7
Carbohydrate Research, 213 (1991) 7-17 Elsevier Science Publishers B.V., Amsterdam
Isolation and characterization of a seed lectin from elderberry (Sambucus nigra L.) and its relationship to the bark lectins* Willy J, Peumans, Jan T. C. Kellens, Anthony K. Allen+, and Els J. M. Van Damme Katholieke
Universiteit
Plantenbeschertning,
Leuven. Fakulteit der Lundbouwwetenschappen, Willem De Croylaan
Laboratorium
voor Fytopathologieen
42, B-3030 Leuven (Belgium)
(Received December 28th, 1989; accepted for publication, March 5th, 1990)
ABSTRACT
A third elderberry (Sombucus nigru L.) lectin (SNA-III) has been isolated from dry seeds by athnity chromatography on immobilized 2-acetamido-2-deoxy-Dgalactose. This lectin is a blood-group, nonspecific glycoprotein containing 21% of carbohydrate, and is rich in asparagine (or aspartic acid), serine, glutamine (or glutamic acid), and glycine. Gel filtration on Superose 12 yielded a single symmetrical peak corresponding to mol. wt. 50 000. SDSpoly(acrylamide) gel (SDS-PAGE) electrophoresis showed a single polypeptide band of 33 kDa, indicating that the native protein is a dimer of identical subunits. Hapteninhibition assays of the agglutination of red blood cells showed that 2-acetamido-2-deoxy-n-galactose is the best inhibitor, being twice as potent as D-galactose, melibiose, and 2-amino-2-deoxy+galactose. A comparison of SNA-111 to the previously described elderberry-bark lectins, SNA-I and SNA-II, indicated that the seed lectin is well distinct from them. INTRODUCTION
In the past, a large number of plant lectins have been purified from seeds (especially legume seeds) and a whole variety of vegetative tissues, such as leaves, stems, bark, bulbs, tubers, rhizomes, and fruits ‘J . Although detailed information about the molecular structure and carbohydrate-binding properties of over 100 such lectins isolated from species belonging to different taxonomic groups exists, their physiological role in the plant remains enigmatic. Moreover, the function of these particular plant proteins becomes even more enigmatic when one realizes that several plants contain two or more lectins having different sugar-binding specificity. Well-known examples are these from Griffonia simplicifoliu (I, II, and IV)~‘, L.aburnum alpinum (I and 11)6, Phaseolus uulgaris (E-PHA and L-PHA)‘, UIex europaeus (I and II)*, Vicia crucc$, Vicia villosa”, and WistariaJloribunda”“*. Another striking example to be added to this list is the elderberry (Sumbucus nigra L.) tree, a member of the plant family Caprifoliaceae. Bark tissue of this tree contains large amounts of two completely different lectins referred to as SNA-I and SNA-II. SNA-I, which was tist isolated13, is a tetrameric * Dedicated to Professors Toshiaki Gsawa and Nathan Sharon. ’ Department of Biochemistry, Charing Cross and Westminster Medical School, Hammersmith, London W6 SRF, United Kingdom. 0008~6215/91/$03.50
@ 1991- Elsevier Science Publishers B.V.
8
w. J. PEUMANS
et al.
glycoprotein (M, 150 000) consisting of two different subunits and recognizes specifically the a-NeuSAc-(2-6)~D-Gal or -D-GalNAc sequence’4. SNA-II, which has been isolated recently15, is a dimeric glycoprotein (M, 68 000) built up of identical subunits and has a binding site most complementary to a-o-GalpNAc-(1+2,3, or 6)-D-Gal disaccharide units. We described herein the isolation of a third elderberry lectin (SNA-III) from dry seeds, and its biochemical characterization and sugar-binding properties. In addition, a comparison is made of the properties of the three elderberry &tins. EXPERIMENTAL
Materials. - Elderberry seeds were prepared from ripe berries which were collected locally. They were extensively washed with tap water, surface sterilized with a 1% solution of hypochlorite (to destroy any contaminating lectin from the juice), again extensively washed with tap water, dried, and stored in the cold until use. Sepharose 4B-GalNAc was preparedI by coupling 2-acetamido-2-deoxy-D-galactose to divinylsulphone-activated Sepharose 4B. The coupling solution contained 10 mg of 2-acetamido-2-deoxy-D-galactose/mL of Sepharose 4B. Phenyl-Sepharose, and columns of Superose 12 and Mono-Q were from Pharmacia LKB Biotechnology Inc. All sugars and glycoproteins used in the hapten-inhibition assays were purchased from Sigma, Chemical Co. Isolationandpurijcation of WA-III. - Dry seeds were powdered in a coffee mill and extracted with petroleum ether (20 mL/g of seeds) to remove excess fat (elderberry seeds are very rich in oil). After extraction, the meal was air-dried and stirred overnight at 4” in phosphate-buffered saline solution (PBS; 1.5mM KH,P04-l~ Na,HPO,, pH 7.4; 3mM KC1;14Om~ NaCl). The homogenate was centriguged (200009) for 20 min, and the supernatant was removed, filtered through filter paper (Whatmann 3MM), and applied onto a column (lo-mL bed volume) of Sepharose 4B-GalNAc. Unbound proteins were eluted with PBS until the A,, fell below 0.0 1, and the lectin was desorbed with 1.2~ D-galactose in PBS. Since the affinity-purified lectin still had a slightly brown color, an additional purification step based on hydrophobic-interaction chromatography was included. Therefore, lectin-containing fractions desorbed from the affinity column were pooled, solid (NH&SO, was added to a final concentration of lo, and the mixture was loaded onto a column (20-mL bed volume) of phenyl-Sepharose equilibrated with M (NH&SO,. Under these conditions, the lectin was retained on the column. After the column had been washed with M (NH&SO, (50 mL), the lectin was desorbed with distilled water, leaving the adsorbed brown contaminants (which have a much higher affinity for phenyl-Sepharose). Finally, the purified lectin was dialyzed against water and lyophilized. A summary of the purification is given in Table I. By use of a combination of affinity chromatography on immobilized 2-acetamido-2-deoxy-D-galactose and hydrophobic-interation chromatography, essentially pure SNA-III was obtained with a high recovery, and a total yield of - 10 mg of lectin per 100 g of seeds.
9
SEED LECTIN PROM ELDERBERRY
TABLE I Purification of elderberry-seed
lectin
Purt&ation step
Volume (mU
Total protein (wl
Extract in PBS’
2000
6210
Affinity chromatography
50
29
Hydrophobic-interactions chromatography
50
25
Agglutination titef
Total agglutinalion wlid
Recovery
5000
100
80
4008
80
80
4000
80
2.5
(%)
’ Highest dilution which still showed agglutination with trypsin-treated, human blood-group A erythrocytes. b An agglutination unit is defined as the amount of l&in required to agglutinate 1 mL of a 1% suspension of trypsin-treated human blood-group A erythrocytes. ’ Defatted meal from 250 g of seeds was Used.
Assays and analytical methods. - Agglutination assays were done in small glass tubes in a final volume of 0.1 mL containing a 1% suspension of erythrocytes (80 pL) and crude extracts or lectin solution (20 ,uL; each serially diluted in two-fold increments). Sodium dodecylsulphate-poly(acrylamide) gel electrophoresis (SDSPAGE) used the discontinuous system described by Laemmli4’ on a 12.5-25% (w/v) gradient gel. Gel filtration was done on a Superose 12 column (type HR lo/30 from Pharmacia) using a Pharmacia FPLC system. Protein was estimated by the Lowry et al.‘* method. Amino acid and carbohydrate were analyzed as described by Van Damme et a1.‘9 Ion-exchange chromatography. - Purified elderberry lectins were analyzed by ion-exchange chromatography using a Pharmacia FPLC system (type GP 250). Lectin samples, dissolved in 2Om~ TrkHCl (pH 8.7), were loaded on a Mono-Q column (type HR 5/5) equilibrated with the same buffer. After the column had been washed with buffer (4 mL), the lectins were eluted with a linear gradient (30 mL) of increasing NaCl concentration (0-0.3~) in this buffer. Production of antiserum. -Antibodies against SNA-I and SNA-II were raised in rabbits immunized with highly purified lectin preparations. The lectins (1 mg in PBS) were emulsified in complete Freund’s adjuvant and injected subcutaneously. Six booster injections were given at IO-day intervals. Ten days after the final injection, blood was collected from an ear marginal vein and allowed to clot before collecting the serum by centrifugation (30009 for 5 min). The antisera were partially purified by repeated precipitation with (NH4),S04 (40% relative saturation). The final precipitate was dissolved in PBS (in a volume equal to that of the original serum) and extensively dialyzed against the same buffer. Double immunodzflwion. - Double immunodiffusion was performed in Petri dishes filled with 1% agarose in PBS containing 4% (w/v) poly(ethylene)glycol(6000), 0.1~ 2-acetamido-2-deoxy-o-galactose, and 2 mg/mL of fetuin hydrolysate, which was
10
W. J. PEUMANS et al.
prepared by adding proteinase K (300 pg/mL) to a solution of 10 mg/mL of fetuin and shaking the mixture for 24 h at 30”. Before use, the hydrolysate was autoclaved to destroy the protease. Both 2-acetamido-2-deoxy-D-galactose and fetuin hydrolysate have to be included to prevent aspecific binding of the lectins to serum proteins (including immunoglobulins) . RESULTS AND DISCUSSION
Occurrence and purification of the elderberry-seed lectin. - The elderberry-seed lectin (SNA-III) was isolated by affinity chromatography on immobilized %-acetamido-2-deoxy-D-galactose, followed by hydrophobic-interaction chromatography on phenyl-Sepharose. As shown in Table I, - 10 mg of pure lectin was obtained from 100 g of seeds with an overall recovery of N 80%. SNA-III represents only - 0.1% of the seed protein. It is apparently much less abundant than the bark lectins, SNA-I and SNA-II, with both represent - 5% of the total bark protein’3~‘5. Molecular structure. - SNA-III was eluted from a Superose 12 column with an apparent mol. mass of 50 kDa (Fig. 1). It is worth mentioning that the elution buffer contained 0.2~ D-galactose to prevent possible interactions of the lectin with the gel matrix. SDS-PAGE of reduced (with 2% B-mercaptoethanol) and unreduced lectin yielded single polypeptide bands of mol. mass of 33 and 31 kDa, respectively (Fig. 2a), which indicated that SNA-III is a dimeric protein built up of two identical subunits. The slightly lower mobility of the reduced lectin, as compared to that of the unreduced lectin, could be the result of the presence of an intramolecular disulphide bound. In this respect, SNA-III resembles SNA-II which is also a dimer composed of two identical subunits of 35 kDa, but differs completely from SNA-I which is much larger (140 kDa), and in addition contains two different subunits of 33 and 35 kDa, respectively (Fig. 2a).
A280
1.0
0.f
-I
11
12 Elution
13
14
15
volume
(mL
16
1
Fig. 1. Gel filtration of SNA-III in a Superose-12 (type 12HR 10/30) column using a Phararacia FPLC system and PBS containing 0.2~ o-galactose as running buffer. The flow rate was 20 mL/h. Molecular mass markers were: (1) Aldolase (158 kDa), (2) bovine serum albumin (67 kDa), (3) ovalbumin (43 kDa), (4) chymotrypsiaogen (25 kDa), and (5) cytochrome c (12.5 kDa). The elution position of these marker proteins was determined in a separate run. SNA-I and SNA-II were also run in the Same column. Their respective elutioa position is indicated by the arrows I and II, respectively.
11
SEED LECTIN FROM ELDERBERRY
RDa
94 67
R
1
2
3
4
5
6
R
kDa 94 67
30
20.1
14.4
Fig. 2. (a) SDS-PAGE of SNA-I, SNA-II, and SNA-III: Unreduced samples (20 pg each) of SNA-I, SNA-II, and SNA-III were loaded on lanes 1,2, and 3, respectively; and reduced (with 2% B-mercaptoethanol) samples of SNA-I, SNA-II, and SNA-III were loaded on lanes 4, 5, and 6, respectively. Lanes R contained molecular-mass markers: lysozyme, 14.4 kDa; soybean trypsin inhibitor, 20.1 kDa; carbonic anhydrase, 30 kDa; ovalbumin, 43 kDa; bovine serum albumin, 67 kDa; and phosphorylase b, 94 kDa. The high-molecular-weight band in lane 1 is typical for unreduced SNA-I”. (b). SDS-PAGE of individual SNA-III isolectins. Lane 1 was loaded with 20 pg of total SNA-III. Samples (20 pg each) of individual isolectins (indicated on Fig. 3) were run on lanes 2-6. All samples were reduced with 2% bmercaptoethanol. Molecular-mass marker proteins are as in (a).
The amino acid composition of SNA III is characterized by high contents of asparagine or aspartic acid, serine, glutamine or glutamic acid, and glycine (Table II). It resembles that of SNA-II reasonably well but differs from the SNA-I amino acid composition. Like the bark &tins, SNA-III is a glycoprotein. However, its carbohydrate content (21%) is higher than that of SNA-I (16%) and SNA-II (7.8%). Multiple molecularforms of SNA-III. -Analysis of SNA-III by high-resolution, ion-exchange chromatography on a Mono Q column revealed the occurrence of multiple molecular forms (Fig. 3). Five individual isolectins were isolated (by repeated chromatography on the Mono Q column) and analyzed by SDS-PAGE and gel
12
W. J. PEUMANS et al.
TABLE II Amino acid“ and carbohydrateb composition of k&ins from elderberry. Amino acid
SNA-F
SNA-Ild
SNA-III
10.9 6.6 10.5 9.5 5.2 7.3 6.3
Phe His LYS Trp ArS
10 0.6 5.0 10 3.0 3.6 1.2 2.1 0 6.4
14.0 6.1 6.5 1.7 4.9 9.0 6.2 2.8 7.1 2.3 6.2 1.2 1.6 2.4 0.2 2.3 c 5.5
16.1 6.6 10.6 10.2 4.1 9.6 5.3 2.8 1.2 0.5 6.3 6.9 0.9 2.1 0.9 1.9 4.4 3.0
Total
100.0
100.0
100.0
Asx Thr Ser GlX pro
GlY Ala l/2 cys Val Met Ile Leu
Tyr
1.1
Sugar
2-Amino-2-deoxyglucose Fucose Mannose Galactose Xylose Glucose Total
3.0 1.0 5.3 0.0 1.0 0.0
4.4 0.5 2.4 0.5 0.0 0.0
2.6 2.7 5.3 0.5 4.1 3.0
16.0
7.8
21.0
’ Residues per 100 residues. * Percent (w/w). ’ Calculated from ref. 13. d Calculated from ref. 15. ’ Not determined.
filtration. The elution of all five isolectins at exactly the same position from the Superose 12 column (results not shown) to yield a single polypeptide band of 33 kDa upon SDS-PAGE (Fig. 2b), implied that the isolectins have the same molecular structure and most likely differ slightly in amino acid composition. It is interesting to note here that SNA-I also consists of a mixture of isolectins, whereas SNA-II could not be resolved in multiple molecular forms (Fig. 3). The elution patterns shown in Fig. 3 were obtained with a column of Mono Q (anion-exchange) resin (which can be used for all three elderberry lectins). Unlike SNA-I and SNA-III, SNA-II binds also to a column of Mono-S (cation-exchange) resin from which it is eluted in a single symmetrical peak, indicating that it represents a single molecular species (results not shown). The two other lectins, SNA-I and SNA-III, do not bind to the Mono S column.
13
WED LECTIN FROM ELDERBERRY
*2l30
SNA-III
SNA-II
SNA-I
/ 0’
0
6
16
24
32 Elution
16
8
0 volume
(mL)
24
L 0
L
1
/
L
Fig. 3. Ion-exchange chromatography of elderberry lectins: Purified SNA-I, SNA-II, and SNA-III were chromatographed on a Mono-Q column using a Pharmacia FPLC system. About 2 mg of each &tin was loaded on the column. The flow rate was 2 mL/min. Peaks from which SNA-III isolectins were purified are indicated.
Carbohydrate-binding specificity and agglutination properties. - Hapten-inhibition assays of the agglutination of human-type A erythrocytes indicated that 2acetamido-2-deoxy-D-galactose is the best inhibitor, being twice a potent as D-galactose, melibiose, and 2-amino$deoxy-D-galactose. Lactose is 2.5fold less inhibitory, and D-fucose and raffinose 5-fold (Table III). All other sugars tested were much less inhibitory. Of the five glycoproteins tested, only mucin was a potent inhibitor of the agglutination activity of SNA-III (Table III). Fetuin, asialofetuin, ovomucoid, and thyroglobulin were hardly inhibitory at all. With respect to its carbohydrate-binding specificity, SNA-III is completely different from SNA-I which recognizes the ctNeuSAc-(2-6)~D-Gal or D-GalNac sequence, and resembles SNA-II in that both SNA-III and SNA-II are best inhibited by 2-acetamido-2-deoxy-D-galactose. It is evident, however, that 2-acetamido-2-deoxy_D-galactose is a much better inhibitor of SNA-II (32-fold) than D-galactose, whereas the difference in inhibitory potency between both sugars is only by a factor of two in the case of SNA-III. Differences in carbohydrate-binding specificity between the three elderberry lectins are also reflected in their agglutination properties with different types of red blood cells. As shown in Table IV, SNA-I has a much higher specific agglutination activity than both SNA-II and SNA-III with human, rabbit, and pigeon erythrocytes. SNA-II and SNA-III exhibited almost equal activities with human and pigeon erythrocytes. However, when tested with rabbit red blood cells, SNA-III was about one order of magnitude less active than SNA-II. Heat andpH-stability. - Since some plant lectins (e.g., the Urtica dioica agglutinirQzOwithstand conditions such as heating or an extreme pH, which lead to complete inactivation of most of the proteins, the heat and pH-stability of the three elderberry
W. J. PEUMANS et cd
14 TABLE III Comparison of the carbohydrate-binding
specificity of the three elderberry lectins
Relative inhibitory potency Sugar
SNA-P
SNA-If
SNA-IIF
D-Galactosed
1 (19mM) 0.44 I.2 1.6 0.7
1 (0.58mar) 1.3 32.2 5.8 1.7
1 (O.SmM) 0.4 2 0.8 1 1 0.4
D-Fucose
n-GalNAc Lactose Melibiose GalN Raffinose Glycoprotein Mucind Fetuin Asialofetuin Ovomucoid Thyroglobulin
1 (250rdW
IQ.5 M3bL)
5 0.5 1 10
1 (5 &mL)
0.0025 0.25 0.005 0.025
< 0.0025 0.05 < 0.0025 < 0.0025
’ Data calculated from ref. 14 except for the glycoproteins. * Data calculated from ref. 15 except for the glycoproteins. ’ Results obtained with hapten-inhibition assays of the agglutination of trypsin-treated, human blood-group A erythrocytes at a lectin concentration of 25 pg/mL. d In parentheses, concentration required for 50% inhibition of precipitation of glycoproteins (for SNA-I and SNA-II) or agglutination (for SNA-III).
TABLE IV Agglutinating activity of SNA-I, SNA-II, and SNA-III in assays with rabbit, pigeon, and human erythrocvtes Erythrocyte source
iUinima1 lectin concentration requiredfor agglutination (pg/mL)” SNA-I
Rabbit Pigeon Human TypeA TypeB Type AB TvneO
SNA-II
SNA-III
u
T
u
T
u
T
1.6 0.2
0.2 0.02
160 160
80 40
10 160
0.7 30
0.4 0.4 0.4 0.4
0.2 0.2 0.2 0.2
80 160 160 40
10 20 20 10
80 40 40 30
5 10 5 5
’ Untreated (U) and trypsin-treated
(T) erythrocytes were used.
15
SEED LECTIN PROM ELDERBERRY
0
2
4
6
6
10
12
14
PH
Fig. 4. Stability of elderberry lectins under different conditions of pH. Portions of SNA-I (-O-O-), SNA-II (-&A-), and SNA-III (-l-J-O--) solutions (0.1 mg/mL in PBS) were adjusted to different pH values (by adding 0.1~ HCl or 0.1~ NaOH). After being kept for 1 h at room temperature, samples were adjusted to 0.1~ TrisHCl (pH 7.4) and assayed For agglutination activity. Results are expressed as percentages of the control (samples kept in PBS) value.
-.- .--:- _*/ Fig. 5. Double immunodiFussion of purified elderberry leetins against SNA-I and SNA-II antiserum. Antisera against SNA-I (A, central well) and SNA-II (B, central well) werechalIenged with S pg of SNA-I (I), SNA-II (II), and SNA-III (III). As a control, preimmune serum (C, central well) was also challenged with the three lectins. After formation of the preeipitin lines, unprecipitated proteins were eluted by washings the gels with PBS overnight. The gels were Fixed and stained with Coomassie Blue as For SDS-poly(acrylamide) gels. All experiments were done with total lectin preparations (not with single isolectins oFSNA-I and SNA-III).
16
W. J. PEUMANS
et al.
lectins were studied in some detail. Thermal-stability assays indicated that both bark lectins as well as the seed lectin are rapidly inactivated when heated to temperatures above 50” (virtually no activity was left after heating at 70” for 10 min) (results not shown). A comparison of the pH stability, however, revealed some striking differences. Indeed, whereas SNA-I and SNA-III are stable only between pH 5 and 12, SNA-II appeared to be very acid resistant (down to pH 0.8) (Fig. 4). Serological relationshipbetween the elderberry lectins.-To test possible serological relationships between the elderberry lectins, they were challenged with antisera against SNA-I and SNA-II in double-immunodiffusion assays, As shown in Fig. 5, antiserum against SNA-I did not react with either SNA-II or SNA-III. However, when antiserum against SNA-II was used, all three lectins formed precipitin lines. Figure 5 indicates that the seed lectin cross-reacted strongly with SNA-II, whereas SNA-I reacted only weakly with the antiserum against SNA-II. These observations suggest that SNA-III and SNA-II are much more closely related serologically than SNA-I and SNA-II, or SNA-I and SNA-III. As a control, the three lectins were challenged with preimmune serum. No precipitin lines could be detected, indicating that the serological reaction was specific. The observed serological relationships are in good agreement with the results described above. Indeed, SNA-I differs totally from SNA-II and SNA-III with respect to its molecular structure and carbohydrate-binding specificity, whereas SNA-II and SNA-III have a similar molecular structure and, in addition, can both be considered as 2-acetamido-2-deoxy-o-galactose specific lectins. It appears, therefore, that the elderberry-seed lectin is much more closely related to bark lectin SNA-II than both bark &tins are related to each other. ACKNOWLEDGEMENTS
This work was supported by grants from “De Nationale Bank” and the National Fund for Scientific Research (Belgium). W. J. P. is Senior Research Associate, and E. J. M. V. D. Research Assistant of this Fund. J. T. C. K. received a fellowship from the Belgian Instituut tot Aanmoediging van het Wetenschappelijk Onderzoek in Nijverheid en Landbouw. REFERENCES 1 I. J. Goldstein, in I. E. Liener, N. Sharon, and I. J. Goldstein @Is.), The L.ecfins. Properties, Functions, and Applications in Biology and Medicine, Academic Press, Orlando, 1986, pp. 35-264. 2 H. Rudiger, Ah. L.ectin Res., 1 (1988) 26-72. 3 C. E. Hayes and 1. J. Goldstein, J. Biol. Chem., 249 (1974) 1904-1914. 4 P. N. S. Iyer, K. D. Wilkinson, and I. J. Goldstein, Arch. Biochem. Biophys., 177 (1976) 33&333. 5 S. Shibata, I. J. Goldstein, and D. A. Baker, J. Biol. Chem., 257 (1982) 9324-9329. 6 Y. Konami, K. Yamamoto, T. Tsuji, I. Matsumoto, and T. Osawa, Hoppe-Seyler’s Z. Physiof. Chem., 364 (1983) 397-405. 7 D. A. Rigas and C. Head, B&hem. Biophys. Res. Commtm., 34 (1969) 633-639. 8 I. Matsumoto and T. Osawa, Biochem. Biophys. Acta, 194 (1969) 180-189. 9 H. Rudiger, Eur. J. Biochem., 72 (1977) 317-322.
SEED LECTIN FROM ELDERBERRY
17
10 S. E. Tollefsen and R. Komfeld, J. Biol. Gem., 258 (1983) 5165-5171. 11 B. E. Barker and P. Fames, Nufure (London), 215 (1967) 6%660. 12 S. Toyoshima, Y. Akiyama, K. Nakano, A. Tonomura, and T. Osawa, Biochemistry, 10 (1971) 44574463. 13 W. F. Broekaert, M. Nsimba-Lubaki, B. Peeters, and W. J. Peumans, Biochem. J., 221(1984) 163169. 14 N. Shibuya, I. J. Goldstein, W. F. Broekaert, M. Nsimba-Lubaki, B. Peeters, and W. J. Peumans, J. Biol. Chem., 262 (1987) 19561601. 15 H. Kaku, W. J. Peumans, and I. 3. Goldstein, Arch. Biochem. Biophys., (1990) 25S262. 16 H. J. Allen and E. A. Z. Johnson, Carbohydr. Res., 58 (1977) 253-265. 17 U. K. Laemmli, Nature (London) 277 (1970) 680-685. 18 0. H. Lowry, N. J. Rosenbrough, N. J. Farr and R. J. Randall, J. Biol. C/tern., 193 (1951) 265-275. 19 E. J. M. Van Damme, A. K. Allen, and W. J. Peumans, Physiol. Plant., 73 (1988) 52-57. 20 W. J. Peumans, M. De Ley, and W. F. Broekaert, FEBS Letr., 177 (1984) 99-103.
Carbohydrate Research, 213 (1991) 7-17 Elsevier Science Publishers B.V., Amsterdam
Isolation and characterization of a seed lectin from elderberry (Sambucus nigra L.) and its relationship to the bark lectins* Willy J, Peumans, Jan T. C. Kellens, Anthony K. Allen+, and Els J. M. Van Damme Katholieke
Universiteit
Plantenbeschertning,
Leuven. Fakulteit der Lundbouwwetenschappen, Willem De Croylaan
Laboratorium
voor Fytopathologieen
42, B-3030 Leuven (Belgium)
(Received December 28th, 1989; accepted for publication, March 5th, 1990)
ABSTRACT
A third elderberry (Sombucus nigru L.) lectin (SNA-III) has been isolated from dry seeds by athnity chromatography on immobilized 2-acetamido-2-deoxy-Dgalactose. This lectin is a blood-group, nonspecific glycoprotein containing 21% of carbohydrate, and is rich in asparagine (or aspartic acid), serine, glutamine (or glutamic acid), and glycine. Gel filtration on Superose 12 yielded a single symmetrical peak corresponding to mol. wt. 50 000. SDSpoly(acrylamide) gel (SDS-PAGE) electrophoresis showed a single polypeptide band of 33 kDa, indicating that the native protein is a dimer of identical subunits. Hapteninhibition assays of the agglutination of red blood cells showed that 2-acetamido-2-deoxy-n-galactose is the best inhibitor, being twice as potent as D-galactose, melibiose, and 2-amino-2-deoxy+galactose. A comparison of SNA-111 to the previously described elderberry-bark lectins, SNA-I and SNA-II, indicated that the seed lectin is well distinct from them. INTRODUCTION
In the past, a large number of plant lectins have been purified from seeds (especially legume seeds) and a whole variety of vegetative tissues, such as leaves, stems, bark, bulbs, tubers, rhizomes, and fruits ‘J . Although detailed information about the molecular structure and carbohydrate-binding properties of over 100 such lectins isolated from species belonging to different taxonomic groups exists, their physiological role in the plant remains enigmatic. Moreover, the function of these particular plant proteins becomes even more enigmatic when one realizes that several plants contain two or more lectins having different sugar-binding specificity. Well-known examples are these from Griffonia simplicifoliu (I, II, and IV)~‘, L.aburnum alpinum (I and 11)6, Phaseolus uulgaris (E-PHA and L-PHA)‘, UIex europaeus (I and II)*, Vicia crucc$, Vicia villosa”, and WistariaJloribunda”“*. Another striking example to be added to this list is the elderberry (Sumbucus nigra L.) tree, a member of the plant family Caprifoliaceae. Bark tissue of this tree contains large amounts of two completely different lectins referred to as SNA-I and SNA-II. SNA-I, which was tist isolated13, is a tetrameric * Dedicated to Professors Toshiaki Gsawa and Nathan Sharon. ’ Department of Biochemistry, Charing Cross and Westminster Medical School, Hammersmith, London W6 SRF, United Kingdom. 0008~6215/91/$03.50
@ 1991- Elsevier Science Publishers B.V.
8
w. J. PEUMANS
et al.
glycoprotein (M, 150 000) consisting of two different subunits and recognizes specifically the a-NeuSAc-(2-6)~D-Gal or -D-GalNAc sequence’4. SNA-II, which has been isolated recently15, is a dimeric glycoprotein (M, 68 000) built up of identical subunits and has a binding site most complementary to a-o-GalpNAc-(1+2,3, or 6)-D-Gal disaccharide units. We described herein the isolation of a third elderberry lectin (SNA-III) from dry seeds, and its biochemical characterization and sugar-binding properties. In addition, a comparison is made of the properties of the three elderberry &tins. EXPERIMENTAL
Materials. - Elderberry seeds were prepared from ripe berries which were collected locally. They were extensively washed with tap water, surface sterilized with a 1% solution of hypochlorite (to destroy any contaminating lectin from the juice), again extensively washed with tap water, dried, and stored in the cold until use. Sepharose 4B-GalNAc was preparedI by coupling 2-acetamido-2-deoxy-D-galactose to divinylsulphone-activated Sepharose 4B. The coupling solution contained 10 mg of 2-acetamido-2-deoxy-D-galactose/mL of Sepharose 4B. Phenyl-Sepharose, and columns of Superose 12 and Mono-Q were from Pharmacia LKB Biotechnology Inc. All sugars and glycoproteins used in the hapten-inhibition assays were purchased from Sigma, Chemical Co. Isolationandpurijcation of WA-III. - Dry seeds were powdered in a coffee mill and extracted with petroleum ether (20 mL/g of seeds) to remove excess fat (elderberry seeds are very rich in oil). After extraction, the meal was air-dried and stirred overnight at 4” in phosphate-buffered saline solution (PBS; 1.5mM KH,P04-l~ Na,HPO,, pH 7.4; 3mM KC1;14Om~ NaCl). The homogenate was centriguged (200009) for 20 min, and the supernatant was removed, filtered through filter paper (Whatmann 3MM), and applied onto a column (lo-mL bed volume) of Sepharose 4B-GalNAc. Unbound proteins were eluted with PBS until the A,, fell below 0.0 1, and the lectin was desorbed with 1.2~ D-galactose in PBS. Since the affinity-purified lectin still had a slightly brown color, an additional purification step based on hydrophobic-interaction chromatography was included. Therefore, lectin-containing fractions desorbed from the affinity column were pooled, solid (NH&SO, was added to a final concentration of lo, and the mixture was loaded onto a column (20-mL bed volume) of phenyl-Sepharose equilibrated with M (NH&SO,. Under these conditions, the lectin was retained on the column. After the column had been washed with M (NH&SO, (50 mL), the lectin was desorbed with distilled water, leaving the adsorbed brown contaminants (which have a much higher affinity for phenyl-Sepharose). Finally, the purified lectin was dialyzed against water and lyophilized. A summary of the purification is given in Table I. By use of a combination of affinity chromatography on immobilized 2-acetamido-2-deoxy-D-galactose and hydrophobic-interation chromatography, essentially pure SNA-III was obtained with a high recovery, and a total yield of - 10 mg of lectin per 100 g of seeds.
9
SEED LECTIN PROM ELDERBERRY
TABLE I Purification of elderberry-seed
lectin
Purt&ation step
Volume (mU
Total protein (wl
Extract in PBS’
2000
6210
Affinity chromatography
50
29
Hydrophobic-interactions chromatography
50
25
Agglutination titef
Total agglutinalion wlid
Recovery
5000
100
80
4008
80
80
4000
80
2.5
(%)
’ Highest dilution which still showed agglutination with trypsin-treated, human blood-group A erythrocytes. b An agglutination unit is defined as the amount of l&in required to agglutinate 1 mL of a 1% suspension of trypsin-treated human blood-group A erythrocytes. ’ Defatted meal from 250 g of seeds was Used.
Assays and analytical methods. - Agglutination assays were done in small glass tubes in a final volume of 0.1 mL containing a 1% suspension of erythrocytes (80 pL) and crude extracts or lectin solution (20 ,uL; each serially diluted in two-fold increments). Sodium dodecylsulphate-poly(acrylamide) gel electrophoresis (SDSPAGE) used the discontinuous system described by Laemmli4’ on a 12.5-25% (w/v) gradient gel. Gel filtration was done on a Superose 12 column (type HR lo/30 from Pharmacia) using a Pharmacia FPLC system. Protein was estimated by the Lowry et al.‘* method. Amino acid and carbohydrate were analyzed as described by Van Damme et a1.‘9 Ion-exchange chromatography. - Purified elderberry lectins were analyzed by ion-exchange chromatography using a Pharmacia FPLC system (type GP 250). Lectin samples, dissolved in 2Om~ TrkHCl (pH 8.7), were loaded on a Mono-Q column (type HR 5/5) equilibrated with the same buffer. After the column had been washed with buffer (4 mL), the lectins were eluted with a linear gradient (30 mL) of increasing NaCl concentration (0-0.3~) in this buffer. Production of antiserum. -Antibodies against SNA-I and SNA-II were raised in rabbits immunized with highly purified lectin preparations. The lectins (1 mg in PBS) were emulsified in complete Freund’s adjuvant and injected subcutaneously. Six booster injections were given at IO-day intervals. Ten days after the final injection, blood was collected from an ear marginal vein and allowed to clot before collecting the serum by centrifugation (30009 for 5 min). The antisera were partially purified by repeated precipitation with (NH4),S04 (40% relative saturation). The final precipitate was dissolved in PBS (in a volume equal to that of the original serum) and extensively dialyzed against the same buffer. Double immunodzflwion. - Double immunodiffusion was performed in Petri dishes filled with 1% agarose in PBS containing 4% (w/v) poly(ethylene)glycol(6000), 0.1~ 2-acetamido-2-deoxy-o-galactose, and 2 mg/mL of fetuin hydrolysate, which was
10
W. J. PEUMANS et al.
prepared by adding proteinase K (300 pg/mL) to a solution of 10 mg/mL of fetuin and shaking the mixture for 24 h at 30”. Before use, the hydrolysate was autoclaved to destroy the protease. Both 2-acetamido-2-deoxy-D-galactose and fetuin hydrolysate have to be included to prevent aspecific binding of the lectins to serum proteins (including immunoglobulins) . RESULTS AND DISCUSSION
Occurrence and purification of the elderberry-seed lectin. - The elderberry-seed lectin (SNA-III) was isolated by affinity chromatography on immobilized %-acetamido-2-deoxy-D-galactose, followed by hydrophobic-interaction chromatography on phenyl-Sepharose. As shown in Table I, - 10 mg of pure lectin was obtained from 100 g of seeds with an overall recovery of N 80%. SNA-III represents only - 0.1% of the seed protein. It is apparently much less abundant than the bark lectins, SNA-I and SNA-II, with both represent - 5% of the total bark protein’3~‘5. Molecular structure. - SNA-III was eluted from a Superose 12 column with an apparent mol. mass of 50 kDa (Fig. 1). It is worth mentioning that the elution buffer contained 0.2~ D-galactose to prevent possible interactions of the lectin with the gel matrix. SDS-PAGE of reduced (with 2% B-mercaptoethanol) and unreduced lectin yielded single polypeptide bands of mol. mass of 33 and 31 kDa, respectively (Fig. 2a), which indicated that SNA-III is a dimeric protein built up of two identical subunits. The slightly lower mobility of the reduced lectin, as compared to that of the unreduced lectin, could be the result of the presence of an intramolecular disulphide bound. In this respect, SNA-III resembles SNA-II which is also a dimer composed of two identical subunits of 35 kDa, but differs completely from SNA-I which is much larger (140 kDa), and in addition contains two different subunits of 33 and 35 kDa, respectively (Fig. 2a).
A280
1.0
0.f
-I
11
12 Elution
13
14
15
volume
(mL
16
1
Fig. 1. Gel filtration of SNA-III in a Superose-12 (type 12HR 10/30) column using a Phararacia FPLC system and PBS containing 0.2~ o-galactose as running buffer. The flow rate was 20 mL/h. Molecular mass markers were: (1) Aldolase (158 kDa), (2) bovine serum albumin (67 kDa), (3) ovalbumin (43 kDa), (4) chymotrypsiaogen (25 kDa), and (5) cytochrome c (12.5 kDa). The elution position of these marker proteins was determined in a separate run. SNA-I and SNA-II were also run in the Same column. Their respective elutioa position is indicated by the arrows I and II, respectively.
11
SEED LECTIN FROM ELDERBERRY
RDa
94 67
R
1
2
3
4
5
6
R
kDa 94 67
30
20.1
14.4
Fig. 2. (a) SDS-PAGE of SNA-I, SNA-II, and SNA-III: Unreduced samples (20 pg each) of SNA-I, SNA-II, and SNA-III were loaded on lanes 1,2, and 3, respectively; and reduced (with 2% B-mercaptoethanol) samples of SNA-I, SNA-II, and SNA-III were loaded on lanes 4, 5, and 6, respectively. Lanes R contained molecular-mass markers: lysozyme, 14.4 kDa; soybean trypsin inhibitor, 20.1 kDa; carbonic anhydrase, 30 kDa; ovalbumin, 43 kDa; bovine serum albumin, 67 kDa; and phosphorylase b, 94 kDa. The high-molecular-weight band in lane 1 is typical for unreduced SNA-I”. (b). SDS-PAGE of individual SNA-III isolectins. Lane 1 was loaded with 20 pg of total SNA-III. Samples (20 pg each) of individual isolectins (indicated on Fig. 3) were run on lanes 2-6. All samples were reduced with 2% bmercaptoethanol. Molecular-mass marker proteins are as in (a).
The amino acid composition of SNA III is characterized by high contents of asparagine or aspartic acid, serine, glutamine or glutamic acid, and glycine (Table II). It resembles that of SNA-II reasonably well but differs from the SNA-I amino acid composition. Like the bark &tins, SNA-III is a glycoprotein. However, its carbohydrate content (21%) is higher than that of SNA-I (16%) and SNA-II (7.8%). Multiple molecularforms of SNA-III. -Analysis of SNA-III by high-resolution, ion-exchange chromatography on a Mono Q column revealed the occurrence of multiple molecular forms (Fig. 3). Five individual isolectins were isolated (by repeated chromatography on the Mono Q column) and analyzed by SDS-PAGE and gel
12
W. J. PEUMANS et al.
TABLE II Amino acid“ and carbohydrateb composition of k&ins from elderberry. Amino acid
SNA-F
SNA-Ild
SNA-III
10.9 6.6 10.5 9.5 5.2 7.3 6.3
Phe His LYS Trp ArS
10 0.6 5.0 10 3.0 3.6 1.2 2.1 0 6.4
14.0 6.1 6.5 1.7 4.9 9.0 6.2 2.8 7.1 2.3 6.2 1.2 1.6 2.4 0.2 2.3 c 5.5
16.1 6.6 10.6 10.2 4.1 9.6 5.3 2.8 1.2 0.5 6.3 6.9 0.9 2.1 0.9 1.9 4.4 3.0
Total
100.0
100.0
100.0
Asx Thr Ser GlX pro
GlY Ala l/2 cys Val Met Ile Leu
Tyr
1.1
Sugar
2-Amino-2-deoxyglucose Fucose Mannose Galactose Xylose Glucose Total
3.0 1.0 5.3 0.0 1.0 0.0
4.4 0.5 2.4 0.5 0.0 0.0
2.6 2.7 5.3 0.5 4.1 3.0
16.0
7.8
21.0
’ Residues per 100 residues. * Percent (w/w). ’ Calculated from ref. 13. d Calculated from ref. 15. ’ Not determined.
filtration. The elution of all five isolectins at exactly the same position from the Superose 12 column (results not shown) to yield a single polypeptide band of 33 kDa upon SDS-PAGE (Fig. 2b), implied that the isolectins have the same molecular structure and most likely differ slightly in amino acid composition. It is interesting to note here that SNA-I also consists of a mixture of isolectins, whereas SNA-II could not be resolved in multiple molecular forms (Fig. 3). The elution patterns shown in Fig. 3 were obtained with a column of Mono Q (anion-exchange) resin (which can be used for all three elderberry lectins). Unlike SNA-I and SNA-III, SNA-II binds also to a column of Mono-S (cation-exchange) resin from which it is eluted in a single symmetrical peak, indicating that it represents a single molecular species (results not shown). The two other lectins, SNA-I and SNA-III, do not bind to the Mono S column.
13
WED LECTIN FROM ELDERBERRY
*2l30
SNA-III
SNA-II
SNA-I
/ 0’
0
6
16
24
32 Elution
16
8
0 volume
(mL)
24
L 0
L
1
/
L
Fig. 3. Ion-exchange chromatography of elderberry lectins: Purified SNA-I, SNA-II, and SNA-III were chromatographed on a Mono-Q column using a Pharmacia FPLC system. About 2 mg of each &tin was loaded on the column. The flow rate was 2 mL/min. Peaks from which SNA-III isolectins were purified are indicated.
Carbohydrate-binding specificity and agglutination properties. - Hapten-inhibition assays of the agglutination of human-type A erythrocytes indicated that 2acetamido-2-deoxy-D-galactose is the best inhibitor, being twice a potent as D-galactose, melibiose, and 2-amino$deoxy-D-galactose. Lactose is 2.5fold less inhibitory, and D-fucose and raffinose 5-fold (Table III). All other sugars tested were much less inhibitory. Of the five glycoproteins tested, only mucin was a potent inhibitor of the agglutination activity of SNA-III (Table III). Fetuin, asialofetuin, ovomucoid, and thyroglobulin were hardly inhibitory at all. With respect to its carbohydrate-binding specificity, SNA-III is completely different from SNA-I which recognizes the ctNeuSAc-(2-6)~D-Gal or D-GalNac sequence, and resembles SNA-II in that both SNA-III and SNA-II are best inhibited by 2-acetamido-2-deoxy-D-galactose. It is evident, however, that 2-acetamido-2-deoxy_D-galactose is a much better inhibitor of SNA-II (32-fold) than D-galactose, whereas the difference in inhibitory potency between both sugars is only by a factor of two in the case of SNA-III. Differences in carbohydrate-binding specificity between the three elderberry lectins are also reflected in their agglutination properties with different types of red blood cells. As shown in Table IV, SNA-I has a much higher specific agglutination activity than both SNA-II and SNA-III with human, rabbit, and pigeon erythrocytes. SNA-II and SNA-III exhibited almost equal activities with human and pigeon erythrocytes. However, when tested with rabbit red blood cells, SNA-III was about one order of magnitude less active than SNA-II. Heat andpH-stability. - Since some plant lectins (e.g., the Urtica dioica agglutinirQzOwithstand conditions such as heating or an extreme pH, which lead to complete inactivation of most of the proteins, the heat and pH-stability of the three elderberry
W. J. PEUMANS et cd
14 TABLE III Comparison of the carbohydrate-binding
specificity of the three elderberry lectins
Relative inhibitory potency Sugar
SNA-P
SNA-If
SNA-IIF
D-Galactosed
1 (19mM) 0.44 I.2 1.6 0.7
1 (0.58mar) 1.3 32.2 5.8 1.7
1 (O.SmM) 0.4 2 0.8 1 1 0.4
D-Fucose
n-GalNAc Lactose Melibiose GalN Raffinose Glycoprotein Mucind Fetuin Asialofetuin Ovomucoid Thyroglobulin
1 (250rdW
IQ.5 M3bL)
5 0.5 1 10
1 (5 &mL)
0.0025 0.25 0.005 0.025
< 0.0025 0.05 < 0.0025 < 0.0025
’ Data calculated from ref. 14 except for the glycoproteins. * Data calculated from ref. 15 except for the glycoproteins. ’ Results obtained with hapten-inhibition assays of the agglutination of trypsin-treated, human blood-group A erythrocytes at a lectin concentration of 25 pg/mL. d In parentheses, concentration required for 50% inhibition of precipitation of glycoproteins (for SNA-I and SNA-II) or agglutination (for SNA-III).
TABLE IV Agglutinating activity of SNA-I, SNA-II, and SNA-III in assays with rabbit, pigeon, and human erythrocvtes Erythrocyte source
iUinima1 lectin concentration requiredfor agglutination (pg/mL)” SNA-I
Rabbit Pigeon Human TypeA TypeB Type AB TvneO
SNA-II
SNA-III
u
T
u
T
u
T
1.6 0.2
0.2 0.02
160 160
80 40
10 160
0.7 30
0.4 0.4 0.4 0.4
0.2 0.2 0.2 0.2
80 160 160 40
10 20 20 10
80 40 40 30
5 10 5 5
’ Untreated (U) and trypsin-treated
(T) erythrocytes were used.
15
SEED LECTIN PROM ELDERBERRY
0
2
4
6
6
10
12
14
PH
Fig. 4. Stability of elderberry lectins under different conditions of pH. Portions of SNA-I (-O-O-), SNA-II (-&A-), and SNA-III (-l-J-O--) solutions (0.1 mg/mL in PBS) were adjusted to different pH values (by adding 0.1~ HCl or 0.1~ NaOH). After being kept for 1 h at room temperature, samples were adjusted to 0.1~ TrisHCl (pH 7.4) and assayed For agglutination activity. Results are expressed as percentages of the control (samples kept in PBS) value.
-.- .--:- _*/ Fig. 5. Double immunodiFussion of purified elderberry leetins against SNA-I and SNA-II antiserum. Antisera against SNA-I (A, central well) and SNA-II (B, central well) werechalIenged with S pg of SNA-I (I), SNA-II (II), and SNA-III (III). As a control, preimmune serum (C, central well) was also challenged with the three lectins. After formation of the preeipitin lines, unprecipitated proteins were eluted by washings the gels with PBS overnight. The gels were Fixed and stained with Coomassie Blue as For SDS-poly(acrylamide) gels. All experiments were done with total lectin preparations (not with single isolectins oFSNA-I and SNA-III).
16
W. J. PEUMANS
et al.
lectins were studied in some detail. Thermal-stability assays indicated that both bark lectins as well as the seed lectin are rapidly inactivated when heated to temperatures above 50” (virtually no activity was left after heating at 70” for 10 min) (results not shown). A comparison of the pH stability, however, revealed some striking differences. Indeed, whereas SNA-I and SNA-III are stable only between pH 5 and 12, SNA-II appeared to be very acid resistant (down to pH 0.8) (Fig. 4). Serological relationshipbetween the elderberry lectins.-To test possible serological relationships between the elderberry lectins, they were challenged with antisera against SNA-I and SNA-II in double-immunodiffusion assays, As shown in Fig. 5, antiserum against SNA-I did not react with either SNA-II or SNA-III. However, when antiserum against SNA-II was used, all three lectins formed precipitin lines. Figure 5 indicates that the seed lectin cross-reacted strongly with SNA-II, whereas SNA-I reacted only weakly with the antiserum against SNA-II. These observations suggest that SNA-III and SNA-II are much more closely related serologically than SNA-I and SNA-II, or SNA-I and SNA-III. As a control, the three lectins were challenged with preimmune serum. No precipitin lines could be detected, indicating that the serological reaction was specific. The observed serological relationships are in good agreement with the results described above. Indeed, SNA-I differs totally from SNA-II and SNA-III with respect to its molecular structure and carbohydrate-binding specificity, whereas SNA-II and SNA-III have a similar molecular structure and, in addition, can both be considered as 2-acetamido-2-deoxy-o-galactose specific lectins. It appears, therefore, that the elderberry-seed lectin is much more closely related to bark lectin SNA-II than both bark &tins are related to each other. ACKNOWLEDGEMENTS
This work was supported by grants from “De Nationale Bank” and the National Fund for Scientific Research (Belgium). W. J. P. is Senior Research Associate, and E. J. M. V. D. Research Assistant of this Fund. J. T. C. K. received a fellowship from the Belgian Instituut tot Aanmoediging van het Wetenschappelijk Onderzoek in Nijverheid en Landbouw. REFERENCES 1 I. J. Goldstein, in I. E. Liener, N. Sharon, and I. J. Goldstein @Is.), The L.ecfins. Properties, Functions, and Applications in Biology and Medicine, Academic Press, Orlando, 1986, pp. 35-264. 2 H. Rudiger, Ah. L.ectin Res., 1 (1988) 26-72. 3 C. E. Hayes and 1. J. Goldstein, J. Biol. Chem., 249 (1974) 1904-1914. 4 P. N. S. Iyer, K. D. Wilkinson, and I. J. Goldstein, Arch. Biochem. Biophys., 177 (1976) 33&333. 5 S. Shibata, I. J. Goldstein, and D. A. Baker, J. Biol. Chem., 257 (1982) 9324-9329. 6 Y. Konami, K. Yamamoto, T. Tsuji, I. Matsumoto, and T. Osawa, Hoppe-Seyler’s Z. Physiof. Chem., 364 (1983) 397-405. 7 D. A. Rigas and C. Head, B&hem. Biophys. Res. Commtm., 34 (1969) 633-639. 8 I. Matsumoto and T. Osawa, Biochem. Biophys. Acta, 194 (1969) 180-189. 9 H. Rudiger, Eur. J. Biochem., 72 (1977) 317-322.
SEED LECTIN FROM ELDERBERRY
17
10 S. E. Tollefsen and R. Komfeld, J. Biol. Gem., 258 (1983) 5165-5171. 11 B. E. Barker and P. Fames, Nufure (London), 215 (1967) 6%660. 12 S. Toyoshima, Y. Akiyama, K. Nakano, A. Tonomura, and T. Osawa, Biochemistry, 10 (1971) 44574463. 13 W. F. Broekaert, M. Nsimba-Lubaki, B. Peeters, and W. J. Peumans, Biochem. J., 221(1984) 163169. 14 N. Shibuya, I. J. Goldstein, W. F. Broekaert, M. Nsimba-Lubaki, B. Peeters, and W. J. Peumans, J. Biol. Chem., 262 (1987) 19561601. 15 H. Kaku, W. J. Peumans, and I. 3. Goldstein, Arch. Biochem. Biophys., (1990) 25S262. 16 H. J. Allen and E. A. Z. Johnson, Carbohydr. Res., 58 (1977) 253-265. 17 U. K. Laemmli, Nature (London) 277 (1970) 680-685. 18 0. H. Lowry, N. J. Rosenbrough, N. J. Farr and R. J. Randall, J. Biol. C/tern., 193 (1951) 265-275. 19 E. J. M. Van Damme, A. K. Allen, and W. J. Peumans, Physiol. Plant., 73 (1988) 52-57. 20 W. J. Peumans, M. De Ley, and W. F. Broekaert, FEBS Letr., 177 (1984) 99-103.
