ARTICLE IN PRESS

0041-0101/$ - see

doi:10.1016/j.tox

Abbreviations

performance liq

thiogalactopyra

ionization; MS,

reaction; PQM,

sodium dodecyl

tetraethylammo

flight.$Ethical stat

potential conflic�Correspondi

fax: +7495 330

E-mail addre

Toxicon 50 (2007) 993–1004

www.elsevier.com/locate/toxicon

o-Lsp-IA, a novel modulator of P-type Ca2+ channels$

Kirill Pluzhnikova, Alexander Vassilevskia,�, Yuliya Korolkovaa,Alexander Fisyunovb, Olena Iegorovab, Oleg Krishtalb, Eugene Grishina

aShemyakin and Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences,

Miklukho-Maklaya, 16/10, 117997 Moscow, Russian FederationbBogomoletz Institute of Physiology, 4 Bogomoletz Street, Kyiv 01024, Ukraine

Received 13 May 2007; received in revised form 16 July 2007; accepted 17 July 2007

Available online 28 July 2007

Abstract

A novel polypeptide, designated o-Lsp-IA, which modulates P-type Ca2+ channels, was purified from the venom of the

spider Geolycosa sp. o-Lsp-IA contains 47 amino acid residues and 4 intramolecular disulfide bridges. It belongs to a

group of spider toxins affecting Ca2+ channels and presumably forms the inhibitor cystine knot (ICK) fold. Peculiar

structural features (a cluster of positively charged residues in the C-terminal loop of the peptide and a regular distribution

of hydrophobic residues) that may play a decisive role in the o-Lsp-IA mechanism of action were located. Recombinant

o-Lsp-IA was produced in prokaryotic expression system and was shown to be structurally and functionally identical to

the native toxin. At saturating concentration (10 nM), the peptide clearly slows down the activation kinetics and partially

inhibits the amplitude of P-current in rat cerebellar Purkinje neurons. Prominent deceleration of the activation kinetics

is manifested as the appearance of a five-fold slower component of the current activation. The specificity of action of

o-Lsp-IA on different Ca2+ channel types was studied in isolated hippocampal neurons of rat. o-Agatoxin IVA

completely removed the effect of o-Lsp-IA on the whole-cell Ca2+ current. Therefore, o-Lsp-IA appears to act specifically

on P-type Ca2+ channels.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Spider toxin; Calcium channels; Inhibitor cystine knot; Channel activation kinetics; Expressed sequence tag; Recombinant

protein

front matter r 2007 Elsevier Ltd. All rights reserved

icon.2007.07.004

: EST, expressed sequence tag; HPLC, high-

uid chromatography; IPTG, isopropyl b-D-1-noside; MALDI, matrix-assisted laser desorption

mass spectrometry; PCR, polymerase chain

processing quadruplet motif; SDS-PAGE,

sulfate polyacrylamide gel electrophoresis; TEA,

nium; TFA, trifluoroacetic acid; TOF, time-of-

ement: We, the authors, certify that there are no

ts of interest to disclose.

ng author. Tel.: +7495 336 6540;

7301.

ss: avas@ibch.ru (A. Vassilevski).

1. Introduction

Voltage-gated Ca2+ channels are ubiquitous toall electro-excitable cells. These channels have theunique property to transform the electrical activityof the cell into an intracellular chemical signal andthereby regulate the vital processes of musclecontraction, hormone secretion and synaptic trans-mission (Hille, 2001). In the latter process, voltage-gated Ca2+ channels of N- and P/Q-types play acentral role. These channels appear to dominate the

.

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004994

pre-synaptic terminals of central and peripheralneurons. Therefore, modulation of N- and P/Q-typechannels allows affecting a number of signalingpathways.

Pain sensation has attracted considerable interestin the past few years. Selective inhibition of painsignals into the central nervous system is thought tobe efficient in terms of producing the analgesic effect(Julius and Basbaum, 2001). Afferent signals carry-ing pain (nociceptive signals) can be selectivelyblocked on the level of synaptic transmission. Thisapproach was proven efficacious with a selectiveblocker of N-type Ca2+ channels approved by theFDA as a novel drug for treatment of pain in 2004(Miljanich, 2004). Impaired synaptic transmissioncaused by specific mutations of the P/Q-type Ca2+

channels results in generation of migraine headacheand ataxia (Cao and Tsien, 2005; Jeng et al., 2006).Novel specific modulators of Ca2+ channels areneeded to unravel the molecular mechanismsinvolved. They will help to develop new therapeuticsfor the treatment of pain and other pathologies. Todate, the only known potent specific inhibitors ofP/Q-type Ca2+ channels are o-agatoxins IVA andIVB, which act through alteration of channel gating(Mintz et al., 1992; Adams et al., 1993; Adams,2004). In this paper, we describe o-Lsp-IA, a newpolypeptide toxin isolated from spider venom. Atnanomolar concentrations, it appears to act speci-fically on P-type channels; the mechanism of o-Lsp-IA action significantly differs from those describedfor o-agatoxins IVA and IVB as well as other Ca2+

channel inhibitors.

2. Materials and methods

2.1. o-Lsp-IA toxin isolation

Crude venom of the spider Geolycosa sp.1

(Araneae, Lycosidae) was obtained from FaunaLaboratories, Ltd. (Almaty, Republic of Kazakh-stan). Pure polypeptide toxin o-Lsp-IA2 wasisolated by size exclusion and reverse-phase chro-matography as described previously (Fisyunovet al., 2005).

1This novel spider species was previously referred to as Lycosa

sp. (Fisyunov et al., 2005). SPP Code—A267TDLS2-KZARNA.

Classification provided by Andrey Feodorov from Fauna

Laboratories, Ltd. (Almaty, Republic of Kazakhstan).2o-Lsp-IA was previously referred to as Lsp-1, before the

primary structure was established (Fisyunov et al., 2005).

2.2. Reduction of disulfide bonds and modification of

thiol groups

Two nanomoles of purified o-Lsp-IA weredissolved in 0.1ml of 0.2M Tris–HCl (pH 8.0),7.5M guanidine hydrochloride and 2mM EDTA.The sample was degassed with nitrogen andincubated at 60 1C for 1 h. A 150-fold molar excessof dithiothreitol in the same solution buffer wasadded. The reduction reaction was performed atroom temperature for 18 h. Alkylation was per-formed by adding 4-vinylpyridine (a three-foldexcess with respect to dithiothreitol) in 10 ml ofmethanol. The modification reaction was carriedout at room temperature for 15min in the dark. Themodified polypeptide was separated by reverse-phase high-performance liquid chromatography(HPLC) on a Jupiter C5 column (2� 150mm,5 mm; Phenomenex) using a 25min linear gradientof acetonitrile (15–50%) in 0.1% (v/v) aqueoustrifluoroacetic acid (TFA) at a flow rate of 0.3ml/min. Eluate absorbance was monitored at 280 nm.

2.3. Polypeptide sequencing

The N-terminal amino acid sequence of theS-pyridylethylated o-Lsp-IA was determined ona Procise Model 492 protein/peptide sequencer(Applied Biosystems) according to the manufac-turer’s protocol.

2.4. Hydrolysis of o-Lsp-IA by endoproteinase

Asp-N

After the determination of the N-terminal aminoacid sequence, S-pyridylethylated o-Lsp-IA (1 nmol)was hydrolyzed with endoproteinase Asp-N (Sigma)at an enzyme–protein ratio 1:40 (w/w) in 50mMsodium phosphate buffer (pH 8.0) at 37 1C for 6 h.Peptide fragments were fractionated by reverse-phase HPLC on a Jupiter C5 column (2� 150mm,5 mm; Phenomenex) using a 60min linear gradient ofacetonitrile (0–60%) in 0.1% (v/v) aqueous TFA ata flow rate of 0.3ml/min. Eluate absorbance wasmonitored at 210 nm. The fragments were subjectedto automated sequencing.

2.5. Mass spectrometry (MS)

Peptides were analyzed by matrix-assisted laserdesorption ionization time-of-flight (MALDI-TOF)mass spectrometry (MS). M@LDI LR (Micromass

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004 995

UK Ltd.) and Ultraflex TOF-TOF (Bruker Dalto-nik GmbH) instruments were used. Calibration wasperformed using either a ProteoMass peptide andprotein MALDI-MS calibration kit (mass rangeof 700–66,000Da) or a ProteoMass peptide MAL-DI-MS calibration kit (mass range of 700–3500Da)(both from Sigma). Molecular masses were deter-mined in a linear positive ion mode (average mole-cular masses were determined) using samplesprepared by the dried droplet method with a2,5-dihydroxy benzoic acid (10mg/ml in 70%acetonitrile with 0.1% TFA) or a-cyano-4-hydro-xycinnamic acid (10mg/ml in 50% acetonitrile with0.1% TFA) matrix.

2.6. o-Lsp-IA gene synthesis

The DNA encoding o-Lsp-IA was constructedfrom synthetic oligonucleotides using the polymer-ase chain reaction (PCR) method. The PCRfragment was finally amplified using a forwardprimer E1, 50-C TCC TTA GAT CTG GAC GAC

GAC GAC AAG GAA AAG TCC TGT ATC ACTTG-30, containing a Bgl II restriction enzyme site(underlined), codons encoding an enterokinasecleavage site (italics) and six N-terminal residuesof o-Lsp-IA, and a reverse primer E2, 50-CG GGATCC TTA TTC CAC AGA CGA CAC TGG CA-30, containing a Bam H1 restriction site (underlined)and a stop codon (italics), and encoding six C-terminal residues of o-Lsp-IA. The PCR-fragmentencoding mature o-Lsp-IA was gel purified, di-gested with Bgl II/Bam H1 and cloned into theexpression vector pET-32b (Novagen). Clones werescreened for the presence of the inserts by PCR. Theresulting constructs were checked by sequencingand used to transform Escherichia coli Origami Bcells for protein production.

2.7. Production and purification of recombinant o-

Lsp-IA

Recombinant o-Lsp-IA was produced as a fusionprotein with thioredoxin. E. coli Origami B cellsharboring the expression vector were cultured at37 1C in LB medium containing 70 mg/ml ampicillin,15 mg/ml kanamycin and 12.5 mg/ml tetracycline upto reaching the culture density of OD600 �0.4–0.6.Expression was induced by 0.1mM isopropyl b-D-1-thiogalactopyranoside (IPTG). The cells werecultured at 25 1C for 12–14 h, harvested, resuspen-ded in the start buffer for affinity chromatography

(300mM NaCl, 50mM sodium phosphate buffer,pH 8.0) and incubated for 30min with 0.2mg/ml oflysozyme, followed by ultrasonication. Next, themixture was centrifuged for 15min at 15,000 rpm toremove any insoluble particles. The supernatant wasapplied to a TALON Superflow Metal AffinityResin (Clontech), and the fusion protein waspurified according to the protocol supplied by themanufacturer. Purity of the hybrid protein waschecked by sodium dodecyl sulfate polyacrylamidegel electrophoresis (SDS-PAGE).

The hybrid protein was quickly desalted on aJupiter C4 semi-preparative column (250� 10mm,5 mm; Phenomenex) using a step gradient ofacetonitrile (0–80%) in 0.1% TFA. The proteinwas then dried on a vacuum concentrator andredissolved in 20mM Tris–Cl (pH 9.0); proteinconcentration was determined by monitoring ab-sorbance at 280 nm using a calculated molarextinction coefficient. Fusion protein cleavage byhuman enterokinase catalytic subunit (Gasparian etal., 2003) was performed (one unit per 1mg offusion protein, 18 h at room temperature), andrecombinant o-Lsp-IA was separated by means ofreverse-phase HPLC. A Jupiter C5 column(4.6� 150mm, 5 mm; Phenomenex) was used.o-Lsp-IA was eluted with a linear gradient ofacetonitrile (0–60% for 60min) in 0.1% aqueousTFA at a flow rate of 1ml/min. Eluate absorbancewas monitored at 280 nm.

The purity of the recombinant o-Lsp-IA waschecked by MALDI-TOF MS as well as byN-terminal sequencing and analytical reverse-phaseHPLC on a Jupiter C5 column (2� 150mm, 5 mm;Phenomenex) using a linear gradient of acetonitrile(20–60% for 60min) in 0.1% aqueous TFA at aflow rate of 0.3ml/min. The identity of recombinantand native o-Lsp-IA was proven by co-injectingsamples (0.5 nmol) onto the same column andvisualizing a single peak on the chromatogram.

2.8. Isolation of cerebellar Purkinje and hippocampal

CA1– CA3 neurons

Cerebellum or hippocampus was dissected, res-pectively, from 9–10 or 14–15-day-old Wistar rats(15–20 g; WAG/GSto) in the saline of the followingcomposition: 5mM KCl, 120mM NaCl, 26mMNaHCO3, 1mM MgCl2 and 22mM glucose(pH 7.4). The tissues were cut into 200–400 mmthick slices. Then the slices were incubated for30min in Ca2+-free Dulbecco’s phosphate-buffered

ARTICLE IN PRESS

3The protein sequence data reported in this paper has been

submitted to the UniProt databank with the accession number

P85079. The nucleotide sequences data reported in this paper

have been submitted to the GenBank database with the accession

numbers EF187331–EF187337.

K. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004996

saline with addition of 2mM MgCl2, 26mMNaHCO3, 30mM glucose and 1mM CaCl2 forhippocampal slices (pH 7.4). The solutions werecontinuously saturated with a 95% O2 and 5%CO2 gas mixture. Then the slices were transferred tothe same solution with addition of 2.4mg/mlprotease XXIII from Aspergillus oryzae and incu-bated for 35min at 23 1C. After enzyme treatmentthe slices were rinsed in the same solutions with-out the protease. Dissociation of the cells wasperformed in normal saline of the followingcomposition: 150mM NaCl, 5mM KCl, 1mMMgCl2, 2mM CaCl2, 20mM HEPES and 10mMglucose (pH 7.4). Single cells were isolated bysuccessive trituration of the slices through severalfire-polished pipettes with opening diameters of0.1–0.3mm.

2.9. Electrophysiology

Currents through voltage-gated Ca2+ channelswere recorded at room temperature (20–22 1C) inthe whole-cell configuration of the patch clamptechnique using an A-M Systems 2400 patch-clampamplifier (Bio-Medical Products) connected to a100 kHz Lab Master DMA board (Scientific Solu-tions) in an IBM PC. Patch pipettes were pulledfrom borosilicate glass tubes (Sutter InstrumentCo.) on a P97 Flaming/Brown micropipette puller(Sutter Instrument Co.). Pipettes had resistances of2–4MO when filled with the intracellular solutioncontaining 70mM Tris–phosphate, 40mM tetra-ethylammonium chloride (TEA-Cl), 5mM MgCl2,20mM Tris–HCl, 5mM EGTA, 5mM ATP and0.5mM GTP, adjusted to pH 7.3 with Tris–OH.Liquid junction potentials were compensated. Aftermembrane rupture, series resistance (8–12MO) wascompensated (70–90%). Only the cells with negli-gible leaks (o50 pA) were used, so the currentrecords were not leak subtracted.

The high-voltage activated (HVA) P-type Ca2+

current in Purkinje cells was measured at a holdingpotential of �70mV to ensure complete inactivationof the low-voltage activated (LVA) T-type Ca2+

current (Panchenko et al., 1993; Regan, 1991).Under these experimental conditions the whole-cellcurrent was completely blocked by o-agatoxin IVA(Mintz et al., 1992; McDonough et al., 1997a). Toexclude Ca2+-dependent current inactivation pro-cesses, Ba2+ was used as a charge carrier throughCa2+ channels. The external control solutioncontained 2mM BaCl2, 2mM MgCl2, 20mM

TEA-Cl, 100mM choline-Cl and 20mM Tris–HCl(pH 7.4).

Whole-cell Ca2+ channel currents in hippocam-pal pyramidal neurons from the CA3–CA1 regionwere measured at a holding potential of �100mV.Under these experimental conditions measurable T-,L-, R-, N- and P/Q-type Ca2+ currents wereactivated (Avery and Johnston, 1996; Ishibashi etal., 1995). The external control solution contained10mM BaCl2, 2mM MgCl2, 20mM TEA-Cl,100mM choline-Cl and 20mM Tris–HCl (pH 7.4).

The drug-containing solutions were applied by‘‘concentration clamp’’ technique using a ‘‘jumpingtable’’ set-up (Pharma Robot). The currents weredigitized every 140 ms and filtered at 3 kHz. Datawere analyzed using ‘‘jumping table’’ software(Pharma Robot) running on an IBM PC.

2.10. Electrophysiological data analysis

All curve fitting and statistics were done withMicrocal Origin software. The amplitude of cur-rents was measured from the baseline to the peakvalue. The inhibitory action of the substance wasmeasured as the mean ratio I/Icon, where I is thepeak amplitude of the current under the action ofthe substance and Icon is the peak amplitude of thecurrent in control saline at test stimulus correspond-ing to the maximum of the I/V curves. Theactivation kinetics of the current was fitted by adouble-exponential function:

I ¼ I fast expð�t=tfastÞ þ I slow expð�t=tslowÞ; x (1)

where Ifast and Islow were the current amplitudeswith fast and slow activation kinetics, while tfast andtslow were the fast and slow activation timeconstants. The effect of the substance on theinvestigated current was averaged for at least fourcells. Cumulative data were calculated as mean7SD(number of experiments) throughout the study.

3. Results

3.1. Structure determination3

The purified active peptide component from thevenom of the wolf spider Geolycosa sp. (see footnote 1)

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004 997

(Araneae, Lycosidae) that showed modulatoryeffects on P-type Ca2+ currents in rat cerebellarPurkinje neurons (Fisyunov et al., 2005) wasnamed o-Lsp-IA (see footnote 2) (measured averagemolecular mass of 5623.7Da). Alkylation ofnon-reduced o-Lsp-IA revealed no free thiolgroups in the native peptide as indicated byMS. Direct peptide sequencing provided 27N-terminal amino acid residues of o-Lsp-IA:1EKS ITWRNS MHNDKG FPWS V W27.Prior to Edman degradation, the disulfide bondswere reduced with dithiothreitol and thiol groupswere modified with 4-vinylpyridine to allow un-ambiguous determination of cysteine residues.Among the first 27 amino acid residues ofo-Lsp-IA only one aspartic acid residue in position15 was observed. For this reason, hydrolysis withendoproteinase Asp-N was performed for thepolypeptide fragmentation. Two major peptidefragments were isolated by reverse-phase HPLC.One exhibited a measured molecular mass of1919.5Da corresponding to the N-terminal (1–14)fragment: 1EKS ITWRNS MHN14 (molecularmasses were calculated using GPMAW software).The other fragment was sequenced and its 32 N-terminal amino acid residues were determined:1DKG FPWS V WSQTVSRNSSRKEKK Q-RL32. The calculated molecular mass (4385.2Da)differed from the measured value (4571.7Da) by186.5Da corresponding to a tryptophan residuewith a free acid (no amidation) at the C-terminus ofthe peptide. Thus, o-Lsp-IA was determined as asingle-chain polypeptide containing 47 amino acidresidues including eight cysteines involved in fourdisulfide bridges and four tryptophan residues. Themeasured molecular mass of the native peptide(5623.7Da) matched the calculated value (5623.5).The isoelectric point and charge at pH 7.0 of

Fig. 1. o-Lsp-IA precursor protein structure and homologous peptides

signal peptide is in italics, the prosequence is underlined and the prope

peptide sequences (1–6) retrieved from the EST database are shown

indicated.

o-Lsp-IA were calculated (GPMAW software): pI

�11.5, charge +6.1EKS IT RNS MHNDKG FP S VSQTVSRNSSRKEKK Q RL 47

3.2. Expressed sequence tag (EST) database

analysis(see footnote3)

Analysis of the Geolycosa sp. venom gland ESTdatabase (obtained in collaboration with DuPontAgriculture & Nutrition) was performed, and o-Lsp-IA full precursor protein sequence was identi-fied (Fig. 1). This protein represents a conventionalprepropeptide structure common to other knownspider toxin precursors (Kozlov et al., 2005) andconsists of (1) a typical N-terminal signal pep-tide that was identified using SignalP 3.0 soft-ware (available at http://www.cbs.dtu.dk/services/SignalP/); (2) an acidic prosequence that terminateswith the processing motif known as the ProcessingQuadruplet Motif (PQM) that specifies propeptidecleavage in spider toxin precursors (Kozlov et al.,2005); and (3) a C-terminal mature chain thatcorresponds to o-Lsp-IA toxin. These resultsunequivocally confirmed the primary structure ofmature o-Lsp-IA determined using protein chem-istry methods.

Sequences of six peptides homologous to o-Lsp-IA were also deduced (Fig. 1). We note that thenumber of EST fragments coding for o-Lsp-IA wellexceeds the number of fragments coding for otherhomologous peptides. This may result from (1) aconsiderably higher rate of expression of the o-Lsp-IA gene as compared with its homologues; (2) allelicvariation of the o-Lsp-IA gene (venom glands fromseveral individuals were used to obtain the ESTdatabase) and (3) mistakes during the databaseassembly.

deduced from the Geolycosa sp. venom gland EST dataset. The

ptide processing motif (PQM) is shown in boldface. Homologous

aligned with o-Lsp-IA and changes in amino acid sequence are

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004998

3.3. Recombinant peptide production

To provide sufficient material for functionalinvestigations, recombinant o-Lsp-IA was pro-duced in the prokaryotic expression system(Fig. 2). Thioredoxin (Trx) was chosen as the fusion

Fig. 2. Production and purification of recombinant o-Lsp-IA. (A) pET

IA gene. (B) Expression and purification of Trx-o-Lsp-IA fusion protein

Origami B cells carrying the plasmid pET-32b-o-Lsp-IA before IPTG

purified by metal-affinity chromatography on Talon Superflow resin; (

protein Trx-o-Lsp-IA digested with a recombinant light chain of ente

used. Elution was performed with a 60min linear gradient of acetoni

Fractions corresponding to o-Lsp-IA and thioredoxin (Trx) are indicate

(0.5 nmol). A Jupiter C5 column (2� 150mm, Phenomenex) was use

(20–60% in 60min) in 0.1% aqueous TFA at a flow rate of 0.3ml/min

partner for expression, since it is known to ensurehigh yields of cysteine-containing polypeptides withnative conformation. A synthetic gene coding for o-Lsp-IA was prepared and cloned into pET-32bexpression vector, and the resulting plasmid (pET-32b-o-Lsp-IA, Fig. 2A) was used to transform

-32b-o-Lsp-IA plasmid chart showing the inserted cloned o-Lsp-as followed by SDS-PAGE (12%). (1) Whole-cell lysate of E. coli

treatment; (2) Induced with 0.1mM IPTG; (3) Fusion protein

M) Molecular mass markers. (C) Reverse-phase HPLC of fusion

rokinase. A Jupiter C5 column (4.6� 150mm, Phenomenex) was

trile (0–60%) in 0.1% aqueous TFA at a flow rate of 1ml/min.

d. (D) Reverse-phase HPLC of recombinant and native o-Lsp-IAd. Elution was performed with a linear gradient of acetonitrile

.

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004 999

E. coli Origami B cells. This strain is characterizedby altered cytoplasmic RedOx potential that favorsthiol group oxidation and is therefore convention-ally used for production of cystine-containingpeptides and proteins. Trx-o-Lsp-IA fusion proteinproduction and purification was followed by SDS-PAGE (Fig. 2B). The chimeric protein was treatedwith enterokinase and the recombinant o-Lsp-IAwas purified by reverse-phase HPLC (Fig. 2C). Therecombinant peptide had the same retention timeand co-eluted with the native o-Lsp-IA whenanalyzed by analytical reverse-phase HPLC(Fig. 2D); it also had the expected N-terminalamino acid sequence as determined by direct Edmansequencing. The molecular mass of the recombinantproduct obtained by MALDI-TOF MS was equalto the mass measured previously for the nativeo-Lsp-IA. The final yield of purified recombinanto-Lsp-IA was estimated to be �200 mg/l of cellculture.

3.4. P-type current modulation

The effect of recombinant o-Lsp-IA toxin onP-type Ca2+ channels in rat cerebellar Purkinje cellswas tested. External application of 10 nM o-Lsp-IAclearly decelerated the activation kinetics andslightly decreased the amplitude of P-current with-out affecting deactivation kinetics of this currentover the whole range of activation voltages (Fig. 3).Subsequent increase of the toxin concentration upto 100 nM had no further effect. At saturatingconcentrations (410 nM), the development ofo-Lsp-IA action on P-current was rapid (o10 s),whereas at lower concentrations (o1 nM) it wasslow (45min). These effects were not use depen-dent and were not removed by strong (+50mV)depolarizing pre-pulses. Subsequent wash-out re-sulted in a partial recovery of both activationkinetics and amplitude of P-current within severalminutes. The data obtained in these experimentsconfirmed that the recombinant o-Lsp-IA toxinproduced the same effects on P-current as the nativeo-Lsp-IA toxin (Fisyunov et al., 2005).

Really dramatic slow-down of the activationkinetics was manifested as the appearance of asecond, slower component of the current activation.The curve fitting of toxin-modified currents bybiexponential function resulted in the voltagedependence of the fast time constant (tfast) similarto the time constant in control currents (tcon) andalso revealed a slow component with the time

constant (tslow) �five-fold larger than in control(tslow/tfastE5) (Fig. 3D). Voltage dependence of thenormalized amplitude of this slow component wassimilar to the control I–V curve, while the normal-ized amplitude of the fast component was practi-cally constant over the range of activation voltages(Fig. 3E). Consequently, for toxin-modified cur-rents, the peak value of the amplitude ratio of theslow component (Islow) to the fast component (Ifast)was observed near the maximum of the I–V curve(Islow/IfastE5).

3.5. Selectivity of action

To test the selectivity of o-Lsp-IA action onCa2+ channels, we used hippocampal neuronsexpressing the main types of Ca2+ channels (Averyand Johnston, 1996; Ishibashi et al., 1995). Applica-tion of 10 nM of o-Lsp-IA resulted in a rapid(developing in less than 10 s) alteration in theactivation kinetics and amplitude of the whole-cellCa2+ current (n ¼ 5) (Fig. 4A, B). These changeswere consistent with the slow-down of the P-typechannel-mediated component of Ca2+ current. Theeffects were partially reversed upon the return to thecontrol solution.

We also used o-agatoxin IVA (o-Aga-IVA),which is the selective blocker of P-type Ca2+

channels (Mintz et al., 1992). After incubationwith 100 nM of o-Aga-IVA for 3min, applicationof o-Lsp-IA toxin at saturating concentration didnot produce any changes in the residual Ca2+

current (n ¼ 8) (Fig. 4C, D). Therefore, the effect ofo-Lsp-IA appears to be limited to the action on P-type Ca2+ channels.

4. Discussion

Animal peptide toxins are among the mostefficient tools for the investigation of Ca2+ chan-nels. Voltage-gated Ca2+ channels of N- and P/Q-types deliver a major part of Ca2+ influx into thepre-synaptic terminals of central neurons initiatingsynaptic transmission (Regehr and Mintz, 1994;Wheeler et al., 1994; Wu et al., 1999). Several toxinsare known to interact directly with these channels(Dos Santos et al., 2002; McDonough et al., 2002).Previously, we have described isolation of a novelpeptide toxin affecting P-type Ca2+ currents inmammalian central neurons (Fisyunov et al., 2005).In the present work, we focused on structuralinvestigation and selectivity of action of this new

ARTICLE IN PRESS

Fig. 3. Effects of the recombinant o-Lsp-IA toxin on the activation kinetics and the amplitude of P-current. (A) Typical families of Ba2+

currents through P-type Ca2+ channels recorded in control (-J-) and after the action of the saturating concentration (10 nM) of o-Lsp-IA(-K-). The voltages of the test pulses are indicated near the current traces. The neuron was held at �70mV and stimulated every 5 s by

50ms-long voltage steps in a 5mV increment. (B) Current–voltage (I/V) relationships for the recordings shown in (A). (C) The activation

kinetics of the currents at test voltage of �25mV was fitted by a single exponential curve with a time constant tcon ¼ 2.3370.01ms in

control (-J-) and a biexponential curve with a fast time constant tfast ¼ 2.3470.04ms and a slow time constant tslow ¼ 14.170.04ms in

10 nM o-Lsp-IA (-K-). The inset illustrates the normalized tail currents for comparison. (D) Voltage dependence of the mean time

constants for the activation kinetics in control (-J-) and after the action of 10 nM of o-Lsp-IA for the fast (-K-) and slow (-’-)

components of P-current. (E) The same data as in (D) analyzed for the voltage dependence of the mean normalized amplitude in control

(empty blocks) and after incubation with the toxin for the fast (filled blocks) and slow (hatched blocks) components of P-current calculated

using Eq. (1). Vertical bar: mean7SD, n ¼ 4.

K. Pluzhnikov et al. / Toxicon 50 (2007) 993–10041000

toxin (named o-Lsp-IA) and its recombinantanalog.

The primary structure of o-Lsp-IA was estab-lished by protein chemistry methods (see Section3.1). Similarity searches in protein sequence data-

bases revealed that o-Lsp-IA shows considerablehomology (41% identity) only with neurotoxinTx3-5, a Ca2+ channel inhibitor from the Brazilianarmed spider Phoneutria nigriventer that predomi-nantly affects L-type channels (Cordeiro Mdo et al.,

ARTICLE IN PRESS

Fig. 4. Effects of o-Lsp-IA toxin on the whole-cell Ca2+ current in hippocampal neurons. (A) o-Lsp-IA toxin produced modulation of

the whole-cell current. Typical currents recorded in control (-J-) and after the application of 10 nM of o-Lsp-IA (-K-). The neuron was

kept at �100mV and currents were elicited by step depolarization to �10mV. (B) The traces demonstrated in (A) were normalized. (C) o-Aga-IVA toxin prevented the changes in the activation kinetics of whole-cell currents produced by o-Lsp-IA. Typical currents recorded in

control (-J-), after the incubation with 100 nM of o-Aga-IVA for 3min (-&-) and following application of 10 nM of o-Lsp-IA (-K-). The

neuron was held at �100mV. Currents were evoked by depolarization to �10mV. (D) The traces demonstrated in (C) were normalized.

Fig. 5. Amino acid sequence alignment of o-Lsp-IA and other Ca2+ channel modulators. Tx3–5, Tx3–6—toxins from the spider

Phoneutria nigriventer (Cordeiro Mdo et al., 1993; Cardoso et al., 2003); o-Aga-IIIA, o-Aga-IVA—toxins from the spider Agelenopsis

aperta (Mintz et al., 1991, 1992); SNX-325, SNX-482—toxins from the spider Segestria florentina (Newcomb et al., 1995, 1998); o-GTx-

SIA—o-grammotoxin SIA from the spider Grammostola spatula (Lampe et al., 1993); o-CTx-GVIA—o-conotoxin GVIA from Conus

geographus (Olivera et al., 1984); o-CTx-MVIIA, o-CTx-MVIIC—o-conotoxins MVIIA, MVIIC from Conus magus (Olivera et al., 1987;

Hillyard et al., 1992). The number of amino acid residues in the polypeptide chain of the corresponding toxin is indicated in the right

column. Cysteine residues are shaded dark gray. Residues identical to o-Lsp-IA are shaded light gray. Clusters of positively charged

residues in the C-terminal parts of sequences are underlined. Residues in o-Lsp-IA suggested to contribute to toxin–membrane

interactions according to Lee and MacKinnon (2004) are shown in boldface.

K. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004 1001

1993; Cardoso et al., 2003) (Fig. 5). The structure ofo-Lsp-IA toxin exhibits some similarity to Tx3-6neurotoxin, also from the venom of P. nigriventer

(Cardoso et al., 2003) (Fig. 5). Tx3-6 is a potentblocker of voltage-gated Ca2+ channels, mainlyP-type channels with a minor effect on L- andN-type channels (Vieira et al., 2003). Negligible

homology (28% identity) was found betweeno-Lsp-IA and o-agatoxin IVA (o-Aga-IVA), a48-amino acid polypeptide from the spider Agele-

nopsis aperta, a classical selective inhibitor of P-typechannels, effective at nanomolar concentrations(Mintz et al., 1992). Similar to Ca2+ channelblockers Tx3-6 and o-Aga-IVA, o-Lsp-IA contains

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–10041002

eight conserved half-cystines and therefore possiblyhas the same disulfide-bonding pattern and similarthree-dimensional structure resembling the inhibitorcystine knot (ICK) fold common to most spiderneurotoxins.

Fig. 5 shows sequence alignment of o-Lsp-IAwith other known Ca2+ channel antagonists iso-lated from spider and Conus venoms. Among these,o-Aga-IVA selectively targets P-type channels(Mintz et al., 1992), SNX-325 and o-conotoxinsGVIA and MVIIA are selective towards N-typechannels (Newcomb et al., 1995; Olivera et al., 1984,1987; Nielsen et al., 2000), SNX-482 blocks R-typechannels (Newcomb et al., 1998), whereaso-agatoxin IIIA (o-Aga-IIIA), o-grammotoxinSIA and o-conotoxin MVIIC show broader speci-ficity and affect both N- and P-type channels(Mintz et al., 1991; Lampe et al., 1993; Hillyardet al., 1992).

Two principal mechanisms of action have beensuggested for Ca2+ channel antagonists. Oneis direct pore blockage, which operates witho-conotoxins GVIA, MVIIA, MVIIC and spidertoxin o-Aga-IIIA. The other is gating modificationby antagonizing voltage-sensor movement sug-gested in the case of o-Aga-IVA, SNX-482 ando-grammotoxin SIA (McDonough, 2007).

We note that a peculiar cluster of positivelycharged amino acid residues is located in theC-terminal parts of some Ca2+ channel antagonists,which act through pore occlusion. This cluster mayeffectively interact with the conserved negativelycharged residues in the channel turret. A similarcluster was found in o-Lsp-IA (Fig. 5). This toxin,on the other hand, produces activation kineticsalteration, the effect that considerably differs fromsimple pore blockage and resembles gating modi-fiers. Moreover, peculiar regular distribution ofhydrophobic residues (tryptophan, in particular)suggested to be involved in the membrane-accessmechanism of voltage-sensor toxins (Lee andMacKinnon, 2004; McDonough, 2007) was alsoobserved in o-Lsp-IA structure (see Fig. 5). At thesame time, in contrast to gating modifiers o-Aga-IVA and o-grammotoxin SIA, the modulatoryeffects of o-Lsp-IA were not altered by strongdepolarizing pre-pulses (Fisyunov et al., 2005). Wetherefore suggest that o-Lsp-IA binding may affectboth the pore region and the gating mechanism.

o-Lsp-IA appears to be selective for P-type Ca2+

channels. Tentatively it may be added to the listof specific Ca2+ channel toxins. Modulation of

P-current by o-Lsp-IA clearly differs from inhibi-tory effects of other toxins (Mintz, 1994; McDo-nough et al., 1997a, b, 2002; Dos Santos et al., 2002)that only decrease the amplitude of Ca2+ currents.By changing the kinetics of P-type Ca2+ current,o-Lsp-IA toxin can be used as a unique tool formodulating the synaptic transmission. It can beused to develop small mimetic molecules withanalogous actions. When acting in vivo, suchmolecules will modulate rather than inhibit synaptictransmission by affecting bursting activity ofneurons, etc. Voltage-gated Ca2+ channels arepotential targets for therapeutics directed againstmigraine and ataxia (Ophoff et al., 1996; Krauset al., 2000; Cao and Tsien, 2005; Jeng et al., 2006),intractable pain (Bowersox and Luther, 1998) andsome forms of epilepsy (Burgess et al., 1997). Thismakes o-Lsp-IA toxin attractive from both scien-tific and clinical perspectives.

Acknowledgments

This work was supported in part by the RussianFoundation for Basic Research (Grant no. 06-04-48425), the Program of Molecular and Cell Biologyof Russian Academy of Sciences (E. Grishin) andCRDF (O. Krishtal).

We thank Sergey Kozlov for the analysis of thevenom gland expressed sequence tag database, VeraTsintsadze for the preparation of the neurons, BillMcCutchen and Rafi Herrmann for the cDNAlibrary construction and sequencing, and AndreyFeodorov for expert opinion concerning spidertaxonomy and biology.

References

Adams, M.E., 2004. Agatoxins: ion channel specific toxins from

the American funnel web spider, Agelenopsis aperta. Toxicon

43, 509–525.

Adams, M.E., Mintz, I.M., Reily, M.D., Thanabal, V., Bean,

B.P., 1993. Structure and properties of omega-agatoxin IVB,

a new antagonist of P-type calcium channels. Mol. Pharma-

col. 44, 681–688.

Avery, R.B., Johnston, D., 1996. Multiple channel types

contribute to the low-voltage-activated calcium current in

hippocampal CA3 pyramidal neurons. J. Neurosci. 16,

5567–5582.

Bowersox, S.S., Luther, R., 1998. Pharmacotherapeutic potential

of omega-conotoxin MVIIA (SNX-111), an N-type neuronal

calcium channel blocker found in the venom of Conus magus.

Toxicon 36, 1651–1658.

Burgess, D.L., Jones, J.M., Meisler, M.H., Noebels, J.L., 1997.

Mutation of the Ca2+ channel beta subunit gene Cchb4 is

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–1004 1003

associated with ataxia and seizures in the lethargic (lh) mouse.

Cell 88, 385–392.

Cao, Y.Q., Tsien, R.W., 2005. Effects of familial hemiplegic

migraine type 1 mutations on neyronal P/Q-type Ca2+

channel activity and inhibitory synaptic transmission. Proc.

Natl. Sci. USA 102, 2590–2595.

Cardoso, F.C., Pacifico, L.G., Carvalho, D.C., Victoria, J.M.N.,

Neves, A.L.G., Chavez-Olortegui, C., Gomez, M.V., Kala-

pothakis, E., 2003. Molecular cloning and characterization of

Phoneutria nigriventer toxins active on calcium channels.

Toxicon 41, 755–763.

Cordeiro Mdo, N., de Figueiredo, S.G., Valentim Ado, C., Diniz,

C.R., von Eickstedt, V.R., Gilroy, J., Richardson, M., 1993.

Purification and amino acid sequences of six Tx3 type

neurotoxins from the venom of the Brazilian ‘armed’ spider

Phoneutria nigriventer (Keys). Toxicon 31, 35–42.

Dos Santos, R., Van Renterghem, C., Martin-Moutot, N.,

Mansuelle, P., Cordeiro, M.N., Diniz, C.R., Mori, Y., De

Lima, M.E., Seagar, M., 2002. Phoneutria nigriventer omega-

phonetoxin IIA blocks the Cav2 family of calcium channels

and interacts with omega-conotoxin-binding sites. J. Biol.

Chem. 277, 13856–13862.

Fisyunov, A., Pluzhnikov, K., Molyavka, A., Grishin, E.,

Lozovaya, N., Krishtal, O., 2005. Novel spider toxin slows

down the activation kinetics of P-type Ca2+ channels in

Purkinje neurons of rat. Toxicology 207, 129–136.

Gasparian, M.E., Ostapchenko, V.G., Schulga, A.A., Dolgikh,

D.A., Kirpichnikov, M.P., 2003. Expression, purification, and

characterization of human enteropeptidase catalytic subunit

in Escherichia coli. Protein Exp. Purif. 31, 133–139.

Hille, B., 2001. Ion Channels of Excitable Membranes, third ed.

Sinauer Associates, Inc., Sunderland, pp.

Hillyard, D.R., Monje, V.D., Mintz, I.M., Bean, B.P., Nadasdi,

L., Ramachandran, J., Miljanich, G., Azimi-Zoonooz, A.,

McIntosh, J.M., Cruz, L.J., et al., 1992. A new Conus peptide

ligand for mammalian presynaptic Ca2+ channels. Neuron 9,

69–77.

Ishibashi, H., Rhee, J.S., Akaike, N., 1995. Regional difference of

high voltage-activated Ca2+ channels in rat CNS neurones.

Neuroreport 6, 1621–1624.

Jeng, C.J., Chen, Y.T., Chen, Y.W., Tang, C.Y., 2006.

Dominant-negative effects of human P/Q-type Ca2+ channel

mutations associated with episodic ataxia type 2. Am.

J. Physiol. Cell Physiol. 290, C1209–C1220.

Julius, D., Basbaum, A.I., 2001. Molecular mechanisms of

nociception. Nature 413, 203–210.

Kozlov, S., Malyavka, A., McCutchen, B., Lu, A., Schepers, E.,

Herrmann, R., Grishin, E., 2005. A novel strategy for the

identification of toxinlike structures in spider venom. Proteins

59, 131–140.

Kraus, R.L., Sinnegger, M.J., Koschak, A., Glossmann, H.,

Stenirri, S., Carrera, P., Striessnig, J., 2000. Three new

familial hemiplegic migraine mutants affect P/Q-type Ca(2+)

channel kinetics. J. Biol. Chem. 275, 9239–9243.

Lampe, R.A., Defeo, P.A., Davison, M.D., Herman, J.L.,

Spreen, R.C., Horn, M.B., Mangono, T.J., 1993. Isolation

and pharmacological characterization of omega-grammotox-

in SIA, a novel peptide inhibitor of neuronal voltage-sensitive

calcium channel responses. Mol. Pharmacol. 44, 451–460.

Lee, S.Y., MacKinnon, R., 2004. A membrane-access mechanism

of ion channel inhibition by voltage sensor toxins from spider

venom. Nature 430, 232–235.

McDonough, S.I., 2007. Gating modifier toxins of voltage-gated

calcium channels. Toxicon 49, 202–212.

McDonough, S.I., Mintz, I.M., Bean, B.P., 1997a. Alteration of

P-type calcium channel gating by the spider toxin omega-Aga-

IVA. Biophys. J. 72, 2117–2128.

McDonough, S.I., Lampe, R.A., Keith, R.A., Bean, B.P., 1997b.

Voltage-dependent inhibition of N- and P-type calcium

channels by the peptide toxin omega-grammotoxin-SIA.

Mol. Pharmacol. 52, 1095–1104.

McDonough, S.I., Boland, L.M., Mintz, I.M., Bean, B.P., 2002.

Interactions among toxins that inhibit N-type and P-type

calcium channels. J. Gen. Physiol. 119, 313–328.

Miljanich, G.P., 2004. Ziconotide: neuronal calcium channel

blocker for treating severe chronic pain. Curr. Med. Chem.

11, 3029–3040.

Mintz, I.M., 1994. Block of Ca channels in rat central neurons

by the spider toxin omega-Aga-IIIA. J. Neurosci. 14,

2844–2853.

Mintz, I.M., Venema, V.J., Adams, M.E., Bean, B.P., 1991.

Inhibition of N- and L-type calcium channels by the spider

venom toxin omega-Aga-IIIA. Proc. Natl. Sci. USA 88,

6628–6631.

Mintz, I.M., Venema, V.J., Swiderek, K.M., Lee, T.D., Bean,

B.P., Adams, M.E., 1992. P-type calcium channels blocked by

the spider toxin omega-Aga-IVA. Nature 355, 827–829.

Newcomb, R., Palma, A., Fox, J., Gaur, S., Lau, K., Chung, D.,

Cong, R., Bell, J.R., Horne, B., Nadasdi, L., Ramachandran,

J., 1995. SNX-325, A novel calcium antagonist from the

spider Segestria florentina. Biochemistry 34, 8341–8347.

Newcomb, R., Szoke, B., Palma, A., Wang, G., Chen, X.,

Hopkins, W., Cong, R., Miller, J., Urge, L., Tarczy-Hornoch,

K., Loo, J.A., Dooley, D.J., Nadasdi, L., Tsien, R.W.,

Lemos, J., Miljanich, G., 1998. Selective peptide antagonist of

the class E calcium channel from the venom of the tarantula

Hysterocrates gigas. Biochemistry 37, 15353–15362.

Nielsen, K.J., Schroeder, T., Lewis, R., 2000. Structure–activity

relationships of omega-conotoxins at N-type voltage-sensitive

calcium channels. J. Mol. Recognit. 13, 55–70.

Olivera, B.M., McIntosh, J.M., Cruz, L.J., Luque, F.A., Gray,

W.R., 1984. Purification and sequence of a presynaptic

peptide toxin from Conus geographus venom. Biochemistry

23, 5087–5090.

Olivera, B.M., Cruz, L.J., de Santos, V., LeCheminant, G.W.,

Griffin, D., Zeikus, R., McIntosh, J.M., Galyean, R., Varga,

J., Gray, W.R., et al., 1987. Neuronal calcium channel

antagonists. Discrimination between calcium channel sub-

types using omega-conotoxin from Conus magus venom.

Biochemistry 26, 2086–2090.

Ophoff, R.A., Terwindt, G.M., Vergouwe, M.N., van Eijk, R.,

Oefner, P.J., Hoffman, S.M., Lamerdin, J.E., Mohrenweiser,

H.W., Bulman, D.E., Ferrari, M., Haan, J., Lindhout, D.,

van Ommen, G.J., Hofker, M.H., Ferrari, M.D., Frants,

R.R., 1996. Familial hemiplegic migraine and episodic ataxia

type-2 are caused by mutations in the Ca2+ channel gene

CACNL1A4. Cell 87, 543–552.

Panchenko, V.A., Krishtal, O.A., Tegtmeier, F., Tsyndrenko,

A.Y., 1993. R56865 as Ca(2+)-channel blocker in Purkinje

neurons of rat: comparison with flunarizine and nimodipine.

Neuroscience 54, 587–594.

Regan, L.J., 1991. Voltage-dependent calcium currents in

Purkinje cells from rat cerebellar vermis. J. Neurosci. 11,

2259–2269.

ARTICLE IN PRESSK. Pluzhnikov et al. / Toxicon 50 (2007) 993–10041004

Regehr, W.G., Mintz, I.M., 1994. Participation of multiple

calcium channel types in transmission at single climbing fiber

to Purkinje cell synapses. Neuron 12, 605–613.

Vieira, L.B., Kushmerick, C., Reis, H.J., Diniz, C.R., Cordeiro,

M.N., Prado, M.A.M., Kalapothakis, E., Romano-Silva, M.A.,

Gomez, M.V., 2003. PnTx3-6 a spider neurotoxin inhibits K+-

evoked increase in [Ca2+](i) and Ca2+-dependent glutamate

release in synaptosomes. Neurochem. Int. 42, 277–282.

Wheeler, D.B., Randall, A., Tsien, R.W., 1994. Roles of N-type

and Q-type Ca2+ channels in supporting hippocampal

synaptic transmission. Science 264, 107–111.

Wu, L.G., Westenbroek, R.E., Borst, J.G., Catterall, W.A.,

Sakmann, B., 1999. Calcium channel types with

distinct presynaptic localization couple differentially to trans-

mitter release in single calyx-type synapses. J. Neurosci. 19,

726–736.