J. Mol. BioE. (1991) 219, 591-592
Preliminary Crystallographic Analysis of Cardiotoxin with Major Fusion Activity from Taiwan Cobra (Naja naja atra) Venom Chun-Jung
V
Chen1v3, John Rose1y2,Chwan-Deng Hsiao19’, Tsong-Jen Lee3 Wen-guey Wu3 and Bi-Cheng Wang1y2t ‘Departments of Crystallography and Biological Sciences University of Pittsburgh, Pittsburgh, PA 15260, U.S.A. ‘Crystallography Laboratory, Institute of Molecular Biology Academia Sinica, Taipei, Taiwan 11529, Republic of China 31nstitute of Life Sciences, National Tsing Hua University Hsinchu, Taiwan 30043, Republic of China (Received 3 December 1990; accepted 28 February
1991)
Crystals of a cardiotoxin from Taiwan cobra venom have been obtained by the va,por diffusion method using methyl pentanediol as precipitant. The crystals belong to the hexagonal space group P6,22 (or P6,22), with cell dimensions a = b = 47-5 A, c = 111.3 A. c1 = /l= 90” and y = 120” and diffract to a resolution of 2.2 A. There is one molecule per asymmetric unit and the solvent content is estimated to be 53%.
Keywords: crystals;
cardiotoxin;
Cardiotoxins, major protein components of Elapidae snake venom, consist of a single polypeptide chain of 60 to 63 amino acid residues and four disulfides (Dufton & Hider, 1988). The species naja (the cobras) contain four to five similar cardiotoxin isomers, some differing by only a single amino acid substitution, represent 50% of the snake’s venom. Due partly to difficulty in separating the isomers by ion exchange chromatography investigation of their structure-function properties is problematical. For instance, the previous functional assay based on hemolysis has been questioned because of the possible contamination of phospholipase A,, which may act synergetically with the cardiotoxins to increase their lytic power (Louw & Visser, 1978; Harvey, 1985). In addition, the separation of the cardiotoxin isomers by high performance liquid chromatography (h.p.1.c.S) has in some cases proved to be difficult as evidenced by two-dimensional nuclear magnetic resonance studies (Otting et al., 1987). Indeed, sample purity may explain the lack of good diffracting cardiotoxin crystals (Wang
cobra;
venom;
fusion
assay
& Yang, 1981), since only one X-ray structure of a cardiotoxin has been determined at 2.5 A! resolution (1 A = 61 nm) (Rees et al., 1990) despite its abundance in several snake species. Most activity studies of cardiotoxin center on t,hree assays: erythrocyte lysis, LD,,, value and muscle cell depolarization (Hodges et al., 1987). Although its target remains obscure, it is clear that cardiotoxins are able to perturb membrane structures (lipid and/or protein), having an effect on both excitable and non-excitable membranes. Recently a more specific method has been designed in vitro for the study of cardiotoxin-lipid interactions (Chien et al., 1991). The fusion activity of sphingomyelin vesicles has been used to assay cardiotoxins from Baja naja atra. This method can not only be applied during protein purification, since it is not affected by phospholipase A, contamination, but, also for the ’ comparative study of the structure-function for the purified proteins. A new cardiotoxin, named cardiotoxin V (CTXV) according to its elution position in the fusion assay, has been identified. It consists of 62 amino acid residues determined by primary sequence studies (Chien et al., 1991). CTXV exhibits major activity in inducing fusion of sphingomyelin vesicles but has little hemolysis activity in erythrocytes. This is in sharp contrast to the other cardiotoxins such as cardiotoxin III, which shows high
t Author to whom all rorrespondencr should br addressed at: Department of Crystallography. llniversitv of Pittsburgh, Pittsburgh. PA 15260. 1‘.S.;\. $ Abbriviations used: h.P.1.c.. high performance liquid chromatography: CTXV. cardiotoxin 717. 591 0022-2836/91/12059lbo2
$03.00/O
0
1991
Academic
Press
Limited
592
C.-J.
Chen
hemolysis activity but has low fusion activity. A better knowledge of the three-dimensional structure and a well-defined functional study of cardiotoxin should help address the mechanism of cardiotoxin actions. In addition to cardiotoxin Vf (Rees et al., 1990), crystallographic results have been reported for four a-neurotoxins: neurotoxin B (Tsernoglou &, Petsko, 1977), erabutoxin B (Smith et al., 1988), ol-cobratoxin (Walkinshaw et al., 1980) and a-bungarotoxin (Love & Stroud, 1986). The folding of these structures is essentially the same, which consists of three loops of antiparallel P-sheets, called loops I, II and III (Rees et al., 1990). While there is considerable overall structural similarity between the cardiotoxin Vf: and the a-neurotoxins, the orientations of the /?-loops, noticeably loop I and loop II of the cardiotoxin Vy are quite different to those observed in the a-neurotoxins. A comparison of the chemical sequence (Chein et al., 1991) of CTXV and that of cardiotoxin Vf: shows that they differ by 18 residues, and of these 18 residues 13 are located within loop I and loop II. In addition, CTXV has an extra residue, hi&dine, inserted between residues 4 and 5 (in loop I) and an extra proline inserted between residues 29 and 30 (tip of loop II). Therefore, it is reasonable to assume that even within the same cardiotoxin family the largest conformational difference may occur in loop I and loop II also. The detailed three-dimensional structure of another cardiotoxin, such as the CTXV. could provide the needed information to verify this assumption as well as to correlate their activities. We report here the crystallization and preliminary X-ray diffraction results of CTXV. Purification of CTXV was done using the procedure described by Yang et al. (1981) with slight modification. It consisted of repeated SP-Sephadex column chromatography of the crude venom obtained from Sigma (V 9250) followed by an h.p.1.c. purification step using an ultrapore reversephase C-8 5 PM column (4.5 mm x 250 mm; Beckman) and a acetonitrile/phosphate buffer gradient (10 m&i-NaH,PO,, 10 mM-Na,SO,, pH 4). Crystals of CTXV were grown by vapor diffusion (McPherson, 1982) at 4°C from 20 ~1 droplets of protein solution (10 mg/ml) containing 20% (v/v) 2-methyl-2,4-pentanediol and in @05 M-potassium phosphate buffer (pH 68) placed in wells of a g-well spot plate. The plate was placed in a sandwich box containing a reservoir of the above buffer solution containing 50% 2-methyl-2,4-pentanediol. Crystals appeared in about three days and grew to a size of @5 mm x O-3 mm x @3 mm in two weeks. For X-ray diffraction analysis a crystal was mounted in a thin-walled glass capillary containing a small amount of mother liquor, to prevent dehydration, and sealed with diffusion pump oil. Oscillation diffraction images, taken on a Siemens Xl000 area detector system using double-mirrorfocused 5 kW CuKa X-rays generated from a Edited
et al.
Rigaku RU300 rotating anode, show good diffraction to 2.5 A resolution, with some reflections observed to a resolution of 2.2 A. A data set of 2.5 .& resolution was collected. The data collection was performed using the Harvard COLLECT routines (Blum et al., 1987) with each data frame (@25”) exposed for 150 seconds. Crystal orientation, integration and scaling were carried out using the XENGEN (Howard et al., 1987) program suite. Analysis of the three-dimensional data set indicates a hexagonal space group with cell parameters a = b = 47.5 A, c = 111.3 A, a = j3 = 90” and y = 120”. Systematic absences(001except for 1 = 6n) and Laue-symmetry of the diffraction pattern (6/ mmm) suggest that the space goup is either P6,22 or P6,22. Assuming one molecule per asymmetric unit,, the V, value (Matthews, 1968) is calculated to be 2.59 and the solvent content of the carystal estimated to be 53%, which is in the normal range for protein crystals. A full crystallographic study of CTXV is underway. This work was supported by grants NS(’ 7%0208-,2100510, N‘SC 79-020%MOO7-120. NSC 79-0203-b(N)l-15 from the National Science Council and research program AS 6602101132-8 of the Academia Sinica. Republic of China. Support from the Faculty of Arts and Sciences. University of Pittsburgh, U.S.A. is also acknowledged. References Blum, M.. Metcalf. P.. Harrison. S. (‘. & 1Vilry. I). (‘. (1987). J. Appl. (‘rystallogr. 20. 235-442. Chien, K.-Y., Huang, W.-N., Jean, J.-H. BE Wu. W. (1991). J. Biol. Chem. 266, 3252-3259. Dufton, M. *J. & Hider. R. C. (1988). Phnmlor. Thur. 36. l-40. Harvey, A. L. (1985). J. Toxicol. Toxin Rec. 4. 41-69. Hodges, 8. J.. Agbaji. A. S.. Harvey. A. L. & Hider. K. (‘. (1987). EUT. J. Bioehem. 165, 373-383. Howard. A. J., Gilliland, G. L.. Finzel. 13. (I.. Poulos. T. L., Ohlendorf. D. H. & Salemme. F. R. (1987). .I. Appl. CT?JSta@T. 20, 383-387. Louw, A. I. & Visser. L. (1978). &o&in?. Niophys. Ac+r/. 512. 163-171. Love, R. & Stroud. R. M. (1988). Protein h%y. 1. 37-46. Matthews, B. M. (1968). J. Mol. Biol. 33. 491-497. McPherson. A. (1982). Preparation and Analysis oj’ f’rotui~, Crystals, pp. 96-97, John Wiley. New York. Otting, G., Marchot, P., Bougis, P. E., Rochat, H. & Wiithrich, K. (1987). EUT.J. B&hem. 168, 603-607. Rees, B., Bilwes. A., Samama. *J. P. & Moras. D. ( 1990). J. Mol. Biol. 214. 281-297. Smith, ,J. L.. Corfield. P. W. R., Hendrickson. 12’. A. & Low. B. W. (1988). Acta Crystallogr. .swt..-I. 44. 357-368. Tsernoglou. D. & Petsko. G. 4. (1977). I’roc. Srrf. dc~rl. Sci., C1.S.A. 74, 971-974. Walkinshaw, M. D., Saenger, W. clr Maelikr. A. (1980). Proc. Nat. Acad. Sk., I7.R.A. 77. 2400.-2404. Wang, A. H.-.J. & Yang. (‘. C. (1981). J. Rio/. C’h~nc. 256. 9279-9282. Yang, C. (:., King, K. & Sun. T. P. (1981). ToCon. 19. 645-659.
by R. Iluber
Preliminary Crystallographic Analysis of Cardiotoxin with Major Fusion Activity from Taiwan Cobra (Naja naja atra) Venom Chun-Jung
V
Chen1v3, John Rose1y2,Chwan-Deng Hsiao19’, Tsong-Jen Lee3 Wen-guey Wu3 and Bi-Cheng Wang1y2t ‘Departments of Crystallography and Biological Sciences University of Pittsburgh, Pittsburgh, PA 15260, U.S.A. ‘Crystallography Laboratory, Institute of Molecular Biology Academia Sinica, Taipei, Taiwan 11529, Republic of China 31nstitute of Life Sciences, National Tsing Hua University Hsinchu, Taiwan 30043, Republic of China (Received 3 December 1990; accepted 28 February
1991)
Crystals of a cardiotoxin from Taiwan cobra venom have been obtained by the va,por diffusion method using methyl pentanediol as precipitant. The crystals belong to the hexagonal space group P6,22 (or P6,22), with cell dimensions a = b = 47-5 A, c = 111.3 A. c1 = /l= 90” and y = 120” and diffract to a resolution of 2.2 A. There is one molecule per asymmetric unit and the solvent content is estimated to be 53%.
Keywords: crystals;
cardiotoxin;
Cardiotoxins, major protein components of Elapidae snake venom, consist of a single polypeptide chain of 60 to 63 amino acid residues and four disulfides (Dufton & Hider, 1988). The species naja (the cobras) contain four to five similar cardiotoxin isomers, some differing by only a single amino acid substitution, represent 50% of the snake’s venom. Due partly to difficulty in separating the isomers by ion exchange chromatography investigation of their structure-function properties is problematical. For instance, the previous functional assay based on hemolysis has been questioned because of the possible contamination of phospholipase A,, which may act synergetically with the cardiotoxins to increase their lytic power (Louw & Visser, 1978; Harvey, 1985). In addition, the separation of the cardiotoxin isomers by high performance liquid chromatography (h.p.1.c.S) has in some cases proved to be difficult as evidenced by two-dimensional nuclear magnetic resonance studies (Otting et al., 1987). Indeed, sample purity may explain the lack of good diffracting cardiotoxin crystals (Wang
cobra;
venom;
fusion
assay
& Yang, 1981), since only one X-ray structure of a cardiotoxin has been determined at 2.5 A! resolution (1 A = 61 nm) (Rees et al., 1990) despite its abundance in several snake species. Most activity studies of cardiotoxin center on t,hree assays: erythrocyte lysis, LD,,, value and muscle cell depolarization (Hodges et al., 1987). Although its target remains obscure, it is clear that cardiotoxins are able to perturb membrane structures (lipid and/or protein), having an effect on both excitable and non-excitable membranes. Recently a more specific method has been designed in vitro for the study of cardiotoxin-lipid interactions (Chien et al., 1991). The fusion activity of sphingomyelin vesicles has been used to assay cardiotoxins from Baja naja atra. This method can not only be applied during protein purification, since it is not affected by phospholipase A, contamination, but, also for the ’ comparative study of the structure-function for the purified proteins. A new cardiotoxin, named cardiotoxin V (CTXV) according to its elution position in the fusion assay, has been identified. It consists of 62 amino acid residues determined by primary sequence studies (Chien et al., 1991). CTXV exhibits major activity in inducing fusion of sphingomyelin vesicles but has little hemolysis activity in erythrocytes. This is in sharp contrast to the other cardiotoxins such as cardiotoxin III, which shows high
t Author to whom all rorrespondencr should br addressed at: Department of Crystallography. llniversitv of Pittsburgh, Pittsburgh. PA 15260. 1‘.S.;\. $ Abbriviations used: h.P.1.c.. high performance liquid chromatography: CTXV. cardiotoxin 717. 591 0022-2836/91/12059lbo2
$03.00/O
0
1991
Academic
Press
Limited
592
C.-J.
Chen
hemolysis activity but has low fusion activity. A better knowledge of the three-dimensional structure and a well-defined functional study of cardiotoxin should help address the mechanism of cardiotoxin actions. In addition to cardiotoxin Vf (Rees et al., 1990), crystallographic results have been reported for four a-neurotoxins: neurotoxin B (Tsernoglou &, Petsko, 1977), erabutoxin B (Smith et al., 1988), ol-cobratoxin (Walkinshaw et al., 1980) and a-bungarotoxin (Love & Stroud, 1986). The folding of these structures is essentially the same, which consists of three loops of antiparallel P-sheets, called loops I, II and III (Rees et al., 1990). While there is considerable overall structural similarity between the cardiotoxin Vf: and the a-neurotoxins, the orientations of the /?-loops, noticeably loop I and loop II of the cardiotoxin Vy are quite different to those observed in the a-neurotoxins. A comparison of the chemical sequence (Chein et al., 1991) of CTXV and that of cardiotoxin Vf: shows that they differ by 18 residues, and of these 18 residues 13 are located within loop I and loop II. In addition, CTXV has an extra residue, hi&dine, inserted between residues 4 and 5 (in loop I) and an extra proline inserted between residues 29 and 30 (tip of loop II). Therefore, it is reasonable to assume that even within the same cardiotoxin family the largest conformational difference may occur in loop I and loop II also. The detailed three-dimensional structure of another cardiotoxin, such as the CTXV. could provide the needed information to verify this assumption as well as to correlate their activities. We report here the crystallization and preliminary X-ray diffraction results of CTXV. Purification of CTXV was done using the procedure described by Yang et al. (1981) with slight modification. It consisted of repeated SP-Sephadex column chromatography of the crude venom obtained from Sigma (V 9250) followed by an h.p.1.c. purification step using an ultrapore reversephase C-8 5 PM column (4.5 mm x 250 mm; Beckman) and a acetonitrile/phosphate buffer gradient (10 m&i-NaH,PO,, 10 mM-Na,SO,, pH 4). Crystals of CTXV were grown by vapor diffusion (McPherson, 1982) at 4°C from 20 ~1 droplets of protein solution (10 mg/ml) containing 20% (v/v) 2-methyl-2,4-pentanediol and in @05 M-potassium phosphate buffer (pH 68) placed in wells of a g-well spot plate. The plate was placed in a sandwich box containing a reservoir of the above buffer solution containing 50% 2-methyl-2,4-pentanediol. Crystals appeared in about three days and grew to a size of @5 mm x O-3 mm x @3 mm in two weeks. For X-ray diffraction analysis a crystal was mounted in a thin-walled glass capillary containing a small amount of mother liquor, to prevent dehydration, and sealed with diffusion pump oil. Oscillation diffraction images, taken on a Siemens Xl000 area detector system using double-mirrorfocused 5 kW CuKa X-rays generated from a Edited
et al.
Rigaku RU300 rotating anode, show good diffraction to 2.5 A resolution, with some reflections observed to a resolution of 2.2 A. A data set of 2.5 .& resolution was collected. The data collection was performed using the Harvard COLLECT routines (Blum et al., 1987) with each data frame (@25”) exposed for 150 seconds. Crystal orientation, integration and scaling were carried out using the XENGEN (Howard et al., 1987) program suite. Analysis of the three-dimensional data set indicates a hexagonal space group with cell parameters a = b = 47.5 A, c = 111.3 A, a = j3 = 90” and y = 120”. Systematic absences(001except for 1 = 6n) and Laue-symmetry of the diffraction pattern (6/ mmm) suggest that the space goup is either P6,22 or P6,22. Assuming one molecule per asymmetric unit,, the V, value (Matthews, 1968) is calculated to be 2.59 and the solvent content of the carystal estimated to be 53%, which is in the normal range for protein crystals. A full crystallographic study of CTXV is underway. This work was supported by grants NS(’ 7%0208-,2100510, N‘SC 79-020%MOO7-120. NSC 79-0203-b(N)l-15 from the National Science Council and research program AS 6602101132-8 of the Academia Sinica. Republic of China. Support from the Faculty of Arts and Sciences. University of Pittsburgh, U.S.A. is also acknowledged. References Blum, M.. Metcalf. P.. Harrison. S. (‘. & 1Vilry. I). (‘. (1987). J. Appl. (‘rystallogr. 20. 235-442. Chien, K.-Y., Huang, W.-N., Jean, J.-H. BE Wu. W. (1991). J. Biol. Chem. 266, 3252-3259. Dufton, M. *J. & Hider. R. C. (1988). Phnmlor. Thur. 36. l-40. Harvey, A. L. (1985). J. Toxicol. Toxin Rec. 4. 41-69. Hodges, 8. J.. Agbaji. A. S.. Harvey. A. L. & Hider. K. (‘. (1987). EUT. J. Bioehem. 165, 373-383. Howard. A. J., Gilliland, G. L.. Finzel. 13. (I.. Poulos. T. L., Ohlendorf. D. H. & Salemme. F. R. (1987). .I. Appl. CT?JSta@T. 20, 383-387. Louw, A. I. & Visser. L. (1978). &o&in?. Niophys. Ac+r/. 512. 163-171. Love, R. & Stroud. R. M. (1988). Protein h%y. 1. 37-46. Matthews, B. M. (1968). J. Mol. Biol. 33. 491-497. McPherson. A. (1982). Preparation and Analysis oj’ f’rotui~, Crystals, pp. 96-97, John Wiley. New York. Otting, G., Marchot, P., Bougis, P. E., Rochat, H. & Wiithrich, K. (1987). EUT.J. B&hem. 168, 603-607. Rees, B., Bilwes. A., Samama. *J. P. & Moras. D. ( 1990). J. Mol. Biol. 214. 281-297. Smith, ,J. L.. Corfield. P. W. R., Hendrickson. 12’. A. & Low. B. W. (1988). Acta Crystallogr. .swt..-I. 44. 357-368. Tsernoglou. D. & Petsko. G. 4. (1977). I’roc. Srrf. dc~rl. Sci., C1.S.A. 74, 971-974. Walkinshaw, M. D., Saenger, W. clr Maelikr. A. (1980). Proc. Nat. Acad. Sk., I7.R.A. 77. 2400.-2404. Wang, A. H.-.J. & Yang. (‘. C. (1981). J. Rio/. C’h~nc. 256. 9279-9282. Yang, C. (:., King, K. & Sun. T. P. (1981). ToCon. 19. 645-659.
by R. Iluber
