Pergamoo

roxr~n, vm . 32. No . ~, ~. a~nz i9sa

0041-0101(94)FA001-N r~ua~c~é~~nEooai-0ioi~ szoo+o.ao

IMMUNOCHEMICAL CHARACTERIZATION OFMICRURUS NIGROCINCTUS NIGROCINCTUS

VENOM WITH MONOCLONAL AND POLYCLONALANTIBODIES

ALBERTO ALAPE-GmBN,~~ BJÖRN GUSTAFSSON,3' 4 BRUNO LOMONTE,~ MONICA THELESTAM4and JOSE MARIA GUTIéRREZ'

'Instituto Clodomiro Picado, Facultad de Microbiologia, and zDepartamento de Bioquimica, Facultad deMedicine, Universidad de Costa Rica, San José, Costa Rica ; 'Department of Vaccine Production, NationalBacteriological Laboratory, Stockholm, Sweden; and 'Department of Bacteriology, Karolinska Institute,

Stockholm, Sweden

(Receiaed 8 September 1993; accepted 21 December 1993)

A. ALAPE-GIRtSN, B . GUSTAFSSON, B. LOMONTE, M. THELESTAM and J . M. GUTIf?R-REZ. lamlunochemical characterization of Micrurus nigrocinctus nigrocinctusvenom with monoclonal and polyclonal antibodies . Toxicon 32, 695-712,1994.-Eleven marine monocloncal antibodies (MAbs) against Micrurus ni-grocinctus nigrocinctus venom were produced and partially characterized . WhenM. n . nigrocinctus venom proteins were separated by SDS-PAGE under non-re-ducing conditions four sharp and three diffuse bands were observed. The sharpbands had migration rates comparable to reduced standards of 10, 12, 50 and72 kDa. The diffuse bands migrate in the range of reduced standards from 14.5to 32 kDa. When venom proteins were separated under reducing conditions thesame sharp bands and an additional prominent 14.5 kDa band were observed .Three antibodies (MAbs 4, 21 and 28) recognized the diffuse bands in westernblots of non-reducing SDS-PAGE, whereas MAbs 7G, 22 and 26 reacted withonly the 72 kDa protein . MAbs 21 and 28 reacted with the 14.5 kDa bandwhereas MAb7G recognized the 72 kDa band in blots ofreducing SDS-PAGE.Two M. nigrocinctus antivenoms cross-reacted by ELISA against nine neuro-toxic snake venoms, as well as with y-toxin from Naja nigricollis and notexin .One antibody (MAb 9A) was used to affinity purify a fraction (called nigroxin)from M. n . nigrocinctus venom . Nigroxin showed phospholipase and myotoxicactivities and appeared as a single 15 kDa band in SDS-PAGE under reducingconditions . However, three bands with slight differences in charge were resolvedby urea-PAGE, representing isoforms named nigroxin a, b, and c. Nigroxininduced a dose-dependent release of peroxidase trapped in negatively chargedliposomes . Nigroxin induced myonecrosis and increased the plasma creatinekinase levels in mice, when injected intramuscularly . The plasma membrane ofcultured L6 myoblasts was permeabilized by nigroxin, as evidenced by therelease of 'H-uridine nucleotides from prelabelled cells . This effect was com-pletely abolished after preincubation with MAb 9A, although this antibodyfailed to neutralize the enzymatic activity of nigroxin. Nigroxin was alsorecognized by MAbs 4, 7H, 21, 27 and 28 . Additionally, the epitope recognizedby MAb 27 is also present in notexin and ß-bungarotoxin .

595

696

A. ALAPE-GIRbN et al .

INTRODUCTIONSNAKES belonging to the genus Micrurus (New World coral snakes) are widely distributedfrom the South-eastern United States to Central Argentina (CAMPBELL arid LAMAR, 1989).All of the 53 Micrurus species are poisonous and several of them have accounted forhuman fatalities (RussELL, 1983; BOLAIVOS, 1984). Micrurus nigrocinctus, with six subspe-cies, is the most abundant and medically important coral snake in Central America(BoLArios, 1984).The signs and symptoms of envenomation by Micrurus are the result of a progressive

blockade at the neuromuscular junction and in severe cases, death is due to respiratoryarrest (RussELt, 1983; BOLANOS, 1984). Experimental studies suggest the presence of aconsiderable spectrum of pharmacological activities in Micrurus venoms. They induceneurophysiological changes similar to those induced by a-neurotoxins and some of themalso show presynaptic effects (WETS and McIsAAC, 1971 ; VrrAt-BRAZIL et al., 1976 ;VITAL-BRAZIL and FONTANA, 1983 ; VITAL-BRAZIL, 1987; GOULARTE et al., 1993). Wheninjected intramuscularly, most of them are myotoxic (GIrr~RREZ et al., 1980, 1983, 1986,1992), and when injected intravenously, they may have cardiotoxin-like activity (RAMSEYet al., 1972; VITAL-BRAZIL et al ., 1976). Ultrastructural changes in muscle fibres are evidenta few minutes after venom injection (GuT~RREZ et al., 1980, 1983, 1986, 1992). Maximumlevels of creatine kinase in serum and myoglobinuria are detected 3 hr postinjection of thevenom (GuTt~RREZ et al., 1986). Almost all Micrurus venoms have a high enzymaticactivity ofphospholipase A2 (PLAZ), but different profiles for other enzymes (KOCHOLATYet al., 1971 ; AmD and DA SILVA, 1991; Tax and PONNUDURAI, 1992). Some ofthese venomshave anti-cholinesterase (KUMAR et al., 1973) and anti-coagulant activities in vitro (TANand POxxunuRAI, 1992).

Currently, five laboratories produce anti-Micrurus horse immunoglobulins for serother-apy in humans (TY-IEAKSTON and WARRELL, 1991). Nevertheless, Micrurus venoms havebeen poorly characterized relative to other elapid venoms because of the difficulty ofmaintaining the snakes and their low venom yields .Venom from Micrurus and other elapids contains cross-reacting components . Immuno-

logically related venom components in Naja naja, Bungarusfasciatus, Enhydrina schistosaand M. fulvius venoms were detected by immunoprecipitation using antivenoms againstthe first three species (MUNJAL and ELLIOT, 1972; MINTON, 1979). From cross-neutraliz-ation studies with 17 elapid antivenoms, only Notechis and Pseudechis antivenoms partiallyneutralized the lethal effect of M. fulvius venom (MnvTON, 1967). A bivalent antivenomproduced against M. frontalis and M. corallinus venoms partially neutralized the lethaleffect of Bungarus fasciatus and Naja naja oxiana venoms (MINTON, 1967). Recently, amonoclonal antibody reactive with Notechis scutatus notexin and Naja nigricollis migexinewas shown to cross-react with M. corallinus venom (MOLLIER et al ., 1990). These resultspoint out the possible presence of toxins in Micrurus venoms, which may have immuno-logical similarities to the well-characterized neurotoxins present in other elapid venoms(Mi NEZ, 1991).Antisera against venom from one species of Micrurus do not always neutralize venoms

from other species (COHEN et al., 1971 ; BoLAr~os et al., 1975 ; BoLAr~os et al ., 1980).Nevertheless, it has been reported that M. nigrocinctus antivenom neutralized M. fulviusand M. dumerilü venoms, indicating the presence of antigenically similar toxins in thesevenoms (BOLANOS et al., 1973).

In this paper we report the production and partial characterization of 11 monoclonalantibodies (MAbs) against M. n. nigrocinctus venom components, as well as their

Characterization of M. n . nigrocinctus Venom

697

cross-reactivities with nine neurotoxic snake venoms and several purified toxins . One ofthe MAbs was used to affinity purify three phospholipase AZ isoforms from the venom.Some of the biological activities of these proteins are also reported. Additionally, twoM. n. nigrocinctus antivenoms were also characterized in terms of their cross-reactivitiesagainst the panel of snake venoms and toxins .

MATERIALS AND METHODS

Snake venoms and toxinsApool ofM. n. nigrocinctus venom was obtained from more than 100 specimens collected in We Pacific region

ofCosta Rica . Pools ofvenoms from Crotalus durissus (newborns), Polamis platurus, and Naja naja kaouthia werealso obtained from the Instituto Clodomiro Picado serpentarium (San José, Costa Rica). All venoms werelyophiliad and stored at -20°C.Venom from Notechis scutatas, and notexin, a- and y-toxin from N. nigricollis were purchased from Latoxan

(Rosana, France). Venoms from Naja nigricollis, Laticauda semifasciata, Dendroaspis jamesonü, Bungarusfasciatus and Pseudechis australis, Bungarus rrtulticinctur a- and ß-bungarotoxin, and bovine pancreaticphoapholipase AZ were purchased from Sigma Chemical Co . (St Louis, MO, U.S.A.) .

ImmioeizationFemale BALB/c mice (6-10 weeks of age) were immunized intraperitoneally (i.p .) with M. n. nigrocinctus

venom dissolved in 10 mM phosphate-buffered saline, pH 7.2 (PBS), as follows : day 0 (5 Ng), days 14 and 28(10 pg), and day 52 (15 hg).

Monoclonal antibodiesHybridomas were established as described previously (ICofu.Ert and Mtrs~rx, 1976). Briefly, splenocytes from

immunved mice were fused with the mouse myeloma cell line SP2/0-Ag l4 using polyethylene glyco14000 (Merck,Darmstadt, Germany) as the fusion agent . Hybridomas were grown in 96-well cell plates (Greiner, Germany)containing HAT medium (Lrrrt~r.n, 1964) supplemented with 10% fetal calf serum (Sigma) . Plates wereincubated at 37°C, in an atmosphere of 80% humidity and 5% C02 . Supernatants were screened for anti-M.n . nigrocinctus venom antibody production by enzyme-liked immunosorbent assay (ELISA), and hybrid cells ofinterest were cloned at least twice by limited dilution, using thymocytea as fender cells (Gvsr~ssox, 1991a, b),MAbs were produced by two methods. (I) Culture medium was collected from hybridomas grown in 260 ml

cell culture flasks (Nunc, Denmark) in RPMI 1640 medium (GIBCO Laboratories, Glasgow, U.K.) supplementedwith 10% fetal calf serum, L-glutamine (1 mM), penicillin (100 U/ml), and streptomycin (100 pg/ml). (2) Ascitesfluid was produced in pristane primed mice injected i.p . with 3-5 x l(I6 cells (Po1-r~a et al., 1972) . Antibody~on-taining culture medium and ascites fluid were stored at -70°C. MAb 9A was purified from ascites fluid by anionexchange chromatography on a DEAE column as described by Gonnva (1983) and purity verified by SDS-PAGE .

Detection and titration ofantibodies by ELISAMicrotitre plates (Flow Laboratories, Irvine, U.K .) were coated overnight at room temperature with 0.125 pg

M. n. nigrocinctus venom/well, dissolved in 100 p10.05 Mcarbonate buffer(pH 9.6) . Remaining binding siteswereblocked with 150 pl of 2% bovine serum albumin in PBS (blocking solution) for 30 min at room temperature.Control wells were saturated only with blocking solution. For cross-reactivity studies wells were coated with0.5 Kg venom or 0.2 hg purified toxin/well as described above. Samples (100 pl/well) were incubated overnightat 4°C. For titrations, supernatants or ascites $aids were serially diluted (twofold) in blocking solution . Eitherculture medium or ascitic fluid from an unrelated hybridoma or normal horse or mouse sera was used as control .Plateswere washed six times with PBS-0.OS% Tween-20 . This was followed by the addition of 100 pl ofa dilution(in blocking solution) of rabbit anti-mouse immunoglobulins (Dako, Copenhagen, Denmark), or rabbitanti-horse IgG (Sigma) conjugated to horseradish peroxidase . Plates were incubated for 1 hr at 37°C beforerinsing as described above. Substrate solution (100 pl/well) consisted of 0.012% HZO~ (Mack) and 2 mg/mlo-phenylenediamine (Sigma) dissolved in 40 mM Tris-HCI, pH 7.6. After 5 min at room temperature, the reactionwas stopped by the addition of 50 pl/well of 1 M HZSO,. Absorbances at 492 nm were measured by a TitertekMultiskan Plus spectrophotometer (Flow Laboratories) .

Immunoglobulin class, subclass and light chainImmunoglobulin class, subclass and light chain were determined by irnmunodiffusion in 1% agarose

(Pharmacia Fine Chemicals, Uppsala, Sweden) dissolved in 10 mM PBS (pH 7.2) using spocific rabbit sensorsto kappa and lambda chains ofmouse IgM, IgA, IgGI, IgG2a, IgG2b and IgG3 (Bionetics Laboratory Products,Charleston, U.S .A) as described earlier (GUSTAFS90N, 1991c) .

698

A. ALAPE-GIRbN et al.

Equine and marine antivenomsAntivenom against M. nigrocinctas venom (Instituto Cladomiro Picado, batch 207A) was produced in horses,

as described by BotrWos and Ciaen~s (1980) or BALB/c mice as described above.

Electrophoresis and densitometric analysisElectrophoretic separations in polyacrylamide gels were done using various systems, including native

conditions (PAGE; 12% w/v; Devts, 1964), denaturing conditions on a discontinuous cathodic electrophoreticsystem for basic proteins (urea-PAGE 12% w/v; Tw+UH et al., 1971), and reducing or non-reducing conditionson SDS-PAGE (IS% w/v; L~t.t, 1970). The gels were run in a double-slab electrophoresis cell (Miniprotean,Bio-Rad Laboratories, Richmond, CA, U.S.A .) for 1 hr at I50 V. Venom was dissolved in distilled water(10 ng/m1) then mixed with an equal volume of the corresponding sample buffer, and 5 pl (25 l+g) of the mixturewas applied to each lane.

Isoelectrofocusing was carried out in precasted minigels (pH range 3-9; Pharmacia) using a Phast systemaccording to the manufacturer's instructions. The minigels were fixed and stained with a silver stain kit(Pharmacia).

Densitometric analysis was performed using a GS300 Densitometer and the program GS365W ElectrophoresisData System (Hcefer Scientific Instruments, San Francisco, CA, U.S .A .) .

ImmunoblottingSamples separated by SDS-PAGE or urea-PAGE were electroblotted (Towetx et al., 1979) onto 0.45pm

nitrocellulose sheets (Bio-Rad Laboratories) using a mini gaol-blot cell (Bio-Rad Laboratories). Blotted proteinswere visualized according to the non-denaturing staining procedure ofSW and Ket~uty (1987) . The nitrocellulosesheets were saturated with blocking solution for 2-4 hr at room temperature before a 10-12 hr incu~tion withthe ascites fluid (1 :25) in blocking solution . Unbound antibodies were removed by washing four times (5-10 mineach) with a solution of 0.1% BSA in PBS with 0.5% Tween-20, before incubation with the horseradish-peroxi-dase conjugates for 4-6 hr at room temperature . The nitrocellulose sheets were washed as described above andincubated with a solution containing 0.5 ng/ml of 4-C1-1-naphthol (Sigma) and O.OlS% H~OZ(Merck) in bufferTris 20 mM, 0.5 M NaCI, pH 7.5 .

Immunoaffinity chromatographyA column was prepared by wupliag 40 mg of MAb 9A to 1 .5 g of CNBr-activated Sepharose 4H (Sigma)

according to the manufacturer's instructions . Five milligrams of M. n. nigrocinctus venom was dissolved in 2.5 mlof PBS and then applied onto the column . The bound fraction was eluted with 0.1 M glycine-HCl (pH 3.0) andcollected into tubes containing 0.5 ml of 0.5 M Tris-HCI buffer (pH 8.8) . A pool of several purifications wasdialysed against distilled water, lyophilized, and stored at -20°C. Venom was also applied onto a Sepharose4B column lacking MAb 9A, as a control for non-specific binding of venom components .

ImmunodrffusivnThe ability of the equine andvenom or MAbs to precipitate components of M. n. nigrocinctus venom, the

affinity purified material, and three purified PLA2 (notexin, ß-bungarotoxin and bovine pancreatic PLAZ) wastested by immunodiffusion in 1 % agarose-PBS plates (OvcH~tt[,onnr and Nttssox, 1978). Samples included 7 ie 1of venom, toxins, or the purified material (5 mg/ml) and equine M. n. nigrocinctus antivenom (8 pl). After 48 hrofinwl~tion at room temperature, gels were extensively washed with saline and stained with Coorrtassie brilliantblue R-250.

Phospholipase Ai activityPhospholipase A2 activity and neutralization by equine antivenom or MAb 9A was tested using an indirect

haemolytic assay in gels (Grtr>é~ez et al., 1988). PBS dilutions ofthe afünity purified fraction were preincubatedwith an equal volume of each ascitic fluid for 30min at 37°C . A volume of 15 pl of each mixture was placedin duplicate 4 mm wells is gels containing 4% sheep erythrocytes, 4% egg yolk as a source of lecithin, and 10 mMof CaCl z . After a 20 hr incubation in a humidity chamber at 37°C the diameter of haemolytis was measured .Gels without lecithin were used as controls.

Effect on multilamellar negatüxly charged liposomesNegatively charged multilamellar liposomes (Sigma), consisting of t.-a-phosphatidylcholine :diacetyl phos-

phate: cholesterol (molar ratio 7 : 2 :1) and entrapped horseradish peroxidase, were prepared as described by Duzet al. (1991). The liposome suspension in microtitre plates (20 pl/well) was mixed with different amounts of theaffinity purified fraction dissolved in 20 p l PBS. After a 30 min incubation at 37°C, 40 p l/well of the substratesolution was added. This consisted of 0.025% HzO~ and 2.5 mM 5-aminosalicylic acid dissolved in 40 mMTris-HCI, pH 7.6. The enzymatic reaction was stopped by the addition of IOltl/well of 12 M Hz SO, followed

Characterization of M. rt . nigrocirtctus Venom

699

by 10pl of 10% Triton X-100. Absorbances at 492 nm were measured by a Titertek Uniskan II spectropho-tometer (Flow Laboratories). Data are expressed as a percentage, taking the absorbance of liposomes incubatedwith 0.2% Triton X-100 as 100% . To correct for spontaneous release, controls where Gposomes were incubatedwith PBS were carried out in parallel and the absorbance was subtracted from sample readings.

Myotoxic activitywebster mice (18-20 g) were injected in the gastrocnemius with 5 pg of the affinity purified fraction dissolved

is 50 pl of PBS. Control mice were injected with 50 pl PBS. Mice were bled from the tail after 3 hr, blood wascollected in heparinized capillary tubes and centrifuged. The creatine kinase activity in plasma was determinedusing a commontial tolorimelrit assay (Sigma).

Mice were euthanized 24 hr after injection and a sample of the injected muscle fixed in Karnovsky's fixative(2.5% glutaraldehyde, 2% paraformaldehyde, 0.1 M phosphate buffer, pH 7.4) for 2 hr. After washing in PBS,pH 7.2, the tissue was dehydrated, embedded in Spurr resin, and sectioned for histological examination ofmyonecrosis .

Assay of cell membrane permeabilizatiort and neutralizationL6 myoblasts (ATCCCRL 1458, kindly provided by Dr T. SerEa.~tv, CMB, Katolinska Institutet) were grown

to confluency in 96-well plates endlabelled with 1 pCi/m15- 3H-uridine for 1 hr at 37°C in DMEM without serumaccording to Trmr.r~r~r (1988) . After three rinses with Hanks balanced salt solution to remove extracelluhuradioactivity, varying concentrations of the immunoaffinity purified fraction were applied dissolved in 250 plTris-buffered saline, pH 7.4 (THS). The ability of antibodies to neutralize membrane permeabilizing activity wastested by preincubationofthe material (dissolved in 150 pl TBS) with 100 pl of ascites fluid or antivenom (100pl)for 30min at 37°C . Equine serum and ascites fluid from an unrelated hybridoma were used as wntrols. Onehundred microlitres of the supernatant was added to 5 ml of scintillation liquid and radioactivity was measuredin a gamma counter. Data are expressed as percentages, using the release from cells incubated with 1 % TritonX-100 as 100% . To correct for spontaneous release, parallel cultures were incubated with only TBS, ascites orantivenom and marker release from those cultures were subtracted from the corresponding experimental values .

RESULTS AND DISCUSSION

Production of anti-M . n . nigrocinctus venom MAbs and isotypesEleven hybridomas producing antibodies directed against M. n. nigrocinctus venom were

established from four different fusions . All established clones produced IgGI with kappachains .

ELISA titresMAbs were produced by cultivating the hybridomas in tissue culture flasks . Super-

natants from those cultures had titres ranging from 320 to 10,240, as determined by ELISA(Table 1). MAbs were also produced in mouse ascites fluid . The titres of such preparationsranged from 10,240 to 1,310,720 . Equine and marine antivenoms showed a titre of 10,240 .

Electrophoresis and densitometric analysisThe electrophoretic mobility ofM. n . nigrocinctus venom components was studied under

various conditions . Iscelectric focusing (IEF) of M. n . nigrocinctus venom, followed bysilver staining of the gel, revealed several bands (Fig. 1) . In the acidic part of the gel, bandswere observed corresponding to iscelectric points between 4.5 and 6.5 . Bands were alsoobserved in the basic part corresponding to isoelectric points between 7.4 and 8.1 and oneprominent band corresponded to an isoelectric point above 9.0 .

In PAGE (Fig . 2A) seven bands were separated, while in urea-PAGE (Fig. 2B) 11 bandswere observed. According to densitometric analysis (data not shown), the three bandsmigrating nearest to the cathode in urea-PAGE represented more than 30% of the venomproteins . A higher number of bands was expected in PAGE, since only the proteins havingan isoelectric point higher than 8.3 would not migrate into these gels . The result may be

700

TABLE I . CHARACTERLSIICS OF MONOCLONAL AN71BODiES llIRECTED AGAINSTMicrurus nigrocinctus nigrocinctus VENOM

'Endpoint titres of cell culture supernatants or ascites fluid in ELISA.Endpoints were defined as the highest dilution in a twofold serial dilution withan absorbance at 492 nm of >0.2 above background .

tPositive: the presence of at least one band aRer developing with HRP-anti-murine immunoglobulins conjugate.$n .t., Not tested.

A B C

FIG . 1 . ISOELECTRIC FOCUSING OF Micrurlts nigrocinctusnigrOCirlCtifr VENOM FOLLOWED BY SILVER STAINING .

(A) Iscelectric point markers are, from the top: 9 .3,8.65, 8 .45, 8 .15, 7 .35, 6 .55, 5 .85, 4 .55, and 3 .5 ; (B) M.n. nigrocinctus venom (1.25 hg); (C) M. n . nigrocinctttsvenom (1 .0 t+g) . Arrowhead indicates a band with pI

higher than 9 .3 .

A . ALAPE-GIR6N et al.

FIG . 2. COOMASSIE BLUE STAINING OF POLYACRYLAMIDEGELS .

Micrurus nigrocinctus nigrocinctus venom was separ-ated in (A) PAGE, (H) urea-PAGE, and (C) SDS-PAGE . Wells were loaded with 25 ug of non-reduced(1) and reduced (2) venom preparations . Arrowheadsin (B) point to the three proteins with most cathodicmigrations, and in (C) signal the proteins bands separ-

ated under reducing conditions .

HybridomaTitre'

Supernatant AscitesSDS-PAGE

Non red .

Reactivity in

Red .

blotst

Urea-PAGE

1 640 n.t.$ - -4 10,240 1,310,720 + - +5 640 81,290 - - -7G 5120 1,310,720 + + +7H 5120 40,960 - - +9A 320 10,240 - - -

21 10,240 1,310,720 + + +22 1280 81,290 + - +26 1280 81,290 + - +27 320 10,240 - -28 5120 n .t . + + n .t.

Characterization of M. n. nigrocinctus Venom

70l

explained either by the presence of multimeric proteins in the venom or alternatively bysome of the proteins having different isoelectric points but identical migration in PAGE.When venom proteins were separated by SDS-PAGE under non-reducing conditions

(Fig . 2C, lane 1), four sharp and three diffuse bands were observed . Two of the sharpbands, with migration rates comparable to reduced standards of 50 and 72 kDa,represented approximately 10% of the venom proteins. The diffuse bands, havingmigration rates comparable to reduced standards of 14.5 to 32 kDa, collectively rep-resented 47% of the venom proteins . Two other sharp bands, having migration ratescomparable to reduced standards of 10 and 12 kDa, represented 27% and 16% of thevenom proteins, respectively . When venom proteins were separated under reducingconditions the same sharp bands and an additional prominent 14.5 kDa band, represent-ing 48% of the venom proteins, were observed . These results suggest that M. n .nigrocinctus venom contains proteins) with mol. wt(s) of 14.5 kDa or less, which eitherform part of multimers or tend to form aggregates .The presence of a higher number of bands in PAGE, urea-PAGE and IEF, as

compared with the number of bands resolved in SDS-PAGE, suggests the presence in thisvenom of isoforms of proteins with the same mol. wt . Coexistence of charge variants ofPLAZ and a-neurotoxins, as well as their respective analogues, has been described inseveral elapid venoms (RosExBEttG, 1990 ; Errno and TnMnrn, 1991) . In the case ofphospholipases, it has been shown that they may have similar mol. wts but a largevariability in charge (DuBotzn~u et al., 1987).Snake phospholipases (either acidic or basic) have at least one subunit of 14-16 kDa

(RosEtas~tG, 1990). An acidic myotoxic phospholipase representing 13% of the venomproteins was isolated from this venom using reverse phase HPLC (AtzxoYO et al., 1987) .The band pattern of M. n. nigrocinctus venom obtained by SDS-PAGE under reducingconditions showed a prominent 14.5 kDa band . According to the densitometric analysisthat band represented 48% of venom proteins . Thus, it is very likely that this venomcontains other phospholipase isoforms which may be separated by charge-dependentelectrophoretic systems.

Cardiotoxins

and

short-chain

a-neurotoxins

(mol.

wt

7 kDa)

or

long-chaina-neurotoxins (mol . wt 8 kDa) are basic proteins having isoelectric points above9 (K.~1ti.ssoN, 1979). Lyophilization of short-chain a-neurotoxins often produces con-siderable amounts of dimers, trimers and higher aggregates (Knxt.ssox, 1979). Theseproteins often display abnormal mol. wt on SDS-PAGE, migrating with an apparentmol. wt of approximately 10 kDa (Msuxi~x, 1972). A band with an isoelectric pointabove 9.0, and a 10 kDa band, were observed by IEF and SDS-PAGE of M. n .nigrocinctus venom. Furthermore, the electrophoretic patterns obtained in SDS-PAGEare compatible with the presence of aggregates in the lyophilized venom. Taken togetherthese results may suggest the presence of short-chain a-neurotoxins inM. n . nigrocinctusvenom.

ImmunoblottlngThe urea-PAGE separated venom components were transferred to nitrocellulose filters

and detected with both polyclonal and monoclonal antibodies . The equine antivenomrecognized all the bands detected in the gel by Coomassie stain, whereas the mouseantivenom did not. Reactivity of marine antivenom was stronger against the threecomponents migrating nearest to the cathode (data not shown) . Six of the 11 MAbs reacted

702

A. ALAPE-GIRtSN et al.

A B

1 2 3

FIO. 3. (A) COOIrtAS4E BLUE STARdING OF UREA-FOLYACRYLAAt1DE GEL.Wells were located with 25 tag of Micrurus nigrocirutus nigrocinctus venom. (B) Immunoblotsshowing the reaction of Mab 22 (1), Mab7G (2), and Mab 26 (3) against M. n. nigrocinctus venom

components separated by urea-PAGE.

in this type of blot (Table 1) . MAbs 7G, 22, and 26 all recognized the same band (Fig . 3),whereas MAbs 21, 4 and 7H recognized the most cathodic bands (data not shown) .The equine antivenom detected all Coomassie stained bands separated in non-reducing

SDS-PAGE (Fig . 4A), whereas the murine antivenom did not react with the two bandshaving the lowest mol. wts (data not shown) . MAbs 4, 21 and 28 recognized the diffusebands, whereas MAbs 7G, 22, and 26 reacted with only the 72 kDa band (Table 1 andFig. 4A). Immunodetection of venom proteins separated by reducing SDS-PAGE, with theantivenoms, showed a pattern similar to that obtained with non-reducing SDS-PAGE: theequine antivenom recognized all bands (Fig . 4B), and the murine did not recognize thelowest mol wt components (data not shown) . MAb 7G recognized the 72 K band, andMAb 21 slightly reacted with the 14.5 kDa (Fig. 4B). MAb 28 reacted as MAb 21 (datanot shown) . Normal horse and mouse sera, as well as a MAb of unrelated specificity,showed no reaction in any of the blots described above. The reactivities of MAbs 4, 21and 28 with 14.5 kDa bands support a possible specificity for phospholipase(s) . Thepotential use of these MAbs in immunoaffinity chromatography for the isolation ofphospholipases from the venom is presently under investigation. Reactivity patterns ofMAbs in blots ofM. n. nigrocinctus venom separated in different polyacrylamide gels aresummarized in Table 1 .

ImmurtodffusionWhen MAbs were tested, either individually or pooled, in Ouchterlony double-diffusion

test against M. n. nigrocinctus venom, no precipitate was obtained. In contrast, the equineantivenom formed five precipitation bands with the venom, but none with notexin,ß-bungarotozin, or pancreatic PLAZ (data now shown) .

Characterization of M. n . nigrocinctus Venom

703

1 2 S 4 5 1 2 9 4 S 8

Fla . 4. Fs>1~uxoal,ols of M. n . ntgrocinctus vENO~t (20 pg PER wEl.l.) sErARATED sY SDS-PAGEUNDER (A) I~DN-REDUCINß AND (B) REDUCING CONDITIONS .

Blots show the reaction of equine antivcnom (1), Mab 21 (2), Mab 7G (3), Mab 22 (4), Mab 26(5). Lane (6) shows the Coomassie blue stained gel .

Antibodies cross-reactivity with neurotoxic snake venomsThe reactivities by ELISA of the antivenoms and MAbs against nine neurotoxic snake

venoms were studied. Both equine and marine antivenoms reacted with all the elapidvenoms tested (Table 2) . Furthermore, the equine antivenom cross-reacted with Crotalusdurissus durissus venom from newborn snakes, which is also neurotoxic (GuTIlhzxEZ et al.,1991). Three of the 11 MAbs (MAb 21, 26 and 2~ recognized heterologous venoms

TABLE 2 . ELISA REACTIVITY OF DIFFERENT VENOAtS WITH EQUINE AND MURINE ArrrIS81tA, AND WTfHMARS 21, 26 AND 27 DIItECIED AGAIIY4[ M. n. nigrceinctus vENOr~

"Normal equine and marine sera gave abaorbances <0.1 with all venoms.tResults are the mean of throe detenminaüons f S.D .$Absorbance <0.1 .

Venom

Abaorbance (492Antisera"

Equine Marine MAb 21

nm)MAbaMAb 26 MAb 27

M. n . nigrocinctus 1 .46 f 0.04t 3 .94 t 0.04 1 .89 f 0.05 2 .39 f 0.04 0.96 t 0 .2P. platwus 1 .49 t 0.14 2 .44 f 0.12 -$ - -L. semijasciata 1 .25 f 0 .01 3.25 f 0.05 2 .91 f 0.02 - -N. scutatus 1 .27 f 0 .01 3 .86 f 0.03 1 .34 f 0.12 - -P. a>rstralis 1 .36 t 0.07 3 .95 t 0.05 - 2.05 t 0.04 -D. jamtsonü 1.24 t 0.07 2.67 f 0.12 - 1 .95 f 0.05 0.18 f 0.02N. nigricollis 1 .46 f 0.04 3 .75 f 0.05 - 0.35 f 0.01 0.18 f 0.02N. naja kaouthia 1 .38 t 0.07 3 .06 t 0.05 - 0.21 f 0.06 -B . rradticinctus 1 .16 t 0.06 3.82 f 0.06 - 0.43 f 0.02 0.22 f 0.04C. d. durissus 1 .44 t 0.04 0 .80 t 0.08 - - -(newborns)

704

A. ALAPE-GIRbN et al.

A

EcNTv

L

a

B

EcNT_~

L

a

Log. dilution

Log. dilution

t

M. n. nigro.Alfa-Tox (N. nigrlcoL)Alfa-Bung.Gamma-Tox. (N. nigrlcd)

M. n. nlgro.Panc. PLANotexlnBeta-Bung.

FIG . 5. TmurtoNS of M. nigrocinctus EQUINE ANTIVENOM AGAIN3r PURIFIED TOXINS.Plates were coated with either M. nigrocinctur venom (0.5 pg/well) or the toxin indicated(0 .2 pg/well), followed by incubation with the antivenom at different dilutions. Anti-equineIgG-peroxidase conjugate was added, followed by substrate, and the absorbante at 492 nm wasrecorded . Each point represents the mean of three determinations . Standard deviations were less

than 10%. Normal equine serum gave readings lower than O.l at all dilutions tested.

(Table 2), indicating the presence of antigens similar to those present in M. n. nigrocinctusvenom. Control readings were less than 0.1 with all venoms .

Cross-reactivity with purled toxinsCross-reactivities of M. n. nigrocinctus antivenoms against several toxins were studied

by ELISA. Figure SA shows the titration curve of the equine antivenom against twoa-neurotoxins and a cardiotoxin . The antivenom obtained from hyperimmunized horsesrecognized both a-neurotoxins, suggesting the presence of components immunologicallyrelated to these toxins in M. n. nigrocinctus venom. A higher reactivity was observed

against a short-chain a-neurotoxin (a-toxin from N. nigricollis) than against a long-chain a-neurotoxin (a-bungarotoxin). However, the marine antivenom did notrecognize these toxins (Fig . 6A). Both antivenoms recognized y-toxin from N. nigricollis(Figs SA and 6A), but we cannot rule out the possible presence of minimal amounts ofphospholipases, not detectable by SDS-PAGE, contaminating this commercial prep-aration . None of the MAbs cross-reacted with either these a-neurotoxins or thiscardiotoxin (not shown) .The equine antivenom also showed reactivity against notexin and ß-bungarotoxin

(Fig . SB), whereas the marine antivenom had a high titre against notexin only (Fig . 6B).These results suggest the presence of components immunologically related to notexin in

La

0.02 .2

2.8

3.4

4.0

Characterization of M. n. nigrocinctw Venom

705

Log. dilution

Log. dilution

t-o--

t

M. n. nigro.Panc . PLANotexlnBeta-Burg.

Fro . 6. ~Yraenoxs of M. nigrocinctus Mvxnve wNnveNO~ ea~ttvsr t~trxrnn:n xotarvs .Plates were coated with either M. nigrocinctw venom (0 .5 pg/well) or the toxin indicated(0 .2 pg/well), followed by incubation with the antivenom at different dilutions. Anti-marineimatunoglobulinrperoxidase conjugate was added, followed by substrate, and the absorbance at492 nm recorded . Each point represents the mean of three determinations . Standard deviationswere less than 10% . Normal marine serum gave readings lower than 0.1 at all diluüons tested .

A t M. n. ngro .Alfa-Tox . (N. nlgrlcoL)

-o"-- Alfa-Bunç.3.0

.. tGamma-Tox. (N. nlgrlcoL)EcNaSI

Ng of the immuraaffinity isolated material

- TBSt Antivenan--~- MAb 9A

FIG. 7 . PHOSFHOLIPASE AC'I7Y1TY OF THE AFFRQITY PURIFIED MATERIAL, A3 MEASURED BY AN WDIAEGTHAEMOLYTIC ASSAY IN AGARO~ GEIS.

Neutralization experiments were performed by preincubating with the antibodies for 30 min at37°C . Results are presented as the mean oftwo detenminations. Standard deviations were less than

5% .

M. n. nigrocinctus venom. Since this venom has presynaptic neurotoxicity (GouLARTEet al., 1993) it would be of interest to determine whether it contains neurotoxicphospholipases .

Immunoa~nity chromatographyThe immunoaffinity purified fraction, applied in wells ofagarose contenting lecithin and

erythrocytes, induced a dose-dependent haemolytis (Fig . 7) . No direct haemolytic activitywas detected in gels without lecithin. The phospholipase activity of the affinity purified

-94 kD

-87 kD

-43 kD

"30 kDrï ,it~ .

~r.~ .

~ ; a1K'.:-20 kD

-ta kD

~a b~C

FIG. H . COOMASSiE BLUE STAINING OF POLYACRYLAMIDE GELS .Six micrograms of the material purified by immunoatfinity with MAb 9A was applied to (A)SDS-PAGE and (B) urea-PAGE gels. Line 2 in (A) shows the migration of mol. wt markers.

707

1 .25 2.5 5 10 20

Nigroxin microg/ml

FIG . 9 . PEROXNASQ nF~ °~g°, FROM NEGATIVELY CHARGED LIPOBOME$ INCUBATED WITH DIIrFERENTCONCENTRATIONS OF NIGAOXiN .

Total release (100%)corresponds to samples where liposomea were incubated with 0.2% TritonX-100 . Results are presented as the mean t S.D ., n =3.

fraction, as determined in the indirect haemolytic assay, was completely neutralized whenpreincubated with the horse antivenom, but not with MAb 9A (Fig. ~. Ouchterlonydouble-ditiusion tests with the affinity purified material formed one sharp precipitationband with equine M. nigrocinctus antivenom (data not shown) .The affinity purified fraction migrated as a single band close to 15 kDa in SDS-PAGE

under reducing conditions (Fig. 8A), whereas under non-reducing conditions it migratedas a diffuse band with a migration rate comparable to reduced standards of 25 to 32 kDa(data not shown) . Three bands were resolved upon urea-PAGE separation (Fig . 8B) whichcorresponded to the venom proteins with the fastest cathodic migration (Fig . 2B). Theseproteins represent major components of the venom, as determined by densitometricanalysis, and are also conserved among individual specimens (ALAPE-GmtSx et al., 1994).No migration was obtained in PAGE, indicating they have iscelectric points higher than8.3 . Only one band with an isoelectric point >8.3 was observed in iscelectrofocusing ofM. n . nigrocinctus venom (Fig . 1), which probably contains all three immunopurifiedproteins .The immunoaffinity purified fraction has phospholipase activity and contains three

proteins which form one precipitation band with the equine antivenom. Based on theirmigration in urea-PAGE, these three proteins probably have similar isoelectric points . Inaddition they have the same mol. wt and antigenic similarity, and therefore they are likelyrelated isoforms . Whether these three proteins are non-covalently associated forming amultimeric protein, or share a common epitope recognized by MAb9A, will be determinedin further studies. We named the affinity purified fraction nigroùn, and its three isoformswill be referred to as nigroùn a, b and c, as labelled in Fig. 8A .

Effect on multilamellar negatively charged liposomesIn the presence of calcium, nigroùn induced a dose-dependent release of peroxidase

trapped in negatively charged liposomes (Fig. 9), indicating the ability of nigroùn todisrupt phospholipid bilayers .

Characterization of M . n. nigrocinctus Venom

80

..be.. 60 _

ô40

8 20

0

708

A. ALAPE-GIR6N et al .

Myotoxicity in vivoThree hours after i.m . injection of 5 hg of nigroxin in mice, a significant (P < 0.05)

increase in plasma levels of creatine kinase was observed (168 f 62 U/ml), compared tocontrol mice receiving PBS only (44 f 20) . Histological observations of muscle samplesobtained 24 hr after injection corroborated myotoxicity (Fig . 10) . At this time, aninflammatory infiltrate was evident in the necrotic tissue . This result showed that M. n .nigrocinctus venom contains other myotoxins in addition to that described by ARROYOet al. (1987) .

Cell membrane permeabilization and neutralizationThe effect of nigroxin on plasma membranes from cultured L6 myoblasts was evaluated

by measuring the release of 3 H-uridine nucleotide from prelabelled cells . Exposure of thecells for 1 hr at 37°C induced a dose-dependent membrane permeabilization (Fig . 11) . Asimilar effect on cell membranes has been observed (GEr~t~, J . A ., personal communication;BuL~rRbN et al., 1993) for two basic myotoxins isolated from Bothrops riper venom :myotoxin II (LOMONTE and GUTIERREZ, 1989) and myotoxin III (KAisER et al., 1990) .One of these toxins (Myotoxin III) is also a phospholipase, whereas the other is a Lys-49variant devoid of enzymatic activity (FRANCIS et al., 1991).The membrane permeabilizing effect of nigroxin was completely inhibited by MAb 9A

(Fig. 11) . The observation that MAb 9A completely neutralized this activity of nigroxinwithout affecting its enzymatic activity may suggest that the structural features responsiblefor these two effects are on separate regions of the toxin . In the case of myotoxin I fromB . riper (GuTIt RREZ et al., 1984), monoclonal antibodies have been used to dissociate thedomain responsible for myotoxicity from the catalytic domain (LOMONTE et al., 1992) . Itwould be of interest to ascertain whether MAb 9A can neutralize the myotoxic effect ofnigroxin in vivo .

Reactivity of MAbs against nigroxin and other phospholipasesELISA reactivity of the MAbs against a preparation containing the three nigroxin

isoforms was studied . Six MAbs (4, 7H, 9A, 21, 27 and 28) recognized these preparationsby ELISA. Three also reacted with them in blots from urea-PAGE gels . In these blots itwas observed that MAb 7H recognized only nigroxin b and c, whereas MAb 21 recognizedall three isoforms . Additionally, MAb 21 recognized another two less catholic bands (datanot shown) . MAb 4 showed a reactivity pattern similar to MAb 21, but gave very faintbands.The reactivity of the six MAbs against three purified PLAZ was determined by ELISA

(Table 3) . MAb 27 recognized notexin and ß-bungarotoxin, whereas the other MAbs didnot recognize these enzymes. According to amino acid sequence homology there are twomain subclasses of elapid PLAZ : enzymes from the Asian/African snakes and PLAZ frommarine/Australian elapids (Dur-roN and Hmr:It, 1983 ; DAVIDSON and DsNNIS, 1990;KOTE4KY, 1991) . MAb 27 recognizes PLAZ of different subclasses, indicating that therecognized epitope is conserved among elapid phospholipases . Other MAbs recognizingconserved epitopes among neurotoxic elapid phospholipases have also been described(MOLLIER et al., 1990; MIDDLEBROOK, 1991) . Since M. n . nigrocinctus venom haspresynaptic neurotoxicity (GOULARTE et al., 1993), and nigroxins share at least one epitopewith notexin and ß-bungarotoxin, it would be interesting to determine whether nigroxinisoforms are neurotoxic .

Characterization of M. n . nigrocinctus Venom

709

t

FIG . 10. PHOTOMICROGRAPHS OF TL44UE SECTIONS FROM MOUSE GASTRONECMIU3 MUSCLE 24 hr errrERi.m . INIEGTION OF S /Ig OF NIGROJON (a) OR PBS ro).

Notice prominent myonecrosis in (a) with abundant inflammatory infiltrate . Bar represents 50 hem.

710 A. ALAPE-GIR6N et al .

10 20 30 40 50 60

Nigroxin pg/ml

FYG. I1 . ~H-URroIxE aFrF.~ FROe~ rREUaELL® L-6 MYOSLASIS INCUBATED WITH DIFFERENT CON-CENTRATIONS OF NIGROXIN.

Both antivenom and MAb 9A neutralized the membrane~isrupting effect of Ngoxin. Total release(100%) was obtained from cultures incubated with 1% Triton X-100. ResWts are presented as the

mean of four determinations. Standard deviations were less than 10% .

TABLE 3 . ELISA REAC'rMTY OF DQ+PERENT FHOBPHOLIPASES WITH MAH8 RECOCiNiZING NiGROXIN9

'Control readings were lower than 0 .1 with all phospholipases .tResWts are the mean of two determinations f S.D.$Absorbance <0.1 .

Acknowledgements-We thank G. LEON and K . MIRANDA for skilfW technical assistance and Dr B . STYLES forcritically reviewing the manuscript . We are also glstefW for the valuable help of C . Duz and E. BuLradN . Thisstudy was supported by the ViceITettoria de Investigation of the Universidad de Costa Rica (Project No741-90-049) and by the Swedish Agency for Research Cooperation with Developing Countries (SAREC). A .ALAPE is a participant of the Karolinska International Research Training Program (KIRT) . B . LostoxrE andJ. M. GuT~RRFZ are recipients of a research career award from the Costa Rican National Scientific andTechnological Research Council (CONICIT) .

REFERENCES

AIRD, S . D . and DA SILVA, N . J . (1991) Comparative enzymatic composition ofbrezilian coral snake (Micrurus)venoms. Comp. Biochem. Physiol. 99B, 287-294 .

ALArE-GIRbrr, A ., LoxorrrE, B ., GvsrAFSSOx, B ., DA SILVA, N. J . and THEt~srAM, M . (1994) Electrophoretit andimmunachemical studies of Micrurvs snake vcnoms . Toxicon 32, 713-723 .

ARROYO, O ., Rasso, J . P., VARGAS, O ., Gtm~tREZ, J . M . and CERDAS, L . (1987) Skeletal muscle necrosis inducedby a phospholipase A3 isolated from the venom of the coral snake Micrunv nigrocinctus nigrocinctus. Comp .Blochem . Physiol. 87B, 949-952.

BoLAAas, R. (1984) Serplentes, Venenosy Ofidismo en Centroamerica. San José, Costa Rita : Editorial Universidadde Costa Rita.

HoLAAos, R . and CERDAS, L . (1980) Production and wntrol of antivenin sera in Costa Rita . Bol. Of. sanit .pan-am. 88, 189-196 .

100

80 ~_l~

60 -ôé

40 -it TBSx ANTIVF_NOMM 20 - -i~ MAb 9A

Phospholipase MAb 4 MAb 7HAbsorbanceMAb 9A

(492 nm)'MAb 21 MAb 27 MAb 28

Nigroxin 3.32 t 0.03t 3 .36 f 0.01 1 .07 t 0.04 3.97 t 0.01 0 .58 f 0.02 3 .93 t 0.02Notexin -$ - - - 1 .21 t 0.04 -ß-Bungarotoxin - - - - 0.93 t 0.03 -Pancreatic PLAz - - - - 0.13 t 0.01 -

Characterization of M. n. nigrocinctus Venom

711

Hor.erlos, R., Ceanes, L. and Tevt.oa, R. (1973) Immunological studies of Micrurua venons from the moatimportant species of North America, Central America, Panama and Colombia. Mtioquia Med. 23, 518.

Bor.ei3as, R., Cranes, L. and Tevr.ox, R. (1975) The production and characteristics ofa coral snake (Micrurusmipartitua hertwigi) antivenom . Toxicon 13, 139-142.

Bur.~txox, E., Tr~reM,M. and Gvr~exez, J. M. (1993) Effects on cultured mammalian cells of myotoxin IIIisolated from B. asps (terciopelo) venom. Biochim . Biophys Acta. 1179, 253-259.

Cezdra©.r., J . A. and Let~un, W.W. (1989) The Venomous ReptAes of Latin America, pp . 90-153. New York :Custom .

Cormv, P., Bear~,W. and SEr.raarecvcv, E. B. (1971) Coral snake venons . In vitro relation of neutralizing andprecipitating antibodies . Am. J. trop. Med. Hyg. 20, 64649.

Devn~sox, F. and DeNrns, E. (1990) Evolutionary relationships and implications for the regulation ofphospholipase A2 from snake venom to human secreted fonms. J. mol. Evol. 31, 228-238.

Devrs, B. J. (1964) Disc electrophoresis II . Method and application to human serum proteins . Ann. N.Y. Aced.Sci. 121, 404-427.

Dtez, C., Gu~xEZ, J. M., LOMONIE, B. and Gervi , J . A. (1991) The effects of myotoxins isolated from Bothropssnake venons on multilamellar liposomea : relationship to phospholipase AZ, anticoagulant and myotoxicactivities. Biochim. Biophys. Acts 1070, 455-460.

D~rox, M. J. and Hroax, R. C. (1983) Classification of phospholipases Az according to sequence . Evolutionaryand pharmawlogical implications . Eur. J. Biochem. 137, 545-551 .

DvBoxnmv, D. J., KAWAGUCHr, H. and Stoma, W. T. (1984) Molecular weight variations in the diversity ofphospholipase Az in reptile venons . Toxicon 25, 333-343.

Exno, T. and Tesmre, N. (1991) Structure-function relationships of postsynaptic neurotoxins from snakevenons. In : Snake Toxins, pp. 165-222 (Iiexvev, A. L., Ed.) . New York : Pergamon Press.

Fraexcrs, B., Gvx~axt:z, J. M., LoMOxtE, B. and Kets£a, I. (1991) Myotoxin II from Bothrops riper (terciopelo)venom is a lysine-49 phospholipase A2 . Archs Biochem. Biophys . 284, 352-359.

Gonuva, J. (1983) Monoclonal Antibodies: Principles and Practice. London: Academic Press.Gout.ea~, F. C. L., Coax, J. C., G

z, J. M. and RODRIGiIF$-Su~norty L. (1993) Effects of Micrwurnigrocinctus snake venom on mouse and chick neuromuscular preparations. Toxicon 31, 135-136.

Gusrerssox, B. (1991a) Fusion protowl for the production of mouse hybridomas . In: Methods in MolecularBiology, Vol. S, Animal Cell Culture, pp. 601-607. New Jersey: Humans Press.

Gusrerssox, B. (1991b) Cloning ofhybridomas . In : Methods in Molecular Biology, Vol. 5, Animal Cell Culture,pp . 609-611 . New Jersey : Hnmana press.

Gusret~ox, B. (1991c) Isotype determination of monoclonal antibodies by immunodiffusion. In : Methods inMolecular Biology, Vol. 5, Animal Cell Culture, pp . 623-625. New Jersey: Hnmanx Press.

Gu~frexPZ, J. M., Ctieves, F., Rotes, E. and Hor.eAas, R. (1980) Local effects induced by Micrurus nigrocinctusvenom in white mice. Toxicon 18, 633-639.

Gur~z, J. M., Losfoxre, B., POATILLA, E., Cesues, L. and Roes, E. (1983) Local effects induced by coralsnake venons : evidence of myonecrosis after experimental inoculations of veaoms from five species. Toxicon21, 777-783.

GUx~EZ, J. M., OweraY, C. andOna.t ., G. (1984) Isolation of a myotoxin from Bothrops riper venom: partialcharacteri7stion and action on skeletal muscle . Toxicon 22, 115-128 .

Gur~rxEZ, J. M., AxaOVO, O., CxeVas, F., Lo~oxre, B. and Cz:rtnes, L. (1986) Pathogenesis of myonecrosisinduced by coral snake (Micrurur nigrocinctus) venom in mice . Br. J. exp. Path . 67, 1-12 .

Gvrtfaenr:z, J. M., Avrr.e, C., Roes, E. and Ceases, L. (1988) An alternative in vitro method for testing thepotency of the polyvalent antivcnom produced in Costa Rica . Toxicon 26, 411-413.

Gv

z, J. M., Dos SeN'ros, M. C., De FemKe Fuareno, M. and Roes, G. (1991) Biochemical andpharmacological similarities between the venons of newborn Crotahis durissus durissus and adult Crotalusdurissus tend rattlesnakes. Toxicon 29, 1273-1277.

Gvr~z, J. M., Roves, G., De SILVe N. J. and Nufiez, J. (1992) Experimental myonecrosis induced by thevenons of South American Micrurvs (coral snakes) . Toxicon 30, 1299-1302.

Ke>sEx, I. L, Gu

z, J. M., Pr.tra~, D., Ann, S. D. and Onta.r., G. V. (1990) The amino acid sequence ofa myotoxic phospholipase from the venom of Bothrops riper. Archs. Biochem. Biophys. 278, 319-325.

Kears~orr, E. (1979) Chemistry ofprotein toxins in snake venons. In : Snake Venoms, Handbook ofExperimentalPharmacology, Vol. 52 (Lee, C. Y., Ed.). Berlin : Springer.

Kaor.ße, G. and Mrrs~rx, C. (1975) Continuous cultures of fused cells secreting antibody of predefinedspecificity . Nature 256, 495-497.

KOCHOLeTY, W. F., Lenr+oren, E. B., DALY, J. G. and Bu.t.tivas, T. A. (1971) Toxicity and some enzymaticproperties and activities in the venons of Crotalidae, Elapidae and Viperidae. Toxicon 9, 131-138.

Koe'rersrcv, P. V., Attx~ove, S. F. and Vt.enrameove, R. R. (1991) Conservative and variable regions ofhomologous snake phospholipases A, sequences : prediction of the taxonspecific peptides structure. J. prot .Chem . 10, 593-601 .

Kusux, V., RF~nv'r, T. and Eu.ror, W. (1973) AntichoGnesterase activity of elapid venons . Toxicon 11, 131-138 .Lemma, U. (1970) Cleavage of structural proteins during the assembly of the head ofbacteriophage T4 . Nature

227, 680-685.

712

A. ALAPE-GIR6N et al.

LITREFIELD, J . (1964) Selection ofhybrids from coatings of fibroblasts in vitro and their presumed recombinants.Science 145, 709-711 .

LostorrtE, B. and Gur~rexez, J . M. (1989) A new muscle damaging toxin, myotoxin II, from the venom of thesnake Bothrops riper (terciopelo). Toxicon 27, 725-733 .

IAMONIE, B., Gur~rexpz, J . M ., RALlfQR7 , M, and Duz, C . (1992) Neutralization of myotoxic phospholipasesA~ from We venom of the snake Bothrops riper by monoclonal antibodies . Toxicon 30, 239-245 .

M~vez, A. (1991) Immunology of snake toxins. In: Snake Toxins, pp . 35-90 (HARVEY, A . L., Ed .) . New York :Pergamon Press.

Mevxn:aa, J ., O~v, R . W., M6~z, A . and Faosieoeor, P . B. P . (1972) Some physical properties of the cholinergicreceptor protein from Electrophorus electricus revealed by a tritiated a-toxin from Naja nigricollis venom .Biochemistry 11, 1200-1210.

MIDDLEBROOR, J . L . (1991) Preparation and characterization of monoclonal antibodies against pseudexin .Toxicon 29, 359-370 .

M~xrotv, S . A. (1967) Paraspecific protection by elapid and sea snake antivenins . Toxicon S, 47-55 .MINTON, S . A. (1979) Common antigens in snake venoms . In : Snake Venorns, Handbook of Experimental

Pharmacology, Vol . 52, pp . 845-862 (LEE, C . Y., Ed .) . Berlin : Springer.MoLLtEa, P., CxwLrrzoFF, S . and M~vEZ, A. (1990) A monoclonal antibody recognizing a wnserved epitope ina group of phospholipases Az . Mol. Immun. 27, 7-15 .

MLTNJAL, D . and ELiroTr, W. B. (1972) Immunological and histochemical identity of esterases and other antigensin elapid venoms . Toxicon 10, 47-54 .

OucxrEßLOrnr, O. and NtLSSOx, L . (1978) Immunodiffusion and immuncelectrophoresis . In : Handbook ofExperimental Immunology. Oxford : Blackwell.

Pormt, M., PusterneEt, J . and Wwt.reas, J. (1972) Growth of primary plasmacytomas in the mineral oil-con-ditioned peritoneal environment. J. natn. Cancer Inst. 49, 305-308 .

Restset, H . W., TAYLOR, W . J., BOAUCHOW, I. B . and NmeR, G . K. (1972) Mechanism of shock produced by anelapid snake (Micrurus f. fulutus) venom in dogs . Am . J. Physiol. 222, 782-786.

RoserrHEaa, P. (1990) Phospholipases. In : Handbook of Toxinology, pp. 67-277 (StttEa, W . and MEes, D ., Eds) .New York : Marcel Dekker .

RusseLL, F. E. (1983) Snake Venom Poisoning. New York: Scholium .$YU,W. and KAEUw, L . (1987) Use of protein stained immunoblots for unequivocal identification of antibody

specificities . J. Immwt . Meth . 103, 247-252.Tex, N.-H . and PoxxtJnuwU, G. (1992) The biological properties of venoms of some American coral snakes

(genus Micrurus). Comp. Biochem . Physiol. lO1B, 471-474 .TFIEAK.41'ON, R . D . G . and WARRELL, D . A. (1991) Antivenoms : a list of hyperimmune sera currently available

for the treatment of envenoming by bites and stings . Toxicon 29, 1419-1470 .merest, M. (1988) Assay of pore forming toxins in cultured cells using radioisotopes. Meth . Enzym . 165,

278-297 .Towenv, H ., STAEFIELIN, T. and GORDON, J . (1979) Electrophoretic transfer of proteins from polyacrylamide gels

to nitrocellulose sheets : procedure and some applications. .Proc. natn . Aced. Sci. U.S.A . 76, 4350-4354 .Tttette, P ., MtzusEQe~, S ., LOWRt, C . and Nosnrw+, M . (1971) Reconstitution of ribosomes from subribosomal

components. Meth . Enzym . 20, 391-417 .VITAL HxeztL, O. (1987) Coral snake venoms: mode of action and pathophysiology ofexperimental envenoma-

tion. Reu. Inst. Med. Trop . S. Paulo 29, 119-126 .Vrrwz. BRAZIL, O. and ForrreNe, M. D. (1983/84) Presynaptic and postsynaptic actions of the venom from the

coral snake Micrurus corallinus in neuromuscular junctions . Mem . Inst. Butantan 47/48, 13-26 .VrrAL BRAZIL, O ., ForrrArIA, M . D . and FILIto, A. P . (1976/77) Physiopathologie et therapeutique de

l'envenomation experimentale causee par le venin de Micrurus frontalis . Mem . Inst. Butantan 40/41, 221-240.WEts, R . and McL4AeC, R. J . (1971) Cardiovascular and muscular effects of venom from coral snake, Micrurusfuluius. Toxicon 9, 219-228 .