Chemical synthesis and 1H-NMR 3D structure determination of AgTx2-MTX chimera, a new potential blocker for Kv1.2 channel, derived from MTX and AgTx2 scorpion toxins.
ABSTRACT Agitoxin 2 (AgTx2) is a 38-residue scorpion toxin, cross-linked by three disulfide bridges, which acts on voltage-gated K(+) (Kv) channels. Maurotoxin (MTX) is a 34-residue scorpion toxin with an uncommon four-disulfide bridge reticulation, acting on both Ca(2+)-activated and Kv channels. A 39-mer chimeric peptide, named AgTx2-MTX, was designed from the sequence of the two toxins and chemically synthesized. It encompasses residues 1-5 of AgTx2, followed by the complete sequence of MTX. As established by enzyme cleavage, the new AgTx2-MTX molecule displays half-cystine pairings of the type C1-C5, C2-C6, C3-C7, and C4-C8, which is different from that of MTX. The 3D structure of AgTx2-MTX solved by (1)H-NMR, revealed both alpha-helical and beta-sheet structures, consistent with a common alpha/beta scaffold of scorpion toxins. Pharmacological assays of AgTx2-MTX revealed that this new molecule is more potent than both original toxins in blocking rat Kv1.2 channel. Docking simulations, performed with the 3D structure of AgTx2-MTX, confirmed this result and demonstrated the participation of the N-terminal domain of AgTx2 in its increased affinity for Kv1.2 through additional molecular contacts. Altogether, the data indicated that replacement of the N-terminal domain of MTX by the one of AgTx2 in the AgTx2-MTX chimera results in a reorganization of the disulfide bridge arrangement and an increase of affinity to the Kv1.2 channel.
- SourceAvailable from: Jon-Paul Bingham[Show abstract] [Hide abstract]
ABSTRACT: Scorpion toxins have been central to the investigation and understanding of the physiological role of potassium (K+) channels and their expansive function in membrane biophysics. As highly specific probes, toxins have revealed a great deal about channel structure and the correlation between mutations, altered regulation and a number of human pathologies. Radio- and fluorescently-labeled toxin isoforms have contributed to localization studies of channel subtypes in expressing cells, and have been further used in competitive displacement assays for the identification of additional novel ligands for use in research and medicine. Chimeric toxins have been designed from multiple peptide scaffolds to probe channel isoform specificity, while advanced epitope chimerization has aided in the development of novel molecular therapeutics. Peptide backbone cyclization has been utilized to enhance therapeutic efficiency by augmenting serum stability and toxin half-life in vivo as a number of K+-channel isoforms have been identified with essential roles in disease states ranging from HIV, T-cell mediated autoimmune disease and hypertension to various cardiac arrhythmias and Malaria. Bioengineered scorpion toxins have been monumental to the evolution of channel science, and are now serving as templates for the development of invaluable experimental molecular therapeutics.Toxins 01/2012; 4(11):1082-119. · 2.13 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: During the development of selective peptides against highly homologous targets, a reliable tool is sought that can predict information on both mechanisms of binding and relative afﬁnities. These tools must ﬁrst be tested on known proﬁles before application on novel therapeutic candidates. We therefore present a comparative docking protocol in HADDOCK using critical motifs, and use it to "predict" the various selectivity proﬁles of several major αKTX scorpion toxin families versus K(v)1.1, K(v)1.2 and K(v)1.3. By correlating results across toxins of similar proﬁles, a comprehensive set of functional residues can be identiﬁed. Reasonable models of channel-toxin interactions can be then drawn that are consistent with known afﬁnity and mutagenesis. Without biological information on the interaction, HADDOCK reproduces mechanisms underlying the universal binding of αKTX-2 toxins, and K(v)1.3 selectivity of αKTX-3 toxins. The addition of constraints encouraging the critical lysine insertion conﬁrms these ﬁndings, and gives analogous explanations for other families, including models of partial pore-block in αKTX-6. While qualitatively informative, the HADDOCK scoring function is not yet sufﬁcient for accurate afﬁnity-ranking. False minima in low-afﬁnity complexes often resemble true binding in high-afﬁnity complexes, despite steric/conformational penalties apparent from visual inspection. This contamination signiﬁcantly complicates energetic analysis, although it is usually possible to obtain correct ranking via careful interpretation of binding-well characteristics and elimination of false positives. Aside from adaptations to the broader potassium channel family, we suggest that this strategy of comparative docking can be extended to other channels of interest with known structure, especially in cases where a critical motif exists to improve docking effectiveness.Toxins 02/2012; 4(2):110-38. · 2.13 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: High conductance calcium-activated potassium (BK(Ca)) channels can modulate cell excitability and neurotransmitter release at synaptic and afferent terminals. BK(Ca) channels are present in primary afferents of most, if not, all internal organs and are an intriguing target for pharmacological manipulation of visceral sensation. Our laboratory has a long-standing interest in the neurophysiological differences between myelinated and unmyelinated visceral afferent function. Here, we seek to determine whether there is a differential distribution of BK(Ca) channels in myelinated and unmyelinated vagal afferents. Immunocytochemistry studies with double staining for the BK-type K(Ca)1.1 channel protein and isolectin B4 (IB4), a reliable marker of unmyelinated peripheral afferents, reveal a pattern of IB4 labeling that strongly correlates with the expression of the K(Ca)1.1 channel protein. Measures of cell size and immunostaining intensity for K(Ca)1.1 and IB4 cluster into two statistically distinct (P < 0.05) populations of cells. Smaller diameter neurons most often presented with strong IB4 labeling and are presumed to be unmyelinated (n = 1,390) vagal afferents. Larger diameter neurons most often lacked or exhibited a very weak IB4 labeling and are presumed to be myelinated (n = 58) vagal afferents. Complimentary electrophysiological studies reveal that the BK(Ca) channel blockers charybdotoxin (ChTX) and iberiotoxin (IbTX) bring about a comparable elevation in excitability and action potential widening in unmyelinated neurons but had no effect on the excitability of myelinated vagal afferents. This study is the first to demonstrate using combined immunohistochemical and electrophysiological techniques that K(Ca)1.1 channels are uniquely expressed in unmyelinated C-type vagal afferents and do not contribute to the dynamic discharge characteristics of myelinated A-type vagal afferents. This unique functional distribution of BK-type K(Ca) channels may provide an opportunity for afferent selective pharmacological intervention across a wide range of visceral pathophysiologies, particularly those with a reflexogenic etiology and pain.AJP Cell Physiology 02/2011; 300(6):C1393-403. · 3.71 Impact Factor
Chemical synthesis and1H-NMR 3D structure
determination of AgTx2-MTX chimera, a new potential
blocker for Kv1.2 channel, derived from MTX
and AgTx2 scorpion toxins
CYRIL PIMENTEL,1,4SARRAH M’BAREK,2,4VIOLETTAVISAN,3STEPHAN GRISSMER,3
FRANC xOIS SAMPIERI,2JEAN-MARC SABATIER,2HERVE´DARBON,1
AND ZIAD FAJLOUN2,5
1Architecture et fonction des Macromole ´cules Biologiques (AFMB), Centre National de la Recherche Scientifique
UMR6098, Aix-Marseille Universite ´s, Parc Scientifique et Technologique de Luminy Case 932,
13288 Marseille Cedex 09, France
2ERT62, IFR Jean Roche, Faculte ´ de Me ´decine Nord, Cedex 15, France
3Universita ¨t Ulm, 89081 Ulm, Germany
(RECEIVED July 23, 2007; FINAL REVISION September 27, 2007; ACCEPTED October 1, 2007)
Agitoxin 2 (AgTx2) is a 38-residue scorpion toxin, cross-linked by three disulfide bridges, which acts on
voltage-gated K+(Kv) channels. Maurotoxin (MTX) is a 34-residue scorpion toxin with an uncommon
four-disulfide bridge reticulation, acting on both Ca2+-activated and Kv channels. A 39-mer chimeric
peptide, named AgTx2-MTX, was designed from the sequence of the two toxins and chemically synthesized.
It encompasses residues 1–5 of AgTx2, followed by the complete sequence of MTX. As established by
enzyme cleavage, the new AgTx2-MTX molecule displays half-cystine pairings of the type C1–C5, C2–C6,
C3–C7, and C4–C8, which is different from that of MTX. The 3D structure of AgTx2-MTX solved by
1H-NMR, revealed both a-helical and b-sheet structures, consistent with a common a/b scaffold of scorpion
toxins. Pharmacological assays of AgTx2-MTX revealed that this new molecule is more potent than both
original toxins in blocking rat Kv1.2 channel. Docking simulations, performed with the 3D structure of
AgTx2-MTX, confirmed this result and demonstrated the participation of the N-terminal domain of AgTx2
in its increased affinity for Kv1.2 through additional molecular contacts. Altogether, the data indicated that
replacement of the N-terminal domain of MTX by the one of AgTx2 in the AgTx2-MTX chimera results in a
reorganization of the disulfide bridge arrangement and an increase of affinity to the Kv1.2 channel.
Keywords: maurotoxin; agitoxin 2; scorpion toxin; K+channels; synthetic peptide; NMR; solution
structure; molecular docking
4These authors contributed equally to this work.
5Present address: Unite ´ des Rickettsies, Marseille Cedex 5, France.
Reprint requests to: Herve ´ Darbon, AFMB CNRS UMR6098 and
Aix-Marseille Universite ´s, Parc Scientifique et Technologique de
Luminy Case 932, 13288 Marseille Cedex 09, France; e-mail: herve.
email@example.com; fax: 33-491-266-720; or Ziad Fajloun,
Unite ´ des Rickettsies, Faculte ´ de Me ´dicine, 27 Boulevard Jean Moulin,
13385 Marseille Cedex 5, France; e-mail: firstname.lastname@example.org; fax:
Abbreviations: CNS, crystallography and NMR system; DMSO,
Dimethylsulfoxide; HPLC, high pressure liquid chromatography;
HsTx1, Toxin 1 from the scorpion Heterometrus spinnifer; IKCa1
channel, intermediate-conductance Ca2+-activated K+channel; Kv
channel, voltage-gated K+channel; LD50, 50% lethal dose; MALDI-
TOF, matrix-assisted laser desorption ionization-time-of-flight; NMP,
N-methylpyrrolidone; Pi1, Pi4, and Pi7, Toxin 1, 4, and Toxin 7 from
the scorpion Pandinus imperator, respectively; SKCa channel, small-
conductance Ca2+-activated K+channel; TFA, trifluoroacetic acid;
RMSD, root mean square distance; Eelec, electrostatic energy; Evdw,
van Der Waals energy; Eacs, ambiguous chemical-shift energy.
Article published online ahead of print. Article and publication date
are at http://www.proteinscience.org/cgi/doi/10.1110/ps.073122908.
Protein Science (2008), 17:107–118. Published by Cold Spring Harbor Laboratory Press. Copyright ? 2008 The Protein Society
Agitoxin 2 (AgTx2) is a 38-residues toxin that has been
isolated from the venom of leiurus quinquestriatus
hebraeus scorpion (Garcia et al. 1994). This toxin was
shown to reversibly block voltage-gated K+channels
(Shaker B, Kv1.1, Kv1.2, and Kv1.3) (Hidalgo and
MacKinnon 1995; Lipkind and Fozzard 1997; Gao and
Garcia 2003). MTX is another small toxin isolated from
the venom of the scorpion Scorpio maurus palmatus
(Kharrat et al. 1997). It is a basic, C-terminal amidated,
34-mer peptide cross-linked by four disulfide bridges.
The solid-phase technique has been used to obtain
synthetic MTX (sMTX), and it was found that both the
natural and synthetic MTX are equally lethal to mice by
ICV inoculation (LD50of 80 ng/mouse). sMTX has been
shown to be active in the nanomolar range on both
voltage-gated Kv1.2 and IKCa channels (Fajloun et al.
2000a). AgTx2 is cross-linked according to C1–C4, C2–
C5, and C3–C6, while MTX possesses C1–C5, C2–C6,
C3–C4, and C7–C8 disulfide bridge organization (vs.
C1–C5, C2–C6, C3–C7, and C4–C8 for other K+channel-
selective four-disulfide-bridged scorpion toxins) (Bontems
et al. 1991; Darbon et al. 1999). MTX is the sole toxin
that exhibits an uncommon disulfide bridge organization
among the four-disulfide-bridged scorpion toxins (e.g.,
Pi1, Pi4, Pi7, and HsTX1) (Olamendi-Portugal et al.
1996; Rogowski et al. 1996; Lebrun et al. 1997; Savarin
et al. 1999; Fajloun et al. 2000b,c; M’Barek et al. 2003a),
but yet it adopts the classical a/b scaffold (Blanc et al.
1997; M’Barek et al. 2003b). This motif, from which arises
the wide functional diversity of scorpion toxins, is mainly
composed of a short a-helix connected to a b-sheet by two
disulfide bridges. However, some equally Cys-stabilized a/
b-motifs composed of three antiparallel b-strands and one
a-helix are found in insect and plant defensins (Spelbrink
et al. 2004) that acts on the Ca2+channel.
Here, we used MTX as a molecular template to design
and chemically synthesize a scorpion toxin-derived chi-
mera cross-linked by four disulfide bridges, AgTx2-
MTX. This peptide encompasses residues 1–5 of AgTx2
at its N terminus, and the sequence of MTX at its C ter-
minus. The N-terminal portion of AgTx2 was chosen to
form the N-terminal extremity of AgTx2-MTX, because
this domain is structured as being part of the three-
stranded b-sheet that could be observed in the 3D struc-
ture of AgTx2 in solution (Krezel et al. 1995). In this
contribution, we describe the synthesis by an optimized
Fmoc/t-butyl strategy (Merrifield 1986), disulfide bridge
organization, 3D structure in solution by1H-NMR means,
and electrophysiological activity on both the Ca2+-acti-
vated intermediate conductance and voltage-gated K+
channels subtypes (IKCa and Kv1.2, respectively).
Because AgTx2-MTX was found to be highly potent in
blocking the Kv1.2 channel, we detailed the AgTx2-MTX
to the mammalian Kv1.2 channel interaction by computed
docking simulations using the HADDOCK program
(High Ambiguity Driven DOCKing program) in order to
better understand the structural features responsible for
this difference in bioactivity.
Materials and Methods
N-a-Fluoren-9-ylmethyloxycarbonyl (Fmoc)-L-amino acids,
Fmoc-amide resin, and reagents used for chemical syn-
thesis of AgTx2-MTX were purchased from Perkin-Elmer.
Solvents were analytical grade products from SDS (Pey-
pin). Trypsin and chymotrypsin were purchased from
Solid-phase synthesis of AgTx2-MTX
The AgTx2-MTX was synthesized by a solid-phase
method (Merrifield1986) using an automated peptide
synthesizer (Model 433A, Applied Biosystems, Inc.).
The peptide chain was assembled by stepwise synthesis
on 0.3 mmol of Fmoc-amide resin (1% cross-linked; 0.65
mmol of amino group/g) using 1 mmol of Fmoc-amino
acid derivatives (Kharrat et al. 1996). The side-chain
protecting groups of trifunctional residues were: tert-
butyl for Ser, Thr, Tyr, and Asp; trityl (Trt) for Cys and
Asn; pentamethylchroman for Arg; and tert-butyloxycar-
bonyl for Lys. N-a-amino groups were deprotected by
treatments with 18% and 20% (v/v) piperidine/N-meth-
ylpyrrolidone (NMP) for 3 and 8 min, respectively. The
peptide-resin was washed with NMP (5 3 1 min), and
then Fmoc-amino acid derivatives were coupled (20 min)
as their hydroxybenzotriazole active esters in NMP (3.3-fold
excess). After the peptide-chain assembly was completed
and the N-terminal Fmoc group removed, the peptide-
resin (ca. 2.1 g) was treated, under stirring, for 3 h at room
temperature, with a mixture of trifluoroacetic acid (TFA)/
H2O/thioanisole/ethanedithiol (88:5:5:2, v/v) in the pres-
ence of crystalline phenol (2.5 g), in a final volume of
30 mL/g of peptide resin. The peptide mixture was filtered
to remove the resin, and the filtrate was precipitated and
washed twice in cold diethylether. The crude peptide was
then pelleted by centrifugation (3000g; 10 min) and the
supernatant was discarded. The peptide was finally dis-
solved in H2O, freeze-dried, and lyophilized. The crude
peptide was then dissolved in 0.2 M Tris-HCl buffer (pH
8.4) to a final concentration of ;2 mM in peptide,
then gently stirred under air to allow folding/oxidation
(48 h, 25°C). The AgTx2-MTX was purified to homoge-
neity by semipreparative reversed-phase high-pressure
liquid chromatography (HPLC) (Perkin-Elmer, C18 Aqua-
pore ODS 20 mm, 250 3 10 mm), by means of a 60-min
Pimentel et al.
Protein Science, vol. 17
linear gradient from 0% to 35% of buffer B (0.08% [v/v]
TFA/ acetonitrile) in buffer A (0.1% [v/v] TFA/H2O), at a
flow rate of 6 mL/min (l ¼ 230 nm). The identity and high
degree of homogeneity of the chimera (as well as of syn-
thetic toxins), were verified by: (1) analytical C18 reversed-
phase HPLC (Chromolith RP18, 5 mm, 4.6 3 100 mm)
using a 40-min linear gradient from 0% to 60% of buffer B
(0.08% [v/v] TFA/ acetonitrile) in buffer A (0.1% [v/v]
TFA/H2O), at a flow rate of 1 mL/min; (2) amino acid com-
position after hydrolysis (6 N HCl/2% [w/v] phenol, 20 h,
118°C, N2 atmosphere); (3) Edman sequencing; and (4)
molecular mass analysis by matrix-assisted laser desorption
ionization-time-of-flight (MALDI-TOF) mass spectrometry.
Assignment of half-cystine pairings of peptides
The peptides (600 mg) were each added to a mixture of
10% (w/w) of trypsin and chymotrypsin in 0.2 M Tris-
HCl buffer (pH 7.4) (14 h, 37°C). The resulting peptide
fragments were then separated by analytical reversed-
phase HPLC (Chromolith RP18, 5 mm, 4.6 3 100 mm) in
a 60-min linear gradient from 0% to 60% of buffer B
(0.08% [v/v] TFA/acetonitrile) in buffer A (0.1% [v/v]
TFA/H2O), at a flow rate of 1 mL/min (230 nm). The
peptide fragments were also analyzed by mass spectrom-
etry (RP-DE Voyager, Perseptive Biosystems).
Stable transfected mammalian cell lines expressing either
Kv1.2 (Kv1.2) channel were used (Grissmer et al. 1994).
The cell lines were maintained in Dulbecco’s modified
Eagle’s medium containing 4 mM L-glutamine, 1 mM
sodium pyruvate, (GIBCO), 10% (v/v) heat-inactivated fetal
calf serum (PAA). The tsA cell line expressing human
IKCa1 (hIKCa1) channel was a kind gift from Dr. Devor
(University of Pittsburg, Pennsylvania).
All of the experiments were carried out at room
temperature (22°C–25°C) using the whole-cell recording
mode of the patch-clamp technique (Hamill et al. 1981;
Rauer and Grissmer 1996). Cells were bathed with
mammalian Ringer’s solution containing (in millimolars):
160 NaCl, 4.5 KCl, 2 CaCl2, 1 MgCl2, and 10 HEPES (pH
7.4) (with NaOH), with an osmolarity of 290–320 mOsm.
When peptides were applied, 0.1% bovine serum albumin
was added to the Ringer’s solution. A simple syringe-
driven perfusion system was used to exchange the bath
solution in the recording chamber. Electrodes were pulled
from glass capillaries (Science Products) in three stages,
and fire-polished to resistances measured in the bath of
2.5–5 MV. The internal pipette solution used for measur-
ing voltage-gated K+currents contained (in millimolars):
155 KF, 2 MgCl2, 10 HEPES, and 10 EGTA (pH 7.2)
(with KOH), with an osmolarity of 290–320 mOsm. For
measuring K+currents through IKCa1 channels, an
internal pipette solution containing (in millimolars): 135
Kaspartate, 8.7 CaCl2, 2 MgCl2, 10 EGTA, 10 HEPES
(pH 7.2) (with KOH), with an osmolarity of 290–320
mOsm, (free [Ca2+]i ¼ 10?6M), was used. Membrane
currents were measured with an EPC-9 patch-clamp
amplifier (HEKA Elektronik) interfaced to a Macintosh
computer running acquisition and analysis software
(Pulse and PulseFit). Capacitive and leak currents were
subtracted using the P/10 procedure. Series-resistance
compensation (>80%) was used if the current exceeded
2 nA. The holding potential in all experiments was ?80
mV. Data analysis was performed in IgorPro, and Kd
values were deduced by fitting a modified Hill equation
(Itoxin/Icontrol ¼ 1/[1 + ([toxin]/Kd)]), with a Hill
coefficient of 1, to normalized data points obtained at
more than four different peptide concentrations. The
value of each peptide concentration was the mean of at
least three measurements.
A 1 mM sample of synthetic AgTx2-MTX in 0.5 mL of
H2O/D2O (90/10 by vol.) at pH 3.0 was used for NMR
spectra recordings. The amide proton exchange rate was
determined after lyophilization of this sample and dis-
solution in 100% D2O.
All1H spectra were recorded on a BRUKER DRX500
spectrometer equipped with a HCN probe, and self-
shielded triple axis gradients were used. Two-dimensional
NOESY and TOCSY spectra were acquired at 290 K and
300 K in order to solve assignment ambiguities. The
spectra collected at 290 K provided the optimal resolution
of overlapping NMR signals of AgTx2-MTX; therefore,
this temperature was used for further studies of the
protein. Two-dimensional spectra were acquired using
states-TPPI method (Marion et al. 1989) to achieve F1
quadrature detection (Marion and Wu ¨thrich 1983). The
spectral width in both dimensions was 6000 Hz. NOESY
and TOCSY experiments were recorded with 2048 data
points for t2 and 512 points for t1 increments, with 64
transients per experiment. A DQF-COSY experiment was
recorded with 4096 data points in t2 and 1024 data points
in t1. Water suppression was achieved using presaturation
during the relaxation delay (1.3 s), and during the mixing
time in the case of NOESY experiments, or using a
Structure-activity of AgTx2-MTX chimera
Watergate 3-9-19 pulse train (Piotto et al. 1992) using a
gradient at the magic angle obtained by applying simul-
taneous x-, y-, and z-gradients prior to detection. NOESY
spectra were acquired using a mixing time of 80 ms.
TOCSY was performed with a spin-locking field strength
of 8 kHz and spin-lock time of 80 ms. The amide proton
exchange experiments were recorded immediately after
dissolution of the peptides in D2O. A series of NOESY
spectra with a mixing time of 80 ms were recorded at 290
K, the first one for 1 h, followed by spectra of 12 h each.
Amide protons still giving rise to nuclear Overhauser
effect (NOE) correlations after 40 h of exchange were
considered as slowly exchanging and therefore engaged
in a hydrogen bond.
Spectra were processed with the XWIN-NMR version
2.1. The matrices were transformed to a final size of
2.048 points in the acquisition dimension and to 1.024
points in the other, except for a coupling constant
determination for which a 8.192 3 1.024 matrix was
used in the COSY spectrum. The signal was multiplied by
a shifted sine bell window in both dimensions prior to a
Fourier transform, and then a fifth-order polynomial
baseline correction was applied.
Identification of amino acid spin systems and sequen-
tial assignment were achieved using the two-step standard
strategy described by Wu ¨thrich (1986) and applied with
graphical software, XEASY (Bartels et al. 1995). The
comparative analysis of COSY and TOCSY spectra
recorded in water gave the spin system signatures of the
protein. The spin systems were then sequentially con-
nected using the NOESY spectra.
The assignment and integration of nOe data using
manual integration in the XEASY software allowed us
to obtain a list of volumes that were automatically
translated into upper limit distances by the calibration
routine of the ARIA software (Linge et al. 2003). The F
torsion angles constraints resulted from the
coupling constant measurements that measured on the
COSY spectrum. Alternatively, they have been estimated
by the INFIT program (Szyperski et al. 1992). For a given
residue, separated NOESY cross-peaks with the backbone
amide proton in the v2 dimension were used. Several
cross-sections through these cross-peaks were selected
that exhibited a good signal-to-noise ratio. They were
added up, and only those data points of the peak region
that were above the noise level were retained. The left and
right ends of the peak region were then brought to zero
intensity by a linear baseline correction. After extending
the baseline-corrected peak region with zeros on both
sides, which is equivalent to over sampling in the time
domain, an inverse Fourier transformation was per-
formed. The value of the3JHN-Hacoupling constant was
obtained from the first local minimum. The F angles
were restrained to ?120 6 40° for a3JHN-Ha$8 Hz and
to ?65 6 25° for a3JHN-Ha#6 Hz. No angle constraint
was assigned to a
as ambiguous. Determination of the amide proton
exchange rates led us to identify protons involved in
hydrogen bonding. The oxygen partners were then iden-
tified by visual inspection of the preliminary calculated
3JHN-Ha¼ 7 Hz, a value considered
The distance restraints (from measured NOE volumes),
dihedral angles (from
restraints from disulfides bridges, and from hydrogen
bonds were used in structural calculations to determine
toxin conformation. These restraints were introduced as
input in ARIA implemented in CNS 1.1. (Brunger et al.
1998). In the first run, the calculation was initiated using
the NOESY peak list, dihedral angles restraints, and
the assignment of experimentally determined disulfide
bridges. This first run allowed us to correct NOE assign-
ment and gave rise to a preliminary fold that was used to
detect hydrogen-bond carbonyl partners. In the second
run, distance restraints, dihedral angles, hydrogen bonds,
and disulfide bridges were used. We calculated 100
structures in the final iteration and 50 structures were
kept for minimization in water. The 20 best structures were
finally kept as defining the conformation of AgTx2-MTX
chimera. Visual analysis of the final selected structures was
carried out with PyMOL software (DeLano Scientific) and
the geometric quality of the resulting structures was
assessed with PROCHECK 3.4 and PROCHECK-NMR
software (Laskowski et al. 1993). The 3D structure
of AgTx2-MTX was deposited at the Protein Data Bank
(code 2Z3S). The assigned chemical shifts of the chimera
are deposited at the BioMagResBank (BMRB accession
3JHN-Ha coupling constants),
Docking simulations were made with HADDOCK (High
Ambiguity Driven DOCKing) (Dominguez 2003), which
allows one to drive the docking process with previously
obtained experimental data such as mutagenesis or NMR
chemical-shift perturbation. This software is a compilation
of python script based on Aria (Linge et al. 2003) using
CNS (Brunger et al. 1998) to calculate structure with
various types of constraints. The docking takes place in
three steps: a randomization of orientations and rigid body
Pimentel et al.
Protein Science, vol. 17
energy minimization, a semi-rigid simulated annealing in
torsion angle space, and a final refinement in Cartesian
space with explicit solvent. In the first stage, HADDOCK
randomly orientates the two partners and performs a rigid
body energy minimization with rotation and translation of
each molecule. In the second stage, the best solutions
resulting from the preceding energy minimization are then
refined with three steps of simulated annealing refinements.
The first step optimizes the orientation of the partners; the
second optimizes the configuration of side chains at the
interface; and the third permits some conformational rear-
rangements, where both backbone and side chains are
allowed to move. In the third and last stage, the structures
are exposed to the solvent during a steepest descent energy
minimization performed in an 8 A˚shell of TIP3P water
molecules. The resulting structures are then clustered
according to their pairwise backbone RMSD at the inter-
face. These clusters are analyzed and ranked according to
their average interaction energies (sum of Eelec, Evdw, EACS)
and their average buried surface area.
To design a scorpion toxin-derived chimera, we focused on
AgTx2 and MTX, two small scorpion toxins active on Kv
channels and reticulated by three and four disulfide bridges,
respectively. Their amino acid sequences and half-cystine
pairings are shown in Figure 1A. MTX display a ‘‘non-
conventional’’ arrangement of the type C1–C5, C2–C6, C3–
C4, C7–C8 (Fajloun et al. 2000a). Therefore, they appeared
to be good candidates for the production of a chimeric pep-
tide, since they possess variant disulfide bridge organiza-
tions, while they fold according to the same regular a/b
scaffold of scorpion toxins (Bontems et al. 1991; Darbon
Figure 1. Sequence comparison between MTX, AgTx2, and AgTx2-MTX. Chemical synthesis and disulfide bridged organization of
AgTx2-MTX chimera. (A) Amino acid sequences (one-letter code) and half-cystine pairings of MTX and AgTx2. Half-cystine residues
are numbered by order of appearance from the N to the C terminus. The relative positioning of secondary structures (helix and b
strands of the b-sheet structure) is indicated for each peptide. Disulfide bridges are depicted by solid lines. For AgTx2-MTX, the amino
acid sequence of the 39-mer is shown. The amino acid sequence of this chimera derived from AgTx2 and MTX are shaded in dark and
light gray, respectively. (B) AgTx2-MTX at different stages of its chemical synthesis. HPLC profiles of the crude reduced peptide (left),
crude peptide after oxidative folding (middle), and purified AgTx2-MTX (right). (C) Half-cystine pairings of the AgTx2-MTX
chimera. Assignment of the half-cystine pairings was achieved by analysis of the peptides yielded by enzyme cleavage (trypsin and
chymotrypsin) of AgTx2-MTX.
Structure-activity of AgTx2-MTX chimera
et al. 1999). Moreover, AgTx2 and MTX display distinct
profiles of pharmacological activities (high affinity of AgTx2
for Kv1.1, Kv1.3 channels, and high affinity of MTX for
Kv1.2 channel). Since, contrary to MTX, AgTx2 possesses an
extended N-terminal extremity (GVPINVS motif vs. VS motif
of MTX), we added the N-terminal portion (residues 1–5) of
AgTx2 to the N-terminal sequence of MTX to assess the glo-
bal effect of the addition of this motif on MTX conformation
and bioactivity. Figure 1A illustrates the amino acid sequences
of MTX, AgTx2, and AgTx2-MTX chimeric peptide.
Synthesis and physicochemical characterization
Stepwise assembly of AgTx2-MTX was achieved on 0.30
mmol Fmoc-amide resin by means of optimized Fmoc/
t-butyl chemistry (Merrifield 1986). Double-coupling of
Fmoc-amino acids was used for the N-terminal portion of
the chimera (residues 1–7). The overall yield of peptide
chain assembly was 80%. The profiles of elution, by ana-
lytical C18 reversed-phase HPLC, of the crude reduced
peptide after final acidolysis are shown in Figure 1B, left.
The crude peptide was folded/oxidized by air exposure in
Tris/HCl buffer (pH 8.3), for 72 h (Fig. 1B, middle), and
the main oxidized product AgTx2-MTX was finally
purified to >95% homogeneity by semipreparative HPLC
(Fig. 1B, right). Mass spectrometry analysis of AgTx2-
MTX gave an experimental Mr (M+H)+of 4420.2, in
good agreement with the deduced Mr (M+H)+of 4420.5
for this chimera. Amino acid analysis of AgTx2-MTX
after acidolysis provides an amino acid content that
agreed with the calculated values. To determine the
pattern of half-cystine connections, AgTx2-MTX was
cleaved by a mixture of trypsin and chymotrypsin. The
resulting proteolytic fragments were purified to homoge-
neity by HPLC and characterized by means of amino acid
analysis, mass spectrometry, and Edman-sequencing
techniques (data not shown). The half-cystine pairings
were thereby mapped as Cys8–Cys29, Cys14–Cys34,
bridged organization of the MTX portion in AgTx2-
MTX chimera corresponds to a ‘‘conventional’’ disulfide
bridge organization of the type C1–C5, C2–C6, C3–C7,
and C4–C8 (Pi1/Pi4/HsTx1 type) (Fig. 1C).
Pharmacology of AgTx2-MTX on IKCa and
In patch-clamp experiments, we evaluated the effects of
AgTx2, MTX, and AgTx2-MTX on human intermediate
conductance Ca2+-activated K+type 1 channel (hIKCa1)
currents, expressed in tsA cell line (Fig. 2A). The three
peptides fully inhibited hIKCa1 K+currents, with Kd values
of 1152.48 6 156 nM (AgTx2), 2.2 6 0.13 nM (MTX), and
7.36 6 0.62 nM (AgTx2-MTX). We observed that AgTx2 is
thus 523-fold less active than MTX for interacting with
hIKCa1, and AgTx2-MTX is thus 158-fold more active than
AgTx2. This observation indicates that AgTx2 is not a
specific ligand for IKCa1 and its N-terminal extremity does
not affect the fixation of MTX on this channel. We also
compared the effects of AgTx2, MTX, and AgTx2-MTX on
K+currents from rat Kv1.2 channel (Fig. 2B). The Kd value
obtained for AgTx2 is 26.8 6 5.2 nM (Kv1.2). For MTX, it
is 0.51 6 0.04 nM (Kv1.2). The dose/response curves of
current inhibition induced by AgTx2-MTX show a Kd value
of 0.14 6 0.01 nM for rat Kv1.2 channel. It thus appears
that swapping the N-terminal VS motif of MTX by the
GVPINVS motif of AgTx2 results in a 3.6-fold increase in
blockage efficacy of Kv1.2 channel.
Structural properties of AgTx2-MTX
Determination of the 3D solution structure of
For NMR resonance assignment and secondary struc-
tures, the spin systems were identified on the basis of
both COSY and TOCSY spectra. Once the sequential
assignment was achieved, almost all protons were iden-
tified, and their resonance frequencies determined. The
distribution of the Hai/HNi+1, Hbi/HNi+1, and HNi/HNi+1
Figure 2. Pharmacological activity of AgTx2-MTX, AgTx2, and MTX.
(A) Concentration-dependent inhibition curves of IKCa1 currents by
AgTx2-MTX (u), AgTX2 (j), and MTX (d). Fits of the data yield IC50
values of 7.3 6 0.6 nM (AgTx2-MTX), 2.2 6 0.2 nM (MTX), and 1152 6
156 nM (AgTx2) for IKCa1. (B) Concentration-dependent inhibition curves
of Kv1.2 currents by AgTx2-MTX (u), AgTx2 (j), and MTX (d). Fits of
the data yield IC50values of 0.14 6 0.01 nM (AgTx2-MTX), 0.51 6 0.04
nM (MTX), and 26.8 6 5.2 nM (AgTx2) for Kv1.2.
Pimentel et al.
Protein Science, vol. 17
NOE correlation and the coupling constants are presented
in Figure 3A. Characteristic features of secondary struc-
tures such as a-helix and b-sheet have been found in
AgTx2-MTX. The HNi/HNi+1 correlations associated
with hydrogen bonds and small coupling constants clearly
demonstrate the presence of an a-helix, including resi-
dues Ser11to Gln21, while strong Hai/HNi+1correlations,
associated with large coupling constants indicate an
extended region from Asn26to Tyr37. The structure of
AgTx2-MTX was solved by using 516 NOE-based dis-
tance restraints, including 248 intra-residue, 158 sequen-
tial, 40 medium-range, and 70 long-range restraints. The
distribution of these NOE is presented in Fig. 3B. In
addition, included were 19 hydrogen bond restraints,
derived from proton exchange, and 23 dihedral angle
restraints, derived from the measurement of coupling
constants, as well as 12 distance restraints derived from
the disulfide bridges. Altogether, the final experimental
set corresponded to 14.6 constraints per residue on
average. The calculation using the whole set of restraints
and water solvent minimization led to a single family of
20 structures. Structural statistics are given in Table 1. All
of the solutions have good nonbonded contacts and good
covalent geometry, as shown by the low values of CNS
energy terms and low RMSD values for bond lengths,
valence, and improper dihedral angles. Correlation with the
experimental data shows no NOE-derived distance violation
greater than 0.2 A˚. The analysis of the Ramachandran plot
for the whole set of the 20 calculated structures reveals that
78.1% of the residues are in the most favored regions,
21.8% in the additional allowed regions, 0.2% in the
generously allowed regions, and none in the disallowed
regions (data not shown). The convergence of the 20 final
structures (Fig. 4A) gives the three-dimensional structure of
AgTx2-MTX, consisting of a compact disulfide-bonded
core. This structure shows a typical a/b scaffold, in which
an a-helix is connected to a three-stranded b-sheet by two
disulfide bridges. In AgTx2-MTX chimera, the a-helix
(Ser11to Gln21) is connected to the two strands of the b-
sheet (Asn26to Tyr37) by the Cys14–Cys34and Cys18–Cys36
disulfide bridges (Fig. 4B). These two strands are con-
nected by a type II-b turn, like that described for MTX.
Two others bridges stabilize the protein. One connects the
N-terminal strand with the second one of the b-sheet
(Cys8–Cys29bridge) and the other connects the C terminus
with the loop located between the a-helix and the b-sheet
(Cys24–Cys34bridge). The best-fit superimposition of the
backbone traces of the 20 best structures is presented in
Figure 4. The RMSD calculated on the whole structure is
1.43 A˚for the backbone and 2.01 A˚for all heavy atoms.
Figure 3. Statistical properties of AgTx2-MTX structure calculation. (A) Sequence of AgTx2-MTX and sequential assignments. Filled
circles (d) represent3JHN-Hacoupling constants $8 Hz and open circles (s) those #6 Hz. Collected sequential nOe are classified into
strong, medium, and weak nOe, and are indicated by thick, medium, and thin lines, respectively. The last line indicates the secondary
elements (extended regions). (B) nOe (left) and RMSD (right) distribution vs. sequence of AgTx2-MTX. Intraresidue nOe are in black,
sequential nOe in dark gray, medium nOe in light gray, and long-range nOe in white. RMSD values for backbone and all heavy atoms
are in black and gray, respectively.
Structure-activity of AgTx2-MTX chimera
These values fall to 1.03 A˚and 1.84 A˚, respectively, if
region 2–38 is solely considered. Despite the fact that the
disulfide pattern is different for MTX and AgTx2-MTX, we
observed that the fold is conserved for the common 34
amino acid residues. In a similar way, with a different
sequence but an identical disulfide pattern, AgTx2-MTX
adopts a fold very close to AgTx2. Together, these results
suggest that a great number of residues of AgTx2-MTX
have identical positions in MTX and/or AgTx2 (Fig. 5).
Docking of ChTX onto KcsA channel
In this study we first focused on the docking between
charybdotoxin scorpion toxin (ChTX) with the KcsA
potassium channel. Indeed, although docking methods
are meant to propose structural models of bimolecular
protein–protein or protein–ligand complexes, they do not
guarantee that the resulting complexes will be close to the
experimentally obtained solution by crystallography or
NMR. Thus, in order to ensure that the used parameters
are correctly set up, which is crucial for a correct
interpretation of the obtained result, we first tested our
docking methodology on an experimentally determined
structure of protein–protein complex (Yu et al. 2005)
(2A9H). It represents the only toxin–channel complex
determined by direct structural information (NMR con-
straints). First, we made the docking calculation with the
two proteins taken apart from the native complex and
randomly orientated. The experiment was made under
constraint, which results in the positioning of lysine 27 in
the center of the pore. This residue is conserved in all
members of the a-KTX family. Therefore, the only
experimental biological data in the docking computation
was the position of the side chain of the critical lysine 27
as have been described in the literature (Park and Miller
1992b; Goldstein and Miller 1993; Fu et al. 2002) with
mutagenesis or thermodynamic mutant cycle studies. The
obtained solution was very close to the experimental
complex, showing an RMSD at the interface <0.5 A˚for a-
carbons and 1.8 A˚for all heavy atoms. With the aim of
testing the capacity of the software to find the conforma-
tional changes taking place during the formation of the
complex, another docking simulation was performed with
the channel part issued from the complex (PDB code:
2A9H) and the 12 NMR solutions of ChTX taken from
the Protein Data Bank (2CRD). This represented the real
conditions in which we made the following docking
simulations: Again, we found a solution close to the
native complex. The RMSD of the interface for a-carbons
is <1.3 A˚and 2.3 A˚for all heavy atoms. This approach
provides convergent results close to the experimental
complex, which allows us to use it to dock AgTx2-
MTX chimera onto the potassium channel.
Docking of AgTx2, MTX, and AgTx2-MTX onto
We performed the docking of the chimera onto the Kv1.2
channel because, with a Kd value of 0.14 nM, AgTx2-MTX
Table 1. Structural statistics of the 20 best structures
RMSD (A˚)Residues 1–39 Residues 2–38
All heavy atoms
van Der Waals (repel)
Most favored and additional
Generously allowed (%)
Disallowed region (%)
1.43 6 0.36
2.01 6 0.25
1.03 6 0.21
1.84 6 0.23
?923.57 6 46.90
9.82 6 0.53
46.55 6 5.69
168.01 6 28.24
176.27 6 3.02
?72.069 6 17.70
?1252.15 6 32.26
13.737 6 1.436
0.37 6 0.19
0.0042 6 0.0002
0.5473 6 0.0322
1.9947 6 0.1685
40.64 6 0.41
0.0254 6 0.0048
0.8170 6 0.2332
Figure 4. Structural properties of AgTx2-MTX. (A) Stereo pair view of
the best fit of 20 structures of AgTx2-MTX. (B) PyMOL ribbon drawing of
the averaged minimized AgTx2-MTX structure (DeLano Scientific). The
eight half-cysteine residues are numbered according to their positions in
the AgTx2-MTX amino acid sequence.
Pimentel et al.
Protein Science, vol. 17
was 3.6 times more active on Kv1.2 channels than MTX
and 191 times more active than AgTx2. Furthermore, no
other docking has already been made onto the recently
elucidated structure of the mammalian potassium channel
Kv1.2 (Long et al. 2005) (PDB ID: 2A79). For a better
understanding of possible interactions responsible for such a
high activity, we also performed the docking of AgTx2 and
MTX onto the Kv1.2 channel beside that of AgTx2-MTX.
Docking simulations have often been driven according
to constraints derived from biological data issued from
the literature about critical residues involved in the toxin/
channel interactions. The only constraint used in this
work was the one derived from the conserved lysine
responsible for the pore occlusion (Park and Miller
1992b; Goldstein and Miller 1993; Fu et al. 2002)
(Lys27in AgTx2, Lys23in MTX, and Lys28in AgTx2-
MTX). Docking simulations indicate that AgTx2, MTX,
and AgTx2-MTX globally share a similar interaction map
with Kv1.2 (Fig. 6), although some differences could be
observed, which may explain their distinct blocking
efficiency. All of these toxins seem to be stabilized onto
the channel by sharing contacts with both the pore
entryway residues and the external vestibule of the
channel. First, the side chain of the central lysine residue
of the b-sheet makes several hydrogen bonds with the
carbonyl group of pore residues Gly376and Tyr377.
Furthermore, the orientation of either of the three toxins
is similar and consists of an alignment of their b-sheet
with the interface between two dimers of the channel. As
a consequence, this presents residues Asn30of AgTx2,
Asn21of MTX, and Asn26of AgTx2-MTX in front of the
highly conserved Asp363of the channel. Such a residue
contact pattern has been identified earlier for various
toxins of the a-KTX family in experiments of thermody-
namic mutant cycles (Hidalgo and MacKinnon 1995;
Ranganathan et al. 1996; Yu et al. 2005), NOE contacts
and chemical-shift perturbations (Yu et al. 2005), trans-
ferred cross-saturation (Keuchi et al. 2003), chemical-
shift perturbations in solid-state NMR (Lange et al.
2006), and mutagenesis studies (Park and Miller 1992a;
Krezel at al. 1995; Lange et al. 2006). Such an orientation
of toxin molecules also implies that Arg24of AgTx2,
Arg27of MTX, and Arg32of AgTx2-MTX make salt
bridges with the channel opposite Asp363. Another com-
mon feature between these toxins is the presence of an
aromatic residue on the b-sheet (His34of AgTx2, Tyr32of
MTX, and Tyr37of AgTx2-MTX), which makes several
hydrophobic contacts with Gly378and Asp379of the pore
entry and Val381of the adjacent monomer. This aromatic
residue and the central lysine responsible for the pore
occlusion form the so-called functional dyad (Dauplais
et al. 1997; M’Barek et al. 2003a; Pimentel et al. 2006).
These three toxins also have a lysine residue on the
second strand of the b-sheet (Lys32of AgTx2, Lys30of
MTX, and Lys35of AgTx2-MTX), which makes salt
bridges with the Asp355of the extracellular loop, even
if this lysine does not have exactly the same position in
the b-strand. It should be noted that Lys32and Lys35of
AgTx2-MTX correspond to Lys27and Lys30of MTX, two
residues that have been shown to play a pivotal role for
MTX bioactivity, including its recognition of the Kv1.2
Several strong interactions are present in the complexes
MTX/Kv1.2 and AgTx2-MTX/Kv1.2, while they are
lacking in the AgTx2/Kv1.2 complex. This could explain
the increase of 191-fold in affinity of these toxins. One of
the major differences between MTX/AgTx2-MTX and
AgTx2 is the presence of a lysine residue at the beginning
of the a-helix (Lys7of MTX and Lys12of AgTx2-MTX),
which makes a salt bridge with the Asp355of a channel
extracellular loop. Another important difference is the
Figure 5. Structural alignment of AgTx2 with AgTx2-MTX (left) and
MTX with AgTx2-MTX (right). Only amino acids conserved between the
toxins and in interaction with the channel are shown.
Figure 6. Important contacts (strong interaction) between AgTx2-MTX
and Kv1.2 channel. Map detailing major molecular contacts between
AgTx2-MTX and Kv1.2 channel (for clarity, not all the contacts are
shown). The docking of AgTx2-MTX onto the Kv1.2 channel can be
imagined by a 180° vertical rotation of AgTx2-MTX from right to left.
Structure-activity of AgTx2-MTX chimera
presence of a tyrosine residue on the a-helix (Tyr10of
MTX and Tyr15of AgTx2-MTX), which makes hydrogen
bonds with the close Val381. Finally, Asn26of MTX and
Asn31of AgTx2-MTX make several hydrogen bonds with
Asp355and Thr383, while the corresponding residue Met29
in AgTx2 makes only some hydrophobic contacts with
Val381and Trp366. The chimera AgTx2-MTX presents
some features derived from the AgTx2 toxin, such as the
N-terminal part, which makes a third strand of the b-
sheet. This five additional amino acids sequence permits
us to enhance the affinity of the toxin in comparison to its
counterpart MTX by allowing Gly1and Pro3of AgTx2-
MTX to make several hydrogen bonds with Asp355
residues of two adjacent monomers of the channel.
The aim of this study was to evaluate whether it is
possible to obtain novel reticulated peptides with variable
pharmacological profiles with regard to the blockage of
specific ion channel subtypes by using part of the amino
acid sequences of scorpion toxins. Such an approach
should help to unravel the molecular basis of the toxin
recognition of ion channel through the production of
various chimeric compounds, followed by careful inves-
tigation of their structural features and pharmacological
properties. Our study was focused on AgTx2 and MTX
scorpion toxins since, (1) AgTx2 has a conventional
organization of the disulfide bridges associated at an
extended N-terminal extremity, forming a third b?sheet,
which does not exist in the MTX; (2) MTX has a short N-
terminal extremity associated with atypical half-cystine
pairing arrangement. Thus, we were curious to know
which structural and pharmacological impact the grafting
of the N-terminal extremity of AgTx2 onto MTX could
have on the chimera AgTx2-MTX. By the herein
described results, we showed that the replacement of
the N-terminal VS motif of MTX by the N-terminal
GVPINVSdomain of AgTx2
resulted in a change of the MTX disulfide pairing pattern
to the Pi1/Pi4/Pi7/HsTx1-like one. Such a change was
previously observed to be induced by point mutations in
the MTX structure (Fajloun et al. 2000b). Furthermore,
we showed that the transfer of the N-terminal extremity
affects the activity of the molecule on the Kv1.2 channel,
on which the AgTx2-MTX chimera was found 191-fold
more active than AgTx2 and 3.6-fold more active than
MTX, which preserves its primary sequence within
AgTx2-MTX. Such an increase in activity of blocking
the Kv1.2 channel was previously observed with another
chimera obtained by grafting the N-terminal extremity of
Butantoxin (BuTX) to MTX (M’Barek et al. 2005). In the
present work, we observed that AgTx2 was 523-fold less
active than MTX for interacting with the human inter-
Ca1)—this was the first data evaluating AgTx2 activity
regarding the hIKCa1—while AgTx2-MTX was 158-fold
more active than AgTx2. These observations indicate that
AgTx2 is not a specific ligand for IKCa1 and/or its N-
terminal extremity does not affect the fixation of MTX on
From a structural properties point of view, the 3D
structure of AgTx2-MTX solved in solution by1H-NMR
showed that the conformation of AgTx2-MTX is similar
to that of MTX (Blanc et al. 1997). As expected, the main
structural difference between AgTx2-MTX and MTX
relies on the existence of a third N-terminal b-strand,
which, according to our docking simulations, carried out
with 3D NMR structure of AgTx2-MTX and the 3D
crystallography structure of Kv1.2, directly mediates the
recognition of this channel. The 3D structure of AgTx2-
MTX was used in docking simulations with the kv1.2
channel. The binding of the chimera seems to be stabi-
lized over the channel pore by sharing contacts with
residues of both the pore entryway and the extracellular
loop. AgTx2-MTX appears to be stabilized onto the
Kv1.2 channel by a number of close contacts (H-bonds
and electrostatic interactions), in particular, Lys12/Asp355,
Tyr15/Val381, Lys28/Gly376, Asn26/Asp363, Asn31/Asp355,
and Arg32/Asp363(Fig. 6). It should be noted that Lys28of
AgTx2-MTX corresponds to Lys23of MTX, Lys27 of
CTX, Lys22 of ShK, Lys25 of BgK, and Lys42 of KP4, a
common basic residue that was shown to be pivotal for
the scorpion toxin, sea anemone toxin, and plant defensin
bioactivities (Goldstein et al. 1994; Rauer et al. 1999;
Fajloun et al. 2000a,b; Gilquin et al. 2002; Spelbrink
et al. 2004).
Moreover, the docking of AgTx2-MTX over Kv1.2
suggests that new specific residues peculiar to the
structural AgTx2 domain of the chimera (e.g., Gly1and
Pro3) contributed significantly to the chimera/channel
interaction, thereby giving a structural basis of the strong
blockade efficiency of this channel by the chimera. In
light of these results, it seems that the high affinity of the
chimera could result from a combination of, on one hand,
new contacts established by the AgTx2 N-terminal
domain (mainly a stabilization with extracellular loop)
and, on the other hand, specific interactions that take
place with the MTX moiety (mainly close contacts over
the pore region).
conductance Ca2+-activated channel(hIK-
We described herein the chemical synthesis of a four-
disulfide-bridged peptide derived from the structures of
two scorpion toxins (AgTx2 and MTX). The pharmaco-
logical properties (binding affinity and K+current block-
age efficacy) of the AgTx2-MTX chimera were evaluated
Pimentel et al.
Protein Science, vol. 17
on both K+channel subtypes (IKCa and Kv1.2) and
highlight some differences as compared with its natural
counterparts. The implant of the N-terminal domain of
AgTx2 onto MTX produced a chimera that was provided
with a greater affinity for the Kv1.2 channel. Docking
simulations that were achieved with the AgTx2-MTX
NMR structure—determined in this work—and the crys-
tallographic structure of Kv1.2, were performed and
found to be in line with the experimental data. They
suggest that new specific residues from the structural
AgTx2 domain of the chimera (e.g., Gly1and Pro3)
contributed significantly to the chimera/channel recogni-
tion, giving thereby a structural basis of the high-block-
ade efficiency of the chimera toward this channel.
Therefore, we concluded that the replacement of the N-
terminal VS motif of MTX by the GVPINVS domain of
AgTx2 within AgTx2-MTX resulted in a change of the
MTX-like disulfide pairing pattern to the Pi1/Pi4/Pi7/
HsTx1like one. The structural outcome of this approach
is also of interest, since the chimera still folds according
to the canonical a/b scaffold. We conclude that a selec-
tive ‘‘gain of function’’ strategy represents an interesting
way to produce toxin analogues or chimera with novel
We thank N. Andreotti, E. Doria, and X. Morelli for helpful
discussions. This work was supported by funds from the CNRS
and the Universite ´ de la Me ´diterrane ´e.
Bartels, C., Xia, T.H., Billeter, M., Guntert, P., and Wu ¨thrich, K. 1995. The
program XEASY for computer-supported NMR spectral analysis of bio-
logical macromolecules. J. Biomol. NMR 5: 110–118.
Blanc, E., Sabatier, J.M., Kharrat, R., Meunier, S., El Ayeb, M., Van
Reitschoten, J., and Darbon, H. 1997. Solution structure of maurotoxin, a
scorpion toxin from Scorpio maurus, with high affinity for voltage-gated
potassium channels. Proteins 29: 321–333.
Bontems, F., Roumestand, C., Gilquin, B., Me ´nez, A., and Toma, F. 1991.
Refined structure of charybdotoxin: Common motifs in scorpion toxins and
insect defensins. Science 254: 1521–1523.
Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-
Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., et al.
1998. Crystallography & NMR system: A new software suite for macro-
molecular structure determination. Acta Crystallogr. D Biol. Crystallogr.
Darbon, H., Blanc, E., and Sabatier, J.M. 1999. Animal toxins and potas-
sium channels. In Perspectives in drug discovery and design. (eds. H.
Darbon and J.M. Sabatier), Vol. 15, pp. 40–60. Kluwer/Dordrecht, The
Dauplais, M., Lecoq, A., Song, J., Cotton, J., Jamin, N., Gilquin, B.,
Roumestand, C., Vita, C., de Medeiros, C.L., Rowan, E.G., et al. 1997.
On the convergent evolution of animal toxins. Conservation of a diad of
functional residues in potassium channel-blocking toxins with unrelated
structures. J. Biol. Chem. 272: 4302–4309.
Dominguez, C. 2003. HADDOCK: A protein–protein docking approach based
on biochemical and/or biophysical information. J. Am. Chem. Soc. 125:
Fajloun, Z., Ferrat, G., Carlier, E., Fathallah, M., Lecomte, C., Sandoz, G., di
Luccio, E., Mabrouk, K., Legros, C., Darbon, H., et al. 2000a. Synthesis,
1H-NMR structure, and activity of a three-disulfide-bridged maurotoxin
analog designed to restore the consensus motif of scorpion toxins. J. Biol.
Chem. 275: 13605–13612.
Fajloun, Z., Carlier, E., Lecomte, C., Geib, S., di Luccio, E., Bichet, D.,
Mabrouk, K., Rochat, H., De Waard, M., and Sabatier, J.M. 2000b. Chemical
synthesis and characterization of Pi1, a scorpion toxin from Pandinus
imperator active on K+channels. Eur. J. Biochem. 267: 5149–5155.
Fajloun, Z., Mosbah, A., Carlier, E., Mansuelle, P., Sandoz, G., Fathallah, M., di
Luccio, E., Devaux, C., Rochat, H., De Waard, M., et al. 2000c. Maurotoxin
versus Pi1/HsTx1 scorpion toxins. Toward new insights in the under-
standing of their distinct disulfide bridge patterns. J. Biol. Chem. 15:
Fu, W., Cui, M., Briggs, J.M., Huang, X., Xiong, B., Zhang, Y., Luo, X.,
Shen, J., Ji, R., Jiang, H., et al. 2002. Brownian dynamics simulations of the
recognition of the scorpion toxin maurotoxin with the voltage-gated
potassium ion channels. Biophys. J. 83: 2370–2385.
Gao, Y.D. and Garcia, M.L. 2003. Interaction of agitoxin2, charybdotoxin, and
iberiotoxin with potassium channels: Selectivity between voltage-gated and
Maxi-K channels. Proteins 52: 146–154.
Garcia, M.L., Garcia-Calvo, M., Hidalgo, P., Lee, A., and MacKinnon, R. 1994.
Purification and characterization of three inhibitors of voltage-dependent
K+channels from Leiurus quinquestriatus var. hebraeus venom. Biochem-
istry 33: 6834–6839.
Gilquin, B., Racape, J., Wrisch, A., Visan, V., Lecoq, A., Grissmer, S.,
Menez, A., and Gasparini, S. 2002. Structure of the BgK-Kv1.1 complex
based on distance restraints identified by double mutant cycles. Molecular
basis for convergent evolution of Kv1 channel blockers. J. Biol. Chem. 277:
Goldstein, S.A. and Miller, C. 1993. Mechanism of charybdotoxin block of a
voltage-gated K+channel. Biophys. J. 65: 1613–1619.
Goldstein, S.A., Pheasant, D.J., and Miller, C. 1994. The charybdotoxin
receptor of a Shaker K+channel: Peptide and channel residues mediating
molecular recognition. Neuron 12: 1377–1388.
Grissmer, S., Nguyen, A.N., Aiyar, J., Hanson, D.C., Mather, R.J.,
Gutman, G.A., Karmilowicz, M.J., Auperin, D.D., and Chandy, K.G.
1994. Pharmacological characterization of five cloned voltage-gated K+
channels, types Kv1.1, 1.2, 1.3, 1.5, and 3.1, stably expressed in
mammalian cell lines. Mol. Pharmacol. 45: 1227–1234.
Hamill, O.P., Marty, A., Neher, E., Sakmann, B., and Sigworth, F.J. 1981.
Improved patch-clamp techniques for high-resolution current recording
from cells and cell-free membrane patches. Pflugers Arch. 391: 85–100.
Hidalgo, P. and MacKinnon, R. 1995. Revealing the architecture of a K+
channel pore through mutant cycles with a peptide inhibitor. Science 268:
Keuchi, K., Yokogawa, M., Matsuda, T., Sugai, M., Kawano, S., Kohno, T.,
Nakamura, H., Takahashi, H., and Shimada, I. 2003. Structural basis of the
KcsA (K+) channel and agitoxin2 pore-blocking toxin interaction by using
the transferred cross-saturation method. Structure 11: 1381–1392.
Kharrat, R., Mabrouk, K., Crest, M., Darbon, H., Oughideni, R., Martin-
Eauclaire, M.F., Jacquet, G., El Ayeb, M., Van Rietschoten, J., Rochat, H.,
et al. 1996. Chemical synthesis and characterization of maurotoxin, a short
scorpion toxin with four disulfide bridges that acts on K+channels. Eur. J.
Biochem. 242: 491–498.
Kharrat, R., Mansuelle, P., Sampieri, F., Crest, M., Oughideni, R., Van
Rietschoten, J., Martin-Eauclaire, M.-F., Rochat, H., and El Ayeb, M.
1997. Maurotoxin, a four disulfide bridge toxin from Scorpio maurus
venom: Purification, structure and action on potassium channels. FEBS
Lett. 406: 284–290.
Krezel, A.M., Kasibhatla, C., Hidalgo, P., MacKinnon, R., and Wagner, G.
1995. Solution structure of the potassium channel inhibitor agitoxin2:
Caliper for probing channel geometry. Protein Sci. 4: 1478–1489.
Lange, A., Giller, K., Hornig, S., Martin-Eauclaire, M.F., Pongs, O., Becker, S.,
and Baldus, M. 2006. Toxin-induced conformational changes in a potas-
sium channel revealed by solid-state NMR. Nature 13: 959–962.
Laskowski, R.A., MacArthur, M.W., Moss, D.M., and Thornton, J.M. 1993. A
program to check the stereochemical quality of protein structures. J. Appl.
Crystallogr. 26: 283–291.
Ishiguro, M., Oyama, Y., Pongs, O., and Nakajima, T. 1997. A four-
disulphide-bridged toxin, with high affinity towards voltage-gated K+
channels, isolated from Heterometrus spinnifer (Scorpionidae) venom.
Biochem. J. 328: 321–327.
Linge, J.P., Habeck, M., Rieping, W., and Nilges, M. 2003. ARIA: Automated
nOe assignment and NMR structure calculation. Bioinformatics 19: 315–316.
Lipkind, G.M. and Fozzard, H.A. 1997. A model of scorpion toxin binding to
voltage-gated K+channels. J. Membr. Biol. 158: 187–196.
Structure-activity of AgTx2-MTX chimera
Long, S.B., Campbell, E.B., and Mackinnon, R. 2005. Crystal structure of a
mammalian voltage-dependent Shaker family K+channel. Science 5: 897–
Marion, D. and Wu ¨thrich, K. 1983. Application of phase sensitive two-
dimensional correlated spectroscopy (COSY) for measurements of1H-1H
spin-spin coupling constants in proteins. Biochem. Biophys. Res. Commun.
Marion, D., Ikuto, M., Tschudin, R., and Bax, A. 1989. Rapid recording of 2D
NMR spectra without phase cycling. Application to the study of hydrogen
exchange in proteins. J. Mag. Res. 85: 393–399.
M’Barek, S., Lopez-Gonzalez, I., Andreotti, N., Di Luccio, E., Visan, V.,
Grissmer, S., Judge, S., El Ayeb, M., Darbon, H., Rochat, H., et al. 2003a. A
maurotoxin with constrained standard disulfide bridging: Innovative strat-
egy of chemical synthesis, pharmacology, and docking on K+channels. J.
Biol. Chem. 15: 31095–31104.
M’Barek, S., Mosbah, A., Sandoz, G., Fajloun, Z., Olamendi-Portugal, T.,
Rochat, H., Sampieri, F., Guijarro, I., Mansuelle, P., Delepierre, M., et al.
2003b. Synthesis and characterization of Pi4, a scorpion toxin from
Pandinus imperator that acts on K+channels. Eur. J. Biochem. 270:
M’Barek, S., Chagot, B., Andreotti, N., Visan, V., Mansuelle, P., Grissmer, S.,
Marrakchi, M., El Ayeb, M., Sampieri, F., Darbon, H., et al. 2005.
Increasing the molecular contacts between maurotoxin and Kv1.2 channel
augments ligand affinity. Proteins 15: 401–411.
Merrifield, R.B. 1986. Solid phase synthesis. Science 232: 341–347.
Olamendi-Portugal, T., Gomez-Lagunas, F., Gurrola, G.B., and Possani, L.D.
1996. A novel structural class of K+-channel blocking toxin from the
scorpion Pandinus imperator. Biochem. J. 315: 977–981.
Park, C.S. and Miller, C. 1992a. Mapping function to structure in a channel-
blocking peptide: Electrostatic mutants of charybdotoxin. Biochemistry 1:
Park, C.S. and Miller, C. 1992b. Interaction of charybdotoxin with permeant
ions inside the pore of a K+channel. Neuron 9: 307–313.
Pimentel, C., Choi, S.J., Chagot, B., Guette, C., Camadro, J.M., and Darbon, H.
2006. Solution structure PcFK1, a spider peptide active against Plasmodium
falciparum. Protein Sci. 3: 628–634.
Piotto, M., Saudek, V., and Sklenar, V. 1992. Gradient-tailored excitation for
single-quantum NMR spectroscopy of aqueous solutions. J. Biomol. NMR
Ranganathan, R., Lewis, J.H., and MacKinnon, R. 1996. Spatial localization of
the K+channel selectivity filter by mutant cycle-based structure analysis.
Neuron 16: 131–139.
Rauer, H. and Grissmer, S. 1996. Evidence for an internal phenylalkylamine
action on the voltage-gated potassium channel Kv1.3. Mol. Pharmacol. 50:
Rauer, H., Pennington, M., Cahalan, M., and Chandy, K.G. 1999. Structural
conservation of the pores of calcium-activated and voltage-gated potassium
channels determined by a sea anemone toxin. J. Biol. Chem. 274: 21885–
Rogowski, R.S., Collins, J.H., O’Neill, T.J., Gustafson, T.A., Werkman, T.R.,
Rogawski, M.A., Tenenholz, T.C., Weber, D.J., and Blaustein, M.P.
1996. Three new toxins from the scorpion Pandinus imperator selec-
tively block certain voltage-gated K+channels. Mol. Pharmacol. 50: 1167–
Savarin, P., Romi-Lebrun, R., Zinn-Justin, S., Lebrun, B., Nakajima, T.,
Gilquin, B., and Menez, A. 1999. Structural and functional consequences
of the presence of a fourth disulfide bridge in the scorpion short toxins:
Solution structure of the potassium channel inhibitor HsTX1. Protein Sci.
Spelbrink, R.G., Dilmac, N., Allen, A., Smith, T.J., Shah, D.M., and
Hockerman, G.H. 2004. Differential antifungal and calcium channel-
blocking activity among structurally related plant defensins. Plant Physiol.
Szyperski, T., Guntert, P., Otting, G., and Wu ¨thrich, K. 1992. Determination of
scalar coupling constants by inverse Fourier transformation of in-phase
multiples. J. Magn. Reson. Imaging 99: 552–560.
Yu, L., Sun, C., Song, D., Shen, J., Xu, N., Gunasekera, A., Hajduk, P.J., and
Olejniczak, E.T. 2005. Nuclear magnetic resonance structural studies of a
potassium channel-charybdotoxin complex. Biochemistry 44: 15834–
Wu ¨thrich, K. 1986. NMR of proteins and nucleic acids, pp. 2–199. John Wiley
& Sons, New York.
Pimentel et al.
Protein Science, vol. 17