PROTEIN
EXPRESSION
AND
PURIFICATION
1, 33-39
(1990)
HPLC-Based Two-Step Purification of Fibrinolytic Enzymes from the Venom of Agkistrodon contortrix con tortrix and Agkistrodon piscivorus conan ti’ Anastassios D. Retzios and Francis S. Markland’ Department 1303 North
Received
of Biochemistry, University of Southern California School of Medicine, Mission Road, CRL 104, Los Angeles, California 90033
January
2, 1990,
and
in revised
form
April
24, 1990
In investigations aimed at characterizing snake venom blood clot-dissolving enzymes, we have developed a rapid two-step high-performance chromatography method for the isolation of these fibrinolytic enzymes from the venoms of Agkistrodon contortrix contortrix and Agkistrodon piscivorus conanti. The first step consisted of hydrophobic interaction chromatography on a propyl-aspartamide column. Fractions containing the fibrinolytic activity were then concentrated and applied to a hydroxylapatite column. The resulting preparation, assessed for purity by reverse-phase chromatography and sodium dodecyl sulfate-polyacrylamide gel electrophoresis, was homogeneous. The molecular weight of both venom fibrinolytic enzymes was approximately 23,000 and amino acid analysis, immunological cross-reaction, cyanogen bromide, and tryptic digestion indicate a significant degree of structural similarity. However, the general proteolytic activity of the A. p. conanti venom enzyme was significantly lower than the corresponding activity of the A. c. contortrix venom, whereas their fibrinolytic activities were quite similar. OlSSOAcademic Press,Inc.
A number of enzymes that influence blood coagulation have been isolated from various snake venoms (1). These enzymes can either promote or inhibit coagulation. Fibrinolytic activities have received special attention because of their possible therapeutic role in dissolution of blood clots (2). Fibrinolytic agents have been identified in the venom of several Agkistrodon species (contortrix andpiscivorus) (3,4). A fibrinolytic enzyme from the venom of Agkistroi This work has been supported in part by NIH Grant from the USPHS and by Cortech, Inc., Denver, CO. * To whom correspondence should be addressed.
HL
31389
don contortrix contortrix (southern copperhead snake) has been purified by conventional liquid chromatography methods (5) and by isoelectric focusing (6) and characterized in this laboratory (7). This enzyme is a zinc metalloproteinase (one mole zinc per mole enzyme) with a molecular weight of 23,000. It readily cleaves the Ao-chain of fibrin and fibrinogen between Lys413 and Leu414 without activating or degrading plasminogen and protein C (8). In the present study we focused on developing a rapid method, basedon high-performance liquidchromatography, for the purification of fibrinolytic enzymes from the venoms of A. c. contortrix and the related species, A. p. conanti (Florida cottonmouth). Additionally, we investigated and compared the enzymes structurally, immunologically, and functionally.
MATERIALS
AND
METHODS
Materials. HPLC-grade solvents and chemicals were purchased from VWR Scientific, Cerritos, California; HPLC-grade trifluoroacetic acid was purchased from Pierce, Redford, Tennessee. Venoms were obtained from Biotoxins, Inc., St. Cloud, Florida. Plasminogen-free human fibrinogen (Kabi Diagnostica, grade L) was obtained from Helena Laboratories, Beaumont, Texas. Standard plasmin was obtained from The World Health Organization, International Laboratory for Biological Standards and Control, London, United Kingdom. L-(Tosylamido 2-phenyl)ethyl chloromethyl ketone (TPCK)3-treated trypsin was obtained from 3 Abbreviations used: AC&b, fihrlnolytic enzyme from the venom of Agkistrodon contortrix contortrti; Apcjib, fibrinolytic enzyme from the venom of Agkistrodon piscivorus conanti; TPCK, L-(tosylamido 2phenyl)ethyl chloromethyl ketone; SDS, sodium dodecyl sulfate; Tris, 2-amino-2-hydroxymethylpropane-1,3-diol; TBST, 0.02 M Tris buffer, pH 8.0, containing 0.14 M NaCl and 0.05% Tween 20; BSA, bovine serum albumin; DTT, dithiothreitol; TFA, trifluoroacetic
34
RETZIOS
AND
MARKLAND
TABLE
Summary
Step number
of Purification
Purification
1 2 3
Crude After After
venom hydrophobic hydroxylapatite
Note. activity
Crude After After In both of these
venom hydrophobic hydroxylapatite
cases hemorrhagic enzymes results
Agkistrodon
contortrir
contortrix
pisciuorus
enzymes that display limited in an underestimate of overall
venom
(a) and Apcfib
AC@
Total activity (PIU) fibrinolytic
conanti
venom
9.8 6.5 5.7 fibrinolytic
Worthington Biochemical Corp., Freehold, New Jersey. Reagents for immunoblotting were purchased from Sigma Chemical Co., St. Louis, Missouri. All other reagents were of analytical grade. Instruments. Purification of the fibrinolytic enzymes was performed on a Perkin-Elmer 410 Bio HPLC system equipped with an LC-95 variable-wavelength detector and an LC-100 integrator. Reverse-phase chromatography for the evaluation of purity of preparations was performed with a Beckman 334 dual-pump HPLC system equipped with a Hitachi 100-40 variable-wavelength detector and a Shimadzu CRlA integrator. Reverse-phase chromatography of cyanogen bromide and tryptic digests was performed on a Spectra-Physics 8800 HPLC system equipped with Spectra-Physics 8450 variable-wavelength detector and a Spectra-Physics WINner data acquisition module. Columns. For hydrophobic interaction chromatography, a propyl-aspartamide column (9.4 X 250 mm) from PolyLC, Columbus, Ohio, was employed. Hydroxylapatite chromatography was performed on a HPHT-BioGel column (7.2 X 100 mm) purchased from Bio-Rad, Richmond, California. Reverse-phase chromatography for the determination of preparation purity was performed on a C, 214TP54 column (4.1 X 250 mm) from Vydac, Hesperia, California. Amino acid analysis. Amino acid analyses were kindly performed by Dr. Anne Randolph, Chiron Corp., Emeryville, California. Amino acid compositions were obtained by the method described by Bidlingmeyer etal. (9). Cyanogen bromide digestion. The fibrinolytic enzymes purified from both snake venoms (3 mg) were to alkaline phosphatase; TCA, trichloroacetic acid.
capacity
Specific activity (PIU/mg)
Activity recovery (o/o)
0.052 0.46 0.71
100 66 58
0.062 0.31 0.5
100 81.5 65
enzyme
148 24 12 fibrinolytic yields.
(b)
enzyme
187 14 8
interaction chromatography chromatography
acid, IgG-AP, immunoglobulin complexed PAGE, polyacrylamide gel electrophoresis;
Enzymes Total protein (md
interaction chromatography chromatography (b)
1 2 3
of the Fibrinolytic
step
(a) Agkistrodon
1
9.2 7.5 6 are removed
at step
2. The
exclusion
of the
fibrinolytic
dialyzed against H,O, freeze-dried, and dissolved in 2 ml of 0.2 M Tris buffer, pH 8.2, containing 8 M urea. The enzymes were reduced and carboxymethylated as follows: 1 mg of dithiothreitol (DTT) was added to each solution (3.2 mM DTT solution) and the mixture was incubated at 37°C for 1 h; 2 mg of iodoacetamide was then added to each tube (5.4 mM iodoacetamide solution); the solutions were gassed with oxygen-free nitrogen; and the tubes were sealed and incubated for 1 h in the dark at 37°C. This procedure was repeated once more to complete the reaction. Following carboxymethylation the proteins were desalted by Sephadex G-15 chromatography and freeze-dried. The dried material was dissolved in 70% formic acid, and cyanogen bromide equal to the weight of the dried protein was added. The reaction was allowed to proceed with continuous stirring, under oxygen-free nitrogen and in the dark for 48 h. After 48 h, more cyanogen bromide was added (half of the dried protein weight) and the reaction was continued for an additional 12 h. Following cleavage, 100 ~1 from the digest of each fibrinolytic enzyme was removed and added to 900 ~1 of HPLC-grade water. The resulting sample was then added to a Vydac 218TP54 C,, reversephase chromatography column (4.1 X 250 mm) and the fragments were eluted according to the following protocol: (a) 5 min isocratically at 90% solvent A (0.1% TFA in water), 10% solvent B (0.1% TFA in 80% acetonitrile); (b) a 40-min linear gradient to 65% B. Tryptic digestion. Dried carboxymethylated fibrinolytic enzymes from both venoms were dissolved in 0.2 M Tris buffer, pH 8.7, containing 2 M urea. TPCK-treated trypsin was added at a ratio of 1:40 (trypsin:fibrinolytic enzyme, w/w). Digestion was allowed to proceed at room temperature, with continuous stirring for 24 h. Following digestion, 100 ~1 of each digest was added to a Vydac 218TP54 C,, column (4.1 X 250 mm) and fragments were eluted according to the following protocol:
_.:
HPLC PURIFICATION OF FIBRINOLYTIC ENZYMES
35
1 ~ I
-----F-l Al
A3
II
I
m---J 10 ?r-T--rI7
10
30 50 Fraction number
70
10
20 30
50
a
10
Time
40
3040
50
(min)
-r
___-Fraction number
Bl
10
2o
50
10
--.T-l
30 50 Fraction number
20
I
3040
I
1
/ 10
I 2o
I I 3040
(min)
/ a0
1
,I 10
1
k.
B3
Time - n
baseline
I
,r
A-. , 3040
10
-
2o
ao40
i
50
i.-,.-1
70
10 20 30 40 Fraction number
FIG. 1. Purification of the fibrinolytic enzymes from Agkistrodon contort& contortrix venom and Agkistrodon piscivorus conanti venom. Al and A2 show elution profiles for A. c. contortrix and A. p. conunti venoms, respectively, during hydrophobic interaction chromatography. Bl and B2 show elution profiles for A. c. contort& and A. p. conanti venoms, respectively, during hydroxylapatite chromatography. In all profiles solid lines indicate absorption at 280 nm whereas dashed lines indicate fibrinolytic activity.
(a) 5 min isocratically at 90% solvent A (0.1% TFA in water), 10% solvent B (0.1% TFA in 80% acetonitrile); (b) an 85min linear gradient to 70% solvent B. Western immunoblotting. Western blotting was performed using a polyclonal antibody prepared against the purified fibrinolytic enzyme from A. c. contortrix venom (10). Both Accfib and A&b were applied to a SDS polyacrylamide gel. After electrophoresis the proteins were transferred from the SDS polyacrylamide gel to a nitrocellulose membrane. Blotting was achieved by applying 300 mA overnight in 0.025 M Tris, pH 8.3, containing 0.192 M glycine, 0.1% SDS, and 40% methanol, according to the method of Towbin et al. (11). Half of the membrane, containing duplicate samples, was stained with a 0.05% solution of amido black to locate protein bands and then destained in 40% ethanol-5% acetic acid. Before a coating with the primary antibody, the other half of the membrane (not stained with amido black) was submerged in Tris-saline buffer (0.02 M Tris,
FIG. 2. Reverse-phase chromatography of crude venoms and fibrinolytic pools during purification of the fibrinolytic enzymes from Agk~trodon contortrix contort& venom (Acefib) and Agk~t~o~ pisciVOFUS conanti venom (Apcfiib). Al, crude A. c. contortrix venom; Bl, crude A. p. conanti venom; A2, AC&& following hydrophobic interaction chromatography; B2, Apefib following hydrophobic interaction chromatography; A3, Accfib following hydroxylapatite chromatography (baseline profile is shown underneath); B3, Apcfib following hydroxylapatite chromatography. On the latter profiles small peaks can be seen following the main protein peak. These are baseline artifacts and not real contaminants.
0.14 M NaCl), pH 8.0, containing 0.05% Tween 20 (TBST), and then incubated for 30 min with TBST containing 1% bovine serum albumin (BSA) to saturate the excess protein binding sites. The BSA-containing solution was then replaced with TBST containing antibody to fibrinolytic enzyme from A. c. contort& (1:lOOO dilution) and incubated for 30 min. The membrane was then washed with TBST five times for 10 min to remove unbound antibody, transferred to TBST containing anti-rabbit IgG-AP (alkaline phosphatase conjugate, 1:2000 dilution), and incubated for 30 min. After a coating by the second antibody, the membrane was washed with TBST five times for 10 min each time. Following the wash the membrane was transferred to the color development solution (66 ~1 of nitroblue tetrazolium in 10 ml of 0.1 M Tris, 0.1 M NaCl, 5 mM MgCl,, pH 9.5). After mixing, 33 ~1 of 5-bromo-4-chloro-3-indolyl phosphate (50 mglml in dimethylformamide) was added. The development of the color reaction was complete
36
RETZIOS
AND
MARKLAND TABLE
2
Amino Acid Analysis of the Fibrinolytic Enzymes from Agkistrodon contortrix contortrix Venom (Accfib) and Agkistrodon piscivorus conanti Venom (Apcfib) Fibrinolytic Amino acid
FIG. 3. SDS-polyacrylamide gel electrophoresis of purified Accfib. Lanes 1 and 2 contain Accfib (41 and 20.5 Fg, respectively) that was incubated with reducing SDS-PAGE sample buffer at 37°C for 5 min. Lanes 3 and 4 contain Accfib (41 and 20.5 pg, respectively) that was incubated with reducing SDS-PAGE sample buffer at 100°C for 10 min. Lanes 5 and 6 contain Accfib (41 and 20.5 pg, respectively) that was incubated in 17 mM EDTA for 10 min before addition of reducing SDS-PAGE and incubation at 100°C. Treatment with EDTA and boiling eliminate the lower molecular bands which are due to autolysis during SDS-PAGE sample preparation. Apcfib also displays autolysis, but upon EDTA treatment only a single band at 23,000 MW is evident, in accordance with HPLC-reverse-phase chromatography results.
within 10 min. Purple staining indicated the location where protein antigen interacted with the antibody. This was SDS-polyacrylamide gel electrophoresis. carried out according to the method of Laemmli (12). Isoelectric focusing. This was performed in a Pharmacia FBE 3000 flatbed electrophoresis apparatus, employing a constant power supply and using precast gels (Pharmacia-LKB, Piscataway, NJ) and Pharmalyte ampholyte solutions (Pharmacia-LKB). This was assessed by azocasein Proteolytic activity. hydrolysis. Azocasein was synthesized from cy-casein and diazotized sulfanilamide as described by Charney and Tomarelli (13). To measure proteolytic activity, 0.75 ml of the azocasein solution (about 50 mg/ml) was added to a 1.5-ml microcentrifuge tube. Test solution (50 ~1) was added, and the tube was incubated at 37°C for 30 min. The reaction was stopped by the addition of 0.75 ml of 1.16 M perchloric acid, and the precipitate was removed by centrifugation. Hydrolysis of azocasein was measured as increased absorbance at 390 nm in the supernatant, using as reference a blank treated in an identical manner. To accurately estimate Accfib and Apcfib specific activity (in absorbance units at 390 nm/mg enzyme), various amounts of the enzyme were added in order to construct a response curve. The linear part of the curve was used for the estimation of specific proteolytic activity. Fibrinolytic activity. Fibrinolytic activity was measured with the fibrin-plate-clearance assay described by
CYS Asx Glx Ser GUY His Arg Thr Ala Pro Tyr Val Met Ile Leu Phe LYS Trp
Accfib 6 31 20 13 14 10 9 13 11 6 5 12 12 22 7 8 ND
enzyme &Mb 3 35 21 14 14 9 8 14 12 6 6 14 6 11 22 7 9 ND
Note. The molecular weight assumed was 23,000. This estimate was derived from SDS-polyacrylamide gel electrophoresis. The results shown represent the average of two 24-h hydrolyses of each fibrinolytic enzyme (ND, not determined).
Bajwa et al. (14). For the preparation of fibrin plates two buffers were necessary: fibrinogen buffer and gelatin buffer. Fibrinogen buffer was prepared by adding 20.62 g of sodium barbital, 100 ml of 0.1 M HCl, 100 ml of “salt solution” (4.89 g CaCl, * 2H,O, 2.79 g MgCl, * 6H,O, 109.12 g NaCl, and H,O to a total volume of 1 liter), and 1350 ml of H,O. The pH of the buffer was adjusted to 7.75 with concentrated HCl and the volume was brought up to 2 liters with H,O. The gelatin buffer was prepared
FIG. 4. Western blotting of Accfb and Apcfib. The procedure is described under Materials and Methods. As can be seen, there is significant interaction of Apcfib with the antibody to Accfib.
HPLC
0.00
10
PURIFICATION
OF
30
4L
30
40
Timez~min)
f0 Lo 1.00 % : 20.50 i
22
j&L+!LJ!I 0.00
10
TimJrnin)
FIG. 5. Cyanogen bromide cleavage of the fibrinolytic enzymes from Agkistrodon contortriz contortrix venom (top) and Agkistrodon piscivorus conanti venom (bottom). The products of the digestion were separated by reverse-phase chromatography on a Vydac Cls column (218TP54) as described under Materials and Methods.
by adding 5.85 g NaCl, 10.31 g sodium barbital, 200 ml of 0.1 N HCl, 25 g gelatin, and 725 ml H,O. After the preparation of the buffers the following solutions were made: (a) 32 mg fibrinogen (76.2 g of dry weight of KABI plasminogen-free fibrinogen) was dissolved in 5.25 ml of fibrinogen buffer and 0.75 ml of H,O; (b) 20 ,ul of 1000 units/ml bovine thrombin (USP Thrombostat, from Parke-Davis, Morris Plains, NJ) was added to 0.5 ml of gelatin buffer. Fibrin plates were readied in disposable petri dishes (100 X 15 mm) by adding, in each one, 200 ~1 of thrombin solution and 6 ml of fibrinogen solution. The plates were incubated for 2-3 h at 37°C and eight equidistant holes, approximately 1 cm from the edge of the plate, were punched in each one. Ten microliters of sample was added per hole and the plates were then incubated for 18 h at 37°C. For accurate measurements of the areas of lysis, at the end of the incubation period the plates were flooded with 10% TCA solution. The diameters of the lysis areas were then measured against a dark background. As with the proteolytic assay, several enzyme dilutions were tested in order to construct a response curve. A response curve was also constructed using a solution of standard plasmin in 0.1 M Tris, 0.05 M NaCl, pH 7.8 (0.5 Plasmin International Unit (PIU)/ ml) and fibrinolytic activity was expressed in terms of PIU. RESULTS
AND
DISCUSSION
Hydrophobic interaction chromatography. Dry venom (300 mg of dry weight corresponding to 145-190 mg of protein) was dissolved in 3.5 ml of 0.1 M phos-
FIBRINOLYTIC
ENZYMES
37
phate buffer, pH 6.8, containing 1.0 M ammonium sulfate and 0.02% sodium azide. Purification was performed at room temperature but effluent was collected at 5°C. Following injection into the propyl-aspartamide column, elution was achieved as follows: 15 min isocratically at 100% buffer A (0.1 M sodium phosphate, 1.0 M ammonium sulfate, 0.02% NaN,, pH 6.8); a 60-min linear gradient to 100% buffer B (0.1 M sodium phosphate, 0.02% NaN,, pH 6.8); 30 min isocratically at 100% buffer B. The flow rate was 1 ml/min and the fraction size was 2 ml. Protein elution was monitored at 280 nm, and fibrinolytic activity was assessed by the fibrin plate assay (Table 1, a and b). Elution profiles are shown in Fig. 1. It should be noted that more than one fibrinolytic peaks is detected in each venom. However, the minor fibrinolytic peaks may be due to hemorrhagic enzymes which exist in both venoms and which display limited fibrinolytic activity. Hydroxylapatite chromatography. The fibrinolytic fractions resulting from hydrophobic interaction chromatography were pooled, concentrated, and dialyzed against 10 mM sodium phosphate buffer, pH 6.8. The dialyzed material was subsequently applied to the HPHT-Bio-Gel column, and elution was achieved by (a) a 12-min isocratic elution at 100% buffer A (10 mM sodium phosphate, 0.3 mM CaCl,, 0.02% NaN,, pH 6.8), and (b) a 60-min linear gradient to 60% buffer B (350 mM sodium phosphate, 0.3 mM CaCl,, 0.02% NaN,, pH 6.8). The flow rate was 1 ml/min, and the fraction size was 2 ml. It is important to note that in hydroxylapatite chromatography both enzymes displayed identical retention times, indicating similarity in their interaction with the hydroxylapatite matrix (Fig. 1).
E = 0.30 Lo z %.20 8 6 go.10 v) 2 0.00
L
El.00 100.80 z GO.60 EO.40 cl -P go.20 z 0.00
FIG. 6. Tryptic digestion of the fibrinolytic enzymes from Agkistrodon contortrix contortrix venom (top) and Agkistrodon piscivorus conanti venom (bottom). The products of the digestion were separated by reverse-phase chromatography on a Vydac C,, column (218TP54) as described under Materials and Methods. Arrows mark peptides with identical elution positions in both digests.
38
RETZIOS
AND
Reverse-phase chromatography. Reverse-phase chromatography was used to determine the effectiveness of the fractionation procedure and to assess the purity of the final preparations. Aliquots (100 ~1 of solution) were added to the column equilibrated with a mixture of 90% 0.1% trifluoroacetic acid in water (solvent A) and 10% 0.1% trifluoracetic acid in 80% acetonitrile-20% water (solvent B). Elution was achieved at a flow rate of 1 ml/min as follows: 5 min isocratically at 90% solvent A and 10% solvent B; a 45min linear gradient to 80% solvent B. Results for both venoms are shown in Fig. 2 and reveal the efficacy of this two-step HPLC procedure in preparing highly purified enzymes. Peak integration indicates that the fibrinolytic enzymes comprise 8-9% of the A. c. contortrix venom and 12% of the A. p. conanti venom. SDS-polyacrylamide gel electrophoresis assessment of purity. Whereas reverse-phase chromatography revealed homogeneous preparations of the enzymes, SDS-PAGE showed a number of low-molecular-weight bands below the main 24,000 MW band of the Accfib. However, these bands disappeared (Fig. 3) when the enzyme was pretreated with EDTA prior to denaturation and reduction by addition of SDS-PAGE sample buffer and boiling of the sample. It should be noted that these enzymes are in active configuration, but that Accfib preparations showed negligible loss of activity (less than 5%) when stored in solution at 5°C for a month. This fact indicates that the preparations are not contaminated with small amounts of unknown proteinases (as reverse-phase chromatography indicates) and that there is a degree of self-recognition. Degradation, observed in SDS-PAGE, is probably the result of autolysis, possibly resulting from a slow rate of inactivation during incubation with SDS-PAGE sample buffer. A similar phenomenon was reported by Bjarnason and Fox with another venom metalloprotease, bothropasin (15). Amino acid analysis. Amino acid analysis results (shown in Table 2) indicate significant similarity between the two fibrinolytic enzymes. The only important difference between the two compositions is in the number of half-cystine residues detected. Physicochemical properties. SDS-polyacrylamide gel electrophoresis shows that both venoms have a molecular weight of approximately 23,000. Isoelectric focusing shows that whereas the pI for the A. c. contortrix fibrinolytic enzyme is 6.5, the pI of the A. p. conanti equivalent is lower (approximately 5-5.5). The lower pl for Apcfib may be attributed to the slightly higher amounts of acidic amino acid residues as revealed by amino acid analysis. Western Structural and immunological comparisons. blotting (Fig. 4) revealed that Apcfib interacts with antibody raised against Accfib. This result suggests a significant degree of sequence and conformational similarity
MARKLAND
between the two enzymes. This suggestion was further reinforced with results obtained from cyanogen bromide and tryptic digestion. As can be seen in Fig. 5 the cyanogen bromide digestion products revealed by reverse-phase chromatography are very similar. Tryptic digestion profiles reveal more variety (Fig. 6), as would be expected. However, approximately 40% of the tryptic peptides elute in identical or similar positions. The results of these comparisons strongly suggest that Apcfib and Accfib are closely related enzymes. Comparison of fibrinolytic and azocaseinolytic activities. Although there is significant structural similarity, the functional properties of Accfib and Apcfib differ. Purified Apcfib possesses much weaker (about 23%) azocaseinolytic activity than Accfib. However, Apcfib preserves 70% of the fibrinolytic activity of Accfib (sp act 0.5 PIU/mg for Apcfib versus 0.71 PIU/mg for Accfib). CONCLUSIONS The method described above for the rapid purification of fibrinolytic enzymes from the venoms of A. c. contortrix and A. p. conanti produces homogeneous solutions of both enzymes. Although the composition of the crude venoms vary and few similarities are detected by SDS-PAGE and reverse-phase chromatography, it appears that the fibrinolytic components of both venoms are quite similar, as retention times in hydroxylapatite chromatography, amino acid analysis, comparison of tryptic and cyanogen bromide digests, and crossimmunoreaction indicate. ACKNOWLEDGMENT The authors acknowledge performing the immunoblotting
the
assistance studies.
of Dr. Hai-ming
Chen
in
REFERENCES 1. Pirkle, H., and Markland, F. S. (Eds.) (1988) “Hemostasis and Animal Venoms,” Decker, New York. 2. Didisheim, P., and Lewis, J. H. (1956) Fibrinolytic and coagulant activity of certain snake venoms and proteases. Proc. Sot. Exp. Biol. Med. 93, 10-13. 3. Kornalik, F. (1966) The influence of snake venoms on fibrinogen conversion and fibrinolysis. Mem. Inst. Butuntan Simp. Znt. 33, 179-188. 4. Moran, J. B., and Geren, C. R. (1981) Characterization of a fibrinogenase from northern copperhead (Agkistrodon contortrix mokasen) venom. Biochim. Biophys. Acta 659, 161-168. 5. Guan, A. L., Retzios, A. D., Henderson, G. N., and Markland, F. S. Purification and properties of a fibrinolytic enzyme from the venim of Agkistrodon contortrix contort& Manuscript in preparation. 6. Egen, N. B., Russell, F. E., Sammons, D. W., Humphreys, R. C., Guan, A. L., and Markland, F. S. (1987) Isolation by preparative isoelectric focusing of a direct acting fibrinolytic enzyme from the venom of Agkistrodon contortriz contort& (southern copperhead). Toxicon 25,1189-1198. 7. Markland, F. S., Reddy, K. N. N., and Guan, A. L. (1988) Purifica-
HPLC
PURIFICATION
OF
tion and characterization of a direct-acting fibrinolytic enzyme from southern copperhead venom, in “Hemostasis and Animal Venoms” (Pirkle, H., and Markland, F. S., Eds.), pp. 173-189, Dekker, New York. 8. Retzios, A. D., and Markland, F. S. (1988) A direct acting fibrinolytic enzyme from the venom of Agkistrodon contoFtFk contoFtFir.’ Effects on various components of the human blood coagulation and fibrinolysis systems. Thromb. Res. 62, 541-552. 9. Bidlingmeyer, B. A., Cohen, S. A., and Tarvin, T. L. (1988) Rapid analysis of amino acids using pre-column derivatization. J. Chromatogr. 336, 93-104. 10. Chen, H., Guan, A. L., and Markland, ties of the fibrinolytic enzyme from and its purification by immunoaffinity script in preparation.
F. S. Immunological southern copperhead chromatography.
propervenom Manu-
FIBRINOLYTIC 11.
39
ENZYMES
Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic transfer of proteins from PAG’s to nitrocellulose sheets: Procedure and some applications. PFOC. Natl. Acad. Sci. USA 76,4350-
4354. 12. Laemmli, assembly
U. K. (1970) Cleavage of structural proteins during the of the head of bacteriophage T4. Nature (London) 227,
680-685. 13. Charney, J., and Tomarelli, R. M. (1947) A calorimetric method for the determination of the proteolytic activity of duodenal juice. J. Biol. Chem. 171,501-505. 14. Bajwa, S. S., Markland, F. S., and Russell, lytic enzymes in western diamondback atrox) venom. Toxicon 18, 285-290.
F. E. (1980) rattlesnake
15. Bjarnason, J. B., and Fox, J. W. (1989) Hemorrhagic snake venoms. J. Toxicol. Toxin Reu. 7,121-178.
Fibrino(Crotulm
toxins
from
EXPRESSION
AND
PURIFICATION
1, 33-39
(1990)
HPLC-Based Two-Step Purification of Fibrinolytic Enzymes from the Venom of Agkistrodon contortrix con tortrix and Agkistrodon piscivorus conan ti’ Anastassios D. Retzios and Francis S. Markland’ Department 1303 North
Received
of Biochemistry, University of Southern California School of Medicine, Mission Road, CRL 104, Los Angeles, California 90033
January
2, 1990,
and
in revised
form
April
24, 1990
In investigations aimed at characterizing snake venom blood clot-dissolving enzymes, we have developed a rapid two-step high-performance chromatography method for the isolation of these fibrinolytic enzymes from the venoms of Agkistrodon contortrix contortrix and Agkistrodon piscivorus conanti. The first step consisted of hydrophobic interaction chromatography on a propyl-aspartamide column. Fractions containing the fibrinolytic activity were then concentrated and applied to a hydroxylapatite column. The resulting preparation, assessed for purity by reverse-phase chromatography and sodium dodecyl sulfate-polyacrylamide gel electrophoresis, was homogeneous. The molecular weight of both venom fibrinolytic enzymes was approximately 23,000 and amino acid analysis, immunological cross-reaction, cyanogen bromide, and tryptic digestion indicate a significant degree of structural similarity. However, the general proteolytic activity of the A. p. conanti venom enzyme was significantly lower than the corresponding activity of the A. c. contortrix venom, whereas their fibrinolytic activities were quite similar. OlSSOAcademic Press,Inc.
A number of enzymes that influence blood coagulation have been isolated from various snake venoms (1). These enzymes can either promote or inhibit coagulation. Fibrinolytic activities have received special attention because of their possible therapeutic role in dissolution of blood clots (2). Fibrinolytic agents have been identified in the venom of several Agkistrodon species (contortrix andpiscivorus) (3,4). A fibrinolytic enzyme from the venom of Agkistroi This work has been supported in part by NIH Grant from the USPHS and by Cortech, Inc., Denver, CO. * To whom correspondence should be addressed.
HL
31389
don contortrix contortrix (southern copperhead snake) has been purified by conventional liquid chromatography methods (5) and by isoelectric focusing (6) and characterized in this laboratory (7). This enzyme is a zinc metalloproteinase (one mole zinc per mole enzyme) with a molecular weight of 23,000. It readily cleaves the Ao-chain of fibrin and fibrinogen between Lys413 and Leu414 without activating or degrading plasminogen and protein C (8). In the present study we focused on developing a rapid method, basedon high-performance liquidchromatography, for the purification of fibrinolytic enzymes from the venoms of A. c. contortrix and the related species, A. p. conanti (Florida cottonmouth). Additionally, we investigated and compared the enzymes structurally, immunologically, and functionally.
MATERIALS
AND
METHODS
Materials. HPLC-grade solvents and chemicals were purchased from VWR Scientific, Cerritos, California; HPLC-grade trifluoroacetic acid was purchased from Pierce, Redford, Tennessee. Venoms were obtained from Biotoxins, Inc., St. Cloud, Florida. Plasminogen-free human fibrinogen (Kabi Diagnostica, grade L) was obtained from Helena Laboratories, Beaumont, Texas. Standard plasmin was obtained from The World Health Organization, International Laboratory for Biological Standards and Control, London, United Kingdom. L-(Tosylamido 2-phenyl)ethyl chloromethyl ketone (TPCK)3-treated trypsin was obtained from 3 Abbreviations used: AC&b, fihrlnolytic enzyme from the venom of Agkistrodon contortrix contortrti; Apcjib, fibrinolytic enzyme from the venom of Agkistrodon piscivorus conanti; TPCK, L-(tosylamido 2phenyl)ethyl chloromethyl ketone; SDS, sodium dodecyl sulfate; Tris, 2-amino-2-hydroxymethylpropane-1,3-diol; TBST, 0.02 M Tris buffer, pH 8.0, containing 0.14 M NaCl and 0.05% Tween 20; BSA, bovine serum albumin; DTT, dithiothreitol; TFA, trifluoroacetic
34
RETZIOS
AND
MARKLAND
TABLE
Summary
Step number
of Purification
Purification
1 2 3
Crude After After
venom hydrophobic hydroxylapatite
Note. activity
Crude After After In both of these
venom hydrophobic hydroxylapatite
cases hemorrhagic enzymes results
Agkistrodon
contortrir
contortrix
pisciuorus
enzymes that display limited in an underestimate of overall
venom
(a) and Apcfib
AC@
Total activity (PIU) fibrinolytic
conanti
venom
9.8 6.5 5.7 fibrinolytic
Worthington Biochemical Corp., Freehold, New Jersey. Reagents for immunoblotting were purchased from Sigma Chemical Co., St. Louis, Missouri. All other reagents were of analytical grade. Instruments. Purification of the fibrinolytic enzymes was performed on a Perkin-Elmer 410 Bio HPLC system equipped with an LC-95 variable-wavelength detector and an LC-100 integrator. Reverse-phase chromatography for the evaluation of purity of preparations was performed with a Beckman 334 dual-pump HPLC system equipped with a Hitachi 100-40 variable-wavelength detector and a Shimadzu CRlA integrator. Reverse-phase chromatography of cyanogen bromide and tryptic digests was performed on a Spectra-Physics 8800 HPLC system equipped with Spectra-Physics 8450 variable-wavelength detector and a Spectra-Physics WINner data acquisition module. Columns. For hydrophobic interaction chromatography, a propyl-aspartamide column (9.4 X 250 mm) from PolyLC, Columbus, Ohio, was employed. Hydroxylapatite chromatography was performed on a HPHT-BioGel column (7.2 X 100 mm) purchased from Bio-Rad, Richmond, California. Reverse-phase chromatography for the determination of preparation purity was performed on a C, 214TP54 column (4.1 X 250 mm) from Vydac, Hesperia, California. Amino acid analysis. Amino acid analyses were kindly performed by Dr. Anne Randolph, Chiron Corp., Emeryville, California. Amino acid compositions were obtained by the method described by Bidlingmeyer etal. (9). Cyanogen bromide digestion. The fibrinolytic enzymes purified from both snake venoms (3 mg) were to alkaline phosphatase; TCA, trichloroacetic acid.
capacity
Specific activity (PIU/mg)
Activity recovery (o/o)
0.052 0.46 0.71
100 66 58
0.062 0.31 0.5
100 81.5 65
enzyme
148 24 12 fibrinolytic yields.
(b)
enzyme
187 14 8
interaction chromatography chromatography
acid, IgG-AP, immunoglobulin complexed PAGE, polyacrylamide gel electrophoresis;
Enzymes Total protein (md
interaction chromatography chromatography (b)
1 2 3
of the Fibrinolytic
step
(a) Agkistrodon
1
9.2 7.5 6 are removed
at step
2. The
exclusion
of the
fibrinolytic
dialyzed against H,O, freeze-dried, and dissolved in 2 ml of 0.2 M Tris buffer, pH 8.2, containing 8 M urea. The enzymes were reduced and carboxymethylated as follows: 1 mg of dithiothreitol (DTT) was added to each solution (3.2 mM DTT solution) and the mixture was incubated at 37°C for 1 h; 2 mg of iodoacetamide was then added to each tube (5.4 mM iodoacetamide solution); the solutions were gassed with oxygen-free nitrogen; and the tubes were sealed and incubated for 1 h in the dark at 37°C. This procedure was repeated once more to complete the reaction. Following carboxymethylation the proteins were desalted by Sephadex G-15 chromatography and freeze-dried. The dried material was dissolved in 70% formic acid, and cyanogen bromide equal to the weight of the dried protein was added. The reaction was allowed to proceed with continuous stirring, under oxygen-free nitrogen and in the dark for 48 h. After 48 h, more cyanogen bromide was added (half of the dried protein weight) and the reaction was continued for an additional 12 h. Following cleavage, 100 ~1 from the digest of each fibrinolytic enzyme was removed and added to 900 ~1 of HPLC-grade water. The resulting sample was then added to a Vydac 218TP54 C,, reversephase chromatography column (4.1 X 250 mm) and the fragments were eluted according to the following protocol: (a) 5 min isocratically at 90% solvent A (0.1% TFA in water), 10% solvent B (0.1% TFA in 80% acetonitrile); (b) a 40-min linear gradient to 65% B. Tryptic digestion. Dried carboxymethylated fibrinolytic enzymes from both venoms were dissolved in 0.2 M Tris buffer, pH 8.7, containing 2 M urea. TPCK-treated trypsin was added at a ratio of 1:40 (trypsin:fibrinolytic enzyme, w/w). Digestion was allowed to proceed at room temperature, with continuous stirring for 24 h. Following digestion, 100 ~1 of each digest was added to a Vydac 218TP54 C,, column (4.1 X 250 mm) and fragments were eluted according to the following protocol:
_.:
HPLC PURIFICATION OF FIBRINOLYTIC ENZYMES
35
1 ~ I
-----F-l Al
A3
II
I
m---J 10 ?r-T--rI7
10
30 50 Fraction number
70
10
20 30
50
a
10
Time
40
3040
50
(min)
-r
___-Fraction number
Bl
10
2o
50
10
--.T-l
30 50 Fraction number
20
I
3040
I
1
/ 10
I 2o
I I 3040
(min)
/ a0
1
,I 10
1
k.
B3
Time - n
baseline
I
,r
A-. , 3040
10
-
2o
ao40
i
50
i.-,.-1
70
10 20 30 40 Fraction number
FIG. 1. Purification of the fibrinolytic enzymes from Agkistrodon contort& contortrix venom and Agkistrodon piscivorus conanti venom. Al and A2 show elution profiles for A. c. contortrix and A. p. conunti venoms, respectively, during hydrophobic interaction chromatography. Bl and B2 show elution profiles for A. c. contort& and A. p. conanti venoms, respectively, during hydroxylapatite chromatography. In all profiles solid lines indicate absorption at 280 nm whereas dashed lines indicate fibrinolytic activity.
(a) 5 min isocratically at 90% solvent A (0.1% TFA in water), 10% solvent B (0.1% TFA in 80% acetonitrile); (b) an 85min linear gradient to 70% solvent B. Western immunoblotting. Western blotting was performed using a polyclonal antibody prepared against the purified fibrinolytic enzyme from A. c. contortrix venom (10). Both Accfib and A&b were applied to a SDS polyacrylamide gel. After electrophoresis the proteins were transferred from the SDS polyacrylamide gel to a nitrocellulose membrane. Blotting was achieved by applying 300 mA overnight in 0.025 M Tris, pH 8.3, containing 0.192 M glycine, 0.1% SDS, and 40% methanol, according to the method of Towbin et al. (11). Half of the membrane, containing duplicate samples, was stained with a 0.05% solution of amido black to locate protein bands and then destained in 40% ethanol-5% acetic acid. Before a coating with the primary antibody, the other half of the membrane (not stained with amido black) was submerged in Tris-saline buffer (0.02 M Tris,
FIG. 2. Reverse-phase chromatography of crude venoms and fibrinolytic pools during purification of the fibrinolytic enzymes from Agk~trodon contortrix contort& venom (Acefib) and Agk~t~o~ pisciVOFUS conanti venom (Apcfiib). Al, crude A. c. contortrix venom; Bl, crude A. p. conanti venom; A2, AC&& following hydrophobic interaction chromatography; B2, Apefib following hydrophobic interaction chromatography; A3, Accfib following hydroxylapatite chromatography (baseline profile is shown underneath); B3, Apcfib following hydroxylapatite chromatography. On the latter profiles small peaks can be seen following the main protein peak. These are baseline artifacts and not real contaminants.
0.14 M NaCl), pH 8.0, containing 0.05% Tween 20 (TBST), and then incubated for 30 min with TBST containing 1% bovine serum albumin (BSA) to saturate the excess protein binding sites. The BSA-containing solution was then replaced with TBST containing antibody to fibrinolytic enzyme from A. c. contort& (1:lOOO dilution) and incubated for 30 min. The membrane was then washed with TBST five times for 10 min to remove unbound antibody, transferred to TBST containing anti-rabbit IgG-AP (alkaline phosphatase conjugate, 1:2000 dilution), and incubated for 30 min. After a coating by the second antibody, the membrane was washed with TBST five times for 10 min each time. Following the wash the membrane was transferred to the color development solution (66 ~1 of nitroblue tetrazolium in 10 ml of 0.1 M Tris, 0.1 M NaCl, 5 mM MgCl,, pH 9.5). After mixing, 33 ~1 of 5-bromo-4-chloro-3-indolyl phosphate (50 mglml in dimethylformamide) was added. The development of the color reaction was complete
36
RETZIOS
AND
MARKLAND TABLE
2
Amino Acid Analysis of the Fibrinolytic Enzymes from Agkistrodon contortrix contortrix Venom (Accfib) and Agkistrodon piscivorus conanti Venom (Apcfib) Fibrinolytic Amino acid
FIG. 3. SDS-polyacrylamide gel electrophoresis of purified Accfib. Lanes 1 and 2 contain Accfib (41 and 20.5 Fg, respectively) that was incubated with reducing SDS-PAGE sample buffer at 37°C for 5 min. Lanes 3 and 4 contain Accfib (41 and 20.5 pg, respectively) that was incubated with reducing SDS-PAGE sample buffer at 100°C for 10 min. Lanes 5 and 6 contain Accfib (41 and 20.5 pg, respectively) that was incubated in 17 mM EDTA for 10 min before addition of reducing SDS-PAGE and incubation at 100°C. Treatment with EDTA and boiling eliminate the lower molecular bands which are due to autolysis during SDS-PAGE sample preparation. Apcfib also displays autolysis, but upon EDTA treatment only a single band at 23,000 MW is evident, in accordance with HPLC-reverse-phase chromatography results.
within 10 min. Purple staining indicated the location where protein antigen interacted with the antibody. This was SDS-polyacrylamide gel electrophoresis. carried out according to the method of Laemmli (12). Isoelectric focusing. This was performed in a Pharmacia FBE 3000 flatbed electrophoresis apparatus, employing a constant power supply and using precast gels (Pharmacia-LKB, Piscataway, NJ) and Pharmalyte ampholyte solutions (Pharmacia-LKB). This was assessed by azocasein Proteolytic activity. hydrolysis. Azocasein was synthesized from cy-casein and diazotized sulfanilamide as described by Charney and Tomarelli (13). To measure proteolytic activity, 0.75 ml of the azocasein solution (about 50 mg/ml) was added to a 1.5-ml microcentrifuge tube. Test solution (50 ~1) was added, and the tube was incubated at 37°C for 30 min. The reaction was stopped by the addition of 0.75 ml of 1.16 M perchloric acid, and the precipitate was removed by centrifugation. Hydrolysis of azocasein was measured as increased absorbance at 390 nm in the supernatant, using as reference a blank treated in an identical manner. To accurately estimate Accfib and Apcfib specific activity (in absorbance units at 390 nm/mg enzyme), various amounts of the enzyme were added in order to construct a response curve. The linear part of the curve was used for the estimation of specific proteolytic activity. Fibrinolytic activity. Fibrinolytic activity was measured with the fibrin-plate-clearance assay described by
CYS Asx Glx Ser GUY His Arg Thr Ala Pro Tyr Val Met Ile Leu Phe LYS Trp
Accfib 6 31 20 13 14 10 9 13 11 6 5 12 12 22 7 8 ND
enzyme &Mb 3 35 21 14 14 9 8 14 12 6 6 14 6 11 22 7 9 ND
Note. The molecular weight assumed was 23,000. This estimate was derived from SDS-polyacrylamide gel electrophoresis. The results shown represent the average of two 24-h hydrolyses of each fibrinolytic enzyme (ND, not determined).
Bajwa et al. (14). For the preparation of fibrin plates two buffers were necessary: fibrinogen buffer and gelatin buffer. Fibrinogen buffer was prepared by adding 20.62 g of sodium barbital, 100 ml of 0.1 M HCl, 100 ml of “salt solution” (4.89 g CaCl, * 2H,O, 2.79 g MgCl, * 6H,O, 109.12 g NaCl, and H,O to a total volume of 1 liter), and 1350 ml of H,O. The pH of the buffer was adjusted to 7.75 with concentrated HCl and the volume was brought up to 2 liters with H,O. The gelatin buffer was prepared
FIG. 4. Western blotting of Accfb and Apcfib. The procedure is described under Materials and Methods. As can be seen, there is significant interaction of Apcfib with the antibody to Accfib.
HPLC
0.00
10
PURIFICATION
OF
30
4L
30
40
Timez~min)
f0 Lo 1.00 % : 20.50 i
22
j&L+!LJ!I 0.00
10
TimJrnin)
FIG. 5. Cyanogen bromide cleavage of the fibrinolytic enzymes from Agkistrodon contortriz contortrix venom (top) and Agkistrodon piscivorus conanti venom (bottom). The products of the digestion were separated by reverse-phase chromatography on a Vydac Cls column (218TP54) as described under Materials and Methods.
by adding 5.85 g NaCl, 10.31 g sodium barbital, 200 ml of 0.1 N HCl, 25 g gelatin, and 725 ml H,O. After the preparation of the buffers the following solutions were made: (a) 32 mg fibrinogen (76.2 g of dry weight of KABI plasminogen-free fibrinogen) was dissolved in 5.25 ml of fibrinogen buffer and 0.75 ml of H,O; (b) 20 ,ul of 1000 units/ml bovine thrombin (USP Thrombostat, from Parke-Davis, Morris Plains, NJ) was added to 0.5 ml of gelatin buffer. Fibrin plates were readied in disposable petri dishes (100 X 15 mm) by adding, in each one, 200 ~1 of thrombin solution and 6 ml of fibrinogen solution. The plates were incubated for 2-3 h at 37°C and eight equidistant holes, approximately 1 cm from the edge of the plate, were punched in each one. Ten microliters of sample was added per hole and the plates were then incubated for 18 h at 37°C. For accurate measurements of the areas of lysis, at the end of the incubation period the plates were flooded with 10% TCA solution. The diameters of the lysis areas were then measured against a dark background. As with the proteolytic assay, several enzyme dilutions were tested in order to construct a response curve. A response curve was also constructed using a solution of standard plasmin in 0.1 M Tris, 0.05 M NaCl, pH 7.8 (0.5 Plasmin International Unit (PIU)/ ml) and fibrinolytic activity was expressed in terms of PIU. RESULTS
AND
DISCUSSION
Hydrophobic interaction chromatography. Dry venom (300 mg of dry weight corresponding to 145-190 mg of protein) was dissolved in 3.5 ml of 0.1 M phos-
FIBRINOLYTIC
ENZYMES
37
phate buffer, pH 6.8, containing 1.0 M ammonium sulfate and 0.02% sodium azide. Purification was performed at room temperature but effluent was collected at 5°C. Following injection into the propyl-aspartamide column, elution was achieved as follows: 15 min isocratically at 100% buffer A (0.1 M sodium phosphate, 1.0 M ammonium sulfate, 0.02% NaN,, pH 6.8); a 60-min linear gradient to 100% buffer B (0.1 M sodium phosphate, 0.02% NaN,, pH 6.8); 30 min isocratically at 100% buffer B. The flow rate was 1 ml/min and the fraction size was 2 ml. Protein elution was monitored at 280 nm, and fibrinolytic activity was assessed by the fibrin plate assay (Table 1, a and b). Elution profiles are shown in Fig. 1. It should be noted that more than one fibrinolytic peaks is detected in each venom. However, the minor fibrinolytic peaks may be due to hemorrhagic enzymes which exist in both venoms and which display limited fibrinolytic activity. Hydroxylapatite chromatography. The fibrinolytic fractions resulting from hydrophobic interaction chromatography were pooled, concentrated, and dialyzed against 10 mM sodium phosphate buffer, pH 6.8. The dialyzed material was subsequently applied to the HPHT-Bio-Gel column, and elution was achieved by (a) a 12-min isocratic elution at 100% buffer A (10 mM sodium phosphate, 0.3 mM CaCl,, 0.02% NaN,, pH 6.8), and (b) a 60-min linear gradient to 60% buffer B (350 mM sodium phosphate, 0.3 mM CaCl,, 0.02% NaN,, pH 6.8). The flow rate was 1 ml/min, and the fraction size was 2 ml. It is important to note that in hydroxylapatite chromatography both enzymes displayed identical retention times, indicating similarity in their interaction with the hydroxylapatite matrix (Fig. 1).
E = 0.30 Lo z %.20 8 6 go.10 v) 2 0.00
L
El.00 100.80 z GO.60 EO.40 cl -P go.20 z 0.00
FIG. 6. Tryptic digestion of the fibrinolytic enzymes from Agkistrodon contortrix contortrix venom (top) and Agkistrodon piscivorus conanti venom (bottom). The products of the digestion were separated by reverse-phase chromatography on a Vydac C,, column (218TP54) as described under Materials and Methods. Arrows mark peptides with identical elution positions in both digests.
38
RETZIOS
AND
Reverse-phase chromatography. Reverse-phase chromatography was used to determine the effectiveness of the fractionation procedure and to assess the purity of the final preparations. Aliquots (100 ~1 of solution) were added to the column equilibrated with a mixture of 90% 0.1% trifluoroacetic acid in water (solvent A) and 10% 0.1% trifluoracetic acid in 80% acetonitrile-20% water (solvent B). Elution was achieved at a flow rate of 1 ml/min as follows: 5 min isocratically at 90% solvent A and 10% solvent B; a 45min linear gradient to 80% solvent B. Results for both venoms are shown in Fig. 2 and reveal the efficacy of this two-step HPLC procedure in preparing highly purified enzymes. Peak integration indicates that the fibrinolytic enzymes comprise 8-9% of the A. c. contortrix venom and 12% of the A. p. conanti venom. SDS-polyacrylamide gel electrophoresis assessment of purity. Whereas reverse-phase chromatography revealed homogeneous preparations of the enzymes, SDS-PAGE showed a number of low-molecular-weight bands below the main 24,000 MW band of the Accfib. However, these bands disappeared (Fig. 3) when the enzyme was pretreated with EDTA prior to denaturation and reduction by addition of SDS-PAGE sample buffer and boiling of the sample. It should be noted that these enzymes are in active configuration, but that Accfib preparations showed negligible loss of activity (less than 5%) when stored in solution at 5°C for a month. This fact indicates that the preparations are not contaminated with small amounts of unknown proteinases (as reverse-phase chromatography indicates) and that there is a degree of self-recognition. Degradation, observed in SDS-PAGE, is probably the result of autolysis, possibly resulting from a slow rate of inactivation during incubation with SDS-PAGE sample buffer. A similar phenomenon was reported by Bjarnason and Fox with another venom metalloprotease, bothropasin (15). Amino acid analysis. Amino acid analysis results (shown in Table 2) indicate significant similarity between the two fibrinolytic enzymes. The only important difference between the two compositions is in the number of half-cystine residues detected. Physicochemical properties. SDS-polyacrylamide gel electrophoresis shows that both venoms have a molecular weight of approximately 23,000. Isoelectric focusing shows that whereas the pI for the A. c. contortrix fibrinolytic enzyme is 6.5, the pI of the A. p. conanti equivalent is lower (approximately 5-5.5). The lower pl for Apcfib may be attributed to the slightly higher amounts of acidic amino acid residues as revealed by amino acid analysis. Western Structural and immunological comparisons. blotting (Fig. 4) revealed that Apcfib interacts with antibody raised against Accfib. This result suggests a significant degree of sequence and conformational similarity
MARKLAND
between the two enzymes. This suggestion was further reinforced with results obtained from cyanogen bromide and tryptic digestion. As can be seen in Fig. 5 the cyanogen bromide digestion products revealed by reverse-phase chromatography are very similar. Tryptic digestion profiles reveal more variety (Fig. 6), as would be expected. However, approximately 40% of the tryptic peptides elute in identical or similar positions. The results of these comparisons strongly suggest that Apcfib and Accfib are closely related enzymes. Comparison of fibrinolytic and azocaseinolytic activities. Although there is significant structural similarity, the functional properties of Accfib and Apcfib differ. Purified Apcfib possesses much weaker (about 23%) azocaseinolytic activity than Accfib. However, Apcfib preserves 70% of the fibrinolytic activity of Accfib (sp act 0.5 PIU/mg for Apcfib versus 0.71 PIU/mg for Accfib). CONCLUSIONS The method described above for the rapid purification of fibrinolytic enzymes from the venoms of A. c. contortrix and A. p. conanti produces homogeneous solutions of both enzymes. Although the composition of the crude venoms vary and few similarities are detected by SDS-PAGE and reverse-phase chromatography, it appears that the fibrinolytic components of both venoms are quite similar, as retention times in hydroxylapatite chromatography, amino acid analysis, comparison of tryptic and cyanogen bromide digests, and crossimmunoreaction indicate. ACKNOWLEDGMENT The authors acknowledge performing the immunoblotting
the
assistance studies.
of Dr. Hai-ming
Chen
in
REFERENCES 1. Pirkle, H., and Markland, F. S. (Eds.) (1988) “Hemostasis and Animal Venoms,” Decker, New York. 2. Didisheim, P., and Lewis, J. H. (1956) Fibrinolytic and coagulant activity of certain snake venoms and proteases. Proc. Sot. Exp. Biol. Med. 93, 10-13. 3. Kornalik, F. (1966) The influence of snake venoms on fibrinogen conversion and fibrinolysis. Mem. Inst. Butuntan Simp. Znt. 33, 179-188. 4. Moran, J. B., and Geren, C. R. (1981) Characterization of a fibrinogenase from northern copperhead (Agkistrodon contortrix mokasen) venom. Biochim. Biophys. Acta 659, 161-168. 5. Guan, A. L., Retzios, A. D., Henderson, G. N., and Markland, F. S. Purification and properties of a fibrinolytic enzyme from the venim of Agkistrodon contortrix contort& Manuscript in preparation. 6. Egen, N. B., Russell, F. E., Sammons, D. W., Humphreys, R. C., Guan, A. L., and Markland, F. S. (1987) Isolation by preparative isoelectric focusing of a direct acting fibrinolytic enzyme from the venom of Agkistrodon contortriz contort& (southern copperhead). Toxicon 25,1189-1198. 7. Markland, F. S., Reddy, K. N. N., and Guan, A. L. (1988) Purifica-
HPLC
PURIFICATION
OF
tion and characterization of a direct-acting fibrinolytic enzyme from southern copperhead venom, in “Hemostasis and Animal Venoms” (Pirkle, H., and Markland, F. S., Eds.), pp. 173-189, Dekker, New York. 8. Retzios, A. D., and Markland, F. S. (1988) A direct acting fibrinolytic enzyme from the venom of Agkistrodon contoFtFk contoFtFir.’ Effects on various components of the human blood coagulation and fibrinolysis systems. Thromb. Res. 62, 541-552. 9. Bidlingmeyer, B. A., Cohen, S. A., and Tarvin, T. L. (1988) Rapid analysis of amino acids using pre-column derivatization. J. Chromatogr. 336, 93-104. 10. Chen, H., Guan, A. L., and Markland, ties of the fibrinolytic enzyme from and its purification by immunoaffinity script in preparation.
F. S. Immunological southern copperhead chromatography.
propervenom Manu-
FIBRINOLYTIC 11.
39
ENZYMES
Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic transfer of proteins from PAG’s to nitrocellulose sheets: Procedure and some applications. PFOC. Natl. Acad. Sci. USA 76,4350-
4354. 12. Laemmli, assembly
U. K. (1970) Cleavage of structural proteins during the of the head of bacteriophage T4. Nature (London) 227,
680-685. 13. Charney, J., and Tomarelli, R. M. (1947) A calorimetric method for the determination of the proteolytic activity of duodenal juice. J. Biol. Chem. 171,501-505. 14. Bajwa, S. S., Markland, F. S., and Russell, lytic enzymes in western diamondback atrox) venom. Toxicon 18, 285-290.
F. E. (1980) rattlesnake
15. Bjarnason, J. B., and Fox, J. W. (1989) Hemorrhagic snake venoms. J. Toxicol. Toxin Reu. 7,121-178.
Fibrino(Crotulm
toxins
from
