Lucifensin, a Novel Insect Defensin of Medicinal Maggots: Synthesis and Structural Study

Article abstract of DOI:10.1002/cbic.201100066

Recently, we identified a new insect defensin, named lucifensin that is secreted/excreted by the blowfly Lucilia sericata larvae into a wound as a disinfectant during the medicinal process known as maggot therapy. Here, we report the total chemical synthesis of this peptide of 40 amino acid residues and three intramolecular disulfide bridges by using three different protocols. Oxidative folding of linear peptide yielded a peptide with a pattern of disulfide bridges identical to that of native lucifensin. The synthetic lucifensin was active against Gram-positive bacteria and was not hemolytic. We synthesized three lucifensin analogues that are cyclized through one native disulfide bridge in different positions and having the remaining four cysteines substituted by alanine. Only the analogue cyclized through a Cys16-Cys36 disulfide bridge showed weak antimicrobial activity. Truncating lucifensin at the N-terminal by ten amino acid residues resulted in a drop in antimicrobial activity. Linear lucifensin having all six cysteine residues alkylated was inactive. Circular dichroism spectra measured in the presence of α-helix-promoting compounds showed different patterns for lucifensin and its analogues. Transmission electron microscopy revealed that Bacillus subtilis treatment with lucifensin induced significant changes in its envelope. Lucifensin is the key antimicrobial peptide of the green bottle fly larvae Lucilia sericata. This defensin protects the larvae when they are exposed to the infectious environment of a wound during maggot therapy and it also contributes as a disinfectant and healing factor. Three disulfide bridges keep this 40-amino-acid peptide in the specific conformation required for its antimicrobial activity.

Full text of DOI:10.1002/cbic.201100066

DOI: 10.1002/cbic.201100066
Lucifensin, a Novel Insect Defensin of Medicinal Maggots:
Synthesis and Structural Study
ˇ
ˇ
ˇ
ˇ
ˇ
´
Vꢀclav Cerovsky,* Jirina Slaninovꢀ, Vladimꢁr Fucꢁk, Lenka Monincovꢀ, Lucie Bednꢀrovꢀ,
[a]
ˇ
Petr Malon, and Jitka Stokrovꢀ
Recently, we identified a new insect defensin, named lucifensin
that is secreted/excreted by the blowfly Lucilia sericata larvae
into a wound as a disinfectant during the medicinal process
known as maggot therapy. Here, we report the total chemical
synthesis of this peptide of 40 amino acid residues and three
intramolecular disulfide bridges by using three different proto-
cols. Oxidative folding of linear peptide yielded a peptide with
a pattern of disulfide bridges identical to that of native lucifen-
sin. The synthetic lucifensin was active against Gram-positive
bacteria and was not hemolytic. We synthesized three lucifen-
sin analogues that are cyclized through one native disulfide
bridge in different positions and having the remaining four
cysteines substituted by alanine. Only the analogue cyclized
through a Cys16–Cys36 disulfide bridge showed weak antimi-
crobial activity. Truncating lucifensin at the N-terminal by ten
amino acid residues resulted in a drop in antimicrobial activity.
Linear lucifensin having all six cysteine residues alkylated was
inactive. Circular dichroism spectra measured in the presence
of a-helix-promoting compounds showed different patterns
for lucifensin and its analogues. Transmission electron micros-
copy revealed that Bacillus subtilis treatment with lucifensin in-
duced significant changes in its envelope.
Introduction
The beneficial use of fly larvae in healing chronic, infected
wounds has been known since ancient times,[1] but reintroduc-
tion of standard, routine maggot debridement therapy in clini-
cal practice at hundreds of hospitals around the world dates
back only to the 1990s and the systematic work of Sherman.[1,2]
The action of wound-healing larvae results in removal of ne-
crotic tissue (debridement), elimination of infecting microor-
ganisms, disinfection of the wound, and stimulation of wound
granulation and repair.[2,3] The application of green bottle fly
larvae (Lucilia sericata) has become particularly important in
the treatment of non-healing wounds infected with such multi-
drug-resistant pathogens as methicillin-resistant Staphylococcus
aureus (MRSA).[4,5] Prompted by successful clinical experience,
many researchers have focused on the fundamental healing
principles of maggot therapy. A main focus of interest has
been to investigate antimicrobial activity of the components of
larval secretions and fecal waste products.[5–9] Early research
has shown that larval excretions/secretions (ES) of Lucilia seri-
cata contain a variety of alkaline components that inhibit bac-
terial growth and that the pH increase provides optimal condi-
tions for the activity of larvae-secreted proteolytic enzymes
that liquidize necrotic tissues.[2,3] It also has been proposed
that larvae release antimicrobial substances into the wound as
a response to infection. Some of these are bacteriostatic low-
molecular-weight compounds that were already characterized
some time ago.[2,3,10] In the past decade, several researchers
have suggested that antimicrobial peptides originating from
the larval immune system might be released into the wound
and thus could contribute to wound healing.[5,7,11] These pep-
tides belong to the large group of insect defensins.[12] Defen-
sins of dipteran species are 4 kDa cationic peptides containing
three disulfide bridges. They differ from one another by a few
amino acid residues and are preferentially active against Gram-
positive bacteria.[13–15] Surprisingly, the structural characteriza-
tion of Lucilia defensin had not been reported until 2010,
when we purified lucifensin from an extract of larval guts, de-
termined its structure by mass spectrometry and Edman degra-
dation (Figure 1), and then identified it in other larval tissues
as well as in the larval ES.[16] Soon after, the primary sequence
Figure 1. Primary structure of lucifensin. The basic amino acid residues (K, R)
that play an important role in its antimicrobial activity are shown in blue;
cysteines forming disulfide bridges are shown in yellow.
ˇ
ˇ
´
ˇ
[a] Dr. V. Cerovsky, Dr. J. Slaninovꢀ, Dr. V. Fucꢁk, L. Monincovꢀ,
ˇ
ˇ
Dr. L. Bednꢀrovꢀ, Dr. P. Malon, Dr. J. Stokrovꢀ
Institute of Organic Chemistry and Biochemistry, Academy of Sciences of
the Czech Republic
Flemingovo 2, 166 10 Prague 6 (Czech Republic)
Fax: (+420)2-2018-3578
E-mail: cerovsky@uochb.cas.cz
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201100066.
1352
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1352 – 1361

Study of Lucifensin
of lucifensin was also published by Andersen and co-work-
ers.[17] They had used molecular biology methods to confirm its
sequence. We assume lucifensin to be the key antimicrobial
component that protects the maggots when they are exposed
to the highly infectious environment of a wound during
maggot therapy, and that it contributes as a disinfectant and
healing factor. To accelerate progress in lucifensin research, it
will be necessary to produce sufficient quantities of lucifensin
through chemical synthesis or recombinant methods. In this
study, we report total chemical synthesis of lucifensin by using
three different protocols employing 9-fluorenylmethyloxycar-
bonyl (Fmoc) solid-phase peptide synthesis (SPPS).
Several reports show that within the family of mammalian
defensins, the order of disulfide connectivity or presence of
three disulfide bridges is not necessary for the antimicrobial
activity.[18,19] Mammalian defensins have been reported to
retain antimicrobial activity even without disulfide bridges.[20–22]
Inspired by those reports, we studied the importance of disul-
fide bridges and the importance of the N-terminal part of the
lucifensin sequence for antimicrobial activity.
Results
Synthesis of linear lucifensin peptides
In the first approach, the linear precursor of lucifensin was pre-
pared manually by stepwise Fmoc SPPS chemistry by using the
N,N’-diisopropylcarbodiimide
(DIPC)/1-hydroxybenzotriazole
(HOBt) method in N,N’-dimethylformamide (DMF). The conver-
sion of free amino groups during the coupling was monitored
by bromphenol blue indicator. Because the coupling reaction
slowed during growth of the peptide chain, the coupling times
had to be extended from several minutes at the beginning of
peptide assembly to several hours, thereby reaching comple-
tion of the 40-residue peptide within three weeks. This proce-
dure resulted, after deprotection and cleavage from the resin,
in a crude peptide showing an analytical HPLC profile dominat-
ed by a discrete peak of the required 40-residue peptide
within a bunch of closely related impurities (Figure 2A). The
determined monoisotopic molecular mass of the linear pep-
tide, 4120.1 Da, was in good agreement with its calculated
value of 4119.93 Da (Figure 3A). The N-terminally shortened
linear precursor of Luc[des1–10, Ala30] analogue (Figure 4D)
was prepared similarly (for HPLC and MS spectrum see Figures
S7 and S8 in the Supporting Information). Secondly, to avoid
the difficulties associated with sluggish couplings and the
large amounts of impurities from Fmoc deprotections, we im-
plemented fragment condensation of the protected N-terminal
octapeptide (ATCDLLSG) fragment onto a 32-residue resin-
bound peptide. The analytical HPLC profile of the deprotected
condensation product still showed the presence of the un-
reacted 32-peptide (Figure 2B) in addition to the desired pep-
tide. Moreover the presence of deletion peptides indicated
their formation even before cycle 32 (Figure 2B). Thirdly, the
linear lucifensin as well as the linear precursors of analogues
with one disulfide bridge (Figure 4A–C) were prepared by
using an automated peptide synthesizer and working with a
Figure 2. RP-HPLC profiles of the crude linear lucifensin peptide prepared by
different SPPS synthetic strategies at 220 nm. A) Manual stepwise synthesis;
B) 8+32 fragment condensation on the resin; C) Automated peptide syn-
thesis by using Applied Biosystems peptide synthesizer. The dominant peak
represents the required peptide of 40 amino acid residues. The peak labeled
with an asterisk in panel B represents the unreacted 9–40 peptide fragment.
An elution gradient ranging from 80% of solvent A to 50% of solvent B was
applied for 60 min at a flow rate 1 mLminÀ1. Solvent A: 5% MeCN/H2O/0.1%
TFA, solvent B: 70% MeCN/H2O/0.1% TFA.
standard protocol based on the 2-(1H-benzotriazol-1-yl)-1,1,3,3-
tetramethyluroniumhexafluorphosphate (HBTU)/HOBt/N,N-di-
isopropylethylamine (DIPEA) activation and N-methyl-2-pyrroli-
done (NMP) as a solvent. Although, the assembly of the 40-res-
idue lucifensin peptide was accomplished within a substantial-
ly shorter time (one week), the purity of the final crude pep-
tide (Figure 2C) was not as good as that of the manually syn-
thesized peptide. See also Figures S3—S6 for HPLC and MS
spectra of analogues containing one disulfide bridge in Fig-
ure 4A–C.
ChemBioChem 2011, 12, 1352 – 1361
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1353

ˇ
ˇ
´
V. Cerovsky et al.
Oxidative folding
The time course of lucifensin folding as followed by HPLC is
shown in Figure 5. As clearly demonstrated, the conversion of
the linear peptide (peak 2) into lucifensin (peak 1) proceeded
via an intermediate that eluted at a longer retention time
(peak 3). ESI-MS analysis of lucifensin (Figure 3B) as well as the
intermediate (not shown) revealed a loss of six mass units
compared to the linear peptide (Figure 3A), thus suggesting
the formation of three disulfide bridges; in the case of inter-
mediate (peak 3) with incorrect pairing. This intermediate
showed activity against Micrococcus luteus in a drop–diffusion
test almost equal to that of lucifensin (not shown). After lyo-
philization of the mixture, the folded lucifensin was purified by
preparative HPLC. The final product was analyzed by HPLC and
ESI-MS (Table 1, and Figures S1 and S2). The determined molec-
ular mass of 4113.5 Da was in good agreement with the calcu-
lated mass of 4113.89 Da (Figure 3B). As expected, in the case
of the analogues with one disulfide bridge, (for their structures
see Figure 4A–C) we did not observe formation of any inter-
mediate during the oxidation of linear peptides into cyclic
peptide (MS data of these analogues in Figures S4–S6). The
folding of the Luc[des1–10, Ala30] analogue (structure Fig-
ure 4D) proceeded similarly to that of lucifensin; that is, via an
incorrectly folded intermediate (for MS spectrum of final prod-
uct see Figure S9).
Determination of disulfide bridges
To determine the connectivity of disulfide bridges, the lucifen-
sin was digested with thermolysin, as described elsewhere.[23]
Peptide fragments obtained by the digestion were separated
by HPLC and then analyzed by ESI-MS. Among numerous frag-
ments, we clearly identified three peptides containing cysteine
(single disulfide bond; Figure 6). Their sequences as well as the
sequences of other identified fragments clearly fit to the se-
quence of lucifensin, which has the expected pattern of disul-
fide bridges consistent with the conserved pattern of disulfide
bridges found in other insect defensins (i.e., Cys1–Cys4, Cys2–
Cys5, and Cys3–Cys6).
Figure 3. ESI-QTOF mass spectra of A) linear lucifensin and B) lucifensin.
Alkylation of linear lucifensin peptide
The alkylation of linear lucifensin peptide was performed on a
sub-milligram scale in order to obtain the material for the anti-
microbial microassay illustrated in Figure 7. The reaction of
linear peptide with iodoacetamide resulted in complete alkyla-
tion of all six SH groups over one hour (Figures S10 and S11).
Antibacterial activity
The antimicrobial activities of synthetic lucifensin and its ana-
logues were tested preferentially against Gram-positive bacte-
ria, Micrococcus luteus (M.l.), Bacillus subtilis (B.s.), and Staphylo-
coccus aureus (S.a.) and one Gram-negative one Escherichia coli
(E.c.). Lucifensin was highly active against M. luteus and B. sub-
tilis, whereas lower but significant activity was observed
Figure 4. Primary structures of lucifensin analogues. A) Luc[C3–C30];
B) Luc[C16–C36]; C) Luc[C20–C38]; D) Luc[des1-10, Ala30]. Basic amino acid
residues (K, R) playing an important role in peptides’ antimicrobial activity
are shown in blue, cysteines forming disulfide bridges are shown in yellow
and their replacements by alanines are shown in red.
1354
www.chembiochem.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1352 – 1361

Study of Lucifensin
Petri dish (Figure 7A). These re-
sults indicate that the presence
of disulfide bridges in lucifensin
is essential for its antimicrobial
activity because it is necessary
for preserving its three-dimen-
sional structure. Similar require-
ments have been reported for
other insect defensins.[23,24] This
is not valid for the defensins be-
longing to the mammalian
family, however, in which the
presence of disulfide bonds is
not a prerequisite for antimicro-
bial activity.[20,21,22] In this con-
text, we tested the antimicrobial
activity of linear nonfolded luci-
fensin and linear lucifensin
having the SH group of all Cys
residues blocked by alkylation.
The results of this test (Fig-
ure 7B, C) clearly show that the
antimicrobial activity of linear
lucifensin (free ÀSH groups) was
almost comparable to that of
lucifensin, whereas the alkylated
species was completely inactive
when tested on two bacteria
strains. These results indicate
Figure 5. Time course for oxidative folding of lucifensin monitored by RP-HPLC at 220 nm. A) 5 min; B) 1 h; C) 2 h;
D) 4 h. Peak 1, lucifensin; peak 2, linear peptide; peak 3, incorrectly folded lucifensin as an intermediate (active
against M. luteus). An elution gradient ranging from 80% of solvent A to 50% of solvent B was applied for 60 min
at flow rate 1 mLminÀ1. Solvent A: 5% MeCN/H2O/0.1% TFA, solvent B: 70% MeCN/H2O/0.1% TFA.
that linear lucifensin was oxidatively folded while carrying out
the assay. HPLC analysis of linear lucifensin incubated in LB
broth gave multiple peaks, one of which corresponded to luci-
fensin (not shown).
Table 1. Antimicrobial activities of lucifensin and its analogues.
Peptide
Antimicrobial activity MIC [mm]
B. subtilis
M. luteus
S. aureus
C. albicans
Lucifensin
Luc[C3–C30]
Luc[C16–C36]
Luc[C20–C38]
Luc[des1-10, Ala30]
1.2
>100
>100
>100
>100
0.6
>100
23
>100
10
41
~100
n.d.
n.d.
n.d.
n.d.
The Luc[des1–10, Ala30] analogue showed only negligible
activity against M. luteus in the drop–diffusion assay but dis-
played noticeable activity when M. luteus growth was observed
in a microdilution assay (Table 1).
n.d.[a]
n.d.
n.d.
n.d.
[a] Not determined.
against S. aureus (Table 1). No activity was detected against
E. coli, thus confirming the generally recognized fact that
insect defensins are more active against Gram-positive than
Gram-negative bacteria.[13,15,24,25] The peptide showed slight
antifungal activity against Candida albicans (C.a). Lucifensin,
having 87–90% sequence identity with other dipteran defen-
sins, such as Phormia terranovae defensins A and B[14] and Sar-
cophaga peregrina sapecin,[13] shows a similar profile in antimi-
crobial activity (pathogenic S. aureus is not the most sensitive
bacterium among those tested). This was also reported by
Andersen et al. in his recent study on lucifensin.[17] As shown in
Table 1, the lucifensin analogues that folded through one disul-
fide bridge were practically inactive even against the most sen-
sitive bacteria, M. luteus and B. subtilis. The only exception is
the Luc[C16–C36] analogue, which still shows noticeable activi-
ty as clearly demonstrated by the drop–diffusion assay on a
Figure 6. Lucifensin and its peptide fragments identified after thermolysin
digestion by MS. The numbers denote calculated molecular masses of pep-
tides that were identical to those obtained experimentally. The existence of
three cystine-containing peptide fragments (bold letters) shown in the
Figure confirmed the correctness of disulfide pairing.
ChemBioChem 2011, 12, 1352 – 1361
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1355

ˇ
ˇ
´
V. Cerovsky et al.
(1688, 1678, 1641, and 1633 cmÀ1) in addition to the a-helix
and random structure (1658 cmÀ1) was clearly identified in the
amide I spectral region (not shown). The analogues of lucifen-
sin containing one disulfide bridge showed a lower a-helical
fraction (aH ꢀ10%) under the same conditions and a higher b-
structure fraction (ab ꢀ65%) in their spectra. The differences
seem to be a consequence of secondary structure stabilization
caused by the presence of disulfide bridges in the primary
structure of a natural peptide. Surprisingly, in the presence of
TFE, lucifensin exhibited almost identical CD spectra as did the
analogues containing one disulfide bridge (Figure 8B), with
the helix fraction aH at about 25%. This could mean that all
these peptides possess very similar ability to form an a-helical
conformation.
On the other hand, in a membrane-mimicking environment
with different concentrations of SDS the studied peptides
showed distinctly different behavior (Figure 8C–F). Although
the peptides were partially a-helical in water, the addition of
SDS at a concentration lower than the critical micelle concen-
tration (cmc) changed the conformation of the peptides in
favor of the b-structure. Further continuous addition of SDS re-
sulted again in a partial formation of a-helical structure. For
the lucifensin itself (Figure 8C), the a-helical fraction increases
to its maximum (aH ꢀ30%) already at around the cmc of 2–
4 mm. With further increase in the SDS concentration above
cmc (16 mm), the a-helical content decreases again, probably
in favor of a structure combining polyproline II and b-sheet
conformations. Knowing that by solely using CD spectra one
cannot reliably demonstrate b-structure and distinguish be-
tween unordered and polyproline conformations, we obtained
additional information by using infrared spectroscopy. IR spec-
tra of lucifensin with different SDS concentrations below and
above cmc were measured. When increasing the concentration
of SDS, a similar trend was observed: an increase of a-helical
structure was accompanied by a decrease of b-structure and b-
turn conformations. The observed spectral changes confirm
the formation of b-structure at concentrations of SDS below
the cmc and the presence of polyproline structure for SDS con-
centrations above cmc (16 mm; not shown). The cause of poly-
proline structure formation was not investigated in detail. The
analogues containing one disulfide bridge (Figure 8D–F) be-
haved similarly, displaying the increase of a-helical content
around the cmc. In these cases, however, the relatively high
a-helical content also remained above the cmc (16 mm). This
indicates that in the absence of constraints imposed by three
disulfide bridges, the analogues with one disulfide bridge are
more flexible in adopting the a-helical conformation dictated
by the environment.
Figure 7. Antimicrobial activities of lucifensin, Luc[C16-C36], linear lucifensin
and linear alkylated lucifensin in the drop–diffusion test. A) Micrococcus
luteus, lane 1: lucifensin in twofold serial dilution starting at concentration
1 mgmLÀ1; lane 2: Luc[C16-C36] at the same dilution but starting at the con-
centration 10 mgmLÀ1. Luc[C3-C30] and Luc[C20-C38] were inactive in this
test. B) Micrococcus luteus, lane 1: linear lucifensin with all six cysteines alky-
lated, lane 2: linear lucifensin, lane 3: lucifensin. The concentrations of the
peptides were equal; in the right column ten times diluted. C) Bacillus subti-
lis, the same order as in (B).
CD analyses and structural features
UV CD spectra of lucifensin and of its analogues containing
one disulfide bridge, Luc[C3–C30], Luc[C16–C36], and
Luc[C20–C38], were acquired 1) in water, 2) in the presence of
the helix-promoting additive 2,2,2-trifluoroethanol (TFE), and
3) in water containing increasing quantities of SDS as a very
simple system intended to mimic a membrane environment.
At first glance, the CD spectrum of lucifensin in water is some-
what different from those of its analogues containing only one
disulfide bridge (Luc[C3–C30], Luc[C16–C36], and Luc[C20–
C38]) (Figure 8A). A qualitative estimate taking into account
just the general spectral shape would tempt one to interpret
this difference as a slightly different content of random and a-
helical conformations.
This approximate picture seems to be supported by the ob-
servation that upon addition of helix-supporting additives like
TFE (Figure 8B) or SDS (Figure 8C–F) the conformation differ-
ences disappear, and all peptides adopt a conformation more
resembling an a-helix. However, the absolute intensity of these
spectra remains generally less than 15000 degcm2 molÀ1 per
unit of residue and thus contradicts such an interpretation. On
the contrary, it gives evidence that even in these situations the
a-helical content is not very high. We attempted to clarify this
picture in at least a semiquantitative way by using a computa-
tional analysis of the spectra by following a numerical proce-
dure (Dichroweb[26]). In accordance with our assumptions, the
result for lucifensin showed a-helical content of about 15%
and random conformation of about 30%, whereas the family
of b-structures (parallel, antiparallel and turn; ab ꢀ60%) was
demonstrated to be the major structural component. Results
of the numerical analysis were further supported by the IR
spectra of lucifensin, in which the presence of b-structure
Transmission electron microscopy (TEM)
To analyze the mechanism of lucifensin’s action, we used TEM
to study the structural changes on bacterial membrane after
peptide treatment. Bacillus subtilis was used as a model bacte-
rium and treated with lucifensin for 10 or 60 min. Bacterial
cells were visualized by negative staining. Untreated bacteria
revealed a normal shape with a smooth, electron-dense surface
1356
www.chembiochem.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1352 – 1361

Study of Lucifensin
tive complex circular structures
(not shown), sometimes bound-
ed by flagella, were often ob-
served to be released from bac-
teria.
Discussion
In 2010, we published the pri-
mary structure of lucifensin, the
key antimicrobial factor of the
Lucilia sericata immune sys-
tem.[16] Later, the production of
recombinant peptide resulting
in a few milligrams of active
peptide was reported by Ander-
sen.[17] Although, numerous syn-
thetic studies in the expanding
field of defensins have been
published during the last two
decades, only a few of these
papers report the synthesis of
insect defensins.[24,27,28] Despite
the generally accepted assump-
tion that the synthesis and pu-
rification of a 40-residue pep-
tide can be achieved routinely,
the synthesis of a peptide with
multiple disulfide bridges is not
so straightforward. Challenged
by our discovery of lucifensin
(Figure 1) and seeing the need
to fill a gap as to its structural
study, we focused on the total
chemical synthesis of lucifensin
as well as on the synthesis of its
three analogues containing only
one native disulfide bond (Fig-
Figure 8. UV CD spectra of lucifensin and its one disulfide bridge containing analogues in A) H2O, B) in 50% TFE,
and in the presence of various concentrations of SDS (C–F). C) lucifensin; D) Luc[C3–C30]; E) Luc[C16–C36];
F) Luc[C20–C38].
and well-preserved membrane structure. The bacterial surface
was well-connected with flagella (Figure 9A). After treatment
with lucifensin, significant structural changes in the bacterial
membrane were observed. After 10 min of treatment (Fig-
ure 9B), most of the bacterial cells lost their electron-dense
character and the bacterial envelopes, the integrity of which
was still preserved, revealed many small pores. The bacterial
surface was still connected with numerous flagella. Also, slight
leakage of cytoplasmic components was observed. Longer in-
cubation with the peptide (60 min) led to further damage to
the bacteria (Figure 9C, D). Even when some bacterial enve-
lopes retained a certain degree of integrity, many more pores,
but only rarely flagella, were found on the envelope surface.
Figure 9. Electron micrographs of negatively stained Bacillus subtilis either
A) untreated, or B) treated by lucifensin for 10 min or 60 min (C, D). Scale
bars represent 1 mm.
Extensive ruptures in the cell membrane of many bacteria led
to their lysis and leakage of their cytoplasmic content. Distinc-
ChemBioChem 2011, 12, 1352 – 1361
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1357

ˇ
ˇ
´
V. Cerovsky et al.
ure 4A–C). We also synthesized an N-terminally shortened ana-
logue with two native disulfide bridges preserved (Figure 4D).
Generally, the presence of intramolecular disulfide bridges in
most biologically active peptides is essential for their confor-
mation and biological activity. However, the requirement of
the presence of disulfide bridges in cationic antimicrobial pep-
tides for their activity is not fully clarified, particularly within
the family of defensins. Some linear analogues of defensins
with cysteines substituted by other amino acids, omitted, or
side-chain modified have exhibited biological activities similar
to those of their native cyclic peptides. Some others, however,
have not. For example, several linear analogues of human b-
defensins 3 with cysteines substituted by various amino acid
residues retained potent antimicrobial activity and exhibited
decreased cytotoxicity to human epithelial cells compared to
wild-type defensin.[21] Interestingly, reduced synthetic human
b-defensin 1 (in the presence of dithiothreitol) was potent
against some Gram-positive bacteria and Candida albicans
whereas its native oxidized form was inactive.[22] Similarly, the
synthetic linear human a-defensin HNP1 with six cysteine resi-
dues omitted from its sequence also displayed significant anti-
microbial activity.[20] However, the linear analogue of HNP1
with all cysteines having side-chain residues protected by an
acetamidomethyl group was inactive.[29] On the other hand,
few reports show that, in the group of insect defensins which
have different connectivity of disulfide bridges (Cys1–Cys4,
Cys2–Cys5, Cys3–Cys6) than do mammalian defensins, the
presence of disulfide bridges is essential for their antimicrobial
activity. For example, the antimicrobial activity of sapecin was
reduced by one order of magnitude when its disulfide bonds
were cleaved.[23]
cysteine-stabilized ab (CSab) motif and which is important for
antimicrobial activity.[30,31]
The percentage of a-helical fraction in lucifensin and its ana-
logues that was calculated from CD spectra (25%) measured in
the presence of TFE is in good agreement with the secondary
structure of sapecin that was determined from NMR spectro-
scopic data, just as the presence of b-sheet conformations as-
sumed from lucifensin CD spectra when measured in water or
a membrane-mimicking environment.
Slight differences are noticeable in the CD spectra between
analogues in which the disulfide bond connects the a-helical
part with the 34–40 strand (Luc[C16–C36] and Luc[C20–C38];
Figure 8E, F) versus the Luc[C3–C30] analogue (Figure 8D) in
which the disulfide bond connects the terminal of the loop
with the 24–31 strand. The CD spectra of the Luc[C3–C30] ana-
logue underwent a slower change of helical formation upon
increasing the SDS concentration around the cmc than did
those of the other two analogues. This might indicate a sup-
portive role of disulfide bonds 16–36 and 20–38 in maintaining
the CSab motif, which also involve formation of the 15–23 hel-
ical segment.
Based on the three-dimensional model of sapecin, the N-ter-
minal flexible loop (sequence 4–14) is supposed to adopt sev-
eral orientations with respect to the ab motif.[30] In lucifensin,
the presence of Lys12 within the loop contributes to lucifen-
sin’s overall net cationic charge being +4 whereas the sapecin
with Asn in position 12 has a net charge of +3. The impor-
tance for antimicrobial activity of the N-terminal loop in the
molecule of insect defensins has been studied marginally.
Chemically synthesized tenecin 1, for example, having a delet-
ed N-terminal loop, did not exhibit antibacterial activity. It was
shown that this loop plays an important role in increasing the
molecule’s a-helical content in a membrane-mimicking envi-
ronment, which relates to the activity.[24] To follow the role of
the lucifensin molecule’s N-terminal loop in antibacterial activi-
ty, we synthesized an analogue of lucifensin (Luc[des1–10,
Ala30]) that was truncated at its N terminus by ten amino acid
residues. Because it is generally accepted that the presence of
positively charged amino acid residues in antimicrobial pep-
tides is a prerequisite for their activity, we retained the 11–12
part of the loop containing the cationic Lys12 residue in the
truncated analogue to preserve its net cationic charge at +4.
Briefly, the cationic antimicrobial peptides interact with the
anionic phospholipids of bacterial cell membrane, then inte-
grate into the lipid bilayer and disrupt the membrane structure
by forming transmembrane pores or ion channels that lead to
leakage of cytoplasmic components and ultimately cell death.
Although deletion of the N-terminal part of the loop did not
affect the net cationic charge of the analogue, its antimicrobial
activity decreased. On the basis of NMR spectroscopy experi-
ments, a putative mechanistic model for membrane permeabi-
lization caused by sapecin has been proposed.[32] According to
this model, sapecin oligomerizes in the bacterial membrane
and thus forms the channels that result in membrane permea-
bilization. This putative model of sapecin oligomerization is
based on electrostatic interaction between Asp4 of one sape-
cin molecule and Arg23 of another sapecin molecule because
We have proven that linear lucifensin with cysteine residues
blocked by alkylation was inactive in a drop–diffusion antimi-
crobial test. The main issue we wished to resolve by our study
was to elucidate which of those three disulfide bridges in the
lucifensin molecule plays the most important role in maintain-
ing its conformation, reflected in its antimicrobial activity. We
expected that the omission of the other two bridges would
certainly negatively affect the antimicrobial activity of lucifen-
sin. We supposed that the three-dimensional structure of luci-
fensin is similar to that of sapecin,[30] because the primary se-
quence of sapecin (40 residues) differs from the sequence of
lucifensin by only four amino acid residues (Ile11, Asn12, Lys33,
Val35). The conformation of sapecin was determined by
1H NMR spectroscopy and simulated annealing calculations,[30]
resulting in a structure that includes an N-terminal flexible
loop (residues 4–12), a helical part (residues 15–23), and an
antiparallel b-structure (residues 24–31 and 34–40). Significant
antimicrobial activity of the Luc[C16–C36] analogue against
the most sensitive bacterium M. luteus (Figure 7A) in compari-
son with the other two analogues containing one disulfide
bridge clearly indicates that the disulfide bond between Cys16
and Cys36, which connects the 15–23 helical part of the mole-
cule with the 34–40 strand is the most important one for main-
taining the ordered structure of lucifensin. The a-helix and b-
structure connected by two disulfide bridges form a common
structural element typical for insect defensins known as the
1358
www.chembiochem.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1352 – 1361

Study of Lucifensin
solvents, and HPLC-grade MeCN were of the highest purity avail-
able from commercial sources.
these two residues are situated at opposite ends of the oligo-
merization site. We suppose that the absence of the Asp4 resi-
due in truncated lucifensin analogue might be the key reason
that its antimicrobial activity decreased. Because the putative
model of sapecin oligomers provides innovative insights into
membrane permeabilization caused by insect defensins, we
also speculate that the rigidity of insect defensin molecules
represented by the CSab motif and supported by three disul-
fide bridges is the prerequisite for their oligomerization in
microbial membrane. The softening of the lucifensin rigidity by
omitting the disulfide bridges (the studied analogues) can
make the oligomerization impossible in microbial membrane
because the Asp4 and Arg23 residues cannot come into
proper contact with one another to provide the necessary
electrostatic interaction.
Instruments: The HPLC was carried out on an Agilent Technolo-
gies 1200 Series module with a Vydac C-18, 250ꢃ4.6 mm, 5 mm,
column (Grace Vydac, Hesperia, California) at a 1 mLminÀ1 flow
rate using various solvent gradients ranging from solvent A (5%
MeCN/H2O/0.1% TFA) to solvent B (70% MeCN/H2O/0.1% TFA)
over 60 min. The elution was monitored by absorption at 220, 254,
and 280 nm by using a diode-array detector. ChemStation software
was used to control the instrument and evaluate UV spectra. Mass
spectra (ESI-MS) of the peptides were acquired on a Micromass Q-
Tof micro mass spectrometer (Waters) equipped with an electro-
spray ion source or on LTQ Orbitrap XL hybrid mass spectrometer
(Thermo Fisher Scientific, Waltham, MA, USA) in the service facility
of the institute. Bioscreen C instrument (Oy Growth Curves AB Ltd,
Helsinki, Finland) was used for quantitative antimicrobial activity
determination.
Peptide synthesis of linear precursor of lucifensin
Conclusions
Method A, Manual stepwise synthesis: The peptide was synthesized
on the preloaded Fmoc-Asn(Trt)-Wang resin (100 mg, substitution
on the resin: 0.47 mmolgÀ1) in a 5 mL polypropylene syringe with
a Teflon filter on the bottom by using the protocol of Na-Fmoc
chemistry. After swelling the resin in DMF and the Fmoc
deprotection with 22% piperidine in DMF, the resin was acylated
with a mixture of Fmoc-Arg(Pbf)-OH/HOBt/DIPC (4:5:7) in DMF. To
provide a lower substitution on the resin, the reaction was
quenched after 7 min by washing the resin (substitution on the
resin 0.33 mmolgÀ1), and the remaining amino groups were acety-
lated with HOAc/HBTU/N-methylmorpholine in DMF. Protected
amino acids were coupled in fourfold excess in DMF as solvent
and DIPC (7 equiv)/HOBt (5 equiv) as coupling reagents while utiliz-
ing nondestructive monitoring of the conversion of the free amino
groups with bromphenol blue indicator. With the growing peptide
chain, the reaction times of the coupling (2 h) as well as of the
Fmoc deprotection (1ꢃ5 min+1ꢃ20 min) were gradually pro-
longed. Val, Ile, and Thr(But) were double coupled. After the Fmoc
deprotection of the 32-residue peptide, the peptide–resin was
equally divided into two syringes. The resin in one syringe was left
for completion of the synthesis via fragment condensation (Meth-
od B), and the peptide elongation by “stepwise” synthesis was
completed in the other syringe (Method A). The linear lucifensin
peptide was deprotected and cleaved from the resin by using a
mixture of TFA/1,2-ethanedithiol/H2O/thioanisol/triisopropylsilane
(2 mL; 90:2.5:2.5:3:2) for 3.5 h and precipitated with tert-butyl
methyl ether, yielding crude product (88 mg). The crude peptide
was repeatedly purified by preparative HPLC on a Vydac C-18
column (250ꢃ10 mm) by using a gradient ranging from 80% of
Lucifensin is a key antimicrobial peptide involved in the de-
fense system of Lucilia sericata larvae against infection. Its dis-
covery as a crucial disinfectant secreted/excreted by maggots
to the wound broadened the understanding of the healing
process of maggot therapy. Currently, in collaboration with
clinicians, we continue in the detailed study of lucifensin with
respect to wound healing. Nevertheless, the determination of
lucifensin’s three-dimensional structure by NMR spectroscopy,
which we plan, is a prerequisite to correlate lucifensin structure
and the antimicrobial activity and to understand its mode of
action definitively. To continue the study and explore its phar-
macological potential, the production of lucifensin in larger
quantities either by total chemical synthesis or a recombinant
method will be necessary. In this report, we have shown that
its production by total chemical synthesis by using SPPS meth-
odology is feasible in principle. The overall yield of the final
product might be enhanced by optimizing the procedure or
by using alternative synthetic approaches such as native chem-
ical ligation.[33] As a homologue to such other dipteran defen-
sins as sapecin and defensin A, lucifensin has a similar effect
on Gram-positive bacteria but it is inactive against Gram-nega-
tive bacteria. The presence of three disulfide bridges in the lu-
cifensin molecule is essential for maximal activity and points to
the importance of the Cys16–Cys36 bridge as a main supporter
of the CSab motif stabilization. Our results also indicate that
the order of the disulfide connectivity is not so important for
its activity. The antimicrobial activity of lucifensin was dramati-
cally affected by truncation of the N-terminal loop. Lucifensin
might find application as a new agent in the treatment of such
serious non-healing wounds as diabetic foot ulcers, pressure
sores, and varicose ulcers.
solvent A to 50% of solvent B over 60 min at
a flow rate
3.0 mLminÀ1 resulting in 8.5 mg of material available for oxidative
folding. The 30-residue linear peptide (Figure 4D), Luc[des1–10,
Ala30], was also synthesized as described above resulting in
152 mg of crude peptide, which was purified by HPLC (19 mg)
prior to oxidative folding. Measured molecular masses of the pure
linear peptides were in good agreement with their calculated
values as shown in Table S1.
Method B, Manual synthesis by 8+32 fragment condensation on the
resin: The N-terminal 1–8 side-chain-protected peptide fragment
was synthesized on 2-chlorotrityl chloride resin (200 mg) in the sy-
ringe by using the same protocol for the coupling steps as above.
Boc-Ala-OH was used as the N-terminal amino acid residue. The
protected peptide was cleaved from the resin by using a mixture
Experimental Section
Materials: Fmoc-protected l-amino acids and the Fmoc-l-Asn(Trt)-
Wang resin were purchased from IRIS Biotech GmbH (Marktredwitz,
Germany). LB (Luria–Bertani) broth, LB agar and thermolysin were
bought from Sigma–Aldrich. All other reagents, peptide synthesis
ChemBioChem 2011, 12, 1352 – 1361
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1359

ˇ
ˇ
´
V. Cerovsky et al.
of CH2Cl2/TFE/HOAc (4 mL; 7:2:1) at RT for 60 and then again for
15 min. The combined solution was concentrated under vacuum,
and the residue triturated with a mixture of tert-butyl methyl ether
and n-hexane yielding the protected peptide (165 mg). The purity
and identity of the peptide was checked by HPLC and MS after
deprotection of the sample (about 1 mg). ESI-MS: m/z calcd for
C31H54N8O13S: 778.35, found: 778.3. The mixture of Boc-1–8 side-
chain-protected peptide fragment (82 mg, 0.064 mmol) and HOBt
(11 mg, 0.08 mmol) in NMP (0.4 mL) was mixed with 32-residue
peptide–resin in the syringe and then the resin was soaked with
2m solution of DIPC in DMF (56 mL, 0.112 mmol) and agitated for
24 h. After washing the resin, the condensation reaction was re-
peated under the same conditions. The deprotection and cleavage
of the 40-residue peptide from the resin was done as described in
method A, yielding crude product (87 mg). This product was puri-
fied by HPLC prior to oxidative folding (13 mg).
by HPLC as a single symmetrical peak. Its identity was confirmed
by ESI-MS: m/z calcd for C178H296N66O57S6: 4462.06, found: 4462.0.
Determination of antimicrobial activity: A simple qualitative esti-
mate of antimicrobial properties was undertaken by using the
drop–diffusion test on Petri dishes by the double-layer tech-
nique.[34] The Petri dishes (90 mm in diameter) contained 20 mL of
LB agar (Sigma). Melted “soft” agar (2 mL), prepared from LB broth
(Sigma) and 0.5% agar (Difco) containing about 107 colony-forming
units of various bacteria (M. luteus, B. subtilis, S. aureus) was poured
over the surface. Fresh bacterial cultures were always prepared in
the LB broth and added when the melted soft agar cooled to
about 458C. Antimicrobial-activity-containing materials (evaporated
HPLC fractions) were diluted in H2O (10 mL) and dropped (2 mL)
onto the surface of the solidified upper layer. To compare the activ-
ity of lucifensin with its analogues (Figure 7) their amounts were
determined from the HPLC peak areas. The plates were incubated
at 378C. Clear zones of inhibition appeared within a few hours and
remained clear for days. The potency was semiquantitatively esti-
mated according to the diameter and clarity of the zones of inhibi-
tion. Quantitative minimum inhibitory concentrations (MICs) were
established by observing bacterial growth in multiwell plates.[35,36]
Mid-exponential phase bacteria were added to individual wells
containing solutions of the peptides at different concentrations in
LB broth (final volume 0.2 mL, final peptide concentration in the
range of 0.5 to 100 mm). The plates were incubated at 378C for
20 h while being continuously shaken in a Bioscreen C instrument
(Helsinki, Finland). The absorbance was measured at 540 nm every
15 min and each peptide was tested at least three times in dupli-
cate. Routinely, 5ꢃ103–10ꢃ103 CFU of bacteria were used per well.
Tetracycline in a concentration range of 0.5–50 mm was tested as a
standard.
Method C, Automated peptide synthesis: Crude linear 40-residue
peptides of lucifensin (Figure 1), Luc[C3–C30], Luc[C16–C36] and
Luc[C20–C38] (Figure 4A, B, and C, respectively), were synthesized
on 100 or 200 mg of preloaded Fmoc-Asn(Trt)-Wang Resin in an
Applied Biosystems 433A peptide synthesizer by using the HBTU/
HOBt/DIPEA activation protocol of Fmoc chemistry. The protected
amino acids were coupled in tenfold excess in NMP as solvent. The
resin substitution was reduced as described above prior to placing
it into the instrument. The deprotection and cleavage as described
in method A, yielded on average 160 mg of crude product per
100 mg of the resin. Measured molecular masses of the pure pep-
tides were in a good agreement with their calculated values
(Table S1).
Oxidative folding: The lyophilized linear peptides prepurified by
HPLC were dissolved (at concentration 1 mg/4 mL) in 0.1m
NH4OAc buffer, pH 7.8 (prepared by bubbling gaseous NH3 into
0.1m acetic acid) and stirred under open air at RT. The time course
of the disulfide bond formation was monitored by HPLC. Typically,
after 4–6 h of the folding reaction the solvent was removed by lyo-
philization and the desired folded peptides were further purified
by preparative HPLC by using a gradient of solvents ranging from
80% of solvent A to 50% of solvent B at a flow rate 3 mLminÀ1
over 60 min. The final products were then lyophilized. Typically, we
obtained 3–4 mg of final folded peptides starting from 10 mg of
corresponding linear precursors.
CD spectra measurement: The circular dichroism (CD) experi-
ments were carried out on a Jasco 815 spectropolarimeter (Tokyo,
Japan). All peptide samples were measured in H2O, in 50% (v/v)
TFE, and in the presence of SDS at concentrations of 0.16 mm to
16 mm (below and above the critical micelle concentration) with
the final peptide concentration 0.25 mgmLÀ1. The spectra were
collected from 190 nm to 300 nm as averages over four scans at
RT by using a 0.1 cm path length. A 0.5 nm step resolution,
10 nmminÀ1 speed, 32 s response time, and 1 nm bandwidth were
generally used. Following baseline correction, the final spectra
were expressed as molar ellipticity q [degcm2 dmolÀ1] per residue.
The secondary structure representation was calculated by using
the circular dichroism analysis program Dichroweb.[26]
Determination of disulfide bridges: Lucifensin and Luc[des1–10,
Ala30] (about 0.05 mg) were dissolved in 0.1m Mes buffer, pH 6.5
containing 2 mm CaCl2 (50 mL). After addition of of thermolysin
stock solution (1 mL; 1 mgmLÀ1), the mixture was incubated at
358C for 16 h. The reaction was quenched by adding 10% TFA
(2 mL) and then the entire mixture was subjected to fractionation
by HPLC on a Vydac C-18 column (250ꢃ4.6 mm). The material was
at first eluted with 2% MeCN/H2O/0.1% TFA for 10 min and then
with a linear gradient of solvents from 2% to 35% MeCN/H2O/
0.1% TFA over 60 min at a 1 mLminÀ1 flow rate. The elution was
monitored by absorption at 220, 254, and 280 nm. The fractions
were collected, the solvent evaporated in a Speed-Vac, and the
peptide fragments were analyzed by MS to identify their sequen-
ces.
Transmission electron microscopy: Bacillus subtilis cells treated
with lucifensin for either 10 or 60 min and untreated control cells
were used for negative staining examination. Bacteria were ad-
sorbed on parlodion-carbon-coated copper grids for 5 min. After
short washing, the samples were negatively stained by floating on
a drop of 0.25% phosphotungstic acid (PTA) with 0.01% BSA in
dH2O for 30 s. Excess stain was blotted off with a piece of filter
paper and samples were air-dried. A JEOL JEM/1200EX transmission
electron microscope operating at 60 kV was used for analysis.
Acknowledgements
Alkylation of linear lucifensin peptide: HPLC-purified linear luci-
fensin peptide was dissolved in a solution of iodoacetamide (1 mg)
in 0.05m NH4HCO3 (50 mL). The mixture was shaken at RT in dark-
ness for 1 h and then the reaction was quenched by 10% TFA
(3 mL). The fully alkylated peptide (all 6 Cys residues) was isolated
This work was supported by the Czech Science Foundation, Grant
No. 203/08/0536, and by Research Project No. Z40550506 of the
Institute of Organic Chemistry and Biochemistry, Academy of Sci-
ences of the Czech Republic. We thank our technical assistant
1360
www.chembiochem.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1352 – 1361

Study of Lucifensin
[18] E. Klꢅver, S. Schulz-Maronde, S. Scheid, B. Meyer, W. G. Forssmann, K.
Adermann, Biochemistry 2005, 44, 9804–9816.
[19] M. Mandal, M. V. Jagannadham, R. Nagaraj, Peptides 2002, 23, 413–418.
[20] J. Varkey, R. Nagaraj, Antimicrob. Agents Chemother. 2005, 49, 4561–
4566.
ˇ
Hana Hulacovꢀ for the help with peptide synthesis. We also
thank Gale A. Kirking at English Editorial Services, s.r.o., for assis-
tance with the English.
[21] S. Liu, L. Zhou, J. Li, A. Suresh, C. Verma, Y. H. Foo, E. P. H. Yap, D. T. H.
Tan, R. W. Beuerman, ChemBioChem 2008, 9, 964–973.
[22] B. O. Schroeder, Z. Wu, S. Nuding, S. Groscurth, M. Marcinowski, J. Beis-
ner, J. Buchner, M. Schaller, E. F. Stange, J. Wehkamp, Nature 2011, 469,
419–423.
Keywords: antimicrobial activity · disulfide bridge · insect
defensin · peptides · solid-phase synthesis
[1] R. A. Sherman, M. J. R. Hall, S. Thomas, Annu. Rev. Entomol. 2000, 45,
55–81.
[2] Y. Nigam, E. Dudley, A. Bexfield, A. E. Bond, J. Evans, J. James, Adv. Insect
Physiol. 2010, 39, 39–81.
[3] A. Parnꢄs, K. M. Lagan, Int. J. Clin. Pract. 2007, 61, 488–493.
[4] Y. Nigam, A. Bexfield, S. Thomas, N. A. Ratcliffe, eCAM 2006, 3, 223–227.
[5] A. Kerridge, H. Lappin-Scott, J. R. Stevens, Med. Vet. Entomol. 2005, 19,
333–337.
[6] S. Thomas, A. M. Andrews, N. P. Hay, S. Bourgoise, J. Tissue Viability
1999, 9, 127–132.
[23] T. Kuzuhara, Y. Nakajima, K. Matsuyama, S. Natori, J. Biochem. 1990, 107,
514–518.
[24] K. H. Lee, S. Y. Hong, J. E. Oh, FEBS Lett. 1998, 439, 41–45.
[25] P. Bulet, R. Stçcklin, Protein Pept. Lett. 2005, 12, 3–11.
[26] L. Whitmore, B. A. Wallace, Biopolymers 2008, 89, 392–400.
[27] J. I. Kim, H. Iwai, S. Kurata, M. Takahashi, K. Masuda, I. Shimada, S.
Natori, Y. Arata, K. Sato, FEBS Lett. 1994, 342, 189–192.
[28] X. Lauth, A. Nesin, J.-P. Briand, J.-P. Roussel, C. Hetru, Insect Biochem.
Mol. Biol. 1998, 28, 1059–1066.
[7] A. Bexfield, Y. Nigam, S. Thomas, N. A. Ratcliffe, Microbes Infect. 2004, 6,
1297–1304.
[8] A. Bexfield, A. E. Bond, E. C. Roberts, E. Dudley, Y. Nigam, S. Thomas,
R. P. Newton, N. A. Ratcliffe, Microbes Infect. 2008, 10, 325–333.
[29] M. Mandal, R. Nagaraj, J. Peptide Res. 2002, 59, 95–104.
[30] H. Hanzawa, I. Shimada, T. Kuzuhara, H. Komano, D. Kohda, F. Inagaki, S.
Natori, Y. Arata, FEBS Lett. 1990, 269, 413–420.
[31] B. Cornet, J.-M. Bonmatin, C. Hetru, J. A. Hoffmann, M. Ptak, F. Vovelle,
Structure 1995, 3, 435–448.
ˇ
ˇ ˇ
[9] D. Jaklic, A. Lapanje, K. Zupancic, D. Smrke, N. Gunde-Cimerman, J. Med.
Microbiol. 2008, 57, 617–625.
[10] L. Huberman, N. Gollop, K. Y. Mumcuoglu, E. Breuer, S. R. Bhusare, Y.
Shai, R. Galun, Med. Vet. Entomol. 2007, 21, 127–131.
[32] K. Takeuchi, H. Takahashi, M. Sugai, H. Iwai, T. Kohno, K. Sekimizu, S.
Natori, I. Shimada, J. Biol. Chem. 2004, 279, 4981–4987.
[33] P. E. Dawson, S. B. H. Kent, Annu. Rev. Biochem. 2000, 69, 923–960.
ˇ
[11] Y. Nigam, A. Bexfield, S. Thomas, N. A. Ratcliffe, eCAM 2006, 3, 303–308.
[12] J. A. Hoffmann, C. Hetru, Immunol. Today 1992, 13, 411–415.
[13] K. Matsuyama, S. Natori, J. Biol. Chem. 1988, 263, 17112–17116.
[14] J. Lambert, E. Keppi, J.-L. Dimarcq, C. Wicker, J.-M. Reichhart, B. Dunbar,
P. Lepage, A. Van Dorsselaer, J. Hoffmann, J. Forthergill, D. Hoffmann,
Proc. Natl. Acad. Sci. USA 1989, 86, 262–266.
ˇ
´
ˇ
ˇ
[34] V. Cerovsky, O. Hovorka, J. Cvacka, Z. Voburka, L. Bednꢀrovꢀ, L. Borovick-
ˇ
ovꢀ, J. Slaninovꢀ, V. Fucꢁk, ChemBioChem 2008, 9, 2815–2821.
[35] O. Lequin, A. Ladram, L. Chabbert, F. Bruston, O. Convert, D. Vanhoye,
G. Chassaing, P. Nicolas, M. Amiche, Biochemistry 2006, 45, 468–480.
[36] Z. Oren, Y. Shai, Biochemistry 1997, 36, 1826–1835.
[15] K. Yamada, S. Natori, Biochem. J. 1993, 291, 275–279.
ˇ
ˇ
ˇ
ˇ
´
ˇ
[16] V. Cerovsky, J. Zdꢀrek, V. Fucꢁk, L. Monincovꢀ, Z. Voburka, R. Bꢄm, Cell.
Mol. Life Sci. 2010, 67, 455–466.
[17] A. S. Andersen, D. Sandvang, K. M. Schnorr, T. Kruse, S. Neve, B. Joergen-
sen, T. Karlsmark, K. A. Krogfelt, J. Antimicrob. Chemother. 2010, 65,
1646–1654.
Received: January 27, 2011
Published online on May 10, 2011
ChemBioChem 2011, 12, 1352 – 1361
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1361