Polybitoxins: a group of phospholipases A2 from the venom of the neotropical social wasp paulistinha (Polybia paulista)

PII: S0041-0101(97)00053-6

Toxicon Vol. 36, No. 1, pp. 189±199, 1998 # 1998 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0005-7967/98 $19.00 + 0.00

POLYBITOXINS: A GROUP OF PHOSPHOLIPASES A2 FROM THE VENOM OF THE NEOTROPICAL SOCIAL WASP PAULISTINHA (POLYBIA PAULISTA) MARCIA REGINA DE OLIVEIRA and MARIO SERGIO PALMA* Department of Biology, Institute of Biosciences of Rio Claro, Center for The Study of Venomous Animals (CEVAP), University of SaÄo Paulo State (UNESP), Rio Claro, SP, CEP 13506-900, Brazil

(Received 6 January 1997; accepted 24 March 1997)

M. R. de Oliveira and M. S. Palma. Polybitoxins: a group of phospholipases A2 from the venom of the neotropical social wasp paulistinha (Polybia paulista). Toxicon 36, 189±199, 1998.ÐThe neotropical wasp Polybia paulista is very aggressive and endemic in south-east Brazil, where it frequently causes stinging accidents. By using gel ®ltration on Sephadex G-200, followed by ion-exchange chromatography on DEAE-Cellulose under a pH gradient, a group of four toxins (designated as polybitoxins-I, II, III and IV) presenting phospholipase A2 (PLA2) activities was puri®ed. These toxins are dimeric with mol. wts ranging from 115,000 to 132,000 and formed by di€erent subunits. The four toxins contain very high sugar contents attached to their molecules (22±43% w/w) and presented di€erent values of pH optimum from 7.8 to 9.0; when dissociated, only residual catalytic activities were maintained. The catalytic activities of polybitoxins (from 18 to 771 mmoles/mg per minute) are lower than that of PLA2 from Apis mellifera venom and hornetin from Vespa basalis. The polybitoxins presented a non-linear steady-state kinetic behavior for the hydrolysis of phosphatidylcholine at pH 7.9, compatible with the negative co-operativity phenomena. All of the polybitoxins were very potent direct hemolysins, especially the polybitoxins-III and IV, which are as potent as the lethal toxin from V. basalis and hornetin from Vespa ¯avitarsus, respectively; polybitoxin-IV presented hemolytic action 20 times higher than that of PLA2 from A. mellifera, 17 times higher than that of neutral PLA2 from Naja nigricolis and about 37 times higher than that of cardiotoxin from Naja naja atra venom. # 1998 Elsevier Science Ltd. All rights reserved

* Author to whom correspondence should be addressed. 189

190

M. R. de OLIVEIRA and M. S. PALMA INTRODUCTION

Hymenoptera venoms are complex mixtures of biochemically and pharmacologically active components such as biogenic amines, peptides and proteins (Nakajima, 1986). The composition of vespid venoms has been subjected to few investigations, since the production of venoms by the social wasps is very small and there is a limited availability of vespid venoms as raw materials. It has been shown that the Vespinae venoms contain many di€erent components such as phospholipases A and B, hyaluronidases, acid phosphatases, proteases and nucleotidases (Nakajima, 1986). The compositions of the various Vespinae venoms are similar to each other (Habermann, 1972); however, almost nothing is known about the neotropical Polistinae venoms. It has been demonstrated that vespid venoms frequently cause allergic reactions in humans (Ho€man, 1985; Reisman and Osur, 1987; Castro et al., 1994) and the phospholipases A2 (PLA2) are recognized as one of the majors allergens from these venoms (Ho€man, 1978, 1985). PLA2 (E.C. 3.1.1.4) catalyses the speci®c hydrolysis of ester bonds at the C2 position of 1,2-diacyl-3-sn-glycerophospholipids into their corresponding lyso compounds with the release of free fatty acids. Thus, PLA2 is able to disrupt the phospholipid packings from several types of biological membranes, leading to pore formation and/or cell lysis (Dotimas and Hider, 1987). The present paper reports the puri®cation and biochemical characterization of the polybitoxins (PbTX), a group of toxins from the venom of a neotropical social wasp Polybia paulista, presenting PLA2 acitivity and very potent hemolytic actions in washed red cells.

MATERIALS AND METHODS Biological material and venom extraction Workers of P. paulista were captured in the University Campus, at Rio Claro, SP, south-east of Brazil. The freshly collected wasps were immediately frozen and dissected. The venom reservoirs were removed from the sting apparatus by pulling with forceps and cutting with microscissors, under a stereomicroscope, minimizing contamination from extraneous tissues. The reservoirs were then carefully washed in a small volume of isotonic solution, thawed and punctured, followed by several washings with distilled water to extract the venom and centrifugation at 12,000 g, for 15 min at 48C. The supernatant was freeze-dried and kept at ÿ808C until used. Protein assay Protein was determined by the method of Lowry (Hartree, 1972), using bovine serum albumin (BSA) as standard. Determination of phospholipase speci®city In order to determine the type of phospholipase activity the crude venom of P. paulista and the puri®ed toxins were incubated at 378C in the presence of natural and synthetic phospholipids as substrates: egg phosphatydylcholine, egg lysophosphatydylcholine, 1-stearoyl-2-oleoyl-3-sn-glycerophosphoryl choline and 1-oleoyl-2stearoyl-3-sn-glycerophosphorylcholine (Sigma Chemical Co., St Louis, MO, U.S.A.). Fine suspensions of phospholipids were prepared by sonication in 1 mM Tris±HCl (pH 7.9), containing 100 mM sodium chloride, 20 potassium chloride, 10 mM calcium chloride and 0.5% (v/v) Triton X-100 in an ultrasonic bath. For the purpose of identi®cation of products formed by phospholipase digestion, the same bu€er as described above was used, but the concentration of Tris±HCl was increased to 50 mM. The digests were examined by thinlayer chromatography (TLC) on silica gel plates (Whatman LK6DF) as described by King et al. (1984): 35 ml (1.5 mg/ml phospholipid) was applied to the preadsorbent zone and the plate was developed in chloroform± methanol±0.1 N HCl (60 : 35 : 5). The spots were visualized by exposure to iodine vapor for detection of monoacyl phospholipids. The digests were also examined directly for the presence of saturated fatty acids by chro-

Phospholipases A2 from Paulistinha Venom

191

matography on freshly prepared silica gel plates which had been dipped in 5% (w/v) AgNO3 and then dried at 1108C for 1 hr. After developing in presence of hexane±diethyl ether±acetic acid (70 : 30 : 1), the spots were visualized under UV light after spraying with 0.2% (w/v) dichloro¯uorescein in ethanol. The procedure described above was also applied to PLA2 from honey bee venom (Sigma) for use in a control experiment. PLA2 activity The assays were routinely carried out using a spectrophotometric method based on pH change due to the liberation of fatty acids, as described by Araujo and Radvanyi (1987). The reaction medium contained 15 mmoles phosphatidylcholine, 18 mmoles Triton X-100, 5 mmoles calcium chloride, 80 nmoles phenol red and 7.5 mmoles Tris in a ®nal volume of 2.5 ml, at pH 7.9. The absorbance was initially read at 558 nm against a proper reference; the reaction was initiated by the addition of either crude/puri®ed venom or puri®ed toxins. The decrease in the absorbance of phenol red, caused by the acidi®cation of medium, was measured after 5 min of incubation at 378C. The DA558 was proportional to the liberation of fatty acids in the assay conditions. One unit of PLA2 activity was de®ned as the amount of enzyme necessary to hydrolyse 1 mmole phosphatidylcholine/hr in 1 ml of the reaction medium at pH 7.9 and 378C. Hemolysis Direct hemolytic activity was assayed on washed mouse red cells with slight modi®cations to the procedure described by Ho and Ko (1988). The toxins were dissolved in 0.14 M saline Tris (0.01 M) bu€ered at pH 7.4 and the red cells, suspended in the same solution at the hematocrit 50%, were mixed and incubated at 378C for 60 min. Hemolysis was stopped by the addition of cooled (48C) Tris-bu€ered saline to a ®nal volume of 5 ml and the degree of hemolysis was determined by measuring the released hemoglobin at 545 nm. Similarly, control samples were incubated in the absence of the toxins, by hemolysing the red cells in water under similar conditions. The hemolytic potency was expressed as per cent hemolysis, assuming the lysis in water to be 100% in the given incubation time. The above determinations were run in triplicate at the end of ®ve independent preparations. The results are expressed as means2 S.E. PLA2 puri®cation All puri®cation steps were carried out at 0±48C, unless otherwise speci®ed. The freeze-dried venom (30 mg) was solubilized in 5 mM ammonium acetate, pH 6.8, and applied to a Sephadex G-200 column (51.0  2.5 cm) previously equilibrated with the same bu€er. Elution was performed with 5 mM ammonium acetate, pH 6.8, at a ¯ow rate of 12 ml/hr and fractions of 3 ml were collected in the presence of 100 ml of 10% (v/v) glycerol (previously added to each collection tube). Protein elution was monitored along the pro®le of elution by measuring the absorbance at 280 nm and PLA2 activity assayed as described above for each collected fraction. The fractions presenting phospholipasic activity were pooled and freeze-dried. The pool containing the PbTX (2.4 mg protein) was then solubilized in 10 mM sodium acetate bu€er, pH 4.5, and applied to a DEAECellulose column (280  10 mm), previously equilibrated with the same bu€er. The elution was performed under gradient from the equilibrium bu€er to 10 mM sodium citrate, pH 6.0, containing 0.1 M sodium chloride, at a ¯ow rate of 12 ml/hr; fractions of 3 ml were collected in the presence of 100 ml of 10% (v/v) glycerol (previously added to each collection tube). Protein elution was monitored along the pro®le of elution by measuring the absorbance at 280 nm and PLA2 activity assayed as described above for each collected fraction. The fractions presenting phospholipasic activity were pooled, freeze-dried and kept at ÿ168C until used. This procedure was carried out in at least ®ve independent preparations. Homogeneity After puri®cation the homogeneity of PbTX was examined by using reduced (pretreated with 5% mercaptoethanol) samples on sodium dodecyl sulfate±polyacrylamide gel electrophoresis (SDS±PAGE) (5±20%) as described by Weber and Osborn (1969). Molecular weight determination The mol. wts of the PbTX were estimated both by molecular exclusion chromatography (Andrews, 1964) and by SDS±PAGE as described above. Kinetic analysis Initial velocities (V) were plotted as function of substrate concentration in Lineweaver±Burke plots and the interaction constants for the substrate (h) were determined by the Hill procedure as described by Koshland (1970). Since all the kinetics of the present paper were biphasic, the Km values were expressed as S0.5 and determined from the Hill plotts (when V = Vmax/2), while the Vmax values given in this paper were obtained

192

M. R. de OLIVEIRA and M. S. PALMA

from linear-square analysis for one of the components of the curve, calculated in at least three independent experiments.

RESULTS

Puri®cation and molecular properties of PLA2 Qualitative estimation by TLC showed that after 4 hr digestion both the phospholipase from the crude venom and puri®ed toxins presented the same results: only unsaturated fatty acid (oleic acid) was released from 1-stearoyl-2-oleoyl-3-snglycerophosphoryl choline, only saturated fatty acid (stearic acid) was released from 1oleoyl-2-stearoyl-3-sn-glicerophosphorylcholine, only unsaturated fatty acids were released from phosphatidylcholine and no fatty acid was released from the digestion with lysophospatidylcholine (results not shown). The analysis of digests of PLA2 from honey bee venom revealed very similar results; since the hydrolytic and speci®city of this enzyme is well established the results of these control experiments may be used as structural proof of the substrates used and products delivered during the digestions. The venom from 12,000 workers (29.9 mg protein) was fractionated on a Sephadex G-200 gel-®ltration column (Fig. 1). Approximately 2.4 mg of proteins was eluted into a large peak presenting PLA2 activity, with an estimated 125,000 mol. wt. PLA2 was further puri®ed by ion-exchange chromatography on a DEAE-Cellulose column with a

0.08

96

I

6.0

88 0.07 80

II 0.06

72

III 5.5

)

64

40

IV

0.03

32

) pH (

48

0.04

PLA 2 activity (

A 280 (

)

56

U.ml -1.h -1)

0.05

5.0

24

0.02

16 0.01 8

0

5

10

15

20

25

30

35

40

45

0

4.5

Fraction number Fig. 1. Molecular exclusion chromatography of crude venom from P. paulista on a Sephadex G200 column (51.0  2.5 cm). The venom (29.9 mg) was applied to the column and eluted with 5 mM ammonium acetate, pH 6.8. Fractions of 3 ml were collected at a ¯ow rate of 12 ml/hr.

Phospholipases A2 from Paulistinha Venom

193

linear pH gradient (Fig. 2). Four di€erent fractions presenting PLA2 activities were eluted at pH 4.6, 4.7, 4.8 and 5.6 and designated as PbTX-I, II, III and IV, respectively. The yield of total PLA2 activity was about 95.7% in relation to the crude venom and the four PbTX together represented 1.1% of the total protein of crude venom (Table 1). PbTX-I, II, III and IV were apparentely homogeneous on SDS±PAGE; however, each of them dissociated into two di€erent subunits under reduced form (Fig. 3). Thus, the mol. wt values estimated for the subunits of each toxin were: 64,000 and 55,000 for PbTX-I; 70,000 and 62,000 for PbTX-II; 70,000 and 55,000 for PbTX-III; and 62,000 and 53,000 for PbTX-IV. In order to understand the origin of the heterogeneity seen on gel electrophoresis among the four PbTX, the content of carbohydrate of each toxin was determined as being 29, 43, 32 and 22% (w/w) for PbTX-I, II, III and IV, respectively. After 1 week's storage at 48C, PbTX lost up to 40% of PLA2 acitivity. However, when kept at 48C in the presence of glycerol, the phospholipasic activity was maintained for at least 1 month. To remove glycerol, the preparartion was quickly dialysed in the presence of 40 mM Tris, pH 7.8±8.3. The speci®c activities for the puri®ed PbTX in the presence of 2.5 mM Ca2+ ions were 1235, 64, 372 and 30 mmoles/mg per minute for PbTX-I, II, III and IV, respectively. Kinetic characterization of PLA2 activity The PLA2 activities of the four toxins are dependent on a low concentration of Ca2+ ions (lower than 1 mM) and activated by increasing concentrations of this ion until 10 mM. However, when the assays were repeated in the presence of both 1 mM Ca2+

96

0.225

32

0.075

PLA 2 activity (

) A 280 (

48

U.ml -1.h -1)

64

0.150

)

80

16

0

30

75

105

135

165

0

Volume of elution (ml) Fig. 2. Ion-exchange chromatography of the PbTX (2.4 mg) on a DEAE-Cellulose column (28  1 cm). Elution was performed with a one-step linear gradient from 10 mM sodium acetate, pH 4.5, to 10 mM sodium citrate, pH 6.0. Fractions of 3 ml were collected at a ¯ow rate of 12 ml/hr.

PbTX-I PbTX-II PbTX-III PbTX-IV

6.0 12.0 5.7 8.7

3.0 3.5

Crude venom Sephadex G-200 + freeze-drying 0.9 16.7 2.0 10.0

8634.0 692.3

Protein (mg/ml)

Activity (U/ml)

29902.0 549.4 2423.0 458.0 DEAE-Cellulose 5.4 66.7 200.4 64.3 11.4 44.6 87.0 17.6

Total protein (mg)

400 771 254 153

1648 1603

Total units (U)

74.11 3.85 22.30 1.80

0.05 0.66

Speci®c activity (U/mg)

Table 1. Puri®cation records of PbTX from the venom of the social wasp P. paulista Volume (ml)

Fraction

24.2 46.8 15.4 9.3

100.0 97.3

Yield (%)

1482 77 446 36

Ð 13

Puri®cation

194 M. R. de OLIVEIRA and M. S. PALMA

Phospholipases A2 from Paulistinha Venom

195

Fig. 3. SDS±PAGE under reducing conditions of puri®ed PbTX, obtained in a 5±20% linear gradient gel. The lane of standards contained the following mol. wt markers: bovine albumine serum (66,000), egg ovalbumin (45,000), lactic dehydrogenase (37,000), trypsinogen (24,000), b-lactoglobulin (18,000) and lysozyme (14,000).

and 1 mM EDTA, no activity was observed. Assays in the presence of 2 mM of Cu2+, Zn2+ and Co2+ inhibited the PLA2 activities of PbTX. Double-reciprocal plots of initial velocities, at pH 7.9, as a function of phosphatidylcholine concentration, are shown in Fig. 4. It can be seen that the double-reciprocal plots did not give straight lines and were characterized by Hill coecients lower than 1. Since the kinetics of phosphatidylcholine hydrolysis were not Michaelian, the constant values expressing the anity for the substrate binding to PbTX were expressed as S0.5 instead of Km; the magnitude of S0.5 values ranged from 10ÿ4 to 10ÿ8 M (Table 2). Hemolysis The hemolytic activities of PbTX were not dependent on Ca2+, Mg2+ or Ba2+ ions. However, these ions were activators of the hemolytic activity (10 mM causes a rate activation of 35%). It can be seen from Table 3 that all the PbTX were extremely potent in lysing red cells in Tris-bu€ered saline (direct hemolysis). It must be emphasized that PbTX were capable of lysing the membrane phospholipids per se, i.e. without combination with any other components. The most active were PbTX-III and IV; EDTA concentrations from 1 to 10 mM had no e€ect on the hemolysis of erythrocytes induced by PbTX. However, 0.5 mM of Cu2+ or Zn2+ ions inhibited 50% of the hemolytic action. DISCUSSION

Phospholipases A have been detected in several di€erent organisms, especially the enzymes of A2 type from animal venoms that are responsible for several myotoxic and/

196

M. R. de OLIVEIRA and M. S. PALMA

5

-0.5

h = 0.44

8

-1.0 -5.5

-5.0

-4.5

6

log [phosphatidylcholine]

2

4

1

2

V V max - V

-1.5 -6.0

0.6 0.3

h = 0.52 0 -6.0

-5.5

-5.0

-4.5

log [phosphatidylcholine]

10

18

27

36

0

45

8

24

6

18

log 2

18

10

PbTX-III

4

9

12

h = 0.52

0.2 0 -6.0

-5.5

-5.0

6

-4.5

9

18

27

36

45

0.6 0.4

h = 0.20

0.2 0 -6.0

-5.5

-5.0

-4.5

log [phosphatidylcholine]

log [phosphatidylcholine]

0

36

PbTX-IV

0.6 0.4

27

V V max - V

9

log

0

V V max - V

1/V (µMoles.ml-1.h -1) [x 102]

3

PbTX-II

log

V V max - V

log

4

10

PbTX-I 0

45

0

9

18

27

36

1/[phosphatidylcholine] [M]-1 (x 102)

Fig. 4. Double-reciprocal plots of initial velocity of PLA2 activity from the PbTX (150 ng/tube) as a function of phosphatidylcholine concentration at pH 7.9. The inserts represent the Hill plotts for each kinetic assay.

or neurotoxic e€ects of these venoms (Ho and Ko, 1988). However, most PLA2 from hymenopteran venoms have been poorly characterized from a biochemical point of view and most of the attention has been focused on their immunological properties. Table 2. Summary of some steady-state kinetic parameters for the hydrolysis of phosphatidylcholine as substrate at pH 7.9 by the PbTX from the venom of P. paulista Enzyme form PbTX-I PbTX-II PbTX-III PbTX-IV

S0.5 (M) 8  10ÿ4 2  10ÿ6 9  10ÿ5 7  10ÿ8

Vmax (mmoles mlÿ1 hrÿ1) 320 5 40 3

Hill coecient 0.44 0.52 0.52 0.20

45

Phospholipases A2 from Paulistinha Venom

197

Table 3. Comparison of the direct hemolytic activity of PbTX with toxins from honey bee, wasps and snakes Toxin

PbTX-I PbTX-II PbTX-III PbTX-IV Hornetin from Vespa ¯avitarsus* Lethal toxin from Vespa basalis* PLA2 from Apis mellifera Neutral PLA2 from Naja nigricolis* Basic PLA2 from Naja nigricolis* Cobra cardiotoxin from Naja naja atra*

Concentration (mg/ml) PbTX

100 100 100 1 1 100 100 100 100 100

Direct hemolysis (%) mouse RBC 6224 (15) 5723 (15) 9026 (15) 6023 (15) 1322 9621 2723 (15) 35210 9221 1621

Data are shown as means2S.E. The number of experiments is given in parentheses. *From Ho and Ko (1988).

From the characterization of phospholipase type experiments performed in the present study, is evident that the site of hydrolysis of the acyl group is at the 2-position, both for the phospholipase activity in the crude venom and for the puri®ed toxins. Thus, we puri®ed and characterized one family of toxins presenting PLA2 activity from the venom of a neotropical social wasp, P. paulista. About 96% of the original activity was recovered from the crude venom with a degree of puri®cation that changed from 36- to 1482-fold, depending on the PbTX type (I,II,III or IV) (Table 1). The total amount of the four PbTX in the crude venom of P. paulista may be considered very reduced (1.1%) when compared with those observed in the venoms from Apis mellifera (10±12%) (Dotimas and Hider, 1987) and Vespa basalis (6%) (Ho and Ko, 1988). The mol. wt of PbTX (from 115,000 to 132,000) is very di€erent from those previously described for the phospholipases from other hymenopteran venoms, such as 36,000 for PLA2 of Polistes exclamans (King et al., 1984), 43,000 for hornetin of Vespa ¯avitarsus (Ho and Ko, 1988) and 32,000 for the lethal protein of V. basalis (Ho and Ko, 1988). In addition, all of these other toxins are apparently monomers, while PbTX are heterodimers. The di€erences in the mol. wt observed among the PbTX may be due to the di€erent content of carbohydrates attached to the toxin molecules, which ranged from 22 to 43% (w/w) depending on the PbTX type. The observed values are much higher than the 8% (w/w) previously described for the PLA2 from the venom of A. mellifera (Banks and Shipolini, 1986). These values are large enough to a€ect the mol. wt estimatimation if either molecular exclusion chromatography or SDS±PAGE had been used alone. However, the sum of mol. wt of each subunit obtained by electrophoresis is consistent with the rate value obtained for the dimers of PbTX in the chromatographic method. In order to obtain more evidence to determine that PbTX are really dimers, the subunits of each PbTX were dissociated (in the presence of 0.1 M NaCl) and separated from each other on a Sephadex G-100 column (results not shown). In these conditions, only a residual activity remained with the larger subunits of each PbTX, while the smaller subunits were inactive, suggesting that PLA2 activity is dependent on a dimeric state of aggregation.

198

M. R. de OLIVEIRA and M. S. PALMA

Since the PbTX eluted from the DEAE-Cellulose column at di€erent pH values between 4.6 and 5.6, it seems that each toxin presents a di€erent number of negative charges on its surface. Their di€erent carbohydrate contents may also contribute to the di€erential degree of these charges. The speci®c activities of the PbTX for hydrolysis of phosphatidylcholine were lower than those described for hornetin from V. basalis (3800 mmoles/mg per minute) and PLA2 from honey bee venom (1471 mmoles/mg per minute). However, PbTX-I and PbTX-III (1235 and 372 mmoles/mg per minute, respectively) are more active than the basic PLA2 from Naja nigricolis venom (323 mmoles/mg per minute) (Ho and Ko, 1988). The PLA2 activities of PbTX are dependent on calcium ions and weakly activated by other divalent cations, such as magnesium, barium and strontium. Copper and zinc ions were potent inhibitors of the catalytic activities of PbTX, as previously described for the PLA2 orientotoxin from the venom of the giant hornet Vespa orientalis (Tuichibaev et al., 1987). The hydrolysis kinetics of the four PbTX for phosphatidylcholine at pH 7.9 (Fig. 4) deviated signi®cantly from Michaelis±Menten behavior and the double-reciprocal plots presented two linear sections with downwards concavity and Hill coecients lower than 1. These data indicate the possibility of negative co-operativity kinetics. The values obtained for the Hill coecients suggest that there is only one catalytic site. Thus the negative co-operativity phenomenon may due to the interaction between the two subunits of each PbTX. The PbTX exhibited extremely potent hemolytic actions on red cells incubated in Tris-bu€ered saline (direct hemolysis) (Table 3). All the PbTX are hemolysins, more potent than the most lytic factors from honey bee and snake venoms, especially PbTXIII, which is as active as the lethal toxin from V. basalis and more potent than hornetin from V. ¯avitarsus. It should be noted that PbTX-IV exhibits hemolytic action 20 times higher than that of PLA2 from A. mellifera, 17 times higher than that of neutral PLA2 from N. nigricolis and about 37 times than that of cardiotoxin from Naja naja atra. The high hemolytic activities of PbTX against washed erythrocytes suggest that these toxins are capable of lysing biological membranes without the presence of exogenous phospholipids or special polypeptides, as is required by the PLA2 from honey bee venom (Banks and Shipolini, 1986; Lawrence and Moores, 1975). This activity is not in¯uenced by EDTA, but is inhibited by Cu2+ and Zn2+. Thus, with regard to these aspects, the PbTX seem to be similar to PLA2 orientotoxin in the venom of V. orientalis (Tuichibaev et al., 1987). The above results suggest that PbTX constitute a group of four toxins puri®ed from the venom of the social wasp P. paulista, presenting PLA2 activity and some molecular properties, such as mol. wt, aggregation state and content of sugar attached to the molecules, which are very di€erent from all known hymenopteran PLA2. However, PbTX present several kinetic properties similar to the PLA2 orientotoxin from the venom of V. orientalis. A comparison of the direct hemolytic activities of PbTX with the hemolytic factors from other sources revealed that the PbTX are among the most potent hemolysins from animal venoms. Polybia paulista is a very aggressive social wasp found in south-east Brazil, which causes several stinging accidents throughout the year. In spite of this and considering the high hemolytic activity of PbTX, the low incidence of fatal stingings by P. paulista may be explained by the reduced amount of PbTX in this venom (1.1% of the total protein from the crude venom). However, cases of multiple stingings are not uncommon and are generally followed by severe hemolysis and myocardial dysfunction. In this situ-

Phospholipases A2 from Paulistinha Venom

199

ation, the described symptoms may be caused by the accumulation of high levels of PbTX. Acknowledgements ÐThe authors acknowledge the following ®nancial support: MSP is a researcher of Conselho Nacional de Desenvolvimento CientõÂ ®co e TecnoloÂgico (CNPq proc. 300658/86-9); MRO was a fellow from Fundac° aÄo de Amparo aÁ Pesquisa do Estado de SaÄo Paulo (FAPESP proc. 91/4267-0).

REFERENCES Andrews, P. (1964) The gel ®ltration behavior of proteins to their molecular weights over a wide range. Biochemical Journal 96, 595±606. Araujo, A. L. and Radvanyi, F. (1987) Determination of phospholipase A2 activity by a colorimetric assay using a pH indicator. Toxicon 25, 1181±1188. Banks, B. E. C. and Shipolini, R. A. (1986) Chemistry and pharmacology of honey bee venom. In Venoms of the Hymenoptera, ed. T. Piek, pp. 329±416. Academic Press, London. Castro, F. F. M., Palma, S. P., Brochetto-Braga, M. R., Malaspina, O., Lazaretti, J., Baldo, M. A. B., Antila, M. A., Zuppi, L. and Cossermelli, W. (1994) Biochemical properties and study of antigenic cross-reactivity between Africanized honey bee and wasp venom. Journal of Investigational Allergology and Clinical Immunology 4, 37±41. Dotimas, E. and Hider, R. C. (1987) Honeybee venom. Bee World 68, 51±71. Habermann, E. (1972) Bee and wasp venoms. Science 177, 314±322. Hartree, E. F. (1972) Determination of protein: modi®cation of Lowry method that gives a linear photometric response. Analytical Biochemistry 48, 422±427. Ho, C. L. and Ko, J. L. (1988) Puri®cation and characterization of a lethal protein with phospholipase A1 activity from the hornet (Vespa basalis) venom. Biochim. biophys. Acta 963, 414±422. Ho€man, D. R. (1978) Allergens in Hymenoptera venom. V: Identi®cation of some of the enzymes and demonstration of multiple allergens in yellow jacket venoms. Annals of Allergy 40, 171±176. Ho€man, D. R. (1985) Allergens in Hymenoptera venom. XIII: Isolation and puri®cation of protein components from three species of vespid venoms. Journal of Allergy and Clinical Immunology 75, 599±605. King, P. T., Kochoumian, L. and Joslyn, A. (1984) Wasp venom proteins: phospholipase A1 and B. Archives of Biochemistry and Biophysics 230, 1±12. Koshland, D. E., Jr (1970) The molecular basis for enzyme regulation. In The Enzymes, ed. P. D. Boyer, Vol. 1, pp. 342±395. Academic Press, New York. Lawrence, A. J. and Moores, G. R. (1975) Activation of honeybee venom phospholipase A2 by fatty acids, aliphatic anhydrides and glutaraldehyde. FEBS Letters 49, 287±291. Nakajima, T. (1986) Pharmacological biochemistry of vespid venoms. In Venoms of the Hymenoptera, ed. T. Piek, pp. 309±327. Academic Press, London. Reisman, R. E. and Osur, S. L. (1987) Allergic exposures following ®rst insect sting exposure. Annals of Allergy 59, 429±432. Tuichibaev, M. U., Akhmedova, N. U. and Muksinov, F. A. (1987) Hemolytic e€ect of phospholipase A and orientotoxin from the venom of the giant hornet Vespa orientalis. Biokhymiya 53, 434±443. Weber, K. and Osborn, M. (1969) The reliability of molecular weight determinations by dodecyl sulfate gel electrophoresis. Journal of Biological Chemistry 244, 4406±4412.