Isolation and characterization of a phospholipase B from venom of collett's snake, Pseudechis colletti

Tarko+r. Vd . 25 . No . S, PD . Primed in Great Erkain.

547-554, 1987.

004i-oiova~ ss .oo+ .ao ® 19a7 Perpmon laumala Ltd .

ISOLATION. AND CHARACTERIZATION OF A PHOSPHOLIPASE B FROM VENOM OF COLLETT'S SNAKE, PSEUDECHIS

COLLETTI

A. W. BERNHEIMER, i R . LINDER,2 S. A. WEINSTEINi and K.-S. KIM i 'Department of Microbiology, New York University School of Medicine, 550 First Avenue, New York, NY 10016, U .S .A ., and =Eiunter College School of Health Sciences, 425 East 25th Street, New York, NY 10010, U .S .A . (Amepted jor publication 2 DrcYmber 1986) A . W . BEata ~tt~Q , R . Lwnr~e, S . A . WetxsrP.tN and K .-S . Kut . Isolation and characterization of a phospholipaae B from venom of Collett'a snake, Psrudcrhis colletti. Toxicon 25, 547-554, 1987 .-Phospholipase B in the venom of the Australian elapid snake, Psruderhis colletti, was purified to near homogendty . By means of gd filtration it had an M, of about 35,000, and by SDSpolyacryhunide gd dectrophoreais an M, of about 16,500. These presumably arc dimeric and monomeric forms of the enzyme . It was isoeletaric at pH 6 .2 as compared to 7 .8 for phospholipaae A= from which it was readily separated . It was rchttively thermostable . As determined by release of water-soluble phosphorous, it degraded phoaphatidylcholine and phosphatidylethanolamine, but did not degrade other phospholipida listed . The purified enzyme was strongly hemolytic in vitro for rabbit and human erythrocytes, but not for bovine or ovine erythrocytes . Hemolytic of rabbit erythrocytes gave rise to membranes showing ultrastructural changes that may be unique for this enzyme. The protein was highly active in producing turbidity in dilute solutions of egg yolk. It was cytotoxic for cultured rhabdomyoaarcoma cells and was lethal for mice in which death was preceded by massive myoglobinuria .

INTRODUCTION

AN EARLIER study (BERNHEIMER et al., 1986) of the venom of Pseudechis colletti suggested that phospholipase B (PLB) is responsible for the venom's hemolytic activity as well as for its capacity to produce turbidity in diluted egg yolk. However, the enzyme preparations used were impure, and the results were therefore considered tentative. We have since obtained the PLB in a more purified form, and we have studied some of its physico-chemical and biological properties . The results described here confirm and extend the earlier findings . The term phospholipase B is used here to designate ''phospholipase B activity" without implying the mechanism whereby both fatty acids are split from substrates such as phosphatidylcholine and phosphatidylethanolamine . The approaches to this problem have been discussed by $HILOAH et al. (1973) in connection with the PLB activity of Vipers palestinae venom. MATERIALS AND METHODS Lyophilized venom was purchased from Sigma Chemical Company, St Louis, MO . Hemolytic activity and egg yolk activity (the latter u a measure of PLB) were asasyed as described (Bt:atvt~t~a et al., 1986). Protein was estimated dther by 280 nm absorption or by both 280 nm and 235 nm absorption (Wttrrwt®e and GRwrtuM, 1980) . 547

S48

A. W. BERNHEIMER et al. 3.3 3.0

~ 2 .0

200

O aD N a

150 E ô n 100 w 50

70 Fraction

No .

flc. 1. ELUilON PROFILE OF SEPItwcRn S-200 cottrntlv. PLB as egg yolk activity (" ) doted as a :iagle peals coinciding with a :mall shoulder on a large peak of 280 nm absorbance (p). Re&~ase by PLB of w~ater~olubk phosphorous front phaspholipidr Chromatographically pure phospholipids were purchased from Supelco, Inc., Bdkfonte, PA . Four aliquots of each phosphollpid in organic solvent, containing approx . 230 Yg lipid each, were driedat the bottom of tubes. Each aliquot was suspended in 400 pl of 0.15 M NaC 1, 0.01 M CaCI, a~ 0.1 ~ bovine serum albumin buffered at pH 7.2 with 0.01 M Tria-HCI . The tubes were scouted and vortex mixed to obtain smooth auapensio~. taste tuoe of each phospholipid was used for daermination of total P, twô for incubation with PLB and one for incubation without PLH. After incubation at 37°C for 30 min the reaction was stopped by addition of chloroform-mahanol (2 :1), and total lipid extraction wu carried out by the method of Btlae and IiYER (1959). One milliliter samples of the aqueous phase (total volume, 1.5 mn were removed and dried in a boiling water ~th. The residues were ached using lOsFs MgNO, in ethanol, and orthophosphate was daermined by the method of AIKt?s and DUBIN (1960) . FJectron nlicramopy Negative staining was performed with 2S~ ammonium molybdate in distilled water at pH 5.2 . Samples prepared on Formar, carbon~oated grids as described by COWELL et al. (1978) were examined in a Siemens Elmiskope IA . RESULTS Pur~cation and molecular weight of PLB

Oae-hundred milligrams of dried venom were dissolved in 2 ml of 0.1 M KCI and 0.03 M sodium borate at pH 8.2 (buffer 1). One-half milliliter.of 0.4% blue dextran 2000 (LKB Instruments, Rockville, MD) in 10% glycerol was added. Size-fractionation was done at 4°C in a 2.5 x 30 cm column of Sephacryl 5-200 superfine (Pharmacia) using buffer 1 . The flow-rate was 43 .6 ml per hr, and fractions of about 4 ml each were collected. The distribution of PLB as egg yolk activity and that of protein as 280 nm absorbance, are plotted in Fig. 1 . It can be seen that egg yolk activity coincided with a small shoulder

349

Phospholipase B of Pseudtchis colletti

centering about Fractions 28 and 29 . The elution volume of the shoulder relative to those of several standard proteins indicated a molecular weight of about 35,000 . Fractions 26-30 were pooled and dialyzed at 4°C overnight against one liter of 5% glycerol . The resulting solution was subjected to electrofocusing under the conditions described for Austrelaps superbus venom (BERNHEII~It et al., 1986) using reversed electrodes (Fig. 2). The two fractions showing strongest egg yolk activity (Fractions 21 and 22 of Fig. 2) were pooled and dialyzed against 200 ml of saturated enzyme-grade ammonium sulfate. The resulting precipitate was collected by centrifugation and suspended in 2 ml of saturated ammonium sulfate. The specific activity and yield at each step are shown in Table 1 . Purification of a second 100 mg lot of venom yielded 24% of starting activity . It contained 1370 egg yolk units and 2400 hemolytic units per mg protein. SDS-polyacrylamide gel electrophoresis was done according to the method of LAE1KMLl (1970) in 10% acrylamide gels and 6% stacking gels . Samples were heated at 100°C for 3 min in 1% SDS and 2% 2-mercaptoethanol. Ten micrograms of the purified protein previously freed of ammonium sulfate by dialysis against distilled water, were

E Y

a E

C

Ô E Y

a

W

0 07 N

2

300

0

400

Y V C

er

300

ô w Q

100

d

200

IS

16 I7 18 19 20 21

22 23 24 23 26 27 26 29 30

Fraction No . FIO. 2 . DISTRIBUTION OF PLB ~IVO HeMOt.rnc Acrrvrrv AMONG FRwcnolvs OBTAINED aY ELECrao~snvG . The hemolytic activity ((y coincides with PLB as measured by egg yolk activity ( t ) . Distribution of protein is shown as 280 nm abaorbana (p) . The pH of the peak fraction wan 6.1 . TABLE 1 . PURIFICATION oF PLB Stage Crude venom solution Dialyzed pool from Sephacryl 5200 Electrofocuaed fractions 21 and 22 Ammonium sulfate precipitate

Volume (ml)

Egg yolk activity (units/ml)

2

3000

20

260

8 2

103 375

Protrin (mg/ml)

egg yolk activity

(units/mg protein)

Total egg yolk units

Per cent of starting activity

60

6000

100

0 .38

684

3200

87

1 .0 0 .80

420 470

840 940

14 16

SO

Spocific

SSO

A . W. BERNHEIMER et al.

applied to the gel. The Coomassie blue stained gel showed a single heavy band and two relatively faint bands, the latter in the region of M~ 60,000 . The enzyme, assumed to be the heavy band, appeared to have a purity considerably greater than 90% . It had an M~ of 16,500 as determined by comparison to standard proteins (Fig . 4) . The clean separation of PLB from phospholipase A by electrofocusing is shown in Fig 3. The input material and the conditions of electrofocusing were identical with those described, but 280 nm absorbance was continuously recorded instead of being measured for each fraction . The larger peak was identified as phospholipase A (a) by the fact that as little as 0.1 lrg protein caused a SO% reduction in turbidity of diluted egg yolk in less than 2 min (MARINETTI, 1965) and (b) by finding that it liberated a large amount of fatty acid from labelled Escherichia coli phosphatidylethanolamine using the method of WEISS et al. (1979) . The smaller peak was identified as PLB by its high activity in increasing the turbidity of egg yolk and by the liberation of water-soluble phosphorous from phospholipids (described in the next section). Substrate specificity of the PLB

PLB (13 Ng) was added to samples of phospholipids prepared as described in Materials and Methods. After incubation and extraction of the reaction mixtures, it was found that PLB released 2 .5 pg P when phosphatidylcholine was the substrate, 1 .2 Fig P when phosphatidylethanolamine was the substrate, and none when phosphatidylserine, o.e

0.6 E c O m N

t 0 u c 0 a

0.4

ô

a

0.2

I I I I I I 1 I I I I I 1 I I I 1 I 1 I I I I I I I I I I I 30 26 26 24 22 20 18 16 14 12 10 9 6 4 2 F~oction No . Fyo . 3 . SEPARATION OF PLB Arro PLA Ex IsoELECrROPOCVSIrIa . The conditions of isoelectric focusing and the methods of identifying the phospholipases are given in the text .

Phospholipase B of Psrirdechis rnNttti

551

SO 40 ~ 30 Y J=

20 IS 10

Phospholipasa B ~- lactal bumin I 20

I 30

I 40

Migration

I SO

I 60

I ?0

I 80

distance in mm .

FYo . 4. ESTIMATION OF Mot.ecut.,~tt Weta~tT of PLB. The migration distance of PLB in SDS-polyacrylamide gel electrophoresis is compared to those of standard proteins, namely : ovalbumin (43 kd), carbonic anhydrase (30 kd), soy bean trypsin inhibitor (20 kd) and o-(actalbumin (14.4 kd). An M, value for PLB of 16,500 is obtained .

Fla . 5 . Ececraox Mtcnosoorlc ArpFwAwNCx of Mel~awxes Dealveo FaoM PLB-TaswTen

Rwaarr EaxTlatocrr~ . Washed rabbit erythrocytes were lysed with PLB (90 Pg per ml) for 90 min at 37°C in a 0.2~ gelatin wlution, 0 .15 M NaCI, 0.01 M MgCI=, 0.01 M CaCI, buffered at pH 7.2 with 0.01 M Tris-HCI . The reaction mixture was then chilled in ico-water for 10 min. The resulting membranes were collected by centrifugation and then examined by staining with 2~fi ammonium molybdate. The membrane shows large numbers of angular aructurea having an average diameter of about 35 nm . Bar represents 100 nm .

ssz

n . w . sB~~i~R ~

er.

phosphatidylinositol, sphingomyelin or cardiolipin were used as substrates . The values obtained with duplicate reaction mixtures were in agreement to t 7 .S%. When the substrate concentrations were reduced by one-half, the same amounts of water-soluble P were released, indicating that the reactions were carried out under conditions of substrate excess . The results are in agreement with those obtained by thin layer and paper chromatography (BERNHEIMER et al., 1986) except for the finding that phosphatidylserine appeared to be susceptible to hydrolysis in the earlier study. The explanation of this discrepancy is not clear . Thermostability Samples of purified PLB diluted in 0.15 M NaCI, 0.01 M MgCIZ, 0.01 M CaCIZ and 0.2% gelatin buffered at pH 7 .2 with 0.01 M Tris-HCl (buffer 2), to contain 100 HU per ml were placed in baths at 0, 37, 60 and 100°C for 30 min. After chilling they were titrated for hemolytic activity . Full activity was retained at 0, 37 and 60°C whereas 47% was recovered after 30 min at 100°C. Sensitivity of different species of erythrocytes The hemolytic activity of purified PLB was titrated using erythrocytes from five different rabbits . Activities of 400, 500, 500, 1000 and 1250 HU per ml (average 730 HU per ml)were obtained, respectively . The same PLB preparation contained 440 HU per ml when tested against human erythrocytes, and <1S HU per ml when tested against ovine or bovine erythrocytes . The resistance of sheep and cow erythrocytes to lysis is consistent with the fact that the phospholipid in the outer half of the erythrocyte membrane of these animals is almost exclusively sphingomyelin, and sphingomyelin as shown earlier (BERNHEIMER et al., 1986) is not a substrate for the PLB under study. Ultrastructural changes in erythrocyte membranes accompanying hymolysis Phase-contrast examination of preparations resulting from treatment of washed rabbit erythrocytes with partially purified P. colletti PLB revealed ghosts of reduced diameter containing many vesicles of various sizes (BERNHEIMER et al., 1986). In contrast, repetition of this work using the more highly purified preparation of PLB described here revealed ghosts showing relatively little size reduction, and the presence of internal vesicles was relatively inconspicuous . However, numerous internal vesicles were readily seen when the ghosts were examined electron-microscopically both in negatively stained and in sectioned preparations . Aggregates of many small vesicles (SO-200 nm in diameter) which had escaped from the ghosts were also observed . High resolution electron microscopic examination of negatively stained preparations revealed membranes bearing large numbers of irregularly shaped structures (Fig. S). These structures had a diameter of about 35 nm (10-100 nm) and their outline which was rarely circular or oval, suggested three-dimensional polyhedral units having rounded angles. As can be seen in the figure, the majority of these "lesions" were relatively uniform in siu. To our knowledge, structures resembling them have not been described as a result of the action of other lytic agents on erythrocytes . Toxicity of PLB for cultured rhabdomyosarcoma cells To see whether PLB is toxic for cells other than erythrocytes, two-fold dilutions of PLB (27 Ng per ml) in serum-free Dulbecco's modified Eagle medium (St~tt~t et al., 1960) were added to rhabdomyosarcoma cells A673 in microtest plates . The mixtures were incubated

Phospholipaae B of PsYUdechts rnlletti

553

in S% C0 2 at 40°C for 2 days . Percentage of survivng cells was determined by intravital dye uptake according to the method of F~INMAN et al . (1987) . Two-fold dilutions of PLB were used and each dilution series was done in triplicate. The percentage of cells surviving after exposure to enzyme concentrations of 28, 14, 7, 3.5 and 0 ~tg per ml were 11, 87, 86, % and 100, respectively. The results indicate that 7 ~g PLB per ml exhibited a significant degree of cytotoxicity . In contrast, no clearcut effect was seen when the enzyme was tested in the same concentrations against human foreskin fibroblasts FS4. Human platelets were not lysod by PLB at a concentration of 3.5 pg per ml, tested according to the method of BERNHEIMER and AVIGAD (1976) . Higher concentrations were not tested .

Fffects in mice

Five Swiss-Webster mice weighing about 20 g each were injected i .p . with 120 hg purified PLB. They displayed ataxia and shallow breathing followed by prostration within 30 min. What appeared to be copious hematuria was seen approx. 40 mice after injection, and throe of the animals displayed intermittent tachypnea with pink fluid exuding from the nose . Death which occurred 3 - S hr after injection was preceded by loss of the righting reflex and by a semi-comatose state. Autopsy revealed no obvious gross Qatllological changes. It has repeatedly been shown that venons from a considerable variety of snakes, including P. colletti, contain phospholipases A that cause extensive destruction of skeletal muscle with consequent excretion of myoglobin (HARws et al., 1975 ; FOHLMAN and BAKER, 1977 ; MESS and SAMEJIMA, 1980, 1986) . In order to find out whether PLBinjected mice were excreting myoglobin (M~ 16,500) or hemoglobin (M~ 64,000) the criterion of ICELNER and ALEXANDER (1985) was applied. Urine was collected on filter paper, and the red pigment was extracted with 0.9% NaCI . The rod color of the extract readily passed through a YM-type membrane (Amicon Company, Danvers, MA) which has an M~ cutoff of 30,000 whereas an authentic specimen of hemoglobin was retained by the same membrane . Hence myoglobinuria is a conspicuous effect of intoxication not only by PLA= but also by PLB. DISCUSSION Our earlier report suggested that a specific component of P. colletti venom is responsible for the capacity of venom (a) to cause hemolytis, and (b) to produce turbidity in dilute egg yolk . On the basis of appreciable circumstantial evidence the venom component was considered to be PLB. The data contained in the present paper confirm this belief; our best preparations of purified PLB proving very potent in respect both to hemolytic activity (2400 U/mg) and to egg yolk activity (1370 U/mg). A PLB from the venom of Vipers palestinae was isolated and partially characterized by $FiILOAH et al. (1973) . The PLB from P. colletti and that from V, palestinae resemble each other in thermostability and in having a requirement for Cat+ . In addition, both enzymes seem capable of existing either as monomers (M, about 16,500) or dieters (M, about 33,000). All of these properties are common also to phospholipases A2 occurring in the venons of a great variety of snakes . To our knowledge the biological effects of PLB have not bcen previously examined . Our results indicate that the enzyme not only is lytic for erythrocytes but also is cytotoxic in relatively low concentration for cultured rhabdomyosarcoma cells. It is lethal for mice, but the i.p . LDP although not precisely determined, is relatively large; of the order of 40 pg

554

A. W. BERNHEIMER et al.

for a 20 g mouse. Mice injected with PLB develop massive myoglobinuria, an effect known to be produced by phospholipases A 2 and one resulting from their destructive action on smooth muscle . Acknowledgements - We thank DOROTHY HeNattcseN-DESTEFANO for testing PLB on rhabdomyosarcoma cells and on fibroblasts, and HANNA C. KEI YFY for measurement of release of fatty acid from phosphatidylethanolamine . REFERENCES Amps, B. N. and DuatN, D. T. (1960) The role of polyamine in the neutralization of bacteriophage deoxyribonucleic acid . J. biol. Chem . 235, 769 BeaNHetMEtt, A. W. and Avtawu, L. S. (1976) Properties of a toxin from the sea anemone Stoichactis helianthus, including specific binding to sphingomyelin. Proc. natn. Acad. Sci . U.S.A . 73, 467 . BeaNHettNea, A. W. WetNSTetN, S. A. and LtNOER, R. (1986) Isoelectric analysis of some Australian elapid snake vrnoms with special reference to phoapholipase B and hemolysis. Toxirnn 24, 841. BuoH, E. G. and DYea, W. J. (1939) A rapid method of total lipid extraction and purification. Can . J. Biochem . Physiol. 37, 911. Coweu., J. L., KtM, K.-S. and BeaNHett~a, A. W. (1978) Alteration by cereolysin of the structure of cholesterol-containing membranes. Biochim . biophys. Acts 307, 230. FetNMAN, R., Hnvnttcsmv-Des~repANO, D., TsurtMOTO, M. and Vtt.cetc, l . (1987) Tumor necrosis factor is an important mediator of tumor all killing by human monocytes. J. Immun . (in press) . Fottutwiv, J . and EAttea, D. (1977) Isolation and characterization of a lethal myotoxic phospholipase A from the vrnom of the common sea snake Enhydrina schistose causing myoglobinuria in min. Toxicon 15, 385 . HAaats, J. B., JOHNSON, M. A, and KAatssoN, E. (1975) Pathological responses of rat skeletal muscle to a single subcutaneous injection of a toxin isolated from the vrnom of the Australian tiger snake, Notxhis scutatus srutatus . Clin. exp. Pharmac. Physiol. 2, 283. Keuvea, M. J. and At .eXANDI7t, N. M. (1985) Rapid separation and identification of myoglobin and hemoglobin in urine by cattrifugation through a microconcentrator membrane . Clin . Chem . 31, 112. 1 .~RMMLI , U .K. . (1970) Cleavage of structural proteins during assembly of the head of batteriophage T4 . Natare 227. 680. MAAINE'1-I't, G. V. (1965) The action of phospholipase A on lipoproteins . Biochim . biophys. Acts 98, SS4. Mess, D. and $AMEJIMA, Y. (1980) Purification from Australian elapid vrnoms, and properties of phospholipases A which cause myoglobinurla in mice . Toxicon 18, 443. Mess, D. and $AMEJIMA, Y. (1986) Isolation and characterization of myotoxic phospholipases A, from crotalid vrnoms . Toxirnn 24, 161 . $HtLOAH, J., KLIBANSKY, C., De VAIES, A. and Beaoea, A. (1973) Phospholipase B activity of a purified phospholipase A from Vlpera palestinae vrnom. J. Lipid Res . 14, 267 . SMrrH, J. D., FtteewtwN, G., Voor, M. and Dut.secco, R. (1960) The nucleic acid of polyoma virus. Virology 12, 185. Welss, J., Bectcexnrre-Qu~ct.twTw, S. aad Etss.~cH, P. (1979) Determinants of the action of phospholipases A on the envelope phospholipids of Fsrherichia toll. J. biol. Chem . 254, 11010. WHrfAKER, J. R. and GRANUM, P. E. (1980) An absolute method for protein determination based on difference in absorbante at 235 and 280 nm . Analyt . Biochem . 109, 156.