Báo cáo khoa học: MR solution structure of the precursor for carnobacteriocin B2, an antimicrobial peptide fromCarnobacterium piscicola Implications of the a-helical leader section for export and inhibition of type IIa bacteriocin activity pdf

NMR solution structure of the precursor for carnobacteriocin B2,
an antimicrobial peptide from
Carnobacterium piscicola
Implications of the a-helical leader section for export and inhibition of type IIa
bacteriocin activity
Tara Sprules
1
, Karen E. Kawulka
1
, Alan C. Gibbs
2
, David S. Wishart
2
and John C. Vederas
1
1
Department of Chemistry and
2
Faculty 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 with
15
Nand
13
CforNMR
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:
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)
Eur. J. Biochem. 271, 1748–1756 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04085.x
these 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
piscicola LV17B [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 50 °C,
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
MNSVKELNVKEMKQLHGG VNYGNGVSCSKTKCSVNWGQAFQERYTAGINSFVSGVASGAGS
IGRRP
[19]
Sakacin G
MKNAKSLTIQEMKSITGG KYYGNGVSCNSHGCSVNWGQAWTCGVNHLANGGHGVC [20]
Plantaricin 423
MMKKIEKLTEKEMANIIGG KYYGNGVTCGKHSCSVNWGQAFSCSVSHLANFGHGKC [21]
Piscicolin 126
MKTVKELSVKEMQLTTGG KYYGNGVSCNKNGCTVDWSKAIGIIGNNAAANLTTGGAAGWNKG [22]
CbnBM1 MKSVKELNKKEMQQINGG AISYGNGVYCNKEKCWVNKAENKQAITGIVIGGWASSLAGMGH [19]
Leucocin A
MMNMKPTESYEQLDNSALEQVVGG KYYGNGVHCTKSGCSVNWGEAFSAGVHRLANGGNGFW [7]
Pediocin
MKKIEKLTEKEMANIIGG KYYGNGVTCGKHSCSVDWGKATTCIINNGAMAWATGGHQGNHKC [23,24]
Mesentericin

Y105
MTNMKSVEAYQQLDNQNLKKVVGG KYYGNGVHCTKSGCSVNWGEAASAGIHRLANGGNGFW [25]
Sakacin A
MNNVKELSMTELQTITGG ARSYGNGVYCNNKKCWVNRGEATQSIIGGMISGWASGLAGM [26]
Divercin V41
MKNLKEGSYTAVNTDELKSINGG TKYYGNGVYCNSKKCWVDWGQASGCIGQTVVGGWLGGAIPGKC [27]
Sakacin P
MEKFIELSLKEVTAITGG KYYGNGVHCGKHSCTVDWGTAIGNIGNNAAANWATGGNAGWNK [28]
Ó FEBS 2004 Solution structure of precarnobacteriocin B2 (Eur. J. Biochem. 271) 1749
stirring the reaction mixture with elemental iodine (0.1
M
in
methanol) under argon for 2 h. The oxidation was
quenched by addition of aqueous ascorbic acid (1
M
), 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 37 °C
in M9 minimal medium as described previously [14].
For labeling experiments, [
15
NH
4

]
2
SO
4
, or alternatively
[
15
NH
4
]
2
SO
4
and
D
-[U-
13
C]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 m
M
isopropyl thio-b-
D
-galactoside when A
600
of the cell culture
reached 0.5. The culture was incubated for a further 3 h at
37 °C, and the cells were harvested by centrifugation. The
cell pellet was resuspended in column buffer (20 m

M
Tris/
HCl, 200 m
M
NaCl, 1 m
M
EDTA, 1 m
M
NaN
3
and 1 m
M
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,
andelutedwith10m
M
maltose, 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 m
M
Tris/HCl/100 m
M
NaCl/1 m
M
CaCl
2
[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
15
N-labeled preCbnB2(1–64) derived from [

15
NH
4
]
2
SO
4
showed a predominant mass peak at 6825.8 ± 0.8, corres-
ponding to complete
15
N-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-d
3
(99.9%)/H
2
O or 70 : 30 trifluoroetha-
nol-d
3
(99.9%)/D
2
O (100%) at a concentration of  1m
M
.
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
(k ¼ 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 J
700 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 25 °C.
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 (degreesÆcm
2
Ædmol
)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 30 °C. Serial
twofold dilutions of preCbnB2 (in Luria–Bertani broth)
werespotted(20mL)ontoaBHIhardagarplate,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 37 °C.
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.
15
N
HSQC [29],
15

N HSQC-TOCSY [30],
15
NHSQC-NOESY
[30],
13
CHSQC,
13
C HSQC-NOESY [31] and a 2D
NOESY spectrum were recorded at 35 °C.
15
NHSQC,
15
N HSQC-NOESY, HNHA [32–34],
13
CHSQCand
13
C HSQC-NOESY spectra were recorded at 20 °C.
15
N
HSQC-NOESY mixing times were 200 ms; the
13
CHSQC-
NOESY experiments were recorded with 150 ms mixing
1750 T. Sprules et al.(Eur. J. Biochem. 271) Ó FEBS 2004
times. 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 from
15
Nand
13
C 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 d
Na
/d
aN
ratios
[38]. The w angle restraint was set to )30° ±110° for d
Na
/
d

aN
ratios less than 1, and to 120° ±100° for d
Na
/d
aN
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°)withnoresiduesinthedisallowedregionofthe
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
15
Nand
13
Cof
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 (
15
NH

4
)
2
SO
4
, or alternatively (
15
NH
4
)
2
SO
4
and
D
-[U-
13
C]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 combination 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
15
N HSQC-TOCSY,
15
N HSQC-NOESY and
13
CHSQC
spectra. Sequential assignments were made by following the
pattern of d
N,N±1
NOEs. Analysis of
3
J
HNHa
coupling
constants, Ha,Ca and Cb chemical shifts, and NOEs
Ó FEBS 2004 Solution structure of precarnobacteriocin B2 (Eur. J. Biochem. 271) 1751
indicated 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
35 °CHa-Ha and Ha-Hb NOEs are seen across the
disulfide bridge, although at lower temperature (20 °C) 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
13
C 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.
15
N HSQC of preCbnB2. Spectra
recorded in 70 : 30 trifluoroethanol/H
2
Oat
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.
1752 T. Sprules et al.(Eur. J. Biochem. 271) Ó FEBS 2004
account 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 Ômature portionÕ 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 EII
Man
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 ÔjackknifeÕ
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.
Ó FEBS 2004 Solution structure of precarnobacteriocin B2 (Eur. J. Biochem. 271) 1753
long-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.
1
H-
1
Hstripsfroma3D
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.
1754 T. Sprules et al.(Eur. J. Biochem. 271) Ó FEBS 2004
10. 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)
Backbone
1
Hand
15
N 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
Cand
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
Ó FEBS 2004 Solution structure of precarnobacteriocin B2 (Eur. J. Biochem. 271) 1755
bacteriocins 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.
1756 T. Sprules et al.(Eur. J. Biochem. 271) Ó FEBS 2004