Elsevier Science Publishers B.V., Amsterdam
Simple fractionation of phospholipase A, analogues from
snake venom by high-performance liquid chromatography
CHIH-HUNG LO and SHYH-HORNG CHIOLJ*
Institute of Biochemical Sciences, National Taiwan University and Instirute of Biological Chemistry, P.O. Box 23-106 Academia Sinica, Taipei IO098 f Taiwan)
(First received January 29th, 1990; revised manuscript received April 6th, 1990)
The venoms of the elapid snakes have been studied extensively both chemically and pharmacologically during the past four decades [ 1,2]. The components isolat- ed from the crude venoms, in addition to some miscellaneous proteins/peptides and enzymes, generally fall into three major categories based on their structures and activities, i.e. (A) phospholipases AZ, (B) neurotoxins, and (C) cardiotoxins (or cytotoxins). All these biologically active proteins have been used widely as tools in the studies of various biological phenomena of molecular and cell biol- ogy.
Previous fractionation and purification of snake venom proteins involved mul- tiple chromatographic steps, including both ion-exchange chromatography and gel permeation techniques, which seemed laborious and tedious. Although most venom toxins can be obtained by conventional open-column techniques, some new methods employing the recent advances in instrumentation for high-per- formance liquid chromatography (HPLC) are highly desirable to improve the speed of analysis and resolution of various isotoxins from the venom.
This paper describes a rapid and semipreparative HPLC method using cation exchange for the isolation and purification of phospholipase A2 (PLA2) from Thailand cobra (Naja naja siamensis), which can be used as the alternative en- zyme source in the studies of inflammatory responses associated with PLA:! of mammalian tissues.
Materials and samples
The lyophilized venom powder was obtained from Biotoxins (St. Cloud, FL, U.S.A.). Cation-exchange resin, TSK CM-650 (S), ammonium acetate, sodium dodecyl-sulphate (SDS) and Coomassie blue were from E. Merck (Darmstadt, F.R.G.). The semipreparative HPLC SynChropak CM-300 column (250 x 10
mm I.D.) and the reversed-pase Cl8 column (300 x 4.0 mm I.D., SynChropak RP-P, 6.5 pm) were from SynChrom (Lafayette, IN, U.S.A.). Molecular mass standards for SDS-PAGE were from Sigma (St. Louis, MO, U.S.A.). Constant- boiling 6 M HCl and 4 A4 methanesulphonic acid containing 0.2% 3-(2-ami- noethyl)indole were from Pierce (Rockford, IL, U.S.A.). High-purity deinonized water was prepared with a Mini-Q purification system (Millipore, Bedford, MA, U.S.A.).
Apparatus and chromatography
Various toxins were first isolated by cation-exchange chromatography on an open column (15 x 2.5 cm I.D.) packed with TSK CM-650 (S). Dissolved venom powder in 0.05 A4 ammonium acetate (pH 5.7) starting buffer (20-50 mg/ml) was applied to a TSK CM-650 (total 5 ml) open column and then eluted in a linear gradient of 0.05-0.5 M ammonium acetate, followed by two stepwise elutions in 0.5 M and 1.0 A4 ammonium acetate (pH 5.9) buffers. For semipreparative sep- aration of venom toxins, an ISCO biocompatible, buffer-impervious HPLC chro- matograph (ISCO, Lincoln, NE, U.S.A.) was used, connected to a CM-300 col- umn adapted with a 250-~1 sample injection coil. A. 100-200 ~1 crude toxin extract was injected each time. The solvent and elution conditions for the HPLC system are described in the figure legend. Reversed-phase HPLC was also carried out with a Hitachi liquid chromatograph (Hitachi, Tokyo, Japan) with a Model L-6200 pump and a variable-wavelength UV monitor. This step was used to purify and desalt the toxin fractions isolated from the ion-exchange chromato- graphies.
Polyacrylamide gel electrophoresis and amino acid analysis
The purities of the isolated toxins were checked by SDS-polyacrylamide slab gel (5% stacking/l4% resolving gel) as described before , with some mod- ifications (5% crosslinking N,N’-methylenebisacrylamide in the gel solution). The amino acid compositions were determined with a Beckman 6300 amino acid analyser and a single-column system based on conventional ion-exchange chro- matography. The special rapid procedure for the preparation of protein hydroly- sates in 6 M HCI or 4 M methanesulphonic acid using microwave irradiation before amino acid analysis was essentially according to the previous reports [4,5].
N-Terminal sequence analysis
The N-terminal sequences of the peak fractions isolated from the HPLC col- umn were carried out by automated Edman degradation with a pulsed-liquid phase protein sequencer (Model 477A, Applied Biosystems, Foster City, CA, U.S.A.). The samples, each containing cu. l-5 nmol of protein, were dissolved in 100 ~1 of 0.1% trifluoroacetic acid (TFA), and 5 ~1 each were used for sequence determinations.
Pharmacological assays for neurotoxicity and PLA2 activity were as described previously . The synthetic substrate L-z-lecithin was used for phospholipase assay by the fatty-acid titration method .
RESULTS AND DISCUSSION
It is well known that various isotoxins or PLAz variants are present in the snake venom even from a single species. Investigation of the biochemical or genetic basis for the generation of multiple toxin isoforms in the same or closely related species remains a great challenge in venom research. It is imperative to isolate and characterize various isoforms of these toxins or enzymes in order to gain some insight into the mechanism underlying the process of sequence var- iation among these proteins. The present study was performed as part of the endeavour to isolate and characterize various PLA2.
Fig. 1 shows the general elution pattern of the crude venom on the TSK CM-650 cation-exchange column. The poorly resolved peaks eluted in the first part of chromatogram represent the PLA2 fractions with strong enzymic activity, and constitute ca. 20% of total venom proteins. The relative percentage yields for each toxin fraction and their respective enzymic or toxic properties were reported
Fig. I, Cation-exchange chromatography on the TSK CM-650(S) column of crude venom from Nuja nq@
siamensis. About 200 mg of lyophilized crude venom dissolved in the starting buffer of 0.05 A4 ammonium
acetate with 0.01% 2-mercaptoethanol (pH 5.7) was applied to the column equilibrated in the same buffer. Elution was carried out in four steps: (A) elution with starting buffer; (B) elution with a linear gradient of 0.05-0.5 M ammonium acetate in 0.01% 2-mercaptoethanol (pH 5.9); (C) 0.5 IM ammonium acetate (pH 5.9); (D) 1.0 M ammonium acetate (pH 5.9). The column eluates (2.8 ml/tube per 3.2 min) were monitored for absorbance at 280 nm. The peak fractions were collected, lyophilized and used for chemical and pharmacological studies. Fraction N indicates the peak of a major long-chain neurotoxin. which consti- tutes ca. 35% of total crude venom. The region labelled as PLA, encompasses the fractions showing the highest phosphohpase A, activity.
Fig. 2. Preparative cation-exchange chromatography of crude venom on the CM-300 column. About 3-5 mg of lyophilized crude venom dissolved in 100-200 ~1 of starting buffer of 0.05 M ammonium acetate with 0.01% 2-mercaptoethanol (pH 5.7) was applied to the column equilibrated in the same buffer. Elution was carried out in three steps: (A) 30 min elution with starting buffer; (B) a linear gradient of ~100% of 1.0 M ammonium acetate CpH 5.7) for 60 min; (C) a final isocratic run of 1.0 M ammonium acetate for 30 min. The flow-rate was set at 1 .O ml/min for the complete cycle, and the fractions were detected with a variable- wavelength UV monitor set at 280 nm (a.u.f.s. 0.512). Fractions 1-5 were collected manually, dried under vacuum and used for SDS-PAGE (Fig. 3) and amino acid analysis (Tables I and II). Fraction N indicates the peak of the same major long-chain neurotoxin as in Fig. 1.
Fig. 3. Gel electrophoresis ofthe fractionated PLA, fractions under denaturing conditions (SDS-PAGE) in the presence of 5 mM dithiothreitol. Lane 6 contains standard proteins used as molecular mass markers (in kDa): transferrin (80), bovine serum albumin (66), ovalbumin (45), carbonic anhydrase (30), soybean trypsin inhibitor (20.1) and lysozyme (14). Lanes l-5 correspond to the five numbered fractions indicated in Fig. 2. The gel was stained with Coomassie blue. The arrows point to the electrophoretic positions of lysozyme and bromophenol blue dye.
previously . The PLA2 fractions require a second chromatographic step before sequence analysis. We have found that these multiple PLAz components are very solubl? in TFA and acetonitrile, and can be easily desalted and purified by re- versed-phase (RP) HPLC without loss of activity (data not shown). At least three distinguishable PLAl with different retention times could be obtained by rechro- matography of the pooled PLA2 fractions on RP-HPLC. These observations prompted us to analyse the crude venom on an HPLC column packed with cation-exchange CM-300, a material similar to TSK CM-650 used in Fig. 1. It is noteworthy that by using cation-exchange HPLC the PLA2 fractions could be separated into at least five well-defined fractions (Fig. 2).
The purities of each fraction were also checked with SDS-PAGE (Fig. 3). The patterns for most fractions are quite clear-cut regarding the homogeneity of each separated PLA2, based on their molecular size. The fraction F-l apparently con- tains several components with molecular masses greater than the typical 14 OOO- 16 000 range for most elapid PLAz identified on SDS-PAGE. Therefore the
AMINO ACID COMPOSITIONS OF FRACTIONS WITH PHOSPOLIPASE A, ACTIVITY
Fractions 2, 3 and 5 correspond to the fractions labelled in Fig. 2. Data are expressed as the number of residues per molecule of protein using alanine as the reference to calculate the residues of other amino acids. Values represent the mean of duplicate determinations. The hydrolysis condition is microwave irradiation for 5 min using 6 M HCI of 4 A4 methanesulphonic acid containing 0.2% 3-(2-aminoethyl) indole. Amino acid F-2 F-3 F-5 1,‘2 cys Asx Thr Ser Glx Pro Gly Ala Val Met Ile Leu Tyr Phe His Lys Arg Trp Total residues 12.8 (14) 22.2 (22) 4.1 (5) 4.9 (5) 7.6 (8) 3.8 (4) 8.8 (9) 11 3.1 (4) 0.8 (1) 4.3 (4) 4.9 (5) 8.7 (9) 3.8 (4) 0.7 (1) 4.8 (5) 5.3 (5) 2.8 (3) (119) 13.1 (14) 21.4 (21) 4.3 (5) 4.2 (5) 7.8 (8) 3.8 (4) 8.9 (9) 11 3.9 (4) 0.6 (1) 4.4 (4) 4.7 (5) 8.6 (9) 3.2 (3) 0.9 (1) 4.9 (5) 5.8 (6) 2.7 (3) (118) 13.4 (14) 21.3 (21) 4.6 (5) 5.3 (6) 7.2 (7) 3.7 (4) 9.3 (9) 11 3.8 (4) 0.9 (1) 4.5 (5) 4.7 (5) 8.7 (9) 3.6 (4) 0.8 (1) 4.7 (5) 5.7 (6) 2.5 (3) (120)
characterization of PLAz is restricted to F-2, F-3 and F-5. Fraction 4 (F-4) was found to be contaminated with a toxin component of lower molecular mass.
The amino acid compositions of F-2, F-3 and F-5 with apparent electropho- retie homogeneity are shown in Table I. They were all shown to be similar to PLAz of Nuja nnja kaoutlzia [Xl, which is supposed to be a species closely related or identical to Baja naja siamensis . It is of interest to note that the amino acid compositions are almost identical, except for some minor variation in the con- tents of aspartic acid, glutamic acid, isoleucine, phenylalanine and arginine. Since F-4 showed a major component of ea. 8 kDa instead of 15 kDa for PLA2, its amino acid composition was determined and compared with that of the major long-chain neurotoxin denoted as Fr. N (Table 11). The slight difference between these two fractions is caused by the contaminant PLA2, as indicated in SDS- PAGE of Fig. 3. Notably this toxin possesses smaller amounts of aromatic amino acids, such as tyrosine and tryptophan, than PLAz (Table I).
The sequence changes among these PLA2 isoforms (F-2, F-3 and F-5) must be small, probably indicating only limited variation of the amino acid residues in
AMINO ACID COMPOSITIONS OF FRACTIONS WITH NEUROTOXICITY
Fractions 4 and N correspond to the fractions labelled m Fig. 2. Data are expressed as the number of residues per molecule of protein using alanine as the reference to calculate the residues of other amino acids. Values represent the mean of duplicate determinations. The hydrolysis condition is microwave irradiation for 5 min using 6 M HCl of 4 M methanesulphonic acid containing 0.2% 3-(2-aminoethyl) indole. Amino acid Fr. N F-4 1;:2 cys Asx Thr Ser Glx Pro GUY Ala Val Met Ile Leu Tyr Phe His Lys Arg Trp Total residues 9.5 (IO) 8.6 (9) 8.5 (9) 2.5 (3) 0.8 (1) 5.6 (6) 3.9 (4) 3 3.8 (4) 0.2 (0) 4.5 (5) 0.9 (1) 0.8 (1) 2.9 (3) 0.8 (1) 4.8 (5) 4.9 (5) 0.8 (1) (71) 9.8 (10) 9.6 (10) 7.6 (8) 2.7 (3) 2.2 (2) 4.7 (5) 4.9 (5) 3 3.8 (4) 0.8 (1) 4.7 (5) 1.7 (2) 1.3 (I) 2.8 (3) 0.6 (1) 4.7 (5) 4.7 (5) 0.5 (I) (74)
their primary structures. The most common one is likely to be the deamidation of asparagine of glutamine to aspartic acid of glutamic acid, respectively. Indeed, N-terminal sequence analysis of these three PLAz by microsequencing indicated only one amino acid difference among the first 20 amino acid, with Asn being identified as the 20th residue in F-5 in contrast to Asp present in F-2 and F-3 (Table III). We have also sequenced neurotoxic F-4 and Fr. N by straightforward Edman degradation: the partial sequences are shown in Table III for comparison. Microsequencing of F-4 revealed two residues at each stap of Edman degrada- tion, which is reflective of the two toxins in the sample as indicated by SDS- PAGE (Fig. 3).
With the advances in HPLC instrumentation and microsequencing, the com- plete sequences of these small toxins and PLA2 could be carried out with the fractions obtained from single-step semipreparative HPLC on the automatic pro- tein sequencer in a relatively short time. The upshot of this report is to establish venom separation and purification by HPLC as a convenient means of toxin isolation, with special regard to obtaining sufficient pure material for sequence
THE AMINO-TERMINAL SEQUENCES OF ISOLATED FRACTIONS FROM CM-300 HPLC
The residues with more than one amino acid denote those positions where more than one phenylthiohy- dantoin derivative was detected by automatic sequencing. Note that F-4 has a composite sequence consist- ing of F-5 and Fr. N in a molar ratio of 1:8, as estimated from the yield ratio of detected amino acids in each step by microsequencing.
Amino acid F-3 F-5 F-4 Fr. N 1 Asn 2 Leu 3 TY~’ 4 Gln 5 Phe 6 Lys 7 Asn 8 Met 9 Be 10 Gln 11 Cys I2 Thr 13 Val 14 Pro 15 Asn 16 Arg 17 Ser 18 Trp 19 Trp 20 Asp Asn Leu Tyr Gln Phe Lys AStl Met Be Gln cys Thr Val Pro Asn Arg Ser Trp Trp Asn AsnIle Leu/Arg TyrCys GlniPhe PheIle Lys,‘Thr AsnjPro Met/Asp Ile Gln;Thr CysiSer Thr/Lys ValiAsp ProjCys Asn;Pro ArgiAsn Ser.‘Gly Trp;His TrpiVal AsnCys Ile Arg Cys Phe Ile Thr Pro Asp Ile Thr Ser Lys Asp cys Pro Asn Gly His Val Cys
analysis. The preparative HPLC method has shortened the overall analysis time for protein separation to less than a few hours for most crude extracts of venom samples. Coupled with current microsequencing techniques, it facilitates the ver- ification of published sequences and the determination of the new enzyme or toxin analogues from the same or closely related snake species.
This work was supported in part by Academia Sinica and the National Science Council, Taipei, Taiwan, Republic of China.
1 C. Y. Lee, in B. Ceccarelli and F. Clementi (Editors). Adwzces in C~topharmacolog~. Vol 3. Raven Press, New York, 1979. pp. 1-16.
2 E. Karlsson, in C. Y. Lee (Editor), Snake Venoms, Hundhook ofExperimental Pharmacology, Springer, Berlin, Vol. 52, 1979, pp. 159-212.
3 U. K. Laemmli, Nature, 227 (1970) 680.
4 S.-H. Chiou and K.-T. Wang, /. Chromatogr., 448 (1988) 404. 5 S.-H. Chiou and K.-T.- Wang, J. Chromatogr., 491 (1989) 424.
6 S.-H. Chiou and W.-W. Lin and W.-P. Chang, Int. /. Pepride Protein Res., 34 (1989) 148.
7 M. A. Wells and D. J. Hanahan, Methods Emymol.. 14 (1969) 178. 8 F. J. Joubert and N. Taljaard. Eur. J. Biochem., 112 (1980) 493.