Biochemistry 1993, 32, 2131-2136
2131
Art ides Solution Conformation of Cobrotoxin: A Nuclear Magnetic Resonance and Hybrid Distance Geometry-Dynamical Simulated Annealing Study? Chin Yu,*J R. Bhaskaran,fs Li-Chin Chuang,t and Chen-Chung Yangll Chemistry Department and Institute of Life Science, National Tsing Hua University, Hsinchu, Taiwan, Republic of China Received August 18, 1992; Revised Manuscript Received November 30, 1992
ABSTRACT: The solution conformation of cobrotoxin has been determined by using proton nuclear magnetic
resonance spectroscopy. With the combination of various two-dimensional N M R techniques, the 'H-NMR spectrum of cobrotoxin was completely assigned (Yu et al., 1990). A set of 435 approximate interproton distance restraints was derived from nuclear Overhauser enhancement (NOE) measurements. These N O E constraints, in addition to the 29 dihedral angle constraints (from coupling constant measurements) and 26 hydrogen bonding restraints (from the pattern of short-range NOEs), form the basis of 3-D structure determination by the hybrid distance geometry-dynamical simulated annealing method. The 23 structures that were obtained satisfy the experimental restraints, display small deviation from idealized covalent geometry, and possess good nonbonded contacts. Analysis of converged structures indicated that there are two antiparallel fl sheets (double and triple stranded), duly confirming our earlier observations. These are well defined in terms of both atomic root mean square (RMS) differences and backbone torsional angles. The average backbone R M S deviation between the calculated structures and the mean structure, for the fl-sheet regions, is 0.92 A. The mean solution structure was compared with the X-ray crystal structure of erabutoxin b, the homologous protein. This yielded information that both structures resemble each other except at the exposed loop/surface regions, where the solution structure seems to possess more flexibility.
Cobrotoxin is a neurotoxic protein isolated in a crystalline state from thevenom of Taiwan cobra (Naja naja atra) (Yang, 1965, 1967). It is a small, basic protein consisting of a single polypeptide chain of 62 amino acid residues, cross-linked by 4 disulfide bonds (Yang et al., 1969a,b). Thecompleteamino acid sequence and the positions of disulfide bonds in cobrotoxin have been established (Yang et al., 1969a,b, 1970). Cobrotoxin binds specifically to the nicotinic acetylcholine receptor on the postsynaptic membrane and thus blocks neuromuscular transmissions. Disulfide bonds and Tyr25, which are buried in the molecule, form a central core to maintain and stabilize the active conformation of the toxin (Chang et al., 1971a,b; Yang et al., 1974; Yang, 1987). Selective and stepwise chemical modifications of cobrotoxin indicate that at least two cationic groups, and the €-amino groupof Lys47 and a quanidino group of Arg33, both of which are common to all known postsynaptic snake neurotoxins, held at a certain critical distance in the molecule, are functionallyimportant for its neuromuscular blocking activity (Chang et al., 1971a,b; Yang, 1974, 1988). Recently, we reported 'H-NMR resonance assignments to specific protons in cobrotoxin for over 95% of the backbone amide and C a protons (Yu et al., 1990). These results were achieved by use of the sequential assignment technique, a two-dimensional NMR method developed by Wuthrich and This work was supportedby research grants from the National Science Council (NSC 81-0208-M-007-62 and NSC 81-0208-M-007-81) of the Republic of China. * To whom correspondence should be addressed. Chemistry Department. On leave from the Department of Physics, Bharathidasan University, Tiruchirapalli, India. 11 Institute of Life Science.
*
0006-2960/93/0432-2131$04.00/0
co-workers (Anil Kumar et al., 1981; Wagner & Wuthrich, 1982; Rance et al., 1983; Wiithrich, 1986). The secondary structures of the antiparallel triple- and double-stranded /3 sheets were also determined by observing the NOEs for cobrotoxin (Yu et al., 1990). The investigation was further extended in order to determine the tertiary structure of cobrotoxin in aqueous solution,the results of which are reported here.
MATERIALS AND METHODS Nuclear Magnetic Resonance Spectroscopy. Cobrotoxin was isolated and purified from the venom of Taiwan cobra (Naja naja atra) by chromatography on a SP-Sephadex (2-25 column as described by Yang et al. (1981). It was further purified by reverse-phase HPLC on a column of TSK gel ODS-120T (preparative, 21.5 mm X 30 cm) obtained from Toyo Soda Manufacturing Co., Tokyo. Cobrotoxin samples were prepared in three different ways to attain various exchanged conditions for the amide protons of the protein. Nonexchanged cobrotoxin was prepared by dissolving lyophilized cobrotoxin (70 mg) in 90% H20/10% D2O ( O S mL, Merck) at 20 OC; this sample produces the spectrum of all amide protons. We prepared fully exchanged cobrotoxin by dissolving protein (70 mg) in D2O and heating at 50 OC for 2 h; this sample produces the spectrum without labile amide protons. We prepared partially exchanged cobrotoxin by dissolving the protein in 99.95% D20 at 5 "C;the spectrum of such a sample, measured at 5 OC, contains resonances from slowly exchanging amide protons. All NMR spectra were recorded on a 400-MHz spectrometer (Bruker AM-400) equipped with an Aspect 3000 computer; 4,4-dimethyl-4-silapentane-l-sulfonate was used as an internal standard. NOESY (Jeener et al., 1979; Anil 0 1993 American Chemical Society
2132 Biochemistry, Vol. 32, No. 9, 1993 Kumar et al., 1981; Macura et al., 1981) and DQF-COSY (Shaka & Freeman, 1983; Rance et al., 1983) spectra were recorded in D20 with four different mixing periods, 50, 80, 130, and 300 ms for the NOESY experiments. Spectra taken in 90% Hz0/10% D20 include DQF-COSY and NOESY experiments. Quadrature detection in 0 1 was achieved using time-proportional phase incrementation (Marion & Wiithrich, 1983). All homonuclear 2-D experiments (Ernst et al., 1987) were performed on a 20 mM protein sample; 768 t l increments were recorded, with 2048 complex points. For each free induction, the decays were Fourier-transformed on w2 using a Kaiser window and on w1 using a 45' phase-shifted sine-bell window. All two-dimensional data were processed on a p V a x computer (MV 3600) using the FTNMR software kindly provided by Dr. D. Hare. Distance Restraints. About 95% of the assignments of the N H and C a protons in the 'H-NMR spectra of cobrotoxin have been reportedearlier (Yuet al., 1990). During thecourse of this present study, the rest of the N H and C a protons and a few additional assignments for side chain protons were made. Totally, a set of 435 interproton distance restraints were obtained from the analysis of NOESY spectra. The restraints are comprised of 192 intraresidue distances, 147 short-range [(i - J ) < 51 interresidue distances, and 96 long-range [(i J ) > 51 interresidue distances. The NOEs were obtained by plotting the NOE build-up curves, for which the volumes of cross-peaks of the NOESY spectra were used for interpretation. The interproton distances, dj,, were calculated according to the equation (Gondol & van Binst, 1986):
d, = di,,,(c,m/aij)1'6 in which dImstands for a known calibrated distance (1.78 8, for geminal protons) and uim and ai, stand for the crossrelaxation (NOE build-up) rate for spins l,m and i j , respectively (obtained from the initial slope of the NOE buildup curve). The distance constraints involving the pseudoatoms were taken into consideration as described by Wiithrich et al. (1983). Dihedral Angle Constraints. 3 J ~coupling ~ a constants were obtained in the fingerprint region of the DQF-COSY spectrum using the peak to peakseparation of the antiphase fine structure components in the contour plots. For J H Nsmaller ~ than 5.5 Hz, 4 was restricted to the range -9Oo,-4O0; for J H Nequal ~ to 8.0 Hz, 4 was restricted to -17Oo,-7O0, and for large couplings with J H N greater ~ than 10.0 Hz, the 4 angle constraints were to the range -140°,-100'. Thus, the backbone dihedral I$ angle restraints were obtained for 29 residues by using the Karplus (1 963) relationship (Pardi et al., 1984). Hydrogen Bonding Restraints. Backbone hydrogen bonds within the antiparallel @ sheets were identified according to the criteria laid out by Wagner et al. (1987) which consists of the presence of NOES between residues on opposite strands of an antiparallel @ sheet in conjunction with slowly exchanging backbone amide protons; 13 slowly exchanging backbone N H protons [identified earlier by Yu et al. (1 990)] could be easily assigned unambiguously in the above manner to hydrogen bonds within antiparallel @-sheetsegments. These NH(i)COO') hydrogen bonds are as follows: i j = 5,13; 15,3; 13,5; 26,53; 28,51; 29,36; 27,38; 36,29; 38,27; 42,23; 52,28; 53,26; and 55,24 where i and j represent the residue positions of the two amino acids involved. In subsequent calculations, two restraints were used for each NH(i)
