Characterization of a metallo-proteinase from Bothrops Asper (terciopelo) snake venom

No . 7, Printed in Great Britain . Toxinon, Vol . 25.

0041-0101/67 53 .00+ .00 O 1987 PerNown Journah Ltd .

pp. 739-766, 1987 .

CHARACTERIZATION OF A METALLO-PROTEINASE FROM BOTHROPS ASPER (TERCIOPELO) SNAKE VENOM F. ARAGON-ORTIZ 1 and F. GUBEN§EK2" 'Department of Biochemistry, School of Medicine, University of Costa Rica, and =Department of Biochemistry, J. Stefan Institute, Ljubljana, Yugoslavia (Accepted for publication 12 February 1987)

F. ARAGON-OR77z and F. GuaENSEx . Characterization of a metallo-proteinase from Bothrops asper(terciopelo) snake venom. Toxicon 25, 759 - 766, 1987 .-Metalloproteinase from the venom of Bothrops asper (proteinase G) is a glycoprotein with 194 neutral hexose and 3.5 moles of sialic add per mole of protein. It hydrolyses a number of protein substrates such as casein, hemoglobin, gelatin and fibrinogen, whose alpha chain is degraded preferentially. The pH optimum of hydrolysis of casein is approximately 9 .5 . The protease is devoid of hemorraghic, esterolytic and anridolytic activities. Theproteolytic activity of the enzyme increases by about 20% in the presence of 0.2 mM Ca'* and Mg'* . Among the other ions tested, only Cd'* and Fe'* markedly decreased its activity. EDTA and cysteine are also strong inhibitors . In the presence of Ca'* and EDTA, Zn=* ions restored 50% of the activity. The amino acid composition shows fewer acidic residues than in related proteinases from other snake venoms .

INTRODUCTION THE vENom of Bothrops asper,

known in Costa Rica as terciopelo, contains a number of proteolytic enzymes . Enzymes have been characterized belonging either to serine-type proteinases with coagulant activity (ORTIZ and GUBENSEK, 1976) or to endoproteinases devoid of coagulant activity (ARAGON et al., 1978). Recently, the effects of the venom on blood coagulation were reported (BARRANTES et al., 1985). Metallo-proteinases are frequently found in snake venoms. Those isolated from Crotalidae and Viperidae venoms were identified either as hemorrhagins or endoproteases (IWANAGA et al., 1976; TAKAHASHI and OHSAKA, 1970; KuRECxi et al., 1978; KURECKI and KRESS, 1985 ; ASsAKuRA et al., (1985) . We reported previously (ARAGON et al., 1978) on the purification and some properties of two fibrinolytic enzymes from the venom of Bothrops asper. In this paper we further characterize the proteinase G, a metallo-proteinase which may play an important role in the fibrinolytic effects of Bothrops asper venom. MATERIALS AND METHODS The venom was kindly supplied by Instituto Clodomiro Picado, University of Costa Rica . It consists of pooled lyophilized samples obtained from Bothrops riper snakes originating from the Atlantic zone . Proteinase G was purified as described earlier (ARAGON et al., 1978). Tosyl-arginine methyl ester HCl(TAME), tosyl-lysine methyl ester HCI (TLME), benzoyl-arginine ethyl ester HCl (BABE), benzoyl-DL-arginine-beta-naphythylamide HCl (BANA), N-t-butoxycarbonyl-L-alanine p-nitrophenyl ester, soybean trypsin inhibitor (Kunitz), orcinol and carbohydrate standards were from Sigma Chemical Co ., St. Louis, MO (U .S .A .) . Proteins used as molecular weight standards were from Serva, Heidelberg F.R .G. Trig-HCI, 2-thiobarbituric add, casein, hemoglobin, *To whom reprint request should be addressed. 759

760

F. ARAGON-ORTIZ ând F. GUBENSEK

fibrin and gelatin were from Merck, Darmstadt (F .R .G .) . Meta sodium arsenite, meta sodium periodate and EDTA were from Kemika, Zagreb (Yugoslavia) . Trasylol (BPTI) was from Bayer-Pharma, Ljubljana (Yugoslavia) . The gel electrophoresis apparatus, GE-2/4 LS, was from Pharmacia Fine Chemicals . Analytical isoelectric focusing was performed on a Desaga Desaphor . Amino acid analysis was performed on a Beckman 118 CL amino acid analyzer . Polyacrylamide gel electrophoresis in Tris-glycine, pH 4.4, was performed according to the method of DAvis (1964) . Polyacrylamide gel electrophoresis in the presence of SDS was performed according to the method of LAEMMLI (1970) in a 4-30% (w/v) gradient gel. Analytical isoelectric focusing was on a 5% polyacrylamide plate with carrier ampholines of pH 3 =10. Proteolytic activity was measured by the method of ANSON (1939), using 1% casein, 2% hemoglobin and 1% fibrinogen . The substrates were dissolved in the universal buffer of Johnson and Linsday of pH 6 - 11 . The hydrolysis of 1 % gelatin was studied in the same pH range. After incubating the enzyme (0.5 mg/ml) for 90 min at 37°C, the reaction was monitored with ninhydrin . Fibrinolytic activity was tested on fibrin plates prepared according to the method of AsrRUP (1952), using 0.3 01o plasminogen-free fibrinogen in veronal buffer (pH 7.5) and bovine thrombin . Fibrinogenolytic activity was determined by incubating fibrinogen (10 mg/ml in 0.05 M Tris, pH 8.5) with 20 pg/ml of the enzyme at 37°C for various time intervals. At 15, 30, 60 and 120 min, 100 pl of theincubated solution were removed, lyophilized and dissolved in 1 nil of 8 M urea in the same buffer . Aliquots of 30 ;d were boiled in a water bath in the presence of 10% beta mercaptoethanol and 1 % SDS. Electrophoresis was then carried out. Esterolytic activity was measured according to the method of HtnmmEL (1959), using TAME, TLME, BAEE, BTEE and BAPNA. The hydrolysis of benzoyl DL-arginine-beta-naphthylamide (BANA) and leucine-beta-naphthylamide HCl (LeuNA) was determined according to the method of BARRE T (1972) . Hydrolysis of N-t-butoxycarbonyl-L-alanine pnitrophenyl ester was carried out according to the method of VisER and BLouT (1972) . Hemorrhagic activity was measured according to the method of Kom)o (1960) . Sialic acid content was quantified by the method of WARREN (1959) . Protein bound hecose was determined according to the procedure of WINZLER (1%1), using galactose and mannose in a 1:1 ratio as a standard . The absorbante coefficient of the enzyme was determined after exhaustive dialysis against pyridine-acetic acid (pH 5.0) and subsequent lyophilvation. Weighed amounts of the enzyme were then dissolved in 0.1 M Tris (pH 7.0) and the optical density recorded at 280 nm . A JobinYvon dichrographe III was used for the measurement of circular dichroism spectra. The apparatus was calibrated with a solution of epiandrosterone in dioxane. Protein was dissolved in 0.1 M Tris - HCl (pH 7.0) at a concentration of 1 mg/ml. Cells of 1 .0, 0.2 and 0.05 cm path length were used for the measurements . The mean residue ellipticity was calculated using an average residue molecular weight of 110. The secondary structure was estimated according to the procedure of CHEN et al., (1974) . The effect of protein inhibitors and chelating ions was studied by mixing the proteinase G (35 pg in a final volume of 2 ml) with : EDTA (1 .0 to 0.01 mM); cysteine (5 .0 to 0.05 mM); trasylol (6 .7 pg/ml) ; soybean trypsin inhibitor (150 pg/ml) . Both protein inhibitors were pre-incubated for 10 min at room temperature prior to determination of enzymatic activity . Separately, the crude venom (50 Mg) was mixed with EDTA (50 - 0.05 mM). The enzyme was added to EDTA - divalent ions (0 .067 and 0.33 mM) and to cysteine - Ca" (0 .067 and 0.33 mM), mixed and pre-incubated for 10 min before the Zn" (0 .25 or 2.5 mM) was added. The effect of cations and anions was studied by incubating the enzyme with the following ions (0.2 or 10 mM): Ca=', Mg'., Ni'', C0'', Fe", Fe", Al", Hg'., CN - and S' -. All cations were present as chlorides, anions were sodium salts. The incubation was carried out for 30 min at room temperature . Subsequently the proteolytic activity was measured against 1 % casein at pH 8.1 . RESULTS

Polyacrylamide gel electophoresis at pH 4.5 of purified proteinase G is shown in Fig. 1 . It has a molecular weight of 18,000 as estimated by SDS electroporesis and the specific activity toward casein is 0.83 OD750/mg/min (ARAGON et al., 1978). The enzyme is devoid of esterolytic and amidolytic activities towards all small molecular weight substrates mentioned in the previous section. The optimal pH values for proteolysis of casein, hemoglobin, fibrinogen and gelatin were 9.5, 9.6, 9.2 and 8.0, respectively (Fig. 2). The fibrinolytic activity is direct (not mediated by plasmin), as judged by the Astrup fibrin plate method using plasminogen-free fibrinogen (ASTRUP and MUELERTZ, 1952). Fibrinogenolytic activity is shown in Fig. 3 . It was observed that the enzyme quickly degrades the alpha chain of fibrinogen, followed by stepwise degradation of the beta and gamma chains upon prolonged incubation . The effect of different ions on the activity of proteinase G is shown in Table 1 . The inhibition by chelating agents and protein inhibitors is shown in Table 2. The isoelectric point of the enzyme, determined by isoelectric

76 1

Metallo- Proteinase from B. aspen Venom

FIG . 1 . ACRYLAMIDE GEL ELECTROPHORESIS OF PROTEINASE G AT

Cathode is at the bottom .

pH 4.4 .

TABLE 1 . THE EFFECT OF IONS ON THÉ PROTEOLYTIC ACFIVrrY OF PROTEINASE G

Compound Ca"

Concentration (mM) Relative activity (%) 80 10 122 0.2 64 Zn2 . 10 0.2 108 0.2 78 Al'` 0.2 97 Cu=' Co" 0.2 81 Hg=' 0.2 42 Cd=' 0.2 17 Fe=` 0.2 83 Fe* 0.2 4 99 Mi. 0.2 Mn=' 0.2 93 0.2 94 S2. CN20.2 106 Tests were performed in two independent experiments in parallel . The 100% value corresponds to 0.410 Ao units. All determinations were within t 0.010 A, units, which corresponds to approximately t 2% in relative activity. All tests were performed with 35 pg of proteinase G in 2 ml of i % casein at pH 8.1 . Pre-incubation of enzyme with potential inhibitors (10 min) and incubation with the substrate (30 min) was performed at 25°C .

v

762

F. ARAGON-0RTIZ and F. GUBENSEK

0 .15

z

0.10

a m

z m a

0 .05

PH FIG . 2. OPTIMAL pH FOR THE DEGRADATION OF VARIOUS PROTEIN SUSTRATES BY PROTEINASE G . (A) casein, (B) gelatin, (C) hemoglobin and (D) fibrinogen. Measurements were performed in

parallel . Absorbancy was measured at 750 nm for casein, hemoglobin and fibrinogen and at 570 rim for gelatin.

TABLE

2.

THE EFFECT OF CHELATING AGENTS AND PROTEIN INHIBITORS ON THE PROTEOLYTIC ACTIVITY OF PROTEINASE G AND OF CRUDE VENOM

Compound added (A) Proteinase G EDTA Cysteine Trasylol (BPTI) Soybean trypsin inhibitor (Kunitz) EDTA/Ca=' EDTA/Cu=' EDTA/Zn2 ' EDTA/Ca2' Zn2' Zn" Cysteine/Ca" Zn=' Zn" (B) Bothrops EDTA

riper

venom

Concentration (mM)

Relative activity (%)

1 0.1 0.01 5 0.5 0.05 6.7 Ng/ml 150 pg/nil

4 9 88 0 0 96 108 112

0.067/0.33 0.067/0.33 0.067/0.33 0.067/0.33 2.5 0.25 0.067/0.33 2.5 0.25

4 36 76 50 2 20 19

0 50 11 5 13 0.5 0.05 100 All tests were performed twice in parallel . All determinations were within f 0.010 A D units, which corresponds to approximately t 2% in relative values shown. For additional details see footnote to Table 1 .

Metallo-Proteinase from B. riper Venom

76 3

*w N*Mi

FIG . 3 . FIBRINOOENOLYIC ACTIVITY OF PROTEINASE G .

Protein standards, (2) reduced fibrinogen, (3)-(6) fibrinogen incubated with proteinase G for 15, 30, 60 and 120 min, respectively, at 37°C . Anode is at the bottom . Fibrinogen (10mg/ml) was incubated for the indicated periods with 20 mg/ml of proteinase G in 0.05 M Tris pH 8.5 .

(1)

focussing, is approximately pH 7.1 . The amino acid composition and the carbohydrate content are shown in Table 3. The circular dichroic spectrum (Fig. 4) reveals 24% of alpha and 30% of beta structure. Absorbance of a 0.1 % solution of the enzyme was 0.946 at 280 run. DISCUSSION

Proteinase G exhibits many of the proteolytic properties of the crude venom, namely an alkaline pH of hydrolysis of the protein substrates and an inhibition of activity by chelating agents such as EDTA. In contrast to the crude venom, proteinase G does not display any coagulant activity nor any detectable hemorrhagic activity . Our enzyme differs from Moojeni proteinase A, which exhibited 50% of the hemorrhagic activity of the crude venom (ASSAKURA et al., 1985). The difference from trypsin-like proteinases

76 4

F. ARAGON-ORTIZ and F. GUBEN§EK TABLE 3. AMINO Amino acid

ACID COMPOSITION AND CARBOHYDRATE CONTENT OF PROTEINASE

Calculated 14 .3 8.1 10 .5 12 .8 6.2 11 .4 6.1 1 .7 11 .8 3.8 9.1 17 .4 6.8 5.7 10 .9 12 .2 3.6 12 .5

Asx Thr Ser Glx Pro Gly Ala Cys Val Met Ile Leu Tyr Phe His Lys Trp Arg

Total

Residues per mole

G

Integer 14 8 11 13 6 11 6 2 12 4 9 17 7 6 11 12 4 13 166

Neutral carbohydrates Sialic acid

1ß'o 3.5 moles/mole

was demonstrated by the inability of protein trypsin inhibitors to inhibit proteinase G. The enzyme shows fibrinogenolytic activity, degrading preferentially the alpha chain of fibrinogen, and may thus be considered to be primarily an alpha fibrinogenase. A similar alpha fibrinogenase has also been purified from Trimeresurus mucrosquamatus venom (OUYANG and TENG, 1975).

0 ö E v N E -4000

20

-40

-8000 210 FIG .

4.

230

250 A[nmI

270

CIRCULAR DICHROIC SPECTRUM OF PROTEINASE

290

310

G IN 0.1 M Tris pH 7.0 .

Metallo-Proteinase from B . riper Venom

765

It was observed that certain divalent cations such as Fe" and Cd2 + are more efficient inhibitors of the enzyme than trivalent cations . Cat' and Zn2+ ions seem to stabilize the native structure of the enzyme, while at higher concentrations they reduce its activity . From inhibition studies (Table 2) it can be inferred that the enzyme is a metallo-proteinase which requires an optimal concentration of Cat' for its stability. The enzyme also seems to require a Zn2 + ion for its activity . This is inferred from the fact that EDTA-Ca21 inhibits its caseinolytic activity, which can be partially restored by an approximately 40fold excess of Zn2+ over EDTA . Since the enzyme is inhibited by EDTA alone and this inhibition could not be reversed by any of the cations studied, it means that in the presence of EDTA the enzyme undergoes an irreversible structural change . Ca" probably stabilizes the active conformation of the proteinase. The role of the Zn2 + ion is presumably related to the catalytic mechanism, as suggested by SPICKERMAN et al. (1973) for the metallo-proteinase from Agkistrodon piscivores leucostoma venom. The nature of the inhibition by cysteine may be explained by its chelating properties. Only 20% of the inhibitory effect of cysteine - Cat' was restored with Zn2' . In this regard the enzyme is very similar to other neutral proteinases from snake venoms (VAN DER WALT, 1972; SPICKERMAN et al., 1973 ; ASSAKuRA et al., 1985 ; KURECKI and KRESS, 1981). Recent purifications and characterizations of metallo-proteinases from snake venoms (AssAKuRA et al., 1985; KURECKI and KRESS, 1985), as well as inhibition studies with soybean trypsin inhibitor and DFP, show that in addition to proteases, esterolytic enzymes without proteolytic activity are also present in these venoms. Comparison of the amino acid composition of proteinase G with similar snake venom proteinases (KuRECKI and KRESS, 1985 ; AssAKURA et al., 1985) shows that enzyme G contains fewer aspartyl and glutamyl residues . The low content of cysteine in proteinase G seems to be a typical characteristic of most of these metallo-proteinases . Bothrops asper proteinase G appears to be one of the venom enzymes that plays a significant role in envenomenation of humans by degrading, among other plasma proteins, fibrinogen in particular and presumably also other clotting factors (BARRANTES et al., 1985). Due to its properties of being preferentially an alpha fibrinogenase with no detectable hemorraghic activity, this metallo-proteinase is valuable for understanding the action of fibrinolytic enzymes and may also be of use in the treatment of venous thrombosis . Acknowledgements - This work was supported by the Research Community of Slovenia, by CONICIT and

Vicerrectoria de Investigation, University, Costa Rica, and by the U.S. - Yugoslav Joint Fund for Scientific and Technological Cooperation under grant no . YOR 80/018 . REFERENCES

ANSON, M. L. (1939) Estimation of pepsin, trypsin, papain and cathepsin with hemoglobin . J. gen . Physiol. 22, 79 . ARAGON, F., KOPITAR, M., BABNIK, J. and GuBENtEK, F. (1978) Some properties of two fibrinolytic enzymes from the venom of B . asper. Period. Biol. 8 0 (Suppl . 1), 91 . AssAKuRA, M. T., RErcHL, A. P., ASPERTI, M. C. A. and MANDELBAum, F. R. (1985) Isolation of the major proteolytic enzyme from the venom of the snake Bothrops moojeni (caissaca) . Toxicon 23, 691. Asraup, T. and MuLERTz, S. (1952) The fibrin plate method for the estimation of fibrinolytic activity. Archs Biochem . Biophys . 40, 346 . BARRANTES, A., Sows, V. and Bot.ARos, R. (1985) Alteration de los mecanismos de la coagulation en et envenenamiento por Bothrops asper (terciopelo) . Toxicon 23, 399. BARRerr, A. J . (1972) A new assay for cathepsin BI and other thiol proteinases. Analyt . Biochem . 47, 280. CHEN, Y., YANG, Y. T. and CHAD, K. H. (1974) Determination of the helix and beta-form of proteins in aqueous solution by circular dichrioism . Biochemistry 13, 3350 . DAvrs, B. J. (1964) Disc electrophoresis 2. Method and application to human serum proteins . Ann . N. Y. Acad. Scl. 121, 404.

76 6

F. ARAGON-ORTIZ and F. GUBENSEK

HUMMEL, B. C. W. (1959) A modified spectrophotometric determination of chymotrypsin, trypsin and thrombin . Can . J. Biochem . Physiol. 37, 1393 . IwANAOA, S., OstttmA, G. and SuzuKi, T. (1976) Proteinases from the venom of Agkistrodon halys blomhojfri . In: Methods in Enzymology, Vol . XLV, p. 459 (LORAND, L., Ed .) . New York : Academic Press. KONDO, H., KoNDO, S., IKEZAWA, H., MURATA, R. and OKASAKA, A. (1960) Studies on the quantitative method for determination of hemorrhagic activity of habu snake venom. Jpn . J. med. Sci. Biol . 13, 43 . KUREcKI, T. and KRESS, L. F. (1985) Purification and partial characterization of the hemorrhagic factor from the venom of Crotales adamanteus (eastern diamondback rattlesnake) . Toxicon 23, 657. KuREcia, T., LAsowsxt, M. . SR and KRESS, L. F. (1978) Purification and some properties of two proteinases from Crotalus adamanteus venom that inactivate human alpha-proteinase inhibitors . J. biol. Chem . 253, 8340 . LAEMmLi, U. K. (1970) Cleavage of structural proteins during the assembly of the head of . bacteriophage T4 . Nature 277, 680. ORTIz, F. A. and GuBEAEK, F. (1976) Isolation and some properties of blood clotting enzyme from the venom of Bothrops asps . Bull. Inst . Pasteur 74, 145 . OUYANG, C. and TNN, C.-M. (1976) Fibrinolytic enzymes of Trimeresurus mucrosquarnatus venom. Biochim .

biophys. Acta 420, 298 .

SrtcxewmAN, A. M., FREz)REtcxs, K. K., WAoNER, F. W. and PREscow, J . M . (1973) Leucostoma peptidase A: a metallo-protease from snake venom. Blochim . biophys. Acta 293, 464. TAKAHASHI, T. and Ott&AKA, A. (1970) Purification and characterization of a proteinase in the venom of Trimeresurus Jkvoviridis . Biochim. biophys. Acta 198, 293. VAN DER WALT, S . J . and JOUBERT, F. J . (1972) Studies on puff adder (Bitis arietans) venom 11 . Toxicon 10, 341. VISER, L. and BLouT, E. R. (1972) The use of p-nitrophenyl N-ten-butyloxycarbonyl-L-alaninate as substrate for elastase . Biochim. biophys. Acta 268, 257. WARREN, L. (1959) The thiobarbituric add assay of sialic acids. J. biol. Chem . 234, 1971 . WINzLER, R. J . (1961) Determination of sentm glyeoproteins. In : Methods of Biochemical Analysis, Vol . 2, p . 290 (GLIcx, D., Ed .) . New York : Interscience .