- Email: [email protected]
Comparative study of three phospholipase A2s from the venom of Vipera aspis
Comp. Biochem. PhysioLVol. 97B, No. 3, pp. 507-514, 1990 Printed in Great Britain
0305-0491/90 $3.00 + 0.00 © 1990 Pergamon Press pie
COMPARATIVE STUDY OF THREE PHOSPHOLIPASE FROM THE VENOM OF VIPERA ASPIS
A2s
YUMIKO KOMORI, TOSHAIKI NIKAI and HISAYOSHISUGIHARA Department of Microbiology, Faculty of Pharmacy, Meijo University, Tenpaku-ku, Nagoya 468, Japan
(Received 30 April 1990) Abstract--1. Three phospholipase A2s, PLA2-I, PLA2-II and PLA2-III, were isolated from Vipera aspis venom by gel filtration and ion exchange chromatography. 2. Purified PLA2-I, -II and -III have mol. wts of 30,200, 16,000 and 13,500, and isoelectric points of 9.45, 7.65 and <4.1, respectively. 3. PLA2-I consists of an acidic subunit (mol. wt 13,700, pI: <3.5) and a basic subunit (mol. wt 16,500, pI: 10.6), which can be separated under highly acidic conditions. 4. PLA2-I possessed lethal activity and LDso for this preparation was estimated to be 0.288 (0.209-0.397)/~g/g, while lethality was not observed when PLA2-II, -III or each subunit of PLA2-I were administered. 5. Capillary permeability-increasing activity was found in the samples which possessed basic isoelectric points. Additionally, PLA2-I and its basic subunit drastically prolonged activated partial thromboplastin time of platelet rich plasma. 6. Intramuscular injections of PLA2-I, -II and -III increased serum creatine phosphokinase activity in mice, indicating that damage in muscle was caused by these enzymes. 7. NH2-terminal sequences of the three PLA2 s were compared with other phospholipase A2s from snake venoms. Furthermore, antigenicities were tested using antiserum prepared against each sample.
Bordetella pertusis were obtained from Wako Pure Chemical Industries, Osaka, Japan, and desiccated Mycobacterium butyricum and Freund's complete adjuvant were from Difco Laboratories. Other chemicals used were of analytical grade from commercial sources.
INTRODUCTION Phospholipase A2 (EC 3.1.1.4) catalyzes the specific hydrolysis of fatty acid ester bonds at the C2 position of 1,2-diacyl-phosphoglycerides. This enzyme has been isolated from various animal toxins (Iwanaga and Suzuki, 1979; Shipolini, 1984) and m a m m a l organs (de Hass et al., 1971; Meijer et al., 1978), and its properties have been investigated. Snake venoms are known to be source of phospholipase A2, and various enzymes have been isolated and studied for their structure (Botes and Viljoen, 1974; Aird et al., 1985; Joubert, 1975; Randolph and Heinrikson, 1982) and biological activities (Verheij et al., 1980; Fletcher et al., 1981; H o w g o o d and Smith, 1977; H o et al., 1984; Kouyoumdjian et al., 1986). In this report we present findings on the purification of three phospholipase A2s from the venom of Vipera aspis (Aspic viper) and discuss their biological and physiological properties. Additionally, their chemical and structural characteristics are compared with phospholipases A : s from other snake venoms. MATERIALS
Enzyme assay Phospholipase A activity was determined using phospholipid emulsions as substrate. The formed free fatty acids were measured quantitatively using an enzymatic colorimetric assay described by Simizu et al. (1980). Indirect hemolytic assay with a 4% suspension of saline-washed rabbit erythrocytes (Hendon and Fraenkel-Conrat, 1971) was also employed. Toxicity Biological assays were carried out using white mice (ddY strain 20 g). A 0.1 ml sample of the toxin solution containing different protein concentrations was used for intravenous injection. LDso determinations were carried out using eight mice/dose level with a 96 hr time interval. Intravenous injections were characterized by the method of Litchfield and Wilcoxon (1949). The effect of sample on the isolated mouse diaphragm was tested according to Kobayashi et al. (1981). The mouse hemi-diaphragm was suspended in a Krebs-Ringer bicarbonate solution (NaCI: 120 mM, KCI: 4.8mM, CaC12: 1.2raM, MgSO4: 1.3raM, KH2PO4: 1.2mM, NaHCO3: 25.2mM and glucose: 5.8mM at pH 7.4), which was aerated with 95% 02/5% CO 2. The preparation was stimulated directly with 5 msec pulses at a frequency of 0.1 Hz. Isometric contractions were measured by a force displacement transducer and recorded on a polygraph.
AND METHODS
Materials Lyophilized crude venom from Vipera aspis was purchased from Latoxan, Rosans, France. Sephadex G-75, S-Sepharose and DEAE-Sephacel were the products of Pharmacia Fine Chemicals. "Phospholipase A setTM'' and "free fatty acidsrM'' for enzymatic colorimetric assay for phospholipase A were obtained from Boehringer Mannheim Biochemica. Egg lecithin was purchased from Serdary Research Laboratories, and 1-acyl-2(N-4-nitrobenzo-2oxa-1,3-diazole) animoeaproyl phosphatidyl choline (NBDphosphatidyl choline) was from Avanti polar-lipids. CPK-UV test and inactive microorganism suspensions of
Other biological assays Capillary permeability-increasing activity was tested by .the method described by Miles and Wilhelm (1955) using white rabbits weighing 2-2.5 kg. The effects of the purified enzyme on activated partial thromboplastin time (APTT) 507
508
Y. KOMORI et al.
were determined using citrated human plasma. Various concentrations of the enzyme were incubated with 0.1 ml of platelet rich plasma at 37°C for 15 min. The clot formation was initiated by adding 0.1 ml of 0.5% kaolin and 0.1 ml of 25 mM CaC12, and the amount of time to the first appearance of fibrin strands was recorded.
Serum creatine phosphokinase (CPK) activity Purified samples (5-10/~g in 100#1 of saline) were injected intramuscularly into ddY strain mice (25g). Controls using saline-injected mice were also assayed. Blood was extracted from the ventriculus dexter cordis at various time intervals following the injections and centrifuged. The effect of each sample was examined by using five mice for each injection type. According to the method of Oliver (1955), 20-100 #1 of the serum was added to a 3.0 ml of CPK assay reagent containing creatine phosphate, ADP, glucose, hexokinase, glucose-6-phopsphate and AMP, and incubated at 30°C. The increase in the absorbance at 340 nm, which was based on the produce of NADPH, was monitored from 5 to 8 min after the addition of the serum. The quantity of creatine phosphokinase which will phosphorylate 1/~mol of creatine phosphate/min was defined to be 1 U of activity. Molecular weight and amino acid analysis The molecular weight was measured by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. Amino acid compositions of the samples were determined with a Hitachi high speed automatic amino acid analyzer, model 835. Samples were hydrolyzed with 6N HCI at ll0°C for 24, 48 and 72 hr. Sequence analysis Samples were reduced and S-pyridylethylated as described by Aketagawa et al. (1986), and amino terminal sequence analysis of the samples were performed with an Applied Biosystems model 477A sequencer. The phenylthiohydantoin derivatives of the amino acids were identified with an Applied Biosystems model 120A PTH-analyzer. Preparation of antiserum Sample solution (0.1 mg in 0.2ml saline) was mixed with 0.2 ml of Freund's complete adjuvant and injected
E tO 8.C
into a New Zealand White rabbit once a week until the antibody titer in serum was sufficiently increased. To obtain a higher titer, intradermal injection of a 0.5 ml of suspension of B. pertussis (2 x 101°) in saline was given simultaneously with the first immunization, and at the second time 5 mg of M. butyricum was mixed with the sample solution. A y-globulin fraction was prepared from the antiserum by 33.3% ammonium sulfate fractionation. RESULTS
Isolation One gram of crude venom was dissolved in 3.0ml of saline and applied to a Sephadex G-75 column (2.5 x 95 cm). The column was developed with saline at a flow rate of 12.0 ml/hr at 4°C, and fractions of 3.0 ml/tube were collected (Fig. IA). The fraction possessing indirect hemolytic activity was lyophilized and purified by a S-Sepharose column (1.5 x 45cm) equilibrated with 0.01 M T r i s - H C l buffer (pH 7.4) containing 0.01 M NaC1. Elution was performed with a linear gradient from 0.01 to 0.5 M NaC1 in a total vol of 600 ml at a flow rate of 19.6ml/hr (Fig. 1B). Indirect hemolytic activity was found in fractions 1, 3 and 4. Fractions 3 (PLA2-II) and 4 (PLA2-I) migrated single bands on polyacrylamide gel electrophoresis (pH 4.3) (photograph is shown in Fig. 2), while fraction 1 was further purified by a D E A E - S e p h a c e l column ( 1 . 5 × 4 5 c m ) equilibrated with 0 . 0 1 M T r i s - H C l buffer (pH 7.2) containing 0.01 M NaCI. Elution was performed with a linear gradient from 0.01 to 0.25 M NaC1 in a total vol of 600 ml at a flow rate of 18.9 ml/hr (Fig. IC). The fraction possessing indirect hemolytic activity (PLA2-III) was collected and examined for purity by disc gel electrophoresis (pH 8.3 and 4.3, photograph is shown in Fig. 2).
A
CO
4.0
i
o
g n
C
o .I0
o
¢q
5O 100 150 fraction number
=E
1.0
/I"
10
0.5 B
0
~®°2i
4
50 100 15O fraction number 10
0.1
O-
•
s-o loo 1sO fraction number
Fig. 1. Purification scheme of PLA2-I, -II and -III from V. aspis venom. • • , protein fractions detected by absorbance at 280 nm; . . . . . , conductivity of eluents; - - - , indirect hemolytic activity. (A) The first step using a Sephadex G-75 column. (B) The second step using a S-Sepharose column. (C) The third step of the purification for PLA2-III using a DEAE-Sephacel column.
Three phospholipase A2s from V. aspis venom
i
509
A
PLA2-111
PLA2-11
PLA,-I
:0 acidic subunit a basic subunit
E ¢,1.0
A B
0,1
C
D
0
o
t-
e.J m e,,
n
<
a
b a+b
B I
0
I
4
F
I
8
I
I
I
I
12 16
retention time (min) Fig. 2. HPLC profiles of PLA2-I, -II and -III. PLA2-I (fraction 4) and -II (fraction 3) from the S-Sepharose column (Fig. I B), and PLA2-III from the DEAE-Sephacel column (Fig. 1C) were lyophilized and applied to a Develosil 300 ODS-7 column (4.6 × 250mm) with the following eluent systems. (I) 0.1% TFA in H20, (II) 0.1% TFA in CNCH 3. A linear gradient from 30-55% in (II) over 18 min at a flow rate of 1 ml/min. Disc gel electrophoresis (pH 4.3) of each sample is shown in the figure. (A) PLA2-I, (B) PLA2-II, (C) PLA2-III, (D) mixture of PLA2-I and -II and -III.
n Chemical properties The elution profiles for reversed-phase HPLC of PLA2-I, -II and -III are shown in Fig. 2. PLA2-II and -III were eluted as a single peak, while PLA2-I was separated into two peaks. Polyacrylamide gel electrophoresis of PLA2-I and its subunits indicates that the subunit eluted precedingly from H P L C is an acidic protein which does not migrate on pH 4.3 gel [Fig. 3B(a)] but possesses a single band on a pH 8.3 gel [Fig. 3A(a)], while the other subunit is a basic protein, which revealed a single band on a pH 4.3 gel [Fig. 3B(b)]. Incubation of the acidic and basic subunits at a molar ratio of 1 : 1 in a 5 m M Tris-HC1 buffer (pH 7.2) induced a recombination of subunits, and resulted in the appearance of a band possessing the same mobility as that of native PLA2-I (Fig. 3A, B). The electrophoresis patterns on pH 2.3 gel shows that PLA2-I can be separated under highly acidic conditions [Fig. 3C(n), a + b]. SDS--polyacrylamide gel electrophoresis of the purified PLA2-I, -II and -III revealed a single band with mol. wt 13,700 for the acidic subunit and 16,500 for the basic subunit of PLA2-I, respectively, 16,000 for PLA2-II and 13,500 for PLA2-III. The isoelectric
a
b a+b
C
n
a
b
a+b
Fig. 3. Disc gel electrophoresis of PLA2-I and its subunit. Gels used were (A) pH 8.3, (B) pH 4.3, and (C) pH 2.3. Samples are: native PLA2-1 (n), acidic subunit (a), basic subunit (b), and mixture of acidic and basic subunit (a + b).
Y. KOMORIet al.
510
Table 1. Amino acid compositions of phospholipase A s from V. aspis v e n o m PLA,-I Acidic Basic subunit subunit
Amino acid
PLA2-1I
PLA2-III
16 6 7 13 11 9 4 12 1 4 5 8 4 4 2 5 0 3
18 5 11 9 14 7 5 16 1 5 6 8 8 9 2 9 1 3
18 6 11 I1 15 7 6 14 2 4 7 9 4 10 2 5 1 3
14 4 10 9 16 5 5 10 2 5 6 7 4 7 2 3 3 5
114
137
135
117
Asp Thr Ser Glu Gly Ala Val cm-Cys Met lie Leu Tyr Phe Lys His Arg Pro Trp* Total
*Tryptophan content was determined by the spectroscopic method of Edelhoch (1967).
point was estimated to be 9.45 for native PLA2-I, 7.65 for PLA2-II and <4.1 flor PLA2-III by isoelectric focusing electrophoresis. The acidic and basic subunits possessed isoelectric points of <3.5 and 10.6, respectively. The amino acid composition of each sample is shown in Table 1. Based on the mol. wts, reduced and carboxymethylated PLA2-I, -II and -III consists of 251, 135 and 117 residues.
Phospholipase A 2 activity The purified samples possessed indirect hemolytic activity on rabbit erythrocytes, which was based on the formation of lysolecithin. This indicates that the samples have phospholipase A activity, and liberate 1 mol free fatty acid from I mol substrate. 1-Acyl2(N-4- nitrobenzo- 2-oxa- 1,3- diazole) aminocaproyl
phosphatidyl choline was employed as a substrate to determine the fatty acid ester bond hydrolyzed by these enzymes. The extraction of NBD-aminocaproic acids released was performed according to the method of Dole and Meinertz (1960). The fluorescence of NBD-aminocaproic acids (2 ex: 465 nm and 2 em: 525 nm) in the extract was increased with increasing enzyme concentrations, indicating that the isolated samples catalyze the specific hydrolysis of ester bonds at the C2-position of phospholipids. The enzyme activity was tested using lecithin emulsion as a substrate, and the formed free fatty acids were measured (65.43 U/mg protein for PLA2-I, 67.03 U/mg for PLA2-II and 46.92 U/mg for PLA2-III)(Table 2). The basic subunit of PLA2-I also had phospholipase A2 activity (85.27 U/mg) however, the acidic subunit possessed only traces of it. Indirect hemolytic assay also indicates that the acidic subunit lacks the activity.
Toxicity A series of intravenous injections demonstrated that PLA2-I was more toxic than P L A : I I and -III. An LDho for the purified PLA2-I was estimated to be 0.288 (0.209-0.397) /tg/g (Table 2). However, no lethality was observed when the acidic and basic subunits were used in the assay. In the mouse diaphragm, twitch responses to direct muscle stimulation were suppressed by PLA2-I at a concentration of 10 -7 M. Figure 4 shows representative inhibition of the twitch response following treatment with 10 -6 M of the purified sample. The rates of suppression were dependent on the concentration of the toxin used. This inhibitory effect was not removed by washing the muscle with fresh medium, indicating that the action of PLA2-I is irreversible. Neither PLA2-II nor PLA2-III possessed the inhibitory effects on contractile responses to electrical stimulation.
Table 2. Comparison of three phospholipase A s from V. aspis venom
Mol. wt p! Biological activity Phospholipase A* activity (U/mg) Indirect hemolytic activity Lethal activity (LDs0) CPI activity~ Prolongation of APTT:~
PLA2-I native
Acidic subunit
Basic subunit
PLA2-II
PLA2-III
30,200 9.45
13,700 <3.5
16,500 10.6
16,000 7.65
13,500 <4.1
0.79 -
85.27 + -
67.03 + -
65.43 + + (0.288 gg/g) + + + + (30min<)
+ + + + + (1-2min) ( 3 0 m i n < )
+ + (l-2min)
46.92 +
+ (1-2 rain)
*Phospholipase A activity was determined by using an enzymatic colorimetric assay of Simizu et al. (1980). tCPI; capillary permeability increasing activity. :~APTT; activated partial thromboplastin time.
~ ' ~ " " ~ ' : ~ - - - ;
~
0 ~ --
:- - :" . . . .
~ --
--T "
!..
1
....
, ~"""~""-"-~
'. .........
~_-..-].---~
10 . . . . . .
-....
-" ~ ! ---:J~-==
7---T20
.....
I. . . . . . . . . . .
' -.
[ -2 ~£40
-;--
. :
: 60
.
~
[ . . . . . . . . . . .
-
:
i, 90"':
.
.
120
Fig. 4. Effect of PLA2-I in the isolated diaphragm. Inhibition pattern of contractile responses of the isolated mouse hemi-diaphragrn to direct electrical stimulation (0.1 Hz, 5 msec, supramaximal voltage).
Three phospholipase A2s from V. aspis venom Table 3. Effect of three phospholipase A2s on the serum creatine phosphokinase level CPK activity PLA2-1 PLA2-II PLA2-III (mU/ml) (5 pg) (10 pg) (10 pg) Time after the injection* 30 min 156.3 145.1 93.6 60 min 374.2 167.6 116.5 Values were presented as average, obtained from three experiments. The normal CPK levelsin serumweretested usingsaline-injected mice (78.0mU/ml). *Blood samples were extracted 30 or 60min following the i.m. injections.
observed when the acidic subunit of PLA2-I and PLA2-III were administered at a dose of 40 pg. The effect of samples on activated partial thromboplastin time (APTT) is shown in Table 2. With a dose of 5.0 #g, the native PLA~-I and basic subunit of PLA2-I drastically prolonged clotting time, while PLA2-II, -III and the acidic subunit of PLA2-I possessed relatively weak activity. These results indicate that biological activities of PLA2-I described above are due to the action of the basic subunit. The serum creatine phosphokinase (CPK) levels in mice following the intramuscular injections of PLA2 s were monitored and shown in Table 3. Injections of purified samples induced higher C P K levels in mice, and they reached a maximum 60 min after administration. The injections of PLA2-I induced the highest
O t h e r biological a c t i v i t y
Native PLA2-I, the basic subunit of PLA2-I and PLA2-II, increased capillary permeability in rabbits at a m i n i m u m dose of l/~g, while no activity was [ Viperidae ] Vipers aspis PLAs-I acidic subunit basic subunit PLA.-II PLAz-III
5 N L-F OFI N L F QFI N L Y OPl H L S OPl
Bitis caudalia Bitis gabonica [ Crotaridae ] Triseresurua okinavensis Crotalus adaaanteus Crotalus durissua t e r r i f i c u s Crotoxin
15 L QKY-G-
20
N D
N L T QFI D L T QFI
N N
S AMI-G-KSSLA-YA N KM--GQ-S-VFDY
H L M OFI S L V OFI
T
MK - [ A G - R S G V W W Y G KT-AK-RSGLLgYS
-
basic subunit
10
KEAVHSYA I N GKL-GAFS-VIINY F KM-TK-KSALLSYS F-GIMSY I N KK-TGI
- - H L L UFI
a c i d i c subunit
511
- i
D
L
°
.
I
.
FE-TR-KNAIPFYA
I
[ Elapidae ] Na_ia naja atra Heaacbatus besacbatus
N L Y !O F[ N N L Y 'O F] N
Q CTVP-SRS-II|DFA K CTVP-SRS-gWHFA
[ Mammals ] Porcine pancreas
A LWIO FJ
l CAIPGSHP-LMDFN
S
40 45 25 30 35 DATDRCCF' I IYGCYCGWG S !YGCYCG|G GIGTPK] DATDRCCF N I~GCYCOWG GKGI[PQ[ DATDRCCF
Y
(i) (Z) (3)
V H D C C YIG K A D K V H D C C YIG I M G T
(5)
GCYCG~G GCYCG|G
G Q G R P Q I D~'-~D R C C F V H D C C YIG l V - T GIGRPQI DATDRCCF V HDCCYGKA-T
(7) (8)
GCYCG~G GCYCG|G
GQG!|PQI
GCYCG|G YGCYCGIG GCYCG|G
S P
50
AXDCCYJ V H D C CL..~ v H D C C Y._JG VHDCC L
GgGI(PL
DATDRCCF
GI(GQPK D~TDRCCF GKGKPII DATDRCCF
I
ZG]RPXl
! D N
GCYCG~ GCYCG G
SGTPVD S G!T P V[D
N
GCYCG
sGtev~ I
A~DRC ATDRC
t
DRC
LDRC DRC
F F HDCCYAKLTG
(4) (8)
FV H D C C YIG K L A l
(9) (10)
Q V HDNCYNEAEK Q T HDNCYSDAEK
(11) (12)
E T HDNC
(13)
YJRDAKN
Fig. 5. Comparison of the NH2-terminal sequences of P1A2-I, -II and -III. The identical sequences are boxed. References for the sequences are as follows: (1)-(4) present study; (5) Viljoen et al., 1982; (6) Bates and Viljoen, 1974; (7) Joubert and Haylett, 1981; (8) Randolph and Heinrikson, 1982; (9) Aird et al., 1985; (10) Aird et al., 1986; (11) Tsai et al., 1981; (12) Joubert, 1975; (13) de Hass et al., 1970.
512
Y. KOMORIet aL
Fig. 6. Immunodiffusion of various antiserums against PLA2-I, -II and -III. Abbreviations used are: I, PLA2-I; If, PLAz-II; III, PLA2-III; C, crude venom of V. aspis; AS-I, AS-II, AS-III, AS-Ia and AS-Ib: antiserum of PLA2-1, PLA~-II, PLA2-III, acidic subunit of PLA2-I, and basic subunit of PLA2-I. (A) Effect of the antiserum of V. aspis crude venom (center well) against PLA2-I, -II and -IIl. (B) Effect of AS-I (center well) against crude venom, PLAz-I, II and III. (C) Effect of AS-II and AS-III against PLA2-II and -III. (D) Effect of AS-I, AS-Ia and AS-Ib (peripheral wells) against PLA2-I (center well). CPK levels with a half dose of PLA2-II and -III. Although the CPK levels gradually returned to normal (78.04mU/ml), it required 6-24hr. Since the serum CPK level is an important indicator of damage in muscle, the increase in the levels of venom-injected mice can be attributed to damage resulting from the intramuscular injection. Sequence analysis The amino-terminal sequences of the three phospholipase AzS from V. aspis venom were compared with other enzymes from snake venoms and porcine pancreas (Fig. 5). The acidic and basic subunits of PLA2-I, -II and -III possessed similarities, and show common residues with the enzymes from snake venoms. The comparison of PLA2-I with crotoxin from C, d. terrificus venom indicates that both the basic subunits are similar in their sequence; however, the acidic subunits are different because of a lack of NH2-terminal residues in the subunit of crotoxin. Immunodiffussion test The purified samples were investigated for antigenicities by means of the Ouchterlony double diffusion test. PLA2-I, -II and -III formed precipitin lines with the antiserum of crude venom (Fig. 6A) or antiserum of PLA2-I (Fig. 6B), which cross-reacted
each other. Antiserum of PLA2-II and -III also formed precipitin lines with PLA2-II and -III, respectively (Fig. 6C). These results indicate that the three phospholipase A2s from V. aspis venom have similarities in their structure of antigenic determinant. Since PLA2-I consists of acidic and basic subunits, it forms double precipitin lines with the antiserum of PLA2-I (Fig. 6D). DISCUSSION Phospholipase A 2 is one of the main components of snake venoms (Iwanaga and Suzuki, 1979), and exhibits a variety of potent biological effects. Indirect hemolytic activity, blockage of neuromuscular transmission, anticoagulant property, myonecrotic activity and capillary permeability increasing activity are examples of activities present in this venom enzyme (Hawgood and Smith, 1977; Rosenberg, 1979; Condrea, 1979; Kouyoumdjian et al., 1986; Rothschild and Rothschild, 1979). However, there is no correlation between these toxicological activities and in vitro enzymatic activity, the results obtained by many investigators suggest that basic phospholipase A2s are relatively toxic when compared to acidic enzymes (Rosenberg, 1979). These pharmacologic differences may be due to differing rates of phospholipid
Three phospholipase Ass from V. asp/s venom hydrolysis or to selective differences in ability to attack particular phosphlipids. In this paper we have reported that the toxicity of the strongly basic PLA2-I from V. aspis venom was greater than that of PLAs-II and -III, although these three enzymes possessed similarities in their in vitro enzymatic activities. The inhibitory effect of PLA2-I on the contractile response to direct stimulation of the diaphragm may partially account for the lethality of this enzyme. Since PLA2-I acts on the cell membrane of muscles and causes an irreversible paralysis of the diaphragm, severe respiratory embarrassment might be induced. Detailed investigations concerning the mechanisms of action by PLA2-I on the cell membrane well elucidate the method whereby this physiological effect occurs. The capillary permeability increasing activity of snake venom phospholipase A2 has been reported as being due to its histamine, or slow reacting substance (SRS), releasing activity from the mast cell membrane (Rothschild and Rothschild, 1979). It seems interesting that only the basic phospholipase A2 preparations (PLA2-I and -II from V. aspis venom) increased capillary permeability. The differences of the activity between the basic and acidic enzymes may be due to the ability of binding to the negative charged phospholipids of the mast cell membrane. The activated partial thromboplastin time of platelet rich plasma was drastically prolonged by PLA2-I, while PLA2-II and -III possessed relatively weak anticoagulant activity. Verheij et aL (1980) reported that phospholipase A2 s which exhibit strong anticoagulant activity were highly basic proteins (pI < 9); however, not all the basic phospholipase A2 s were anticoagulant. They also showed that anticoagulant basic phospholipase Ass possessed not only high affinity for the negative charged surface phospholipids, but also hydrolytic activity for procoagulant phospholipids. Since PLA2-I presented in this study possesses a high isoelectric point (9.45), it possibly penetrates into the bilayer and hydrolyzes procoagulant phospholipids to destroy. The ability of PLA2-II (pI 7.65) and -III (pI < 4.1) to bind to the negative surface is relatively weak even though they possess phospholipid hydrolyzing activity. Therefore, these enzymes induce insignificant effects. The elevation of serum CPK levels in mice following the injection of purified phospholipase A2 fractions indicated that these enzymes caused damage to muscle cells. Gopalakrishnakone et aL (1984) reported that degenerative changes were observed in muscle following injections of phospbolipase As, crotoxin, using light and electron microscopy. Since the primary site of action of crotoxin was concluded to be sarcolemmal membrane, it is conceivable that the myotoxicity of PLA2-I is due to its ability to bind muscle cell membranes. PLA2-I consists of a non-toxic acidic subunit and a relatively toxic basic subunit. The acidic subunit is devoid of enzyme and biological activities, while the basic subunit possesses all the activities observed in native PLA2-I except for the lethality. These results suggest that most of the biological activities of this enzyme are due to the basic subunit. Crotoxin from Crotalus durissus terrificus venom is one of the well studied toxic phospholipase A2s which can be separ-
513
ated into two subunits, a basic phospholipase A2 and an acidic protein with no enzyme activity (Hendon and Fraenkel-Conrat, 1971). Although the mol. wts of these crotoxin subunits (acidic: 8400, basic: 13,000) are smaller than those of PLA2-I (acidic: 13,700, basic: 16,500), these enzymes have many characteristics in common; for example, lethality only found in native toxin, myoencrotic activity, NHs-terminal sequence of basic subunit, and absence of enzymatic and biological activities in acidic subunit. However, PLA2-I from V. aspis venom is clearly different from crotoxin with respect to the NHs-terminal sequence of the acidic subunit. As shown in Fig. 5, crotoxin acidic subunit lacks 22 residues in NH2-terminal region (Aird et al., 1985), while the sequence of PLA2-I acidic subunit is similar to that of other phospholipase A2 preparations from V. aspis venom and other snake venoms. These results suggest that the absence of enzymatic and biological activities in these acidic subunits is not related to the NHs-terminal sequence. The mechanism of action of crotoxin was studied by Hendon and Tu (1979) and Radvanyi and Bon (1984). These investigators suggested that the function of the acidic subunit is to protect the positive charged basic subunit against the nonspecific binding, and increase the probability of the enzyme reaching the neuromuscular junction. Although the role of the PLA2-I acidic subunit has not been clarified, it may possess a similar function to that of the crotoxin acidic subunit. Further investigation will help to lead to a better understanding of the lethality of PLA2-I. The difference in toxicity of three phospholipase Ass from V. aspis venom might be due to the difference in the ability to bind to various cell membrane phospholipids; for example, mast cell membrane, sarcolemmal membrane, and procoagulant phospholipids. A more detailed study about the substrate preference of these enzymes will be necessasry to elucidate the mode of action. Acknowledgements--We thank K. Fukushima, S. Hirano and N. Kuwahara for their excellent technical assistance. REFERENCES
Aird S. D., Kaiser I. L, Lewis R. V. and Krugel W. G. (1985) Rattlesnake presynaptic neurotoxins; primary structure and evolutionary origin of the acidic subunit. Biochemistry/4, 7054-7058. Aird S. D., Kaiser I. I., Lewis R. V. and Krugel W. G. (1986) A complete amino acid sequence for the basic subunit of crotoxin. Archs. Biochem. Biophys. 249, 296-300. Aketagawa J., Miyata T., Ohtsubo S., Nakamura T., Morita T., Hayashida H., Miyata T., Iwanaga S., Takao T. and Shimonishi Y. (1986) Primary structure of limulus anticoagulant antilipopolysaccharide factor. J. biol. Chem. 261, 7357-7365. Botes D. P. and Viljoen C. C. (1974) Bitis gabonica venom: the amino acid sequence of phospholipase A. J. biol. Chem. 249, 3827-3835. Condrea E. (1979) Hemolytic effects of snake venoms. In Handbook of Experimental Pharmocology (Edited by Lee C. Y.), Vol. 52, pp. 448-479. Springer, New York. de Hass G. H., Bonsen P. P. M., Pieterson W. A. and van Deenen L. L. M. (1971) Studies on phospholipase A and its zymogen from porcine pancreas. III. Action of the enzyme on short-chain lecithins. Biochim. biophys. Acta 239, 252-266.
514
Y. KOMORXet al.
de Hass G. H., Slotboom A. J., Bonsen P. P. M., van Deenen L. L. M., Maroux S., Puigserver A. and Desnuelle P. (1970) Studies on phospholipase A and its zymogen from porcine pancreas. I. The complete amino acid sequence. Biochim. biophys. Acta 221, 31-53. Dole V. P. and Meinertz H. (1960) Microdetermination of long-chain fatty acids in plasma and tissues. J. biol. Chem. 235, 2595-2599. Edelhoch H. (1967) Spectroscopic determination of tryptophan and tyrosine in proteins. Biochemistry 6, 1948-1954. Fletcher J. E., Rapuano B. E., Condrea E., Yang C. C. and Rosenberg P. (1981) Relationship between catalysis and toxicological properties of three phospholipase A 2 from Elapid snake venoms. Toxicol. appl. Pharmac. 59, 375-388. Gopalakrishnakone P., Dempster D. W., Hawgood B. J. and Elder H. Y. (1984) Cellular and mitochondrial changes induced in the structure ofmurine skeletal muscle by crotoxin, a neurotoxic phospholipase A 2 complex. Toxicon 22, 85-98. Hawgood B. J. and Smith J. W. (1977) The mode of action at the mouse neuromuscular junction of the phospholipase A-crotapotin complex isolated from venom of the South American rattlesnake. Br. J. Pharmac. 61, 597-606. Hendon R. A. and Fraenkel-Conrat H. (1971) Biological roles of the two components of crotoxin. Proc. natn. Acad. Sci. USA 68, 1560-1563. Hendon R. A. and Tu A. T. (1979) The role of crotoxin subunits in tropical rattlesnake neurotoxic action. Biochem. biophys. Aeta 578, 243-252. Ho C. L., Teng C. M. and Lee C. Y. (1984) Presynaptic and musculotropic effects of a basic phospholipase A 2 from the Formosan Habu (Trimersurus mucrosquamatus) venom. Toxicon 22, 813-816. Iwanaga S. and Suzuki T. (1979) Enzyme in snake venoms. In Handbook of Experimental Pharmacology (Edited by Lee C. Y.), Vol. 52, pp. 61-158. Springer, New York. Joubert F. J. (1975) Haemachjatus haemachatus (ringhals) venom. Purification, some properties and amino-acid sequence of phospholipase A (fraction De-I). Eur. J. Biochem. 52, 539-554. Joubert F. J. and Haylett T. (1981) Purification, some properties and amino acid sequence of a phospholipase A 2 (DE-I) from Trimeresurus okinavensis (Hime-habu) venom. Hoppe-Seyler's Z. Physiol. Chem. 362, 997-1006. Kobayashi J., Ohizumi H., Nakamura H. and Hirata Y. (1981) Pharmacological study on the venom of the marine snail Conus textile. Toxicon 19, 757-762.
Kouyoumdjian J. A., Harris J. B. and Johnson M. A. (1986) Muscle necrosis caused by the sub-units of crotoxin. Toxicon 24, 575-583. Litchfield J. R. and Wilcoxon F. (1949) A simplified method of evaluating dose-effect experiments. J. Pharmac. exp. Ther. 96, 99-I 13. Meijer H., Meddens M. J. M., Dijkman R., Slotboom A. J. and de Haas G. H. (1978) Immunological studies on pancreatic phospholipase A2. J. biol. Chem. 253, 8564-8569. Miles A. A. and Wilhelm D. L. (1955) Enzyme-like globulin from serum reproducing the vascular phenomena of inflammation--I. An activable permeability factor and its inhibitor in guinea-pig serum. Br. J. exp. Path. 36, 71-81. Oliver I. T. (1955) Spectrophotometric method for the determination of creating phosphokinase and myokinase. Biochem. J. 61, 116-122. Radvanyi F. and Bon C, (1984) Investigations on the mechanism of action of crotoxin. J. Physiol., Paris 79, 327-333. Randolph A. and Heinrikson R. L. (1982) Crotalus atrox phospholipase A2: Amino acid sequence and studies on the function of the NH2-terminal region. J. biol. Chem. 257, 2155-2161. Rosenberg P. (1979) Pharmacology of phospholipase A2 from snake venoms. In Handbook of Experimental Pharmacology (Edited by Lee C. Y.), Vol. 52, pp. 403-447. Springer, New York. Rothschild A. M. and Rothschild Z. (1979) Liberation of pharmacologically active substances by snake venoms. In Handbook of Experimental Pharmacology (Edited by Lee C. Y.), Vol. 52, pp. 591-628. Springer, New York. Shipolini R. A. (1984) Biochemistry of bee venom. Handbook o f Natural Toxins (Edited by Tu A. T.), Vol. 2, pp. 49-85. Marcel Dekker, New York. Simizu S., Tani Y., Yamada H., Tabata M. and Murachi T. (1980) Enzymatic determination of serum-free fatty acids: a colorimetric method. Analyt. Biochem. 107, 193-198. Tsai I. H., Wu S. H. and Lo T. B. (1981) Complete amino acid sequence of a phospholipase A 2 from the venom of Naja naja atra (Taiwan cobra). Toxicon 19, 141-152. Verheij H. M., Boffa M. C., Rothen C., Bryckaert M. C., Verger R. and De Haas G. H. (1980) Correlation of enzymatic activity and anticoagulant properties of phospholipase A 2. Eur. J. Biochem. 112, 25-32. Vilojoen C. C., Bores D. P. and Kruger H. (1982) Isolation and amino acid sequence of caudoxin, a presynaptic acting toxic phospholipase A 2 from the venom of the homed puff adder (Bitis caudalis). Toxicon 20, 715-737.
0305-0491/90 $3.00 + 0.00 © 1990 Pergamon Press pie
COMPARATIVE STUDY OF THREE PHOSPHOLIPASE FROM THE VENOM OF VIPERA ASPIS
A2s
YUMIKO KOMORI, TOSHAIKI NIKAI and HISAYOSHISUGIHARA Department of Microbiology, Faculty of Pharmacy, Meijo University, Tenpaku-ku, Nagoya 468, Japan
(Received 30 April 1990) Abstract--1. Three phospholipase A2s, PLA2-I, PLA2-II and PLA2-III, were isolated from Vipera aspis venom by gel filtration and ion exchange chromatography. 2. Purified PLA2-I, -II and -III have mol. wts of 30,200, 16,000 and 13,500, and isoelectric points of 9.45, 7.65 and <4.1, respectively. 3. PLA2-I consists of an acidic subunit (mol. wt 13,700, pI: <3.5) and a basic subunit (mol. wt 16,500, pI: 10.6), which can be separated under highly acidic conditions. 4. PLA2-I possessed lethal activity and LDso for this preparation was estimated to be 0.288 (0.209-0.397)/~g/g, while lethality was not observed when PLA2-II, -III or each subunit of PLA2-I were administered. 5. Capillary permeability-increasing activity was found in the samples which possessed basic isoelectric points. Additionally, PLA2-I and its basic subunit drastically prolonged activated partial thromboplastin time of platelet rich plasma. 6. Intramuscular injections of PLA2-I, -II and -III increased serum creatine phosphokinase activity in mice, indicating that damage in muscle was caused by these enzymes. 7. NH2-terminal sequences of the three PLA2 s were compared with other phospholipase A2s from snake venoms. Furthermore, antigenicities were tested using antiserum prepared against each sample.
Bordetella pertusis were obtained from Wako Pure Chemical Industries, Osaka, Japan, and desiccated Mycobacterium butyricum and Freund's complete adjuvant were from Difco Laboratories. Other chemicals used were of analytical grade from commercial sources.
INTRODUCTION Phospholipase A2 (EC 3.1.1.4) catalyzes the specific hydrolysis of fatty acid ester bonds at the C2 position of 1,2-diacyl-phosphoglycerides. This enzyme has been isolated from various animal toxins (Iwanaga and Suzuki, 1979; Shipolini, 1984) and m a m m a l organs (de Hass et al., 1971; Meijer et al., 1978), and its properties have been investigated. Snake venoms are known to be source of phospholipase A2, and various enzymes have been isolated and studied for their structure (Botes and Viljoen, 1974; Aird et al., 1985; Joubert, 1975; Randolph and Heinrikson, 1982) and biological activities (Verheij et al., 1980; Fletcher et al., 1981; H o w g o o d and Smith, 1977; H o et al., 1984; Kouyoumdjian et al., 1986). In this report we present findings on the purification of three phospholipase A2s from the venom of Vipera aspis (Aspic viper) and discuss their biological and physiological properties. Additionally, their chemical and structural characteristics are compared with phospholipases A : s from other snake venoms. MATERIALS
Enzyme assay Phospholipase A activity was determined using phospholipid emulsions as substrate. The formed free fatty acids were measured quantitatively using an enzymatic colorimetric assay described by Simizu et al. (1980). Indirect hemolytic assay with a 4% suspension of saline-washed rabbit erythrocytes (Hendon and Fraenkel-Conrat, 1971) was also employed. Toxicity Biological assays were carried out using white mice (ddY strain 20 g). A 0.1 ml sample of the toxin solution containing different protein concentrations was used for intravenous injection. LDso determinations were carried out using eight mice/dose level with a 96 hr time interval. Intravenous injections were characterized by the method of Litchfield and Wilcoxon (1949). The effect of sample on the isolated mouse diaphragm was tested according to Kobayashi et al. (1981). The mouse hemi-diaphragm was suspended in a Krebs-Ringer bicarbonate solution (NaCI: 120 mM, KCI: 4.8mM, CaC12: 1.2raM, MgSO4: 1.3raM, KH2PO4: 1.2mM, NaHCO3: 25.2mM and glucose: 5.8mM at pH 7.4), which was aerated with 95% 02/5% CO 2. The preparation was stimulated directly with 5 msec pulses at a frequency of 0.1 Hz. Isometric contractions were measured by a force displacement transducer and recorded on a polygraph.
AND METHODS
Materials Lyophilized crude venom from Vipera aspis was purchased from Latoxan, Rosans, France. Sephadex G-75, S-Sepharose and DEAE-Sephacel were the products of Pharmacia Fine Chemicals. "Phospholipase A setTM'' and "free fatty acidsrM'' for enzymatic colorimetric assay for phospholipase A were obtained from Boehringer Mannheim Biochemica. Egg lecithin was purchased from Serdary Research Laboratories, and 1-acyl-2(N-4-nitrobenzo-2oxa-1,3-diazole) animoeaproyl phosphatidyl choline (NBDphosphatidyl choline) was from Avanti polar-lipids. CPK-UV test and inactive microorganism suspensions of
Other biological assays Capillary permeability-increasing activity was tested by .the method described by Miles and Wilhelm (1955) using white rabbits weighing 2-2.5 kg. The effects of the purified enzyme on activated partial thromboplastin time (APTT) 507
508
Y. KOMORI et al.
were determined using citrated human plasma. Various concentrations of the enzyme were incubated with 0.1 ml of platelet rich plasma at 37°C for 15 min. The clot formation was initiated by adding 0.1 ml of 0.5% kaolin and 0.1 ml of 25 mM CaC12, and the amount of time to the first appearance of fibrin strands was recorded.
Serum creatine phosphokinase (CPK) activity Purified samples (5-10/~g in 100#1 of saline) were injected intramuscularly into ddY strain mice (25g). Controls using saline-injected mice were also assayed. Blood was extracted from the ventriculus dexter cordis at various time intervals following the injections and centrifuged. The effect of each sample was examined by using five mice for each injection type. According to the method of Oliver (1955), 20-100 #1 of the serum was added to a 3.0 ml of CPK assay reagent containing creatine phosphate, ADP, glucose, hexokinase, glucose-6-phopsphate and AMP, and incubated at 30°C. The increase in the absorbance at 340 nm, which was based on the produce of NADPH, was monitored from 5 to 8 min after the addition of the serum. The quantity of creatine phosphokinase which will phosphorylate 1/~mol of creatine phosphate/min was defined to be 1 U of activity. Molecular weight and amino acid analysis The molecular weight was measured by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. Amino acid compositions of the samples were determined with a Hitachi high speed automatic amino acid analyzer, model 835. Samples were hydrolyzed with 6N HCI at ll0°C for 24, 48 and 72 hr. Sequence analysis Samples were reduced and S-pyridylethylated as described by Aketagawa et al. (1986), and amino terminal sequence analysis of the samples were performed with an Applied Biosystems model 477A sequencer. The phenylthiohydantoin derivatives of the amino acids were identified with an Applied Biosystems model 120A PTH-analyzer. Preparation of antiserum Sample solution (0.1 mg in 0.2ml saline) was mixed with 0.2 ml of Freund's complete adjuvant and injected
E tO 8.C
into a New Zealand White rabbit once a week until the antibody titer in serum was sufficiently increased. To obtain a higher titer, intradermal injection of a 0.5 ml of suspension of B. pertussis (2 x 101°) in saline was given simultaneously with the first immunization, and at the second time 5 mg of M. butyricum was mixed with the sample solution. A y-globulin fraction was prepared from the antiserum by 33.3% ammonium sulfate fractionation. RESULTS
Isolation One gram of crude venom was dissolved in 3.0ml of saline and applied to a Sephadex G-75 column (2.5 x 95 cm). The column was developed with saline at a flow rate of 12.0 ml/hr at 4°C, and fractions of 3.0 ml/tube were collected (Fig. IA). The fraction possessing indirect hemolytic activity was lyophilized and purified by a S-Sepharose column (1.5 x 45cm) equilibrated with 0.01 M T r i s - H C l buffer (pH 7.4) containing 0.01 M NaC1. Elution was performed with a linear gradient from 0.01 to 0.5 M NaC1 in a total vol of 600 ml at a flow rate of 19.6ml/hr (Fig. 1B). Indirect hemolytic activity was found in fractions 1, 3 and 4. Fractions 3 (PLA2-II) and 4 (PLA2-I) migrated single bands on polyacrylamide gel electrophoresis (pH 4.3) (photograph is shown in Fig. 2), while fraction 1 was further purified by a D E A E - S e p h a c e l column ( 1 . 5 × 4 5 c m ) equilibrated with 0 . 0 1 M T r i s - H C l buffer (pH 7.2) containing 0.01 M NaCI. Elution was performed with a linear gradient from 0.01 to 0.25 M NaC1 in a total vol of 600 ml at a flow rate of 18.9 ml/hr (Fig. IC). The fraction possessing indirect hemolytic activity (PLA2-III) was collected and examined for purity by disc gel electrophoresis (pH 8.3 and 4.3, photograph is shown in Fig. 2).
A
CO
4.0
i
o
g n
C
o .I0
o
¢q
5O 100 150 fraction number
=E
1.0
/I"
10
0.5 B
0
~®°2i
4
50 100 15O fraction number 10
0.1
O-
•
s-o loo 1sO fraction number
Fig. 1. Purification scheme of PLA2-I, -II and -III from V. aspis venom. • • , protein fractions detected by absorbance at 280 nm; . . . . . , conductivity of eluents; - - - , indirect hemolytic activity. (A) The first step using a Sephadex G-75 column. (B) The second step using a S-Sepharose column. (C) The third step of the purification for PLA2-III using a DEAE-Sephacel column.
Three phospholipase A2s from V. aspis venom
i
509
A
PLA2-111
PLA2-11
PLA,-I
:0 acidic subunit a basic subunit
E ¢,1.0
A B
0,1
C
D
0
o
t-
e.J m e,,
n
<
a
b a+b
B I
0
I
4
F
I
8
I
I
I
I
12 16
retention time (min) Fig. 2. HPLC profiles of PLA2-I, -II and -III. PLA2-I (fraction 4) and -II (fraction 3) from the S-Sepharose column (Fig. I B), and PLA2-III from the DEAE-Sephacel column (Fig. 1C) were lyophilized and applied to a Develosil 300 ODS-7 column (4.6 × 250mm) with the following eluent systems. (I) 0.1% TFA in H20, (II) 0.1% TFA in CNCH 3. A linear gradient from 30-55% in (II) over 18 min at a flow rate of 1 ml/min. Disc gel electrophoresis (pH 4.3) of each sample is shown in the figure. (A) PLA2-I, (B) PLA2-II, (C) PLA2-III, (D) mixture of PLA2-I and -II and -III.
n Chemical properties The elution profiles for reversed-phase HPLC of PLA2-I, -II and -III are shown in Fig. 2. PLA2-II and -III were eluted as a single peak, while PLA2-I was separated into two peaks. Polyacrylamide gel electrophoresis of PLA2-I and its subunits indicates that the subunit eluted precedingly from H P L C is an acidic protein which does not migrate on pH 4.3 gel [Fig. 3B(a)] but possesses a single band on a pH 8.3 gel [Fig. 3A(a)], while the other subunit is a basic protein, which revealed a single band on a pH 4.3 gel [Fig. 3B(b)]. Incubation of the acidic and basic subunits at a molar ratio of 1 : 1 in a 5 m M Tris-HC1 buffer (pH 7.2) induced a recombination of subunits, and resulted in the appearance of a band possessing the same mobility as that of native PLA2-I (Fig. 3A, B). The electrophoresis patterns on pH 2.3 gel shows that PLA2-I can be separated under highly acidic conditions [Fig. 3C(n), a + b]. SDS--polyacrylamide gel electrophoresis of the purified PLA2-I, -II and -III revealed a single band with mol. wt 13,700 for the acidic subunit and 16,500 for the basic subunit of PLA2-I, respectively, 16,000 for PLA2-II and 13,500 for PLA2-III. The isoelectric
a
b a+b
C
n
a
b
a+b
Fig. 3. Disc gel electrophoresis of PLA2-I and its subunit. Gels used were (A) pH 8.3, (B) pH 4.3, and (C) pH 2.3. Samples are: native PLA2-1 (n), acidic subunit (a), basic subunit (b), and mixture of acidic and basic subunit (a + b).
Y. KOMORIet al.
510
Table 1. Amino acid compositions of phospholipase A s from V. aspis v e n o m PLA,-I Acidic Basic subunit subunit
Amino acid
PLA2-1I
PLA2-III
16 6 7 13 11 9 4 12 1 4 5 8 4 4 2 5 0 3
18 5 11 9 14 7 5 16 1 5 6 8 8 9 2 9 1 3
18 6 11 I1 15 7 6 14 2 4 7 9 4 10 2 5 1 3
14 4 10 9 16 5 5 10 2 5 6 7 4 7 2 3 3 5
114
137
135
117
Asp Thr Ser Glu Gly Ala Val cm-Cys Met lie Leu Tyr Phe Lys His Arg Pro Trp* Total
*Tryptophan content was determined by the spectroscopic method of Edelhoch (1967).
point was estimated to be 9.45 for native PLA2-I, 7.65 for PLA2-II and <4.1 flor PLA2-III by isoelectric focusing electrophoresis. The acidic and basic subunits possessed isoelectric points of <3.5 and 10.6, respectively. The amino acid composition of each sample is shown in Table 1. Based on the mol. wts, reduced and carboxymethylated PLA2-I, -II and -III consists of 251, 135 and 117 residues.
Phospholipase A 2 activity The purified samples possessed indirect hemolytic activity on rabbit erythrocytes, which was based on the formation of lysolecithin. This indicates that the samples have phospholipase A activity, and liberate 1 mol free fatty acid from I mol substrate. 1-Acyl2(N-4- nitrobenzo- 2-oxa- 1,3- diazole) aminocaproyl
phosphatidyl choline was employed as a substrate to determine the fatty acid ester bond hydrolyzed by these enzymes. The extraction of NBD-aminocaproic acids released was performed according to the method of Dole and Meinertz (1960). The fluorescence of NBD-aminocaproic acids (2 ex: 465 nm and 2 em: 525 nm) in the extract was increased with increasing enzyme concentrations, indicating that the isolated samples catalyze the specific hydrolysis of ester bonds at the C2-position of phospholipids. The enzyme activity was tested using lecithin emulsion as a substrate, and the formed free fatty acids were measured (65.43 U/mg protein for PLA2-I, 67.03 U/mg for PLA2-II and 46.92 U/mg for PLA2-III)(Table 2). The basic subunit of PLA2-I also had phospholipase A2 activity (85.27 U/mg) however, the acidic subunit possessed only traces of it. Indirect hemolytic assay also indicates that the acidic subunit lacks the activity.
Toxicity A series of intravenous injections demonstrated that PLA2-I was more toxic than P L A : I I and -III. An LDho for the purified PLA2-I was estimated to be 0.288 (0.209-0.397) /tg/g (Table 2). However, no lethality was observed when the acidic and basic subunits were used in the assay. In the mouse diaphragm, twitch responses to direct muscle stimulation were suppressed by PLA2-I at a concentration of 10 -7 M. Figure 4 shows representative inhibition of the twitch response following treatment with 10 -6 M of the purified sample. The rates of suppression were dependent on the concentration of the toxin used. This inhibitory effect was not removed by washing the muscle with fresh medium, indicating that the action of PLA2-I is irreversible. Neither PLA2-II nor PLA2-III possessed the inhibitory effects on contractile responses to electrical stimulation.
Table 2. Comparison of three phospholipase A s from V. aspis venom
Mol. wt p! Biological activity Phospholipase A* activity (U/mg) Indirect hemolytic activity Lethal activity (LDs0) CPI activity~ Prolongation of APTT:~
PLA2-I native
Acidic subunit
Basic subunit
PLA2-II
PLA2-III
30,200 9.45
13,700 <3.5
16,500 10.6
16,000 7.65
13,500 <4.1
0.79 -
85.27 + -
67.03 + -
65.43 + + (0.288 gg/g) + + + + (30min<)
+ + + + + (1-2min) ( 3 0 m i n < )
+ + (l-2min)
46.92 +
+ (1-2 rain)
*Phospholipase A activity was determined by using an enzymatic colorimetric assay of Simizu et al. (1980). tCPI; capillary permeability increasing activity. :~APTT; activated partial thromboplastin time.
~ ' ~ " " ~ ' : ~ - - - ;
~
0 ~ --
:- - :" . . . .
~ --
--T "
!..
1
....
, ~"""~""-"-~
'. .........
~_-..-].---~
10 . . . . . .
-....
-" ~ ! ---:J~-==
7---T20
.....
I. . . . . . . . . . .
' -.
[ -2 ~£40
-;--
. :
: 60
.
~
[ . . . . . . . . . . .
-
:
i, 90"':
.
.
120
Fig. 4. Effect of PLA2-I in the isolated diaphragm. Inhibition pattern of contractile responses of the isolated mouse hemi-diaphragrn to direct electrical stimulation (0.1 Hz, 5 msec, supramaximal voltage).
Three phospholipase A2s from V. aspis venom Table 3. Effect of three phospholipase A2s on the serum creatine phosphokinase level CPK activity PLA2-1 PLA2-II PLA2-III (mU/ml) (5 pg) (10 pg) (10 pg) Time after the injection* 30 min 156.3 145.1 93.6 60 min 374.2 167.6 116.5 Values were presented as average, obtained from three experiments. The normal CPK levelsin serumweretested usingsaline-injected mice (78.0mU/ml). *Blood samples were extracted 30 or 60min following the i.m. injections.
observed when the acidic subunit of PLA2-I and PLA2-III were administered at a dose of 40 pg. The effect of samples on activated partial thromboplastin time (APTT) is shown in Table 2. With a dose of 5.0 #g, the native PLA~-I and basic subunit of PLA2-I drastically prolonged clotting time, while PLA2-II, -III and the acidic subunit of PLA2-I possessed relatively weak activity. These results indicate that biological activities of PLA2-I described above are due to the action of the basic subunit. The serum creatine phosphokinase (CPK) levels in mice following the intramuscular injections of PLA2 s were monitored and shown in Table 3. Injections of purified samples induced higher C P K levels in mice, and they reached a maximum 60 min after administration. The injections of PLA2-I induced the highest
O t h e r biological a c t i v i t y
Native PLA2-I, the basic subunit of PLA2-I and PLA2-II, increased capillary permeability in rabbits at a m i n i m u m dose of l/~g, while no activity was [ Viperidae ] Vipers aspis PLAs-I acidic subunit basic subunit PLA.-II PLAz-III
5 N L-F OFI N L F QFI N L Y OPl H L S OPl
Bitis caudalia Bitis gabonica [ Crotaridae ] Triseresurua okinavensis Crotalus adaaanteus Crotalus durissua t e r r i f i c u s Crotoxin
15 L QKY-G-
20
N D
N L T QFI D L T QFI
N N
S AMI-G-KSSLA-YA N KM--GQ-S-VFDY
H L M OFI S L V OFI
T
MK - [ A G - R S G V W W Y G KT-AK-RSGLLgYS
-
basic subunit
10
KEAVHSYA I N GKL-GAFS-VIINY F KM-TK-KSALLSYS F-GIMSY I N KK-TGI
- - H L L UFI
a c i d i c subunit
511
- i
D
L
°
.
I
.
FE-TR-KNAIPFYA
I
[ Elapidae ] Na_ia naja atra Heaacbatus besacbatus
N L Y !O F[ N N L Y 'O F] N
Q CTVP-SRS-II|DFA K CTVP-SRS-gWHFA
[ Mammals ] Porcine pancreas
A LWIO FJ
l CAIPGSHP-LMDFN
S
40 45 25 30 35 DATDRCCF' I IYGCYCGWG S !YGCYCG|G GIGTPK] DATDRCCF N I~GCYCOWG GKGI[PQ[ DATDRCCF
Y
(i) (Z) (3)
V H D C C YIG K A D K V H D C C YIG I M G T
(5)
GCYCG~G GCYCG|G
G Q G R P Q I D~'-~D R C C F V H D C C YIG l V - T GIGRPQI DATDRCCF V HDCCYGKA-T
(7) (8)
GCYCG~G GCYCG|G
GQG!|PQI
GCYCG|G YGCYCGIG GCYCG|G
S P
50
AXDCCYJ V H D C CL..~ v H D C C Y._JG VHDCC L
GgGI(PL
DATDRCCF
GI(GQPK D~TDRCCF GKGKPII DATDRCCF
I
ZG]RPXl
! D N
GCYCG~ GCYCG G
SGTPVD S G!T P V[D
N
GCYCG
sGtev~ I
A~DRC ATDRC
t
DRC
LDRC DRC
F F HDCCYAKLTG
(4) (8)
FV H D C C YIG K L A l
(9) (10)
Q V HDNCYNEAEK Q T HDNCYSDAEK
(11) (12)
E T HDNC
(13)
YJRDAKN
Fig. 5. Comparison of the NH2-terminal sequences of P1A2-I, -II and -III. The identical sequences are boxed. References for the sequences are as follows: (1)-(4) present study; (5) Viljoen et al., 1982; (6) Bates and Viljoen, 1974; (7) Joubert and Haylett, 1981; (8) Randolph and Heinrikson, 1982; (9) Aird et al., 1985; (10) Aird et al., 1986; (11) Tsai et al., 1981; (12) Joubert, 1975; (13) de Hass et al., 1970.
512
Y. KOMORIet aL
Fig. 6. Immunodiffusion of various antiserums against PLA2-I, -II and -III. Abbreviations used are: I, PLA2-I; If, PLAz-II; III, PLA2-III; C, crude venom of V. aspis; AS-I, AS-II, AS-III, AS-Ia and AS-Ib: antiserum of PLA2-1, PLA~-II, PLA2-III, acidic subunit of PLA2-I, and basic subunit of PLA2-I. (A) Effect of the antiserum of V. aspis crude venom (center well) against PLA2-I, -II and -IIl. (B) Effect of AS-I (center well) against crude venom, PLAz-I, II and III. (C) Effect of AS-II and AS-III against PLA2-II and -III. (D) Effect of AS-I, AS-Ia and AS-Ib (peripheral wells) against PLA2-I (center well). CPK levels with a half dose of PLA2-II and -III. Although the CPK levels gradually returned to normal (78.04mU/ml), it required 6-24hr. Since the serum CPK level is an important indicator of damage in muscle, the increase in the levels of venom-injected mice can be attributed to damage resulting from the intramuscular injection. Sequence analysis The amino-terminal sequences of the three phospholipase AzS from V. aspis venom were compared with other enzymes from snake venoms and porcine pancreas (Fig. 5). The acidic and basic subunits of PLA2-I, -II and -III possessed similarities, and show common residues with the enzymes from snake venoms. The comparison of PLA2-I with crotoxin from C, d. terrificus venom indicates that both the basic subunits are similar in their sequence; however, the acidic subunits are different because of a lack of NH2-terminal residues in the subunit of crotoxin. Immunodiffussion test The purified samples were investigated for antigenicities by means of the Ouchterlony double diffusion test. PLA2-I, -II and -III formed precipitin lines with the antiserum of crude venom (Fig. 6A) or antiserum of PLA2-I (Fig. 6B), which cross-reacted
each other. Antiserum of PLA2-II and -III also formed precipitin lines with PLA2-II and -III, respectively (Fig. 6C). These results indicate that the three phospholipase A2s from V. aspis venom have similarities in their structure of antigenic determinant. Since PLA2-I consists of acidic and basic subunits, it forms double precipitin lines with the antiserum of PLA2-I (Fig. 6D). DISCUSSION Phospholipase A 2 is one of the main components of snake venoms (Iwanaga and Suzuki, 1979), and exhibits a variety of potent biological effects. Indirect hemolytic activity, blockage of neuromuscular transmission, anticoagulant property, myonecrotic activity and capillary permeability increasing activity are examples of activities present in this venom enzyme (Hawgood and Smith, 1977; Rosenberg, 1979; Condrea, 1979; Kouyoumdjian et al., 1986; Rothschild and Rothschild, 1979). However, there is no correlation between these toxicological activities and in vitro enzymatic activity, the results obtained by many investigators suggest that basic phospholipase A2s are relatively toxic when compared to acidic enzymes (Rosenberg, 1979). These pharmacologic differences may be due to differing rates of phospholipid
Three phospholipase Ass from V. asp/s venom hydrolysis or to selective differences in ability to attack particular phosphlipids. In this paper we have reported that the toxicity of the strongly basic PLA2-I from V. aspis venom was greater than that of PLAs-II and -III, although these three enzymes possessed similarities in their in vitro enzymatic activities. The inhibitory effect of PLA2-I on the contractile response to direct stimulation of the diaphragm may partially account for the lethality of this enzyme. Since PLA2-I acts on the cell membrane of muscles and causes an irreversible paralysis of the diaphragm, severe respiratory embarrassment might be induced. Detailed investigations concerning the mechanisms of action by PLA2-I on the cell membrane well elucidate the method whereby this physiological effect occurs. The capillary permeability increasing activity of snake venom phospholipase A2 has been reported as being due to its histamine, or slow reacting substance (SRS), releasing activity from the mast cell membrane (Rothschild and Rothschild, 1979). It seems interesting that only the basic phospholipase A2 preparations (PLA2-I and -II from V. aspis venom) increased capillary permeability. The differences of the activity between the basic and acidic enzymes may be due to the ability of binding to the negative charged phospholipids of the mast cell membrane. The activated partial thromboplastin time of platelet rich plasma was drastically prolonged by PLA2-I, while PLA2-II and -III possessed relatively weak anticoagulant activity. Verheij et aL (1980) reported that phospholipase A2 s which exhibit strong anticoagulant activity were highly basic proteins (pI < 9); however, not all the basic phospholipase A2 s were anticoagulant. They also showed that anticoagulant basic phospholipase Ass possessed not only high affinity for the negative charged surface phospholipids, but also hydrolytic activity for procoagulant phospholipids. Since PLA2-I presented in this study possesses a high isoelectric point (9.45), it possibly penetrates into the bilayer and hydrolyzes procoagulant phospholipids to destroy. The ability of PLA2-II (pI 7.65) and -III (pI < 4.1) to bind to the negative surface is relatively weak even though they possess phospholipid hydrolyzing activity. Therefore, these enzymes induce insignificant effects. The elevation of serum CPK levels in mice following the injection of purified phospholipase A2 fractions indicated that these enzymes caused damage to muscle cells. Gopalakrishnakone et aL (1984) reported that degenerative changes were observed in muscle following injections of phospbolipase As, crotoxin, using light and electron microscopy. Since the primary site of action of crotoxin was concluded to be sarcolemmal membrane, it is conceivable that the myotoxicity of PLA2-I is due to its ability to bind muscle cell membranes. PLA2-I consists of a non-toxic acidic subunit and a relatively toxic basic subunit. The acidic subunit is devoid of enzyme and biological activities, while the basic subunit possesses all the activities observed in native PLA2-I except for the lethality. These results suggest that most of the biological activities of this enzyme are due to the basic subunit. Crotoxin from Crotalus durissus terrificus venom is one of the well studied toxic phospholipase A2s which can be separ-
513
ated into two subunits, a basic phospholipase A2 and an acidic protein with no enzyme activity (Hendon and Fraenkel-Conrat, 1971). Although the mol. wts of these crotoxin subunits (acidic: 8400, basic: 13,000) are smaller than those of PLA2-I (acidic: 13,700, basic: 16,500), these enzymes have many characteristics in common; for example, lethality only found in native toxin, myoencrotic activity, NHs-terminal sequence of basic subunit, and absence of enzymatic and biological activities in acidic subunit. However, PLA2-I from V. aspis venom is clearly different from crotoxin with respect to the NHs-terminal sequence of the acidic subunit. As shown in Fig. 5, crotoxin acidic subunit lacks 22 residues in NH2-terminal region (Aird et al., 1985), while the sequence of PLA2-I acidic subunit is similar to that of other phospholipase A2 preparations from V. aspis venom and other snake venoms. These results suggest that the absence of enzymatic and biological activities in these acidic subunits is not related to the NHs-terminal sequence. The mechanism of action of crotoxin was studied by Hendon and Tu (1979) and Radvanyi and Bon (1984). These investigators suggested that the function of the acidic subunit is to protect the positive charged basic subunit against the nonspecific binding, and increase the probability of the enzyme reaching the neuromuscular junction. Although the role of the PLA2-I acidic subunit has not been clarified, it may possess a similar function to that of the crotoxin acidic subunit. Further investigation will help to lead to a better understanding of the lethality of PLA2-I. The difference in toxicity of three phospholipase Ass from V. aspis venom might be due to the difference in the ability to bind to various cell membrane phospholipids; for example, mast cell membrane, sarcolemmal membrane, and procoagulant phospholipids. A more detailed study about the substrate preference of these enzymes will be necessasry to elucidate the mode of action. Acknowledgements--We thank K. Fukushima, S. Hirano and N. Kuwahara for their excellent technical assistance. REFERENCES
Aird S. D., Kaiser I. L, Lewis R. V. and Krugel W. G. (1985) Rattlesnake presynaptic neurotoxins; primary structure and evolutionary origin of the acidic subunit. Biochemistry/4, 7054-7058. Aird S. D., Kaiser I. I., Lewis R. V. and Krugel W. G. (1986) A complete amino acid sequence for the basic subunit of crotoxin. Archs. Biochem. Biophys. 249, 296-300. Aketagawa J., Miyata T., Ohtsubo S., Nakamura T., Morita T., Hayashida H., Miyata T., Iwanaga S., Takao T. and Shimonishi Y. (1986) Primary structure of limulus anticoagulant antilipopolysaccharide factor. J. biol. Chem. 261, 7357-7365. Botes D. P. and Viljoen C. C. (1974) Bitis gabonica venom: the amino acid sequence of phospholipase A. J. biol. Chem. 249, 3827-3835. Condrea E. (1979) Hemolytic effects of snake venoms. In Handbook of Experimental Pharmocology (Edited by Lee C. Y.), Vol. 52, pp. 448-479. Springer, New York. de Hass G. H., Bonsen P. P. M., Pieterson W. A. and van Deenen L. L. M. (1971) Studies on phospholipase A and its zymogen from porcine pancreas. III. Action of the enzyme on short-chain lecithins. Biochim. biophys. Acta 239, 252-266.
514
Y. KOMORXet al.
de Hass G. H., Slotboom A. J., Bonsen P. P. M., van Deenen L. L. M., Maroux S., Puigserver A. and Desnuelle P. (1970) Studies on phospholipase A and its zymogen from porcine pancreas. I. The complete amino acid sequence. Biochim. biophys. Acta 221, 31-53. Dole V. P. and Meinertz H. (1960) Microdetermination of long-chain fatty acids in plasma and tissues. J. biol. Chem. 235, 2595-2599. Edelhoch H. (1967) Spectroscopic determination of tryptophan and tyrosine in proteins. Biochemistry 6, 1948-1954. Fletcher J. E., Rapuano B. E., Condrea E., Yang C. C. and Rosenberg P. (1981) Relationship between catalysis and toxicological properties of three phospholipase A 2 from Elapid snake venoms. Toxicol. appl. Pharmac. 59, 375-388. Gopalakrishnakone P., Dempster D. W., Hawgood B. J. and Elder H. Y. (1984) Cellular and mitochondrial changes induced in the structure ofmurine skeletal muscle by crotoxin, a neurotoxic phospholipase A 2 complex. Toxicon 22, 85-98. Hawgood B. J. and Smith J. W. (1977) The mode of action at the mouse neuromuscular junction of the phospholipase A-crotapotin complex isolated from venom of the South American rattlesnake. Br. J. Pharmac. 61, 597-606. Hendon R. A. and Fraenkel-Conrat H. (1971) Biological roles of the two components of crotoxin. Proc. natn. Acad. Sci. USA 68, 1560-1563. Hendon R. A. and Tu A. T. (1979) The role of crotoxin subunits in tropical rattlesnake neurotoxic action. Biochem. biophys. Aeta 578, 243-252. Ho C. L., Teng C. M. and Lee C. Y. (1984) Presynaptic and musculotropic effects of a basic phospholipase A 2 from the Formosan Habu (Trimersurus mucrosquamatus) venom. Toxicon 22, 813-816. Iwanaga S. and Suzuki T. (1979) Enzyme in snake venoms. In Handbook of Experimental Pharmacology (Edited by Lee C. Y.), Vol. 52, pp. 61-158. Springer, New York. Joubert F. J. (1975) Haemachjatus haemachatus (ringhals) venom. Purification, some properties and amino-acid sequence of phospholipase A (fraction De-I). Eur. J. Biochem. 52, 539-554. Joubert F. J. and Haylett T. (1981) Purification, some properties and amino acid sequence of a phospholipase A 2 (DE-I) from Trimeresurus okinavensis (Hime-habu) venom. Hoppe-Seyler's Z. Physiol. Chem. 362, 997-1006. Kobayashi J., Ohizumi H., Nakamura H. and Hirata Y. (1981) Pharmacological study on the venom of the marine snail Conus textile. Toxicon 19, 757-762.
Kouyoumdjian J. A., Harris J. B. and Johnson M. A. (1986) Muscle necrosis caused by the sub-units of crotoxin. Toxicon 24, 575-583. Litchfield J. R. and Wilcoxon F. (1949) A simplified method of evaluating dose-effect experiments. J. Pharmac. exp. Ther. 96, 99-I 13. Meijer H., Meddens M. J. M., Dijkman R., Slotboom A. J. and de Haas G. H. (1978) Immunological studies on pancreatic phospholipase A2. J. biol. Chem. 253, 8564-8569. Miles A. A. and Wilhelm D. L. (1955) Enzyme-like globulin from serum reproducing the vascular phenomena of inflammation--I. An activable permeability factor and its inhibitor in guinea-pig serum. Br. J. exp. Path. 36, 71-81. Oliver I. T. (1955) Spectrophotometric method for the determination of creating phosphokinase and myokinase. Biochem. J. 61, 116-122. Radvanyi F. and Bon C, (1984) Investigations on the mechanism of action of crotoxin. J. Physiol., Paris 79, 327-333. Randolph A. and Heinrikson R. L. (1982) Crotalus atrox phospholipase A2: Amino acid sequence and studies on the function of the NH2-terminal region. J. biol. Chem. 257, 2155-2161. Rosenberg P. (1979) Pharmacology of phospholipase A2 from snake venoms. In Handbook of Experimental Pharmacology (Edited by Lee C. Y.), Vol. 52, pp. 403-447. Springer, New York. Rothschild A. M. and Rothschild Z. (1979) Liberation of pharmacologically active substances by snake venoms. In Handbook of Experimental Pharmacology (Edited by Lee C. Y.), Vol. 52, pp. 591-628. Springer, New York. Shipolini R. A. (1984) Biochemistry of bee venom. Handbook o f Natural Toxins (Edited by Tu A. T.), Vol. 2, pp. 49-85. Marcel Dekker, New York. Simizu S., Tani Y., Yamada H., Tabata M. and Murachi T. (1980) Enzymatic determination of serum-free fatty acids: a colorimetric method. Analyt. Biochem. 107, 193-198. Tsai I. H., Wu S. H. and Lo T. B. (1981) Complete amino acid sequence of a phospholipase A 2 from the venom of Naja naja atra (Taiwan cobra). Toxicon 19, 141-152. Verheij H. M., Boffa M. C., Rothen C., Bryckaert M. C., Verger R. and De Haas G. H. (1980) Correlation of enzymatic activity and anticoagulant properties of phospholipase A 2. Eur. J. Biochem. 112, 25-32. Vilojoen C. C., Bores D. P. and Kruger H. (1982) Isolation and amino acid sequence of caudoxin, a presynaptic acting toxic phospholipase A 2 from the venom of the homed puff adder (Bitis caudalis). Toxicon 20, 715-737.