Báo cáo khoa học: Poneratoxin, a neurotoxin from ant venom Structure and expression in insect cells and construction of a bio-insecticide pot

Poneratoxin, a neurotoxin from ant venom
Structure and expression in insect cells and construction of a bio-insecticide
Ewa Szolajska
1
, Jaroslaw Poznanski
1
, Miguel Lo
´
pez Ferber
2
, Joanna Michalik
1
, Evelyne Gout
3
,
Pascal Fender
3
, Isabelle Bailly
3
, Bernard Dublet
3
and Jadwiga Chroboczek
3
1
Institute of Biochemistry and Biophysics (IBB), Polish Academy of Sciences, Warsaw, Poland;
2
Laboratoire de Pathologie
Compare
´
e, UMR 5087, INRA-CNRS-Universite
´

de Montpellier II, St. Christol les Ales;
3
Institute of Structural Biology (IBS),
Grenoble, France
Poneratoxin is a small neuropeptide found in the venom
of the ant Paraponera clavata. It is stored in the venom
reservoir as an inactive 25-residue peptide. Here we des-
cribe both chemically synthesized poneratoxin and pon-
eratoxin obtained by expression in insect cells. When
expressed in insect cells, poneratoxin was observed
attached to cell membranes. Both synthetic and recom-
binant ponerotoxins were soluble below pH 4.5. The
structure of synthetic poneratoxin was characterized by
circular dichroism and solved by nuclear magnetic reso-
nance. In an environment imitating a lipid bilayer, at pH
within the range of insect hemolymph, synthetic ponera-
toxin has a V shape, with two a-helices connected by a
b-turn. Insect larvae were paralyzed by injection of either
of the purified toxins, with the recombinant one acting
faster. The recombinant toxin-producing baculovirus
reduced the average survival time of the insect host by
25 h compared with unmodified virus. Mass spectrometry
analysis showed that the recombinant toxin has an
N-terminal 21-residue extension, possibly improving its
stability and/or stabilizing the membrane-bound state. The
potential use of poneratoxin for the construction of bio-
logical insecticide is discussed.
Keywords: synthetic poneratoxin; recombinant poneratoxin;
baculovirus; insecticide; peptide atomic structure.
Living organisms have developed natural toxins targeting

key metabolic pathways of either their predators or their
prey. These toxins are used in research as molecular probes,
targeting with high affinity selected ion channel subtypes. As
such, they are very useful for understanding the mechanism
of synaptic transmission. Moreover, studies on toxin entry
into cells have been important for unraveling the mechanism
of cell endocytosis and the functioning of membrane
receptors. Many arthropod species such as scorpions and
spiders, as well as insects (bees, wasps and ants) produce
venom, which is a mixture of different neurotoxins,
arthropods’ natural insecticides. Some of these neurotoxins
are peptides.
The insect viruses, baculoviruses, have been used as insect
pest control agents since the last century [1]. They have a
relatively narrow host range, which might allow specific
pests to be targeted. However, the baculovirus life cycle is
complex and long, so it takes several days before the
infected insect dies, leading to considerable damage to
crops. To overcome this limitation, several attempts have
been made to obtain baculoviruses with enhanced toxicity.
Recombinant baculoviruses have been constructed with
genes coding for regulators of insect metabolism such as
hormones and enzymes [2,3], but also for natural toxins of
scorpions, mites, or spiders [4–8]. When compared with the
wild-type virus, some of these recombinants were able to
reduce the life span of infected insects.
The tropical ant Paraponera clavata is a predator of small
animals such as insect larvae. Its venom contains a potent
insect-specific peptide neurotoxin, poneratoxin. Ponera-
toxin affects the voltage-dependent sodium channels and

blocks the synaptic transmission in the insect central
nervous system in a concentration-dependent manner
[9–11]. It appears to be a good candidate for the construc-
tion of a baculovirus insecticide apt to immobilize the
infected insect. In this study, we have solved the atomic
structure of poneratoxin. In addition, we have expressed the
toxin in baculovirus and explored the biological properties
of such recombinant virus.
Experimental procedures
Matrix assisted laser desorption ionization-time of flight
(MALDI-TOF) mass spectrometry analysis
Mass spectra were obtained with a Perseptive Biosystems
(Framingham, MA, USA) Voyager Elite Xl time of flight
mass spectrometer with delayed extraction, operating with a
pulsed nitrogen laser at 337 nm. Positive-ion mass spectra
were acquired using a linear, delayed extraction mode with
Correspondence to J. Chroboczek, Institute of Structural Biology
(IBS), 41 J. Horowitz, 38027 Grenoble, France.
Fax: + 33 4 38785494, Tel.: + 33 4 38789590, E-mail:
Abbreviations: AcMNPV, Autographa californica nuclear polyhedrosis
virus; MOI, multiplicity of infection; PC, phosphatidylcholine; pfu,
plaque-forming units; Px, poneratoxin; SPx, poneratoxin preceded by
signal peptide; TFA, trifluoroacetic acid; TFE, 2,2,2-trifluoroethanol.
(Received 19 December 2003, revised 10 March 2004,
accepted 30 March 2004)
Eur. J. Biochem. 271, 2127–2136 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04128.x
an accelerating potential of 20 kV, a 94% grid potential, a
0.15% guide wire voltage, and a delay time of 100 ns. Each
spectrum is the results of 100 averaged laser pulses. Aliquots
of 1 lL samples and of 1 lL of a saturated solution of

a-cyano-4-hydroxycinnamic acid prepared in 50% aqueous
acetonitrile/0.3% trifluoroacetic acid (TFA; v/v/v) were
mixed on the stainless steel sample plate and dried in air
prior to analysis. External calibration was performed with
standards using the averaged m/z values of 1297.51,
2094.46, 2466.72, 3660.19 and 5734.59.
Synthetic poneratoxin, antibody and
P. clavata
venom
The peptide with the sequence FLPLLILGSLLMTPPVI
QAIHDAQR-NH
2
was synthesized by solid phase synthe-
sis using t-boc chemistry. It was purified by reverse phase
HPLC on Vydac C 18 column with an acetonitrile gradient
of 35 to 90% in 0.1% (v/v) TFA and it eluted at 73% (v/v)
acetonitrile (Fig. 1B). Peptide identity, integrity and purity
were analyzed by MALDI-TOF mass spectrometry. This
analysis revealed a peak with molecular mass of 2756. The
peptide was coupled to ovalbumin by the benzidine method
at a ratio 17/1 and used to immunize rabbits. P. clavata
dried venom was a kind gift of J. O. Schmidt [12].
1
Genes, plasmids, viruses and cells
Synthetic oligonucleotides made with the codon choice
suitable for AcMNPV [13] were used for the construction of
the poneratoxin gene [11]. Two oligonucleotides: forward
5¢-
2
GATCCATGTTTCTTCCGCTTCTGATCCTTGGCT

CTCTTCTGATGAC-3¢ and reverse 5¢-CGGCGTCATCA
GAAGAGAGCCAAGGATCAGAAGCGGAAGAAA
CATG-3¢, were used to synthesize the N-terminal fragment
of the poneratoxin gene. Two others: forward 5¢-GCC
GCCCGTGATACAGGCGATCCACGATGCGCAGA
GGTAGTAATGAG-3¢ and reverse 5¢-AATTCTCATTA
CTACCTCTGCGCATCGTGGATCGCCTGTATCAC
GGG-3¢ were used to construct the C-terminal fragment.
After phosphorylation and annealing, both fragments were
ligated. The full-length gene contained 75 nucleotides with
an ATG codon in front of the gene and three stop codons
at the end of it, and with flanking regions containing the
restriction sites for BamHI and EcoRI. The gene was
inserted into the pFastBac transfer vector (Life Technology)
giving a recombinant plasmid pFastBacPx. The second
construct containing the poneratoxin gene with the up-
stream signal sequence of AcMNPV glycoprotein gp67 [14]
was made by a three-step PCR amplification. For the first
PCR step, an upstream primer containing the 5¢ signal
peptide sequence: 5¢-GAATTC
ATGCTACTAGTAAAT
CAG-3¢ (number 1) and downstream primer with a
sequence complementary to the 5¢ end of poneratoxin gene
and 3¢ end of a signal peptide: 5¢-CAGAAGCGGAA
GAAA
GCATGCAAAGGCAGA-3¢ (number 2), were
used. In the second PCR the plasmid pFastBacPx was used
as a template with upstream and downstream primers
containing, respectively, the first 15 nucleotides of the 3¢ end
of signal peptide sequence and 15 nucleotides of the 5¢ end

of the poneratoxin gene (number 3) and 12 nucleotides
complementary to the 3¢ end of poneratoxin gene (number
4): 5¢-
TCTGCCTTTGCATGCTTTCTTCCGCTTCTG-3¢
and 5¢-GAATTCTCATTACTACCT-3¢. Finally, the mix-
ture of these two PCR reactions was used as a template with
primers numbers 1 and 4 in a 30-cycles run. In all of these
oligonucleotides the sequence of the poneratoxin gene is in
bold letters and that of signal peptide is underlined.
The final DNA products were digested with BamHI and
EcoRI and inserted into the vector pFastBac, yielding,
respectively, pFastBacSPx and pFastBacPx. All constructs
were confirmed by DNA sequencing. The recombinant
baculoviruses with signal peptide (SPx) or without it (Px)
were generated in the Bac-to-Bac Expression System (Life
Fig. 1. Expression of recombinant poneratoxin in the baculovirus system
and its purification. (A)Totalcellextractofponeratoxin-expressing
Sf21 cells (5 · 10
5
) was subjected to 20% SDS/PAGE and Western
blot. Lane 1, crude cell lysate; lane 2, synthetic poneratoxin (0.5 lg);
lane 3, P. clavata venom (62.5 lg). (B) Fractionation on reverse-phase
HPLC C18 column of the poneratoxin pool after Superdex column.
Fractions containing the recombinant poneratoxin are indicated with
the arrow. (C) Purified recombinant poneratoxin was analyzed on
20% SDS/PAGE and revealed with silver stain. Lane 1, fractions no.
49–51; lane 2, synthetic poneratoxin. Molecular mass markers (in kDa)
areshownontheleftofbothgels.
2128 E. Szolajska et al. (Eur. J. Biochem. 271) Ó FEBS 2004
Technology). They were propagated in Sf21 cells (IPBL

from Spodoptera frugiperda), as well as in HF (High Five
or BTI-TN-5B1-4 from Trichoplusia ni)(Invitrogen),main-
tained in TC 100 medium supplemented with 5% (v/v) fetal
bovine serum [15].
Protein expression
Virus stocks were prepared by infection of Sf21 monolayers
at the multiplicity of infection (MOI) 0.1 with the
supernatants obtained after cell transfection with recom-
binant baculovirus DNA. For protein expression, insect
cells (l.5 · l0
6
per 35-mm dish) were infected at MOI 10. For
time course of expression, harvested cells were resuspended
in 50 m
M
Tris/1 m
M
EDTA/100 m
M
NaCl, pH 8.0, and
lysed by three cycles of freezing and thawing. The total cell
extract was analyzed for the presence of poneratoxin by
20% SDS/PAGE followed by Western blot with the
antiponeratoxin antibody (500-fold dilution). Electrotrans-
fer onto poly(vinylidene difluoride) membrane was carried
out in the presence of 10% (v/v) methanol.
Production and purification of recombinant poneratoxin
Three days after infection with the recombinant baculovirus
the insect cells were collected and washed twice with 50 m
M

Tris/1 m
M
EDTA/100 m
M
NaCl,pH8.Thecellswere
suspended in lysis buffer [50 m
M
Tris/1% (v/v) Nonidet
P-40/200 m
M
NaCl/1 m
M
EDTA, pH 8.5] containing
Complete protease inhibitors (Boehringer) and sonicated.
The extract was centrifuged at 10 000 g
3
for 20 min and the
resulting pellet was dissolved in 70% formic acid. It was
fractionated on the Superdex Peptide HR 10/30 column
(Pharmacia) using 70% (v/v) formic acid for elution.
Fractions containing the recombinant peptide were applied
onto C18-218 TP54 column (4.6 · 250 mm, Vydac) and
eluted with acetonitrile/water gradient (35–100%) contain-
ing 0.1% TFA (v/v/v). Recombinant poneratoxin eluted
between 87% and 91% acetonitrile (v/v) (Fig. 1B). Western
blot with antiponeratoxin serum was used to identify
fractions containing poneratoxin.
Confocal microscopy
Recombinant baculovirus-infected cells were collected by
low-speed centrifugation, transferred onto cover slips and

fixed for 6 min in 50% ethanol/0.2% Triton X-100 (v/v/v).
They were washed twice with TBST [20 m
M
Tris/150 m
M
NaCl/1 m
M
EGTA/2 m
M
MgCl
2
/0.4% (v/v) Tween 20,
pH 7.2] and incubated for 1 h with the antiponeratoxin
serum diluted 1/400 in TBST. After TBST wash, the
samples were incubated with FITC-conjugated goat
antiserum against rabbit IgG (Pasteur Institute, Paris) for
1 h at room temperature, washed three times with TBS,
mounted with Citifluor (Citifluor Ltd, UK) and photo-
graphed with Leitz–Wetzlar confocal microscope.
Toxicity assays
S. frugiperda larvae were obtained from a laboratory colony
reared on semiartificial diet [16] at 22.5 ± 0.5 °C, 70%
humidity, with 16 h photoperiod. Groups of 12 fourth
instar larvae aged 11 days, with average weight
148.6 ± 3.7 mg, were injected with 8 lL±0.5lLeach
of infectious supernatant of either unmodified parental
baculovirus bMON14272 (Bac-to-Bac Expression System),
or virus containing poneratoxin gene (Px) and virus with
poneratoxin gene preceded by signal peptide (SPx).
Preliminary experiments indicated that a dose of 10

5
pfu
was sufficient to kill all the larvae, and so the three
baculoviruses were diluted with NaCl/P
i
to reach this dose
in 8 lL. The mock-infected group was injected with 8 lL
NaCl/P
i
. To evaluate the LT
50
, a total of 132 larvae were
injected each with SPx or Px virus supernatants, and 84 with
the unmodified control virus. Survival was scored every 8 h.
The data were analyzed using the Kruskal–Wallis non-
parametric test with the correction for tied ranks [17].
Individual comparisons were carried out using the Dunn
test [18]. Synthetic and recombinant poneratoxins were
dissolved in 50 m
M
acetic acid/NaOH, pH 4.5. Groups of
12 larvae (approximate weight 158 mg) were injected with
10 ng of each peptide and with the P. clavata venom
equivalent to 10 ng of poneratoxin [11] or with 8 lL solvent
as a control. The time needed to paralyze larvae as well as
paralysis duration was monitored.
Circular dichroism (CD) measurements
The spectra were collected at 25 °C in 185–270 nm
wavelength range with a 0.2 nm spectral step size on an
AVIV 202 spectropolarimeter, using 1 cm path-length cell.

Each spectrum was recorded as an average of three scans
and then corrected for the buffer background. For all CD
measurements the same 360 l
M
stock solution of ponera-
toxin prepared in 1% (v/v) aqueous 2,2,2-trifluoroethanol
(TFE) was used. The presence of TFE permitted studies
at pH 5.5. Aqueous solutions of SDS and PC were used
at 10 and 1% (v/v), respectively. CD measurements were
carried out with samples obtained by mixing the adequate
amounts of 1% (v/v) aqueous TFE, peptide stock, and
TFE or SDS or PC solutions. Before each experiment the
exact peptide concentration (set initially at  5 l
M
)was
determined from the absorption at 280 nm using the
extinction coefficient calculated according to the peptide
sequence [19]. Estimations of secondary structure elements
were carried out using deconvolution by back-propagation
of neural networks implemented by Bo
¨
hm et al.[20].All
the CD data were expressed as mean residue ellipticity
given in °Æcm
2
Ædmol
)1
.
Nuclear magnetic resonance (NMR) experiments
NMR measurements were performed on the Bruker AMX

600 MHz spectrometer at 298 K. Peptide solution (3 m
M
)in
the 25 : 65 : 10 (v/v/v) mixture of
2
H-enriched-TFE, H
2
O
and
2
H
2
O was adjusted to pH 5.5. The standard COSY,
TOCSY as well as 100 and 200 ms mixing time NOESY
spectra were accumulated, processed by
NMRPIPE
[21] and
analyzed by
X
-
EASY
program [22]. Structure determination
was obtained with
DYANA
software in the REDAC strategy
mode [23]. Final refinement was carried out by simulated
annealing procedure with help of
X
-
PLOR

[24]. The ponera-
toxin structure was deposited in the Protein Data Bank
(), accession code PDB1G92. Chemical
Ó FEBS 2004 Structure and expression of an ant neurotoxin (Eur. J. Biochem. 271) 2129
shift and coupling constants have been deposited in
BioMagResBank (), accession
code BMRB-4921.
Results
One of the major components of the venom of tropical ant
P. clavata is poneratoxin, a peptide largely responsible for
the venom’s neurotoxic activity [9,10,12]. We were interested
in the structure of this peptide and its anti-insect neuro-
toxicity with the idea of using it for engineering a
baculovirus capable of serving as a bio-insecticide.
Peptide synthesis and expression of recombinant
poneratoxin
Using the published poneratoxin sequence [12], the peptide
FLPLLILGSLLMTPPVIQAIHDAQR-NH
2
was chemic-
ally synthesized and purified by HPLC. Mass spectrometry
analysis of the product gave the correct molecular mass of
2756 (theoretical 2757). This peptide was used for structural
studies and to raise the polyclonal poneratoxin-specific
antibody.
For the poneratoxin expression, two recombinant bacu-
loviruses with the poneratoxin gene were engineered, one
containing a signal peptide and another without it. As the
signal for poneratoxin secretion is not known, the 39 amino
acid signal sequence of the major envelope glycoprotein

gp67 (MLLVNQSHQGFNKEHTSKMVSAIVLYVLLA
AAAHSAFA) was used. The same sequence was success-
fully employed in the baculovirus system for expression of
the insect-specific neurotoxin of the scorpion Androctonus
australis [6]. The sequence includes all the nucleotides from
the first ATG of the open reading frame. During the cloning
procedure, a cysteine residue was introduced between this
signal and the mature poneratoxin. Surprisingly, ponera-
toxin expression was detected only in cells infected with
baculovirus containing the poneratoxin gene with the signal
peptide. The maximum level of expression was seen at three
days post infection, in good agreement with the usual
activity of the polyhedrin promoter. A similar expression
level was observed in both, Sf21 and HF, cell lines.
Therefore Sf21 cells were used for the rest of the studies as
they are easier to grow in suspension.
Toxin expression was analyzed on 20% (w/v) polyacryl-
amide denaturing gel followed by immunoblot. No toxin
was detected in the extracellular medium (tested using
reverse-phase C18 SepPac and Western blot, not shown).
Analysis of fractions derived from the crude lysate revealed
that poneratoxin is expressed in an insoluble form (not
shown). The recombinant toxin was retarded on denaturing
gels in comparison with the synthetic peptide (Fig. 1A) but
this mobility difference cannot be explained by the disulfide
bridge as electrophoresis was run in the presence of reducing
agent. The mass spectroscopy data for the 25-amino acid
synthetic peptide confirmed its integrity (molecular mass
2756). However, the mass spectroscopy of the recombinant
poneratoxin contained in the formic acid extract showed

that it has molecular mass of 4861, compatible with a longer
peptide starting with the methionine in the middle of the
signal peptide (theoretical mass 4861). It is relevant that
when gp64 (called also gp67) is expressed during AcMNPV
infection, the second ATG in the open reading frame is used
as the translation initiation codon and that downstream
sequences encode a functional signal peptide [25]. In
addition, the preparation showed the presence of a second
species with mass of 4888, suggesting the postranslational
modification by formylation of the initiator methionine
(theoretical mass 4889), which explained the difficulties
encountered in the N-terminal sequencing. The possible
dimerization of the recombinant peptide mediated by the
N-terminal cysteine (added during the cloning steps) was
excluded by repeating the mass spectrometry analysis under
reductive conditions, with unchanged results.
The pH of P. clavata venom is very low, due to the high
concentration of formic acid. Accordingly, the synthetic
peptide is soluble below pH 4.5 and such conditions were
used for the extraction and purification of recombinant
poneratoxin. On reverse-phase C18 column the recombin-
ant poneratoxin eluted at higher acetonitrile concentration
than the synthetic peptide (Fig. 1B). It is relevant in this
context that the recombinant peptide has the N-terminal
extension MVSAIVLYVLLAAAAHSAFAC, which will
likely reinforce its hydrophobic character.
Confocal microscopy was used to determine the cellular
localization of the recombinant poneratoxin in insect cells.
The toxin was observed at the cell periphery (Fig. 2), and
treatment with the mild detergent NP-40 did not liberate it

from the insoluble fraction (data not shown). This suggests
that poneratoxin synthesized in the cytoplasm becomes
insoluble upon its transfer towards the cell membrane.
Toxicity studies
ForthetestsonS. frugiperda larvae three baculoviruses
were used: the unmodified parental virus obtained after
infection of insect cells with the initial unmodified shuttle
vector, the recombinant virus with the poneratoxin gene
(Px) and the third virus with the poneratoxin gene preceded
Fig. 2. Confocal microscopy of the Sf21 cells expressing recombinant
poneratoxin. Baculovirus-infected cells were collected on the cover
slips, fixed, incubated with the antiponeratoxin serum and observed
with Leitz–Wetzlar confocal microscope as described in the Materials
and methods. Magnification ·1000.
2130 E. Szolajska et al. (Eur. J. Biochem. 271) Ó FEBS 2004
by signal peptide (SPx). They were obtained with titers of
2 · 10
7
,2.5· 10
7
and 1.5 · 10
7
pfuÆmL
)1
, respectively.
Global analysis of the data (Fig. 3 and Table 1) confirms
the difference in the killing rate of the three viruses
(Kruskal–Wallis H-values equal 59.91 after correction for
tied ranks, with a P ¼ 9.77 · 10
)14

). The highest killing rate
was observed with the baculovirus expressing recombinant
poneratoxin. The killing rate of 50% was reached at about
131 h post injection
4
by SPx and at 160 h post injection by
the parental virus. The difference is statistically significant
(Q ¼ 3.22, P < 0.005). Thus, the expression of poneratox-
in gives virus with improved killing properties. Surprisingly,
the Px baculovirus containing the poneratoxin gene but
unable to express the peptide, killed the larvae 35 h later
than the parental baculovirus. The Px virus seems to be
disabled in its multiplication due to instability; we observed
titers decreasing with time for Px, with constant titers for
two other viruses. Nevertheless, toxicity studies on freshly
obtained viruses resulted in 100% larvae killing by all three
viruses (Fig. 3).
To estimate the paralyzing activity of these neurotoxins,
the larvae were injected with synthetic and recombinant
poneratoxins as well as the P. clavata venom containing an
equivalent amount of poneratoxin (estimated according to
Piek et al. [11]) or with the solvent alone (50 m
M
acetic acid/
NaOH, pH 4.5). The strongest paralyzing effect was exerted
by venom (Table 2), which suggests that other venom
components might also be neurotoxins. Recombinant
poneratoxin was more toxic than the synthetic one. It
should be borne in mind that the recombinant poneratoxin
has the 21 amino acid extension compared with the

synthetic one. Unless this difference in activity is due to
some as yet uncharacterized post-translational modifica-
tions of the recombinant toxin, it seems that the N-terminal
extension increases its neurotoxicity. It is conceivable that
the hydrophobic extension might improve toxin stability
resulting in longer bioavailability.
Structure characterization by CD
All the CD spectra were analyzed at pH 5.5. The CD
spectrum obtained for the synthetic peptide in 1% (v/v)
TFE solution was dominated by a minimum located at
200 nm and exhibited no maximum below 200 nm (Fig. 4),
Fig. 3. Cumulative mortality (in percentage) of S. frugiperda fourth
instar larvae, injected with 10
5
pfu of SPx, Px and control virus or buffer
(mock) injected.
Table 1. Average survival times (in hours) of 4th instar S. frugiperda
larvae. S. frugiperda larvae were injected with 10
5
pfu of the parental
virus (control, shuttle vector bMON14272), the virus expressing pon-
eratoxin (SPx), and the virus unable to express poneratoxin (Px). The
95% interval is the confidence interval for a type 1 error of > 0.05. It
shows that the real average value obtained from our data would be
between the lower and higher values in 95% of the experiments using
the same population.
Virus
95% interval
Median Lower Higher
SPx 136.17 130.95 148.85

Control 161.23 153.36 178.67
Px 196.36 189.36 203.486
Table 2. Direct paralyzing effect of poneratoxin. Groups of 12 S. fru-
giperda larvae were injected with 10 ng of each peptide and with the
P. clavata venom equivalent to 10 ng of poneratoxin. Larvae were
scored as paralyzed if they were unable to right themselves within 30 s
of being placed on their backs. As a control, 12 larvae were injected
with 8 lL of sample solvent. They showed some reduction in mobility
at 2 min after injection and then recuperated.
Toxin Paralysis observed after Recovery after
Venom 30 s 25 min
Recombinant 3 min 7 min
Synthetic 11 min 3 min
Solvent Not observed
Fig. 4. Molar ellipticity of synthetic poneratoxin as a function of SDS
concentration in 1% TFE (A) and of TFE concentration (B). In both
figures, curve (a) was obtained at 1% TFE. Concentrations of SDS in
(A) were 0.2% (b), 1.4% (c) and 2.5% (d); and TFE in (B) 3% (b), 6%
(c), 12% (d), 25% (e) and 50% (f).
Ó FEBS 2004 Structure and expression of an ant neurotoxin (Eur. J. Biochem. 271) 2131
which is characteristic of a highly populated unordered
random conformation [26]. However, CD experiments
performed upon PC addition with the aim of imitating
lipid bilayer environment demonstrated stabilization of the
a-helical conformation (results not shown). Upon addition
of PC the minimum moved toward longer wavelengths
becoming deeper and a maximum appeared at 192 nm,
which is indicative of the stabilization of a-helical structure.
However, due to strong light scattering, the concentration
range of PC was very limited and only qualitative analysis

could be performed. In contrast, use of SDS, a detergent
with micelles exhibiting no significant scattering, permitted
quantitative analysis of the helix stabilization induced by the
micellar phase. The SDS titration of poneratoxin in 1%
(v/v) TFE demonstrated a systematic increase of the positive
peak at 192 nm accompanied by the buildup of the minima
at 208 nm and 223 nm, clearly indicating SDS-induced
stabilization of peptide helical structure (Fig. 4A). The
estimated partition of helical structure exceeded 30% for
peptide in 1.4% (v/v) aqueous SDS.
TFE is known to promote a stable secondary structure of
the polypeptide chain in peptides [27] by strengthening the
internal H-bonds of the peptide [28]. TFE-induced con-
formational change is close to that observed for SDS
(Fig. 4B); the CD spectra recorded in 12% TFE and 1.4%
SDS are almost identical. The maximal effect of the
secondary structure stabilization (63% a-helical, 12%
b-turn, 18% random) was observed in 25% (v/v) TFE
solution. The increase of the TFE concentration above 25%
did not result in any significant change of CD spectra.
Additional titration experiment showed no significant
changes of peptide secondary structure in the pH range of
5.3–7.8 in 35% (v/v) TFE solution (data not presented).
Therefore, in order to minimize amide proton exchange
rates, the 25% (v/v) TFE aqueous solution of the synthetic
poneratoxin at pH 5.5 was used for the NMR analysis.
Structure of synthetic poneratoxin by NMR
The solution structure of poneratoxin was determined on
the basis of 428 experimentally derived distance restraints.
Finally 10 structures exhibiting occasional residual viola-

tions larger than 0.3 A
˚
were accepted. NMR-derived
structure showed the presence of two helical regions:
PLLILGS(3–9) and IQAIHDAQ(17–24), with residues
LLMTPPV(10–16) forming a turn. The structure of the
central LMTPPV(11–16) region of the peptide is almost
identical to the open turn type III conformation of
LMTDPV(151–156) fragment from the haloalkane dehalo-
genase of Xanthobacter autotrophicus [29]. For both helices
Fig. 5. Structural properties of the synthetic poneratoxin in solution. (A)
Sausage model of the mean structure. The thickness of the tube is a
measure of local backbone flexibility. Helical regions are in red. (B)
Hydrophobic potential on peptide surface. Hydrophobic residues are
in red, hydrophilic in blue. (C) The putative structure–function rela-
tion. The apolar N-terminal helix is marked in red, the C-terminal
polar helix is marked in blue. The helical regions are separated by a
turn (in green) The N-terminus is dark green. (D) Stick representation
of poneratoxin. Amino acids participating in the long–range hydro-
phobic interactions stabilizing V-shaped conformation are space-filled.
2132 E. Szolajska et al. (Eur. J. Biochem. 271) Ó FEBS 2004
the experimentally determined pattern of short-range con-
tacts is consistent with the proposed secondary structure.
The overall backbone root-mean-square deviation (rmsd)
is 1.32 A
˚
for 10 lowest-energy structures obtained after
simulated annealing procedure, indicating the high quality
of the obtained model. The average structure of the peptide
is presented in the sausage model (Fig. 5A). The radius of

the tube modeling the Ca backbone is proportional to the
local structure deviations among the ensemble of 10 selected
conformations. Detailed analysis showed that rmsd value
(which is a measure of structure quality) determined
separately for helical regions PLLILGS(3–9) or IQ-
AIHDAQ(17–24) is significantly lower (< 0.4 A
˚
). This
clearly demonstrates the internal stability of the helical
regions with their relative spatial organization weakly
defined, resulting in lowered number of experimental long-
range interhelical restraints.
In conclusion, the solution structure of synthetic poner-
atoxin can be modeled as two loosely interacting helices
with a preferred V-shaped orientation. The central loop is
stabilized by a small, flexible, but well defined hydrophobic
core built from Ile6, Leu10, Pro15, Leu17 and Gln18 side-
chains (Fig. 5D). The helix-break-helix organization of
poneratoxin is similar to that found for other peptides
interacting with the plasma membrane [30,31]. Taking into
account the distribution of polar/apolar residues along the
sequence and the sequence based prediction of peptide
localization [TM
PRED
at />ware/TMPRED_form.html predicted transmembrane
localization of FLPLLILGSLLMTPPVI(1–17) fragment]
it is conceivable that the N-terminal apolar helix favors
transmembrane localization while the C-terminal amphi-
philic helix, is either solvent exposed or interacting with the
membrane surface.

Discussion
Poneratoxin is a potent insect-specific toxin produced by the
predatory ant P. clavata. The primary sequence of poner-
atoxin has been obtained from the peptide mixture stored in
the venom reservoir [10]. The gene has not been character-
ized and nothing is known of the toxin processing during
synthesis and secretion into the venom reservoir.
A peptide was synthesized using the published amino acid
sequence [11]. Synthetic poneratoxin is a very hydrophobic
peptide of 25 amino acid residues with a rather charged
C-terminal part. Our CD data show that under conditions
imitating the membrane surroundings it has a propensity to
acquire an ordered structure. The NMR structure shows the
peptide in the form of two a-helices connected by a b-turn.
The two helices have quite different characters. The first,
PLLILGS(3–9), is apolar, whereas the second, IQ-
AIHDAQR(17–25), contains polar and charged amino
acids. This will result in different interactions with cell
membranes. The extremely hydrophobic N-terminal helix
will easily interact with uncharged lipid bilayers composed
of phosphatidylcholine [32]. The C-terminal helix, slightly
positively charged and terminating with arginine, will be
able to attach to negatively charged cell surfaces similar as
found for other membrane interacting peptides [30,31]. Such
a toxin can thus use two different complementary modes
of interaction to attain its target, cellular membranes.
Moreover, the poneratoxin sequence starts with a bulky
hydrophobic phenylalanine, which enhances the peptide
hydrophobicity index and may be important for membrane
penetration [33].

To analyze the biological activity of poneratoxin, we
constructed two recombinant baculoviruses, one with the
poneratoxin gene and another in which the peptide is
preceded by a secretion signal from a baculovirus gene.
However, no poneratoxin was detected when the gene
devoid of a signal peptide was used (virus Px). It seems likely
that when not exported, this peptide is either destroyed
inside the cell or is toxic for the cells. Sequence analysis
suggested that the part LPLLILGSLLMTPPVIQA(2–19),
which looks like a transmembrane segment is similar to
signal peptides of a variety of proteins [34]. If without an
authentic signal peptide this fragment is seen by the
expressing cell as a signal peptide, it could be degraded by
the appropriate enzymatic system such as one that cleaves
the signal peptide of preprolactine within its hydrophobic
core, between two leucine clusters [35].
When the toxin was preceded by a signal peptide,
baculovirus produced a recombinant toxin containing an
N-terminal extension of 21 amino acid residues. Interest-
ingly, the ATG coding for the middle methionine of the
signal peptide is contained in a perfect Kozak consensus
sequence AAGATGG, ensuring proper translation initi-
ation [36]. Thus, the recombinant poneratoxin seems to be
synthesized through the initiation from the second ATG in
the open reading frame, similar to insect protein gp64 [25], a
source of signal peptide. However, the signal peptide was
not cleaved from the poneratoxin resulting in an uncleaved
intracellular form, with no extra-cellular poneratoxin
detected. It cannot be excluded that the amount of mature
secreted toxin is too low for detection, but important

enough to be responsible for the biological activity.
Alternatively, the toxin liberated by cells dying from the
infection could be responsible for the increased pathogen-
icity observed in the biological assays. The results of the
biological assay demonstrate that the 21 amino acid residue
N-terminal extension improves the paralyzing activity of the
recombinant peptide when compared with the synthetic
one. The extension is quite hydrophobic in character and it
is conceivable that this improves toxin stability and therefore
its bioavailability. Alternatively, the N-terminal extension
could stabilize the active toxin conformation. Additional
experiments are needed to clarify these questions.
The pH of the lepidopteran larvae hemolymph is between
6.6 and 6.8 [37,38], which is the upper pH limit of
poneratoxin solubility. However, the infection of S. fru-
giperda or T. ni larvae with AcMNPV by intrahaemocoelic
injection starts with the progeny virus observed first in fat
bodies and epithelium, with a rather slow build-up in the
hemolymph [39]. It is conceivable that the conditions on the
surface of the epithelium are sufficient to allow partial toxin
solubility promoting its interaction with epithelial cell lipidic
membranes.
The poneratoxin activation is likely to be a multistep
process and the poneratoxin structure studies give some
insights into this process. Secretion of the native ponera-
toxin to the venom reservoir is most probably triggered by
the specific secretion signal. The neurotoxin stored in the ant
venom reservoir should be inactive, preventing damage to
Ó FEBS 2004 Structure and expression of an ant neurotoxin (Eur. J. Biochem. 271) 2133
the ant, suggesting that the acidic conditions present in the

venom reservoir will forbid the structure conformation
necessary for exerting poneratoxin neurotoxicity. Upon
injection of the venom in the hemocoel of the prey, the
membrane insertion and the pH change could trigger
conformational changes, yielding the active neurotoxin.
Few ant toxins are so far described [10,11,40–42] and only
one three-dimensional structure of an ant venom peptide
toxin is known [43]. This toxin, ectatomin, is built from two
chains, each consisting of two a-helices bound by a hinge
region. However, this structure seems to be much more rigid
than that of poneratoxin, as each chain forms a hairpin
stabilized by disulfide bridges. Furthermore, the chains are
connected by a third S–S bridge resulting in a four-alpha-
helical bundle structure. It should be noted that contrary to
the majority of known sequences of venom peptides specific
for sodium channels [44] native poneratoxin does not
contain cysteine and it is conceivable that it has a distinct
mode of action.
Classical arthropod toxins, such as those of scorpions,
form a family of small proteins of 30–70 amino acids
affecting the kinetics of sodium or potassium channels.
Autographa californica baculovirus recombinants expressing
some of these toxins present improved insecticide activity
as compared with wild type virus [45]. A well-documented
example of a recombinant baculovirus is that carrying the
sequence of the toxin from A. australis Hector scorpion
venom [46,47]. Several recombinant baculoviruses with
different toxin synthetic genes have been studied in the
laboratory [6,48,49] and in controlled conditions in the field
[47,50].

We thought that poneratoxin, another insect toxin, might
be a good candidate for the reinforcement of the insecticide
action of a baculovirus, perhaps providing an alternative
insecticide activity with a mechanism of action possibly
different from that of spiders and mite toxins. Toxicity
studies showed that the baculovirus engineered to express
poneratoxin is a better killer than the parental virus. The
gain in time is considerable if we remember that the feeding
period of S. frugiperda larvae extends up to 10 days. This is
shortened with the parental baculovirus infection to about
7 days and to 5–6 days with recombinant SPx baculovirus.
Similar results have been obtained with recombinant
baculoviruses expressing neurotoxins of the scorpion
A. australis or the mite Pyemotes tritici
5
[6,7,49]. We think
that further improvements could be obtained by adjusting
the secretion pathway to imitate the native one. Also
the development of a bio-insecticide expressing in parallel
two toxins targeting different pathways may significantly
increase killing speed [51].
Many concerns have been raised about the risk of using
genetically modified baculoviruses as insecticides in the field.
One is the possible toxicity of the recombinant protein to the
environment. To date, no detailed information exists on the
per os toxicity of poneratoxin to other animals, birds or
mammals, especially when released in nature. Predators will
ingest the almost-dead larvae containing the expressed
toxin. Poneratoxin appears to be soluble in acidic condi-
tions, close to those existing in the stomach of mammals.

However, it is not clear if the protein present in the larvae
cadavers is solubilized and released, and, if so, if it will
remain active. Also, the amounts of toxin released from
ingested larvae may or may not be high enough to have an
effect on the predators. Clearly, more work is required to
understand how this improvement in virus killing rate
occurs and what are its implications for the development of
safe baculovirus recombinant bio-insecticides.
Acknowledgements
This work was supported in part by NATO Linkage Grant no. 940881.
ES was supported in part by a Ôposte rougeÕ of CNRS. We are indebted
to J. O. Schmidt (South-west Venoms, Tucson, AZ, USA) for a sample
of P. clavata venom. The help of the Laboratory of Magnetic
Resonance (IBS) in the structural part of this work, G. Goch (IBB)
in CD spectroscopy, H. Kozlowska (IBB) in HPLC and of M. Jerka-
Dziadosz (Nencki Institute, Warsaw) in immunofluorescence technique
is acknowledged. We are grateful to A. Wyslouch (IBB) and to
M. Jaquinod and J P. Andreini (IBS) for discussions and to R. Wade
for reading our manuscript.
References
1. Entwistle, P.F. & Evans, H.F. (1985) Viral control. In Compre-
hensive Insect Physiology, Biochemistry and Pharmacology
(Gilbert, L.I. & Kerkut, G.A., eds), Vol. 12, pp. 347–412.
Pergamon Press, Oxford, UK.
2. Maeda, S. (1989) Increased insecticidal effect by a recombinant
baculovirus carrying a synthetic diuretic hormone gene. Biochem.
Biophys. Res. Commun. 165, 1177–1183.
3. Hammock, B.C., Bonning, B.C., Possee, R.D., Hanzlik, T.N. &
Maeda, S. (1999) Expression and effects of the juvenile hormone
esterase in a baculovirus vector. Nature 344, 458–461.

4. Carbonell, L.F., Hodge, M.R., Tomalski, M.D. & Miller, M.K.
(1988) Synthesis of a gene coding for an insect-specific scorpion
neurotoxin and attempts to express it using baculovirus vectors.
Gene 73, 409–418.
5. Burden, J.P., Hails, R.S., Windass, J.D., Suner, M.M. & Cory, J.S.
(2000) Infectivity, speed of kill, and productivity of a baculovirus
expressing the itch mite toxin Txp-1 in second and fourth instar
larvae of Trichoplusia ni. J. Invertebr. Pathol. 75, 226–236.
6. Stewart, M.D., Hirst, M., Ferber, M.L., Merryweather, A.T.,
Cayley, P.J. & Possee, R.D. (1991) Construction of an improved
baculovirus insecticide containing an insect-specific toxin gene.
Nature 352, 85–88.
7. Tomalski, D. & Miller, L.K. (1991) Insect paralysis by baculo-
virus-mediated expression of a mite neurotoxin gene. Nature 352,
82–85.
8. Kiyatkin, N.I., Kulikovskaya, I.M., Grishin, E.V., Beadle, D.J. &
King, L.A. (1995) Functional characterization of black widow
spider neurotoxins synthesised in insect cells. Eur. J. Biochem. 230,
854–859.
9. Duval, A., Malecot, C.O., Pelhate, M. & Piek, T. (1992) Poner-
atoxin, a new toxin from an ant venom, reveals an interconversion
between two gating modes of the Na channels in frog skeletal
muscle fibers. Pflugers Arch. 420, 239–247.
10. Piek,T.,Duval,A.,Hue,B.,Karst,H.,Lapied,B.,Mantel,P.,
Nakajima, T., Pelhate, M. & Schmidt, J.O. (1991) Poneratoxin, a
novel peptide neurotoxin from the venom of the ant, Paraponera
clavata. Comp. Biochem. Physiol. C. 99, 487–495.
11. Piek,T.,Hue,B.,Mantel,P.,Nakajima,T.&Schmidt,J.O.(1991)
Pharmacological characterization and chemical fractionation of
the venom of the ponerine ant, Paraponera clavata (F.). Comp

Biochem. Physiol. C. 99, 481–486.
12. Schmidt, J.O. (1986) Chemistry, pharmacology and chemical
ecology of ant venoms. In Venoms of the Hymenoptera (Piek, T.,
ed.), pp. 425–500. Academic Press, London.
2134 E. Szolajska et al. (Eur. J. Biochem. 271) Ó FEBS 2004
13. Ayres,M.D.,Howard,S.C.,Kuzio,J.,Lopez-Ferber,M.&Pos-
see, R.D. (1994) The complete DNA sequence of Autographa
californica nuclear polyhedrosis virus. Virology 202, 586–605.
14. Whitford,M.,Stewart,S.,Kuzio,J.&Faulkner,P.(1989)Iden-
tification and sequence analysis of a gene encoding gp67, an
abundant envelope glycoprotein of the baculovirus Autographa
californica nuclear polyhedrosis virus. J. Virol. 63, 1393–1399.
15. King, L.A. & Possee, R.D. (1992) The Baculovirus Expression
System. A Laboratory Guide. Chapman & Hall, London.
16. Poitout, S. & Bues, R. (1974) Elevage des chenilles de vingt-huit
espe
`
ces de lepidopte
`
res Noctuidae et de deux espe
`
ces dÕArctiidae
sur milieu artificiel simple. – Particularite
´
dÕe
´
levage selon les
espe
`
ces. Ann. Zool. Ecol. Anim. 6, 431–441.

17. Kruskall, W.H. & Wallis, W.A. (1952) Use of ranks in one-
criterion variance analysis. J. Am. Statist. Ass. 47, 583–621.
18. Dunn, O.J. (1964) Multiple contrasts using rank sums. Techno-
metrics 6, 241–252.
19. Gill, S.C. & von Hippel, P.H. (1989) Calculation of protein
extinction coefficients from amino acid sequence data. Anal
Biochem. 182, 319–326.
20. Bo
¨
hm, G., Muhr, R. & Jaenicke, R. (1992) Quantitative analysis
of protein far UV circular dichroism spectra by neutral networks.
Protein Eng. 5, 191–195.
21. 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.
22. Bartels,C.,Xia,T E.,Billeter,M.,Guntert,P.&Wuthrich,K.
(1995) The Program XEASY for the computer-supported NMR
spectral analysis of biological macromolecules. J. Biomol. NMR
5,1.
23. Guntert, P., Mumenthaler, C. & Wuthrich, K. (1997) Torsion
angle dynamics for NMR structure calculation with the new
program DYANA. J. Mol. Biol. 273, 283–298.
24. Brunger, A.T. (1992) The X-Plor Software Manual,Version3.1.
Yale University, New Haven, Connecticut.
25. Jarvis, D.L. & Garcia, A. Jr (1994) Biosynthesis and processing of
the A. californica nuclear polyhedrosis virus gp64 protein. Virology
205, 300–313.
26. Creighton, T.E. (1984) Pathways and mechanisms of protein
folding. Adv. Biophys. 18, 1–20.
27. Sonnichsen, F.D., Van Eyk, J.E., Hodges, R.S. & Sykes, B.D.

(1992) Effect of trifluoroethanol on protein secondary structure:
an NMR and CD study using a synthetic actin peptide. Bio-
chemistry 31, 8790–8798.
28. Luo, P. & Baldwin, R.L. (1997) Mechanism of helix induction by
trifluoroethanol: a framework for extrapolating the helix-forming
properties of peptides from trifluoroethanol/water mixtures back
to water. Biochemistry 36, 8413–8421.
29. Verschueren, K.H., Kingma, J., Rozeboom, H.J., Kalk, K.H.,
Jansse, D.B. & Dijkstra, B.W. (1993) Crystallographic and fluor-
escence studies of the interaction of haloalkane dehalogenase with
halide ions: studies with halide compounds reveal a halide binding
site in the active site. Biochemistry 32, 9031–9037.
30. Wang, G., Sparrow, J.T. & Cushley, R.J. (1997) The helix-hinge-
helix structural motif in human apolipoprotein A-I determined by
NMR spectroscopy. Biochemistry 36, 13657–13666.
31. Veglia, G., Zeri, A.C., Ma, C. & Opella, S.J. (2002) Duterium/
hydrogen exchange factors measured by solution nuclear magnetic
resonance spectroscopy as indicators of the structure and topology
of membrane proteins. Biophys. J. 82, 2176–2183.
32. Langner, M. & Kubica, K. (1999) The electrostatics of lipid sur-
faces. Chem. Phys. Lipids 101, 3–35.
33. Victor, K., Jacob, J. & Cafiso, D.S. (1999) Interactions controlling
the membrane binding of basic protein domains: phenylalanine
and the attachment of the myristoylated alanine-rich C-kinase
substrate protein to interfaces. Biochemistry 38, 12527–12536.
34. Nielsen, H., Engelbrecht, J., Brunak, S. & von Heijne, G. (1997)
Identification of prokaryotic and eukaryotic signal peptides and
prediction of their cleavage sites. Protein Engineering 10, 1–6.
35. Lyko, F., Martoglio, B., Jungnickel, B., Rapoport, T.A. & Dob-
berstein, B. (1955) Signal sequence processing in rough micro-

somes. J. Biol. Chem. 270, 19873–19878.
36. Kozak, M. (1987) An analysis of 5¢-noncoding sequences from 699
vertebrate messenger. RNAs. Nucleic Acids Res. 15, 8125–8148.
37. Moffett, D. & Cummings, S. (1994) Transepithelial potential ad
alkalizationinaninsitupreparationoftobaccohornworm
(Manduca sexta)midgut.J. Exp. Biol. 194, 341–345.
38. Gringorten, J.L., Crawford, D.N. & Harvey, W.R. (1993) High
pH in the ectoperitrophic space of the larval lepidopteran midgut.
J. Exp. Biol. 183, 353–359.
39. Clarke,T.E.&Clem,R.J.(2002)Lackofinvolvementofhae-
mocytes in the establishment and spread of infection in Spodoptera
frugiperda larvae infected with the baculovirus Autographa cali-
fornica M nucleopolyhedrovirus by intrahaemocoelic injection.
J. Gen. Virol. 83, 1565–1572.
40. Pluzhnikov, K., Nosyreva, E., Shevchenko, L., Kokoz, Y.,
Schmalz, D., Hucho, F. & Grishin, E. (1999) Analysis of ectato-
minactiononcellmembranes.Eur. J. Biochem. 262, 501–506.
41. Martinez,T.,Burmester,T.,Veenstra,J.A.&Wheeler,D.(2000)
Sequence and evolution of a hexamerin from the ant Camponotus
festinatus. Insect Mol. Biol. 9, 427–431.
42. Orivel, J. & Dejean, A. (2001) Comparative effect of the venoms of
ants of the genus pachycondyla (Hymenoptera ponerinae). Toxicon
39, 195–201.
43. Nolde, D.E., Sobol, A.G., Pluzhnikov, K.A., Grishin, E.V. &
Arseniev, A.S. (1995) Three-dimensional structure of ectatomin
from Ectatomma tuberculatum ant venom. J. Biomol. NMR 5,
1–13.
44. Possani, L.D., Becerril, B., Delepierre, M. & Tytgat, J. (1999)
Scorpion toxins specific for Na
+

-channels. Eur. J. Biochem. 264,
287–300.
45. Wood, A.H. & Granados, R.R. (1992) Genetically engineered
baculoviruses as agents for pest control. Ann. Rev. Microbiol. 45,
69–87.
46. Darbon, H., Zlotkin, E., Kopeyan, C., van Rietschoten, J. &
Rochat, H. (1982) Covalent structure of the insect toxin of the
North African scorpion Androctonus australis Hector. Int. J. Pept.
Protein Res. 20, 320–330.
47. Zlotkin, E., Fishman, Y. & Elazar, M. (2000) AaIT: from neu-
rotoxin to insecticide. Biochimie 82, 869–881.
48. McCutchen, B.F., Choudary, P.V., Crenshaw, R., Maddox, D.,
Kamita, S.G., Palekar, N., Volrath, S., Fowler, E., Hammock,
B.D. & Maeda, S. (1991) Development of a recombinant baculo-
virus expressing an insect-selective neurotoxin: potential for pest
control. Biotechnology (NY) 9, 848–852.
49. Maeda, S., Volrath, S.L., Hanzlik, T.N., Harper, S.A., Majima,
K., Maddox, D.W., Hammock, B.D. & Fowler, E. (1991)
Insecticidal effects of an insect-specific neurotoxin expressed by a
recombinant baculovirus. Virology 184, 777–780.
50. Cory, J.S., Hirst, M.L., Williams, T., Halls, R.S., Goulson, D.,
Green, B.M., Carthy, T.M., Possee, R.D., Cayley, P.J. &
Bishop, D.H.L. (1994) Insecticidal effects of an insect-specific
neurotoxin expressed by a recombinant baculovirus. Nature 370,
138–140.
51. Regev, A., Rivkin, H., Inceoglu, B., Gershburg, E., Hammock,
B.D., Gurevitz, M. & Chejanovsky, N. (2003) Further enhance-
ment of baculovirus insecticidal efficacy with scorpion toxins that
interact cooperatively. FEBS Lett. 537, 106–110.
Ó FEBS 2004 Structure and expression of an ant neurotoxin (Eur. J. Biochem. 271) 2135

Supplementary material
The following material is available from http://blackwell
publishing.com/products/journals/suppmat/EJB/EJB4128/
EJB4128sm.htm
Table S1. Structural statistics and restraint violations for the
ensemble of 10 structures representing solution structure
of poneratoxin. In parentheses is the range of estimated
values.
Fig. S1. NMR derived restraints analyzed in the terms of
range categories (upper) and position in sequence (lower).
Fig. S2. Distribution of the sequential and short range
NMR constraints. The systematic pattern of i,i+3 NOEs
accompanied by the lowered values of
3
J
HaHN
vicinal
coupling constants permitted the assignment of the secon-
dary structure. The helical regions in the peptide sequence
(top) are marked in bold letters.
Fig. S3. The CD spectra of poneratoxin recorded in 35%
TFE solution at pH 5.3 and 7.8. For comparison the
spectrum obtained for 25% TFE, pH 5.5 adopted from
manuscript is presented.
2136 E. Szolajska et al. (Eur. J. Biochem. 271) Ó FEBS 2004