysteine is 40.7 ± 3.8 p.p.m and those of reduced cysteine
fall in the range of 28.4 ± 2.4 p.p.m The Cb chemical
shifts for both cysteines in PreCbnB2(1–64) fall clearly in the
oxidized region (44.1 and 43.9 p.p.m.), confirming that a
disulfide bridge is present Because of the absence of NOE
and 13C chemical shift data indicative of disulfide bond
formation in mature CbnB2, MS and chemical modification
experiments were used to determine that the disulfide bridge
was present [9]
The structures of the region common to both mature
CbnB2 and its precursor are nearly identical: both possess a
poorly defined reverse turn and relatively unstructured
N-terminus, where only sequential NOEs are observed,
followed by the highly conserved amphipathic helix from
residues 20–38 The a-helix in PreCbnB2(1–64) is very
slightly less well defined at its N-terminus This may be due
to fewer hydrogen bond and dihedral angle restraints observed because of overlap of Ha and HN chemical shifts The slightly increased proportion of water (30% vs 10% for previous experiments with mature CbnB2 [9]) may also
Fig 1. 15N HSQC of preCbnB2 Spectra recorded in 70 : 30 trifluoroethanol/H 2 O at
35 C and 600 MHz.
Fig 2 Solution structure of preCbnB2 in 70% trifluoroethanol The positions of the N-terminus and C-terminus are indicated, as are the residues at the start and finish of each a-helix The position of the double glycine motif where cleavage occurs during maturation of the bacteriocin is indicated by an arrow.
Trang 6account for this very minor difference The side chain
chemical shifts are very similar across the protein (within ±
0.1 p.p.m.), with the most variance observed at the
N-terminus, next to the junction with the leader peptide
Similar patterns of NOEs are observed as well Thus, the
critical a-helix in the mature portion of the bacteriocin is
conserved in lipophilic environments, not only in all type IIa
bacteriocins examined to date [8–10], but also in the
precursor, preCbnB2
Mutations that disrupt the helix of mature CbnB2, for
example replacement of Phe33 by serine, render the
bacteriocin completely inactive and greatly alter its retention
time on RP-HPLC [14] Many additional experiments
support the importance of this helix for membrane-bound
receptor recognition and consequent antimicrobial activity
[2,10] For example, Fimland et al [42] have shown that
interchange of large domains of different type IIa
bacterio-cins, such as pediocin PA-1, sakacin P, and curvacin A,
gives chimeric peptides, the antimicrobial specificity of
which for target organisms corresponds to the C-terminal
region It has also been demonstrated that a C-terminal
fragment of pediocin (residues 20–34) specifically inhibits
pediocin PA-1 activity [43] Interestingly, this short peptide
is not antimicrobially active and does not significantly
inhibit closely related bacteriocins such as leucocin A,
sakacin P, and curvacin A The receptor for type IIa
bacteriocins may involve an extracellular domain present in
the MptD subunit of EIIMan
t of the mannose phospho-transferase system [44] Although these types of experiments
have not yet been performed with prebacteriocins, the
present work shows that the leader peptide does not
interfere with formation of the key C-terminal a-helix, and
as a result, all of the precursors are likely to possess significant antimicrobial activity with specificity similar to the mature type IIa bacteriocin
The leader peptide, which consists of the first 18 amino acids of PreCbnB2(1–64), forms a 10-amino acid a-helix, from Val15 to Gln5, as we had previously predicted [19] Although it is not as clearly defined as the amphipathic a-helix in the C-terminus of the polypeptide, the leader peptide helix presents both a charged and hydrophobic face (Fig 3) The region between the two a-helices is relatively unstructured, permitting them to approach one another (Fig 4), thereby making contact between the hydrophobic faces of the two helices possible to give a closed jackknife structure This could interfere to some extent with recog-nition of the mature bacteriocin a-helix by the putative receptor in membranes of target bacteria and somewhat diminish the activity of the prebacteriocin Although no
Fig 3 Amphipathic a-helical structure in PreCbnB2 (A) Leader
pep-tide a-helix (B) C-terminal a-helix The charged residues are coloured
blue, hydrophilic residues white, and hydrophobic residues purple.
Fig 4 Superimposition of preCbnB2 structures (A) Superimposition
of the backbone of the leader peptide a-helix for 20 structures (B) Superimposition of the backbone of helix 2 Both helix domains are highly conserved in each case, but the flexible linker region displaces their relative positions.
Trang 7long-range NOEs were observed between side chains in the
two a-helices under the conditions used in this study, this
may reflect a relatively weak or short-lived interaction The
hydrophilic positively charged face of the leader helix, which
contains three lysines, could potentially also interact with
the negatively charged membrane of Gram-positive
bac-teria, thereby hindering the ability of the prebacteriocin to
fully interact with the target receptor These observations
are consistent with the maintenance of significant
antibac-terial activity for the prebacteriocin, but at a greatly reduced
level ( 125 fold)
The amphipathic nature of the leader peptide is likely to
be critical for its export and processing Analysis of the
sequences of leader peptides for type IIa bacteriocins
(Table 1) indicates that an a-helix as seen in this section
with preCbnB2(1–64) (Fig 5) is likely to be present in all of
the structures of the prepeptides The sakacins and
carno-bacteriocin BM1 (CbnBM1), which have short leader
peptides, are especially closely related in the arrangement
of hydrophobic and hydrophilic residues in this portion
The ABC transporter protein must recognize this leader
portion not only for export, but also for processing by its
cysteine proteinase domain [17] In the absence of 3D
structures of these transporter proteases, at this stage it is
impossible to ascertain the molecular details of these
recognition events However, it is likely that interactions
of the hydrophobic faces of the leader helix with the ABC transporter play a key role in recognition, as the bacteriocin must pass through the membrane, a helix-inducing envi-ronment, during export and processing
In summary, this work describes the chemical synthesis and properties of the 66-amino acid bacteriocin precursor preCbnB2, and the biochemical production and isotopic labeling of preCbnB2(1–64), along with its detailed NMR analysis and 3D structure The sequence homology with other type IIa prebacteriocins indicates that the structural motif of two amphipathic a-helices connected in the cleavage region by a flexible hinge will be present in all such peptides Knowledge of such structures provides a basis for understanding the protein–protein interactions involved in transport, processing and antimicrobial activity
Acknowledgements
We thank Robert Luty (Department Biochemistry, University of Alberta) for performing all CD experiments Albin Otter (Department
of Chemistry) and Ryan T McKay (NANUC, University of Alberta) are gratefully acknowledged for assistance with NMR experiments We thank Mark Williams (Department of Medical Microbiology & Immunology, University of Alberta) for helpful discussions These investigations were supported by the Natural Sciences and Engineering Research Council of Canada, the Alberta Heritage Foundation for Medical Research, CanBiocin Ltd (Edmonton, AB), and the Canada Research Chair in Bioorganic and Medicinal Chemistry.
References
1 O’Sullivan, L., Ross, R.P & Hill, C (2002) Potential of bacter-iocin-producing lactic acid bacteria for improvements in food safety and quality Biochimie 84, 593–604.
2 Nes, I.F & Holo, H (2000) Class II antimicrobial peptides from lactic acid bacteria Biopolymers 55, 50–61.
3 Garneau, S., Martin, N.I & Vederas, J.C (2002) Two-peptide bacteriocins produced by lactic acid bacteria Biochimie 84, 577–592.
4 McAuliffe, O., Ross, R.P & Hill, C (2001) Lantibiotics: structure, biosynthesis and mode of action FEMS Microbiol Rev 25, 285–308.
5 van Belkum, M.J & Stiles, M.E (2000) Nonlantibiotic anti-bacterial peptides from lactic acid bacteria Nat Prod Rep 17, 323–335.
6 Delves-Broughton, J., Blackburn, P., Evans, R.J & Hugenholtz,
J (1996) Applications of the bacteriocin, nisin Antonie Van Leeuwenhoek 69, 193–202.
7 Hastings, J.W., Sailer, M., Johnson, K., Roy, K.L., Vederas, J.C.
& Stiles, M.E (1991) Characterization of leucocin A-UAL 187 and cloning of the bacteriocin gene from Leuconostoc gelidum.
J Bacteriol 173, 7491–7500.
8 Fregeau Gallagher, N.L., Sailer, M., Niemczura, W.P., Nakashima, T.T., Stiles, M.E & Vederas, J.C (1997) Three-dimensional structure of leucocin A in trifluoroethanol and dodecylphosphocholine micelles: spatial location of residues critical for biological activity in type IIa bacteriocins from lactic acid bacteria Biochemistry 36, 15062–15072.
9 Wang, Y., Henz, M.E., Fregeau Gallagher, N.L., Chai, S., Gibbs, A.C., Yan, L.Z., Stiles, M.E., Wishart, D.S & Vederas, J.C (1999) Solution structure of carnobacteriocin B2 and implications for structure-activity relationships among type IIa bacteriocins from lactic acid bacteria Biochemistry 38, 15438–15447.
Fig 5 NOESY spectrum of preCbnB2.1H- 1 H strips from a 3D 13 C
HSQC-NOESY taken at planes corresponding to the a-carbon of
the indicated residues b i+3 -a i and cCH3 i+3 -a i NOEs indicative of
a-helical structure are boxed.
Trang 810 Uteng, M., Hauge, H.H., Markwick, P.R., Fimland, G.,
Mant-zilas, D., Nissen-Meyer, J & Muhle-Goll, C (2003)
Three-dimensional structure in lipid micelles of the pediocin-like
antimicrobial peptide sakacin P and a sakacin P variant that is
structurally stabilized by an inserted C-terminal disulfide bridge.
Biochemistry 42, 11417–11426.
11 Yan, L.Z., Gibbs, A.C., Stiles, M.E., Wishart, D.S & Vederas,
J.C (2000) Analogues of bacteriocins: antimicrobial specificity
and interactions of leucocin A with its enantiomer,
carnobacter-iocin B2, and truncated derivatives J Med Chem 43, 4579–4581.
12 Fimland, G., Eijsink, V.G & Nissen-Meyer, J (2002) Mutational
analysis of the role of tryptophan residues in an antimicrobial
peptide Biochemistry 41, 9508–9515.
13 Miller, K.W., Schamber, R., Osmanagaoglu, O & Ray, B (1998)
Isolation and characterization of pediocin AcH chimeric protein
mutants with altered bactericidal activity Appl Environ Microb.
64, 1997–2005.
14 Quadri, L.E., Yan, L.Z., Stiles, M.E & Vederas, J.C (1997) Effect
of amino acid substitutions on the activity of carnobacteriocin B2.
Overproduction of the antimicrobial peptide, its engineered
var-iants, and its precursor in Escherichia coli J Biol Chem 272,
3384–3388.
15 Ennahar, S., Sashihara, T., Sonomoto, K & Ishizaki, A (2000)
Class IIa bacteriocins: biosynthesis, structure and activity FEMS
Microbiol Rev 24, 85–106.
16 Venema, K., Kok, J., Marugg, J.D., Toonen, M.Y.,
Ledeboer, A.M., Venema, G & Chikindas, M.L (1995)
Func-tional analysis of the pediocin operon of Pediococcus acidilactici
PAC1.0: PedB is the immunity protein and PedD is the precursor
processing enzyme Mol Microbiol 17, 515–522.
17 Havarstein, L.S., Diep, D.B & Nes, I.F (1995) A family of
bacteriocin ABC transporters carry out proteolytic processing of
their substrates concomitant with export Mol Microbiol 16,
229–240.
18 Franke, C.M., Tiemersma, J., Venema, G & Kok, J (1999)
Membrane topology of the lactococcal bacteriocin ATP-binding
cassette transporter protein LcnC Involvement of LcnC in
lac-tococcin A maturation J Biol Chem 274, 8484–8490.
19 Quadri, L.E., Sailer, M., Roy, K.L., Vederas, J.C & Stiles, M.E.
(1994) Chemical and genetic characterization of bacteriocins
produced by Carnobacterium piscicola LV17B J Biol Chem 269,
12204–12211.
20 Simon, L., Fremaux, C., Cenatiempo, Y & Berjeaud, J.M (2002)
Sakacin G, a new type of antilisterial bacteriocin Appl Environ.
Microbiol 68, 6416–6420.
21 van Reenen, C.A., Dicks, L.M & Chikindas, M.L (1998)
Isola-tion, purification and partial characterization of plantaricin 423, a
bacteriocin produced by Lactobacillus plantarum J Appl
Micro-biol 84, 1131–1137.
22 Jack, R.W., Wan, J., Gordon, J., Harmark, K., Davidson, B.E.,
Hillier, A.J., Wettenhall, R.E., Hickey, M.W & Coventry, M.J.
(1996) Characterization of the chemical and antimicrobial
prop-erties of piscicolin 126, a bacteriocin produced by Carnobacterium
piscicola JG126 Appl Environ Microbiol 62, 2897–2903.
23 Henderson, J.T., Chopko, A.L & van Wassenaar, P.D (1992)
Purification and primary structure of pediocin PA-1 produced by
Pediococcus acidilactici PAC-1.0 Arch Biochem Biophys 295,
5–12.
24 Motlagh, A.M., Bhunia, A.K., Szostek, F., Hansen, T.R.,
Johnson, M.C & Ray, B (1992) Nucleotide and amino acid
sequence of pap-gene (pediocin AcH production) in Pediococcus
acidilactici H Lett Appl Microbiol 15, 45–48.
25 Hechard, Y., Derijard, B., Letellier, F & Cenatiempo, Y (1992)
Characterization and purification of mesentericin Y105, an
anti-Listeria bacteriocin from Leuconostoc mesenteroides J Gen.
Microbiol 138, 2725–2731.
26 Holck, A., Axelsson, L., Birkeland, S.E., Aukrust, T & Blom, H (1992) Purification and amino acid sequence of sakacin A, a bacteriocin from Lactobacillus sake Lb706 J Gen Microbiol 138, 2715–2720.
27 Metivier, A., Pilet, M.F., Dousset, X., Sorokine, O., Anglade, P., Zagorec, M., Piard, J.C., Marion, D., Cenatiempo, Y & Fremaux, C (1998) Divercin V41, a new bacteriocin with two disulphide bonds produced by Carnobacterium divergens V41: primary structure and genomic organization Microbiology 144, 2837–2844.
28 Tichaczek, P.S., Vogel, R.F & Hammes, W.P (1994) Cloning and sequencing of sakP encoding sakacin P, the bacteriocin produced
by Lactobacillus sake LTH 673 Microbiology 140, 361–367.
29 Kay, L.E., Keifer, P & Saarinen, T (1992) Pure absorption gradient enhanced heteronuclear single quantum correlation spectroscopy with improved sensitivity J Am Chem Soc 114, 10663–10665.
30 Zhang, O., Kay, L.E., Olivier, J.P & Forman-Kay, J.D (1994) Backbone1H and15N resonance assignments of the N-terminal SH3 domain of drk in folded and unfolded states using enhanced-sensitivity pulsed field gradient NMR techniques J Biomol NMR
4, 845–858.
31 Pascal, S.M., Muhandiram, D.R., Yamazaki, T., Formankay, J.D & Kay, L.E (1994) Simultaneous acquisition of N-15-edited and C-13-edited NOE spectra of proteins dissolved in H 2 O.
J Magn Reson Ser B 103, 197–201.
32 Kuboniwa, H., Grzesiek, S., Delaglio, F & Bax, A (1994) Mea-surement of H-N-H-Alpha J-couplings in calcium-free calmodulin using new 2D and 3D water-flip-back methods J Biomol NMR
4, 871–878.
33 Grzesiek, S., Kuboniwa, H., Hinck, A.P & Bax, A (1995) Mul-tiple-quantum line narrowing for measurement of H-Alpha-H-Beta J-couplings in isotopically enriched proteins J Am Chem Soc 117, 5312–5315.
34 Vuister, G.W & Bax, A (1993) Quantitative J Correlation: a new approach for measuring homonuclear 3-bond J (H (N) H (Alpha) coupling-constants in N-15-enriched proteins J Am Chem Soc.
115, 7772–7777.
35 Wishart, D.S., Bigam, C.G., Yao, J., Abildgaard, F., Dyson, H.J., Oldfield, E., Markley, J.L & Sykes, B.D (1995) 1 H, 13 C and 15 N chemical shift referencing in biomolecular NMR J Biomol NMR
6, 135–140.
36 Delaglio, F., Grzesiek, S., Vuister, G.W., Zhu, G., Pfeifer, J & Bax, A (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes J Biomol NMR 6, 277–293.
37 Johnson, B.A & Blevins, R.A (1994) NMR view: a computer-program for the visualization and analysis of NMR data J Bio-mol NMR 4, 603–614.
38 Gagne, S.M., Tsuda, S., Li, M.X., Chandra, M., Smillie, L.B & Sykes, B.D (1994) Quantification of the calcium-induced sec-ondary-structural changes in the regulatory domain of
troponin-C Protein Sci 3, 1961–1974.
39 Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., Read, R.J., Rice, L.M., Simonson, T & Warren, G.L (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination Acta Crystallogr D Biol Crystallogr 54, 905–921.
40 Laskowski, R.A., Macarthur, M.W., Moss, D.S & Thornton, J.M (1993) Procheck: a program to check the stereochemical quality of protein structures J Appl Crystallogr 26, 283–291.
41 Sharma, D & Rajarathnam, K (2000) 13 C NMR chemical shifts can predict disulfide bond formation J Biomol NMR 18, 165–171.
42 Fimland, G., Blingsmo, O.R., Sletten, K., Jung, G., Nes, I.F & Nissen-Meyer, J (1996) New biologically active hybrid
Trang 9bacteriocins constructed by combining regions from various
pediocin-like bacteriocins: the C-terminal region is important
for determining specificity Appl Environ Microbiol 62, 3313–
3318.
43 Fimland, G., Jack, R., Jung, G., Nes, I.F & Nissen-Meyer, J.
(1998) The bactericidal activity of pediocin PA-1 is specifically
inhibited by a 15-mer fragment that spans the bacteriocin from
the center toward the C terminus Appl Environ Microbiol 64,
5057–5060.
44 Dalet, K., Cenatiempo, Y., Cossart, P & Hechard, Y (2001) A
sigma (54)-dependent PTS permease of the mannose family is
responsible for sensitivity of Listeria monocytogenes to
mesenter-icin Y105 Microbiology 147, 3263–3269.
Supplementary material
The following material is available from http:// blackwellpublishing.com/products/journals/suppmat/EJB/ EJB4085/EJB4085sm.htm
Fig S1 Helical wheel representation of leader peptide residues –4 to –15 for selected type IIa bacteriocins Fig S2 CD spectra of preCbnB2(1–64) (0–60% aqueous trifluoroethanol)
Table S1 Nitrogen and carbon chemical shifts
Table S2 Proton chemical shifts
Table S3 Structure statistics