ilable at ScienceDirect

Toxicon 52 (2008) 255–263

Contents lists ava

Toxicon

journal homepage: www.elsevier .com/locate/ toxicon

Toxic activities of Brazilian centipede venoms

Marılia B. Malta a, Marcela S. Lira a, Sabrina L. Soares a, Guilherme C. Rocha a, Irene Knysak b,Rosana Martins b, Samuel P.G. Guizze b, Marcelo L. Santoro c, Katia C. Barbaro a,*

a Laboratory of Immunopathology, Butantan Institute, Av. Vital Brasil 1500, 05503-900, Sao Paulo, SP, Brazilb Laboratory of Arthropods, Butantan Institute, Av. Vital Brasil 1500, 05503-900, Sao Paulo, SP, Brazilc Laboratory of Pathophysiology, Butantan Institute, Av. Vital Brasil 1500, 05503-900, Sao Paulo, SP, Brazil

a r t i c l e i n f o

Article history:Received 7 May 2008Accepted 23 May 2008Available online 4 June 2008

Keywords:CentipedeVenomScolopendraCryptopsOtostigmus

* Corresponding author. Tel.: þ55 11 3726 7222x223726 1505.

E-mail addresses: kbarbaro@butantan.gov.br, kBarbaro).

0041-0101/$ – see front matter � 2008 Elsevier Ltddoi:10.1016/j.toxicon.2008.05.012

a b s t r a c t

Centipedes have a venom gland connected to a pair of forceps, which are used to arrestpreys. Human victims bitten by centipedes usually manifest burning pain, paresthesiaand edema, which may develop into superficial necrosis. The aim of this work was tocharacterize and compare toxic activities found in venoms of three species of BraziliancentipedesdOtostigmus pradoi, Cryptops iheringi and Scolopendra viridicornis. By SDS–PAGE (4–20%), important differences were noticed among venoms (between 7 and205 kDa). Few bands showed feeble caseinolytic, fibrinogenolytic and gelatinolytic activi-ties by zymography, but strong hyaluronidase activity was observed in S. viridicornis andO. pradoi venoms. In addition, such activities could be inhibited by o-phenanthroline, in-dicating that these enzymes are metalloproteinases. All venoms induced nociception,edema and myotoxicity in mice, but only S. viridicornis induced mild hemorrhagic activity.No coagulant activity was detected in centipede venoms. Low phospholipase A2 activitywas observed exclusively in S. viridicornis and O. pradoi venoms, but these venoms had in-tense direct hemolytic activity on human erythrocytes. Cross-reactivity among venomswas observed using species-specific sera raised in rabbits. Differences were noticed amongcentipede venoms, but S. viridicornis is indeed the most toxic venom and thereby it couldinduce a more severe envenomation.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Centipedes are terrestrial arthropods belonging to theclass Chilopoda (Negrea and Minelli, 1995), being charac-terized by the presence of a head, and an externally-segmented body containing a pair of articulate legs ineach segment (Barnes et al., 1995). The ventral region ofthe head contains a pair of forcipules, whose extremitiesfinish in venom claws. Such structures are connected tothe short and cylindrical venom gland through a venomduct (Jangi, 1984; Pedroso et al., 2007). Centipedes are

78/2134; fax: þ55 11

barbaro@usp.br (K.C.

. All rights reserved.

predators that use venom primarily to arrest or subduepreys. They have noctivagant habits, and are frequentlyfound inside termitaria, under trunks in decompositionand in underground galleries. Moreover, they like to inhabithidden places that allow an easy adaptation, such as thesurroundings and interior of dwellings in urban areas (Kny-sak and Martins, 1999).

About 2800 species are known in the world, and theyare distributed in all continents except in Antarctica(Barnes et al., 1995). The major biological diversity ofcentipedes is found in subtropical and temperate regions(Knysak and Martins, 1999). The order Scolopendromorphacontains the largest centipedes in the world, and severalspecies are medically important. This order is divided intothree families: Scolopendridae (16 genera), Cryptopidae(7 genera) and Scolopocryptopidae (8 genera) (Schileyko,

M.B. Malta et al. / Toxicon 52 (2008) 255–263256

1996; Shelley, 2002). Scolopendromorpha centipedes aregeneralist predators, but they appear chiefly to eat insects,spiders and other arthropods. The larger species alsoeat small vertebrates, such as rodents, birds and snakes(Stankiewicz et al., 1999). The Scolopendra genus may reachmore than 200 mm, and is commonly found throughoutBrazil, especially in the states of the northern and north-eastern regions.

Cases of humans bitten by centipedes have beenreported in Brazil (Knysak et al., 1998; Barroso et al.,2001; Medeiros et al., in press), but limited information isavailable about the biological activities found in venomsof Brazilian centipedes. Three genera of centipedes aremedically important in Brazil: Otostigmus, Scolopendraand Cryptops (Buecherl, 1946; Knysak et al., 1998; Barrosoet al., 2001, Medeiros et al., in press). Centipede bites admit-ted to Hospital Vital Brazil (Butantan Institute, Brazil) aremainly caused by genera Cryptops and Otostigmus(Medeiros et al., in press), and patients usually manifestmild symptoms, especially pain, edema and erythema(Knysak et al., 1998; Barroso et al., 2001; Medeiros et al.,in press). However, severe cases of centipede envenoma-tion have also been reported, especially in patients bittenby Scolopendra spp. (Harada et al., 1999; Acosta and Cazorla,2004; Ozsarac et al., 2004; Wang et al., 2004; Hasan andHassan, 2005; Yildiz et al., 2006). Taking into considerationthe clinical picture of patients admitted to Hospital VitalBrazil who were bitten by centipedes, we investigated theenzymatic, immunological and biological properties ofvenoms of Otostigmus pradoi, Cryptops iheringi and Scolo-pendra viridicornis, species that are the major agents of cen-tipede bites in Brazil.

2. Materials and methods

2.1. Animals and venoms

Swiss mice (18–20 g) and adult rabbits (3–4 kg) wereprovided by Butantan Institute Animal House. Animals re-ceived food and water ad libitum. Specimens of Otostigmuspradoi (n ¼ 20) and Cryptops iheringi (n ¼ 40) centipedeswere collected in Sao Paulo State, Brazil, and of Scolopendraviridicornis (n ¼ 10) in Tocantins State, Brazil. The specimenswere identified by specialists of Laboratory of Arthropods(Butantan Institute). All specimens were adults, and variedbetween 40 and 80 mm for O. pradoi, 60 and 100 mm forC. iheringi, and 160 and 200 mm for S. viridicornis. These an-imals were maintained in captivity for 3–4 years (dead ani-mals were replaced during this period) and milked byelectrical stimulation once a month. The average amountof venom obtained per animal in each extraction was25 mg from O. pradoi, 17.5 mg from C. iheringi and 100 mgfrom S. viridicornis. Venom pools of at least six different ex-tractions were stored at�20 �C, and thawed at the momentof use. The protein content of venom pools was determinedusing bicinchoninic acid according to Smith et al. (1985),using bovine serum albumin as standard. The proceduresinvolving animals were conducted in conformity withnational laws and policies controlled by Butantan InstituteAnimal Investigation Ethical Committee (protocol no. 115/2002).

2.2. Production of specific rabbit antivenom

Sera against S. viridicornis, O. pradoi or C. iheringi centi-pede venoms were obtained by immunization of rabbits.Two hundred micrograms of each venom extract wasdiluted in 500 mL of PBS and added to 500 mL of completeFreund’s adjuvant, and these mixtures were injected i.m.After 1 month, rabbits received five consecutive boostersof antigen emulsified in incomplete Freund’s adjuvant,with 15-day intervals. Blood was collected and sera wasseparated and stored at �20 �C until used.

2.3. ELISA

Rabbit species-specific sera were titrated by ELISA usingS. viridicornis, O. pradoi and C. iheringi centipede venoms(10 mg/mL) to coat plates (Nunc, USA), according to Theak-ston et al. (1977). The reaction was read using an ELISAreader (Multiskan EX) and the titer determined as the re-ciprocal of the highest dilution that causes an absorbancegreater than 0.050 at 492 nm, as non-specific reactionswere observed below this value.

2.4. Sodium dodecyl sulfate–polyacrylamidegel electrophoresis (SDS–PAGE)

Proteins of S. viridicornis, O. pradoi and C. iheringi centi-pede venoms (5 mg) were analyzed by SDS–PAGE (4–20%acrylamide resolution gels, Pierce, USA) under non-reducing and reducing conditions (Laemmli, 1970). Afterseparation of proteins by electrophoresis, gels were silverstained (Blum et al., 1987). Myosin, b-galactosidase, bovineserum albumin, carbonic anhydrase, soybean trypsin inhib-itor, lysozyme and aprotinin were used as the molecularmass markers (Kaleidoscope pre-stained standards, Bio-Rad, Hercules, CA, USA).

2.5. Western blotting

Proteins of S. viridicornis, O. pradoi and C. iheringi centi-pede venoms (20 mg) were first fractionated by SDS–PAGE(10%) as described above. Electroblotting was performedas described by Towbin et al. (1979). Nitrocellulose mem-branes were incubated with rabbit species-specific sera,diluted at 1/200. Immunoreactive proteins were detectedusing peroxidase-labeled anti-rabbit IgG and the blotwas developed with 0.05% (w/v) 4-chloro-1-naphthol in15% (v/v) methanol in the presence of 0.03% (v/v) H2O2.Non-immunized rabbit serum was used as control. Pre-stained molecular mass markers (BioRad, Hercules, CA,USA) were used.

2.6. Protease and hyaluronidase assays

Zymography was employed to evaluate protease and hy-aluronidase activities of centipede venoms, using casein,gelatin and fibrinogen (Heussen and Dowdle, 1980; Barbaroet al., 2005), and hyaluronic acid from rooster comb (Sigma,St. Louis, MO) (Miura et al., 1995; Barbaro et al., 2005), re-spectively, as substrates. Samples of S. viridicornis, O. pradoiand C. iheringi centipede venoms in non-reducing sample

M.B. Malta et al. / Toxicon 52 (2008) 255–263 257

buffer were loaded, and 10% polyacrylamide gels were runat 20 mA/gel. Clear areas in the gel indicated regions of en-zyme activity. When required, the metal chelating agent1,10-phenanthroline (Sigma Chemicals, St Louis, MO) wasadded in a final concentration of 3 mM to every gel washingand incubation buffers, and then gels were stained as usual.Pre-stained molecular mass markers (BioRad, Hercules, CA,USA) were used.

2.7. Nociceptive and edematogenic activities

To detect the nociceptive activity, mice (n ¼ 6) wereinjected in the right hind paw with 30 mL of PBS containingdifferent doses (0.9, 3.8, 15 and 60 mg) of S. viridicornis, O.pradoi or C. iheringi centipede venoms. Animals wereplaced individually under glass funnels on a mirror. After-wards, the reactivity of animals to lick or bite the injectedfoot was measured, in seconds, during 30 min of experi-mental evaluation (Hunskaar et al., 1985). Animals injectedonly with PBS were used as negative controls.

Edema-forming activity was evaluated at different times(0.25, 1, 4, 24, 48 and 72 h) as the difference of thickness(mm) between the right foot pawdinjected with differentdoses (3.8, 15 and 60 mg) of S. viridicornis, O. pradoi or C.iheringi centipede venoms diluted in PBS or vehicle alone(negative control)dand the left paw of mice (not injected).

2.8. Estimation of myotoxic activity

Mice (n ¼ 8) were injected intramuscularly (i.m.) intothe right gastrocnemius muscle with 120 mg of S. viridicor-nis, O. pradoi or C. iheringi centipede venoms in 50 mL of PBS.The control group was injected with PBS alone. After 3 h,blood was collected from the brachial plexus. Sera of micewere separated and immediately assayed for creatine ki-nase (CK) activity (CK-NAC Liquiform, Labtest, Brazil). Oneunit corresponds to the amount of enzyme that hydrolyzes1 mmol of creatine per minute at 25 �C. Myotoxic activitywas expressed as U/mg of venom of three independent ex-periments. Bothrops jararacussu snake venom (120 mg) wasused as a positive control.

2.9. Coagulant activity

Clotting time was performed according Santoro andSano-Martins (1993). Scolopendra viridicornis, O. pradoiand C. iheringi centipede venoms (30, 60 and 120 mg) di-luted in 50 mL of PBS were added to 200 mL of humanplasma. Samples (duplicate) were observed for 5 min at37 �C to determine the clotting time. After this period,50 mL of thrombin (30 U/mL) (Sigma, St. Louis, MO) wasadded to samples to verify fibrinogen hydrolysis. As a posi-tive control, 50 mL samples of two-fold serially dilutedBothrops jararaca snake venom (1.56–200.0 mg) were usedto determine the minimal coagulant dose (MCD). Experi-ments were carried out in duplicate.

2.10. Direct hemolytic activity

Human blood (type O, Rhþ) was collected in the presenceof 0.15 M sodium citrate (9:1), and centrifuged 1900 � g for

15 min at 10 �C. Red blood cells were obtained after threeconsecutive washes with PBS. Samples (50 mL) of 3% redblood cells were mixed with 100 mL of different doses (3.8,7.5, 15, 30, 60 and 120 mg) of S. viridicornis, O. pradoi and C.iheringi centipede venoms. Each sample (50 mL) was placed(duplicates) in microplates. As controls, distilled water(100% hemolysis) and PBS (0% hemolysis) were used. Micro-plates were kept at room temperature for 3 h. The absorbancewas read using an ELISA reader (Multiskan EX) at 595 nm.

2.11. Phospholipase A2 activity

Phospholipase activity was determined as describedelsewhere (Santoro et al., 1999). Scolopendra viridicornis, O.pradoi or C. iheringi (15 mg) venoms, diluted in 15 mL of PBSpH 7.4, were added to 1.5 mL of reaction solution (100 mMNaCl, 10 mM CaCl2, 7 mM Triton X-100, 0.265% soybeanlecithin, 98.8 mM phenol red, pH 7.6) in a spectrophotometercuvette. The solution was immediately homogenized andread at 558 nm. The definition of 1 U of phospholipase A2 ac-tivity was taken as the amount of toxin (mg of protein/assay)producing a decrease of 0.001 absorbance units per minuteunder the conditions described. Crotalus durissus terrificus(6 mg) was used as a positive control. Phospholipase activitywas expressed as U/mg of two independent experiments.

2.12. Hemorrhagic activity

The hemorrhagic activity was evaluated by the methoddescribed by Kondo et al. (1960) with some modifications.Groups of six mice were shaved on the back and then intra-dermically (i.d.) injected with different doses of S. viridicor-nis, O. pradoi or C. iheringi venoms (15, 30, 60 and 120 mg) inphosphate-buffered saline (PBS). Skins were excised 2 hlater and the diameters of hemorrhagic spots were mea-sured on the internal surfaces. Injections of B. jararacasnake venom (1 mg) or PBS were used as positive and neg-ative controls, respectively.

2.13. Statistical analysis

Results are expressed as means � S.D. Two-way ANOVAfollowed by Bonferroni test was used to analyze data,employing SigmaStat 3.0 software. Values with p < 0.05were considered statistically significant.

3. Results

3.1. Analysis of venoms by SDS–PAGE

Electrophoretic profiles of O. pradoi, C. iheringi and S. vir-idicornis venoms showed that the composition of thesevenoms is distinct, with many unique bands in each venom(Fig. 1). Major differences were observed between 156 and40 kDa, and more densely stained bands were present in O.pradoi venom. On the other hand, C. iheringi venom showedfew weakly stained bands below 40 kDa. After incubationwith 2-mercaptoethanol, several bands around 114–88 kDa disappeared in S. viridicornis venom, and manyothers appeared in the range of 82–40 kDa. Bands of O. pra-doi and C. iheringi venoms also displayed some changes

1 2 1 2 1 2

Sv Op Ci

190.56

125.44

82.28

40.39

31.33

16.877.12

Fig. 1. Electrophoretic pattern of S. viridicornis (Sv), O. pradoi (Op) and C. iheringi(Ci) venoms by SDS–PAGE 4–20%. The gel was silver stained. Venom samples(5 mg) were incubated in sample buffer containing (2) or not containing (1)2-mercaptoethanol. Numbers on the right correspond to the position of molec-ular mass markers.

40.28

81.11

197.70126.37

M.B. Malta et al. / Toxicon 52 (2008) 255–263258

after incubation 2-mercaptoethanol. Venom bands below31 kDa in O. pradoi disappeared after reduction. Manyalterations were observed in the range between 61 and30 kDa in C. iheringi venom.

3.2. Cross-reactivity determined by ELISAand Western blotting

Table 1 shows the comparison of antibody titers, deter-mined by ELISA, of antisera against homologous and heter-ologous antigens. All venom extracts were immunogenicand induced the production of high levels of antibodies inrabbits. Cross-reactivity among venom extracts wasdetected, and no significant differences on titers were no-ticed (only variations higher than two-fold dilutions wereconsidered significant) when anti-S. viridicornis serumwas used. However, anti-O. pradoi serum presented low ti-ters against C. iheringi venom, and anti-C. iheringi serumweakly recognized S. viridicornis venom components.

Table 1Antigenic cross-reactivity by ELISA among Scolopendra viridicornis, Otos-tigmus pradoi and Cryptops iheringi centipede venoms using homologousor heterologous rabbit antisera

Antisera Venoms

S. viridicornis O. pradoi C. iheringi

Anti-S. viridicornis 512,000a 512,000 256,000Anti-O. pradoi 64,000 128,000 16,000Anti-C. iheringi 16,000 128,000 512,000

a ELISA titers. Microplates were coated with each venom, and then incu-bated with homologous or heterologous rabbit antiserum. Initial dilutionwas 1/250.

Fig. 2 shows immunoblots of S. viridicornis, O. pradoi andC. iheringi venoms after incubation with species-specificsera produced in rabbits. Many components of all venoms,mainly above 47 kDa, could be detected. Homologous anti-sera strongly detected major components around 12–18 kDa in their respective venoms. Anti-S. viridicornis se-rum recognized many components in heterologousvenoms, but a component around 12 kDa was only recog-nized by the homologous antiserum. On the other hand,anti-O. pradoi serum showed less antigenic cross-reactivityagainst the other two venoms.

3.3. Enzymatic activities of centipede venoms

Enzymatic activities were detected after incorporationof casein, gelatin, fibrinogen and hyaluronic acid in 10%polyacrylamide gels (Fig. 3). Gelatinolytic activity wasdetected in bands around 44 kDa in S. viridicornis venom,35, 23 and 15 kDa in O. pradoi, and 121, 44 and 25 kDa inC. iheringi venoms. Caseinolytic activity was weakly ob-served around 44 and 22 kDa in S. viridicornis, 35, 23 and15 kDa in O. pradoi venom, and 44 kDa in C. iheringi venom.Fibrinogenolytic activity was weakly detected exclusivelyin S. viridicornis venom (44 kDa). Most of these enzymeswere metalloproteinases, since their activities wereabolished after treatment with 1,10-phenanthroline. Hyal-uronidase activity was detected in a diffuse region be-tween 66 and 40 kDa in S. viridicornis and O. pradoivenoms; another band with 32 kDa was detected in O. pra-doi venom.

Low phospholipase A2 activity was detected in venoms(15 mg), 3900 U/mg to S. viridicornis and 3967 U/mg to O.pradoi. No phospholipase A2 activity was detected in C. iher-ingi venom. Crotalus durissus terrificus (6 mg) snake venom(117,000 U/mg) was used as a positive control.

Sv Op Ci Sv Op Ci Sv Op Ci

6.77

17.09

31.27

ASv AOp ACi

Fig. 2. Antigenic cross-reactivity of S. viridicornis (Sv), O. pradoi (Op) and C. iher-ingi (Ci) venoms (20 mg) detected by Western blotting, using rabbit anti-S.viridicornis (ASv), anti-O. pradoi (AOp) and anti-C. iheringi (ACi) sera. Numberson the right correspond to the position of molecular mass markers.

Gelatin Hyaluronic acid

Casein Fibrinogen

Sv Op Ci Sv Sv SvOp Ci Op Ci Op Ci

Sv Op Ci Sv Op Ci Sv Op Ci

A B A B

A B

190.56125.44

82.28

40.39

31.33

16.87

7.12

190.56

190.56 190.56

125.44

125.44125.44

82.28

82.28

82.28

40.39

40.39

40.39

31.33

31.33

31.33

16.87

16.8716.87

7.12

7.127.12

Fig. 3. The technique of substrate SDS–PAGE 10% was used to determine the caseinolytic (30 mg), gelatinolytic (20 mg), fibrinogenolytic (40 mg) and hyaluronidase(1 mg) activities of S. viridicornis (Sv), O. pradoi (Op) and C. iheringi (Ci) venoms. The venoms were incubated observed in the absence (A) or the presence (B) of3 mM 1,10-phenanthroline (final concentration). Numbers on the right correspond to the position of molecular mass markers. Clear areas in the gel indicateregions of enzyme activity.

M.B. Malta et al. / Toxicon 52 (2008) 255–263 259

3.4. Toxic activities

All venoms induced nociceptive activity in a dose-dependent manner, which could be observed from dosesas low as 0.9 mg. No statistically significant differenceswere observed among venoms for the dose of 60 mg(Fig. 4). The highest edematogenic activity of venomextracts was noticed 15 min after venom injection (Fig. 5).Edema was dose dependent and decreased progressivelyuntil 72 h, when significant edema was still observed(dose: 60 mg) (Fig. 5).

Fig. 6 shows that centipede venoms had high myotoxicactivity and could induce a remarkable CK release. Animalsinjected with PBS (965.0 � 285.6 U/L) or B. jararacussusnake venom (120 mg) (3299.3 � 1273.3 U/L) were used asnegative and positive controls, respectively.

No coagulant activity was detected in centipede venomseven when higher doses of venom (120 mg) were used. Afteraddition of thrombin (30 U/mL) to samples previously incu-bated with venoms, plasma clotted within 10–13 s, which isconsidered a normal value, indicating that fibrinogen chainswere not hydrolyzed by proteases of centipede venoms.

Only S. viridicornis and O. pradoi induced direct hemolyticactivity in human O positive red blood cells in a dose-dependent manner, which could be observed even usinglow doses (3.8 mg) of venoms. However, C. iheringi venomshowed low hemolytic activity (13.0% of hemolysis), evenwhen doses as high as 120 mg of venom were used (Fig. 7).

Hemorrhagic activity was observed only in S. viridicornisvenom (Table 2), and it was dose-dependent. However, itwas considered low if compared with B. jararaca snakevenom, used as a positive control.

PBS 0.9 3.8 15 600

50

100

150

****

****

Tim

e (s)

0

50

100

150

Tim

e (s)

0

100

200

300

Tim

e (s)

****

A

B

C

Groups

PBS 0.9 3.8 15 60Groups

PBS 0.9 3.8 15 60Groups

*

**

****

Fig. 4. Nociceptive activity of S. viridicornis (A), O. pradoi (B) and C. iheringi(C) venoms. The reactivity was determined by the time (in seconds) that an-imals lick or bite the injected paw during 30 min of observation. *p < 0.01,statistically different from the control group (PBS). Data are expressed asmean � S.D. (n ¼ 6). The abscissas represent animal injected with vehicle(PBS) and S. viridicornis, O. pradoi or C. iheringi venoms (0.9–60 mg).

∗

∗∗

∗

∗

∗ ∗

∗

∗

∗

∗∗

∗ ∗ ∗

A

∗∗

∗∗

∗

∗

∗∗

∗∗

∗

∗

∗∗ ∗

∗

B

C

0.25 1 4 24 48 72

0.25

0.20

0.15

0.10

0.05

0.00

∗∗

∗

∗

∗

∗∗

∗

∗

∗

∗∗

∗ ∗ ∗∗

0.25

0.20

0.15

0.10

0.05

0.00

0.30

15 µg 60 µg

PBS3.8 µg

15 µg 60 µg

PBS3.8 µg

15 µg 60 µg

PBS3.8 µg

Time (hours)

0.25 1 4 24 48 72Time (hours)

0.25 1 4 24 48 72Time (hours)

Diffe

re

nc

e o

f th

ic

kn

es

s (c

m)

0.25

0.20

0.15

0.10

0.05

0.00

Diffe

re

nc

e o

f th

ic

kn

es

s (c

m)

Diffe

re

nc

e o

f th

ic

kn

es

s (c

m)

Fig. 5. Edematogenic activity (mean � S.D.) of S. viridicornis (A), O. pradoi (B)and C. iheringi (C) venoms. Edema was calculated by the difference (in cm)between the injected and control paw (not injected). *p < 0.05, statisticallydifferent from the control group (PBS).

M.B. Malta et al. / Toxicon 52 (2008) 255–263260

4. Discussion

Centipedes are arthropods that have a venom apparatuscontaining a pair of forcipules connected to a venom glandthrough a venom duct localized in the ventral shield of thehead (postcephalic segment) (Menez et al., 1990). Centi-pede venoms have been poorly characterized in the litera-ture, and the few published studies have reported thepresence of esterases, proteinases, alkaline and acid phos-phatases, cardiotoxins, histamine, and neurotransmitter-releasing compounds in Scolopendra venoms (Gomeset al., 1983; Mohamed et al., 1983; Stankiewicz et al.,1999; Gutierrez et al., 2003). Recently, Rates et al. (2007),using a proteomic approach for Scolopendra viridicornis

nigra and Scolopendra angulata venoms, detected morethan 60 proteins/peptides, and a number of them weretoxic to insects. Herein, the venoms of three important spe-cies of centipedes have been characterized, and thesevenoms were noticed to evoke remarkable nociception,edema and myotoxicity in mice.

Our results showed that the electrophoretic profiles ofS. viridicornis, O. pradoi and C. iheringi were different.

0

1000

2000

3000

4000

5000

Cre

atin

oq

uin

as

e U

/L

PBS Bj Sv Ci Op

∗∗

∗∗

Fig. 6. Myotoxic activity (mean � S.D.) of S. viridicornis (Sv), O. pradoi (Op)and C. iheringi (Ci) venoms (120 mg). *p < 0.05, significantly different fromthe negative control (PBS). Bothrops jararacussu (Bj) snake venom (120 mg)was used as a positive control group.

Table 2Hemorrhagic activity of Bothrops jararaca and Scolopendra viridicornisvenoms

Venom (dose) Area, cm2 (mean � S.D.) Intensity

B. jararaca (1 mg) 152.3 � 55.8 þþþþS. viridicornis (15 mg) 13.2 � 7.9 þS. viridicornis (30 mg) 17.0 � 9.2 þþS. viridicornis (60 mg) 23.6 � 11.9 þþS. viridicornis (120 mg) 26.5 � 9.4 þþþ

Hemorrhagic activity of B. jararaca and S. viridicornis venoms wasexpressed as mean � S.D. (n ¼ 6). PBS was used as a negative control.

M.B. Malta et al. / Toxicon 52 (2008) 255–263 261

Similarities were observed among them, especially below40 kDa for O. pradoi and S. viridicornis venoms, and bothof them presented a more complex banding pattern com-pared to C. iheringi venom. This variability, as well as theinherent venom variation due to the use of venom pools,can explain the differences in toxicity observed amongvenoms. After incubation with 2-mercaptoethanol, a vari-ety of bands disappeared, indicating thereby the presenceof proteins with more than one polypeptide chain.However, further studies are necessary to demonstratethe homology of proteins found in centipede venoms.

Low caseinolytic and gelatinolytic activities weredetected in all venoms, and S. viridicornis and O. pradoivenoms also showed low fibrinogenolytic activity. Most ofthese enzymatic components seems to be metalloprotei-nases, since they were inhibited by incubation with 1,10-phenanthroline. In these venoms, the profile of enzymaticdegradation using different substrates was alike, indicatingthe presence of proteases with broad substrate specificity.We observed an intense hyaluronidase activity, especially

3.8 7.5 15.0 30.0 60.0 120.0

Hem

olysis (%

)

0

20

40

60

80

100

120

Dose

S. viridicornis

O. pradoi

C. ihering

Fig. 7. Direct hemolytic activity in vitro of S. viridicornis, O. pradoi and C. iher-ingi venoms. Distilled water (100% of hemolysis) was used as a positivecontrol, and PBS pH 7.4 as a negative control (0% of hemolysis).

in S. viridicornis venom, which could contribute to amplifythe local damage induced by other proteases and alsofacilitate venom spreading. In fact, as shown for othervenomous animals, these enzymes can cause disturbancesin the extracellular matrix, favoring the establishment oflocal injury and functioning as a diffusion factor (Tan andPonnudurai, 1992; Birkedal-Hansen et al., 1993; Veigaet al., 2000; Barbaro et al., 2005, 2007; Lira et al., 2007).

Cross-reactivity was noticed among all venoms by ELISAand Western blotting using species-specific sera raised inrabbits. Using ELISA, we verified that centipede venomswere immunogenic and could stimulate humoral immuneresponse. By Western blotting, several components ofhigher molecular weight were recognized by all antisera.Anti-S. viridicornis serum recognized many homologouscomponents and higher molecular mass components ofheterologous venoms. Besides, we verified that anti-C. iher-ingi and anti-O. pradoi sera weakly recognized componentsof S. viridicornis venom, indicating the presence of fewcommon epitopes among these venoms. Interestingly, anti-bodies present in the specific anti-S. viridicornis serumcould strongly recognize venom components of O. pradoiand C. iheringi venoms.

The biological activities investigated herein took intoaccount the main symptoms described in victims ofcentipede bites (Knysak et al, 1998; Barroso et al., 2001;Medeiros et al., in press). We observed that all centipedevenoms can induce nociception activity in a dose-depen-dent manner likely due to action of several venomcomponents, but their pharmacodynamics has not beentotally elucidated. Gomes et al. (1982) and Mcfee et al.(2002) reported that the pain manifested by victimsdepends on the size of the centipede and the site of thebite. Our results showed that the edematogenic activityinduced by centipede venoms was similar and of rapid on-set (15 min). However, the pathogenesis of edema inducedby centipede venoms remains to be elucidated. Histamine,found in some centipede venoms (Gomes et al., 1982) couldcontribute to the development of pain and edema.

Unlike S. viridicornis venom, O. pradoi and C. iheringivenoms presented no hemorrhagic activity. However, thelocal hemorrhage evoked by S. viridicornis venom was dis-crete when compared to that caused by B. jararaca snakevenom, the positive control. Hemorrhage might be causedby enzymes found in S. viridicornis venom, inasmuch as itshowed higher gelatinolytic, caseinolytic, hyaluronidaseand fibrinogenolytic activities than O. pradoi and C. iheringivenoms. Studies with Bothrops spp. venoms indicate thatlocal hemorrhage is likely associated with the degradation

M.B. Malta et al. / Toxicon 52 (2008) 255–263262

of proteins that form the basal membrane of capillaries. Be-sides, endogenous inflammatory mediators generated byvenom components can indirectly contribute to blood leak-age (Gutierrez and Lomonte, 2003).

Centipede venoms neither prolonged the coagulationtime nor degraded plasma fibrinogen, indicating that theydo not interfere with coagulation factors as Bothrops spp.snake venom and Lonomia obliqua bristle extract do(Sano-Martins and Santoro, 2003).

In spite of the low phospholipase A2 activity, we verifiedthat S. viridicornis and O. pradoi could directly lyse erythro-cytes, demonstrating that these venoms have potenthemolysins, similar to those found in other venoms(Lopes-Ferreira et al., 1998; Sano-Martins and Santoro,2003). Indirect hemolytic activity is induced by the actionof phospholipases A2 on exogenous lecithin, releasing fattyacids and lysolecithin, which can lyse erythrocyte mem-branes. Nonetheless, S. viridicornis and O. pradoi venomsalso seem to contain a direct lytic factor, which does notrequire exogenous lecithin.

When we used the dose of 120 mg, all centipede venomspresented myotoxic activity, which was statistically differ-ent from the negative control (PBS). The inflammatory lo-cal reaction induced by centipede venoms resulted inedema and tissular damage, which can also provoke mus-cular injury. Like the myotoxicity provoked by Bothropsspp. venoms (Gutierrez and Lomonte, 2003), phospholi-pases A2 could partially contribute to the myotoxicity ob-served in S. viridicornis and O. pradoi venoms, but howimportant they are is speculative. However, myonecrosisdoes not seem to be a major problem in centipede bites,and it occurs only when high doses of centipede venomsare injected.

Our results showed that venoms of S. viridicornis andO. pradoi, which belong to the family Scolopendridae,are more similar than that of C. iheringi (family Cryptopi-dae), likely due to their phylogenetic relationship. Althoughthe differences noticed in the electrophoretic profile andenzymatic activities of S. viridicornis, O. pradoi and C. iheringivenoms, edematogenic, myotoxic and nociceptive activitieswere comparable. On the other hand, hemorrhagic andhemolytic activities were diverse among venoms, suggest-ing that these venoms have non-identical toxic components,which might induce different clinical pictures (Medeiroset al., in press). The comparative morphological analysis ofvenom glands of these three species indicates that they aresimilar, except for the gland size, which is proportional tothe body size of animals. In addition, histochemical analysesreveal that their venom secretion basically containsglycoproteins (Pedroso et al., 2007).

The venom composition and the smaller size of O. pradoiand C. iheringi evidence that envenomation caused by thesespecies is less severe, as demonstrated clinically (Medeiroset al., in press). On the other hand, the variety of active pro-teins found in S. viridicornis venom is likely due to the greatdiversity of preys ingested by them (manuscript in prepara-tion). Furthermore, the larger size, the higher quantity ofvenom released from the venom gland and the more offen-sive behavior of S. viridicornis demonstrate that this speciesis potentially more dangerous to humans, especiallychildren.

Acknowledgments

This work was supported by FAPESP (03/04527-1).M.S.L. and M.B.M. received respectively a FUNDAP andFAPESP fellowship. The authors thank Danieli M. Rangel,Thais A. Oliveira and Letıcia M.P. Martins for technical assis-tance. We also thank CNPq for the grant to K.C.B. (306158/2004-3). IBAMA provided animal collection permission no.02027.002695/2004 and CGEN provided the license for ge-netic patrimony access (221/2006).

Ethical statement

The authors warrant that this manuscript is an originalwork, it has not been published before and it is not submit-ted for publication anywhere else. It contains no libelous orother unlawful statements, and it does not infringe on therights of others. The authors have no relationship withany manufactures or distributors of products used in thismanuscript. This paper reflects our own research and anal-ysis and does so in a truthful and complete manner. All au-thors have contributed significantly to the execution,analysis and writing of the study and all co-authors haveagreed to submit the manuscript to Toxicon. Moreover,the manuscript is appropriately placed in the context ofprior and existing research.

Conflict of interest

The authors declare that there are no conflicts ofinterest.

References

Acosta, M., Cazorla, D., 2004. Centipede (Scolopendra sp.) envenomationin a rural village of semi-arid region from Falcon State, Venezuela.Rev. Invest. Clin 56, 712–717.

Barbaro, K.C., Knysak, I., Martins, R., Hogan, C., Winkel, K., 2005. Enzy-matic characterization antigenic cross-reactivity and neutralizationof dermonecrotic activity of five Loxosceles spider venoms of medicalimportance in the Americas. Toxicon 45, 489–499.

Barbaro, K.C., Lira, M.S., Soares, S.L., Malta, M.B., Garrone-Neto, D.,Cardoso, J.L.C., Santoro, M.L., Haddad Jr., V., 2007. Comparative studyon extracts from the tissue covering the spines of freshwater (Pota-motrygon falkneri) and marine (Dasyatis guttata) stingrays. Toxicon50, 676–687.

Barnes, R.S.K., Calow, P., Olive, P.J.W., 1995. Os Invertebrados: Uma NovaSıntese. Atheneu, Sao Paulo, 526 p.

Barroso, E., Hidaka, A.S.V., Santos, A.X., França, J.D.M., Sousa, A.M.B.,Valente, J.R., Magalhaes, A.F.A., Pardal, P.P.O., 2001. Acidentes porcentopeias notificados pelo ‘‘Centro de Informaçoes Toxicologicas deBelem’’, num perıodo de dois anos. Rev. Soc. Bras. Med. Trop. 34,527–530.

Birkedal-Hansen, H., Moore, W.G., Bodden, M.K., Windsor, L.J., Birkedal-Hansen, B., DeCarlo, A., Engler, J.A., 1993. Matrix metalloproteinases:a review. Crit. Rev. Oral Biol. Med 4, 197–250.

Blum, H., Beier, H., Gross, H.J., 1987. Improved silver staining of plant pro-teins, RNA and DNA in polyacrylamide gels. Electrophoresis 8, 93–95.

Buecherl, W., 1946. Açao do veneno dos escolopendromorfos no Brasil, so-bre alguns animais de laboratorios. Mem. Inst. Butantan 19, 181–198.

Gomes, A., Datta, A., Bandita, S., Kar, P.K., Lahiri, S.C., 1982. Occurrence ofhistamine and histamine release by centipede venom. Indian J. Med.Res. 76, 888–891.

Gomes, A., Datta, A., Sarangi, B., Kar, P.K., Lahiri, S.C., 1983. Isolation, puri-fication, pharmacodynamics of a toxin from the venom of the centi-pede Scolopendra subspinipes dehaani Brandt. Indian J. Exp. Biol. 21,203–207.

M.B. Malta et al. / Toxicon 52 (2008) 255–263 263

Gutierrez, M.C., Abarca, C., Possani, L.D., 2003. A toxic fraction from Scol-opendra venom increase the basal release of neurotransmitters in theventral ganglia of crustaceans. Comp. Physiol. C 135, 205–214.

Gutierrez, J.M., Lomonte, B., 2003. Efeitos locais no envenenamento ofı-dico na America Latina. In: Cardoso, J.L.C., França, F.O.S., Wen, F.H.,Malaque, C.M.S., Haddad Jr., V. (Eds.), Animais Peçonhentos no Brasil.Biologia, Clınica e Terapeutica dos Acidentes. Sarvier, pp. 324–333.

Harada, K., Asa, K., Imachi, T., Yamaguchi, Y., 1999. Centipede inflictedpostmortem injury. J. Forensic Sci. 44, 849–850.

Hasan, S., Hassan, K., 2005. Proteinuria associated with centipede bite.Pediatr. Nephrol 20, 550–551.

Heussen, C., Dowdle, E.B., 1980. Electrophoretic analysis of plasminogenin polyacrylamide gels containing sodium dodecyl sulfate and copoly-merized substrates. Anal. Biochem 102, 196–202.

Hunskaar, S., Fasmer, O.B., Hole, K., 1985. Formalin test in mice, a usefultechnique for evaluating mild analgesics. J. Neurosci. Methods 14, 69–76.

Jangi, B.S., 1984. Centipede venoms and poisoning. In: Tu, A.T. (Ed.), Hand-book of Natural Toxins: Insect Poisons, Allergens and Other Inverte-brate Venoms. Marcel Dekker, New York, pp. 333–368.

Knysak, I., Martins, R., Bertim, C.R., 1998. Epidemiological aspects of cen-tipede (Scolopendromorphae: Chilopoda) bites registered in Greater S.Paulo, SP, Brazil. Rev. Saude Pub 32, 514–518.

Knysak, I., Martins, R., 1999. Myriapoda. In: Brandao, C.R., Cancello, E.M.(Eds.), Biodiversidade do Estado de Sao Paulo, Brasil: Sıntese do Conhe-cimento do Final do Seculo XX, vol. 5: invertebrados terrestres, p. 279.

Kondo, H., Kondo, S., Ikesawa, H., Murata, R., Ohsaka, A., 1960. Studies onthe quantitative method for determination of hemorrhagic activity ofHabu snake venom. Jpn. J. Med. Sci. Biol. 13, 43–51.

Laemmli, U.K., 1970. Cleavage of structural proteins during assembly ofthe head of bacteriophage T4. Nature 227, 680–685.

Lira, M.S., Furtado, M.F., Martins, L.M.P., Lopes-Ferreira, M., Santoro, M.L.,Barbaro, K.C., 2007. Enzymatic and immunochemical characterizationof Bothrops insularis venom and its neutralization by polyspecific Bo-throps antivenom. Toxicon 49, 982–994.

Lopes-Ferreira, M., Barbaro, K.C., Cardoso, D.F., Moura-da-Silva, A.M.,Mota, I., 1998. Thalassophryne nattereri fish venom: biological and bio-chemical characterization and serum neutralization of its toxic activ-ities. Toxicon 36, 405–410.

Mcfee, R.B., Caraccio, T.R., Mofenson, H.C., McGuigan, M.A., 2002. Enveno-mation by the Vietnamese centipede in a Long Island pet store. Clin.Toxicol 40, 573–574.

Medeiros, C.R., Susaki, T.T. Knysak I., Malaque C.M.S., Wen, F.H., Santoro,M.L., França, F.O.S., Barbaro K.C. Epidemiologic and clinical survey ofvictims of centipede bites admitted to Hospital Vital Brazil (Sao Paulo,Brazil). Toxicon, in press. doi:10.1016/j.toxicon.2008.07.009

Menez, A., Zimmerman, K., Zimmerman, S., Heatwole, H., 1990. Venomapparatus and toxicity of the centipede Ethmostigmus rubripes (Chilo-poda, Scolopendridae). J. Morphol 206, 303–312.

Miura, R.O., Yamagata, S., Miura, Y., Harada, T., Yamagata, T., 1995. Analysisof glycosaminoglycan-degrading enzymes by substrate gel electro-phoresis (Zymography). Anal. Biochem 225, 333–340.

Mohamed, A.H., Abu-Sinna, G., El-Shabaka, H.A., El-Aal, A.A., 1983. Pro-teins, lipids, lipoproteins and some enzyme characterizations of thevenom extract from the centipede Scolopendra morsitans. Toxicon21, 371–377.

Negrea, S., Minelli, A., 1995. Chilopoda. In: Decu, V., Juberthie, C. (Eds.).Enciclopedia Bioespeologica, 1. Moulis, France, pp. 249–254.

Ozsarac, M., Karcioglu, O., Ayrik, C., Somuncu, F., Gumrukcu, S., 2004.Acute coronary ischemia following centipede envenomation: case re-port and review of the literature. Wild. Environ. Med 15, 109–112.

Pedroso, C.M., Antoniazzi, M.M., Knysak, I., Martins, R., Guizze, S.P.G.,Barbaro, K.C., 2007. Comparative morphological study of venomglands of Cryptops iheringi, Otostigmus pradoi and Scolopendra viridi-cornis centipede. Mem. Inst. Butantan 64 4.13 (eletronic midia).

Rates, B., Bemquerer, M.P., Richardson, M., Borges, M.H., Morales, R.A.V.,De Lima, M.E., Pimenta, A.M.C., 2007. Venomic analyses of Scolopen-dra viridicornis nigra and Scolopendra angulata (Centipede, Scolopen-dromorpha): Shedding light on venoms from a neglected group.Toxicon 49, 810–826.

Sano-Martins, I.S., Santoro, M.L., 2003. Disturbios hemostaticos em enve-nenamentos por animais peçonhentos no Brasil. In: Cardoso, J.L.C.,França, F.O.S., Wen, F.H., Malaque, C.M.S., Haddad Jr., V. (Eds.), AnimaisPeçonhentos no Brasil Biologia, Clınica e Terapeutica dos Acidentes.Sarvier, pp. 289–309.

Santoro, M.L., Sano-Martins, I.S., 1993. Different clotting mechanisms ofBothrops jararaca snake venom on human and rabbit plasmas. Toxi-con 31, 733–742.

Santoro, M.L., Sousa-e-Silva, M.C., Gonçalves, L.R., Almeida-Santos, S.M.,Cardoso, D.F., Laporta-Ferreira, I.L., Saiki, M., Peres, C.A., Sano-Martins, I.S., 1999. Comparison of the biological activities in venomsfrom three subspecies of the South American rattlesnake (Crotalusdurissus terrificus, C. durissus cascavella and C. durissus collilineatus).Comp. Biochem. Physiol. C Pharmacol. Toxicol. Endocrinol 122, 61–73.

Schileyko, A., 1996. Some problems in the systematics of the order Scolo-pendromorpha (Chilopoda). In: Geoffroy, J.J., Mauries, J.P., NaguyenDuy-Jacquemin, M. (Eds.), Acta Myriapodologica. Mem. Mus. Nat.Hist. Nat, 169, pp. 293–297.

Shelley, R.M., 2002. A synopsis of the North American centipedes of theScolopendromorpha (Chilopoda). Mem. Virginia Mus. Nat. Hist. 5,1–108.

Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H.,Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J., Klenk, D.C.,1985. Measurement of protein using bicinchoninic acid. Anal. Biochem150, 76–85.

Stankiewicz, M., Hamon, A., Benkhalifa, R., Kadziela, W., Hue, B.,Lucas, S., Mebs, D., Pelhate, M., 1999. Effects of a centipede venomfraction on insect nervous system, a native Xenopus oocyte receptorand on expressed Drosophila muscarinic receptor. Toxicon 37, 1431–1445.

Tan, N.H., Ponnudurai, G., 1992. Comparative study of the enzymatic,hemorrhagic, procoagulant and anticoagulant activities of someanimal venoms. Comp. Biochem. Physiol. C 103, 299–302.

Theakston, R.D.G., Lloyd-Jones, M.J., King, L.E., 1977. Micro-ELISA fordetecting and assaying snake venom and anti-venom antibody.Lancet 2, 639–641.

Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets: Proce-dure and some applications. Proc. Natl. Acad. Sci. USA 76, 4350–4354.

Veiga, S.S., da Silveira, R.B., Dreyfuss, J.L., Haoach, J., Pereira, A.M.,Mangili, O.C., Gremski, W., 2000. Identification of high molecularweight serine proteases in Loxosceles intermedia (brown spider)venom. Toxicon 38, 825–839.

Wang, I.K., Hsu, S.P., Lee, K.F., Lin, P.Y., Chang, H.W., Chuang, F.R., 2004.Rhabdomyolysis, acute renal failure and multiple focal neuropathiesafter drinking alcohol soaked with centipede. Renal Fail. 26, 93–97.

Yildiz, A., Biceroglu, S., Yakut, N., Bilir, C., Akdemir, R., Akilli, A., 2006.Acute myocardial infarction in a young man caused by centipedesting. Emerg. Med. J 23, e30.