Centro de Bioproductos Marinos (CEBIMAR), Loma y 37, Alturas del Vedado, CP 10400 Habana, CubaInstituto de Fisiología, Universidad Autónoma de Puebla, 14 sur 6301, CU, San Manuel, Puebla, Puebla CP 72750, MexicoLaboratory of Genetics, Butantan Institute, São Paulo 05503-900, BrazilLaboratory of Toxicology, University of Leuven (KU Leuven), Campus Gasthuisberg O&N2, Herestraat 49, P.O. Box 922, 3000 Leuven, BelgiumCentro Nacional de Investigaciones Oncológicas (CNIO), C/ Melchor Fernández Almagro 3, 28029 Madrid, SpainFundacão Ezequiel Dias-FUNED, Rua Conde Pereira Carneiro 80, CEP 30510-010 Belo Horizonte, MG, BrazilKompetenzzentrum Ulm Peptide Pharmaceuticals (U-PEP), Universität Ulm (West), Albert-Einstein Allee 47, 89081 Ulm, Germany
r t i c l e i n f o
rticle history:eceived 16 April 2013eceived in revised form 1 June 2013ccepted 3 June 2013vailable online 10 June 2013
edicated to professor Lászlo Béress, aioneer in the research on sea anemoneoxins, for his contributions to this field.
eywords:SIC
on channel
a b s t r a c t
Sea anemones produce ion channels peptide toxins of pharmacological and biomedical interest. However,peptides acting on ligand-gated ion channels, including acid-sensing ion channel (ASIC) toxins, remainpoorly explored. PhcrTx1 is the first compound characterized from the sea anemone Phymanthus crucifer,and it constitutes a novel ASIC inhibitor. This peptide was purified by gel filtration, ion-exchange andreversed-phase chromatography followed by biological evaluation on ion channels of isolated rat dorsalroot ganglia (DRG) neurons using patch clamp techniques. PhcrTx1 partially inhibited ASIC currents(IC50 ∼ 100 nM), and also voltage-gated K+ currents but the effects on the peak and on the steady statecurrents were lower than 20% in DRG neurons, at concentrations in the micromolar range. No significanteffect was observed on Na+ voltage-gated currents in DRG neurons. The N-terminal sequencing yielded32 amino acid residues, with a molecular mass of 3477 Da by mass spectrometry. No sequence identityto other sea anemone peptides was found. Interestingly, the bioinformatic analysis of Cys-pattern and
ea anemonenimal venomoxinnhibitor Cystine Knot
secondary structure arrangement suggested that this peptide presents an Inhibitor Cystine Knot (ICK)scaffold, which has been found in other venomous organisms such as spider, scorpions and cone snails.Our results show that PhcrTx1 represents the first member of a new structural group of sea anemonestoxins acting on ASIC and, with much lower potency, on Kv channels. Moreover, this is the first report ofan ICK peptide in cnidarians, suggesting that the occurrence of this motif in venomous animals is moreancient than expected.
. Introduction
Sea anemones are venomous animals that produce a greatumber of bioactive peptides and proteins, comprising cytolysins,
hospholipases, ion channel toxins and protease inhibitors [44].hese proteinaceous molecules are used as weapons for prey-ng on small crustaceans and fishes, and for defense against
predators. The most commonly known bioactive proteins are the20 kDa actinoporins [4], whereas more than a hundred peptidetoxins, mainly comprising voltage-gated ion (Na+ and K+) chan-nel toxins [44], have been isolated and characterized from theseorganisms during the last four decades. Nevertheless, recent trans-criptomic and peptidomic studies have shown that the diversity ofpeptide toxins produced by sea anemones is more complex thanpreviously estimated [36,51,52,66], thus opening up new possibili-ties in the search for novel structures and biological activities fromthese organisms, including new peptide toxins acting on diversereceptors, and voltage-gated and ligand-gated ion channels.
Sea anemone peptide toxins acting on ligand-gated ion chan-
nels, such as acid-sensing ion channel (ASICs), are barely known incontrast with the known large number of voltage-gated ion chan-nel toxins. ASICs are H+-gated Na+ channels that belong to theENaC/degenerin superfamily of sodium channels [48]. ASICs are
nvolved in sensory perception, synaptic plasticity, learning, mem-ry formation, cell migration and proliferation, nociception, andeurodegenerative disorders, among other processes [48]; there-
ore those molecules that specifically target these channels aref growing pharmacological and biomedical interest. Few pro-einaceous ASIC toxins have been isolated and characterized fromenomous animals. Up to date, APETx2 is the only well-knownSIC toxin from a sea anemone species (Anthopleura elegantis-ima), which selectively inhibits the ASIC3 subunit in Xenopus laevisocytes (IC50 = 63 nM) [20] as well as the ASIC3-like current ofRG neurons with an IC50 of 216 nM [20], although it also inhibits
he tetrodotoxin (TTX)-resistant Nav 1.8 currents of DRG neuronesIC50 = 2.6 �M) [9], and channel subtypes Nav1.2 and Nav1.8 in X.aevis oocytes [45]. Recently, another peptide toxin (�-AnmTX Hcrb-1) acting on ASIC3 in X. laevis oocytes was characterized from
sea anemone [37], however its potency (IC50 = 5.5 �M) is muchower than the one exhibited by APETx2. Some other proteinaceousompounds acting on ASICs have been isolated from tarantula [25]nd snake species [10,21]. Therefore the finding of ASIC modulatorsrom animal venoms remains limited considering the large num-er of biologically active peptide and proteins produced by theserganisms.
In the present work we isolated and characterized a new ASICnhibitor from the aqueous extract of the sea anemone P. crucifer.his peptide toxin, named PhcrTx1, inhibited the ASIC currentsn rat sensory neurons and produced a smaller but significantnhibitory effect on voltage-gated K+ currents, with no action onoltage-gated Na+ currents. PhcrTx1 is the first toxin isolated andharacterized from P. crucifer and it has no significant sequenceimilarity to any known sea anemone peptide. Our analysis indi-ated that this peptide presents the Inhibitor Cystine Knot (ICK)otif, which has not been previously identified in cnidarians. These
esults show that PhcrTx1 represents the first isolated and charac-erized member of a new structural group of sea anemone toxins.
. Materials and methods
.1. Sample preparation and chromatographic separation of theioactive peptide
Fifteen specimens of the sea anemone P. crucifer (Le Sueur, 1817)ere collected at the north coast of Havana. A voucher sample (No.NC 03.3.2.001) was deposited at the Cuban National Aquarium.
n the laboratory the specimens were immediately separated fromtones, and then homogenized together with the secreted mucussing a blender. The homogenate was lyophilized and the driedaterial (27 g) stored at −20 ◦C.An amount of 5 g of the dry homogenate was mixed with 100 ml
f 0.1 M ammonium acetate (p.a, Merck, Germany), stirred dur-ng 30 min and centrifuged at 4000 × g during 1 h at 4 ◦C. Theupernatant (350 mg/90 ml) was applied onto a Sephadex G-50 MPharmacia, Sweden) column (5 cm × 93 cm) previously calibrated52] and the separation was done at a flow rate of 2 ml/min using.1 M ammonium acetate as eluent. Fractions of 20 ml each wereollected and manually read at 280 nm on a UV-1201 spectropho-ometer (Shimadzu, Japan).
The biologically active samples from gel filtration were appliedo a Fractogel EMD SO3
− 650 M (Merck, Germany) cation-exchangeolumn of dimensions 1.8 cm × 5 cm. The ion-exchange step waserformed under the following conditions: a 400 ml (31 column
olumes) gradient of ammonium acetate was run at a flow rate of
ml/min, from 0.01 to 1 M, using a gradient mixer GM-1 (Phar-acia, Sweden). Eighty fractions of 5 ml each were collected andanually read at 280 nm.
ides 53 (2014) 3–12
The biologically active samples from ion-exchange chro-matography were submitted to reversed-phase chromatographyin a Hypersil H5 ODS column (Unicam, UK) of dimensions4.6 mm × 250 mm, previously equilibrated in solvent A, 0.1% tri-fluoroacetic acid (HPLC grade, AppliChem, Germany). Elution wascarried out at a flow rate of 0.8 ml/min using stepwise elution at100% A during 10 min, followed by an ascending linear gradient ofsolvent B, 0.05% TFA in acetonitrile (0–80% B in 80 min). The peptidewas re-purified in a Discovery RPC18, 4.6 mm × 250 mm (Supelco,USA) HPLC column, using the gradient 10–20%B in 5 min, and then20–30% in 50 min, at 1 ml/min. Eluting compounds were detectedat 214 nm.
The protein content of the aqueous extract and chromatographicfractions were estimated with a BCA protein assay kit (AppliChem,Germany) [58].
2.2. Patch clamp experiments on isolated neurons
The biological activity screening of chromatographic fractionswas performed on ASIC currents using the whole-cell patch clamptechnique in primary cultured rat dorsal root ganglion (DRG) neu-rons. The pure compound was characterized on the ASIC currentsand also on the voltage-gated (Na+ and K+) currents in DRG neu-rons. Animal care and procedures were in accordance with theNational Institutes of Health Guide for the Care and Use of Labora-tory Animals. All efforts were made to minimize animal suffering.The number of animals used for this work was kept to the minimumnecessary for a meaningful interpretation of the data.
The DRG neurons were isolated from Wistar rats at postna-tal ages P5 to P9 without sex distinction, and cultured accordingto a previously reported procedure [53]. Briefly, for the cell cul-ture the rats were anesthetized and killed with an overdose ofsevofluorane. The dorsal root ganglia were isolated from the ver-tebral column and incubated (30 min at 37 ◦C) in Leibovitz L15medium (L15) (Invitrogen, USA) containing 1.25 mg/ml trypsinand 1.25 mg/ml collagenase (Sigma–Aldrich, USA). After enzymatictreatment, the ganglia were washed three times with sterile L15.Cells were mechanically dissociated using a Pasteur pipette andthen plated on 12-mm × 10-mm glass coverslips (Corning, USA)pre-treated with poly-d-lysine (Sigma–Aldrich, USA) and placedonto 35-mm culture dishes (Corning, USA). Neurons were incu-bated 4–8 h in a humidified atmosphere (95% air, 5% CO2, at 37 ◦C)using a CO2 water-jacketed incubator (Nuaire, USA) to allow theisolated cells to settle and adhere to the coverslips. The plat-ing medium contained L15, with added 15.7 mM NaHCO3 (Merck,Mexico), 10% fetal bovine serum, 2.5 �g/ml fungizone (both fromInvitrogen), 100 U/ml penicillin (Lakeside, Mexico), and 15.8 mMHEPES (Sigma–Aldrich, USA).
Whole-cell recording was carried out using an Axopatch-1Damplifier (Axon Instruments, USA). The pulse generation and thedata sampling were controlled by Pclamp 9.2 software (AxonInstruments, USA) using a 16-bit data-acquisition system, Digidata1320A (Axon Instruments, USA). All experiments were performedat room temperature (23–25 ◦C). The patch electrodes were pulledfrom borosilicate glass and had a resistance in the range of1.5–2.5 M� when filled with the intracellular solution (see Table 1).The capacitance and series resistance (80%) were electronicallycompensated. Experiments were rejected when the voltage errorexceeded 5 mV after compensation for series resistance at the max-imum peak current, but no corrections were made for smallervalues.
ASIC currents were elicited by a fast (about 40 ms) pH change
from 7.4 to 6.1 for 5 s, by shifting one of the three outlets of afast change perfusion system (SF-77B, Warner Inst., USA) whilekeeping the cell at a holding potential (Vh) of −60 mV. The timeinterval between the pH change steps was 1 min to guarantee that
A.A. Rodríguez et al. / Peptides 53 (2014) 3–12 5
Table 1Solutions used in the whole-cell patch clamp measurements of ASIC currents and voltage-gated K+ currents. Concentration values are expressed in mM. MES (pK = 6.15) wasused instead of HEPES (pK = 7.55) for working at pH 6.1.
Intracellular Extracellular normal Extracellular pH 6.1 Intracellular IK Extracellular IK Intracellular INa Extracellular INa
he ASIC current was completely recovered from desensitization.he transient receptor potential V1 (TRPV1) antagonist capsazepine10 �M) (Sigma Chemical Co, USA) was added to the extracellularolution (pH 6.1) to avoid potential activation of TRPV1 receptory pH drop. The parameters measured to characterize the ASICurrents were: (a) maximum peak amplitude (Imax); (b) currentesensitization time constant (�des, determined by fitting the decayhase of the current with a single exponential function), (c) themplitude at the end of the 5 s acid pulse (Iss, it was considered aeasure of the steady state current and it was calculated as theean of the current in the last 100 ms of the acid pulse). The com-
ounds were applied 20 s before the pH change to 6.1 and duringhe whole pH pulse (sustained application). In a subgroup of exper-ments, the compounds were applied only during the 5 s perfusionf the acid solution (during the pH drop). For the study of effectsn voltage-gated Na+ and K+ currents, the toxin was ejected underressure using a microinjector (Baby Bee, USA) from a micropipetteositioned in the vicinity of the cell under recording. This allowedparing toxin since the gravity-driven fast changer requires higherolumes of toxin. K+ currents were elicited by a single-step volt-ge protocol comprising depolarizing pulses from −100 mV (Vhold)o 0 mV during 800 ms every 8 s. Na+ currents were elicited by aingle-step voltage protocol comprising depolarizing pulses from100 mV (Vhold) to −10 mV during 40 ms every 8 s. The solutionssed in the experiments are depicted in Table 1.
The concentration–response curves were fitted to the Hill equa-ion: Y = Ymaxxn/(kn + xn), where Y is the pharmacological effect ofhe substance under study, Ymax is its maximum effect, x is theoncentration of the toxin, k is the concentration of the toxin thatroduces half the maximum effect and n is the Hill coefficient.o define the statistical significance the control recordings wereompared with those obtained after toxin perfusion, using a pairedtudent’s t-test; a P ≤ 0.05 was considered as significant. Unless oth-rwise stated, all numerical data are presented as the mean ± S.E.M.
.3. Crab bioassay
The pure toxin was assayed on male shore crabs Uca thayeri,ased on the crab bioassay widely used for the detection of para-
yzing effects induced by sea anemone toxins [8,55]. Samples werenjected (10 �l/g crab weight) into the base of the third walkingeg. The concentration of toxin used in the assay was 0.2 mg/ml
dose: 2000 �g/kg) and 6 crabs weighing 2–4 g were used per sam-le (2–4 �g were injected to crabs). The toxicity was consideredositive at the inability of crabs (placed upward-facing) to righthemselves within 2 h after toxin administration.
– – 100 –– – 30 –7.2 7.4 7.3 7.4
2.4. Mass spectrometry and N-terminal sequencing
The molecular mass of the toxin was measured with aVoyager DE Pro matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) device (PerSeptiveBiosystems, USA). The matrix solution was prepared with �-cyano-4-hydroxycinnamic acid dissolved in mass buffer (50%acetonitrile + 0.1% TFA) up to saturation. Measurement was per-formed in linear mode. Positive ions were accelerated at 25 kV,and up to 50 laser shots were automatically accumulated. Voy-ager RP BioSpectrometry Workstation version 3.07.1 (PerSeptiveBiosystems, USA) was used as controlling software.
The peptide sample was dissolved in 10% acetonitrile in water(v/v) and spotted on a glass fiber disk (Wako, Japan) whichwas pretreated with Sequa-brene (Sigma, USA). Sequences weredetermined by automated Edman degradation using a ShimadzuPPSQ-30 protein sequencer (Tokyo, Japan) according to the manu-facturer’s instructions.
2.5. Computational analyses
The software GPMAW 9.02 (http://www.welcome.to/gpmaw)[46] was used for the theoretical calculations of isoelectric point,net charge and average molecular mass. Sequences and three-dimensional (3D) structure of peptides were retrieved from theUniProt/Swiss-Prot and the Protein Data Bank (PDB) databases,respectively. PSI-BLAST (http://www.ebi.ac.uk/Tools/sss/psiblast/)[2] and Pfam (http://pfam.sanger.ac.uk/) [47] web-serverswere used to identify related sequences based on sequencesimilarity and domain architecture, respectively. Sequencealignments with bit score greater than 100 and E-value ofless than 0.001 were considered significant. Clustalw 2.1(http://www.ebi.ac.uk/Tools/clustalw2/index.html) [38], wasused for multiple sequence alignment of the toxin and relatedsequences. Independent pairwise alignments were doneby Lalign software (http://www.ebi.ac.uk/Tools/psa/lalign/).Secondary structure prediction was performed withPSSpred (http://zhanglab.ccmb.med.umich.edu/PSSpred/)and PSI-PRED (http://globin.bio.warwick.ac.uk/psipred/)[34]. Comparative protein modeling of the toxin was per-formed in the Swiss-model server using the alignmentmode (http://swissmodel.expasy.org/workspace/) [5]. In
search for alternative structural templates from knownstructures in the PDB, and a sequence-to-structure align-ment of the toxin we used the metaserver LOMETS(http://zhanglab.ccmb.med.umich.edu/LOMETS/) [64]. Dipole
oment was calculated in the Protein Dipole Momentserver (http://bip.weizmann.ac.il/dipol) [27]. The predictedD-models of the toxin were subjected to a series of testsor evaluating its internal consistency and reliability. Back-one conformation was evaluated by the inspection of thesi/Phi Ramachandran plot obtained from PROCHECK anal-sis [39]. Packing quality of the 3D model was investigatedy the calculation of WHATCHECK Z-score value [32]. Finally,equence-structure compatibility was evaluated by VERIFY3D11]. PROCHECK, WHATCHECK and VERIFY3D were executed fromhe structure analysis and verification servers’ website at UCLAhttp://www.doe-mbi.ucla.edu/Services/SV/). In addition, we usedmean Z-score (http://swissmodel.expasy.org/Qmean) for thebsolute quality assessment of the peptide models [6].
. Results
.1. Bioassay-guided purification of the ASIC inhibitor
The resulting aqueous extract (350 mg/90 ml 0.1 M ammoniumcetate), obtained from 5 g of the sea anemone homogenate, wasubsequently loaded onto a Sephadex G-50M gel filtration columnFig. 1A). The Sephadex G-50M profile was divided into four sec-ions (I–IV), from which only II and III covered the peptide fraction,ccording to a previous calibration [52]. Considering that knownroteinaceous ASIC inhibitors have molecular masses of less than0 kDa [20,21,25] only II and III were selected for the evaluation onSICs from dorsal root ganglia neurons.
Both fractions inhibited ASIC currents. However, as fraction IIIVe = 1440–1920 ml) showed a significantly higher inhibitory effect,t was selected as starting sample for the subsequently purificationteps. Its application during the pH drop with the pH 6.1 pulse (5 s)educed the peak amplitude of ASIC currents in 36.8 ± 5.2% (n = 3)n the assayed concentration (500 �g/ml).
Pool III from Sephadex G-50 was submitted to cation-exchangehromatography considering that ASIC peptide toxins PcTx1 [25]nd APETx2 [20] are very basic peptides. Seven pools of cation-xchange fractions were collected (Fig. 1B) and tested on ASICurrents. The application of these fractions at pH 6.1 reducedhe peak amplitude of ASIC currents at the concentration assayed250 �g/ml) but the higher effect was observed for adjacent frac-ions III-4CEX, III-5CEX and III-6CEX (more than 60% currentnhibition), which therefore were selected for reversed-phase chro-
atographic fractionation. III-4CEX, III-5CEX and III-6CEX, wereooled and submitted to reversed-phase C18 chromatographyFig. 1C). A major fraction (III-4,5,6-CEX-RP1), which eluted at2.6 min, showed inhibitory activity on ASIC currents. The amountf toxin obtained was 0.22 mg, representing 0.063% of the total sol-ble protein content (350 mg). The purity of the toxin was tested in
different analytical reversed-phase C18 chromatographic columnsing a very shallow gradient of acetonitrile. The detection of onlyne peak indicated a high purity of the toxin (Fig. 1D).
.2. Biological evaluation of the toxin
A total of 81 DRG neurons were successfully voltage-clampedor a sufficient time to allow the study of the toxin. The neu-ons examined had a capacitance in the range of 18–90 pFmean = 55 ± 14.6 pF), which corresponds to an estimated celliameter in the range of 24–53 �m (mean = 42 ± 6 �m), i.e., the DRGeurons used for the voltage clamp recordings were basically neu-
ons with a large size cell body, according with the values reportedy Lawson et al. [40].
The ASIC current characteristics in DRG neurons were similaro those reported previously by our group and by other authors
ides 53 (2014) 3–12
[19,23,28], namely: (a) fast transient currents with a small steady-state current; (b) currents with a fast desensitization time course(measured by �des) and an obvious steady-state current, and (c)currents with a slow desensitization time course and a large steady-state component. These characteristics suggest that diverse ASICsubunits participate in the macroscopic currents recorded in ourexperiments. No correlation was found between the cell diameter(calculated from cell capacitance measurements), desensitizationtime course or steady-state component and the effect of the toxin,therefore no classification of the ASIC current subtypes was done,and we used the whole population of cells recorded to constructthe concentration–response curves.
The mean values for the parameters measured on ASIC currentsactivated at pH 6.1 were: Imax = 5.3 ± 0.8 nA, �des = 297.6 ± 30.2 ms,Iss = 0.1 ± 0.02 nA.
The sustained application of the toxin (n = 35) 20 s beforethe pH change from 7.4 to 6.1 and during the whole pH 6.1pulse (5 s) reduced the peak amplitude of the ASIC currents(Fig. 2A) in the assayed concentrations, from 0.01 �M to 10 �M,in a concentration-dependent manner (Fig. 2B). The toxin inhi-bition on ASIC current was statistically significant from 0.03 �Mto 10 �M (P ≤ 0.05, Student’s t test). The concentration–responserelationship had an IC50 of 100 ± 18 nM and a Hill coeffi-cient = 1.2 ± 0.2. At the IC50 value the inhibition was statisticallysignificant (P ≤ 0.05, Student’s t-test) with a value of 22 ± 3%,whereas the maximal inhibition was observed at the highest con-centration (10 �M) and it was about 44 ± 7%. Both the currentdesensitization rate and the sustained component were not sig-nificantly affected at any concentration. For example, the toxin at0.3 �M decreased the peak current amplitude 28.7 ± 5.6% whereasno significant effect was observed on the current desensitiza-tion rate (control �des = 300.4 ± 47.4 ms versus toxin presence�des = 328.4 ± 40.9 ms, P > 0.05, Student’s t-test, n = 9), neither on thesustained component (control amplitude = 0.08 ± 0.01 nA versustoxin amplitude = 0.09 ± 0.02 nA, P > 0.05, Student’s t-test, n = 9).The inhibitory effect took place fast, reaching the steady-state inabout 2 min. The effect of the toxin on the ASIC current peak ampli-tude was fully reversible in about 5 min even at the maximumconcentration (10 �M) in which the reversibility was of 97 ± 2.5%.
The application of the toxin during the pH drop at 10 �M (n = 8)had no significant effect on the current peak amplitude, �des, neitheron the sustained component of the ASIC current. At this concen-tration, the toxin diminished the current peak amplitude by only4.5 ± 1.4% (P > 0.05, Student’s t-test) (Fig. 2C), thus indicating thateffect of the toxin takes place preferentially in the closed channelconformation.
Given the large number of Nav and Kv channel toxins so farisolated from sea anemones, and the existence of promiscuous tox-ins in these organisms [41,45], including APETx2 the best knownASIC inhibitor isolated from sea anemone, which also acts on Na+
currents, we explored the activity of the toxin on voltage-gated Na+
or K+ currents in DRG neurons. Voltage-gated Na+ currents (mea-sured at the peak current and at the end of a 40 ms voltage pulsefrom a Vh of −90 mV to a test potential of −10 mV) were not sig-nificantly affected by the toxin even at 10 �M (data not shown),however, it significantly inhibited the K+ current in a concentration-dependent manner with an IC50 = 3.4 ± 0.2 �M and 3.5 ± 0.7 �M, forthe peak current and the steady-state component of the current,respectively (n = 31) (Supplementary Fig. S1 A–C). The maximuminhibitory effect of the toxin on K+ currents was reached in about3 min, showing an inhibition of 16 ± 2.9% at 30 �M, for the peak cur-rent and 19 ± 3.0% for the steady state current, being fully reversible
for the peak current, and in 90 ± 2% for the steady state current5 min after washing.
No paralyzing effect was observed after the administration ofthe toxin to crabs. Moreover, none of the ion-exchange pools of
Fig. 1. Chromatographic profiles corresponding to the bioassay-guided purification of PhcrTx1. The biological screening of chromatographic fractions was done on ASICcurrents from DRG neurons. (A) Gel filtration profile of the sea anemone aqueous extract. The soluble material contained in 5 g of the total homogenate (350 mg/90 ml)was fractionated in Sephadex G-50 (5 cm × 93 cm) at 2 ml/min in 0.1 M ammonium acetate Fractions of 20 ml each were collected and grouped (I to IV) as shown in thefigure. The 1–10 kDa polypeptides were located within II and III, the latter showed the highest inhibitory activity (signalized with *). (B) Chromatographic profile of III,from Sephadex G-50, in the cation-exchanger Fractogel EMD SO3
− 650 M. The separation was performed at 1 ml/min, using a 400 ml gradient from 0.01 to 1 M ammoniumacetate (pH 7). Fractions of 5 ml each were collected and grouped into seven pools (III-1CEX to III-7CEX). III-4CEX, III-5CEX and III-6CEX showed the highest inhibitory effect(*) on ASIC currents. Gradient development is signalized by a discontinuous line. (C) Reversed-phase profile of the mixture of III-4CEX, III-5CEX and III-6CEX from cation-e founda nalytit aphic
ciw
3
md3
FAdPpod
xchange chromatography. A main chromatographic fraction, III-4,5,6CEX-RP1, was discontinuous line. (D) The purity of III-CEX4,5,6-RP1was checked in a different ahe milliabsorbance units at 214 nm. The toxin characterized from this chromatogr
hromatographic fractions, including those showing the highestnhibitory activity on ASIC currents (III-4CEX, III-5CEX and III-6CEX)
as toxic to crabs.
.3. Structural and computational analyses
The pure peptide was analyzed by MALDI-TOF and its averageolecular mass value was 3477 Da. Automated N-terminal degra-
ation of the toxin yielded an amino acid sequence composed of2 residues, CASQGQKCKTKSDCCNGMWCAGTRGHTCYKPK, with a
ig. 2. Biological evaluation of PhcrTx1 on ion channels from DRG neurons by the whole and C, the gray bars above the traces represent the protocol employed for the applicatioenote Student’s t-test significance: *P < 0.05. (A) Representative experiment showing thhcrTx1 and washout of toxin. The toxin was applied 20 s before the pH change and duarticular cell PhcrTx1 reduced the peak amplitude in 48.5%. The recovery after washout
n the ASIC currents. The IC50 was 100 ± 18 nM (n = 35). (C) Representative experiment shrop (no significant effect) and after toxin washout.
to inhibit ASIC currents in rat DRG neurons. Gradient development is signalized bycal reversed-phase column using a shallower gradient. The vertical axis representsfraction was subsequently named PhcrTx1.
theoretical average molecular mass of 3476.03 Da, being in goodagreement with the experimental molecular mass, assuming theformation of three disulfide bridges. The peptide was namedPhcrTx1 and its sequence was deposited at the UniProt Knowledge-base under the accession number C0HJB1. This peptide could benamed as �-PMTX-Pcf1a according to the nomenclature recently
proposed by Oliveira et al. [44].
PhcrTx1 is a very basic peptide (theoretical pI 10.89), as previ-ously observed for ASIC toxins PcTx1 (pI 10.38) [25] and APETx2(pI 9.59) [20] using the same software (GPMAW) for pI calculation.
-cell patch clamp technique. This toxin inhibited ASICs (preferentially) currents. Inn of the toxin, the dotted lines indicate the zero current. In B, E and F, the asteriskse acid gated current under control condition, after sustained application of 10 �Mring the whole pH pulse (pH 6.1, 5 s) holding the cell voltage at −60 mV. In this
was about 94%. (B) Concentration–response curve of the PhcrfTx1 inhibitory effectowing control ASIC current, after the application of 10 �M PhcrTx1 during the pH
8 A.A. Rodríguez et al. / Peptides 53 (2014) 3–12
Fig. 3. Multiple sequence alignment of PhcrTx1 and related spider Inhibitor Cystine Knot (ICK) toxins with 3-D structure determined by NMR: �-theraphotoxin-Gr3a(UniprotKB: P60980), �-hexatoxin-Mg1a (UniprotKB: P83560), U5-theraphotoxin-Hh1a (UniprotKB: Q86C51), �-theraphotoxin-Gr1a (UniprotKB: P60590), �-hexatoxin-Ar1a (UniprotKB: P01478), �-hexatoxin-Hv1a (UniprotKB: P13494), �-theraphotoxin-Pg1a (UniprotKB: P84835), �-theraphotoxin-Gr1a (UniprotKB: Q7YT39) and �/�-theraphotoxin-Hh1a (UniprotKB: P56676). Moreover, �-theraphotoxin-Pc1a (PcTx1, UniprotKB: P60514) was included since it represents the only known ICK toxin withinhibitory activity on ASICs. �-Strand (arrow) and �-helix (helix) are represented above the amino acid sequence segments involved in the formation of these structures.E rTx1, sc of ideo s used
Tutfnnataawtcsthi(t
tGftShh�
taPQt
very PDB code is indicated above its corresponding toxin name. In the case of Phconnectivity of the ICK toxins is represented below, I–IV, II–V, III–VI. The percentagef PhcrTx1 with ICK toxins in Dalign (right side). Jalview, www.jalview.org [63], wa
he search for related sequences annotated in UniProt/Swiss-Protsing PSI-BLAST showed no significant hits with scores better thanhreshold. The search against the Pfam database of protein domainamilies showed 3 hits related to PhcrTx1 sequence: (i) Ion chan-el inhibitory toxin (Toxin 12, PF07740, E-value = 0.0073); (ii) PhTxeurotoxin (Toxin 29, PF08116, E-value = 0.045); and (iii) Deltatracotoxin (Atracotoxin, PF05353, E-value = 0.11). All these pep-ide families are composed of animal venom toxins that containn Inhibitor Cystine Knot (ICK) motif [15]. A multiple sequencelignment was constructed with PhcrTx1 and the ICK peptidesith known 3D structures, from the Pfam families above men-
ioned (Fig. 3). These peptides are mostly voltage-gated (Na+ or K+)hannels toxins isolated from spider venoms, and share 26–50%equence identity with PhcrTx1. PcTx1 (�-theraphotoxin-Pc1a),he only ICK toxin known to inhibit an ASIC channel (specificallyomomultimeric ASIC1a) was also included in the MSA. The result-
ng alignment suggested that PhcrTx1 has a cystine connectivityI–IV, II–V, III–VI) presumably associated to �-strands, similarly tohe other ICK toxins.
Aiming to predict the three-dimensional structure of the peptidehe two most significant templates (Fig. 3), �-therophotoxin-r3a (E = 0.0007) and �-hexatoxin-Mg1a (E = 0.0019), were selected
or constructing structural models in Swiss-model server usinghe alignment mode. Additionally, the LOcal MEta-Threading-erver (LOMETS) automatically generated several models withigh confidence score, based on �-therophotoxin-Gr3a and �-exatoxin-Mg1a among other ICK toxin structures, showing-therophotoxin-Gr3a as the top ranked template.
The quality of the structural models was evaluated by checkinghe similarities between every model and its respective template,
ccording to the calculated parameters obtained with programsROCHECK, WHATCHECK and VERIFY3D, from (Table 2). Themean Z-score was also used to evaluate the absolute quality of
he structural models. In general, the PhcrTx1 model yielded by
econdary structure was predicted by PSSpred, PSI-PRED and Swiss-model. Cystinentity and similarity as well as the E-value were obtained from pairwise alignments
for illustrating conserved amino acids.
Swiss-model based on �-therophotoxin-Gr3a (PDB code 1S6X) wassuperior to the others, taking into account the similarity betweenmodel and template, as well as the absolute quality of the model(Table 2). Fig. 4A shows the PhcrTx1 model, mainly composed of along N-terminal loop, and a hairpin � motif formed by two antipar-allel �-strands, Trp19-Cys20-Ala21 (positions 23–25 in the MSA)and His26-Thr27-Cys28-Tyr29 (positions 33–36 in the MSA), linkedby the hairpin loop Gly22-Thr23-Arg24-Gly25 (positions 26–29 inthe MSA). Both �-strands are approximately located in the sameregions containing �-strands in �-theraphotoxin-Gr3a and theother ICK toxins, as well as in those sequence segments predictedby PSSpred and PSI-PRED (Fig. 3). Disulfide bridges are Cys1-Cys15,Cys8-Cys20 y Cys14-Cys28, corresponding to the cystine connectiv-ity (I–IV, II–V, III–VI) found in ICK toxins.
Several interaction models used for predicting the putativefunctional surface of toxins acting on ion channels have beensummarized by Escoubas et al. [24]. These are based on (1) theorientation of the dipole moment resulting from the electrostaticanisotropy of the toxin, (2) the hot spot model involving a functionaldyad composed of a critical lysine (implicated in pore blocking)and a neighboring aromatic residue such as tryptophan or phen-ylalanine, and (3) another hot spot model involving a patch ofhydrophobic residues forming a contact surface surrounded bycharged residues anchoring the toxin to its target surface throughformation of salt bridges. The orientation of the dipole momenthas been used for predicting the functional surface of the ASICtoxins PcTx1 [24] and APETx2 [13], which target different ASICsubunits and have no sequence homology between them. In bothcases a putative interaction surface mainly comprising a patch ofbasic and aromatic residues was proposed. Recently, the struc-
ture of a 3PcTx1–ASIC1 complex was elucidated, showing that allthree PcTx1 molecules exhibit the same binding mode using ahydrophobic patch and a basic cluster for interaction with ASIC1[17]. Such hydrophobic/basic group includes all residues previously
Table 2Model quality assessment. Different structural models of the toxin PhcrTx1, obtained by Swiss-model and LOMETS servers, were evaluated by comparing several parameterscalculated in PROCHECK, WHATCHECK, VERIFY3D with values obtained from their respective templates, �-theraphotoxin-Gr3a (UniprotKB P60980, PDB 1S6X) and �-hexatoxin-Mg1a (UniprotKB P83560, PDB 2ROO). NA: not applicable to templates, only for structural models.
Program Characteristic 3-D structural models Templates
≥0.2)Normalized Qmean (Z-score) Absolute quality of model
(correct values: from −1 to 1)−0.46
roposed to form the binding surface of PcTx1, thus supporting therediction through the orientation of the dipole moment and theot spot model of functional dyads.
Fig. 4B shows the resulting dipole moment calculated forhcrTx1 model, represented by an arrow that emerges betweenhe residues Lys11 and Trp19 separated by approximately 6.7–7.2 A,
hus being within the measured distance (6.6 ± 1 A) in the dyadys-(Tyr or Phe) proposed to be essential, in scorpion and seanemone toxins, for binding Kv channels [16,35]. Moreover, the
ig. 4. Molecular modeling of PhcrTx1. (A) Cartoon representation of the model. �-Strandnd C14-C28) are signalized. (B) Stick representation of a cluster mainly composed of bao previous studies in ASIC [13,24] and Kv toxins [35]. Blue: basic residues, magenta: arohrough the dyad K11/W19. The distances (A) between C� of the basic residues and the acTx1 structure determined by NMR, respectively. Both ASIC toxins have in common a clcTx1 was supported by the crystallographic structure of 3PcTx1–ASIC1a [17]. PyMol, htttructural representations were rendered by POV-Ray (version 3.6 by Persistence of Visio
−1.08 −0.45 −2.06 NA NA
hydrophobic residue Tyr29 is located between two near basicresidues, Arg24 and Lys32, at approximately 6.2–6.7 A and4.2–4.9 A respectively, from the C� of the basic residue tothe center of the aromatic ring. Another basic residue, Lys32,together with Lys30, Arg24 and Lys11, surrounds the aromaticresidues Trp19 and Tyr29 to form a basic/aromatic surface
(Fig. 4C) as previously described for ASIC toxins PcTx1 (towhich PhcrTx1 is structurally closer) and APETx2. Fig. 4D showsthe basic/aromatic cluster proposed as the main component
s (WCA, 19–21, and HTCY, 26–29) and disulfide bridges (ICK type: C1-C15, C8-C20sic/aromatic dyads, which might be involved in ion channel recognition accordingmatic residues. The dipole moment is represented by a cyan arrow that emerges
romatic rings are indicated. (C and D) Surface representation of PhcrTx1 model anduster of basic/aromatic residues. Recently, the prediction of the binding surface ofp://www.pymol.org [18], was used for viewing the 3-D models and templates. Then Raytracer, Pty., Ltd.).
f the channel recognition surface of PcTx1, comprising theasic residues (Lys25-Arg26-Arg27-Lys28) forming a contiguousositive surface located in the �-turn linking the two �-strands,nd neighboring aromatic residues (Trp7, Trp24, Phe30) [24].
. Discussion
In the present work we discovered a new ASIC toxin, PhcrTx1,hich represents the first compound isolated and characterized
rom the sea anemone species P. crucifer and its family, Phyman-hidae. To date, sea anemone toxins have been isolated from other
different families [44].
.1. PhcrTx1 inhibits ASIC and Kv channels in dorsal root ganglioneurons
This peptide partially inhibited ASIC peak currents in rat DRGeurons with an IC50 ≈ 100 nM, without affecting the sustainedomponent of the current (Iss) nor the current desensitization rate.his is similar to the effect described for amiloride, a non-specificSIC channel inhibitor. It has been suggested that amiloride blocks
he channels by entering into the conducting pathway and com-eting for its binding site with sodium cations [49]. However,he application of PhcrTx1 during the pH drop had no significantffect on the ASIC currents. Under our experimental conditionshe toxin must be applied 20 s before the acidic pulse to exert itsction on ASIC currents, a time that seems too short to allow thenternalization of the toxin. Furthermore, the effect of the toxin
as completely reversible after washout of the preparation. Takenogether, these data suggest that the effects of the toxin on the ASICurrents depend on the channel state (i.e., closed or open), indicat-ng that PhcrTx1 has a higher affinity for the closed state of thehannel, and that it acts on cell-surface targets.
PhcrTx1 has a theoretical isoelectric point of 10.89. At pH 7.4ts net charge would be +5.03. This is in agreement with the calcu-ated charge for ASIC toxins APETx2 (net charge = +2.00 at pH 7.4)nd PcTx1 (net charge = +3.00 at pH 7.4), and also with the chargeeported for other ASIC modulators such as FMRFamide-relatedeptides [14] and aminoglycosides [29]. These positively chargedolecules may interact with some of the functional domains of theSIC extracellular surface that have been related with its sensitiv-
ty to protons, its Ca2+ sensitivity, and its capability to accumulateations in the extracellular vestibule [30,33]. Additional mutagen-sis experiments are necessary to test this hypothesis.
All known ASIC subunits are present in DRG neurons1,3,19,60,61], and they combine to form homomeric or het-romeric channels that contribute (excepting ASIC4 and the ASIC2bomomultimer) to the total ASIC currents [1,7,19,31,60], dependingn the properties of the constituent subunits. Known peptides withnhibitory activity on ASIC channels exert their effects on particularSIC currents in DRG neurons according to their affinity for spe-ific subunits [20,21,25]. PhcrTx1 inhibited ASIC currents (maximalnhibitory action of 44 ± 7%) independently of their characteris-ics, thus in this study it was no possible to establish a correlationetween the kinetic properties of the current and the toxin effect.n fact we mostly recorded from DRG neurons, which are known toxpress ASIC1 subunits, but also ASIC2 and ASIC3 subunits. Somef these neurons coexpress ASIC2 and ASIC3 forming heteromerichannels [3]. Therefore, studies in heterologous expression systemsill be necessary to determine the toxin selectivity to specific ASIC
ubunits.
PhcrTx1 also inhibited voltage-gated K+ (Kv) currents in DRG,
ut with significantly lower potency and efficacy than the inhi-ition on ASIC currents. DRG neurons express a wide varietyf Kv channels, including rapid inactivation (IKA) and delayed
ides 53 (2014) 3–12
rectifier (IKDR) channel groups. We studied the effects of the toxinon total Kv currents in which we identified both, a transientcomponent presumably related to IKA contribution and a sustainedcomponent related to IKDR contribution. The toxin affected bothcurrent components without preferential inhibitory activity on anyof them; no protocols were used for dissecting the effects of thetoxin on a particular Kv current subtype. PhcrTx1 represents a newtoxin with inhibitory activity on ASICs, which constitute its maintarget, and it also exhibits a certain degree of concentration depend-ent inhibitory activity on Kv channels. Other sea anemones toxinsexhibiting activity on distinct ion channels families are APETx1, 2[9,20,45] and BDS-I [22,41,65]. These examples illustrate that toxinspecificity in several cases is not absolute, as previously observedin toxins from other venomous animals, e.g. spiders [26,50].
However, PhcrTx1 had no effect on Na+ voltage-gated currentsin DRG neurons, even at high concentration (30 �M). More-over, no paralyzing effects on crabs at the maximal dose of2000 �g/kg were observed. These results suggest that PhcrTx1 isnot active on crustacean Nav channels neither it is on mammalianNav subtypes expressed in DRG neurons including TTX-sensitive(Nav1.1, -1.2, -1.3, -1.6 and -1.7) and TTX-resistant (Nav1.8, and-1.9) [12]. The inhibitory activity of PhcrTx1 on ASICs does notseem to be related to neurotoxic effects in vivo as previouslyobserved for other ASIC inhibitors [20,21]. A variety of sea anemonepeptides, mostly Nav channels toxins, also Kv channel toxins andprotease inhibitors, as well as several other toxins acting on stillunknown targets, has been detected through the well-known crabbioassay [54]. Although many more sea anemone toxins could bediscovered through this useful bioassay, ASIC inhibitors are likelyto escape from detection and therefore in vitro assays are prefer-able for ASIC inhibition screening in bioassay-guided purificationprotocols.
4.2. PhcrTx1 is a novel sea anemone toxin
PhcrTx1 shares low sequence identity with the ASIC1a toxinPcTx1, and no significant sequence identity to other known ASICinhibitors, neither it is related to sea anemone peptides so farreported. On the other hand, PhcrTx1 sequence is related to spi-der toxins acting on Nav, Kv and Cav channels. Multiple sequencealignment and structure prediction indicated that this toxin has anInhibitor Cystine Knot (ICK) motif, which has been found in tox-ins from other venomous animals such as spiders (including PcTx1,the only known ICK toxin acting on ASICs [24,25]), cone snails andscorpions [67]. Up to date, known structural scaffolds present in ionchannel toxins from sea anemones comprise defensin (type 1 and2 Nav channel toxins, type 3 Kv channel toxins, ASIC3 toxin), ��(type 1 Kv channel toxins), Kunitz (type 2 Kv channel toxins) andthe unique case of Av3, a Na+ channel toxin with no regular �-helixor �-sheet structure [42,43]. Therefore, PhcrTx1 defines a new fam-ily of sea anemone peptide toxins exhibiting inhibitory activity onASIC channels; to the best of our knowledge this peptide representsthe first ICK toxin characterized from a sea anemone.
The structural model of PhcrTx1 contains several featuresresembling binding sites of ASIC and Kv channel toxins. Similarlyto other ASIC toxins such as PcTx1 and APETx2, the PhcrTx1model contains a basic/hydrophobic cluster composed of theresidues Lys11, Arg24, Lys30, Lys32, Trp19 and Tyr29, which mightbe involved in ASIC channels recognition. Most of these residuesare grouped into a basic/aromatic dyad (Lys11/Trp19) and a triad(Arg24/Tyr29/Lys30) with similarities to functional dyads involvedin Kv channel inhibition, present in sea anemone and scorpion
toxins [16,59]. This pattern found in Kv toxins is composed ofLys/Tyr or Lys/Phe residues separated at the distance of 6.6 ± 1 A.Particularly, the dyad formed by consecutive residues Lys and Tyr(Lys30/Tyr29 in PhcrTx1), has been found in Kv toxins such as
/ Pepti
d(rdccaa
cppptgiadwoaauvi
5
atRtttita
A
CJtPSSt
tFloaraPmC1MSJm
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
A.A. Rodríguez et al.
endrotoxin (Lys3/Tyr4), BgK (Lys25/Tyr26), HanatoxinLys22/Phe23), and ShK (Lys22/Tyr23), as previously summa-ized [59]. Although the PhcrTx1 structure was not experimentallyetermined, nor its specificity of action on ASIC subunits and Kv
hannel subtypes, the theoretical structural data herein providedould be used as starting point for selective mutagenesis studiesimed to determine the functional residues involved in inhibitoryctivity on ASIC and Kv channels.
It has been proposed that the ancestral two-disulfide foldalled “disulfide-directed �-hairpin (DDH) fold” is the evolutionaryrecursor of the ICK motif [62]. Recently, this hypothesis was sup-orted in arachnids by the discovery of a unique scorpion-venomeptide (U1-LITX-Lw1a) with a molecular scaffold correspondingo the previously hypothesized DDH fold [57]. Another study sug-ested that the disulfide arrangement (I–III, II–V, IV–VI) presentn a spider toxin (HWTX-II), which is different from ICK motif,lso evolved from the DDH folding [56]. Given that sea anemonesate back even further than other invertebrates in evolution itould be interesting to verify whether ICK motif has the same
rigin (DDH folding) in sea anemone toxins. The presence of thisncient fold (DDH) in toxins isolated from these organisms wouldlso open the possibility of finding several other derived molec-lar scaffolds, for example those similar to HWTX-II, exhibiting aariety of biological activities of pharmacological and biomedicalnterest.
. Conclusions
We isolated and characterized PhcrTx1, a peptide that defines novel class of ASIC inhibitors from sea anemones. This pep-ide represents the first ICK toxin isolated from these organisms.ecent works have shown that sea anemones secretions con-ain a complex mixture including hundreds of peptides. Most ofhis diversity remains unexplored so we expect that other ICKoxins as well as new molecular scaffolds can be found, thus rais-ng the interest in finding novel biologically active peptides fromhese organisms, with potential pharmacological and biomedicalpplications.
cknowledgements
We are grateful to Adys Palmero Colmenares and Estrellauquerella for their technical assistance; the divers Luis Alejandre,
osé Ramón García and José Ramón Martinez Guerra for collectinghe sea anemone specimens. A.A. Rodríguez specially thanks Dr.eter Højrup for his very kind gift of the software GPMAW, Prof.ona Vasudevan for her helpful suggestions, as well as Prof. Joel L.ussman and Dr. Clifford Felder for their kind explanations abouthe dipole moment server.
This work was supported by the International Founda-ion for Science (travel grant and research grants F/4082-1,/4082-2), the European Molecular Biology Organization (fel-owships 167-2007 and 126-2009), the Third World Academyf Sciences (Fellowship for research and advanced trainingpplication and Research grant 06344-2007), the Brazilianesearch agencies FAPESP (07/56525-3), CNPq (490194/2007-9nd 481565/2009-4) and CAPES-Toxinologia, the Ibero-Americanrogram for Science, Technology and Development (CyTED, the-atic network 212RT0467 BIOTOX), CONACyT, Mexico (Mexico-
uban Integral Research Inter Exchange and research grant69835), VIEP-BUAP, Mexico (grant 2012-00316) and PIFI,
exico (grant 2013) and PIFI-PROMEP-Red: “Instrumentación-de-
ensores-para-aplicaciones-de-Fisiología-y-Biomedicina”, Mexico..T. was supported by the FP7 ‘MAREX’ grant. None of the above
entioned organizations were involved in the study design,
[
[
des 53 (2014) 3–12 11
collection, analysis and interpretation of data, the writing ofthe manuscript or the decision to submit the manuscript forpublication.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,in the online version, at http://dx.doi.org/10.1016/j.peptides.2013.06.003.
References
[1] Akopian AN, Chen CC, Ding Y, Cesare P, Wood JN. A new member of the acid-sensing ion channel family. Neuroreport 2000;11:2217–22.
[2] Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, et al. GappedBLAST and PSI-BLAST: a new generation of protein database search programs.Nucleic Acids Res 1997;25:3389–402.
[3] Alvarez de la Rosa D, Zhang P, Shao D, White F, Canessa CM. Functional implica-tions of the localization and activity of acid-sensitive channels in rat peripheralnervous system. Proc Natl Acad Sci USA 2002;99:2326–31.
[4] Anderluh G, Macek P. Cytolytic peptide and protein toxins from sea anemones(Anthozoa: Actiniaria). Toxicon 2002;40:111–24.
[5] Arnold K, Bordoli L, Kopp J, Schwede T. The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics2006;22:195–201.
[6] Benkert P, Biasini M, Schwede T. Toward the estimation of the absolute qualityof individual protein structure models. Bioinformatics 2011;27:343–50.
[7] Benson CJ, Xie J, Wemmie JA, Price MP, Henss JM, Welsh MJ, et al. Heteromulti-mers of DEG/ENaC subunits form H+-gated channels in mouse sensory neurons.Proc Natl Acad Sci USA 2002;99:2338–43.
[8] Béress L, Béress R. Reinigung zweier krabbenlähmender Toxine aus derSeeanemone Anemonia sulcata. Kiel Meeresforsch 1971;27:117–27.
[9] Blanchard MG, Rash LD, Kellenberger S. Inhibition of voltage-gated Na+ currentsin sensory neurones by the sea anemone toxin APETx2. Br J Pharmacol2012;165:2167–77.
10] Bohlen CJ, Chesler AT, Sharif-Naeini R, Medzihradszky KF, Zhou S, King D, et al.A heteromeric Texas coral snake toxin targets acid-sensing ion channels toproduce pain. Nature 2011;479:410–4.
11] Bowie JU, Luthy R, Eisenberg D. A method to identify protein sequences thatfold into a known three-dimensional structure. Science 1991;253:164–70.
12] Catterall WA, Goldin AL, Waxman SG. International Union of Pharmacology.XLVII. Nomenclature and structure-function relationships of voltage-gatedsodium channels. Pharmacol Rev 2005;57:397–409.
13] Chagot B, Escoubas P, Diochot S, Bernard C, Lazdunski M, Darbon H. Solu-tion structure of APETx2, a specific peptide inhibitor of ASIC3 proton-gatedchannels. Protein Sci 2005;14:2003–10.
14] Chen X, Paukert M, Kadurin I, Pusch M, Grunder S. Strong modulation byRFamide neuropeptides of the ASIC1b/3 heteromer in competition with extra-cellular calcium. Neuropharmacology 2006;50:964–74.
15] Craik DJ, Daly NL, Waine C. The cystine knot motif in toxins and implicationsfor drug design. Toxicon 2001;39:43–60.
16] Dauplais M, Lecoq A, Song J, Cotton J, Jamin N, Gilquin B, et al. On the conver-gent evolution of animal toxins. conservation of a diad of functional residuesin potassium channel-blocking toxins with unrelated structures. J Biol Chem1997;272:4302–9.
17] Dawson RJ, Benz J, Stohler P, Tetaz T, Joseph C, Huber S, et al. Structure of theacid-sensing ion channel 1 in complex with the gating modifier psalmotoxin 1.Nat Commun 2012;3:936.
18] DeLano W. The PyMOL molecular graphics system. San Carlos, CA, USA: DeLanoScientific; 2002.
19] Deval E, Noel J, Lay N, Alloui A, Diochot S, Friend V, et al. ASIC3, a sensor of acidicand primary inflammatory pain. EMBO J 2008;27:3047–55.
20] Diochot S, Baron A, Rash LD, Deval E, Escoubas P, Scarzello S, et al. A new seaanemone peptide, APETx2, inhibits ASIC3, a major acid-sensitive channel insensory neurons. EMBO J 2004;23:1516–25.
21] Diochot S, Baron A, Salinas M, Douguet D, Scarzello S, Dabert-Gay AS, et al.Black mamba venom peptides target acid-sensing ion channels to abolish pain.Nature 2012;490:552–5.
22] Diochot S, Schweitz H, Beress L, Lazdunski M. Sea anemone peptides with aspecific blocking activity against the fast inactivating potassium channel Kv3.4.J Biol Chem 1998;273:6744–9.
23] Dube GR, Lehto SG, Breese NM, Baker SJ, Wang X, Matulenko MA, et al. Electro-physiological and in vivo characterization of A-317567, a novel blocker of acidsensing ion channels. Pain 2005;117:88–96.
24] Escoubas P, Bernard C, Lambeau G, Lazdunski M, Darbon H. Recombinant pro-duction and solution structure of PcTx1, the specific peptide inhibitor of ASIC1aproton-gated cation channels. Protein Sci 2003;12:1332–43.
25] Escoubas P, De Weille JR, Lecoq A, Diochot S, Waldmann R, Champigny G, et al.Isolation of a tarantula toxin specific for a class of proton-gated Na+ channels.J Biol Chem 2000;275:25116–21.
26] Escoubas P, Rash L. Tarantulas: eight-legged pharmacists and combinatorialchemists. Toxicon 2004;43:555–74.
[K, et al. Proteomics of the neurotoxic fraction from the sea anemone Buno-
2 A.A. Rodríguez et al.
27] Felder CE, Prilusky J, Silman I, Sussman JL. A server and database for dipolemoments of proteins. Nucleic Acids Res 2007;35:W512–21.
28] Garateix A, Salceda E, Menendez R, Regalado EL, Lopez O, Garcia T, et al.Antinociception produced by Thalassia testudinum extract BM-21 is mediatedby the inhibition of acid sensing ionic channels by the phenolic compoundthalassiolin B. Mol Pain 2011;7:10.
29] Garza A, Lopez-Ramirez O, Vega R, Soto E. The aminoglycosides modulate theacid-sensing ionic channel currents in dorsal root ganglion neurons from therat. J Pharmacol Exp Ther 2010;332:489–99.
30] Gonzales EB, Kawate T, Gouaux E. Pore architecture and ion sites in acid-sensingion channels and P2X receptors. Nature 2009;460:599–604.
31] Hesselager M, Timmermann DB, Ahring PK. pH dependency and desensitizationkinetics of heterologously expressed combinations of acid-sensing ion channelsubunits. J Biol Chem 2004;279:11006–15.
32] Hooft RW, Vriend G, Sander C, Abola EE. Errors in protein structures. Nature1996;381:272.
33] Jasti J, Furukawa H, Gonzales EB, Gouaux E. Structure of acid-sensing ion chan-nel 1 at 1.9 A resolution and low pH. Nature 2007;449:316–23.
34] Jones DT, Protein secondary structure prediction based on position-specificscoring matrices. J Mol Biol 1999;292:195–202.
35] Jouirou B, Mouhat S, Andreotti N, De Waard M, Sabatier JM. Toxin deter-minants required for interaction with voltage-gated K+ channels. Toxicon2004;43:909–14.
36] Kozlov S, Grishin E. The mining of toxin-like polypeptides from EST databaseby single residue distribution analysis. BMC Genom 2011;12:88.
37] Kozlov SA, Osmakov DI, Andreev YA, Koshelev SG, Gladkikh IN, MonastyrnayaMM, et al. Polypeptide toxin from sea anemone inhibiting proton-sensitivechannel ASIC3. Bioorg Khim 2012;38:653–9.
38] Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H,et al. Clustal W. and Clustal X version 2.0. Bioinformatics 2007;23:2947–8.
39] Laskowski RA, Rullmannn JA, MacArthur MW, Kaptein R, Thornton JM. AQUAand PROCHECK-NMR: programs for checking the quality of protein structuressolved by NMR. J Biomol NMR 1996;8:477–86.
40] Lawson SN, Caddy KW, Biscoe TJ. Development of rat dorsal root ganglion neu-rones. Studies of cell birthdays and changes in mean cell diameter. Cell TissueRes 1974;153:399–413.
41] Liu P, Jo S, Bean BP. Modulation of neuronal sodium channels by the seaanemone peptide BDS-I. J Neurophysiol 2012;107:3155–67.
42] Mouhat S, Jouirou B, Mosbah A, De Waard M, Sabatier JM. Diversity of folds inanimal toxins acting on ion channels. Biochem J 2004;378:717–26.
43] Norton RS. Structures of sea anemone toxins. Toxicon 2009;54:1075–88.44] Oliveira JS, Fuentes-Silva D, King GF. Development of a rational nomencla-
ture for naming peptide and protein toxins from sea anemones. Toxicon2012;60:539–50.
45] Peigneur S, Beress L, Moller C, Mari F, Forssmann WG, Tytgat J. A naturalpoint mutation changes both target selectivity and mechanism of action ofsea anemone toxins. FASEB J 2012;26:5141–51.
46] Peri S, Steen H, Pandey A. GPMAW – a software tool for analyzing proteins andpeptides. Trends Biochem Sci 2001;26:687–9.
47] Punta M, Coggill PC, Eberhardt RY, Mistry J, Tate J, Boursnell C, et al. The Pfamprotein families database. Nucleic Acids Res 2012;40:D290–301.
48] Qadri YJ, Rooj AK, Fuller CM. ENaCs and ASICs as therapeutic targets. Am JPhysiol Cell Physiol 2012;302:C943–65.
49] Qadri YJ, Song Y, Fuller CM, Benos DJ. Amiloride docking to acid-sensing ionchannel-1. J Biol Chem 2010;285:9627–35.
50] Redaelli E, Cassulini RR, Silva DF, Clement H, Schiavon E, Zamudio FZ, et al.Target promiscuity and heterogeneous effects of tarantula venom peptidesaffecting Na+ and K+ ion channels. J Biol Chem 2010;285:4130–42.
[
ides 53 (2014) 3–12
51] Rodriguez AA, Cassoli JS, Sa F, Dong ZQ, de Freitas JC, Pimenta AM, et al. Peptidefingerprinting of the neurotoxic fractions isolated from the secretions of seaanemones Stichodactyla helianthus and Bunodosoma granulifera. New membersof the APETx-like family identified by a 454 pyrosequencing approach. Peptides2012;34:26–38.
52] Rodriguez AA, Standker L, Zaharenko AJ, Garateix AG, Forssmann WG, BeressL, et al. Combining multidimensional liquid chromatography and MALDI-TOF-MS for the fingerprint analysis of secreted peptides from the unexplored seaanemone species Phymanthus crucifer. J Chromatogr B Analyt Technol BiomedLife Sci 2012;903:30–9.
53] Salceda E, Garateix A, Soto E. The sea anemone toxins BgII and BgIIIprolong the inactivation time course of the tetrodotoxin-sensitive sodium cur-rent in rat dorsal root ganglion neurons. J Pharmacol Exp Ther 2002;303:1067–74.
54] Shiomi K. Novel peptide toxins recently isolated from sea anemones. Toxicon2009;54:1112–8.
55] Shiomi K, Honma T, Ide M, Nagashima Y, Ishida M, Chino M. An epidermalgrowth factor-like toxin and two sodium channel toxins from the sea anemoneStichodactyla gigantea. Toxicon 2003;41:229–36.
56] Shu Q, Lu SY, Gu XC, Liang SP. The structure of spider toxin huwentoxin-IIwith unique disulfide linkage: evidence for structural evolution. Protein Sci2002;11:245–52.
57] Smith JJ, Hill JM, Little MJ, Nicholson GM, King GF, Alewood PF. Unique scor-pion toxin with a putative ancestral fold provides insight into evolution of theinhibitor cystine knot motif. Proc Natl Acad Sci USA 2011;108:10478–83.
58] Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, ProvenzanoMD, et al. Measurement of protein using bicinchoninic acid. Anal Biochem1985;150:76–85.
59] Srinivasan KN, Sivaraja V, Huys I, Sasaki T, Cheng B, Kumar TKS, et al. �-Hefutoxin1, a novel toxin from the Scorpion Heterometrus fulvipes with uniquestructure and function: importance of the functional diad in potassium channelselectivity. J Biol Chem 2002;277:30040–7.
60] Ugawa S, Ueda T, Yamamura H, Shimada S. In situ hybridization evidence forthe coexistence of ASIC and TRPV1 within rat single sensory neurons. Brain ResMol Brain Res 2005;136:125–33.
61] Voilley N, de Weille J, Mamet J, Lazdunski M. Nonsteroid anti-inflammatorydrugs inhibit both the activity and the inflammation-induced expres-sion of acid-sensing ion channels in nociceptors. J Neurosci 2001;21:8026–33.
62] Wang X, Connor M, Smith R, Maciejewski MW, Howden ME, Nichol-son GM, et al. Discovery and characterization of a family of insecticidalneurotoxins with a rare vicinal disulfide bridge. Nat Struct Biol 2000;7:505–13.
63] Waterhouse AM, Procter JB, Martin DM, Clamp M, Barton GJ. Jalview Version 2 –a multiple sequence alignment editor and analysis workbench. Bioinformatics2009;25:1189–91.
64] Wu S, Zhang Y. LOMETs: a local meta-threading-server for protein structureprediction. Nucleic Acids Res 2007;35:3375–82.
65] Yeung SY, Thompson D, Wang Z, Fedida D, Robertson B. Modulation of Kv3subfamily potassium currents by the sea anemone toxin BDS: significance forCNS and biophysical studies. J Neurosci 2005;25:8735–45.
66] Zaharenko AJ, Ferreira Jr WA, Oliveira JS, Richardson M, Pimenta DC, Konno
dosoma cangicum venom: novel peptides belonging to new classes of toxins.Comp Biochem Physiol D: Genom Proteom 2008;3:219–25.
67] Zhu S, Darbon H, Dyason K, Verdonck F, Tytgat J. Evolutionary origin of inhibitorcystine knot peptides. FASEB J 2003;17:1765–7.
