Comparative Biochemistry and Physiology Part C 129 Ž2001. 151᎐162

Venom of the crotaline snake Atropoides nummifer ž jumping viper/ from Guatemala and Honduras: comparative toxicological characterization, isolation of a myotoxic phospholipase A 2 homologue and neutralization by two antivenoms Ermila Rojasa,c , Patricia Saraviab, Yamileth Angulo a,c , Viviana Arce a , b Bruno Lomonte a , Juan Jose , ´ Chavez ´ b, Ruben ´ Velasquez ´ a,U Monica Thelestam d, Jose Marıa Gutierrez ´ ´ ´ b

a Instituto Clodomiro Picado, Facultad de Microbiologıa, ´ Uni¨ ersidad de Costa Rica, San Jose, ´ Costa Rica Departamento de Bioquımica, Facultad de Ciencias Quımicas y Farmacia, Uni¨ ersidad de San Carlos de Guatemala, ´ ´ Guatemala, Guatemala c Departamento de Bioquımica, Escuela de Medicina, Uni¨ ersidad de Costa Rica, San Jose, ´ ´ Costa Rica d Microbiology and Tumor Biology Center, Karolinska Institutet, Stockholm, Sweden

Received 20 December 2000; received in revised form 16 April 2001; accepted 17 April 2001

Abstract A comparative study was performed on the venoms of the crotaline snake Atropoides nummifer from Guatemala and Honduras. SDS-polyacrylamide gel electrophoresis, under reducing conditions, revealed a highly similar pattern of these venoms, and between them and the venom of the same species from Costa Rica. Similar patterns were also observed in ion-exchange chromatography on CM-Shephadex C-25, in which a highly basic myotoxic fraction was present. This fraction was devoid of phospholipase A 2 activity and strongly reacted, by enzyme-immunoassay, with an antiserum against Bothrops asper myotoxin II, a Lys-49 phospholipase A 2 homologue. A basic myotoxin of 16 kDa was isolated to homogeneity from the venom of A. nummifer from Honduras, showing amino acid composition and N-terminal sequence similar to those of Lys-49 phospholipase A 2 variants previously isolated from other crotaline snake venoms. Guatemalan and Honduran A. nummifer venoms have a qualitatively similar toxicological profile, characterized by: lethal; hemorrhagic; myotoxic; edema-forming; coagulant; and defibrinating activities, although there were significant quantitative variations in some of these activities between the two venoms. Neutralization of toxic activities by two commercially-available antivenoms in the region was studied. Polyvalent antivenom produced by Instituto Clodomiro Picado was effective in the neutralization of: lethal; hemorrhagic; myotoxic; coagulant; defibrinating; and phospholipase

U

Corresponding author. Fax: q506-2920485. E-mail address: [email protected] ŽJ.M. Gutierrez ´ ..

1532-0456r01r$ - see front matter 䊚 2001 Elsevier Science Inc. All rights reserved. PII: S 1 5 3 2 - 0 4 5 6 Ž 0 1 . 0 0 1 9 8 - 3

152

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

A 2 activities, but ineffective against edema-forming activity. On the other hand, MYN polyvalent antivenom neutralized: hemorrhagic; myotoxic; coagulant; defibrinating; and phospholipase A 2 activities, albeit with a lower potency than Instituto Clodomiro Picado antivenom. MYN antivenom failed to neutralize lethal and edema-forming activities of A. nummifer venoms. 䊚 2001 Elsevier Science Inc. All rights reserved. Keywords: Snake venom; Viperids; Atropoides nummifer; Myotoxic phospholipase A 2 ; Lys-49 phospholipase A 2 ; Neutralization; Antivenoms

1. Introduction Crotaline snakes Žfamily Viperidae, subfamily Crotalinae. are widely distributed in Latin America, comprising a large number of species which are responsible for the majority of snakebite envenomations in this region ŽCampbell and Lamar, 1989; Fan and Cardoso, 1995; Gutierrez, 1995.. ´ The venoms of some of these species have been partially characterized, but very little is known about the venom biochemistry and pharmacology of many species, or about the intraspecies venom variability. Atropoides nummifer, known as ‘jumping viper’, ‘mano de piedra’ or ‘timbo’, is a crotaline snake formerly classified within the genus Bothrops. However, more detailed systematic analyses prompted its classification in the new genus Atropoides ŽWerman, 1992.. It is a stout, terrestrial pit viper distributed in low, moderate and intermediate elevations of the Atlantic drainage in Mexico and Central America, from San Luis Potosı´ to Panama, ´ and on the Pacific drainage of Mexico, Guatemala, El Salvador and Costa Rica ŽCampbell and Lamar, 1989.. The actual incidence of bites induced by A. nummifer is not known, partially due to an incomplete epidemiological record and to its resemblance with other pit vipers in the region, therefore making its identification by patients and physicians rather difficult. However, due to its relatively wide distribution, it is likely that this species inflicts a number of accidents in Central America. There have been few studies on the biochemistry and pharmacology of A. nummifer venom. Ž1964, 1967. performed an elecJimenez-Porras ´ trophoretic characterization of the venoms of Costa Rican populations, and other workers have demonstrated: lethal; hemorrhagic; myotoxic; edema-forming; proteolytic; indirect hemolytic; coagulant; and defibrinating activities in A. nummifer venoms from Costa Rica and Honduras ŽBolanos, 1972; Gutierrez and Chaves, 1980; ˜ ´

Gutierrez et al., 1985; Gutierrez et al., 1986c; ´ ´ Rojas et al., 1987; Gene ´ et al., 1989.. In addition, Lys-49, enzymatically inactive phospholipase A 2 homologues which induce myotoxicity, were isolated and characterized from Costa Rican A. nummifer venom ŽGutierrez et al., 1986a; 1989; ´ de Azevedo et al., 1999; Angulo et al., 2000.. However, relatively little is known about intraspecies venom variation in this species. The present work was performed to characterize and compare the venoms of A. nummifer from Guatemala and Honduras, as well as to assess the ability of two antivenoms, currently available in Central America, to neutralize the most relevant toxic activities of these venoms. In addition, a new myotoxic component was isolated and partially characterized from the venom of this species from Honduras.

2. Materials and methods 2.1. Venoms and anti¨ enoms Venoms were obtained from more than 10 adult specimens collected in the Atlantic region of Honduras, and from five adult specimens collected in various locations in Guatemala. In some experiments, a pool of venom obtained from more than 40 adult specimens collected in the Atlantic region of Costa Rica was also used. Venoms were lyophilized and stored at y20⬚C. The antivenoms tested were the liquid polyvalent antivenom from Instituto Clodomiro Picado ŽICP., Costa Rica Žbatch 3161299LQ, expiration date December, 2002. and the MYN lyophilized polyvalent antivenom ŽBioclon S.A. de C.V., Mexico, batch B-8F-01, expiration date June, 2000.. 2.2. Electrophoresis and chromatography Polyacrylamide gel electrophoresis, in the presence of sodium dodecylsulfate ŽSDS-PAGE., was

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

performed on 15% gels ŽLaemmli, 1970.. Gels were stained with Coomassie Brilliant Blue R-250. Analytical ion-exchange chromatography on CMSephadex C-25 columns was also performed. Samples of 10 mg of each venom were dissolved in 1 ml of 0.05 M Tris, 0.1 M KCl, pH 7.0 buffer, and applied to a column previously equilibrated with the same buffer. After elution of unbound material, a linear gradient from 0.1 to 0.75 M KCl in 0.05 M Tris buffer, pH 7.0, was applied to elute the basic components, and the absorbance of the fractions was monitored at 280 nm. After dialysis and lyophilization, fractions were analyzed by SDS-PAGE. In addition, an enzyme immunoassay was used to assess the cross-reactivity between these fractions and Bothrops asper myotoxin II, a Lys49 phospholipase A 2 homologue ŽLomonte and Gutierrez, 1989; Francis et al., 1991.. Frac´ tion proteins Ž0.4 ␮grwell. were adsorbed onto Immulon-2 microplates ŽDynatech. overnight. After a washing step, plates were blocked with PBS containing 2% bovine serum albumin. Then, various dilutions of rabbit antiserum to B. asper myotoxin II were added, followed by a washing step and the addition of anti-immunoglobulin conjugated to alkaline phosphatase ŽSigma, 1:2000.. Color was developed with p-nitrophenylphosphate, and absorbances recorded at 410 nm in a Dynatech MR 5000 microplate reader. Normal rabbit serum was tested in parallel.

153

For amino acid composition analysis, a 600-␮g sample of protein was gas-hydrolyzed under vacuum in a nitrogen atmosphere in 6 M HCl at 110⬚C for 20 h. Amino acids were quantitated using Beckman 6300 equipment. N-terminal amino acid sequence was determined by automatic Edman degradation on a model LF3000 Beckman sequencer, in samples that had been reduced and carboxymethylated. Phospholipase A 2 activity was determined titrimetrically ŽGutierrez et al., 1986b.. N-terminal sequence of ´ a myotoxin previously isolated from the venom of A. nummifer from Costa Rica ŽGutierrez et al., ´ 1986a. was also determined. Cytotoxic activity on mouse C2C12 myoblasts in culture was performed by determining the release of lactic dehydrogenase, as previously described ŽLomonte et al., 1999.. 2.4. Pharmacological characterization 2.4.1. Lethality Groups of six Swis-Webster mice Ž16᎐18 g body wt.. were injected with various amounts of venoms, dissolved in 0.5 ml of 0.12 M NaCl, 0.04 M sodium phosphate, pH 7.2 buffer ŽPBS., by the intraperitoneal route. Controls were injected with PBS. Deaths were recorded during 48 h and the mean lethal dose ŽLD50 . was estimated by the Spearman᎐Karber method ŽWorld Health Organization, 1981..

2.3. Isolation and characterization of a myotoxin Venom of A. nummifer from Honduras Ž500 mg. was fractionated on a CM-Sephadex C-25 preparative column Ž30 = 2.5 cm., as previously described ŽAngulo et al., 2000.. A linear gradient from 0.1 to 0.75 M KCl, in a total volume of 800 ml, was applied, at a flow rate of 0.4 mlrmin. Fractions were tested for myotoxic activity by assessing the increments in plasma creatine kinase ŽCK; EC 2.7.3.2. activity in mice injected intramuscularly Žsee below.. Tubes corresponding to myotoxic fractions were pooled, desalted and rechromatographed on the same column under identical conditions. Homogeneity of the preparation was assessed by: Ža. SDS-PAGE ŽLaemmli, 1970.; Žb. PAGE in the presence of urea, using a cathodic buffer system ŽTraub et al., 1971.; and Žc. reverse phase HPLC ŽWaters. on a C-4 column, eluted at 1 mlrmin with a gradient from 0 to 60% acetonitrile in 0.1% trifluoroacetic acid.

2.4.2. Hemorrhagic acti¨ ity Groups of four mice Ž18᎐20 g. were injected intradermally in the abdominal region with various amounts of venom, dissolved in 0.1 ml PBS ŽKondo et al., 1960; Gutierrez et al., 1985.. Con´ trols were injected with PBS. Mice were killed 2 h after injection and the diameter of the hemorrhagic spots measured. The minimum hemorrhagic dose ŽMHD. corresponds to the amount of venom that induces a hemorrhagic area of 10 mm diameter ŽGutierrez et al., 1985.. ´ 2.4.3. Myotoxic acti¨ ity Groups of four mice Ž18᎐20 g. were injected intramuscularly, in the right gastrocnemius, with various amounts of venom dissolved in 0.1 ml PBS. Controls received 0.1 ml PBS alone. Mice were bled from the tail after 3 h, and blood was collected into heparinized capillary tubes. After centrifugation, plasma CK activity was de-

154

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

termined by the Sigma kit No. 47-UV ŽSigma Chemical Co., MO, USA.. Activity was expressed in unitsrl; one unit defined as the amount of enzyme which produces one ␮mole of NADH per minute under the conditions of the assay. Minimum myotoxic dose ŽMMD. was the amount of venom inducing an increment in plasma CK activity corresponding to four times the activity of samples from mice injected with PBS. 2.4.4. Edema-forming acti¨ ity Groups of four mice Ž18᎐20 g. were injected subcutaneously in the right footpad with various amounts of venom, dissolved in 0.05 ml PBS, whereas the left footpad received 0.05 ml PBS. Edema was assessed 1 h after injection by measuring the increment in the weight of the right footpad, as compared to the left one ŽYamakawa et al., 1976.. Minimum edema-forming dose ŽMED. was the amount of venom inducing 30% edema. 2.4.5. Coagulant acti¨ ity The procedure described by Theakston and Reid Ž1983., and modified by Gene ´ et al. Ž1989., was followed. Briefly, various amounts of venom, dissolved in 0.1 ml PBS, were added to 0.2 ml of human citrated plasma at 37⬚C. Coagulation times were recorded and the minimum coagulant dose ŽMCD. was determined as the venom concentration inducing clotting of plasma in 60 s. Controls included plasma incubated with PBS. 2.4.6. Defibrinating acti¨ ity Groups of four mice Ž18᎐20 g. were injected intravenously with various amounts of venom, dissolved in 0.2 ml PBS, whereas controls received 0.2 ml of PBS alone ŽTheakston and Reid, 1983; Gene ´ et al., 1989.. Mice were bled 1 h after injection from the orbital plexus, and blood was placed in glass tubes and incubated at 22᎐25⬚C for 1 h in order to observe clotting. Minimum defibrinating dose ŽMDD. was the amount of venom that rendered blood unclottable in all mice tested. 2.5. Neutralization by anti¨ enoms Neutralization studies were performed with ICP and MYN polyvalent antivenoms, involving assays in which venom and antivenoms were incubated for 30 min at 37⬚C before testing in the corre-

sponding pharmacological systems ŽGutierrez et ´ al., 1990; Bogarın ´ et al., 1995.. In all cases, a constant amount of venom, corresponding to a ‘challenge dose’ selected on the basis of dose᎐ response studies, was incubated with various dilutions of antivenom ŽGutierrez et al., 1990; Bogarın ´ ´ et al., 1995; Saravia et al., 2001.. The ‘challenge doses’ for the different effects were: Ža. lethal, two or four LD50 values; Žb. hemorrhagic, three MHD; Žc. myotoxic, three MMD; Žd. coagulant, two MCD; Že. defibrinating, two MDD; and Žf. edema-forming, three MED. The number of minimum doses used for each effect as ‘challenge dose’ differed from those used in a previous study with the venom of B. asper ŽSaravia et al., 2001., since doses were selected to be submaximal and to correspond to the linear portion of the dose᎐response curves. Neutralization was expressed as Effective Dose 50% ŽED50 ., defined as the ratio ␮l antivenomrmg venom at which the effect was neutralized by 50% ŽGutierrez et al., ´ . 1990; Bogarın et al., 1995 . In the cases of in vitro ´ coagulant and in vivo defibrinating activities, neutralization was expressed as Effective Dose ŽGene ´ et al., 1989.. 2.6. Statistical analyses The significance of the differences between mean values of two experimental groups was determined by the Student’s t-test. A P value lower than 0.05 was considered significant.

3. Results 3.1. Biochemical characterization and myotoxin isolation SDS-PAGE of venoms from specimens of A. nummifer from Guatemala, Honduras and Costa Rica revealed a similar pattern when run under reducing conditions ŽFig. 1.. Seven predominant bands of M r of 62, 42, 34, 27, 17, 15 and 13 kDa were observed, together with various less conspicuous bands. Venom from Honduran specimens was separated by ion-exchange chromatography on CM-Sephadex C-25 ŽFig. 2.. A large unbound peak was eluted before the start of the KCl gradient. Then, three basic peaks of variable height were eluted after the onset of the gradient. Peak 3 was rechromatographed under identical

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

Fig. 1. SDS-PAGE of venoms of A. nummifer from: Guatemala Ža.; Honduras Žb.; and Costa Rica Žc.. Electrophoresis was run under reducing conditions in 12% acrylamide gels. Lane M corresponds to molecular mass markers Žin kDa.. Gels were stained with Coomassie Blue R-250.

conditions and was shown to be homogeneous by several criteria ŽFig. 3.. This protein, hereby named A.num-Ih, was devoid of phospholipase A 2 activity, and had a molecular mass of 16 kDa, as estimated by SDS-PAGE under reducing con-

155

Fig. 2. Preparative ion-exchange chromatography of A. nummifer venom from Honduras on CM-Sephadex C-25. Venom Ž500 mg. was dissolved in Tris 0.05 M, KCl 0.1 M, pH 7.0 buffer, and applied to the column. After elution of unbound material, a linear gradient from 0.1 to 0.75-M KCl was applied. Fraction three was collected and rechromatographed under identical conditions.

ditions, and of 13 kDa on the basis of amino acid composition. When 50 ␮g of A.num-Ih was injected into the gastrocnemius muscle of mice, an increment of plasma CK activity was observed,

Fig. 3. Homogeneity of the preparation of myotoxin isolated from A. nummifer from Honduras. Ža. SDS-PAGE run under reducing conditions in 12% gels; lane A corresponds to molecular mass markers Žin kDa. and lane B corresponds to the myotoxin. Žb. Reverse phase HPLC run on a C-4 column using a gradient of 0᎐60% acetonitrile in 0.1% trifluoroacetic acid.

156

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162 Table 1 Amino acid composition of the myotoxin of A. nummifer from Honduras ŽA.num-Ih.

Fig. 4. Cytotoxic activity of A. nummmifer myotoxin A.num-Ih on C2C12 myoblasts in culture. Toxin was added to the medium and the release of lactic dehydrogenase to the supernatant was assessed after 3 h of incubation. Cytotoxicity is expressed as percentage release, taking as 100% the lactic dehydrogenase activity in the supernatants of myoblasts incubated with 0.1% Triton X-100. Results are presented as mean " S.D. Ž n s 3..

reaching values of 1920 " 187 unitsrl Ž n s 4., 3 h after injection, whereas control mice injected with PBS had CK activity of 192 " 62 unitsrl Ž n s 4.. Myotoxin induced a dose-dependent cytotoxic effect on C2C12 myoblasts in culture ŽFig. 4..

Amino acid

Moles amino acidrmol protein

Nearest integer

Asx Thr Ser Glx Pro Gly Ala 1r2 Cys Val Met Ile Leu Tyr Phe His Lys Arg

15.41 5.60 5.80 7.71 0.77 8.43 7.76 12.40 2.90 4.30 2.58 7.82 8.32 2.94 1.80 18.00 2.30

15 6 6 8 1 8 8 12 3 4 3 8 8 3 2 18 2

Amino acid composition of A.num-Ih revealed a high amount of Lys, Asx and Cys ŽTable 1.. N-terminal sequence of this myotoxin showed a high level of identity with various Lys49 variants from crotaline snake venoms, whereas less iden-

Fig. 5. N-terminal amino acid sequence of A. nummifer myotoxins from Honduras ŽA. num-Ih. and Costa Rica ŽA. num-I. compared to related proteins isolated from crotaline snake venoms. Residues identical to the A. nummifer myotoxin from Honduras are represented by dots. Sequences correspond to: Acl-K49, Agkistrodon contortrix laticinctus myotoxin ŽSelistre de Araujo et al., 1996.; Cgod-II, Cerrophidion godmani myotoxin II Žde Souza et al., 1998.; B. moo-II, Bothrops moojeni myotoxin II ŽSoares et al., 1998.; Batr, B. atrox PLA 2 ŽMaraganore et al., 1984.; App-K49, A. p. pisci¨ orus K49 PLA 2 ŽMaraganore and Heinrikson, 1986.; A bil PLA-II, A. bilineatus basic PLA 2 ŽNikai et al., 1994.; Basp-IV, B. asper myotoxin IV ŽDıaz ´ et al., 1995.; A. num-II, A. nummifer myotoxin II ŽAngulo et al., 2000.; Basper-II, B. asper myotoxin II ŽFrancis et al., 1991.; Bjsu-I, B. jararacussu bothropstoxin I ŽCintra et al., 1993.; Bsch-I, B. schlegelii myotoxin I ŽAngulo et al., 1997.; Tflv-bpI, Trimeresurus fla¨ o¨ iridis basic protein I ŽYoshizumi et al., 1990.; Vamm-S49, Vipera ammodytes S49 PLA 2 ŽKrizaj et al., 1991.; Bjsu-II, B. jararacussu bothropstoxin II; Bjsu-PLA, B. jararacussu PLA 2 ŽMoura da Silva et al., 1995.; Basp-III, B. asper myotoxin III ŽKaiser et al., 1990.; App-D49, A. p. piscivorus D49 PLA 2 ŽMaraganore and Heinrikson, 1986..

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

tity was observed with Asp49 phospholipases A 2 ŽFig. 5.. For comparison, analytical ion-exchange chromatography on CM-Sephadex was also performed on the venom from Guatemala ŽFig. 6.. Two basic fractions were eluted after the application of the salt gradient. These fractions were myotoxic to mice and lacked phospholipase A 2 activity. Moreover, these basic fractions, and the one purified from the venom from Honduras, cross-reacted strongly with antibodies against Bothrops asper myotoxin II, a Lys49 variant, further evidencing the similarities between these basic components Žresults not shown.. 3.2. Pharmacological characterization The venoms of A. nummifer from Guatemala and Honduras exerted: lethal; hemorrhagic; myotoxic; edema-forming; coagulant; and defibrinating activities, evidencing a qualitatively similar toxicological profile ŽTable 2.. LD50 values did not differ between these venoms, however, there were significant quantitative differences regarding most of the activities tested. Venom from Honduras showed higher hemorrhagic and myotoxic activities, whereas that of Guatemala had higher edema-forming and defibrinating activities.

157

Table 2 Toxic and enzymatic activities of the venom of A. nummifer from Guatemala and Honduras Effect

Guatemala venom

Honduras venom

Lethal ŽLD50 , ␮grmouse.a Hemorrhagic ŽMHD, ␮g.b Myotoxic ŽMMD, ␮g.c Coagulant ŽMCD, ␮g.d Defibrinating ŽMDD, ␮g.e Edema-forming ŽMED, ␮g.f Phospholipase A2 Ž␮Eqrmg min.g

158Ž128᎐195. 57 " 6 27 " 5 19.0" 0.3 20 5.0" 0.4 36 " 5

160Ž143᎐178. U 22 " 1 U 13 " 2 U 12.0" 0.4 U 30 U 12.0" 2.0 30 " 6

Except for lethality, results are presented as mean " S.D. Ž n s 4.. U P - 0.05 when the two venoms were compared. a Lethality was determined in 16᎐18 g mice by the intraperitoneal route; 95% confidence limits are included. b MHD Žminimum hemorrhagic dose.: amount of venom inducing a hemorrhagic area of 10 mm diameter 2 h after intradermal injection in mice. c MMD Žminimum myotoxic dose.: amount of venom that, 3 h after intramuscular injection, induces an increment in plasma CK activity corresponding to four times the CK activity of mice injected with PBS. d MCD Žminimum coagulant dose.: amount of venom which induces coagulation of 0.2 ml of human citrated plasma in 60 s. e MDD Žminimum defibrinating dose.: minimum amount of venom that renders blood unclottable 1 h after intravenous injection in mice. f MED Žminimum edema-forming dose.: amount of venom inducing 30% edema 1 h after subcutaneous injection in mice. g Phospholipase A 2 activity, expressed as ␮Eq fatty acids released per min per mg venom.

3.3. Neutralization by anti¨ enoms

Fig. 6. Analytical ion-exchange chromatography of A. nummifer venom from Guatemala on CM-Sephadex C-25. Venom Ž10 mg. was dissolved in Tris 0.05 M, KCl 0.1 M, pH 7.0 buffer, and applied to the column. After elution of unbound material, a linear gradient from 0.1 to 0.75-M KCl was applied.

Marked differences were observed in the neutralizing ability of antivenoms when tested against venoms of A. nummifer from Guatemala and Honduras. Both antivenoms neutralized: hemorrhagic; myotoxic; edema-forming; coagulant; and defibrinating activities, albeit with different efficacy, since ICP antivenom showed significantly higher neutralizing potency for all these effects ŽTable 3.. On the other hand, ICP antivenom neutralized lethality, with ED50 values varying, depending on the challenge dose of venom used. Lower neutralizing potency was observed when four LD50 values of venom were used than when the challenge dose corresponded to two LD50 values ŽTable 3.. However, despite these quantitative differences, ICP antivenom effectively neutralized lethality of A. nummifer venoms from

158

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

Guatemala and Honduras. In contrast, MYN antivenom failed to neutralize lethality of both venoms in the experimental conditions of this study. Neither antivenom neutralized the edemaforming activity of the venoms from both countries.

4. Discussion Intraspecies variability in venom composition has been described for a number of snakes, a phenomenon which has relevant biological and clinical implications ŽChippaux et al., 1991; Warrell, 1997.. As part of an ongoing effort to characterize the venoms of Central American snakes, this study evaluated intraspecific variability in the venom of A. nummifer, a species distributed from Mexico to Panama, ´ and whose venom has not been characterized in detail. Electrophoretic and chromatographic analyses evidenced strong similarities between venoms of Guatemala, Honduras and Costa Rica. Individual variations in the electrophoretic patterns of venoms of A. nummifer from Costa Rica have been described ŽJimenez´

Porras, 1964.. However, it is likely that such individual variations are masked by the use of venom pools in the present work. The three venoms have the same predominant bands on SDS-PAGE and contain a basic myotoxic fraction of 16 kDa, which lacks phospholipase A 2 activity and cross-reacts with B. asper myotoxin II. Two Lys49 myotoxic phospholipase A 2 homologues have been previously isolated and characterized from A. nummifer from Costa Rica ŽGutierrez et al., 1986a; 1989; de ´ Azevedo et al., 1999; Angulo et al., 2000.. On the basis of amino acid composition and N-terminal sequence, it is suggested that myotoxin A.numnIh, isolated in the present study from the venom of Honduran specimens, is another Lys-49 variant. It shows higher identity with myotoxin I from Costa Rican A. nummifer than to myotoxin II from this venom ŽAngulo et al., 2000.. Previous studies have identified a number of residues that characterize Lys49 phospholipase A 2 homologues, such as Leu5, Gln11 and Asn28 ŽFrancis et al., 1991; Selistre de Araujo et al., 1996., all of which are present in the new m yotoxin variant described. Thus, Lys49 phospholipase A 2 variants are present in the venoms of

Table 3 Neutralization, by polyvalent antivenoms, of effects induced by A. nummifer venom from Guatemala and Honduras Effect

Neutralization ŽED50 , ␮l antivenomrmg venom.a Guatemala venom

Lethalc 2 LD50 values 4 LD50 values Hemorrhagic Myotoxic Coagulant Defibrinating Edema-forming Phospholipase A2 a

Honduras venom

ICP antivenom

MYN Antivenom

ICP antivenom

MYN antivenom

U 245Ž178᎐339. U 833Ž543᎐1282. U 109 " 4 U 594 " 67 U 664 " 83 U 250 N U 707 " 32

Nb N 673 " 106 1979 " 498 929 " 103 500 N 2791 " 180

U 555Ž402᎐769. U 1136Ž781᎐1639. U 81 " 10 U 150 " 20 U 223 " 3 U 500 N NTd

N N 457 " 65 1057 " 307 741 " 26 1000 N NT

Results are presented as ED50 , defined as the antivenomrvenom ratio Ž␮l antivenomrmg venom. at which the effect of the venom is neutralized 50%. In the cases of coagulant and defibrinating activities, neutralization is expressed as Effective Dose. For coagulant activity, this corresponds to the antivenomrvenom ratio at which coagulation time was increased three times when compared with coagulation time of plasma incubated with venom alone. For defibrinating activity, Effective Dose corresponds to the lowest antivenomrvenom ratio at which blood coagulation occurred in the four mice tested. Except for lethality, results are presented as mean " S.D. Ž n s 4.. b N: no neutralization was observed at the highest antivenomrvenom ratio tested Ž1200 ␮lrmg for lethality and 4000 ␮lrmg for edema.. c For neutralization of lethality, experiments were performed using a challenge dose of venom, either 2 LD50 or 4 LD50 values Žsee Section 2.. d NT: not tested. U P- 0.05 when the ED50 values of the antivenoms against the venoms of each country were compared.

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

various populations of A. nummifer in Central America. These components induce muscle cell damage in vitro and in vivo by initially affecting the integrity of the plasma membrane of skeletal muscle cells ŽGutierrez et al., 1989; Bruses ´ ´ et al., 1993; Gutierrez and Lomonte, 1997; Lomonte et ´ al., 1994, 1999.. Our observations evidence a toxicological profile for A. nummifer venoms which is similar to those of the majority of crotaline species, since they induce local effects Žhemorrhage, edema and myonecrosis., as well as systemic alterations Žcoagulopathies and lethality .. Regarding lethality tested by the intraperitoneal route of injection, A. nummifer is one of the least toxic pit viper venoms tested so far in Central and South America ŽBolanos, ˜ 1972; Rojas et al., 1987; Bogarın ´ et al., 2000.. It also shows relatively low hemorrhagic, coagulant and defibrinating activities ŽGutierrez and Chaves, 1980; Gutierrez et al., 1985; ´ ´ . Gene et al., 1989 . In contrast, A. nummifer ´ venoms have higher myotoxic and proteolytic activities than other crotaline snake venoms ŽLomonte and Gutierrez, 1983; Gutierrez et al., ´ ´ 1989.. No qualitative variations were observed when comparing the toxicological profile of venoms of this species from Guatemala and Honduras, being also similar to those previously described for the venom from Costa Rican specimens ŽGutierrez et al., 1985; Gutierrez et al., ´ ´ 1986c; 1989; Bogarın ´ et al., 2000.. This suggests that a similar pathophysiological picture would be expected in envenomations caused by this species in these Central American countries. On the other hand, there were significant quantitative variations between Guatemalan and Honduran venoms in most of the activities tested. The venom from Guatemalan specimens had higher edema-forming, coagulant and defibrinating activities, whereas the venom from Honduras showed higher hemorrhagic and myotoxic effects, but they did not differ in their lethal activity. Parenteral administration of antivenoms constitute the mainstay in treatment of snakebite envenomations worldwide ŽWarrell, 1992.. The two antivenoms tested in this work are the most common therapeutic resources available in Central America and, therefore, it was of interest to assess their ability to neutralize the venom of A. nummifer, especially taking into consideration that this venom is not included in the immunizing mixtures used to manufacture these products.

159

Preclinical evaluation of antivenoms using standard laboratory tests is a relevant task in order to determine the neutralizing and therapeutic profile of these products. Owing to the well-demonstrated intraspecies variation in venom composition, it is important to test antivenoms against venoms from the country where the antivenoms are going to be used, as was the case in the present study. Our observations give evidence that, with the exception of edema-forming activity, ICP polyvalent antivenom was effective in the neutralization of the toxic activities tested, showing a higher neutralizing potency than MYN polyvalent antivenom. MYN antivenom neutralized hemorrhagic, myotoxic, coagulant and defibrinating activities, but failed to neutralize edema-forming and lethal effects. Since the neutralization of lethality is the single most important test in assessing the efficacy of an antivenom ŽWorld Health Organization, 1981; Theakston, 1986; Gutierrez et al., 1996., such lack of neutralization ´ questions the use of this antivenom in the treatment of A. nummifer envenomations in Guatemala and Honduras. The inability of antivenoms to neutralize edema-forming activity deserves consideration. Experimental and clinical observations indicate that edema is a multifactorial pathophysiological event difficult to neutralize by antivenoms in crotaline snake envenomations ŽWarrell, 1992; Otero et al., 1996., even by antivenoms that show efficacy in preincubation-type assays. Edema develops very quickly after a snakebite, thereby making its treatment rather difficult, especially when people take several hours to reach health centers. Our results demonstrate that both antivenoms were ineffective in neutralizing this effect even when incubated with venom prior to injection, suggesting a lack of neutralizing antibodies against edema-forming toxins, in agreement with previous observations performed with ICP antivenom ŽGutierrez et al., 1986c; Rojas et al., 1987.. It ´ would be relevant to investigate if the addition of A. nummifer venom to the mixture of venoms used to immunize horses for antivenom production would improve the ability of antivenoms to neutralize this activity. In conclusion, the venoms of A. nummifer from Guatemala and Honduras have a qualitatively similar biochemical and toxicological profile, which in turn is similar to the one previously

160

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

described for venom from Costa Rican specimens. However, significant quantitative differences were observed between these venoms in most activities tested. A. nummifer venoms contain basic myotoxic Lys49 phospholipase A 2 homologues which probably play a role in muscle tissue damage in these envenomations. Conspicuous differences were observed in the ability of two commercially-available antivenoms to neutralize the various toxic activities induced by these venoms.

Acknowledgements The authors thank the staff of Instituto Clodomiro Picado ŽUniversidad de Costa Rica., Programa de Investigacion ´ en Enfermedades Tropicales ŽEscuela de Medicina Veterinaria, Universidad Nacional, Costa Rica., Escuela de ŽUniBiologıa ´ and Departamento de Bioquımica ´ versidad de San Carlos de Guatemala. and Museo Nacional de Historia Natural ŽGuatemala.. The collaboration of Dr Lourival D. Possani ŽInstituto de Biotecnologıa, U niversidad Nacional ´ Autonoma de Mexico ´ ´ . in the biochemical characterization of the myotoxin is gratefully acknowledged. This investigation was supported by NeTropica, International Foundation for Science Žproject 2677-1., Vicerrectorıa ´ de Investigacion, ´ Universidad de Costa Rica Žprojects 741-89-057 and 741-99-269. and CONCYT of Guatemala Žproject Facyt MS-058-1.. References Angulo, Y., Chaves, E., Alape, A., Rucavado, A., Gutierrez, J.M., Lomonte, B., 1997. Isolation and ´ characterization of a myotoxic phospholipase A 2 from the venom of the arboreal snake Bothriechis Ž Bothrops . schlegelii from Costa Rica. Arch. Biochem. Biophys. 339, 260᎐267. Angulo, Y., Olamendi-Portugal, T., Possani, L.D., Lomonte, B., 2000. Isolation and characterization of myotoxin II from Atropoides Ž Bothrops. nummifer snake venom, a new Lys49 phospholipase A 2 homologue. Int. J. Biochem. Cell Biol. 32, 63᎐71. Bogarın, ´ G., Segura, E., Duran, ´ G., Lomonte, B., Rojas, G., Gutierrez, J.M., 1995. Evaluacion ´ ´ de la capacidad de cuatro antivenenos para neutralizar el veneno de la serpiente Bothrops asper Žterciopelo. de Costa Rica. Toxicon 33, 1242᎐1245.

Bogarın, ´ G., Morais, J.F., Yamaguchi, I.K. et al., 2000. Neutralization of crotaline snake venoms from Central and South America by antivenoms produced in Brazil and Costa Rica. Toxicon 38, 1429᎐1441. Bolanos, R., 1972. Toxicity of Costa Rican snake ˜ venoms for the white mouse. Am. J. Trop. Med. Hyg. 21, 360᎐363. Bruses, ´ J.L., Capaso, J., Katz, E., Pilar, G., 1993. Specific in vitro biological activity of snake venom myotoxins. J. Neurochem. 60, 1030᎐1042. Campbell, J.A., Lamar, W., 1989. The Venomous Reptiles of Latin America. Comstock, Ithaca. Chippaux, J.P., Williams, V., White, J., 1991. Snake venom variability: methods of study, results and interpretation. Toxicon 29, 1279᎐1303. Cintra, A.C.O., Marangoni, S., Oliveira, B., Giglio, J.R., 1993. Bothropstoxin-I: amino acid sequence and function. J. Prot. Chem. 12, 57᎐64. de Azevedo, W.F., Ward, R.J., Gutierrez, J.M., Arni, ´ R.K., 1999. Structure of a Lys49-phospholipase A 2 homologue isolated from the venom of Bothrops nummifer Žjumping viper.. Toxicon 37, 371᎐384. de Souza, M.V., Morhy, L., Arni, R.K., Ward, R.J., Dıaz, J.M., 1998. Amino acid sequence ´ C., Gutierrez, ´ of a myotoxic Lys49-phospholipase A 2 homologue from the venom of Cerrophidion Ž Bothrops. godmani. Biochim. Biophys. Acta 1384, 204᎐208. Dıaz, J.M., ´ C., Lomonte, B., Zamudio, F., Gutierrez, ´ 1995. Purification and characterization of myotoxin IV, a phospholipase A 2 variant, from Bothrops asper snake venom. Nat. Toxins 3, 26᎐31. Fan, H.W., Cardoso, J.L., 1995. Clinical toxicology of snake bites in South America. In: Meier, J., White, J. ŽEds.., Handbook of Clinical Toxicology of Animal Venoms and Poisons. CRC Press, Boca Raton, pp. 667᎐688. Francis, B., Gutierrez, J.M., Lomonte, B., Kaiser, I.I., ´ 1991. Myotoxin II from Bothrops asper Žterciopelo. venom is a lysine-49 phospholipase A 2 . Arch. Biochem. Biophys. 284, 352᎐359. Gene, J.M., Cerdas, ´ J.A., Roy, A., Rojas, G., Gutierrez, ´ L., 1989. Comparative study on the coagulant, defibrinating, fibrinolytic and fibrinogenolytic activities of Costa Rican crotaline snake venoms and their neutralization by a polyvalent antivenom. Toxicon 27, 841᎐848. Gutierrez, J.M., 1995. Clinical toxicology of snakebite ´ in Central America. In: Meier, J., White, J. ŽEds.., Handbook of Clinical Toxicology of Animal Venoms and Poisons. CRC Press, Boca Raton, pp. 645᎐665. Gutierrez, J.M., Chaves, F., 1980. Efectos proteolıtico, ´ ´ hemorragico y mionecrotico de los venenos de serpi´ ´ entes costarricenses de los generos Bothrops, Cro´ talus y Lachesis. Toxicon 18, 315᎐321.

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

Gutierrez, J.M., Gene, ´ ´ J.A., Rojas, G., Cerdas, L., 1985. Neutralization of proteolytic and hemorrhagic activities of Costa Rican snake venoms by a polyvalent antivenom. Toxicon 23, 887᎐893. Gutierrez, J.M., Lomonte, B., Cerdas, L., 1986a. Isola´ tion and partial characterization of a myotoxin from the venom of the snake Bothrops nummifer. Toxicon 24, 885᎐894. Gutierrez, J.M., Lomonte, B., Chaves, F., Moreno, E., ´ Cerdas, L., 1986b. Pharmacological activities of a toxic phospholipase A isolated from the venom of the snake Bothrops asper. Comp. Biochem. Physiol. 84C, 159᎐164. Gutierrez, J.M., Rojas, G., Lomonte, B., Gene, ´ ´ J.A., Cerdas, L., 1986c. Comparative study of the edemaforming activity of Costa Rican snake venoms and its neutralization by a polyvalent antivenom. Comp. Biochem. Physiol. 85C, 171᎐175. Gutierrez, J.M., Chaves, F., Gene, ´ ´ J.A., Lomonte, B., Camacho, Z., Schosinsky, K., 1989. Myonecrosis induced by a basic myotoxin isolated from the venom of the snake Bothrops nummifer Žjumping viper. from Costa Rica. Toxicon 27, 735᎐746. Gutierrez, J.M., Rojas, G., Lomonte, B. et al., 1990. ´ Standardization of assays for testing the neutralizing ability of antivenoms. Toxicon 28, 1127᎐1129. Gutierrez, J.M., Rojas, G., Bogarın, ´ ´ G., Lomonte, B., 1996. Evaluation of the neutralizing ability of antivenoms for the treatment of snake bite envenoming in Central America. In: Bon, C., Goyffon, M. ŽEds.., Envenomings and Their Treatments. Fondation Marcel Merieux, Lyon, pp. 223᎐231. ´ Gutierrez, J.M., Lomonte, B., 1997. Phospholipase A 2 ´ myotoxins from Bothrops snake venoms. In: Kini, R.M. ŽEd.., Venom Phospholipase A 2 Enzymes: Structure, Function and Mechanism. Wiley, United Kingdom, pp. 321᎐352. Jimenez-Porras, J.M., 1964. Intraspecific variations in ´ composition of venom of the jumping viper, Bothrops nummifera. Toxicon 2, 187᎐196. Jimenez-Porras, J.M., 1967. Differentiation between ´ Bothrops nummifer and Bothrops picadoi by means of the biochemical properties of their venoms. In: Russell, F.E., Saunders, P.R. ŽEds.., Animal Toxins. Pergamon Press, Oxford, pp. 307᎐321. Kaiser, I.I., Gutierrez, J.M., Plummer, D., Aird, S.D., ´ Odell, G.V., 1990. The amino acid sequence of a myotoxic phospholipase from the venom of Bothrops asper. Arch. Biochem. Biophys. 278, 319᎐325. Kondo, H., Kondo, S., Ikezawa, H., Murata, R., Ohsaka, A., 1960. Studies on the quantitative method for the determination of hemorrhagic activity of Habu snake venom. Jpn. J. Med. Sci. Biol. 13, 43᎐51. Krizaj, I., Bieber, A.L., Ritonja, A., Gubensek, F., 1991. The primary structure of ammodytin L, a myotoxic

161

phospholipase A 2 homologue from Vipera ammodytes venom. Eur. J. Biochem. 202, 1165᎐1168. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680᎐685. Lomonte, B., Gutierrez, J.M., 1983. La actividad pro´ teolıtica de los venenos de serpientes de Costa Rica ´ sobre la caseına. ´ Rev. Biol. Trop. 31, 37᎐40. Lomonte, B., Gutierrez, J.M., 1989. A new muscle ´ damaging toxin, myotoxin II, from the venom of the snake Bothrops asper Žterciopelo.. Toxicon 27, 725᎐733. Lomonte, B., Tarkowski, A., Hanson, L.A., 1994. Broad cytolytic specificity of myotoxin II, a lysine-49 phospholipase A 2 of Bothrops asper snake venom. Toxicon 32, 1359᎐1369. Lomonte, B., Angulo, Y., Rufini, S. et al., 1999. Comparative study of the cytolytic activity of myotoxic phospholipase A 2 on mouse endothelial ŽtEnd. and skeletal muscle ŽC2C12. cells in vitro. Toxicon 37, 145᎐158. Maraganore, J.M., Merutka, G., Cho, W., Welches, W., Kezdy, F.J., Heinrikson, R.L., 1984. A new class of phospholipases A 2 with lysine in place of aspartate. J. Biol. Chem. 259, 13839᎐13843. Maraganore, J.M., Heinrikson, R.L., 1986. The lysine-49 phospholipase A 2 from the venom of Agkistrodon pisci¨ orus piscivorus. Relation of structure and function to other phospholipases A 2 . J. Biol. Chem. 261, 4797᎐4804. Moura da Silva, A.M., Paine, M.J.I., Diniz, M.R.V., Theakston, R.D.G., Crampton, J.M., 1995. The molecular cloning of a phospholipase A 2 from Bothrops jararacussu snake venom: evolution of venom group II phospholipase A 2 s may imply gene duplications. J. Mol. Evol. 41, 174᎐179. Nikai, T., Komori, Y., Ohara, A., Yagihashi, S., Ohizumi, Y., Sugihara, H., 1994. Characterization and amino terminal sequence of phospholipase A 2-II from the venom of Agkistrodon bilineatus Žcommon cantil.. Int. J. Biochem. 26, 43᎐48. Otero, R., Gutierrez, J.M., Nunez, ´ ´˜ V. et al., 1996. A randomized double-blind clinical trial of two antivenoms in patients bitten by Bothrops atrox in Colombia. Trans. R. Soc. Trop. Med. Hyg. 90, 696᎐700. Rojas, G., Gutierrez, J.M., Gene, M., ´ ´ J.A., Gomez, ´ Cerdas, L., 1987. Neutralizacion ´ de las actividades toxicas y enzimaticas de cuatro venenos de serpi´ ´ entes de Guatemala y Honduras por el antiveneno polivalente producido en Costa Rica. Rev. Biol. Trop. 35, 59᎐67. Saravia, P., Rojas, E., Escalante, T. et al., 2001. The venom of Bothrops asper from Guatemala: toxic activities and neutralization by antivenoms. Toxicon 39, 401᎐405.

162

E. Rojas et al. r Comparati¨ e Biochemistry and Physiology Part C 129 (2001) 151᎐162

Selistre de Araujo, H.S., White, S.P., Ownby, C.L., 1996. cDNA cloning and sequence analysis of a lysine-49 phospholipase A 2 myotoxin from Agkistrodon contortrix laticinctus snake venom. Arch. Biochem. Biophys. 326, 21᎐30. Soares, A.M., Rodrigues, V.M., Homsi-Brandeburgo, M.I. et al., 1998. A rapid procedure for the isolation of the Lys-49 myotoxin II from Bothrops moojeni Žcaissaca . venom: biochemical characterization, crystallization, myotoxic and edematogenic activity. Toxicon 36, 503᎐514. Theakston, R.D.G., 1986. Characterization of venoms and standardization of antivenoms. In: Harris, J.B. ŽEd.., Natural Toxins: Animal, Plant and Microbial. Clarendon Press, Oxford, pp. 287᎐303. Theakston, R.D.G., Reid, H.A., 1983. Development of simple standard assay procedures for the characterization of snake venoms. Bull. Who 61, 949᎐956. Traub, P., Mizushima, S., Lowry, C.V., Nomura, M., 1971. Reconstitution of ribosomes from subribosomal components. Methods Enzymol. 20, 391᎐417. Warrell, D.A., 1992. The global problem of snake bite: its prevention and treatment. In: Gopalakrishnakone, P., Tan, C.K. ŽEds.., Recent Advances in Toxinology Research, 1. National University of Singapore, Singapore, pp. 121᎐153.

Warrell, D.A., 1997. Geographical and intraspecies variation in the clinical manifestations of envenoming by snakes. In: Thorpe, R.S., Wuster, W., Malhorta, A. ŽEds.., Venomous Snakes. Ecology, Evolution and Snakebite. Clarendon Press, Oxford, pp. 189᎐203. Werman, S.D., 1992. Phylogenetic relationships of Central and South American pit vipers of the genus Bothrops Žsensu lato.: cladistic analyses of biochemical and anatomical characters. In: Campbell, J.A., Brodie, E.D. ŽEds.., Biology of the Pitvipers. Selva, Texas, pp. 21᎐40. World Health Organization, 1981. Progress in the characterization of venoms and standardization of antivenoms. WHO, Geneva, 44 pp. Yamakawa, M., Nozaki, M., Hokama, Z., 1976. Fractionation of sakishima habu ŽTrimeresurus elegans. venom and lethal, hemorrhagic and edema-forming activities of the fractions. In: Ohsaka, A., Hayashi, K., Sawai, Y. ŽEds.., Animal, Plant and Microbial Toxins, 1. Plenum Press, New York, pp. 97᎐109. Yoshizumi, K., Liu, S.Y., Miyata, T. et al., 1990. Purification and amino acid sequence of basic protein I, a lysine-49 phospholipase A 2 with low activity, from the venom of Trimeresurus fla¨ o¨ iridis ŽHabu snake.. Toxicon 28, 43᎐54.