Neutralization of proteolytic and hemorrhagic activities of Costa Rican snake venoms by a polyvalent antivenom

Toxicon, ~'ol . J, \o . 6, pp . 88'-893 . I981 . Prineed in Greu 8ricain.

00+1 -0101 . BS ß .00+ .00 . 1985 PerBuewn Press Led .

NEUTRALIZATION OF PROTEOLYTIC AND HEMORRHAGIC ACTIVITIES OF COSTA RICAN SNAKE VENOMS BY A POLYVALENT ANTIVENOM JOSÉ MARIA GUTIÉRREZ, 7051: A. GENÉ, GUSTAVO RODAS and LUIS CERDAS Instituto Clodomiro Picado, Facultad de Microbiologia, Universidad de Costa Rica, San José, Costa Rica (Accepted for publication 12 February 1985) J. ~1 . GUTIERREZ, J. A. Ge1vE, G . RoJns and L. CERDAS . Neutralization of proteolytic and hemorrhagc activities of Costa Rican snake venoms by a polyvalent antivenom. Toxlcon 23, 887-893, 1985 . - The polyvalent antivrnom produced at the Instituto Clodomiro Picado, Costa Rica, was tested for its capacity to neutralize proteolytic and hemorrhaic activities of ten Costa Rican crotaline venoms. In experiments with preincubation of venom and antivrnom, the latter efficiently neutralized proteolytic activities of nine venoms, with eo ranging from SO to 300 NI antivrnom/mg venom. The venom of Bothrops numm(jer was neutralized less efficiently (eo m 760 pVmg). Antivrnom was also very effective in neutralizing hemorrhagc activity, having its lowest neutralizing ability against the vrnom of B. plcadol (en = 430 pl/mg) and its highest towards the venom of B. riper (PaciF~c region) (eo = 47 pl/mg). There was a significant correlation between the ability of antivenom to neutralize proteolytic and hemorrhagc effects . In spite of the ability of antivrnom to neutralize hemorrhage when incubated with venom prior to injection, hemorrhage was only partially neutralized when antivenom was administered i.v. at different time periods after envenomation . This suggests that the rapid development of local hemorrhage, instead of the absence of antivrnom antibodies, is the explanation for the poor neutralization observed in these types of experiments .

INTRODUCTION IN COSTA RICA, the polyvalent

antivenom produced at the Instituto Clodomiro Picado is used to treat envenomations induced by snakes from the genera Bothrops, Crotales and Lachesis (BoLA~(OS, 1982) . This antivenom is prepared in horses immunized with a mixture of venoms of Bothrops ~asper, Lachesls mutes and Crotales durissus (BotAt~os and CERDAS, 1980). BoLAIlos (1971) and BoLAt~os and CERDAS (1980) have demonstrated that it effectively neutralizes the lethal effect of the venoms of 12 Costa Rican species classified in the subfamily Crotalinae. There is a growing body of experimental evidence in support of the concept that neutralization of the lethal effect does not necessarily correlate with neutralization of other toxic activities, since specific pharmacologic properties of venoms depend on different toxins . For instance, it has been demonstrated that although the polyvalent antivenom used in Costa Rica is very effective in neutralizing the lethal effect of venoms (Bot.At3os, 1971), it neutralizes edema-forming and myotoxic activities of some venoms less efficiently (GUTIÉRREZ et al., 1981 ; LOMONTE, 1985) . The World Health Organization (WHO) has recommended the evaluation and standardization of antivenoms based on the neutralization of specific pharmacologic effects (WHO, 1981). Therefore, it is necessary to study the capacity of our antivenom to neutralize several toxic activities of Costa Rican snake venoms, not only of those included in the immunization ss7 -

888

JOSÉ MARfA GUTIÉRREZ et ai.

mixture, but also of the rest of crotaline venoms from this country . This study deals with the neutralization of proteolytic and hemorrhagic activities of the ~~enoms of ten Costa Rican species by this polyvalent antivenom. MATERIALS AND b1ETHODS

Vertoms and antivenom

Venoms were obtained from specimlns collected in Costa Rica and kept at the serpentarium of the Instituto Clodomiro Picado . Once obtained, venoms from a single species were pooled, lyophilized and kept at -70°C until used . The venoms utilized in this study were : Bothrops riper (Atlantic region), B . riper (Pacific region), B. schlegefü, B. picadoi, B. rtummijer, B. godmatti, B. iateralis, B. nasutus, Crotalus durissus durissus and Lachesis muta (Atlantic region) . The antivenom used was the polyvalent, lot No . 132, produced at the Instituto Clodomiro Picado . This antivenom is produced in horses immunized with a mixture of equal parts of venoms of Bothrops riper (equal amounts of venoms from Atlantic and Pacific regions), Lachesis muta and Crotaius durissus durissus (Bot.aAos and CeRnas, 1980).

Neutralization ojproteolytic activity

Initially, the proteolytic activity of the venoms using casein as substrate was determined according to the technique of FRteoetcH and Tu (1971), as modified by LOMONTI: and GurttatRt:z (1983) . Dose-response curves were obtained for each venom, and the amount of venom required to produce a change in absorbance (at 280 nm) of 0.3 was considered a minimum proteolytic dose . Neutralization experiments were performed by incubating one minimum proteolytic dose with several dilutions of the antivenom. In the incubation mixtures, the following antivenom/venom ratios were obtained (in a final volume of L0 ml): 1000, 500, 250, 123, 62 and 31 NI/mg. Two tubes per mixture were used in each determination . Incubation of venom and antivenom was carried out for 30 min at 37°C, after which 2.0 ml of a solution of I o9o casein (dissolved in 0.1 M phosphate buffer, pH 7. l) was added to one of the tubes, whereas the other received 2.0 ml of 0.1 M phosphate buffer, pH 7.1, and was considered a control tube . After 30 min of incubation, at 37°C, the reactions were stopped with 4.0 ml of S°lo trichloroacetic acid . Tubes were kept at room temperature for 30 min and then centrifuged. The absorbance of the supernatants at 280 nm was recorded . The absorbances of the control tubes were subtracted from the absorbances of casein-containing tubes; subtractions were done in order to correct the absorbance due to the phenol present in antivenom, as well as to the hydrolysis of antivenom proteins by venom professes. Results were expressed as percentages, taking as 100% the absorbance of the tubes containing only venom and casein, with no antivenom. Data were plotted and the eo,° (effective dose SOo7o) was calculated from the graph as the antivenom/venom ratio in which proteolytic activity was reduced to SO7o . Experiments were performed three times .

Neutralization of hemorrhagic activity

Two kinds of experiments were performed: (A) Experiments with incubation of venom and antivenom prior to injection ; (B) experiments with inoculation of venom and subsequent administration of antivenom at different time intervals after envenomation . (A) Experiments with incubation of venom axd antivenom prior to injection . Initially, the hemorrhagic activity of the venoms was determined by the technique of KoNt>o et al. (1960) modified as follows: different doses of venom were injected intradermally, in a volume of 0.1 ml, to groups of four mice (18-22 g); 2 hr later they were sacrificed, their skin removed, and the area of the hemorrhagic spot was measured . Diameters were calculated and the minimum hemorrhagic dose was defined as the dose of venom which induced a lesion of 10 mm diameter . In neutralization experiments, venom was incubated with several dilutions of antivenom in such a way that the following antivenom/venom ratios were obtained : 1000, 500, 250, 125, 62 and 31 pl/mg. Nlixtures were incubated for 30 min at 37°C and then O.l ml (containing 10 minimum hemorrhagic doses) was injected intradermally. Two hours later hemorrhagic activity was recorded as described. Four mice (18-22 g) were used in each experimental group. Hemorrhage was expressed as a percentage, taking as 100070 the diameter of the lesions induced by inoculating 10 minimum hemorrhagic doses of venom alone. Results were plotted and the en (effective dose SO%) was defined as the ratio of antivenom/venom that decreased the activity to 50%. (B) Experlmertts in which antivenom was admlrtistered after envenomation . The technique of OwrtaY et ai. (1984) was used . Venom was injected i .m . in the thigh and antivenom was administered i.v . at two time periods after envenomation (immediately and IS min later). Only the venoms of Bothrops riper, B. schiegelü and Lachesis mots were used in this type of experiment . The following doses of these venoms were selected in order to produce a similar extent of hemorrhage : B. riper 100 Ng ; B. schiege(ü 80 Ng ; L, mots SO Ng . Four experimental groups were used with each venom: (1) mice injected i.m . with saline solution; (2) mice injected i .m . with venom; (3) mice receiving an i.m . injection of venom and immediately 0.2 ml of antivenom i.v . ; (4) mice receiving an i .m, injection of venom and 13 min later 0.2 ml of antivenom i .v . The volume of antivenom injected (0.2 mp was selected according to the results of preincubation neutralization experiments, and it is approximately 40 times the volume of antivenom necessary to neutralize hemorrhage if

tieutralization of Snake Venoms by Antivenom

889

venom and antivenom were incubated prior to injection . Four mice were used in each group. The i .v . route was sele~-ted because it is the most effective according to our previous studies (GLTIERREZ et al., 1981). Six hours after venom injections mice were sacrificed and a sample of muscular tissue from the thigh was obtained . It was weighed and immediaCely homogeniud with 2 .0 mi of distilled water. The homogenate was centrifuged at 10.000 x g for IO min and l.8 ml of the supernatant were added to 3 ml of 2X Drabkin solution . The absorbante was recorded at 340 nm and the hemoglobin content determined using a standard curve. The amount of hemoglobin in muscle was corrected on the basis of sample weight and the results were expressed as percentages, taking as IOO~Io the values obtained in mice injected with venom alone and as Oela the values of muscle injected with saline solution . RESULTS

Neutràlization ofproteolytic activity

The minimum proteolytic dose of each venom, as well as the neutralizing capacity of the antivenom (expressed as EDso ), are given in Table 1 . Antivenom neutralized proteolytic activity of all venoms tested, although not equally . Among the 10 venoms tested, B. nummijer required the highest dose of antivenom to be neutralized (EDT = 760 ~tl/mg), whereas the others had ED sa ranging from 50 to 300 ~tl/mg (Table 1). If neutralization is expressed as ~tl antivenom/minimum proteolytic dose of venom, the venom of B. nummifer still required the highest volume of antivenom to be neutralized (EDsa = 380 ul/minimum proteolytic dose) (Table 1).

Neutralization of hemorrhagic activity

In experiments with incubation of venom and antivenom prior to injection, antivenom was highly effective in the neutralization of hemorrhagic effect . The ED,o for B. picadoi venom was the highest (430 ~1/mg) . For the other venoms ED so ranged from 47 to 350 ~l/mg (fable 2) . However, if neutralization is expressed as ~l antivenom/10 minimum hemorrhagic doses of venom, the EDso for B. nasutus is the highest (17 .5 ~tl/10 minimum hemorrhagic doses) (Table 2). Figure 1 shows that there was a significant correlation between neutralization of proteolytic and hemorrhagic effects (r= 0.71; P<0 .05) if neutralization is expressed as ~l antivenom/minimum proteolytic or hemorrhagic dose . However, when neutralizing capacity is expressed as ~l antivenom/mg venom, there is no significant correlation (r= 0.51 ; P>0.1) . TABLE l . PROTEOLYTIC ACTIVITY OF VENOMS AND NEIfrRALIZATION OF PROTEOLYSIS BY ANTlVEN0~1 Venom

Bothrops asper

(Atlantic) B . asper (Pacific)

B . schlegelü B . numm(/er B . plcadoi B . laterolis B . naswtus B . godmanl Lachesis muta Crotalus durissus durissus

Proteolytic activity MPD (mg)'

NCUtraIi2at10R (ED,a)t Nl antivenom/mg venom

NI antivenom/MPD

1 .2

90

I08

l .4

120 SO 760 300 280 160 230 190 I50

168 87 380 330 l68 336 138 171 135

l .7 0.5

1 .1

0.6

2 .1

0.6 0.9 0.9

'MPD (minimum proteolytic dose): amount of venom required to produce a change in absorbante at 280 nm of 0.5 (see Materials and Methods) . tED,a (effective dose S0~) andvrnom/vrnom ratio at which the effect is reduced to SONS .

890

JOSÉ MARIA GUTIÉRREZ et al . TABLE

2.

HEMORRH.4GIC ACTIVITY OF vENOhI .4ND NEUTRALIZATIO!~ OF HE~IORRH .4GE BY AvTIVEi`O~1

Venom Bothrops asper (Atlantic) B. asper (Pacific) B. schlegelü B. nummifer B. picadoi B. /ateralis B . rtasutus B. godmani Lachesis muta Crotalus durissus durissus

!neutralization (ED ,)`

Hemorrhagic activity ~IHD (kg) "

NI antivenom/mg venom

1 .S

13S

2 .0

2 .S l .8 4 .0 2 .0

47 60 300 430 270

l .2 I .? I2 .0 2 .l

1 .0 1 .0

l20 8S

0 .5 S .1

Nl

5 .4

17 .5 l .2 0.8 4 .5

3S0

3 .0

antivenomll0 MHD

1 SO

"MHD (minimum hemorrhagic dose) : dose of venom that induces a hemorrhagic spot of 10 mm diameter (see Materials and Methods) . t ED~ (effective dose SO%) : antivenom/venom ratio at which the effect is reduced to SO% .

On the other hand, when antivenom was administered i.v . after venom injection, neutralization of hemorrhagic activity in local muscular tissue was poor (Table 3). Results demonstrated that there was a conspicuous reduction of hemorrhage only when antivenom was administered immediately after envenomation, although even in these circumstances it was not totally neutralized . When antivenom injection was performed 15 min after envenomation, neutralization was very poor. Qualitatively similar results were

W aVV gd

HEMORRHAGE

tYl/10 MHD)

FIG .

1 . CORRELATION BETWEEN THE ABILITY OF ANTIVENOM TO NEUTRALIZE PROTEOLYTIC AND HEMORRHAGIC EFFECrs (r=0.712 ; FGO .OS) . Neutralization was expressed as ED," (Id antivenom/minimum proteolytic or hemorrhagic dose). (a) B . schlegelli; (b) B . riper (Atlantic) ; (c) B. godmani; (d) C. durEssus; (e) B . ~ !ateralis; (f) L . muta; (g) B. riper (Pacific) ; (h) B. picadol; (i) B. nummjjer, B. nasutus .

U)

Neutraliration of Snake Vrnoms by Antivenom

89 1

TABLE 3. NEUTRALIZATION OF HE.~IORRHAGIC EFFECT IN EXPERIINFMS [N WHICH A.~ITIVENOM W'AS AD~IIYISTERED AFTER ENYENO~IAT[ON

Yenom Bothrops asper

(Atlantic)

B. schlegetü Lachesis muta

Delay in antivrnom administration (min)' No antivenom 0 15 No antivenom 0 15 No antivenom 0 l5

Hemorrhage (W~)` L00 ~ 14 ~ 71 t l00 t 23 t 102 * I00 * 79 * 104 *

7 4* l4s l5 8~ 7 4 l0~ 10

rI'ime interval between venom injection and antivrnom administration ; 0.2 ml of antivenom were inoculated i.v . tExpressed as a percentage, taking as I00~ the extent of hemorrhage induced by venom alone (n=4). Values correspond to mean t S.E .M . }Significant reduction of hemorrhage (PG0 .05) whrn compared to effect induced by venom alone.

obtained with the three venoms, but quantitatively there was a better neutralization of the venoms of B. riper and B. schlegelü (Table 3). DISCUSSION

Bothrops riper is the species responsible for most snakebite cases in Costa Rica 1982). However, many accidents are caused by other crotaline snakes, such as B. schlegelü, B. lateralis and B. nasutus (BoLAI~os, 1982). Accidents due to these species are usually less severe than those due to B. riper, probably because they inject a relatively small amount of venom (BOLAI~OS, 1971). However, despite the fact that systemic poisoning usually does not take place, local tissue damage and hemorrhage occur. Furthermore, although accidents due to Lachesis muta are not abundant, they are extremely severe (BoLAI'vos et al., 1982) . Thus, it is important to determine the capacity of the polyvalent antivenom to neutralize toxic effects induced by crotaline venoms . Hemorrhage was studied because it has been related to the pathogenesis of envenomation by crotaline snakes . Hemorrhage is responsible for local tissue damage (OWNBY, 1982; GuTIERREZ et al., 1982), as well as for systemic alterations such as cardiovascular shock (CARLSON et al., 1975) . Our results show that the antivenôm used in Costa Rica is highly effective in the neutralization of proteolytic and hemorrhagic effects of the 10 crotaline venoms tested, although there are quantitative differences in the EDso values . Interestingly, the antivenom more effectively neutralized the venom of B. schlegelü, which is not used in the immunization mixture, than that of B. riper. The finding that a polyvalent antivenom is highly effective in neutralizing hemorrhage is not new, since RUSSELL et al. (1973) and OwlvsY et al. (1984) described that Wyeth polyvalent antivenom has a high antihemorrhagic titer . Most proteolytic and hemorrhagic components so far isolated from snake venoms are proteins with relatively high molecular weights (usually more than 20,000) ('Tt1, 1977, 1982), thus they are probably good immunogens, able to induce a significant immune response in horses . This may explain the high neutralizing capacity of our antivenom towards proteolytic and hemorrhagic activities . (BoLAi~oS,

89'_

JOSÉ ~[AR1A GUTIÉRREZ et a(.

There was a significant correlation between neutralization of proteolysic and hemorrhagic activities . Despite the fact that most hemorrhagic toxins isolated are metalloproteinases (Tu, 1982), it is well known that there are very active proteolysic enzymes devoid of hemorrhagic activity in snake venoms . For instance, ARAG6N et al . (1978) isolated two proteases from B. asper venom which do not induce hemorrhage. 1~Ioreover, GUTIÉRREZ and CRAVES (1980) described a lack of correlation between hemorrhagic and caseinolytic activities in 10 Costa Rican snake venoms . In the present work, we confirmed this lack of correlation between hemorrhage and proteolysis induced by these venoms (r= 0.35 ; P>0.1) . Thus, the correlation observed in this study is not due to the fact that these two effects are induced by the same toxins . Instead, this correlation may reflect a parallel phylogenetic variation in antigenic propenies of proteolysic and hemorrhagic components in the venoms used in this study. Although the antivenom is very effective in neutralizing hemorrhage in experiments in which it is incubated with venom prior to injection, it is only partially effective in experiments with independent inoculation . Similar results were obtained previously for edema and myonecrosis induced by B. asper venom (GUTIÉRREZ et al., 1981) . The reason behind this apparent discrepancy lies in the fact that local hemorrhage develops very fast once venom has been injected ; for instance, in the case of B. asper venom there is a drastic hemorrhagic effect as early as 30 min after envenomation (GUTIÉRREZ et al., 1984) . Thus, hemorrhage probably starts minutes after injection, as has been demonstrated by the ultrastructural studies of OWNBY et al. (1974) with the venom of Crotalus atrox. Therefore, even when antivenom is administered i .v. immediately after venom injection, local hemorrhage develops before enough antibodies reach the affected tissue. These findings have clinical implications, since they emphasize the need for an early administration of antivenom. It has been proposed that a way to improve antivenoms is by the enrichment of the antigenic mixture with purified toxins . This is particularly valid for low molecular weight toxins, such as elapid neurotoxins (see TAM, 1983) and myotoxin a from Crotalus viridis viridis venom (OWNHY et al., 1983) . However, in our case, results indicate that there is a good immune response to proteolysic and hemorrhagic components in horses immunized with a mixture of venoms of B. asper, C. durissus and L. muta. In regard to local hemorrhage, our findings suggest that the problem in neutralization is not due to the lack of antibodies in the antivenom, but instead to the rapid development of damage to microvasculature . These results stress the need for therapeutic devices which work once local hemorrhage has developed in snakebite cases . Acknowledgments - The authors thank JAVIER NI1f7EZ, ALFRED VARGAS and ABEL ROBLES for their collaboration in the laboratory work . Also, we are grateful to Dr HRUNO LOMONTE for reviewing the manuscript and to Mrs HILDA HERRERA for typing it . This work was supported by the Vicerrector(a de Investigacibn, Universidad de Costa Rica, project 741-84-82 . REFERENCES J . and GUBENSEK, F . (1978) Some properties of two fibrinolytic enzymes from the venom of Bothrops asper. Period. B(ot, gll, 91 . BoLAAos, R . (1971) Nuevos Recursos Contra et Ofuüsmo eR Centroamérica. 29 pp . San José: Ministerio de Salubridad Pûblica y Universidad de Costa Rica . BOLA30S, R . (1982) Las serpientes venenosas de Centroamérica y et problems del ofidismo. Primera parte . Aspectos zooldgicos, epidemiolbgicos y biomédicos . Revta costarr. Cienc. méd. 3, 165 . BoLAAOS, R . and CERDAS, L . (1980) Produccidn y rnntrol de sueros antiof(dicos en Costa Rica . BoIR Oj. saR(t. part-am. 88, 189 . ARACbrt, F., KOPITAR, M ., BABNIK,

Neutralization of Snake Venoms by Antitenom

893

BOLA :~OS, R ., Rotas, O. slid LJtLOA, C. E . (1982) Aspectos biomEdicos de cuatro casos de mordedura de

serpiente por Lachesis muta (Ophidia : Viperidae) en Costa Rica . Revta Biol. trop . 30, 33 . CARtSON, R. W., $CHAEFFER, R. C., Vi~HIGHA~t, H., MICHAELIS, S ., RI;SSELL, F. E. and ~L'Ett, ~I . H. (1971) Rattlesnake venom shock in the rat: development of a method . Am . J. Physiol. 239, 1668 . FRIEDRICH, C . and Tt :, A. T. (1971) Role of metals in snake venoms for hemorrhagic, esterase, and proteolyzic activities . Biochem. Pharmac. 20, 1549 . Gl,T1ERREZ, J. M. and CHAVES, F. (1980) Efectos proteolitico, hemorrl{gico y mionecrdtico de los venenos de serpientes costarricenses de los géneros Bothrops, Crotales y Lachesis. Toxicon 18, 313. GUTIERREZ, J. M ., CHAVES, F., BOLAi~OS, R., CERDAS, L., RODAS, E ., ARROYO, O. and PORTILL.a, E. (l9gl) Neutralization de los efectos locales del veneno de Bothrops riper por un antiveneno polivalente. Toxicon 19, 493 . GI,TIERREZ, J. Ivl., CERDAS, L., ARROYO, O., RODAS, E., LONtO~TE, B. and GENE, J . A. (198) Pa[Ogt:ne515 y neutralization de los efectos locales inducidos por veneno de la serpiente "terciopelo" (Bothrops riper) . Acts mtd. ctutarr. 25, 233. GIJTIERRFZ, J . M., OWHBY, C. L. and ODEtL, G. V. (1984) Pathogenesis of myonecrosis induced by crude venom and myotoxin of Bothrops riper. Exp. moles. Path . 40, 367 . KON[)D, H ., KoNDO, 5., IKEZAWA, 1., N1uRArA, R. and OHSAKA, A. (1960) Studies of the quantitative method for the determination of hemorrhagic activity of Habu snake venom. Jpn. J. med. Scl. Biol. 13, 43 . LOMONTE, B. (1985) Edema-forming activity of bushmaster (Lachesis muta stenophrys) and Central American rattlesnake (Crotales durissus durissus) venoms and its neutralization by a polyvalent antivenom . Toxicon 23, 173. LontorvTE, B. and GIJTIERREZ, J. M . (1983) La actividad proteolitica de los venenos de serpientes de Costa Rica sobre la caseina. Revta Biol. trop. 31, 37 . OWNBY, C. L. (1982) Pathology of rattlesnake envenomation . In : Rattlesnake Venoms, Their Actions and Treatment, p. 163 (Tu, A. T., Ed .) . New York : Marcel Dekker . OWNBY, C. L., KAIHER, R. A. and Tu, A. T. (1974) Pathogenesis of hemorrhage induced by rattlesnake venom. Am. J. Parti. 76, 401. OwxaY, C. L., ODELL, G. V., Wooos, W. M. and CotaeaG, T. R. (1983) Ability of antiserum to myotoxin a from prairie rattlesnake (Crotales viridis viridis) venom to neutralize local myotoxicity and lethal effects of myotoxin a and homologous crude venom. Toxicon 21, 33 . OwNaY, C. L., COLBERG, T. R. and ODELL, . G. V. (1984) A new method for quantitativg hemorrhage induced by rattlesnake venoms: ability of polyvalent antivenom to neutraliu hemorrhagic activity . Toxirnn 22, 227. RUSSELL, F. E. RuziC, N. and GONZÂLEZ, H . (1973) Effectiveness of antivenin (Crotalidae) polyvalent following injection of Crotales venom. Toxicon 11, 461 . TAN, N. H. (1983) Improvement of Malayan cobra (Ngja ngja spurotrix) antivenin. Toxicon 21, 73 . Tu, A. T. (1977) Venoms : Chemistry and Molecular Biology. New York : John Wiley. Tu, A. T. (1982) Chemistry of rattlesnake venoms . In : Rattlesnake Venoms, Their Agitions and Treatment, p. 247 (Tu, A. T., Ed .) . New York :Marcel Dekker . WORLD HEALTH ORGANIZATION (1981) Progress in the characterization of venoms and standardization of antivenoms . WHO Offset Publication No . 58 .