A COMPARISON OF SOME PROPERTIES OF VICILIN AND LEGUMIN ISOLATED FROM SEEDS OF PI SUM SATIVUM, VICIA FAB A AND CICER ARIETINUM

Transcription

1 New Phytol. (1969) 68, A COMPARISON OF SOME PROPERTIES OF VICILIN AND LEGUMIN ISOLATED FROM SEEDS OF PI SUM SATIVUM, VICIA FAB A AND CICER ARIETINUM BY P. JACKSON, D. BOULTER* AND D. A. THURMAN Hartley Botanical Laboratories, University of Liverpool {Received 29 July 1968) SUMMARY A comparison has been made of the globulins, vicilin and legumin, prepared by isoelectric precipitation, of the seeds of three species of the Vicieae, namely Pisum sativum, Viciafaba and Cicer arietinum. Their amino acid composition, band patterns after electrophoresis upon polacrylamide-urea gels, fingerprint patterns of their tryptic digests and N-terminal amino acids were compared. Considerable portions of the amino acid sequences of vicilin and legumin of Pisum sativum are probably the same or very similar and both proteins consist of a number of sub-units. Viciafaba and Cicer arietinum also contain two proteins homologous with the vicilin and legumin of Pisum sativum. The homologous proteins of the three species resemble each other more closely than do the vicilin and legumin of any one species. INTRODUCTION Osborne (1924) extracted the globulins from pea seeds and separated them into two components, vicilin and legumin. Danielsson (1949) investigated methods for the further purification of these two proteins and found that isoelectric precipitation gave preparations with the least cross-contamination. He also showed that vicilin and legumin had molecular weights of 186,000 and 331,000 respectively using the ultracentrifuge. Since he found proteins with molecular weights similar to those of vicilin and legumin in a large number of different genera within the Leguminoseae, and that they constituted the bulk of the protein present in the seeds examined, he concluded that they were characteristic of the storage globulin fraction of the family. Comparative studies of the amino acid composition of legumin and vicilin of peas have been reported by a number of workers (Yemm, 1958; Goa and Strid, 1959; Grant and Lawrence, 1964). These indicated that both proteins have similar amino acid profiles and contain a high content of aspartic acid, glutamic acid and arginine. A comparison of the N-terminal amino acids of vicilin and legum.in of peas has been made by Gofman and Vaintraub (1961) and Grant and Lawrence (1964). The predominant end-groups of vicilin were found to be aspartic acid and serine; of legumin they were glycine, leucine and threonine. In other comparative studies Vaintraub (1962a, b), Vaintraub, Shutov and Klimenko (1962), showed that the N-terminal amino acids of the legumins of the members of Vicieae resembled each other more closely than the N- terminal amino acids of the corresponding vicilins. A similar conclusion for the vicilins of the Vicieae can be drawn from the N-terminal amino acid data of Vaintraub and his colleagues. * Present address: Department of Botany, University of Durham. 25

2 26 p. JACKSON, D. BOULTER AND D. A. THURMAN The present work was carried out to compare the structures of vicilin and legumin isolated from seeds of Pisum sativum, Vicia faba and Cicer arietinum by means of (i) their amino acid composition; (ii) peptide maps of their tryptic digests, and (iii) by a comparison of the protein band patterns produced after electrophoresis of the two proteins on polyacrylamide gels containing urea (8 M). MATERIALS AND METHODS The source of seeds and extraction procedures have been described previously (Jackson, Milton and Boulter, 1967). Separation of the components of the globulin fraction was carried out on a Sephadex G-200 column (140 x 2.5 cm), packed and equilibrated with the eluting buffer, 0.5 M NaCl-o.i M phosphate, ph 8.0 (Jackson et al, 1967). The proteincontaining fractions (8 ml) were located by measuring their absorption at 280 m^u in a Unicam SP 500 spectrophotometer. Vicilin and legumin were also purified by isoelectric precipitation (Danielsson, 1949); the complete procedure was repeated twice. Amino acid analyses were performed using a Technicon Auto Analyzer (Boulter, 1966) after acid hydrolysis of the proteins using the method of Moore and Stein (1963). Partial enzymic hydrolysis, separation and fingerprinting of peptides have been described previously (Jackson et al., 1967). Polyacrylamide gel electrophoresis was carried out using the method of Ornstein and Davis (1961), with slight modifications. N-terminal amino acid analysis N-terminal amino acids were determined using the 2-4-dinitrofiuorobenzene method. The dinitrophenyl-protein (DNP-protein) obtained was hydrolysed in sealed tubes with 6.0 N HCl for 12 hours at 110 C. The ether-soluble amino acids were removed from the hydrolysate and separated by thin layer chromatography using Selica gel G and the solvent systems described by Brenner, Niederwiesser and Pataki (1964). A standard DNP-amino acid mixture was chromatographed in the same tank at the same time as the sample mixture. After identification of the DNP-amino acids present quantitative determinations were made by eluting the DNP-amino acids from the Silica Gel G with acetone, making up the solutions to a known volume and measuring their absorption at 360 m^ m a Unican SP 500 spectrophotometer using i cm cells. Cahbration curves using DNP-serine, DNP-threonine, DNP-glycine, DNP-leucine, DNP-aspartic acid and DJNP-glutamic acid were constructed. Method of presentation of results Due to inherent variation in the fingerprinting method the same peptides do not necessarily move to exactly the same position in duplicate fingerprint patterns. Therefore mstead of comparing the Rf of the separated peptides, the pattern of the peptide spots in different fingerprint patterns were compared. When a number of spots having approximately the same Rf value formed an identical sequence in two different fingerprint patterns the corresponding spot in each fingerprint pattern were assumed to be due to peptides with the same structure. This assumption is, of course, not necessarilv a]wav<j correct. ' ""i.y»

3 Vicilin and legumin 27 RESULTS Five peaks were obtained when the globulin of Pisum sativum was fractionated on Sephadex G-200 and these have been numbered in order of their elution from the column. Electrophoregrams of samples taken from the leading edge of peak 3, Fig. i (a), demonstrated that they contained only legumin, identified as such by comparing its mobility on polyacrylamide gel with that of legumin prepared by Danielsson's (1949) isoelectric (a) (b) (c) Fig. I. (a) Electrophoregram of a sample taken from the leading edge of peak 3. (b) Electrophoregram of a sample taken from the trailing edge of peak 3. (c) Electrophoregram of peak 2. precipitation method. Samples taken from the trailing edge of this peak when electrophorized were shown to contain mainly vicilin (Fig. ib) with a trace only of legumin. Viciliti was identified by comparing its position after electrophoresis with that of vicilin prepared by isoelectric precipitation. Samples taken from other positions in peak 3 were shown by electrophoresis to contain mixtures of the two proteins. Electrophoregrams of peak 2 (which was very small) possessed two finely stained bands (see Fig. ic) and a faster moving, more diffuse band identified as contaminating legumin; the two finely stained bands also occurred on electrophoregrams of samples of peak 3 (vicilin and legumin) which had been stored at 2 C or which had been freeze-dried. Peak i gave no protein bands on electrophoresis, neither did peaks 4 or 5. Compounds responsible for absorption at 280 rnjx present in peak i are unknown, peaks 4 and 5 contained some peptide and amino acid material. Attempts therefore to obtain pure vicilin and legumin in preparative amounts by gel filtration from pea seeds were unsuccessful and all subsequent work has been done with

4 28 p. JACKSON, D. BOULTER AND D. A. THURMAN vicilin and legumin prepared by isoelectric precipitation, even though this method does not remove all of the legumin from the vicilin preparation. This method has also been used to isolate from the seeds of Vicia faba and Cicer arietinum two proteins with the ( j 5 (p) (q ) (r ) p!f»mi^^7 ^'r t>'''^i"f of electrophoregrams of: (a) vicilin and (c) legumin isolated from Fisum sativum by Danielsson's isoelectric precipitation method; (b) globulin of P sativummetho1"retdibu&f ft^ from r.-aa/^^^ by Danielsson'i fsoelectric pfeciphs method, (e) globulin of V. faba; (g) vicilm and i legumin isolated from Cicer arietinum bv Danie sson s isoelectric precipitation method; (h) globulin of C. ar^etinum. ^ eels /^J;""!, ^' ''^""f f ^I'^^rophoregrams of vicilin, globulin and legumin, in 8 M urea on gels 8 M with respect to urea, oi Ptsum sativum, Vicia faba and Cicer arietinum: (j) vicilin (k) fofv^d^irl^l^ ri"t''^^a7\r'^^""?^'^""^ ^'^'''"' (") g' bulin and(o) legumin"(v.wama) (p) vicilm, (q) globulin and (r) legumin (C!c«r anetrnwrn). J""J, same electrophoretic mobilities as vicilin and legumin of Pisum sativum and in the work described here these proteins are regarded as homologous to the vicilin and legumin of P. sativum (see Fig. 2, d-i). Fig. 2 (c, f and i) shows in addition to the main legumin band two other finely stained bands obtained upon storage or freeze-drying peak ^ from the Sephadex fractionation.

5 Vicilin and legumin 29 The amino acid composition of vicilin and legumin isolated from P. sativum, Viciafaba and Cicer arietinum are given in Tables i and 2. These tables show that glutamic acid and aspartic acid residues predominate in both proteins of all three species, whereas Table i. Amino acid composition of vicilins Amino acids Pisum sativum Vicia faba Cicer arietinum Aspartic Threonine Serine Glutamic Glycine Alanine Valine J Cystine Methionine Iso-leucine i Leucine Tyrosine Phenylalanine Lysine Histidine a Arginine Values are amino acid residues as a percentage of the total number of residues present. Percentage recovery of nitrogen as determined by amino acid recovery as compared with percentage nitrogen determined by Conway microdiflfusion method: Viciafaba 96%, Pisum sativum 96%, Cicer arietinum 97%. Table 2. Amino acid composition of legumins Ammo acids Aspartic Threonine Serine Glutamic Glycine Alanine Valine ^ Cystine Methionine Iso-leucine Leucine Tyrosine Phenylalanine Lysine Histidine Arginine Pisum sativum Vicia faba Cicer arietinum Values are amino acid residues as a percentage of the total number of residues present. Percentage recovery of nitrogen as determined by amino acid recovery as compared with percentage nitrogen by Conway microdiflfusion method: Viciafaba 104%, Pisum sativum gg /o, ' ' Cicer arietinum 94%. cystine and methionine residues are inconspicuous. Legumins isolated from all three species show a very similar amino acid composition to one another as do the vicilins of the three species. By contrast, the vicilin and legumin isolated from any one species show different amino acid compositions. Vicilin, legumin and globulin of Fisum sativum, Vicia faba and Cicer arietinum were fingerprinted and the resulting patterns compared. Table 3 is an analysis of the spot

6 30 P. JACKSON, D. BOULTER AND D. A. THURMAN data and shows that thirty-four of the fifty-eight spots, obtained on fingerprinting the globulin fraction of Pisum sativum, are common to the fingerprint patterns of both vicilin and legumin; also more spots were obtained on fingerprinting legumin than vicilin. Ten of the spots common to the fingerprint patterns of globulin and legumin were not present in the vicilin pattern and eight of the spots common to the globulin and vicilin patterns are not present in the legumin pattern. On a percentage basis 70% of the legximin pattern is present in the vicilin pattern and 80% of the vicilin pattern is present in the legumin pattern. Table 3. Pisum sativum: {a) the number of peptides present in flngerprint patterns of tryptic digests of the globulin fraction, vicilin and legumin; (b) the number of peptides common to globulin, vicilin and legumin (a) Globulin Vicilin Legumin Fingerprint patterns No. of pink spots* (b) Globulin and legumin 42 Globulin and vicilin 40 Globulin, vicilin and legumin 32 Vicilin and legumin 32 Globulin only 3 Vicilin only I Legumin only 2 No. of yellow spots* * Peptides were developed using the technique of Atfield and Morris (1961) Table 4. Vicia faba: (a) the number of peptides in flngerprint patterns of tryptic digests of globulin fraction, vicilin and legumin; (b) the number of peptides common to globulin, vicilin and legumin No. of No. of Fingerprint patterns pink spots yellow spots (a) Globulin 57 4 Vicilin 51 i Legumin 54 4 (b) Globulin and vicilin 40 1 Globulin and legumin 44 4 Globulin, vicilin and legumin 31 1 Vicilin and legumin 31 I Globulin only 3 Vicilin only 11 _ Legumin only 9 _ Table 4 is an analysis of the spot data for Vicia faba and shows that thirty-two of the sixty-one spots, that are obtained on fingerprinting the globulin fraction of V. faba, are common to the fingerprint patterns of both vicilin and legumin. Seventeen of the spots common to the fingerprint patterns of globulin and legumin are not present in the vicilin pattern. Nine of the spots common to the globulin and vicilin patterns are not present in the legumin pattern. Fifty-five percent of the spots present in the legumin pattern are present in the vicilin pattern and 61% of the spots present in the vicilin pattern are also present in the legumin pattern.

7 Vicilin and legumin 31 Table 5 is a similar analysis of the spot data for Cicer arietinum and shows that twentyfive of the fifty-four spots, that are obtained on fingerprinting the globulin of C. arietinum, are common to the patterns of vicilin and legumin. Twenty-two of the spots common to the patterns of globulin and legumin are not present in the vicilin pattern, and two of the spots common to the globulin and vicilin patterns are not present in the legumin pattern. On a percentage basis 69% of the legumin pattern is present in the vicilin pattern and 49% of the vicilin pattern is present in the legumin pattern. Table 5. Cicer arietinum: {a) the number ofpeptides in fingerprint patterns oftryptic digest of globulin fraction, vicilin and legumin; (b) the number ofpeptides common to globulin, vicilin and legumin No. of No of Fingerprint patterns pink spots yellow spots (a) Globulin ji 2 Vicilin 36 o Legumin 48 3 (b) Globulin and legumin 44 3 Globulin and vicilin 27 o Globulin, vicilin and legumin 25 o Vicilin and legumin 25 o Globulin only 5 o Vicilin only 9 O Legumin only 4 o When solutions of the globulin fractions and vicilin and legumin of Fisum sativum, Viciafaba and Cicer arietinum in 8 M urea, were electrophorized on polyacrylamide gels, 8 M with respect to urea, the electrophoretic patterns shown in Fig. 2 were obtained. Of the eleven bands present in the pattern of vicilin of Fisum sativum, seven were also present in the pattern of vicilin patterns of Viciafaba and Cicer arietinum, though these were not the same seven. Thus of the eight bands present in the pattern of Vicia faba, only five were present in that of Cicer arietinum. Of the twelve bands present in the legumin pattern of Fisum sativum, that of Vicia faba contained six and that of Cicer arietinum contained five. Of the ten bands present in the pattern of Viciafaba legumin that of Cicer arietinum contained eight. Six bands were common both to the vicilin and legumin of Fisum sativum; five to the vicilin and legumin of Viciafaba and seven to the vicilin and legumin of Cicer arietinum. The analysis of the N-terminal amino acids of vicilin of Fisum sativum showed that serine, threonine and aspartic acid were predominant, for vicilin of Vicia faba serine, aspartic acid and glutamic acid and for the vicilin of Cicer arietinum serine, aspartic acid and threonine. Analysis of the N-terminal amino acids of legumin showed that in all three species glycine was predominant, followed by leucine and threonine in Fisum sativum, serine, and leucine in Vicia faba and serine, leucine, aspartic acid and threonine in Cicer arietinum. DISCUSSION Osborne's original separation of pea seed globulin into vicilin and legumin, which he characterized by isoelectric precipitation and which Danielsson characterized later by differences in sedimentation coefficient have been further characterized here. c N.p.

8 32 p. JACKSON, D. BOULTER AND D. A. THURMAN If one considers the fingerprint data, and assumes that peptides which occupy similar positions on peptide maps are the same, one is led to the conclusion that there are considerable portions of the amino acid sequence of vicilin and legumin which are the same or very similar. This confirms the work of Kloz and Turkova (1963), who showed 30-40% cross-reaction between vicilin and legumin of peas, using serological techniques. The interpretation of protein band patterns obtained when vicilin and legumin were electrophorized on urea gels is made difficult because the mechanism of urea denaturation of proteins is not fully understood. However, it can be safely concluded that both vicilin and legumin are composed of several sub-units even if the exact number cannot be gauged with accuracy. Grant and Lawrence (1964) obtained six bands when an urea globulin was electrophorized on polyacrylamide gel. In the work described here, however, leaving globulin in contact with 8 M urea for 16 hours gave the greatest number of bands with a reproducible pattern, namely sixteen for pea globulin. It may be that the large number of bands detected may be due to the formation of sulphydryl groups, formed by the enfolding of the protein molecule, leading to polymer formation. However, carboxymethylation of the protein did not markedly reduce the number of bands making such an explanation unlikely. Other mechanisms whereby artifacts arise on urea gels have been investigated by Cole and Mecham (1966) and Brewer (1967) and although in our experiments the gels were not heated during casting and purified urea was used, it is not possible to discount that some of the extra bands obtained on urea gels may be due to the occurrence of artifacts. The determination of the N-terminal amino acids of legumin of Pisum sativum agree qualitatively with the results of other workers for this species (Gofman and Vaintraub, i960; Grant and Lawrence, 1964), in that glycine was the predominant N-terminal amino acid followed by leucine and in our case by threonine. Similarly, the most conspicuous N-terminal amino acid of the legumin of Vicia faba and Ciecr arietinum was glycine, though in the case of the latter species, leucine, serine, aspartic acid and threonine were conspicuous. Serine was the most predominant N-terminal amino acid of the vicilin of all three species and the molar ratios of other amino acids present suggested a greater heterogeneity of poly pep tide chains present than in the legumins. REFERENCES BOULTER, D. (1966). An introduction to automatic amino acid analysis with plant extracts. In: Techniques in Amino acid Anaiysis, p. 93. Technicon Instrument Company Limited, Chertsey, England BRENNER, M., NIEDERWIESSER, A. & PATAKI, G. (1964). Characterisation of protein hydrolysates and amino acids m biological materials by thin layer chromatography. In: New Biochemical Separations (Ed. by A. T. James & L. J. Morris), p Van Nostrand, London. BREWER, J. M. (1967). Artifact produced in disc electrophoresis by ammonium persulphate Science N Y 156,256. f f.,., COLE, E. G. & MECHAM, D. K. (1966). Cyanate formation and electrophoretic behaviour of proteins of gels containing urea. Analyt. Biochem., 14, 215. DANIELSSON, C. E. (1949). Seed globulins of the Gramineae and Leguminosae. Biochem J GOA, J. & STRID, C. (1959). The amino acid content of Leguminous proteins as affected by genetic and nutritional factors. III. Arch. MikrobioL, 33, 253. GOFMAN, Y. Y. & VAINTRAUB, I. A. (1961)-N-terminal amino acids of pealegumin and vicilin..bio^^aimjj'a GRANT, D. R. & LAWRENCE, J. M. (1964). Effects of sodium dodecyl sulphate and other dissociatine reagents on the globulins of peas. Archs Biochem. Biophys., 108, 552. JACKSON, P., MILTON, J. M. & BOULTER, D. (1967). Fingerprint patterns of the globulin fraction obtained from various seeds of the Fabaceae. New PhytoL, 66, 47. KLOZ, J. & TURKOVA, V. (1963). Legumin, vicilin and proteins similar to them in the seeds of some sdecies of the Viciaceae family (a comparative serological study). Biologia PL, 5, 29.

9 Vicilin and legumin 33 MOORE, S. & STEIN, W. H. (1963). Chromatographic determination of amino acids by the use of automatic recording equipment. In: Methods of Emymology, Vol. VI (Ed. by S. P. Colowick & N. O. Kaplan). Academic Press, New York. ORNSTEIN, L. & DAVIS, B. J. (1961). Disc electrophoresis. Preprint by Distillation Product Industries (Eastman Kodak Co.), Rochester, New York. OSBORNE, T. B. (1924). The Vegetable Proteins, 2nd edn. Longmans, Green & Co., London. VAINTRAUB, I. A. (1962a). Globulins of Lathyrus seeds. Tr. pokhim. Prirodn. Soedin. Kishinevsk, Gos. Univ., VAINTRAUB, I. A. (1962b). N-terminal amino acids of the globulins of some pea varieties. Tr. pokhim. Prirodn. Soedin. Kishinevsk. Gos. Univ., 5, 515. VAINTRAUB, I. A., SHUTOV, A. D. & KLIMENKO, V. G. (1962). The globulins of vetch seeds. Biochemistry {U.S.S.R.), 27, 298. YEMM, E. W. (1958). The plant proteins and peptides and their localisation in cells and tissues. In: Encyclopedia of Plant Physiology, Vol. VIII (Ed. by W. Ruhland). Springer, Berlin.

10