ARTICLE IN PRESS

0041-0101/$ - see

doi:10.1016/j.tox

$The protein$$Ethical st

version submitte�Correspondi

Brazil. Tel.: +5

E-mail addre

Toxicon 50 (2007) 1095–1104

www.elsevier.com/locate/toxicon

Purification, characterization and homology analysis ofocellatin 4, a cytolytic peptide from the skin secretion

of the frog Leptodactylus ocellatus$,$$

Anna Nascimentoa, Alex Chapeaurougec, Jonas Peralesc, Antonio Sebbenb,Marcelo V. Sousaa, Wagner Fontesa, Mariana S. Castroa,b,�

aBrazilian Center for Protein Research, Department of Cell Biology, University of Brasilia, Brasilia DF 70910-900, BrazilbToxinology Laboratory, Department of Physiological Sciences, University of Brasilia, Brasilia DF 70910-900, Brazil

cLaboratory of Toxinology, Department of Physiology and Pharmacodynamics, Oswaldo Cruz Institute,

Oswaldo Cruz Foundation, Av. Brasil, 4365, Rio de Janeiro RJ 21040-900, Brazil

Received 20 April 2007; received in revised form 27 July 2007; accepted 27 July 2007

Available online 3 August 2007

Abstract

Neobatrachia is the amphibian suborder with the largest number of species and a most important source of bioactive

peptides from frog skin secretions. However, 90% of the studies on this subject have been focused on the frog families

Hylidae and Ranidae, while very little is known about peptides of other families, like Leptodactylidae. Our work reports

for the first time the isolation and characterization of ocellatin 4 (GLLDFVTGVGKDIFAQLIKQI-NH2), a cytolytic

peptide from the skin secretion of the South American frog Leptodactylus ocellatus. While most cytolytic amphibian

skin peptides are selective for microorganisms and harmless for mammalian cells, the HC50 of ocellatin 4 against

human erythrocytes is 14.3mM. The interaction between ocellatin 4 and zwitterionic phospholipids in mammalian

plasma membranes may be favored by its neutral charge at pH 7.0. Ocellatin 4 also shows some antibacterial activity

(MICsE. coli and S. aureus ¼ 64 mM) and its sequence shares similarities with the only six leptodactylid peptides previously

known and with four peptides from Australian hylid frogs of the genus Litoria.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Antimicrobial; Frog; Hemolytic; Leptodactylus; Litoria; Peptide

front matter r 2007 Elsevier Ltd. All rights reserved.

icon.2007.07.014

sequence data reported in this paper will appear in the UniProt Knowledgebase under the accession number P85090.

atement: All co-authors have participated in the planning, execution and analysis of this work and have approved the final

d. All sources of financial support for this study have been disclosed.

ng author. Toxinology Laboratory, Department of Physiological Sciences, University of Brasilia, Brasilia DF 70910-900,

5 61 3307 2160; fax: +55 61 3274 1251.

ss: mscastro@unb.br (M.S. Castro).

ARTICLE IN PRESSA. Nascimento et al. / Toxicon 50 (2007) 1095–11041096

1. Introduction

The largest order of amphibians is the Anura,with about 4899 living species (Hoegg et al., 2004)belonging to at least 26 families (Darst andCannatella, 2004). According to morphologicalfeatures, anurans have been divided into thesuborders Archaeobatrachia and Neobatrachia,and Neobatrachia has been divided into the super-families Hyloidea and Ranoidea. Recent molecularanalyses have supported both divisions (Darst andCannatella, 2004).

Neobatrachia presents greater species diversitythan Archaeobatrachia: about 4693 species against206 (Hoegg et al., 2004). The skin secretions of frogsbelonging to different species contain antimicrobialpeptides with distinct sizes, charges, hydrophobi-city, conformation and spectrum of action (Dudaet al., 2002; Nascimento et al., 2003), so Neoba-trachia is by far the richest source of frogantimicrobial peptides (Vanhoye et al., 2003).Currently, there are 518 entries for Neobatrachiaskin secretion peptides and proteins in the UniProtdatabase. However, 90% of these concern thefamilies Hylidae (258 entries) and Ranidae (210entries), while little is known about peptides of theother families, of which there are at least 16.

Leptodactylidae is one of the neglected families,with only six peptides studied up to now: fallaxin(Leptodactylus fallax—Rollins-Smith et al., 2005);laticeptin (Leptodactylus laticeps—Conlon et al.,2006a); ocellatins 1, 2 and 3 (Leptodactylus ocella-

tus—Nascimento et al., 2004) and pentadactylin(Leptodactylus pentadactylus—King et al., 2005).Leptodactylus laticeps, L. ocellatus and L. pentadac-

tylus are South American species, while L. fallax isCentral American. The genus Leptodactylus contains65 distinct species; however, conclusive evidenceabout its monophyly is still missing (Frost, 2004).

Obtaining peptide datasets of the various frogfamilies may be of great importance to increaseknowledge of amphibian biology and to improvethe design of antimicrobial agents aimed at resistantpathogens. In the light of this, we have beenstudying the peptide contents of the skin secretionsof the South American frog L. ocellatus (‘‘Argusfrog’’—Neobatrachia, Hyloidea, Leptodactylidae,Leptodactylinae). The present work reports acontinuation of our previous investigations thatled to the isolation of ocellatins 1–3. Herein, wedescribe the recent purification and characterizationof a fourth peptide, ocellatin 4.

Most amphibian skin secretion cytolytic peptidesare highly cationic, which causes them to interactwith negative charges exposed on bacterial mem-branes, but not with the zwitterions present in theextracellular layer of mammalian plasma mem-branes (Epand and Vogel, 1999). The interactionwith Gram-negative bacteria (e.g. Escherichia coli)is favored by the presence of negatively chargedlipopolysaccharide (LPS) molecules in the outermembrane, oriented towards its exterior. Gram-positive bacteria (e.g. Staphylococcus aureus), how-ever, do not have an outer membrane or LPS andmay present lower sensitivity to cationic antibacter-ial peptides as a result of the altered structure ofteichoic acids, a major component of their surfacemembrane (Epand and Vogel, 1999; Sitaram andNagaraj, 1999).

In contrast to cationic molecules, peptides with alow net negative charge or a low net positive chargealong their helix backbone are mostly lytic tomammalian cells (Shai, 1999), but are not socommon amongst frog skin secretion peptides.Ocellatin 4, however, belongs to this category andpresents hemolytic activity against human erythro-cytes.

Secretion of hemolytic compounds may contri-bute to make the frog unpalatable to somepredators and help in the control of externalparasitism. Chigger mites (Trobiculidae) and ticks(Ixodidae) are common blood-sucking ectoparasitesof the skin of terrestrial amphibians living in dryforests (Duellman and Trueb, 1994). The DistritoFederal region of Brazil, where we collectedL. ocellatus specimens, is a region of dry vegetation.

2. Materials and methods

2.1. Chemicals

Solvents for chromatographic procedures were ofHPLC grade from several sources. Reagents andsolvents for protein sequencing were purchasedfrom Applied Biosystems (USA). Only analyticalgrade reagents from commercial suppliers were usedthroughout the experiments and solutions wereprepared with Milli-Q water (Millipore ReagentWater System, USA).

2.2. Preparation of skin secretion

Adult specimens of L. ocellatus were collected inDistrito Federal, Brazil (voucher specimens for this

ARTICLE IN PRESSA. Nascimento et al. / Toxicon 50 (2007) 1095–1104 1097

population can be found in the University ofBrasilia Department of Zoology Herpetology Col-lection). Frogs were kept in captivity at theUniversity of Brasilia and their skin secretions wereobtained by mild electric stimulation, a process thatdoes not harm the amphibians (Apponyi et al.,2004). Secretions were thoroughly washed from theskin surface with Milli-Q water, collected in abeaker and lyophilized.

2.3. Peptide purification by RP-HPLC

Aliquots (10.0mg) of lyophilized skin secretionwere dissolved in 200 mL of 0.1% (v/v) trifluoroa-cetic acid (TFA) in water and injected into a10� 250mm Vydac C18 218TP1010 column (TheSeparations Group, USA), equilibrated with 0.1%(v/v) TFA/water. Elution was performed with alinear gradient of 0.1% (v/v) TFA/acetonitrile,increasing from 0% to 60% over 60min, at a flowrate of 2.5mL/min. Eluants were monitored by UVabsorbance at 216Zm. Fractions were manuallycollected and vacuum dried.

Ocellatin 4 was individually chromatographedusing a higher-resolution 4.6� 250mm Vydac C18

218TP54 column (The Separations Group, USA),equilibrated with 0.1% (v/v) TFA/water. Theconcentration of 0.1% (v/v) TFA/acetonitrile inthe eluting solvent was raised to 45% over 5minand elution was performed with a 45–60% lineargradient over 15min, at a flow rate of 1.0mL/minand with UV detection at 216Zm. The pure peptidewas collected, vacuum dried, weighed and used inthe subsequent experiments.

2.4. Antibacterial assays

Reference strains of Staphylococcus aureus

(ATCC 25923) and E. coli (ATCC 25922) were firstgrown in 5mL of Mueller-Hinton broth, at 37 1C, ina shaker. When OD590 nm ¼ 1.0 was reached, thebacterial suspension was diluted in fresh Mueller-Hinton broth (1:50 for Gram-negative and 1:10for Gram-positive bacteria). Serial dilutions of asterile 256 mM ocellatin 4 stock solution wereincubated with the diluted bacterial suspension 1:1(v/v) in sterile 96-well microtiter plates. Forcontrols, we incubated either Milli-Q water or0.4% (v/v) formaldehyde with bacterial suspension(1:1, v/v). After 18 h of incubation at 37 1C, the platewas analyzed at 595Zm in a BioRad microplatereader Model 3550-UV. The assay was performed

in triplicate. Minimum inhibitory concentration(MIC) was taken as the lowest concentration ofocellatin 4 in which no bacterial growth could bedetected.

2.5. Hemolysis and osmotic protection assays

The hemolysis assay protocol was modified fromOnuma et al. (1999). Human O+ red blood cellswere washed three times in 0.01M Tris–HCl pH 7.4buffer containing 0.15M NaCl. A 1% (v/v) cellsuspension and a 128 mM ocellatin 4 solution wereprepared in this same buffer. In 1.5mL plastictubes, serial dilutions of the peptide solution wereused to determine its HC50 (concentration thatcauses 50% hemolysis). The final volume in eachtube was 200 mL: 100 mL of a given peptide dilutionplus 100 mL of the cell suspension. After incubationat RT for 60min, the tubes were centrifuged at3000 rpm for 2min and 100 mL aliquots of thesupernatants were pipetted into 96-well microtiterplates. Absorbance at 405Zm was determined usinga BioRad microplate reader Model 3550-UV. The100% lysis reference was 100 mL of 1% (v/v) TritonX-100 added to 100 mL of the cell suspension; the0% reference was 100 mL of the cell suspension plus100 mL of the Tris buffer. The assay was performedin triplicate. HC50 was determined by logarithmicregression.

For osmotic protection assessment, we tested7 mM (final concentration) ocellatin 4 for hemolysisin the presence of 0.05M arabinose, sucrose orraffinose. The same concentration of carbohydrateswas added to positive and negative controls.

2.6. Mass spectrometry

The value for the molecular mass of ocellatin 4was assessed by MALDI-TOF MS on a BrukerDaltonics Reflex IV mass spectrometer. On astainless steel plate, 1 mL of the sample in Milli-Qwater was mixed into 1 mL of a 10 mg/mL a-cyano-4-hydroxycinnamic acid (HCCA) matrix solution in0.1% (v/v) TFA in 50% (v/v) acetonitrile and left toair dry. Analysis was performed in the reflectorpositive ion mode. The mass scale was calibratedwith Peptide Mix (Bruker Daltonics Inc., USA) andthe spectrum was scanned from m/z 800 to 3000. Toinvestigate whether the native ocellatin 4 presentedan amidated C-terminus, analysis by MS/MS wascarried out on an Applied Biosystems 4700 Proteo-mics Analyzer MALDI-TOF/TOF, at a maximum

ARTICLE IN PRESSA. Nascimento et al. / Toxicon 50 (2007) 1095–11041098

accelerating potential of 20 kV and in the reflectorpositive ion mode. On a stainless-steel plate, 0.4 mLof the peptide dissolved in 0.1% (v/v) TFA in 50%(v/v) acetonitrile was mixed with an equal volume ofa 10 mg/mL HCCA matrix solution in the samesolvent and allowed to air dry. Calibration wasmade with a mixture of peptides of known masses.MS/MS spectra were collected using 2400 lasershots, with the metastable suppressor on and theCID gas off.

2.7. Amino acid sequence determination

Primary structure was determined by automatedEdman degradation using an Applied Biosystems477A sequencer, modified as described by Fontes etal. (1998). Phenylthiohydantoin (PTH) amino acidswere identified using an on-line reversed-phasePTH-C18 column in an Applied Biosystems 120AHPLC apparatus. The chromatography system wascalibrated with PTH amino acid standards prior toeach analysis.

2.8. Bioinformatics

The following programs were used: (1) ComputepI/Mw Tool (http://www.expasy.ch/tools/pi_tool.html), for theoretical molecular mass calcula-

Fig. 1. (a) Chromatographic profile of crude L. ocellatus skin secretio

(b) Further purification of ocellatin 4 (RP-HPLC; Vydac C18 4.6� 25

(MALDI-TOF MS, reflector positive ion mode).

tion; (2) BLAST2 (Altschul et al., 1997—http://au.expasy.org/tools/blast/), for similarity searchesagainst the UniProt Knowledgebase; (3) ClustalW(Thompson et al., 1994—http://www.ebi.ac.uk/clustalw), for multiple sequence alignments, whichwere then manually adjusted; (4) EMBL Gatewayto Isoelectric Point Service (http://www.embl-heidelberg.de/cgi/pi-wrapper.pl), for calculations ofpeptide charges; (5) the applet at http://cti.itc.virginia.edu/�cmg/Demo/wheel/wheelApp.html,for wheel projections of peptides; (6) PAUP 4.0, forbuilding, by distance analysis, bootstrap 50%majority-rule consensus neighbor-joining trees,based on peptide sequence alignments, with char-acters resampling in 1000 replicates; and (7) Tree-View, for displaying the trees.

3. Results

3.1. Biological activity of ocellatin 4

Ocellatin 4 has only two positively chargedresidues, counteracted by two negatively charged,and thus shows a very weak charge of �0.0012 atpH 7.0, which may contribute to its significanthemolytic activity: HC50 of 14.3 mM for human O+

red blood cells. However, it also means that thepeptide has little affinity for negative charges in

n (RP-HPLC; Vydac C18 10� 250mm; flow rate ¼ 2.5mL/min).

0mm; flow rate ¼ 1.0mL/min). (c) Mass spectrum of ocellatin 4

ARTICLE IN PRESS

Fig. 2. (a) Alignments between ocellatin 4 and peptides from

leptodactylids (ocellatins, fallaxin, laticeptin and pentadactylin)

and from Australian hylids (aureins and caeridins). Identical

positions are boxed in black and conserved substitutions are

highlighted in gray. (b) Bootstrap 50% majority-rule consensus

neighbor-joining tree, based upon the sequence alignment of the

indicated peptides. Bootstrap analysis was performed with

character resampling in 1000 replicates and values greater than

50% are shown on branches. The distance scale is drawn below

the tree.

A. Nascimento et al. / Toxicon 50 (2007) 1095–1104 1099

bacterial membranes, which explains its weakerantibacterial activity: the MIC obtained for bothE. coli ATCC 25922 and S. aureus ATCC 25923 was64 mM.

Ocellatin 4 was the last eluant of the RP-HPLCof the crude skin secretion (Fig. 1a). Longerretention times have been correlated with increasedamphipathicity, which is reported as necessaryfor hemolytic activity (Sitaram and Nagaraj,1999). Increased amphipathicity can render thepeptide a strong self-associating ability in solution,which enhances hemolytic activity and hindersantibacterial activity (Chen et al., 2005). It is thuspossible that ocellatin 4 self-associates in aqueoussolution.

In the hemolytic activity assay, it was possible tovisually detect hemolysis right after adding 1% (v/v)Triton X-100 (positive control), by the observationthat a red solution formed, instead of cell sedimen-tation occurring as in the negative control. Thesame was observed after adding the highest con-centration of ocellatin 4 (64 mM, obtained by adding100 mL 128 mM peptide solution to 100 mL cellsuspension), indicating that the peptide canpromptly disturb cell membranes. The results ofthe osmotic protection assay suggest a pore-formingmechanism for the hemolytic activity of ocellatin 4.The carbohydrates arabinose and sucrose, withmolecular diameters of 0.62 and 0.92Zm (Clinken-beard and Thiessen, 1991), caused the hemolyticactivity of 7 mM ocellatin 4 to decrease 50% and75%, respectively. In contrast, raffinose (moleculardiameter ¼ 1.14Zm) did not reduce hemolyticactivity at all. According to these data, thefunctional diameter of pores formed by ocellatin 4in the plasma membrane of erythrocytes is believedto be above 0.92Zm and below 1.14Zm.

3.2. Primary structure characterization and analysis

The primary structure of ocellatin 4 isGLLDFVTGVGKDIFAQLIKQI-NH2. The theo-retical molecular mass, 2274.30 u(monoisotopic),and the value determined by MALDI-TOF MS,2 274.24 u (Fig. 1c) were close. The slight differencebetween them suggested the possibility of aC-terminal amidation, which the MS/MS analysisconfirmed by the fact that C-terminal g-ions showeda �1 u shift, corresponding to the substitution of–OH (17 u) by �NH2 (16 u). C-terminal amidationcan increase cytolytic activity by inducing

C-terminal helix formation or stabilizing alreadyexisting helical conformations (Shalev et al., 2002).

Comparisons between ocellatin 4 and the othersix leptodactylid antimicrobial peptides described inliterature were made (Fig. 2a). Among the otherocellatins, ocellatin 4 is most similar to ocellatin 1,with nine identical positions, and shows eightpositions identical to ocellatins 2 and 3 (Nascimentoet al., 2004). Fallaxin (Rollins-Smith et al., 2005)and pentadactylin (King et al., 2005) present sevenpositions identical to ocellatin 4, while laticeptin(Conlon et al., 2006a) presents six.

A similarity search within all entries for neoba-trachian peptides in the UniProt database revealed afew other molecules that resembled ocellatin 4: fourpeptides from the skin secretions of four Australianhylid species of the genus Litoria. Among these,

ARTICLE IN PRESS

Fig. 3. (a) Alignments between ocellatin 4 and three hemolytic

peptides. Identical positions are boxed in black and conserved

substitutions are highlighted in gray. (b) Bootstrap 50%

majority-rule consensus neighbor-joining tree, based upon the

sequence alignment of the indicated peptides. Bootstrap analysis

was performed with character resampling in 1000 replicates and

values greater than 50% are shown on branches. The distance

scale is drawn below the tree.

A. Nascimento et al. / Toxicon 50 (2007) 1095–11041100

ocellatin 4 was most similar to aurein 5.1 (Litoria

raniformis and Litoria aurea—Rozek et al., 2000),with 10 identical positions, which overtakesthe similarity between ocellatin 4 and the otherleptodactylid peptides (Fig. 2a). The 13-residue-longcaeridin 7.1 (Litoria ewingi—Steinborner et al.,1997) differs from the 1–13 portion of aurein 5.1by just one conserved substitution. Thus, thesimilarity between ocellatin 4 and caeridin 7.1 isthe same observed between ocellatin 4 and the firsthalf of aurein 5.1. We also found nine positionsidentical to aurein 2.4 (L. aurea—Rozek et al.,2000), with the insertion of two gaps in its sequence,and six positions identical to caeridin 4 (Litoria

caerulea—Waugh et al., 1993) (Fig. 2a).Among ocellatin 4 and the 10 similar leptodacty-

lid and hylid peptides, the first four residues seem tohave some conservation, being G1 and D4 unan-imous. ‘‘1GLLD4’’ occurs in ocellatin 4, pentadac-tylin, aurein 5.1, caeridin 7.1 and caeridin 4, whilethe other peptides differ from them by conservedsubstitutions in residues 2 and/or 3 which keep thenonpolar character of both positions. Several otherconserved positions can be observed among thesepeptides (e.g. K11 in all the leptodactylid peptides;A15 and I21 in all ocellatins; G8 in ocellatins 1 and4, aurein 5.1, caeridin 7.1, fallaxin, laticeptin andpentadactylin). Ocellatin 4, aurein 5.1 and caeridin7.1 present a contiguous similarity from position 1to 9, where seven amino acids are identical and twoare conservatively substituted (Fig. 2a).

A bootstrap 50% majority-rule consensus neigh-bor-joining tree based on sequence alignmentsreflects the remarkable similarity between ocellatin4 and the Litoria peptides, which appear within thesame clade. It also indicates that this contiguoussimilarity is unlikely to happen by chance, since theclade is supported by 83% bootstrap value(Fig. 2b).

No remarkable similarity was found betweenocellatin 4 and other hemolysins. The only hemo-lytic peptides that resemble ocellatin 4 to someextent are chrysophsin 3, from the gills of the redsea bream, Chrysophrys major (Iijima et al., 2003),melittin, from the venom of the European honeybeeApis mellifera (Kreil, 1975) and nigrocin 2GRb,from the skin of the Asian frog Rana grahami

(Conlon et al., 2006b) (Fig. 3a). However, bootstrapanalysis (Fig. 3b) indicates that the similaritiesbetween chrysophsin 3, melittin and ocellatin 4occur probably by chance (bootstrap valueso50%,not shown). The similarity between nigrocin

2GRb and ocellatin 4 is more reliable, with abootstrap value of 66% (Fig. 3b), but nigrocin2GRb presents a disulfide bridge, which does notoccur in ocellatin 4.

4. Discussion

Only six leptodactylid skin secretion peptides,from four species of the genus Leptodactylus, hadbeen characterized up to now. Ocellatin 4, hereindescribed for the first time, is different from the bulkof frog skin secretion peptides: it has a very weakcharge of �0.0012 in pH 7.0 and interacts prefer-entially with blood-cell membranes.

A comparison between N-terminal (1–18) Schifferand Edmundson (1967) wheel projections and someproperties of ocellatin 4, fallaxin, laticeptin andpentadactylin evidences some possible causes for thedifferences in their activities (Fig. 4). The wheelprojection of pentadactylin shows the most well-organized polar face and one less aspartate (D)residue than the projections of the other peptides. Itis possible that the negative charge of aspartate sidechains repels the negative charges exposed inbacterial membranes (Lee et al., 2006), which couldpartially explain why pentadactylin is more active

ARTICLE IN PRESS

Fig. 4. Comparison between Schiffer and Edmundson N-terminal (1–18) wheel projections, charges and biological activities of ocellatin 4,

fallaxin (Rollins-Smith et al., 2005), laticeptin (Conlon et al., 2006a) and pentadactylin (King et al., 2005).

A. Nascimento et al. / Toxicon 50 (2007) 1095–1104 1101

against E. coli ATCC 25922 than ocellatin 4 and theother two peptides. The N-terminal wheel projec-tion of ocellatin 4 shows only one positively chargedresidue. The other three peptides, with more positivecharges in their helices, probably have more affinityfor the outer membrane of E. coli, rich in negativelycharged lipopolysaccharides. The projection ofocellatin 4, the only hemolytic peptide of the four,shows the less organized polar and nonpolar faces(Fig. 4).

The same comparison was made between ocella-tin 4 and the hemolytic peptides chrysophsin 3,melittin and nigrocin 2GRb (Fig. 5). In addition tobeing hemolytic, these three peptides present re-markable antimicrobial activities and are cationic.Ocellatin 4 is the only one with neutral charge at pH

7.0, reinforcing the possibility that its weakerantimicrobial activity results from little chargeaffinity for bacterial membranes. We could notobserve any relationship between the presence ofnegatively charged amino acid residues and hemo-lytic activity. Both melittin, the most hemolytic, andnigrocin 2GRb, the least hemolytic, have no acidicresidues in their sequences, while chrysophsin 3 andocellatin 4 present, respectively, one and twoaspartate residues. The weaker hemolytic activityof nigrocin 2GRb may be related to its betterorganized polar face and its smaller number ofhydrophobic residues.

It is uncommon to observe conserved regions inskin secretion peptides of different frog species,since accelerated mutation is favored by the fact

ARTICLE IN PRESS

Fig. 5. Comparison between Schiffer and Edmundson N-terminal (1–18) wheel projections, charges and biological activities of ocellatin 4,

chrysophsin 3 (Iijima et al., 2003), melittin (Kreil, 1975; Conlon et al., 2003) and nigrocin 2GRb (Conlon et al., 2006b).

A. Nascimento et al. / Toxicon 50 (2007) 1095–11041102

that many of these peptides do not participate in thegeneral metabolism of the animal and no deleteriouseffects are expected when a new variant emerges(Vanhoye et al., 2003). Within the several hundredsof frog skin peptide sequences in the UniProtdatabase, aurein 5.1 and caeridin 7.1 were the onlypeptides to show a nine-position contiguous simi-larity to ocellatin 4.

Given this observation, it is interesting to brieflyrecall the paleobiogeography of anurans, mostlyassociated with Gondwanaland, a supercontinentthat began to break up about 150 million years ago(m.y.a.). Three of the four subfamilies of hylids arepredominantly South American, while Pelodryadi-nae, which includes the genus Litoria, is restricted to

Australia and New Guinea. It is possible thatancestors of the Australian hylids, coming fromSouth America, crossed Antarctica to Australiabetween 150 and 130m.y.a., when these continentswere still connected (Vanhoye et al., 2003; Nicolaset al., 2003). The family Leptodactylidae is essen-tially South American but is also present inAustralia (Cochran, 1962), indicating that part ofit followed the same radiation route as the hylidsand that leptodactylids and hylids may have had acommon Gondwanaland ancestor.

When the majority of frog families have had theirpeptide repertoires characterized, making samplesize and species distribution large enough to allowidentification of inherent analysis flaws (Chen et al.,

ARTICLE IN PRESSA. Nascimento et al. / Toxicon 50 (2007) 1095–1104 1103

2006), major advances will be made in the study ofamphibian biology, perhaps including corrobora-tion or refutation of some amphibian paleobiogeo-graphy hypotheses.

Acknowledgements

The authors thank Dr. Cynthia Kyaw for theaccess to the Microbiology Facility and MSc NatanMaciel for the assistance with the PAUP 4.0program. This study was supported by CNPq(Conselho Nacional de Desenvolvimento Cientıficoe Tecnologico), FINATEC (Fundac- ao de Empreen-dimentos Cientıficos e Tecnologicos) and FUB/UnB(Fundac- ao Universidade de Brasılia).

References

Altschul, S.F., Maden, T.L., Schaffer, A.A., Zhang, J., Zhang,

Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-

BLAST: a new generation of protein database search

programs. Nucleic Acids Res. 25, 3389–3402.

Apponyi, M.A., Pukala, T.L., Brinkworth, C.S., Maselli, V.M.,

Bowie, J.H., Tyler, M.J., Booker, G.W., Wallace, J.C.,

Carver, J.A., Separovic, F., Doyle, J., Llewellyn, L.E., 2004.

Host-defence peptides of Australian anurans: structure,

mechanism of action and evolutionary significance. Peptides

25, 1035–1054.

Chen, Y., Mant, C.T., Farmer, S.W., Hancock, R.E.W., Vasil,

M.L., Hodges, R.S., 2005. Rational design of a-helicalantimicrobial peptides with enhanced activities and specifi-

city/therapeutic index. J. Biol. Chem. 280, 12316–12329.

Chen, T., Zhou, M., Rao, P., Walker, B., Shaw, C., 2006. The

Chinese bamboo leaf odorous frog (Rana (Odorrana)

versabilis) and North American Rana frogs share the same

families of skin antimicrobial peptides. Peptides 27,

1738–1744.

Clinkenbeard, K.D., Thiessen, A.E., 1991. Mechanism of action

of Moraxella bovis hemolysin. Infect. Immun. 59, 1148–1152.

Cochran, D.M., 1962. Living Amphibians of the World. Double-

day & Company Inc., New York, USA.

Conlon, J.M., Sonnevend, A., Patel, M., Camasamudram, V.,

Nowotny, N., Zilahi, E., Iwamuro, S., Nielsen, P.F., Pal, T.,

2003. A melittin-related peptide from the skin of the Japanese

frog, Rana tagoi, with antimicrobial and cytolytic properties.

Biochem. Biophys. Res. Commun. 306, 496–500.

Conlon, J.M., Al-Ghaferi, N., Abraham, B., Sonnevend, A.,

King, J.D., Nielsen, P.F., 2006a. Purification and properties

of laticeptin, an antimicrobial peptide from skin secretions of

the South American frog Leptodactylus laticeps. Protein Pept.

Lett. 13, 355–359.

Conlon, J.M., Al-Ghaferi, N., Abraham, B., Jiansheng, H.,

Cosette, P., Leprince, J., Jouenne, T., Vaudry, H., 2006b.

Antimicrobial peptides from diverse families isolated from the

skin of the Asian frog, Rana grahami. Peptides 27, 2111–2117.

Darst, C.R., Cannatella, D.C., 2004. Novel relationships among

hyloid frogs inferred from 12S and 16S mitochondrial DNA

sequences. Mol. Phylogenet. Evol. 31, 462–475.

Duda Jr., T.F., Vanhoye, D., Nicolas, P., 2002. Roles of

diversifying selection and coordinated evolution in the

evolution of amphibian antimicrobial peptides. Mol. Biol.

Evol. 19, 858–864.

Duellman, W.E., Trueb, L., 1994. Biology of Amphibians. The

Johns Hopkins University Press, Baltimore, USA.

Epand, R.M., Vogel, H.J., 1999. Diversity of antimicrobial

peptides and their mechanisms of action. Biochim. Biophys.

Acta 1462, 11–28.

Fontes, W., Cunha, R.B., Sousa, M.V., Morhy, L., 1998.

Improving the recovery of lysine in automated protein

sequencing. Anal. Biochem. 258, 259–267.

Frost, D.R., 2004. Electronic Database accessible at /http://

research.amnh.org/herpetology/amphibia/index.phpS. Amer-

ican Museum of Natural History, New York.

Hoegg, S., Vences, M., Brinkmann, H., Meyer, A., 2004.

Phylogeny and comparative substitution rates of frogs

inferred from sequences of three nuclear genes. Mol. Biol.

Evol. 21, 1188–1200.

Iijima, N., Tanimoto, N., Emoto, Y., Morita, Y., Uematsu, K.,

Murakami, T., Nakai, T., 2003. Purification and character-

ization of three isoforms of chrysophsin, a novel antimicrobial

peptide in the gills of the red sea bream, Chrysophrys major.

Eur. J. Biochem. 270, 675–686.

King, J.D., Al-Ghaferi, N., Abraham, B., Sonnevend, A.,

Leprince, J., Nielsen, P.F., Conlon, J.M., 2005. Pentadactylin:

an antimicrobial peptide from the skin secretions of the South

American bullfrog Leptodactylus pentadactylus. Comp. Bio-

chem. Physiol. C 141, 393–397.

Kreil, G., 1975. The strucutre of Apis dorsata melittin:

phylogenetic relationships between honeybees as deduced

from sequence data. FEBS Lett. 54, 100–102.

Lee, K.H., Lee, D.G., Park, Y., Kang, D., Shin, S.Y., Hahm,

K.S., Kim, Y., 2006. Interactions between the plasma

membrane and the antimicrobial peptide HP (2-20) and its

analogues derived from Helicobacter pylori. Biochem. J. 394,

105–114.

Nascimento, A.C., Fontes, W., Sebben, A., Castro, M.S., 2003.

Antimicrobial peptides from anurans skin secretions. Protein

Pept. Lett. 10, 227–238.

Nascimento, A.C., Zanotta, L.C., Kyaw, C.M., Schwartz, E.N.,

Schwartz, C.A., Sebben, A., Sousa, M.V., Fontes, W., Castro,

M.S., 2004. Ocellatins: new antimicrobial peptides from the

skin secretion of the South American frog Leptodactylus

ocellatus (Anura: Leptodactylidae). Protein J. 23, 501–508.

Nicolas, P., Vanhoye, D., Amiche, M., 2003. Molecular strategies

in biological evolution of antimicrobial peptides. Peptides 24,

1669–1680.

Onuma, Y., Satake, M., Ukena, T., Roux, J., Chanteau, S.,

Rasolofonirina, N., Ratsimaloto, M., Naoki, H., Yasumoto,

T., 1999. Identification of putative palytoxin as the cause of

clupeotoxism. Toxicon 37, 55–65.

Rollins-Smith, L.A., King, J.D., Nielsen, P.F., Sonnevend, A.,

Conlon, J.M., 2005. An antimicrobial peptide from the skin

secretions of the mountain chicken frog Leptodactylus fallax

(Anura:Leptodactylidae). Regul. Pept. 124, 173–178.

Rozek, T., Wegener, K.L., Bowie, J.H., Olver, I.N., Carver, J.A.,

Wallace, J.C., Tyler, M.J., 2000. The antibiotic and anticancer

active aurein peptides from the Australian bell frogs Litoria

aurea and Litoria raniformis. Eur. J. Biochem. 267,

5330–5341.

ARTICLE IN PRESSA. Nascimento et al. / Toxicon 50 (2007) 1095–11041104

Schiffer, M., Edmundson, A.B., 1967. Use of helical wheels to

represent the structures of proteins and to identify segments

with helical potential. Biophys. J. 7, 121–135.

Shai, Y., 1999. Mechanism of the binding, insertion and

destabilization of phospholipid bilayer membranes by a-helical antimicrobial and cell non-selective membrane-lytic

peptides. Biochim. Biophys. Acta 1462, 29–54.

Shalev, D., Mor, A., Kustanovich, I., 2002. Structural con-

sequences of carboxyamidation of dermaseptin S3. Biochem-

istry 41, 7312–7317.

Sitaram, N., Nagaraj, R., 1999. Interaction of antimicrobial

peptides with biological and model membranes: structural

and charge requirements for activity. Biochim. Biophys. Acta

1462, 29–54.

Steinborner, S.T., Bowie, J.H., Tyler, M.J., Wallace, J.C., 1997.

An unusual combination of peptides from the skin glands of

Ewing’s tree frog, Litoria ewingi—sequence determination

and antimicrobial activity. Aust. J. Chem. 50, 889–894.

Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL

W: improving the sensitivity of progressive multiple sequence

alignment through sequence weighting, position-specific gap

penalties and weight matrix choice. Nucleic Acids Res. 22,

4673–4680.

Vanhoye, D., Bruston, F., Nicolas, P., Amiche, M., 2003.

Antimicrobial peptides from hylid and ranin frogs originated

from a 150-million-year-old ancestral precursor with a

conserved signal peptide but a hypermutable antimicrobial

domain. Eur. J. Biochem. 270, 2068–2081.

Waugh, R.J., Stone, D.J.M., Bowie, J.H., Wallace, J.C., Tyler,

M.J., 1993. Peptides from Australian frogs. Structures of the

caeridins from Litoria caerulea. J. Chem. Soc. Perkin Trans. I

1, 573–576.