hadrurin, a new antimicrobial peptide from the venom of the scorpion hadrurus aztecus

Eur. J. Biochem. 267, 5023±5031 (2000) q FEBS 2000

Hadrurin, a new antimicrobial peptide from the venom of the scorpionHadrurus aztecus

Alfredo Torres-Larios, Georgina B. Gurrola, Fernando Z. Zamudio and Lourival D. Possani

Department of Molecular Recognition and Structural Biology, Institute of Biotechnology, Avenida Universidad, Cuernavaca, Mexico

A new antimicrobial peptide, hadrurin, was isolated from the venom of the Mexican scorpion Hadrurus aztecus,

by gel filtration on a Sephadex G-50 column, followed by high performance liquid chromatography. It is a basic

peptide composed of 41 amino-acid residues with a molecular mass of 4436 Da, and contains no cysteines. A

model of the three-dimensional folding of hadrurin is compatible with that of an amphipatic molecule with

two a-helical segments. Hadrurin demonstrates antimicrobial activity at low micromolar concentration, inhibiting

the growth of bacteria such as: Salmonella thyphi, Klebsiella pneumoniae, Enterococcus cloacae, Pseudomonas

aeruginosa, Escherichia coli and Serratia marscences. It also shows cytolytic activity when tested in human

erythrocytes. Hadrurin and two analogs (C-terminal amidated, and all d-enantiomer) were chemically

synthesized. They were used to study the possible molecular mechanism of action by testing their ability to

dissipate the diffusion potential of liposomes of different compositions. The results obtained indicate that there

are no specific receptor molecules for the action of hadrurin, and the most probable mechanism is through a

membrane destabilization activity. It is surmised that hadrurin is used by the scorpion as both an attack and

defense element against its prey and putative invasive microorganisms. It is a unique peptide among all known

antimicrobial peptides described, only partially similar to the N-terminal segment of gaegurin 4 and brevinin 2e,

isolated from frog skin. It would certainly be a model molecule for studying new antibiotic activities and

peptide±lipid interactions.

Keywords: antimicrobial peptide; hadrurin; hemolysis; scorpion venom; synthetic peptide.

Since the discovery in 1981 of an antimicrobial peptide ininvertebrates [1], more than 100 molecules with this propertyhave been isolated, most of them from arthropods, amphibians,plants and mammals (reviewed in [2±5]). Their role in innateimmunity has been proved in some insects [6±8] and mammals[9], as well as their constitutive or inducible expression depend-ing on the tissue, organism, or pathogen [10±15]. Briefly, wecan group them in: (a) a-helicoidal; (b) peptides with one tofour disulfide bridges; (c) peptides rich in certain amino acids,such as proline or tryptophan. Besides their antibacterialactivity, some of them are hemolytic [16±25]. Most of thesepeptides share some common characteristics such as their lowmolecular mass (2±5 kDa), the presence of multiple lysine andarginine residues, and their amphipatic nature. Their site ofaction is the cytoplasmic membrane, where they destabilizeits lipid package and produce transient channels (reviewed in[26±28]). However, it seems that the primary action of some ofthem, such as PR39 [29,30], is not directed towards themembrane, and it is believed that for plant defensins thereexists a receptor [31,32];. Although the minimal inhibitoryconcentrations (1±50 mm) of these molecules are high incomparison to other antibiotics, their broad activity spectra andspeed of action makes them good candidates for delivering

drugs, and a number of possible applications have already beendescribed (reviewed in [33]).

Scorpion venom has been traditionally studied for the pre-sence of neurotoxins that affect ion channels: sodium, potas-sium, calcium and chloride [34], and very little is known aboutpeptides with different activity. Antimicrobial peptides havebeen isolated from the venom of spider [35], hornet [23] andbee [16].

This communication reports the isolation of a novel anti-bacterial and cytolytic peptide constitutively present in thevenom of the scorpion Hadrurus aztecus. It has a uniquesequence, quite different from other antimicrobial peptidesisolated from the hemolymph of scorpions [36,37]; and otherorganisms [2±5]. It inhibits the growth of a variety of bacteria,but also has hemolytic activity. This peptide and several homo-logues were chemically synthesized and their molecular mech-anism of action studied using liposomes made of different lipidcompositions. It was named hadrurin, because of the scorpiongenus from which it was isolated.

E X P E R I M E N T A L P R O C E D U R E S

Source of venom

The crude venom of Hadrurus aztecus was obtained by electricstimulation of the telson of scorpions collected in Iguala, stateof Guerrero, Mexico. The venom was recovered with doubledistilled water and centrifuged in a Beckman Optima TL ultra-centrifuge for 15 min, at 4 8C and 15 000 g. The supernatantwas lyophilized and stored at 220 8C.

Correspondence to L. D. Possani, Instituto de Biotecnologia-UNAM,

Avenida Universidad, 2001, Apartado Postal 510-3, Cuernavaca, 62210,

Mexico. Fax: 1 52 73 172388, Tel.: 1 52 73 171209,

E-mail: [email protected]

Note: a patent application has been presented to the Office of Patent and

TradeMarks in Mexico City, Mexico, regarding the discovery of hadrurin

and its possible uses.

(Received 8 March 2000, revised 8 June 2000, accepted 9 June 2000)

Purification of the peptide

The purification of the soluble components was initially per-formed by gel filtration on Sephadex G-50 column chromato-graphy (grade superfine, Pharmacia Fine Chemicals, Uppsala,Sweden). The lyophilized venom was dissolved in 20 mmammonium acetate buffer, pH 4.7, and applied directly to thecolumn. The fraction number III obtained from this separationwas purified by high performance liquid chromatography(HPLC) using a Cl8 semipreparative column (Vydac, Hysperia,CA), in a Waters 600E equipment with a UV detectorWaters-486, using a linear gradient from 0 to 60% of aceto-nitrile, in the presence of 0.1% of trifluoroacetic acid, during60 min. The most hydrophobic component was re-chromato-graphed in an analytical C18 column in the same conditions andshown to be homogeneous.

Chemical characterization

The chemical composition of the peptide was obtained throughamino-acid analysis in a Beckman analyzer 6300E, after 20 hhydrolysis at 110 8C in vacuum sealed tubes with 6 m HCl and0.05% phenol (v/v). The minimal molecular mass obtainedfrom the amino-acid analysis was confirmed by mass spectro-scopy, using Finnegan-MAT equipment. The purified peptidewas sequenced in the Milligen/Biosearch ProSequencer 6600(Millipore division, Bedford, MA, USA) in covalent-boundmembranes Sequelon-AA, following the company protocol.Initially, the native peptide was used for direct sequencing.Additional data were obtained by sequencing HPLC-isolatedpeptides from the digestion with the endopeptidase Asp-N(Boehringer, Mannheim, Germany). For enzyme digestion,100 mg of peptide was used each time, dissolved in 50 mmphosphate buffer, pH 8.0, and incubated for 4 h at 37 8C. Theratio of enzyme/peptide was 1 : 100.

For determining the recovery values during chromatographicprocedures, it was assumed that one unit of absorbance at280 nm, in 1 cm cuvette width, is equal to 1 mg´mL21 ofprotein. However, for the quantification of the amount ofpeptide in the experiments with bacteria (dose±response curve),the real value obtained by amino-acid analysis was used.

Chemical synthesis and purification of peptides

The solid phase method of Merrifield [38] was used, with Fmoc(N-(9-fluorenyl) methoxycarbonyl) amino acids (Novabiochem,Laufelfingen, Switzerland). The yield amount of incorporationof each amino acid was monitored by ninhydrin reaction. At theend of the synthesis the peptide was freed from the resin bycleavage with trifluoroacetic acid in presence of the scavengersthioenisole and ethanedithiol. The peptide was purified using anHPLC apparatus, equipped with a reverse-phase C4 preparativecolumn, eluted with a linear gradient of acetonitrile from 30 to60%, in the presence of 0.1% of trifluoroacetic acid, for60 min. The second and third steps were performed by HPLCseparation with a C18 reverse-phase semipreparative andanalytical column, respectively, in the same conditions asbefore.

Inhibition assays of bacterial growth in liquid phase

For each fraction to be assayed, 5 mL of peptide dissolved in0.01% acetic acid containing 0.2% bovine serum albumin(BSA) was employed, using different peptide concentrations. A96-well microtiter plate was used for bacterial growth in the

presence of a buffered Luria±Bertani medium. Each well con-tained 45 microliters of culture, with approximately 1 � 105

bacteria´mL21. The microbial growth was monitored by opticaldensity at 492 nm, in an ELISA reader (BioRad, model 2550,Hercules, CA, USA), after incubation for 18 h at 37 8C. Thepositive control was performed using bacteria alone, in thepresence of the same buffer. Negative controls were obtained bythe addition of 0.4% formaldehyde to the culture. The bacteriastrains used were obtained from the collection of the NationalInstitute of Public Health, Mexico.

Hemolytic assays

Before use, freshly collected human blood was rinsed threetimes with NaCl/Pi, pH 7.4, and centrifuged each time for15 min at 900 g. A suspension was made at 0.5% packed cellsin NaCl/Pi. From this suspension, aliquots containing 195 mLeach were incubated with 5 mL of peptide at different con-centrations. As a positive control (100% lysis), a 1% solution ofTriton X-100 and honey-bee venom melittin obtained fromSigma (St Louis, MO, USA) was used. After incubation for 1 hat 37 8C, the sample was centrifuged at 900 g for 2 min, and theabsorbance of the supernatant was determined at 541 nm in aBeckman DU-50 spectrophotometer.

Preparation of liposomes

Liposomes were prepared by the addition, in test tubes, of 1 mgof chloroform solubilized phospholipids, either 1,2-diphy-tanoyl-sn-glycero-3-phosphocholine (DPhPC), or a mixture ofthis and 1,2-diphytanoyl-sn-glycero-3-phosphoserine (DPhPS)(1 : 1, w/w), from Avanti Polar Lipids, Inc. (Alabaster, AL,USA). After evaporation of the solvent under a nitrogen stream,the lipid film was hydrated with 500 mL of a potassium buffer(150 mm KCl, 10 mm Hepes, pH 7.0). The lipid dispersion wasvortexed and then sonicated for 15±20 min in a bath typesonicator (Ney, Yucaipa, CA) until clear.

Membrane permeability study

The membrane permeability tests were performed usingvalinomycin-mediated K1 diffusion potential assay, asdescribed [39,40]. Aliquots containing 14 mL of the liposomesuspension (for a final lipid concentration of 40.2 mm) werediluted in 900 mL of an isotonic, potassium-free buffer[150 mm methyl glucamine (MeGlc), 10 mm Hepes, pH 7.0].To each sample, 1 mL of a potentially sensitive dye wasadded. The dye used was 3,3 0-diethyl-thio-dicarbocyanineiodide (diS-C2-5, MW 492), from Molecular Probes (Eugene,OR, USA). Valinomycin was immediately added to eachsample (1 mL, for a final concentration of 1 nm), in order toslowly create a negative diffusion potential inside the vesicles,monitored as a decrease in the dye's fluorescence. Once thefluorescence had stabilized, 8±10 min later, increasing amountsof peptide were added, until total dissipation of the diffusionpotential was reached, as reflected by a gradual increase influorescence, monitored on a Beckman LD50 spectrofluoro-meter, with excitation set at 620 nm and emission at 670 nm.The percentage of fluorescence recovery, Ft, was defined as

Ft � �It 2 Io/If 2 Io�100

where It is the fluorescence observed after the addition of thepeptide at time t, Io is the fluorescence after the addition ofvalinomycin, and If is the total fluorescence before addingvalinomycin.

5024 A. Torres-Larios (Eur. J. Biochem. 267) q FEBS 2000

R E S U LT S

Peptide purification and primary structure determination

The venom of the scorpion H. aztecus was fractionated bysize-exclusion chromatography on Sephadex G-50, as shown in

Fig. 1A. Of the resulting seven fractions only the third (labeledIII in the figure) inhibited the growth of E. coli (strain ATCC25926) at the concentration tested, using 5 mg of each fractionper assay. This fraction was further purified by HPLC usingreverse-phase chromatography. At least 20 major subfrac-tions were detected (Fig. 1B), where the last one, eluting at

Fig. l. Purification of hadrurin from

H. aztecus venom. (A) The soluble venom

(69 mg in 1.5 mL) was fractionated in a

Sephadex G-50 superfine column

(0.9 � l50 cm) equilibrated and run in 20 mm

ammonium acetate buffer, pH 4.7, with a

flow-rate of 12 mL´hr21. Tubes, with a volume

of 1.5 mL, were collected and grouped

according to the absorbance at 280 nm as shown

(horizontal bars). The total yield was 81%, from

which approximately 25% corresponds to

fraction I, 17% to fraction II, 29% to fraction III

and the 27% left to the other minor fractions.

Fraction III showed antimicrobial effect against

E. coli. (B) Fraction III from Fig.lA (18 mg)

was separated in a Cl8 reverse-phase,

semipreparative column, with a linear gradient

from 100% of solution A (0.12% of

trifluoroacetic acid in water) to 60% of solution

B (0.10% of trifluoroacetic acid in acetonitrile),

for 1 h with a flow-rate of 2 mL´min21. The

component labeled with an asterisk showed

antimicrobial activity against E. coli. The inset

shows the HPLC profile of the pure component

(asterisk of Fig. 1B) using an analytical C18

reverse-column, with the same gradient.

(C) Complete amino-acid sequence of hadrurin

was obtained from native peptide and by

sequencing a peptide fragment purified by

HPLC, after digestion with endopeptidase

AspN. Samples with approximately 1 nmol

were applied to the sequencer. Direct Edman

degradation of hadrurin allowed the sequence

determination of the first 30 residues (glycine

1 up to valine 30), underlined with letter d. One

of the pure peptides obtained by HPLC after the

enzymatic digestion, resolved the amino-acid

position from aspartic acid 21 to alanine 41,

showing an extended overlapping segment

(labeled with AspN).

q FEBS 2000 Scorpion antimicrobial peptide (Eur. J. Biochem. 267) 5025

approximately 50 min and labeled with an asterisk in the figure,had antibacterial activity. The purity of this subfraction wasassessed by its application on an analytic HPLC column (seeinset of Fig. 1B). This peptide represents less than 1.7% of thetotal protein of the venom. It seems to be a constitutivecomponent of the venom, because it was found each time thescorpions were milked, over a period of about 2 years (usually,one extraction per month).

A sample of the pure peptide was applied to the micro-sequencer, and the first 30 amino-acid residues were unequi-vocally identified, as indicated in Fig. 1c (underlined sequencelabeled with d, which stands for direct sequencing). Amino-acid analysis of the native molecule showed that it was apeptide with a minimum molecular mass compatible with thepresence of at least 38 amino-acid residues. Note that acidhydrolysis does not permit identification of Trp residues. Inorder to complete the sequence a digestion of the native peptidewas conducted with endopeptidase AspN and the correspondingsubpeptides were separated by HPLC (data not shown). One ofthe fragments gave the overlapping sequence from residuesnumber 21±41, underlined with dashes in Fig. 1C, labeledAspN. Mass spectrometry analysis of the native peptide gavethe value 4435.3 Da, close to the calculated mass for theprimary structure experimentally determined of 4436.15. Simi-lar analysis performed on the C-terminal fragment, obtained

from the cleavage with AspN, gave a value of 2265.3 Da, whichis virtually identical to the mass predicted for the free acid formof the peptide. In view of these data the primary structure of thispeptide was assumed to be complete and we decided to call ithadrurin, from the genus of the scorpion from which it wasextracted. Thus, hadrurin is a 41 amino-acid long peptide, witha calculated molar absorption coefficient of 20 539 m21´cm21,containing no cysteines among its amino acids, but seven basicresidues, and a calculated isoelectric point of 11.08.

Antimicrobial activity of hadrurin

Figure 2 shows the growth inhibition of a series of bacteriawhen treated with different concentrations of hadrurin in liquidphase assays. The strains of Escherichia coli ATCC 25926,Serratia marscencens ATCC 13880 and Enterococcus cloacae129 were the most sensitive, with a minimum inhibitoryconcentration (MIC) inferior to 10 mm. The effects againstSalmonella thyphi and Klebsiella pneumoniae were lessevident, requiring a concentration near 40 mm for full inhibitoryeffect. The other bacteria tested needed 50 mm concentrationfor complete growth inhibition. Initial experiments were alsoconducted with freshly collected human blood, in order toverify if hadrurin had cytotoxic action. Important hemolyticactivity was obtained with concentration in the order of30 mm, as discussed later. Thus, hadrurin had antimicrobial andcytolytic activity.

Chemical synthesis of hadrurin and analogs

In order to confirm the primary structure of hadrurin and furthercharacterize its function, the chemical synthesis of hadrurin was

Fig. 2. Antibacterial activity of hadrurin. Growth inhibition was esti-

mated by monitoring the absorbance decrement at 492 nm, after incubating

the microorganisms with different concentrations of peptide for 18 h at

37 8C. At time zero, the absorbance was equal to zero. After 18 h of

incubation, the absorbance was maximum for the positive control cultures,

but for the cultures with different concentrations of hadrurin, the absor-

bance decreased as shown in the graph. As a positive control, BSA 0.2%

with 0.01% of acetic acid was used, and as negative control culture treated

with formaldehyde 0.4%, for each type of microorganism assayed, where

closed circles stand for Salmonella thyphi, empty circles for Klebsiella

pneumoniae 9, crosses for Enterococcus cloacae 129, closed squares for

Pseudomonas aeruginosa PG201, open squares for Pseudomonas aerugi-

nosa ATCC9027, closed triangles for Enterococcus feacalis 51, open

triangles for Escherichia coli ATCC 25926, and inverted closed triangles for

Serratia marscencens ATCC13880. Numbers after the scientific names of

microorganisms are different strains, as enlisted by the National Institute of

Public Health of Mexico. Values are mean from duplicated experiments.

Fig. 3. Separation of synthetic and native hadrurin. The profile of HPLC

separation of 20 mg of native hadrurin (labeled native) was superimposed

with that of 50 mg of synthetic hadrurin (labeled synthetic) and a mixture of

both (labeled nat. and synth), in amounts of 10 mg each. The absorbance

scale was adjusted, because for the synthetic hadrurin we overloaded the

column, in order to see any possible contaminants. A reverse-phase Cl8

analytical column was used with a linear gradient from 100% of solution A

(0.12% trifluoroacetic acid in water) to 60% of solution B (0.1 % of

trifluoroacetic acid in acetonitrile) for 1 h, at a flow-rate of 1 mL´min21.


undertaken by the solid-phase method of Merrifield, usingFmoc amino-acids. The free acid, amidated, and d-isomerforms of hadrurin were prepared. After synthesis, the peptideswere purified by HPLC (data not shown) and the amino-acidsequences of their N-terminal regions were confirmed bymicrosequencing. Additional results supporting the authenticityof the synthetic analogs were obtained by HPLC separation.Figure 3 shows the results of HPLC separation of purifiedpeptides in an analytical C18 reverse-phase column. In thisfigure, a chromatogram of the native hadrurin, the syntheticone and a mixture of both (50% each) were compared. Theretention times and the symmetry of peaks suggest that thesesamples are pure and correspond to identical compounds. Theform with an amidated C-terminus and the all d-enantiomerwere synthesized in order to verify the possible differences inactivity of the free acid form vs. the amidated one, and to verifythe possible presence of a specific receptor for the peptideaction, respectively. As Fmoc-d-isoleucine was not readilyavailable, the positions in which isoleucines occur in theprimary structure, were replaced by d-leucines. After synthesisthe analogs were purified by HPLC (data not shown). Pureamidated hadrurin eluted approximately 18 s earlier thanthe native one, whereas the retention time of the purifiedd-enantiomeric hadrurin was delayed by approximately 3 min,as expected by the substitution of isoleucine for a more hydro-phobic residue. This will be discussed later. The N-terminalamino-acid sequence of the d-enantiomer was also confirmedby microsequencing. Note that PTH-amino-acid derivatives ofd- and l-amino acids are not distinguished by the sequencer.

Antimicrobial and hemolytic activity of synthetic analogs

The antimicrobial activity of the analogs were tested in com-parison to native hadrurin, using E. coli ATCC 25926 as modelbacteria. The effects were practically identical (data not shown),including the d-enantiomer. In the case of the amidated formof hadrurin, a slightly lower MIC was obtained (around 5 mm,vs. about 7.5 mm for the native one). The assay with thed-enantiomer was repeated several times, adding new cultureon top of a previous one. The idea was to find out if thed-enantiomeric form would survive the attack of destructivebacterial enzymes, and remain antibiotically active for a longerperiod of time. We were unable to observe any clear differencecompared to the native hadrurin. The results of the hemolytictest, are shown in Fig. 4. In this figure the activity of nativehadrurin was compared to the synthetic ones and to a controlpeptide, melittin, which is a well known hemolytic bee venomcomponent [16]. The hemolytic effect of the native peptidecompared to that of the synthetically prepared free acid andamidated forms, is quite similar (at 20 mm concentration about

Fig. 4. Hemolytic effect of hadrurin and synthetic peptides. The hemo-

lytic activity was estimated by monitoring the increase in the absorbance at

541 nm, after incubating the blood sample in NaCl/Pi with different peptide

concentrations for 1 h at 37 8C. Positive control was estimated using a l%

solution of Triton X-100, whereas negative control contained BSA 0.2% in

0.01% acetic acid. Closed circles stand for native hadrurin, open circles for

synthetic free acid form, closed squares for synthetic amidated form, open

squares for synthetic d-isomer hadrurin and crosses for melittin. Values are

mean of duplicates.

Fig. 5. Effect of peptides on the maximal dissipation of the diffusion

potential in liposomes. (A) Vesicles prepared with phosphatidyl-choline.

The peptides were added to isotonic K1-free buffer containing vesicles pre-

equilibrated with the fluorescent dye diS-C225 and valinomycin. The

plot shows the fluorescence recovery after mixing the peptides with

the vesicles, where closed circles stand for melittin, open circle for

synthetic d-isomer of hadrurin, closed squares for the amidated form,

open squares for native hadrurin, crosses for the acid form. Values are

mean (^ SD) of five independent determinations. (B) Vesicles prepared

with a mixture of PtdCho and PtdSer. Conditions and symbols are the

same as in (A). Values are mean (^ SD) for three independent

determinations.


80% lysis is obtained), but it is certainly less hemolytic thanmelittin. In contrast, the hemolytic action of the d-isomer isvery similar to that of melittin, which shows an hemolysis of100% at a concentration around 5 mm. These results alsosupport the idea that there is no specific receptor for the actionof hadrurin, as the d-enantiomer was also active.

Activity using artificially prepared membranes

In view of the above reported results, it was decided to test theability of hadrurin and its synthetic analogs to disturb the lipidpacking of liposomes, using a diffusion potential assay. Thepeptides were mixed in increasing concentrations with eitherphosphatidyl-choline (PtdCho) or phosphatidyl-serine (PtdSer)plus PtdCho vesicles (in constant concentration) preloaded withthe fluorescent dye diS-C225 and valinomycin, as described in[41]. The fluorescence recovery was recorded and plottedagainst the peptide lipid21 (P/L) molar ratio between 0.002 and0.16 (Fig. 5). A perturbing activity of all the peptides from thelowest ratios [peptide]/[lipid] used was observed, in the caseof PtdCho liposomes (Fig. 5A), whereas the PtdSer/PtdCholiposomes (Fig. 5B) needed more peptide to produce the sameeffect. As an example, the peptide/lipid molar ratio at which20% activity of the synthetic d-isomer was found, is 0.0025 forPtdCho and 0.0168 for PtdSer/PtdCho vesicles, respectively. Atpeptide/lipid ratios higher than 0.12, the activity becomes 100%in both types of vesicles for all the peptides assayed. Anexception to this behavior is that of melittin and the syntheticd-isomer, which produces 100% of fluorescence recovery ata lower ratio (0.0784), for the PtdCho vesicles (see Fig. 5A). Inconclusion, the most active ones are melittin and the d-isomer,the d-isomer of hadrurin probably being slightly moreactive. The least active on the PtdCho liposomes is theamidated form of hadrurin, whereas the effects of all peptideson PtdSer/PtdCho liposomes (Fig. 5b) are quite comparable.

We have estimated that approximately 40 molecules ofpeptide bind to each vesicle for the case of PtdCho vesiclesprepared with a ratio of peptide/lipid equal to 0.008, whereasthe estimated amount of hadrurin is in the order of 500molecules per vesicle made of PtdCho/PtdSer mixture at ratiopeptide/lipid of 0.096.

D I S C U S S I O N

Hadrurin, a rare component of scorpion venom

Some scorpion species often use to spray venom on their ownbodies to clean them from dirty and possible saprophyticorganisms (bacteria and fungi). This observation, originallyreported to us by M. A. Gonzalez-Sponga from the Departmentof Biology and Chemistry of the Instituto Universitario Peda-gogico in Caracas Venezuela, suggested a working hypothesisby which the venom of these scorpions could contain some sortof antibiotic. Because Hadrurus aztecus lives underground intunnels excavated from soil and is directly exposed to micro-organisms, it seemed to us a good model to test the initialhypothesis. Additionally, the pair of venomous glands ofscorpions communicate freely with the exterior, via the aper-tures at the end of telson (stinger), hence leaving open thepossibility of contamination by microorganisms present in thehemocele of the scorpion preys during the sting.

Analysis of the antimicrobial activity constitutively present(not inducible) in the venom of Hadrurus aztecus allowed theidentification, isolation and characterization of hadrurin, thefirst antimicrobial peptide ever purified from scorpion venom.

Hadrurin is a peptide of 41 amino-acid residues, of whichseven are basic, three of them grouped as a triplet of sequenceLys-Arg-Lys. The absence of cysteines is also remarkable,because this residue has been shown to be a ubiquitous aminoacid in scorpion venom components. Cysteine residues are alsopresent in other peptides with antimicrobial activity, isolatedfrom the hemolymph of the scorpion Leiurus quinquestriatus[36] and Androctonus australis [37].

Hadrurin, similarly to melittin from bee [16], crabrolin fromhornet [42] and lycotoxins from spider [35], demonstrates lyticactivity against eukaryotic cells, thus all of them should beconsidered `bona fide' toxic components of the venoms. Physio-logically, these molecules showed a double function: defense(antimicrobial activity) and attack (cytolytic effect), whichwould explain their constitutive presence in the venom. Earlier,the hemolytic activity of certain scorpion venoms were reportedto be present in Nebo hierichonticus [43], Heterometrusfulvipes [44], and Tityus serrulatus [45], but no additionalisolation or characterization of components were conducted.For the work reported here the following species were alsotested: Centruroides limpidus, Centruroides noxius, Pandinusimperator, Anuroctonus pheodactilus, at a concentration of150 mg´mL21, and the antimicrobial and lytic activity waspositive only with the venom of Hadrurus aztecus. As thisassay was performed without calcium, the presence of phos-pholipases A2 activity is discarded. A calcium dependentphospholipase activity was reported for the venom of Pandinusimperator [46]. This fact suggests that Hadrurus aztecusdeveloped another peptide for capturing its prey, in parallel

Fig. 6. Sequence comparison and structural features of hadrurin. (A)

Amino-acid sequence comparison of hadrurin with other similar peptides.

Identical residues are marked with bold letters, and similar ones with italics.

Gaps were introduced in order to enhance similarities. (B) Schiffer-

Edmundson projection of hadrurin shows its probable amphipatic alpha

helicoidal conformation. Hydrophobic residues are in bold. Residues are

numbered, starting from the N-terminal. (C) Results of the prediction of

four different algorithms (Chou-Fasman, SOPMA, PHD and BCM, see

Materials and methods), which show an helicoidal region within the first 11

residues, followed by an undetermined region of 5 residues, and ending

with a second helicoidal region. H, helix; B, b-sheet; t, turn; c, coil.


with the production of toxins directed against ion channels, asalready described [47]. Preliminary results obtained by sepa-rating hemolymph of Hadrurus aztecus in the same conditionsas those used for Fig. 1B gave several components in the sameelution time, but with no antimicrobial activity (data notshown). This was taken as an indication that hadrurin is eitherrestricted to the venom only, or if it is also in the hemolymphthe concentration is very low and does not allow detection ofactivity.

Comparative analysis of sequence and secondary structureprediction

Comparison of the primary structure of hadrurin with otherantimicrobial peptides indicate that it has a unique and unusualsequence. Some sequence similarities were found only with thefirst 10 N-terminal amino acids of gaegurin 4 and brevinin 2e,purified from the skin of the frogs Rana rugosa [48] and Ranaesculenta [19], respectively (Fig. 6A). However, the C-terminalsection of gaegurin 4 and brevinin 2e is completely different,including the presence of the characteristic disulfide bridge,known as `rana box' [49] not present in hadrurin. Consideringthe C-terminal end, hadrurin shows some similar residues inequivalent positions of the sequence, when compared to the pigcecropin P1 [50], shown in Fig. 6A. However, Cecropin P1does not have the corresponding 12 most N-terminal situatedamino-acid residues. The gap program of the University ofWisconsin Genetics Computer Group (GCG) showed a certaindegree of similarity between hadrurin and gaegurin 4, brevinin2e and cecropin P1 of 38%, 39% and 31%, respectively. Thesecondary structure prediction algorithms, from: Chou andFasman [51], Baylor College of Medicine (BCM) [52], SOPMA[53,54], and PHD [55±57], predicted an alpha helix content of68.3, 51.2, 46.3 and 63.4%, respectively, with a confidencefactor of approximately 65%. The helicoidal regions are shownin Fig. 6C, where we can observe that there is a consensusregion, located between residues 12±16, that does not have apredicted conformation. When plotted as alpha-helical wheelsaccording to [58], the regions corresponding to amino acids1±11 and 18±41 present an amphipatic conformation, indicat-ing hydrophobic and hydrophilic residues on opposing sides ofthe helix (Fig. 6B). To show this conformation, the secondhelical region has to rotate 1008 with respect to the first, whichcould be satisfied by the flexibility of the residues 12±16proposed by the algorithms of structure prediction.

Functional effects of hadrurin and its synthetic analogs

Synthetic hadrurin proved to be as effective as the native one.Both the antibacterial and the hemolytic effect were preserved.The activity of the synthetic amidated analog is slightlyless than the free acid form, whereas the d-isomeric form ofhadrurin is six times more potent than the native one. This valuecomes from comparing the 5 mm concentration of the all-d-isomer for 100% hemolysis with that of 30 mm required fornative hadrurin and their analogs (Fig. 4). This result shouldcorrelate with an increase in the hydrophobicity of the syntheticd-isomer, due to the substitution of isoleucine for leucineresidues in four positions: 2,6,9 and 27 of the primary structureof hadrurin. It also explains the delay of the retention timeduring HPLC separation, earlier observed by Regnier [59].However, these substitutions did not alter the antimicrobialaction of the synthetic d-isomer. Thus, further developmentsof this work will certainly include substitution of the morehydrophobic residues for less hydrophobic ones, in an attempt

to decrease the cytolytic action and hopefully to cause nochanges to the antibiotic properties of the hadrurin analogs.

The lytic activity of these types of peptides have beenassociated with certain structural characteristics, namely theamphipatic helices formed by the peptide when in contact withthe biological membranes or with artificially prepared bilayersand multilayers, made of lipids with different compositions.Disruption of the membrane integrity and normal functionalityis caused by peptide±lipid interactions, rather than throughreceptor recognition sites.

Previous studies have shown that non hemolytic anti-microbial peptides, such as the cecropins, dermaseptins andmagainins, have a low perturbing activity against zwitterionicmembranes (PtdCho) compared with that against negativecharged phospholipids (PtdSer) [28,41,60,61]. The oppositeeffect occurs with the peptides that have cytolytic activity, suchas melittin [62], pardaxin [63] and LL37 [25]. The assays madewith hadrurin show that it can lyse preferentially zwitterionicphospholipids at a low concentration, whereas a higher criticalmolar ratio peptide/lipid is required to begin the lysis of acidicliposomes. These results indicate that at least two types ofdistinct interactions are involved in the activity of the lyticpeptides: first, hydrophobic interactions, which are favored bythe presence of PC in the liposome model, and this seems to bethe case of hadrurin, possibly promoting its oligomerizationaand second, electrostatic forces, which are more important forthe interaction with acidic phospholipids. However, the factthat the biological membranes are composed of more than 100different lipid molecules with different polar groups and hydro-phobic characteristics [64] allows the possibility of findingdistinct targets, more sensitive to a particular peptide, depend-ing on the cell type under experimentation. Hadrurin is a goodmodel for studying peptide±lipid interactions and could eventu-ally have a therapeutic use if new derivatives are found withouthemolytic activity.

A C K N O W L E D G E M E N T S

This work was partially supported by grant no. 75197±527107 of the

Howard Hughes Medical Institute and a grant from Glaxo-Wellcome

Mexico to L. D. P. The authors want to acknowledge the National Institute

of Public Health of Mexico for providing the bacterial strains used in this

work, and the support of M.Sc. Timoteo Olamendi Portugal and M.Sc.

Silvia Tenorio during various aspects of this work. A. T. L. is a recipient

of a scholarship from the Consejo Nacional de Ciencia y Tecnologia of

Mexico.

R E F E R E N C E S

1. Steiner, H., Hultmark, D., Engstrom, A., Bennich, H. & Boman, H.G.

(1981) Sequence and specificity of two antibacterial proteins

involved in insect immunity. Nature 292, 246±248.

2. Bulet, P., Hetru, C., Dimarcq, J. & Hoffmann, D. (1999) Antimicrobial

peptides in insects; structure and function. Dev. Comp. Immunol. 23,

329±344.

3. Barra, D. & Simmaco, M. (1995) Amphibian skin: a promising

resource for antimicrobial peptides. Trends Biotechnol. 13, 205±209.

4. Broekaert, W.F., Terras, F.R., Cammue, B.P. & Osborn, R.W. (1995)

Plant defensins: novel antimicrobial peptides as components of the

host defense system. Plant Physiol. 108, 1353±1358.

5. Ganz, T. & Lehrer, R.I. (1998) Antimicrobial peptides of vertebrates.

Curr. Opin. Immunol. 10, 41±44.

6. Lemaitre, B., Meister, M., Govind, S., Georgel, P., Steward, R.,

Reichhart, J.M. & Hoffmann, J.A. (1995) Functional analysis and

regulation of nuclear import of dorsal during the immune response in

Drosophila. EMBO J. 14, 536±545.


7. Lemaitre, B., Nicolas, E., Michaut, L., Reichhart, J.M. & Hoffmann,

J.A. (1996) The dorsoventral regulatory gene cassette spatzle/ Toll/

cactus controls the potent antifungal response in Drosophila adults.

Cell 86, 973±983.

8. Braun, A., Hoffmann, J.A. & Meister, M. (1998) Analysis of the

Drosophila host defense in domino mutant larvae, which are devoid

of hemocytes. Proc. Natl Acad. Sci. USA 95, 14337±14342.

9. Goldman, M.J., Anderson, G.M., Stolzenberg, E.D., Kari, U.P., Zasloff,

M. & Wilson, J.M. (1997) Human beta-defensin-1 is a salt-sensitive

antibiotic in lung that is inactivated in cystic fibrosis. Cell 88,

553±560.

10. Lemaitre, B., Reichhart, J.M. & Hoffmann, J.A. (1997) Drosophila host

defense: differential induction of antimicrobial peptide genes after

infection by various classes of microorganisms. Proc. Natl Acad. Sci.

USA 94, 14614±14619.

11. Stolzenberg, E.D., Anderson, G.M., Ackermann, M.R., Whitlock, R.H.

& Zasloff, M. (1997) Epithelial antibiotic induced in states of disease.

Proc. Natl Acad. Sci. USA 94, 8686±8690.

12. Dimopoulos, G., Richman, A., Muller, H.M. & Kafatos, F.C. (1997)

Molecular immune responses of the mosquito Anopheles gambiae to

bacteria and malaria parasites. Proc. Natl Acad. Sci. USA 94,

11508±11513.

13. Dimopoulos, G., Seeley, D., Wolf, A. & Kafatos, F.C. (1998) Malaria

infection of the mosquito Anopheles gambiae activates immune-

responsive genes during critical transition stages of the parasite life

cycle. EMBO J. 17, 6115±6123.

14. Ferrandon, D., Jung, A.C., Criqui, M., Lemaitre, B., Uttenweiler, J.S.,

Michaut, L., Reichhart, J. & Hoffmann, J.A. (1998) A drosomycin-

GFP reporter transgene reveals a local immune response in

Drosophila that is not dependent on the Toll pathway. EMBO J. 17,

1217±1227.

15. Levashina, E.A., Ohresser, S., Lemaitre, B. & Imler, J.L. (1998) Two

distinct pathways can control expression of the gene encoding the

Drosophila antimicrobial peptide metchnikowin. J. Mol Biol. 278,

515±527.

16. Fennell, J.F., Shipman, W.H. & Cole, L.J. (1968) Antibacterial action

of melittin, a polypeptide from bee venom. Proc. Soc. Exp Biol. Med.

127, 707±710.

17. Mignogna, G., Simmaco, M., Kreil, G. & Barra, D. (1993)

Antibacterial and haemolytic peptides containing d-alloisoleucine

from the skin of Bombina variegata. EMBO J. 12, 4829±4832.

18. Simmaco, M.B.D., Severini, C., Aita, M., Erspamer, G.F., Barra, D. &

Bossa, F. (1990) Purification and characterization of bioactive

peptides from skin extracts of Rana esculenta. Biochim. Biophys.

Acta 1033, 318±323.

19. Simmaco, M., Mignogna, G., Barra, D. & Bossa, F. (1993) Novel

antimicrobial peptides from skin secretion of the European frog Rana

esculenta. FEBS Lett. 324, 159±161.

20. Mor, A., Hani, K. & Nicolas, P. (1994d) The vertebrate peptide

antibiotics dermaseptins have overlapping structural features but

target specific microorganisms. J. Biol. Chem. 269, 31635±31641.

21. Subbalakshmi, C., Krishnakumari, V., Nagaraj, R. & Sitaram, N. (1996)

Requirements for antibacterial and hemolytic activities in the bovine

neutrophil derived 13-residue peptide indolicidin. FEBS Lett. 395,

48±52.

22. Skerlavaj, B., Gennaro, R., Bagella, L., Merluzzi, L., Risso, A. &

Zanetti, M. (1996) Biological characterization of two novel

cathelicidin-derived peptides and identification of structural require-

ments for their antimicrobial and cell lytic activities. J. Biol. Chem.

271, 28375±28381.

23. Krishnakumari, V. & Nagaraj, R. (1997) Antimicrobial and hemolytic

activities of crabrolin, a 13-residue peptide from the venom of the

European hornet, Vespa crabro, and its analogs. J. Pept. Res. 50,

88±93.

24. Charpentier, S., Amiche, M., Mester, J., Vouille, V., Le, C.J., Nicolas, P.

& Delfour, A. (1998) Structure, synthesis, and molecular cloning of

dermaseptins B, a family of skin peptide antibiotics. J. Biol. Chem.

273, 14690±14697.

25. Oren, Z., Lerman, J., Gudmundsson, G., Agerberth, B. & Shai, Y.

(1999) Structure and organizaton of the human antimicrobial

peptide LL-37 in phospholipid membranes: relevance to the

molecular basis for its non-cell-selective activity. Biochem. J. 341,

501±513.

26. White, S.H., Wimley, W.C. & Selsted, M.E. (1995) Structure, function,

and membrane integration of defensins. Curr. Opin. Struct. Biol. 5,

521±527.

27. Bechinger, B. (1997) Structure and functions of channel-forming

peptides: magainins, cecropins, melittin and alamethicin. J. Membr.

Biol. 156, 197±211.

28. Matsuzaki, K. (1998) Magainins as paradigm for the mode of action of

pore forming polypeptides. Biochim. Biophys. Acta 1376, 391±400.

29. Boman, H.G., Agerberth, B. & Boman, A. (1993) Mechanisms of

action on Escherichia coli of cecropin P1 and PR-39, two anti-

bacterial peptides from pig intestine. Infect Immun 61, 2978±2984.

30. Chan, Y.R. & Gallo, R.L. (1998) PR-39, a syndecan-inducing anti-

microbial peptide, binds and affects p130(Cas). J. Biol. Chem. 273,

28978±28985.

31. Thevissen, K., Osborn, R.W., Acland, D.P. & Broekaert, W.F. (1997)

Specific, high affinity binding sites for an antifungal plant defensin

on Neurospora crassa hyphae and microsomal membranes. J. Biol.

Chem. 272, 32176±32181.

32. Fant, F., Vranken, W., Broekaert, W. & Borremans, F. (1998) Deter-

mination of the three-dimensional solution structure of Raphanus

sativus antifungal protein 1 by 1H NMR. J. Mol. Biol. 279, 257±270.

33. Hancock, R.E. & Lehrer, R. (1998) Cationic peptides: a new source of

antibiotics. Trends Biotechnol. 16, 82±88.

34. Possani, L.D., Becerril, B., Delepierre, M. & Tytgat, J. (1999) Scorpion

toxins specific for Na1 channels. Eur. J. Biochem. 264, 287±300.

35. Yan, L. & Adams, M.E. (1998) Lycotoxins, antimicrobial peptides from

venom of the wolf spider Lycosa carolinensis. J. Biol. Chem. 273,

2059±2066.

36. Cociancich, S., Goyffon, M., Bontems, F., Bulet, P., Bouet, F., Menez,

A. & Hoffmann, J. (1993) Purification and characterization of a

scorpion defensin, a 4kDa antibacterial peptide presenting structural

similarities with insect defensins and scorpion toxins. Biochem.

Biophys. Res. Commun. 194, 17±22.

37. Ehret, S.L., Loew, D., Goyffon, M., Fehlbaum, P., van Hoffmann, J.A.,

D.A. & Bulet, P. (1996) Characterization of novel cysteine-rich

antimicrobial peptides from scorpion blood. J. Biol. Chem. 271,

29537±29544.

38. Merrifield, B.R. (1963) Solid phase peptide synthesis I: the synthesis of

a tetrapeptide. J. Am. Chem. Soc. 85, 2144±2154.

39. Loew, L.M., Rosenberg, I., Bridge, M. & Gitler, C. (1983) Diffusion

potential cascade. Convenient detection of transferable membrane

pores. Biochemistry 22, 837±844.

40. Sims, P.J., Waggoner, A.S., Wang, C.H. & Hoffman, J.F. (1974) Studies

on the mechanism by which cyanine dyes measure membrane

potential in red blood cells and phosphatidylcholine vesicles.

Biochemistry 13, 3315±3330.

41. Pouny, Y., Rapaport, D., Mor, A., Nicolas, P. & Shai, Y. (1992)

Interaction of antimicrobial dermaseptin and its fluorescently

labeled analogues with phospholipid membranes. Biochemistry 31,

12416±12423.

42. Argiolas, A. & Pisano, J.J. (1984) Isolation and characterization of two

new peptides, mastoparan C and crabrolin, from the venom of the

European hornet, Vespa crabro. J. Biol. Chem. 259, 10106±10111.

43. Rosin, R. (1969) Note on the alpha-hemolytic effect of the venom of

the scorpion Nebo hierichonticus. Toxicon 6, 225±226.

44. Venkaiah, B., Parthasarathy, P.R. & Krishnaswamy, S. (1983) Haemo-

lytic effect of the scorpion Heterometrus fulvipes venom. Biochem.

Int. 7, 241±245.

45. Correa, M.M., Sampaio, S.V., Lopes, R.A., Mancuso, L.C., Cunha,

O.A., Franco, J.J. & Giglio, J.R. (1997) Biochemical and histopatho-

logical alterations induced in rats by Tityus serrulatus scorpion venom

and its major neurotoxin tityustoxin-I. Toxicon 35, 1053±1067.

46. Zamudio, F.Z., Conde, R., AreÂvalo, C., Becerril, B., Martin, B.M.,

Valdivia, H.H. & Possani, L.D. (1997) The mechanisms of inhibition

of ryanodine receptor channels by imperatoxin I, a heterodimeric


protein from the scorpion Pandinus imperator. J. Biol. Chem 272,

11886±11894.

47. Torres-Larios, A., Olamendi-Portugal, T., GoÂmez-Lagunas, F. &

Possani, L.D. (1998) Isolation and characterization of a novel

K-channel blocking toxin from the venom of the scorpion Hadrurus

aztecus. Toxicon 36, 1281.

48. Park, J.M., Jung, J.E. & Lee, B.J. (1994) Antimicrobial peptides from

the skin of a Korean frog, Rana rugosa. Biochem. Biophys. Res.

Commun. 205, 948±954.

49. Amiche, M., Seon, A.A., Pierre, T.N. & Nicolas, P. (1999) The

dermaseptin precursors: a protein family with a common prepro-

region and a variable C-terminal antimicrobial domain. FEBS Lett.

456, 352±356.

50. Lee, J.Y., Boman, A., Sun, C.X., Andersson, M., Jornvall, H., Mutt, V.

& Boman, H.G. (1989) Antibacterial peptides from pig intestine:

isolation of a mammalian cecropin. Proc. Natl Acad. Sci. USA 86,

9159±9162.

51. Chou, P.Y. & Fasman, G.D. (1978) Empirical predictions of protein

conformation. Annu. Rev. Biochem. 47, 251±276.

52. Solovyev, V. & Salamov, A. (1994) Predicting a-helix and b-strand

segments of globular proteins. Comput. Appl. Biosci. 10, 661±669.

53. Geourjon, C. & Deleage, G. (1994) SOPMA: Significant improvement

in protein secondary structure prediction by prediction from

alignments and joint prediction. Comput. Appl. Biosci. 11, 681±684.

54. Geourjon, C. & Deleage, G. (1995) SOPM: a self optimised prediction

method for protein secondary structure prediction. Protein Eng. 7,

157±164.

55. Rost, B. & Sander, C. (1993) Improved prediction of protein secondary

structure by use of sequence profiles and neural networks. Proc. Natl

Acad. Sci. USA 90, 7558±7562.

56. Rost, B. & Sander, C. (1993) Prediction of protein structure at better

than 70% accuracy. J. Mol. Biol. 232, 584±599.

57. Rost, B., Sander, C. & Schneider, R. (1994) PHD ± an automatic mail

server for protein secondary structure prediction. Comput. Appl.

Biosci. 10, 53±60.

58. 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.

59. Regnier, F.E. (1987) The role of protein structure in chromatographic

behavior. Science 238, 319±323.

60. Gazit, E., Lee, W.J., Brey, P.T. & Shai, Y. (1994) Mode of action of the

antibacterial cecropin B2: a spectrofluorometric study. Biochemistry

33, 10681±10692.

61. Gazit, E., Boman, A., Boman, H.G. & Shai, Y. (1995) Interaction of the

mammalian antibacterial peptide cecropin P1 with phospholipid

vesicles. Biochemistry 34, 11479±11488.

62. Matsuzaki, K., Sugishita, K., Fujii, N. & Miyajima, K. (1995a)

Molecular basis for membrane selectivity of an antimicrobial peptide,

magainin 2. Biochemistry 34, 3423±3429.

63. Oren, Z. & Shai, Y. (1996) A class of highly potent antibacterial

peptides derived from pardaxin, a pore-forming peptide isolated

from Moses sole fish Pardachirus marmoratus. Eur. J. Biochem. 237,

303±310.

64. Matsuzaki, K., Sugishita, K.I., Ishibe, N., Ueha, M., Nakata, S.,

Miyajima, K. & Epand, R.M. (1998) Relationship of membrane

curvature to the formation of pores by magainin 2. Biochemistry 37,

11856±11863.


hadrurin, a new antimicrobial peptide from the venom of the scorpion hadrurus aztecus

Documents