.J.” .Q.’ "C-o PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE l J I ‘I . 4 I MSU Is An Affirmative ActiorVEqual Opportunity Institution CHARACTERIZATION AND FUNCTIONAL ANALYSIS OF 88-1 LECTIN FROM CULTURED SOYBEAN ROOT CELLS by Shahnaz Malek-Hedayat A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1988 In (_) “.4 ABSTRACT CHARACTERIZATION AND FUNCTIONAL ANALYSIS OF 88-1 LECTIN FROM CULTURED SOYBEAN ROOT CELLS by Shahnaz Malek-Hedayat In an attempt to elucidate the mechanism of Rhizobial attachment to soybean roots, a model system was established using a cultured soybean cell line, SB-i, originally derived from roots of Glycine max. Incubation of Rhizobia with the 88-1 cells resulted in adhesion of the bacteria to the plant cells. This heterotypic interaction was strain specific for Rhizobium that normally infect soybean roots. Studying the inhibition of binding with various sugars suggested that the inter- action of Rhizobium and 38-1 cells is mediated through a galactose specific recognition system. A lectin, termed SB-1 lectin, was isolated from cultured SB-1 cells by affinity chromatography on a Sepharose column derivatized with N-caproyl-galactosamine, which is normally used for purification of soybean agglutinin (SBA), a galactose specific lectin found in soy- bean seeds. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting analysis of 88-1 lectin with anti-seed SBA antibody demonstrated a major polypeptide (Mr 3 30,000) which co-migrated with seed SBA. SB-1 lectin was observed in fractions purified from culture medium of 88-1 cells or supernatant fractions of 88-1 cell suspension after enzymatic removal of the cell wall. Moreover, fluorescently-labeled rabbit anti-SBA antibody incubated with 88-1 cultured cells showed specific immunofluorescent staining on =9 hub O». 2‘ the cell'wall and plasma membrane of the 88-1 cells. These results suggest that SB-1 lectin, produced by 88-1 cells, may mediate the recognition between the Rhizobium and the soybean cells. This notion is supported by the observation that rabbit anti-seed SBA antibody blocked Rhizobium-soybean cell adhesion, whereas the control antibody did not. Comparison of 58-1 lectin, derived from culture medium, and seed SBA by gel filtration and peptide mapping after limited proteolysis revealed no detectable difference between the lectins from the two sources. In addition, we found that both SBA and 88-1 lectin, under certain conditions, form highly stable dimers (Mr. 2 60,000) from their basic subunits (Mr 3 30,000). The polypeptide presentation of cell surface SB-1 lectin to the environment was probed by cell surface labeling of intact SB-1 cells with anti-peptide specific antibodies. The results suggest that the MHZ-terminal half of the 88-1 lectin is exposed, while the integration of 88-1 lectin in the cell wall may occur through interactions with the C-terminal half of the molecule. Combined results of fluorescence analysis of intact cell labeling and affinity chromatography methods suggest that the overall geometry of both SB-1 lectin anchored on the wall and SBA in solution may be similar. DEDICATION to my parents for their love iv . V ..U’ (1| [1) ACKNOWLEDGMENTS I would like to express my sincere appreciathmmto myrmnmors, Drs. Melvin Schindler and John L. Wang, for all the support, guidance and encouragement they gave me. I owe a special thanks to my friend and colleague, Dr. Siu-Cheong Ho, for his collaborative efforts and sincere friendship, which made the rough times easier to endure. Many thanks to all my colleagues and friends who shared the ups and downs of this work withlne and made my time in the lab such a pleasure. Special thanks to Linda Lang for her assistance in the prepara- tion of this thesis. Last, but not least, I truly thank my sister, Saidie, for being such a wonderful and caring sister and friend. TABLE OF CONTENTS List Of TableSOOOOOOOOOOOODO ..... OOOOOOOOOOD. ....... .0. 0000000000 Viii List Of FigureSOOOOOOOOOOOOOO0.0.0....0..ODOOOOOOOOOOOOOOOOOOOIII ix List Of AbbreViationS. I O O I O O O O O O I C O I O O O O O O O O O O D O O O O O O ...... O O I O O O Xi CHAPTER I LITERATURE REVIEWOOOOOO..00...OOOOOIOOOOOOOOOOOOOOOOOOOOCOOIOOOOOOOOO 1 Introduction to Lectins.......................................... 2 Specificity of Interaction of Lectins with Glycosylated Ligand(s)........................................................ 2 Endogenous Ligands............................................... A Biosynthesis and Regulation of Lectins........................... 5 Multiple Forms of Lectins........................................ 13 Tissue Distribution and Subcellular Localization................. 15 Endogenous Functions of Plant Lectins............................ 20 References...0.00.0000...DOOOOOOOODOOOOOOOOOO ....... OOOOOOOIOODOOOOOO 26 CHAPTER II ENDOGENOUS LECTIN FROM CULTURED SOYBEAN CELLS: ISOLATION OF A PROTEIN IMMUNOLOGICALLY CROSS-REACTIVE WITH SEED SOYBEAN AGGLUTININ AND ANALYSIS OF ITS ROLE IN BINDING OF RHIZOBIUM JAPONICUM............................................................ 35 Abstract............................................................. 36 Introduction......................................................... 37 Materials and Methods................................................ 39 Cell Culture and Protoplast Isolation............................ 39 Seed Soybean Agglutinin and Antibody Reagents.................... U0 Polyacrylamide Gel Electrophoresis and Immunoblotting .......... .. “2 Isolation of Antibodies by Specific Adsorption to a Polypeptide on Nitrocellulose.................................... A3 Assay of SB-1 Cell Components Reactive with Rabbit Anti-Seed SBA............................................... ..... A3 Isolation of Lectin Activity from SB-1 Cells..................... AU Rhizobium Culture and 88-1 Cell Binding.......................... A5 Histological Studies of 88-1 Callus Infected with Rhizobium...... N6 ReSUItSIIOOOO0.00.....0...00.0.0.0...DOOOOOOOOOOO0.00000000IOOO0....O "7 Characterization of Antibodies Against Seed SBA.................. "7 Binding of Antibodies Directed Against Seed SBA to 88-1 Cells.... 51 Isolation of 88-1 Lectin After Cell Wall Digestion............... 57 Binding of Antibodies Directed Against Seed SBA to Protoplasts... 58 Binding of Rhizobium to 88-1 Cells............................... 58 Evidence for Specificity and Role of Lectin in Rhizobium Binding to 88-1 Cells............................................ 66 Correlation Between Rhizobium Binding and Establishment of in vivo Symbiosis......... ...... ..... ..... ....................... 70 vi 5x» a .L. .~ DiSCUS81onOOOOOOOOOOOOOOOOIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO References.OOOOOOOOOOOOOOOOOOOOO00.00.000.000... CHAPTER III ENDOGENOUS LECTIN FROM CULTURED SOYBEAN CELLS CHEMICAL CHARACTERIZATION OF THE LECTIN OF 88-1 CELLS................ AbstractOOOOOOOOOO0....00.0.0.0...OOOOOOOOOOOOOOOOOOOOOOOOOOOOI IntrOduction.0.00.00.00.000.0.0.... Materials and Methods......................................... Seed SBA and Anti-Seed SBA Antibodies..................... Polyacrylamide Gel Electrophoresis and Immunoblotting............ Culture of 88-1 Cells............................................ Isolation of Lectin Activity from 88-1 Cells..................... Isolation of Lectin Activity from Soybean Seedlings.............. Isolation of Polypeptides from SDS-PAGE.......................... Comparative Peptide Map Analysis.......................... Gel Filtration and Radioimmunoassay...... ..... ...... Results.............................................................. Isolation of Lectin from 88-1 Cells.............................. Comparative Peptide Map Analysis of Seed SBA and 88-1 Lectin..... Interconversion of the 30 and 60 kDa forms of the Lectin......... Characterization of 88-1 Lectin.................................. Identification of the Lectin in Soybean Roots...... DiSCUSSion...DOC...O...OOOOOIOOOOOOOOOOOOOOO ReferenceSODO...0.0.0....OOOOOOOOOIOOOOOOOOIOI...0.... CHAPTER IV ENDOGENOUS LECTIN FROM CULTURED SOYBEAN CELLS SB-1 LECTIN ON THE CELL WALL................. Abstract...D0.0.0.0.0000...OOOOOOOOOOOOOOOOOO Introduction......................................................... Materials and Methods................................................ Preparation of Seed SBA and Anti-Seed SBA Antibodies...... Polyacrylamide Gel Electrophoresis and Immunoblotting............ Enzymatic Digestion of Seed SBA........ Preparation of Anti-Peptide Antibodies by Specific Adsorption to Polypeptides Immobilized on Nitrocellulose Membrane........... Culture and Immunolabeling of 88-1 Cells......................... ReSUItSOOOOOCOOOOOOIOOIOOOOOOOOOD.O0.0.0.0...OOOOOOOOOOOOOIOOCOOOOOI. Digestion of Seed SBA by V-8 Protease.................. Placement of the Peptides in the Primary Structure of seed SBAOOOOOODOOOOOOOOOI.OOOOOOOOOOOOIDO. Antibodies Reactive with Peptides F1-F7.......................... Reactivity of Peptides (F1-F7) with Antibodies Raised to Undenatured SBA.................................................. Immunofluorescence staining of 88-1 Cells with Anti-Peptide Antibodies....................................................... Discussion................ ............. .... ..... ....... References.................... CLOSING STATEMENT............. vii 72 76 81 82 8A 85 85 85 86 87 88 88 88 90 92 92 95 99 105 108 110 115 118 119 120 122 122 122 123 12A 125 126 126 129 132 136 136 M2 12m 1117 LIST OF TABLES Table CHAPTER II 1 Binding of 125I-Labeled Rabbit Anti-Seed Soybean Agglutinin to 88-1 Cells and Protoplasts Prewashed with Various saCCharideSoeeeoeoeeeeeeeeeeeoeeoeeecoo-00000000000000.coeoocoooo 2 Saccharide Inhibition of Rhizobium japonicum Binding to 88“ cells...I.0..DOOOOOOOOOOOOOOOOOOOOOOODOOOOOOOOOODOODOO.DO 3 Correlation Between Bacterial Binding and Symbiotic Infection.... viii Page 56 69 71 LIST OF FIGURES Figure CHAPTER I Diagrammatic representation of the circularly permuted sequence homology that relates concanavalin A to other leguminous lectinSOODOOOO00......I0.0DOOOOODOIOOOOOOOOOOOOOOOO0.0 CHAPTER II SDS-PAGE analysis of extracts of soybean seeds, purified seed SBA, andSB-1lectin...DO...DOOODOOOOOOODOOOOOOOOOO0.0...0.. Fluorescence staining patterns of 88-1 cells..................... Dose-response curve for the binding of 125I-labeled rabbit anti-seedSBAtOSB-1cellSOOOOOOI0.0......DOOIOOOIOOOOODOOOOOOOO Fluorescence staining patterns of protoplasts derived fromSB-1cell-SOOOOODOOOOOOOIOOIOOOOOOOOODOODOOOOOO ..... 0.000.... Representative photomicrographs showing the adhesion of Rhizobium japonicum (R11od) to 88-1 cells after (a) 2 h and (b) 2A h of co-culture at 26°C in the dark................... Histological staining of section derived from callus cultures of 88-1 cells with and without Rhizobium japonicum............... Representative photographs showing the adhesion of Rhizobium japonicum(R110d)tOSB-1cell-SODOOOOOOOOOOOOOOOD.0.000.000.0000. CHAPTER III SDS-PAGE analysis of seed SBA and 88-1 lectin purified by affinity chromatography on gal-Sepharose column.................. Comparative peptide map analysis of purified seed SBA and 88-1 lectin after limited V-8 protease hydrolysis and SDS-PAGE... Two-dimensional peptide map of radioiodinated 30-kDa poly- peptides from seed SBA and 88-1 lectin........................... Interconversion of 60 and 30 kDa polypeptides from seed SBA and purified SB-1 lectin analyzed by SDS-PAGE and immunoblotting ..... ix Page 10 A8 52 5A 59 61 6A 67 93 97 100 10“ Figure Page 5 SDS-PAGE and immunoblotting analysis of seed SBA and 88-1 lectin obtained from gal-Sepharose column upon sequential elution with varioussaccnarideSOOOOIOO...00.0.0.........OOOOOOOOOOOOOODDO....106 CHAPTER IV 1 SDS-PAGE analysis of peptides from seed SBA obtained via digestion with V-8 protease............................ ....... ... 127 2 Mapping of SBA peptides on the linear structure of SBA pelypeptide..0OOO.DIODOOIOOOOOOOOODOOOOOOIIOOOOOOOOIOOOOOOO00.0..130 3 Reactivity of V-8 generated SBA peptides with anti-SBA (30 kDa) and various anti-peptide antibodies as analyzed by immUnODlOttingeoeoee0000.00.00.co...00000000000000.0000.eeoeee 13]4 A SDS-PAGE and immunoblotting analysis of peptides from seed SBA with different preparations of anti-SBA polyclonal antibOdj-es.......OODOOOOOOIOOO0.0IDOOODODOOOOOIOOOOOD0.00.0.00...137 5 Fluorescence staining pattern of intact SB-1 cells with anti-SBA (30 kDa) or anti-peptide antibodies prepared by specific adsorption to nitrocellulose paper...................... 139 S .. n ...m 11 pl Ill? “mo—Pa an L s ... 5C ».~ A J . r In «J» «hum an, . . 2- :C 3. .u. Di Con A ANS TNS SBA WGA ABA 2,N,D PHA EPS LPS Gal GalNAc Gal-Sepharose Glc SDS PAGE BSA PBS HRP AP LIST OF ABBREVIATIONS Concanavalin A anilino-naphthalenesulfonic acid toluidinylnaphthalenesulfonic acid soybean agglutinin wheat germ agglutinin abscissic acid dichlorophenoxy acetic acid phytohemagglutinin extracellular polysaccharide lipopolysaccharide D-galactose N-acetyl-D-galactosamine Sepharose derivatized with N-caproyl-galactosamine D-glucose sodium dodecyl sulfate polyacrylamide gel electrophoresis bovine serum albumin phosphate buffered saline horseradish peroxidase alkaline phosphatase xi Chapter I LITERATURE REVIEW REVIEW OF LITERATURE Introduction to Lectins A century ago, Stilmark made the startling observation that extract of castor bean agglutinated erythrocytes (1). This discovery marked the initiation of research in the field of agglutinins (see Ref. 2 for review). Subsequently, it was found that a large variety of plants contained such agglutinating activity. Because these agglutinins were found to be blood group specific (3), they‘were given the name "Lectins" from the Latin "Legere," to choose (3). Currently, the term lectin describes: "Proteins or glyc0proteins of non- immunoglobulin nature capable of specific recognition and reversible binding to carbohydrate moieties of complex carbohydrates without .altering covalent structure of any of the recognized glycosyl ligands" (3). Even though lectins were originally found in plants, it is now firmly established that these proteins occur in a wide variety of organisms from microbes to humans. This thesis is concerned with pflant lectins. 'The review will predominantly deal with their bio- chemistry, in particular the lectins derived from leguminous plants. Specificity of Interaction of Lectins with Glycosylated Ligand(s) The ability of lectins to bind saccharides and saccharide~ containing proteins in a highly specific manner is onecfl‘thermmm prominent characteristics of this group of proteins. Due to this inuque property, lectins have provided investigators with a useful tool fxn‘ isolating and characterizing carbohydrates and glycoproteins, as well as for probing the molecular architecture of the cell surface and the changes induced therein by transformation. Sugar-lectin specificity is generally determined on the basis of hemagglutination reaction (5,6), affinity chromatography techniques such as passing glycoproteins of known glycoconjugate structure through a lectin column (7.8), or by polysaccharide precipitation (9,10). Lectins can be classified in terms of their specificity for monosaccharides as determined by hemagglutination inhibition or precipitation of carbohydrate-containing polymers (11). Such criteria: put lectins iJHHD defined carbohydrate specificity groups, e.g. N-acetylgalactosamine/galactose-binding lectins, mannose/glucose- binding lectins, the N-acetylglucosamine-binding lectins,1nua L-fucrose-binding lectins, sialic acid-binding lectins, and those witri complex binding sites (12). In general, lectins react with non- reducing, terminal glycosyl groups of polysaccharides or glycoprotein chain-ends. There>are, however, a few exceptions to this. For example, the lectins from pea, lentil, and fava been all react with reducing mannose units of N-acetylglucosamine 8-1.2,mannose (13). Concanavalin A (Con A), the lectin from jack bean, in addition to its interacticniruith terminal groups, binds internal 2-0-a-mannopyranosyl residues (1A). Lectins also exhibit a pronounced difference with regard to their anomeric specificity. Some show significant anomeric specificity (15,16), while other lectins seem to be anomerically indifferent (17). The carbohydrate-binding site of some lectins seems . r. q. _. .. . a. a. -. n..\ .. a 7; AL. r . .J . . w. L. t',ixllr 1 .I’tb. I- {II inulv u to be complementary to a single glycosyl residue, while in some others, the combining site can accommodate between 2-6 sugar residues, or in other words, they have greater specificity for oligosaccharides (18). A few lectins are even more specific and recognize carbohydrate sequence, in conjunction with the linkage amino acid(s) (19). Several lectins have also been observed to interact with non-carbohydrate ligands. This binding is independent of the carbohydrate-binding activity. Edelman and Wang (20) reported binding of Con A to plant auxin B-indoleacetic acid, a non-polar compound. Binding of several legume lectins to hydrophobic fluorescent molecule anilino-naphthalenesulfonic acid (ANS) and 2,6-toluidinylnaphtha1ene- sulfonic acid (TNS) has also been reported (21,22). The presence of binding sites for adenine and cytokinins has been demonstrated (13,23). The role of lectin binding to carbohydrates or hydrophobic ligands for cellular activity is another, predominantly, unanswered question. Endogenous Ligands To understand the biological functions of plant lectins, it is important to gain knowledge concerning potential endogenous ligands. Some early examples of the presence of endogenous lectin ligands were demonstrated by Gansera gt §_l_. (214) and also Gebauer e_t_ 11° (25). They took advantage of affinity chromatography and could isolate components from the seeds of Pisum sativum, Canavalia ensiformis, Vicia faba, Vicia sativa, and Ricinus communis that bound to the appropriate homologous seed lectin attached to the affinity column. These seed components could be specifically eluted with saccharides. .: C~ fif‘ t; V‘ ~—. udn Further investigations on the molecular properties of the isolated components revealed that they were glycoproteins. Bowles and Marcus (26), using an approach similar to Gansera gt gt. (214) and Gebauer gt a_l_. (25), showed that endogenous lectins from soybean and jack bean are highly specific for the glyCOproteins in the seed extracts of the homologous plants. Analysis of jack bean seed extract with gel electrophoresis and treatment of the gel with 1251- labeled Con A indicated that the dominant endogenous Con A binding receptor was the heavy subunit of o-mannosidase (27). In soybean, Bond and Bowles (28) found some factor(s) in the lectin depleted extracts from cotyledon and axis which had the ability to interact with soybean agglutinin (SBA) and inhibit SBA-induced hemagglutination activity. The presence of a receptor for SBA has been shown on the intact plasma membrane of the soybean protoplasts by Metcalf e_t_ at. (29) in studying mobility of membrane receptors (35). Despite these demon- strations of the presence of endogenous lectin-binding ligands, conclusive information about their function awaits further investiga- tion. Biosynthesis and Regulation of Lectins A recurring theme in studying plant lectins has been the control of their synthesis and activity which may eventually shed light on a better understanding of the endogenous function of these proteins. Regulation of lectin biosynthesis and expression may occur at various levels ranging from gene transcription to post-translation. A comparison of amino acid sequence from several legume lectins (30) ‘.v Cu r. c‘. Lulu! 7v. A”. .u' - ilnuiizfii r.- w ..J «\v :4 Flig'p "C‘ 9 strongly argues that these lectins are evolutionarily related. In a way, this suggests the possibility that synthesis of various legume lectins may be regulated via similar types of control mechanisms. Advances in DNA technology and cloning techniques have enabled investigators to provide more information about lectin genes during recent years. Genetic studies using a cDNA probe specific for soybean seed lectin coding sequence have shown the presence of two related lectin genes, Le1 and Le2, in soybean plants (31). The Le1 gene was identi- fied as the gene encoding for the prevalent soybean seed lectin. IWue function of the Le2 gene is unknown; however, it is expressed at a low level in embryo and roots. There are some soybean lines which apparently lack detectable seed lectin. Investigation of these lectin-negative soybean lines at the genetic level showed that they also contain both Le1 and Le2 genes (31,32). However, the Le1 gene is modified by a 3.14 kb DNA segment inserted within the coding region of this gene, which causes dramatic reduction in the transcription of the Le1 gene (31,32). The insert was shown to have structural features of a transposable element (32). Walling _e_t a_l_. (33) have clearly demon- strated that soybean seed Le1 lectin gene transcription is activated during early embryogenesis, maximized during mid-embryogenesis, and is repressed prior to dormancy. The Le1 gene appears to be also expressed at a low level in the roots of mature soybean plants (31,314). but the level of root lectin mRNA is lower (by a factor of 20,000) inuni that observed at mid-embryogenesis. In contrast, transcriptional activity of the soybean lectin gene in root and embryo differ only by a factor of 10 (3A). These observations suggested that A“ y. .‘ .h. p: Z ... Oi A1.— Av ».v (v a g .1 6 u C. b v u . r . L c . 1 . QC .. :1 3.. H. ch 9 4 v $ 4 t s u a. i s a v r. .~. P. a P Cu a v .1 I. D ‘ post-transcriptional processing may be involved in the regulation of gene expression. Interestingly, Okamuro gt FE' (311) have shown soybean lectin gene transferred to tobacco plants undergo the same developmental regulation. Studying the regulation of seed lectin gene transcript levels in Phaseolus vulgaris by hybridization of cotyledon RNA with lectin cDNA showed an accumulation of transcripts in the cotyledons during mid-maturation, and a decrease during late maturation (35). Studies of pea lectin gene transcript indicated the increase in lectin mRNA levels coincide with the time of maximal production of the seed lectin and that the accumulation of the lectin in the seed is regulated at the transcriptional level (36). Both 00- and post-translational processing has been shown to occur in many seed lectins. Comparison of the amino acid sequence of some lectins obtained by protein sequencing and those deduced from the nucleotide sequencing of cDNA revealed that primary translation products of mRNA contained a signal sequence at the NHg-terminus which is absent in mature protein; the signal sequence is cleaved off post-translationally. Vodkin (37) found _i_n v_i_tr_o_ translation of soybean seed mRNA produced a major polypeptide precipitable with antibodies directed against soybean seed lectin. But the molecular weight of this polypeptide, 32,300, was several thousand daltons larger than that of the non-glycosylated soybean lectin subunit, 28,000. This implicated that processing of nascent polypeptide may be involved during _i_n_ 2L0. synthesis. In some other leguminous plants such as lentil, pea, and fava, it has been reported that seed lectin is composed of subunits, each of fiL .C P. a. A» C» .h. it I . ... a v A.» 44 L. Q» 8 which contains a short o-chain and a long B-chain. Basically, this type of two-chain lectin is synthesized as a single chain polypeptide precursor. The precursor chain is cleaved into a and B and held together by non-covalent forces. Combined results of i vivo pulse- chase labeling experiments and _i_n_ 2.1.1322 translation of mRNA from immature pea cotyledons (36,38) indicated that pea lectin is synthe- sized as a single 25,000 pre-pro lectin polypeptide in the form of NHg-signal-B-chain-o-chain-CO0H. This primary precursor is initially associated with rough endoplasmic reticulum and then sequestered into the lumen of endoplasmic reticulum where the signal sequence is removed co-translationally. The polypeptide, ”r 23,000, is then transported to the protein body where it is post-translationally cleaved to yield a (Mr = 6,000) and is (Mr = 17,000) chains. These results were further confirmed by sequencing of two overlapping cDNA clones complementary to pea lectin mRNA, which was shown to code for 8 and 0: subunits (36,38). Similarly, co- and post-translational processing has been shown to be involved in biosynthesis of favin, the lectin from Vicia faba (39). .12 vitro translation of fava bean mRNA showed that favin is first synthesized as a single polypeptide precursor of 29,000. Amino acid sequencing of this precursor indicated the presence of a 29 hydrophobic amino acid residue signal sequence at the NHg-terminal, followed by the sequence of B-and o-chains (39). Comparison of the complete amino acid sequence of several legume seed lectins, either deduced from their nucleotide sequences of cDNA obtained by reverse transcription of mRNA (pea (36), soybean (32), P. vulgaris (35, Dolichos biflorus (A0)) or determined by classical .: LJ' -.. 9 methods using various proteolytic cleavage and protein sequencer (fava bean (A1), lentil (A2), sainfoin (A3), jack bean (AA)), have shown extensive homology exists among these lectins from the same taxonomi- cal grouping. However, to achieve maximum homology, the NHZ-terminal end of single chain lectins (soybean, sainfoin and Dolichos biflorus lectins) or the B-chain of two-chain lectins (fava bean, lentil and pea lectins) should be aligned with residue 120 of jack bean lectin (Con A). The a-chain of two-chain lectins is aligned with residues 70 through 119 of Con A. This relationship of the primary sequence of Con A with other legume lectins has been termed circular permutation (Figure 1) (“S-A7). The circular permutation of Con A may now be explained by the unique way that this protein is processed after synthesis to give rise to the mature lectin (A8). The synthesis and processing of Con A was studied by metabolic labeling of the precursor and pulse-chase experiments. These investigations revealed this protein initially is synthesized in the form of a glycosylated precursor, which is then deglycosylated and cleaved into two smaller peptides of Mr. 18,800 and 1A,000. The two fragments are then re-annealed to form mature Con A, but the alignment of residues 1-118 and 119-237 is reversed as confirmed by NHg-terminal sequencing of the precursor. The annealing of the fragments appears to be involved in a transpeptidation event, as suggested from the pulse-chase experiment. Post-translational processing has also been demonstrated to occur in the biosynthesis of lectin from cereals and rice belonging to the Gramineae family (A9,50). In these plants, lectin is synthesized as a 23,000 molecular weight precursor, which is post-translationally 1O Figure 1. Diagrammatic representation of the circularly permuted sequence homology that relates concanavalin A to other leguminous lectins (from reference 30). c~ bl. . Herc-1 :» .vk .7. is . ..~ .2 by «\v 3:“ .» A a L. p L b.. r . Pa «3 t . 2‘ .... 11 processed in a single step into an 18,000 molecular weight polypep- tide, as was evident by i vivo pulse-chase labeling experiments (119,50). In cereals, the ”r 18,000 polypeptide is the final product, while in rice, the Mr 18,000 polypeptide is further cleaved in a second step into smaller molecular weight polypeptides of 10,000 and 8,000 (50a). Studying _i_n y_i_t_r_g synthesis and the processing of these lectins in cell-free extracts derived from corresponding plants and in Xenopus oocytes have provided further support for the above synthesis and processing mechanism. There is some evidence that the biosynthesis of lectins may be regulated via plant hormones, although the mechanism of such effects is not yet known. Such regulatory effect has been best studied in the biosynthesis of wheat germ agglutinin (WGA) from wheat belonging to Gramineae. Triplett and Quatrano (51) found that if young wheat embryos were removed from grain and cultured, they germinate precoci- ously and concomitantly cease WGA synthesis as determined by measuring lectin levels using radioimmunoassay. The presence of 1-100 uM abscissic acid (ABA), a plant hormone, in the culture medium halts this precocious germination of embryos and not only was WGA synthesis initiated, but also the rate of synthesis was accelerated when compared with the level of lectin synthesis in the intact embryos. Raikhel gt gt. (52) provide evidence on the effect of ABA on the synthesis of WGA in "adult" wheat plants. These investigators found that wheat seedlings, grown under hydroponic conditions in the presence of ABA, showed 2- to 3-fold enhancement in lectin synthesis in the shoot, base and terminal portion of the root system. The effect of ABA on lectin synthesis and accumulation was also studied in 0v (.\ and .a ¢ P. [N ..u . r F. n. ... ... 111' 11'; El .- ‘1'... HPIA'. "r¢.' 0- S 12 the callus culture derived from immature wheat embryo in conjunction with other plant hormones such as 2,A-dichlorophenoxy acetic acid (2,A-D) (52). The highest rate of lectin accumulation was observed in callus grown only in 10 11M ABA. The presence of 2,A-D seemed to have an antagonistic effect and decreased the amount of lectin in culture. Such negative effect of 2,A-D on lectin synthesis has also been reported for Dolichos biflorus (53). Callus cultures derived from hypocotyls of Dolichos biflorus plants was shown to produce lectins similar to those isolated from intact stems and leaves if grown in the culture medium deprived from 2,A-D. Lectin production was completely inhibited in the presence of the exogenous 2,A-D in the culture medium. In contrast, depression of endogenous levels of ABA in wheat seedlings resulted in lowering the level of WGA in shoot base and roots compared to control plants (52). Studies carried out on the effects of hormones on the synthesis of lectin in developing rice embryos grown in culture indicated ABA strongly promoted the level of lectin synthesis, while gibberellic acid suppressed the synthesis of lectin in rice embryos (50). There is no report of how and at which level these hormones exert their effects. More recently, Raikhel and Wilkins (5A), using a cDNA clone as a hybridization probe, have shown that treatment of excised wheat embryo with ABA changes the level of mRNA for WGA. They suggested exogenous ABA may play a regulatory role at the level of gene expression. Further investigation is necessary to explore the mechanism by which these hormones manipulate lectin synthesis and accumulation. 'l I. 4‘ ‘-L V .. . .u. . . , . A. . ..A . . .. .. .1. r. ... ... c. .. C a p. r, r . .3 ml 7. “t .... C a. ...! i c .: o . Mr. ...... . a. we. . 3: . . t c c. . r 2 r . r . . a; , r . .. 5.: -. c . av C J. n. to s. . . . . n. ... ... 0. .. 2» ... . L. n . .G . . w . . a. . pg L. as c. c . . u A. A J I p. I c at» ..a s s . A ‘9 Pic A.“ 13 Multiple Forms of Lectin A number of laboratories have provided evidence on the presence of multiple forms of lectins in plants. These multiple forms are called "isolectins" and usually can be identified by SDS-gel and isoelectric focusing. The basis for such variability is complex and could be due to different factors, including genetic polymorphism, species polymorphism and post-translational modification (36). Lectins from Phaseolus vulgaris, phytohemagglutinin (PHA) , con- tain two forms of subunits which are structurally different from one another by six residues at their amino-terminal sequence (55,56). These two subunits, which are termed E (erythrocyte-reactive) and L (lymphocyte-reactive), associate randomly and yield five different tetrameric isolectins: Lu, L3E1, L2E2, L1E3, and Eu (65). In 11333. cracea seeds, two different lectins have been found, a mannose Spe- cific lectin and an N-acetyl-galactosamine specific lectin (57). Com- bined results from electrophoresis and N-terminal sequencing analysis of the two forms indicated that the mannose specific subunit is com- posed of two chains, while the N-acetyl-galactosamine specific subunit is a single chain. In soybean, the presence of two related lectin genes, Le1 and Le2, has been demonstrated (discussed earlier) (31). Heterogeneity in lectin subunits has also been shown to arise as a result of post-translational modification. For example, in some leguminous plants, seed lectin is composed of subunits of short or and long B-chains (discussed earlier). The lectin from Dolichos biflorus seed is a tetramer composed of two subunit types. Structural studies of these two subunits have shown they are very similar and they only differ at their carboxy-terminal end (58,59). These two subunits have Hy Cu :2 F. PH ,sd (1 .Vd- 1A been termed I and II. Recently, Schenell and Etzler (A0) have shown by molecular cloning studies that both subunits are encoded by a single gene. These investigators provided evidence that subunit II arises by post-translational proteolytic cleavage of 10 amino acid sequences from the carboxy-terminus of subunit I. Materials cross- reactive to seed lectin have also been identified in the stems and leaves of Dolichos biflorus (60). Subunit I from the seed lectin apparently is shared by the lectin(s) identified in leaves and stems. Other kinds of known post-translational modifications, such as deamination of asparagine or glutamine side chains and glycosylation, may also occur after the synthesis of lectins, and lead to multiple forms of these proteins. Multiple forms of lectins have also been observed in plants from Gramineae. Three closely-related isolectins have been described in hexaploid wheat (Triticum aestivum) for WGA (61 ,62). Studies with wheat of different ploidy (diploid, tetraploid and hexaploid) have provided evidence for the existence of diverged triplicate genes for the lectins in hexaploid wheat (63,6A); the synthesis of each of these lectins, which have been termed isolectin I, II, and III, is directed by a different genome, A, D, and B, respectively (6A). The three isolectins have the same molecular weight, but have slightly different amino acid compositions (61,62). Amino acid sequences of these lectins have indicated that isolectin I and II are different at four residues (65,66). “Isolectin III is different from isolectin I and II by 10 and 8 amino acids, respec- tively (5A). Whether multiple forms of lectins in a plant have any physiological role or biological significance remains to be investi- gated. 15 Tissue Distribution and Subcellular Localization Although lectins were originally observed in the seeds of legumi- runna plants, there are now numerous reports to support the ubiquitous occurrence of these proteins/glycoproteins in many plant species (for review see 67 and 68). Two major approaches have been employed for detection and/or localization of lectins in plants, biological assays and immunological assays. Biological assays basically involve nmasuring hemagglutination activity of lectins, since they can specifically react with simple or complex carbohydrateaffhflty chromatography on columns derivatized with rabbit anti-seed SBA. The immunoreactive material yielded a predominant polypeptide band (Mr ~ 30,000) and a minor component (Mr ~ 60,000). I designate this affinity purified material as SB-1 lectin. Preliminary studies indicate that SB-1 lectin is similar if not identical to seed SBA when compared by gel filtration under non-denaturing conditions and by peptide mapping analysis. These results suggest that SB-1 cells produce an endogenous lectin that binds tn) galactose-containing glycoconjugates and that the SB-1 lectin may be on the cell wall and could be released upon degradation of the wall. This conclusion is consistent with our observations on the immunofihxwescence staining of the cell wall with rabbit anti-seed SBA. 3. ' P 9 '~ _ «.W'!‘ 58 Binding of Antibodies Directed Against Seed SBA to Protoplasts Rabbit antibodies directed against seed SBA also bind to proto- plasts derived from SB-1 cells. Indirect immunofluorescence revealed ring-like staining, outlining the periphery of the cell and charac- teristic of surface staining patterns obtained with other spherical objects (Figure Aa). This suggests that the antigenic determinant recognized by rabbit anti-seed SBA is diffusely distributed on the plasma membrane. Preincubation with saccharides failed to alter the staining pattern (Figure Ab,c,d). Preimmune rabbit immunoglobulin yielded little or weak staining (Figure Ae). Similar results were obtained using the radiolabeled antibody probes. In these respects, the molecule immunologically cross-reactive with seed SBA on the plasma membrane appears to be very similar to that found on the outside of the cell wall (see below, however, for possible differences in Rhizobium-binding properties). Binding of Rhizobium to SB-1 Cells When SB-1 cells were mixed with Rhizobium japonicum (R110d) at 26°C for several hours, washed, and sampled under a microscope, the bacteria adhered to certain soybean cells (Figure 5). Initially, there was little bacterial binding and the binding was limited to the tips of some plant cells. A representative photomicrograph, taken after 2 h of co-culture, is shown in Figure 5a. Between 12 and 2A h, however, there appeared to be a sorting out process. When the co-culture of SB-1 cells and Rhizobium was sampled after 2A h, a striking "polar" mode of binding was observed; the Rhizobium adhered to the plant cells in an "end-to-end" fashion (Figure 56). 59 Figure A. Fluorescence staining patterns of protoplasts derived from SB-1 cells treated for 1 h at A°C with rabbit anti-seed SBA (30 ug/ml) or normal rabbit immunoglobulin (30 ug/ml), followed by fluorescein conjugated goat anti-rabbit immunoglobulin (1:100 dilution; 30 min at A°C). (a) protoplasts treated with rabbit anti-seed SBA; (b) proto- pflasts washed with Gal (0.1 M), then treated with rabbit anti-seed SBA; (c) protoplasts washed with GalNAc (0.1 M), then treated with rabbit anti-seed SBA; (d) protoplasts washed with Glc (0.1 M), then treated with rabbit anti-seed SBA; (e) protoplasts treated with normal rabbit immunoglobulin. ph, phase contrast microscopy; fl, fluores- cence microscopy; bar = 5 pm. 6O Figure A 61 Figure 5. Representative photomicrographs showing the adhesion of Rhizobium japonicum (R110d) to SB-1 cells after (a) 2 h and (6) 2A h of co-culture at 26°C in the dark. Bar = 10 um. Arrows indicate bacteria. 62 Figure 5 63 To test whether this interaction between Rhizobium and SB-1 cells leads to penetration into the soybean cell and infection by the bacteria, we carried out histological staining on SB-1 cells in callus culture that had been incubated with Rhizobium for three weeks. Using the Gram stain to reveal the presence of bacteria, we observed staining within the cell wall of certain cells (indicated by the dark arrow in Figure 6a), suggesting bacterial infection of these cells. In addition, there was also staining in the interstitial spaces between cells (indicated by the open arrow in Figure 6a). These observations, particularly the presence of bacteria in areas between adjacent cells, are reminiscent of similar observations on pseudo-infection threads in the establishment of Rhizobium-soybean symbiosis (18). Control sections, derived from cultures without Rhizobium, failed to show bacterial staining (Figure 6b). We have also stained sections of the Rhizobium-SB-1 callus co-culture with hematoxylin-eosin. These sections showed the posi- tions of the nucleus and cytoplasm of the plant cell instead of the bacteria. In cultures containing Rhizobium, there were focal regions containing many cells stained with the reagent, revealing prominent nuclei (Figure 60). In contrast, control sections contained many areas that were not stained, most probably because these areas of the cell were filled with vacuoles (Figure 6d). These observations are similar to those reported previously on the i vivo infection of soybean roots by Rhizobium (2), in which the infection process stimulated cell division. 6A Figure 6. Histological staining of sections derived from callus cultures of SB-1 cells with and without Rhizobium japonicum (R110d). The SB-1 callus was cultured with the bacteria for three weeks, fixed, sectioned, and stained as described in Materials and Methods. (a) SB-1 callus culture plus Rhizobium stained with Gram stain. The black arrows point to infected cells, containing bacteria within the cell wall. The open arrow shows bacteria in the interstitial space between cells, mimicking a pseudo-infection thread. (b) Control SB-1 callus culture without Rhizobium stained with Gram stain. (c) SB-1 callus culture plus Rhizobium stained with hematoxylin-eosin. The black arrow highlights a focal region of proliferative cells. (d) SB-1 callus culture without Rhizobium stained with hematoxylin-eosin. Bar =10 um. 65 Figure 6 66 Evidence for Specificity and Role of Lectin in Rhizobium Binding to SB-1 Cells Several aspects of the specificity of the binding interaction between Rhizobium and SB-1 cells were checked. First, the inclusion of certain saccharides such as Gal during the co-culture inhibited the binding of the Rhizobium to SB-1 cells when the observed polar adherence was assayed at 2A h (Figure 7, a and b). This inhibition was observed at a Gal concentration as low as 3 mM. Similar results were observed with the disaccharide lactose (Table II). In contrast, other saccharide epimers of Gal, such as Glc (0.2 M), failed to yield the same inhibitory effect (Figure 7c). Melibiose, the o anomer of lactose, did not show inhibition. GalNAc was also not inhibitory at the concentration tested (Table II). These results raise the possibil- ity that the adhesion of Rhizobium to SB-1 cells may be mediated via a highly specific carbohydrate recognition system. Second, since we have identified on the cell wall and plasma membrane of the SB-1 cells a lectin that is specific for galactose residues, it was of interest to test whether antibodies reactive against the cell wall lectin could block Rhizobium adhesion. We found that inclusion of rabbit anti-seed SBA (10 ug/ml) during the co-culture inhibited the polar binding of the bacteria to the SB-1 cells (Figure 7d). Normal rabbit immunoglobulin did not yield the same effect (Figure 7e). We also wished to test whether any ligand bound to the cell wall of SB-1 cells would block Rhizobium adhesion. To accomplish this, we took advantage of the availability of rabbit anti-cell wall fragments. This immunoglobulin fraction showed immunofluorescence staining of 67 Figure 7. Representative photographs showing the adhesion of Rhizobium japonicum (R110d) to SB-1 cells after 2A h of co-culture at 26°C in the dark. (a) co-culture; (b) co-culture in the presence of Cal (0.1 M); (c) co-culture in the presence of Glc (0.1 M); (d) co-culture in the presence of rabbit anti-seed SBA (10 ug/ml): (e) co-culture in the presence of normal rabbit immunoglobulin (1 mg/ml); and (f) co-culture in the presence of rabbit anti-cell wall fragments (1 mg/ml). Bar - 10 um. 68 Figure 7 ......sp. 1 I u 4 i n u ‘. 4 n 69 TabhaII. Saccharide inhibition of Rhizobium japonicum binding to SB-1 cells. Inhibition of polar binding Saccharide* to SB-1 Cells Control - Galactose+ + N-acetyl-galactosamine - Lactose+ + Galacturonic acid+ + Gluconic acid - Mannose ’ Glucose - Melibiose - Glucuronic acid - Xylose - * Saccharide concentrations varied from 3 mM to 0.2 M. + At a saccharide concentration of 3 mM or above, polar binding«of Rhizobium to SB-1 cells was inhibited (see text). “A K'U. A? A L D (I (I) 5’71;- vv. A R \‘V-uu,‘ 70 both intact SB-1 cells and the fraction containing cell wall fragments but it did not yield any positive reaction with seed SBA or SB-1 lectin on immunoblots. More importantly, the binding of this immuno- globulin on the outer surface of SB-1 cells did not inhibit Rhizobium binding (Figure 7f). These results provide strong evidence for the specificity and the role of the SB-1 lectin in mediating the initial recognition and adhesion between the Rhizobium and SB-1 cells. However, it should be noted that not all of the soybean cells bound Rhizobium. For example, Figure 7a shows one cell with many bacteria bound, but several adjacent cells devoid of any Rhizobium. In addition, we also found that Rhizobium did not bind to protoplasts derived from SB-1 cells after cell wall removal. Therefore, even though the plasma membrane of SB-1 protoplasts contained a lectin reactive with rabbit anti-seed SBA, no binding of Rhizobium was observed. Correlation Between Rhizobium Binding and Establishment of in vivo Symbiosis The polar binding of bacteria to the SB-1 cells was also specific in terms of the bacterial cells used in the co-culture (Table III). Rhizobium japonicum bound, but Escherichia coli did not. Moreover, the binding was restricted to Rhizobium japonicum and Rhizobium fredii, two strains of bacteria that normally infect soybean roots to form a nitrogen fixing symbiosis. In contrast, Rhizobium meliloti, Rhizobium trifolii, and Rhizobium leguminosarum did not bind to the SB-1 cells. Table 71 Table III. Correlation between bacterial binding and symbiotic infection. Polar binding to Bacterium Normal Host SB-1 cells E. coli ? - R. japonicum R110d soybean + R. fredii PRC 205 str soybean + R. meliloti 102F28 alfalfa - R. trifolii 0A03 clover - R. leguminosarum 128C56 pea - DISCUSSION The data documented in the present study indicate: (a) Incuba- tion of Rhizobium with a cultured cell line derived from roots of Glycine max (SB-1) results in specific adhesion of the bacteria to the plant cell. (b) This binding interaction appears to be mediated via carbohydrate recognition, since Gal can inhibit the heterotypic adhesion whereas Glc failed to inhibit. (c) One likely candidate that may mediate such an interaction is a lectin identified on the cell wall and plasma membrane of the SB-1 cells. This notion is supported by the observation that rabbit anti-seed SBA blocked the Rhizobium- soybean cell adhesion whereas control rabbit immunoglobulin did not. These results are consistent with the "lectin recognition" hypothesis that suggests carbohydrate recognition as a basis for determining legume host-bacterial symbiont interactions (6,15,20). This hypothesis has been supported by experiments carried out in the soybean system (2,3,13,1A,35) and in the clover system (10). There are, however, a number of experiments from various laboratories argu- ing against the acceptance of the hypothesis that lectins play a specific and indispensable role in legume-Rhizobium symbiosis; in the case of soybeans, this viewpoint has been put forth succinctly by Pueppke (29). In light of these circumstances, it is important to 72 73 discuss our data on the SB-1-Rhizobium interaction with respect to the following key points. First, we have obtained definitive evidence for the presence of a lectin in the SB-1 cells. This endogenously produced lectin has been purified to apparent homogeneity on the basis of its carbohydrate- binding activity. Immunofluorescence and binding studies carried out with 125I-labeled antibodies indicate that the lectin is found on the cell wall. Treatment of SB-1 cells with the haptens for seed SBA, Cal and GalNAc, failed to remove the soybean lectin from the cell wall. This implies that the lectin may be anchored on the cell wall with its carbohydrate-binding sites unoccupied and therefore is capable of mediating recognition and binding of external ligands (e.g. Rhizobium). Therefore, the requirement for the presence of lectin molecule at the proximal point of interaction has been fulfilled. Second, it should be noted that the mere presence of the lectin is not sufficient for Rhizobium binding. Two observations make this point particularly clear. The lectin of SB-1 cells is found on the cell wall of all cells examined by immunofluorescence. Yet, only certain cells out of a given pOpulation have Rhizobium adsorbed on them after co-culture of the plant cells and bacterium. This may be related to the growth phase of the SB-1 cells in culture or other phenomena associated with transient susceptibility of root cells to be nodulated by Rhizobium i vivo (A). In addition, protoplasts also have SB-1 lectin exposed outside the plasma membrane but these proto- plasts do not bind Rhizobium at all under conditions used to assay the adhesion of the bacteria to SB-1 cells. These results suggest that lectin-carbohydrate interactions may be a necessary but not sufficient 7A condition for adhesion of the cells. A requirement for dual recogni- tion (involving another set of complementary molecules) has been persuasively demonstrated in the interaction between lymphoid cells and target cells bearing foreign antigens (39). Third, it is important to realize that lectin-carbohydrate binding need not be the only, or even the main, determinant of specificity in soybean root cell-Rhizobium interactions. The notion of dual recognition, invoking other sets of complementary molecules, is consistent with the less absolute "lectin recognition" hypothesis. In any case, the demonstration of saccharide and antibody specificity in blocking Rhizobium adhesion to SB-1 cells strongly suggest that at least one required component is a carbohydrate-binding protein. In this connection, it should be noted that GalNAc, a known hapten for seed SBA (23), did not inhibit Rhizobium adhesion to SB-1 cells. This may reflect a difference between SB-1 lectin and seed SBA. Alterna- tively, it may reflect the fact that the lectin anchored on the cell wall does not bind GalNAc. Because our studies have been carried out in a defined cell culture system, one issue is whether this Rhizobium-SB-1 cell binding is relevant to the i vivo symbiosis. Several phenomenological observations suggest that our system mimics at least the early phase of the process of nodule formation in soybean roots. First, the binding of Rhizobium is polar as had been observed in a number of systems of Rhizobium binding to root cells (7,11,37). Second, there is preliminary evidence, based on histological staining, for the presence of bacteria in the interstitial spaces mimicking a pseudo-infection thread (18). The staining with hematoxylin and eosin 75 also suggest an increase in the size of the nucleus and possibly cell division (2). These observations at the light microscope level must now be extended to the ultrastructural level to confirm that the Rhizobium initially bound to SB-1 cells actually penetrate and infect the target cells. Finally, correlative studies between Rhizobium binding to SB-1 cells and establishment of _i_n_ vivo symbiosis indicate the specificities of the Rhizobium strains and their hosts. There have been several previous reports on the binding of Rhizobium to cultured cells derived from callus of soybean roots (8,16,18,26,27,31,32). In some of these systems, the interaction of Rhizobium with the soybean cells ultimately led to infection of the plant cell and the generation of a nitrogen-fixing symbiosis, as characterized by ultrastructural studies and enzymatic assays. It remains to be demonstrated that our present Rhizobium adhesion to SB-1 cells will lead to a symbiosis and activation of nitrogenase. REFERENCES Allen, A.K. and A. Neuberger. 1975. A simple method for the preparation of an affinity absorbent for soybean agglutinin using galactosamine and CH-Sepharose. FEBS Lett. 50:362-36A. Bauer, W.D. 1981. Infection of legumes by Rhizobia. Ann. Rev. Plant Physiol. 32:A07-AA9. Bhuraneswari, T.V. and W.D. Bauer, 1978. Role of lectins in plant microorganism interactions. ILII. Influence of rhizosphere/rhizoplane culture conditions on the soybean lectin-binding properties of Rhizobia. Plant Physiol. 62:71-7A. Bhuvaneswari, T.V., A.A. Bhagwat, and W.D. Bauer. 1981. Transient susceptibility of root cells in four common legumes to nodulation by Rhizobia. Plant Physiol. 68:11AA-11A9. Bhuvaneswari, T.V., B.G. Turgeon, and W.D. Bauer. 1980. Early events in the infection of soybean (Glycine max (L.) Merr) by Rhizobium japonicum. Plant Physiol. 66:1027-1031. Bohlcxxl, 8.8. and E.L. Schmidt. 197A. Lectins: a possible basis for specificity in the Rhizobium-legume root nodule symbiosis. Science 185:269-271. Calvert, H.E., M. Lalonde, T.V. Bhuvaneswari, and W.D. Bauer. 1978. Role of lectins in plant microorganism interactions. IV. Ultrastructural localization of soybean lectin binding sites on Rhizobium jgponicum. Can. J. Microbiol. 2A:785-793. 76 10. 11. 12. 13. 1A. 15. 77 Child, J.J. and T.A. LaRue. 197A. A simple technique for the establishment of nitrogenase in soybean callus culture. Plant Physiol. 53:88-90. Constabel, 1?. 1975. In Plant Tissue Culture Methods (0.L. Gamborg and L.R. Wetter, eds.) National Research Council of Canada, Saskatoon, Saskatchewan, Canada. pp. 11-21. Dazzo, FHEL., G.A. Napoli, and D.H. Hubbell. 1976. Adsorption of bacteria to roots as related to host Specificity in the Rhizobium-clover symbiosis. Appl. Environ. Microbiol. 32:166-171. Dazzo, F.B., G.L. Truchet, J.E. Sherwood, E.M. Hrabah, M. Abe, and S.H. Pankratz. 198A. Specific phases of root hair attachment in the Rhizobium trifolii-clover symbiosis. Appl. Environ. Microbiol. A8:11A0-1150. Gamborg, O.in, IRLA. Miller, and K. Ojima. 1968. Nutrient requirements of suspension cultures of soybean root cells. Exp. Cell. Res. 50:151-158. Halverson, L.J. and G. Stacey. 198A. Host recognition in the Rhizobium-soybean symbiosis. Detection of a protein factor in soybean root exudate which is involved in the nodulation process. Plant Physiol. 7A:8A-89. Halverson, L.J. and G. Stacey. 1985. Host recognithm1ix1the Rhizobium-soybean symbiosis. Evidence for the involvement of lectin in nodulation. Plant Physiol. 77:621-625. Hamblirh,.1. and S.P. Kent. 1973. Possible role of phytohemagglu- tinin in Phaseolus vulgaris L. Nature New Biol. 2A5:28-30. 16. 17. 18. 19. 20. 21. 22. 23. 2A. 78 Hermina, N. and M. Reporter. 1977. Root hair cell enhancement in tissue cultures from soybean roots: a useful model system. tg vitro Rhizobium symbiosis. Plant Physiol. 59:97-102. Ho, S.-C., H. Izumi, and A.M. Michelakis. 1982. The purifica- tion and characterization of multiple forms of mouse submaxillary gland renin. Biochim. Biophys. Acta 717:A05-A13. Holsten, R.V., R.C. Burns, R.W.F. Hardy, and R.R. Hebert. 1971. Establishment of symbiosis between Rhizobium and plant cells t1_1_ yttgg. Nature 232:173-175. Kowal, R. and R.C. Parsons. 1980. Stabilization of proteins immobilized on Sepharose from leakage by glutaraldehyde cross- linking. Anal. Biochem. 102:72-76. Krflpe, M. 1956. Blutgruppenspezifische Pflanzliche Eiweisskorper (Phytagglutine). Ferdinand Enke, Stuttgart. pp. 131. Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage TA. Nature (Lond.) 227:680-685. Lillie, R.D. 1965. Histopathologic Technic and Practical Histochemistry, 3rd ed. McGraw-Hill Book Co., New York. p. 1A1. Lis, H. and N. Sharon. 1981. Lectins in higher plants. In trtg Biochemistry gt Plants (P.K. Stumpf and E.E. Conn, eds.) Vol. 6, Academic Press, New York. pp. 371-AA7. Luna, L.G. 1968. Manual of Histologic Staining Methods of the Armed Forces Institute of Pathology. 3rd. ed. McGraw-Hill Book Co., New York. p. 38. 25. 26. 27. 28. 29. 30. 31. 32. 79 Metcalf, T.N., Wang,J.L., Schubert, K.R. and M. Schindler. 1983. Lectin receptors on the plasma membrane of soybean cells. Binding and lateral diffusion of lectins. Biochemistry 22:3969- 3975. Ozawa, T. and M. Yamaguchi. 1979. Inhibition of soybean cell growth by the adsorption of Rhizobium Eponicum. Plant Physiol. 6Az65-68. Phillips, D.A. 197A. Factors affecting the reducthmucfi‘ acetylene by Rhizobium-soybean cell associations tg yttgg. Plant Physiol. 53:67-72. Pierce, M. and C.B. Ballou. 1983. Cell-cell recognition in yeast. Characterization of the sexual agglutinin factors from Saccharomyces Kloyveri. J. Biol. Chem. 25:3576-3582. Pueppke, S.G. 1983. Soybean lectin: does it have an essential role in the Rhizobium-soybean symbiosis? In Chemical Taxonomy, Molecular Biology, and Function of Plant Lectins. (I.J. Goldstein and M.E. Etzler, eds.) Alan R. Liss, New York. pp. 225-236. Reitherman, R.W., S.D. Rosen, W.A. Frazier, and S.H. Barondes. 1975. Cell surface species-specific high affinity receptors for discoidin: developmental regulation in Dictyostelium discoideum. Proc. Natl. Acad. Sci. USA 72:35A1-35A5. Reporter, M. 1978. Hydrogen (H2) evolution by Rhizobium after synergetic culture with soybean cell suspensions. Plant Physiol. 61:753’756. Reporter, M., D. Raveed, and G. Norris. 1975. Binding of Rhizobium iaponicum to cultured soybean root cells: morphological evidence. Plant Science Lett. 5:73-76. 33. 3A. 35. 36. 37. 38. 39. 8O Rutishauser, U. 198A. Developmental biology of a neural cell adhesion molecule. Nature 310:5A9-55A. Smith, D.E. and P.An Fisher. 198A. Identification, developmental regulation, and response to heat shock of two antigenically related forms of a major nuclear envelope protein in Drosophila embryos. Application of an improved method for affinity purification of antibodies using polypeptide immobilized on nitrocellulose blots. J. Cell Biol. 99:20-28. Stacey, G., A.S. Paan, and W.F. Brill. 1980. Host recognition in the Rhizobium-soybean symbiosis. Plant Physiol. 66:609-61A. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U.S.A. 76:A350-A35A. Tsien, R.C. and E.L. Schmidt. 1977. Polarity in the exponential phase Rhizobium japonicum cell. Can. J. Microbiol. 23:127A-128A. Vacquier, V.D. and G.W. Moy. 1977. Isolation of bindin: the protein responsible for adhesion of sperm to sea urchin eggs. Proc. Natl. Acad. Sci. USA 7A:2A56-2A60. Zinkernagel, R.M. and P.C. Doherty. 1975. H-2 compatibility requirement for T-cell mediated lysis of target cells infected with lymphocytic choriomeningitis virus. Different cytotoxic T-cell specificities are associated with structures coded for in H-2K or H-2D. J. Exp. Med. 1A1:1A27-1A36. CHAPTER III ENDOGENOUS LECTIN FROM CULTURED SOYBEAN CELLS: CHEMICAL CHARACTERIZATION OF THE LECTIN OF SB-1 CELLS 81 deri‘ Re“ C. ABSTRACT A lectin has been identified in the cell line, SB-1, originally derived from the roots of Glycine max. This lectin, which we shall refer to as SB-1 lectin, was isolated on the basis of its carbohydrate- binding activity (affinity chromatography on Sepharose column derivatized with N-caproyl-galactosamine) and its immunological cross-reactivity (immunoblotting with rabbit antibodies directed against seed soybean agglutinin (SBA)). SDS-PAGE and immunoblotting analysis of SB-1 lectin revealed a major polypeptide (Mr 3 30,000) which co-migrated with seed SBA. This form of the lectin was observed in fractions purified from culture medium of SB-1 cells or supernatant fraction of SB-1 cell suspension after enzymatic removal of cell wall. Extracts of SB-1 cells under some other conditions yielded a major band (Mr = 60,000) as revealed by SDS-PAGE and immunoblotting with rabbit anti-seed SBA; prolonged incubation of these samples in the presence of SDS resulted in the appearance of the 30 kD polypeptide. It appeared that the 60 kD band represented a highly stable, even under SDS-PAGE conditions, dimeric form of the 30 kD subunit. The SB-1 lectin derived from the culture medium was compared with seed SBA by gel filtration and by peptide mapping after limited proteolysis; no difference between the lectins from the two sources was found. Extracts of soybean roots fractionated on N-caproyl-galactosamine Sepharose affinity columns yielded, upon elution with galactose, 82 83 polypeptides of Mr 30,000 and Mr 60,000. These results suggest that soybean roots contain a lectin whose polypeptide composition corres- ponds to that of seed SBA and SB-1 lectin. INTRODUCTION In the previous studies, we have shown that incubation of Rhizobium japonicum with the cultured soybean cell line SB-1, origin- ally derived from the roots of Glycine max, resulted in specific adhesion of the bacteria to the plant cells (1). We had also shown that rabbit antibodies directed against seed soybean agglutinin (SBA) blocked the Rhizobium-soybean cell adhesion, whereas control anti- bodies did not. These results prompted the hypothesis that one likely candidate that may mediate the recognition between the Rhizobium and the soybean cells is a lectin produced endogenously by the SB-1 cells. Consistent with this hypothesis, fluorescently-labeled rabbit anti- seed SBA yielded specific immunofluorescent staining on the cell wall and plasma membrane of the SB-1 cells and a lectin-like activity could be identified in the cell wall fraction of the same cells. Because of its implicated role in Rhizobium-soybean cell adhe- sion, it was of interest to characterize this lectin in some detail. In particular, we wished to determine the relationship of the SB-1 lectin to seed SBA. The present paper documents that the SB-1 lectin is similar, if not identical, to seed SBA on the basis of the follow- ing criteria: (a) carbohydrate-binding activity; (6) immunological cross-reactivity; and (c) peptide mapping patterns after limited proteolysis. 8A MATERIALS AND METHODS Seed SBA and Anti-Seed SBA Antibodies Seed SBA was isolated and purified by affinity chromatography CW1 Sepharose column derivatized with N-caproyl-galactosamine (Gal- Sepharose) (2). The details of this procedure, as well as the characterization of the purified lectin, have been documented previ- ously (1). The generation of antibodies directed against seed SBA and the characterization of the specificity of these antibodies have also been described (1). Seed SBA was labeled with 1251 using the chloramine T procedure described by Ho gt gt. (3). Free 1251 was removed by passing the labeled material over a column (3 x 0.5 cm) of Dowex AG1x8 (Bio Rad, Richmond, CA). The specific activity of the 125I-labeled seed SBA was approximately 2 x 107 cpm/ug. The same procedure was used to label the lectin isolated from SB-1 cells. Polyacrylamide Gel Electrophoresis and Immunoblotting Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE) was done as described (A) using 10% and A% (w/v) acrylamide concentrations in the running and stacking gels, respectively. Following electrophoresis, the proteins were revealed by staining with Coomassie Brilliant Blue, by fluorography of radio- active samples, or by immunoblotting after transfer to nitrocellulose 85 86 paper. Fluorography was carried out according to the method of Bonner and Laskey (5), using XAR-5 film (Kodak, Rochester, NY) and an exposure time of three weeks. For immunoblotting, the proteins were transferred to nitrocellu- lose paper (Schleicher and Schuell, Keene, NH) by the method of Towbin _et _1_. (6) and the immunoreactive material was revealed either via horseradish peroxidase (HRP) conjugated or alkaline phosphatase (AP) conjugated secondary antibodies. In the former case, the immuno- blotted nitrocellulose membrane was incubated with HRP-conjugated goat anti-rabbit immunoglobulin (Bio Rad; 1:2000 dilution), followed by development with A-chloro-1-napthol and hydrogen peroxide (1). In some experiments, the sensitivity of the detection and stability of immuno-reactive materials were enhanced by the use of AP-conjugated goat anti-rabbit immunoglobulin (Sigma, St. Louis, MO; 1:1000 dilu- tion). The nitrocellulose membrane was then developed for alkaline phosphatase activity following the procedure described by Blake _et _a_l_. (7) using 5-bromo-A-chloro-3-indolyl phosphate and Nitro Blue Tetra- zolium as substrates. Culture of SB-1 Cells The SB-1 cell line, derived from soybean roots (Glycine max (L.) Merr. cv. Mandarin) (8), was kindly provided by Dr. G. Lark (Depart- ment of Biology, University of Utah, Salt Lake City, UT) and cultured as previously described (1). For experiments on the metabolic labeling of SB-1 cells with 3580142-, the 185C medium was modified as follows: (a) casein hydrolysate was omitted; (b) magnesium sulfate and ammonium sulfate were replaced with magnesium chloride and 87 ammonium chloride, respectively. Actively growing SB-1 cells were washed and resuspended in the sulfate-free medium described above. H23580u (0.5 mCi/35 ml of medium) was added and the culture was carried out for four days. Isolation of Lectin Activity from SB-1 Cells Two protocols were used to identify and isolate lectin activity owxn SB-1 cells: (a) culture medium, and (b) digestion mixture after enzymatic removal of the cell‘wall. Each procedure is briefly described below. (a) Culture medium: Suspension culture of SB-1 cells were grown in 185C medium for 3-A days. The medium was separated fwwnn the cells by filtration through a Whatman filter paper. The pooled medium was adjusted to pH 7.A, then fractionated on Gal-Sepharose column (2). The material eluted with 0.2 M Gal was concentrated by Amicon ultra- filtration (PM 10 membrane). (i3) Supernatant fraction after cell wall removal: The cell wall was degraded by a modified procedure of Constabel (9). Actively growing EHl-1 cells (A days old) were washed with fresh 185C medium by centrifugathm1(A60 g, A min) and resuspension. The pelleted cells were then readjusted to the same volume with fresh 1B5Crmxfium and digested with an equal volume (100 ml) of enzyme solution containing 0.8 g pectinase (Sigma), 1.6 g cellulysin (Calbiochem, CA) and 10.0 g D-Sorbitol (Sigma), pH 5.5. After 2 hours incubation at 37°C, the digesticnirnixture was centrifuged (10,000 g, 30 min, A°C). The supernatant was adjusted to pH 7.A and fractionated on a Gal-Sepharose 88 column followed by elution with 0.2 M Gal. The eluted material was concentrated by Amicon ultrafiltration and dialyzed against 0.1% SDS. Isolation of Lectin Activity from Soybean Seedlings Soybean seeds, variety Williams, were germinated at rwxnn temper- ature ixl‘the dark for four days. The primary root tips including root hair regions were excised, frozen in liquid N2, and lyophilized. Dried sample was then extracted in PBS buffer (10 mM sodium phosphate, 0.1A M NaCl, A1m41flnq pH 7.A) (50 mg/ml) containing 5% 2-mercapto- ethanol and the clarified supernatant was obtained by centrifugation in a microfuge (Beckman). Isolation of Polypeptides from SDS-PAGE Purified seed SBA (8 mg) was subjected to preparative SDS-PAGE. Proteins were revealed in the gel by a short period of Coomassie-blue staining and destaining. The regions corresponding to the 60 kD and 30 kD polypeptides were sliced from the gel and then the proteins were eluted by incubation in 0.1% SDS (0.5 ml/1 cm slice) on a rotary shaker overnight at room temperature. The eluted material was either maintained in the SDS solution (room temperature) or was dialyzed against 50% methanol to remove SDS. In the latter case, methanol was evaporated and the dried sample was dissolved in PBS buffer and stored at A°C. Comparative Peptide Map Analysis Seed SBA and SB-1 lectin were compared after partial proteolysis following the procedure of Cleveland gt gt. (10). The isolated 89 proteins were subjected to SDS-PAGE in separate gels. Each individual gel was then soaked for 30 min in 15 ml of 125 mM Tris, pH 6.8, 0.1% SDS, and 1 mM EDTA. The gel was next placed horizontally at the top of a new slab of SDS-PAGE (15% acrylamide). Staphylococcus aureus V-8 protease (Miles Laboratories, Elkhart, IN; 250 ill at a concentration of 32 ug/ml) was overlayed on this and the digestion was allowed to continue for 30 min at room temperature. Electrophoresis was then continued subsequently. The protein fragments were revealed by Coomassie Blue staining. This yielded a comparison of the polypep- tides migrating in the Mr 30,000 region of the gel as only this region showed detectable dye staining. To compare the polypeptides migrating in the Mr 60,000 region of the gel, seed SBA and SB-1 lectin were first labeled with 1251 (3) as described above. The radioactive samples were subjected to SDS-PAGE and regions of the gels corresponding to polypeptides of Mr 30,000 and Mr 60,000 were excised from the gel using a gel slicer (2 mm per slice). The proteins in the gel slices were eluted with 0.1% SDS as described above. Samples containing 125I-labeled material from the 60 kD region of gels derived from seed SBA and SB-1 lectin were digested with V-8 protease (32 pg/ml) for 30 min at 37°C (10). After diges- tion, the samples were boiled for 10 min and then subjected to SDS-PAGE (15% acrylamide). The gel was dried and the radioactive polypeptides were revealed by autoradiography. The gel slices containing 125I-labeled 30 kD polypeptides from seed SBA and SB-1 lectin were used for two-dimensional peptide mapping analysis following the procedure of Elder gt Q: (11), with slight modifications. The SDS was first removed from the gel slices by three 9O washes with 50% methanol and then the gel slices were dried by a heat lamp. Each dried slice was treated with 0.25 ml trypsin (1 mg/ml in 50 mM NHuHCO3, pH 8) (Millipore Corporation, Freehold, NJ; 2A0 units/mg) at 37°C overnight. The supernatant fractions were separated from the gel pieces and evaporated to dryness. The dried samples were dissolved in 10 ul of buffer I (acetic acid/formic acid/water, 15:5:80) and 5 to 10 n1 (0.2-1o x 105 cpm) was spotted onto Avicel TLC plates (Analtech, Inc., Newark, DE). ElectrOphoresis was carried out at room temperature on a high voltage electrOphoresis apparatus (Brinkmann, Westbury, NY) in buffer I at 1 KV for about 30 min. The plates were then dried and the peptides were chromatographed in a second dimension in buffer II (butanol/pyridine/acetic acid/water, 32.5:25:5:20) with 7% (w/v) 2,5-diphenyloxazole. The plates were again dried and exposed on XRP-5 film (Kodak) for one week. Gel filtration and Radioimmunoassay Protein samples containing lectin activity were subjected to gel filtration on a column (1.5 x 5A cm) of Sephadex G-200 equilibrated with 10 mM sodium phosphate, 0.1 M NaCl, 0.2 M Gal, pH 7.A at A°C. Fractions of 2 ml were collected and were assayed for material reactive with rabbit anti-seed SBA by a solid phase radioimmunoassay. Microtitre wells (Immulon-2, Dynatech Labs, Alexandria, VA) were coated with 200 pl of rabbit anti-seed SBA (1 ug/ml in 0.1 M NaHCO3, pH 8.3) for 18 h at A°C. The wells were washed twice with Tris- buffered saline (20 mM Tris, 500 mM NaCl, pH 7.5) containing 0.05% Tween 20. Samples (100 111) were then added to the wells, followed by ”SI-labeled seed SBA (1 x 105 cpm) and incubated overnight at A°C. 91 The wells were washed three times with Tris-buffered saline to remove unbound radioactivity, and the amount of radioactivity bound in each well was determined in a gamma counter. The presence ofIHUabeled seed SBA in the sample results in a decreased binding of [125I]seed .SBA to the well. Immunoreactive SB-1 lectin could then be quantitated by the ability to compete with the labeled SBA for binding by the solid phase immunoglobulin. Seed SBA was used as a standard. «I. .r . . . . . all , r , o i . h\ .t L if“ *lv Q t L. C e a a $ ~ flk ‘ v .L - 6 Au To 1 . nu JO t s .1 1-. «a XL t... a u .. x 1 a a v. a . . .... PJ .1. Ru 9. A: .u a» ”u .3 ... v F. no {J 5 P1 mi 0 C d 74 7.. n- 0 n1 8 Tn. Cu (1 .U. a.) RESULTS Isolation of a Lectin from SB-1 Cells In previous studies (1), we had demonstrated the specificity of a rabbit antiserum directed against seed SBA. Using this highly specific antibody, we had shown, by immunofluorescence and by radioactive binding studies, the presence of a molecule cross-reactive with rabbit anti-seed SBA on the cell wall and plasma membrane of the cultured soybean cell line, SB-1. Moreover, a lectin-like activity could be identified in the supernatant fraction after SB-1 cells were digested with cellulysin and pectinase to remove cell wall material. In order to obtain sufficient material for more extensive characteri- zation, we surveyed a number of conditions to identify a source readily amenable to our purification procedures. We found that culture medium which had been exposed to SB-1 cells contained immuno- reactive material. The culture medium of SB-1 cells (72-hour collection) was subjected to affinity chromatography on Gal-Sepharose columns. The material bound by the affinity column was eluted by Gal (0.2M). Upon SDS-PAGE analysis, this material yielded a single band (Mr. - 30,000) (Figure 1B, lane 1), corresponding to that obtained with the 30 kD polypeptide of seed SBA (Figure 1A, lane 1). We have also cultured SB-1 cells in the presence of 3580u2‘ for 96 h to label the cellular components. The culture medium was then 92 93 Figure 1. SDS-PAGE analysis of seed SBA and SB-1 lectin purified by affinity chromatography on Gal-Sepharose column. (A) Purified seed SBA (- 10 pg of sample were electrophoresed in each lane): Lane 1, Coomassie Blue staining; lane 2, immunoblotting with rabbit anti-seed SBA and horseradish peroxidase-goat anti-rabbit immunoglobulin. (B) SB-1 lectin: Lane 1, Coomassie Blue staining; lane 2, immunoblotting with rabbit anti-seed SBA and horseradish peroxidase-goat anti-rabbit immunoglobulin (- 50 pg of sample were electrophoresed in these lanes); lane 3, fluorogram of purified SB-1 lectin (medium) derived from the medium of 3SS-labeled SB-1 cells. Approximately 2000 cpm (0.1 pg protein) were electrophoresed and the fluorogram was exposed for three weeks. (C) Extract of soybean root (- 130 pg of total protein was loaded in this lane) immunoblotted with rabbit anti-seed SBA and alkaline phosphatase-goat anti-rabbit immunoglobulin. The numbers on the left indicate the positions of migration of polypep- tides of Mr 30,000 and 60,000, relative to known molecular weight markers . 9A If N _. . win-1‘“-- Figure 1 coll bout flu: VA (“0 O "7 /\ (,AJ [(7 (T1 95 collected and fractionated by affinity chromatography. The material bound by the Gal-Sepharose column was subjected to SDS-PAGE and fluorography. The results showed one predominant radioactive poly- peptide (Mr. - 30,000; Figure 1B, lane 3). All of these data provide strong evidence that SB-1 cells synthesize a lectin with a subunit molecular weight and carbohydrate-binding capacity similar to seed SBA. We shall hereafter designate the affinity purified material as SB-1 lectin. When purified seed SBA was subjected to immunoblotting analysis with rabbit anti-seed SBA, one major polypeptide band (Mr. - 30,000) and one minor band (Mr - 60,000) were observed (Figure 1A, lane 2). The position of migration of the predominant band at 30 kD corres- ponded to the subunit.molecular weight of seed SBA (Mr - 30,000) (12) and will be hereafter designated seed SBA (30 kD). The material corresponding to the minor band (Mr - 60,000), which accounted for about 1% of the total protein, will be designated seed SBA (60 kD). Immunoblotting analysis of SB-1 lectin with rabbit anti-seed SBA yielded one predominant band (Mr 30,000) and one minor band (Mr 60,000) (Figure 1B, lane 2); this pattern was identical to that seen on immunoblotting analysis of purified seed SBA (Figure 1A, lane 2). For consistency of nomenclature, the material corresponding to the Mr 30,000 and Mr 60,000 region of the gel will be designated SB-1 lectin (30 kD) and ss-1 lectin (60 kD), respectively. Comparative Peptide Mgp Analysis of Seed SBA and SB-1 Lectin Seed SBA and SB-1 lectin were subjected to SDS-gel electrophor- esis. The individual lanes for the seed SBA and SB-1 lectin were 96 overlaid on a second polyacrylamide gel (15% acrylamide). The proteins separated in the original gels were subjected to limited proteolysis with V-8 protease and then electrophoresed and stained with Coomassie Brilliant Blue. This yielded a comparison of the peptide maps of seed SBA (30 kD) and SB-1 lectin (30 kD) since only this region of the gel showed detectable dye staining. The results showed that, in addition to residual undigested material (band e, Mr - 30,000, Figure 2A, lanes 1 and 2), both seed SBA (30 kD) and SB-1 lectin (30 kD) yielded four comparable fragments (bands a-d, Mrs 10,500 to 15,000, Figure 2A, lanes 1 and 2). There was too little material in the Mr 60,000 region of the gel to allow a comparison of seed SBA (60 kD) and SB-1 lectin (60 kD) by the above procedure. To increase the sensitivity of detection, seed SBA and SB-1 lectin were first radiolabeled with 1251. The radio- active samples were subjected to SDS-PAGE and 125I-labeled SBA (60 kD) and SB-1 lectin (60 kD) were extracted from the gel. After V-8 protease digestion and SDS-PAGE (on a 15% acrylamide gel), autoradio- graphic analysis revealed essentially identical patterns for seed SBA (60 kD) and SB-1 lectin (60 kD) (Figure 28, lanes 1 and 2). Both samples yielded four peptide fragments below the Mr 30,000 region (bands a-d, Figure 28, lanes 1 and 2), similar to those fragments generated from the 30 kD material (bands a-d, Figure 2A, lanes 1 and 2). Both seed SBA (60 kD) and SB-1 lectin (60 kD) samples also yielded a Mr 30,000 band (band c, Figure 28, lanes 1 and 2), which most probably corresponded to the undigested 30 kD polypeptide (band e in Figure 2A, lanes 1 and 2). Finally, both samples yielded a band migrating in the Mr A8,000 region (band f, Figure 28, lanes 1 and 2). a... Cw .54 r . 3. 1" a Ea‘ 'Va¢: . Mtg-.1 .— 97 Figure 2. (A) Comparative peptide map analysis of purified seed SBA and SB-1 lectin after limited V-8 protease hydrolysis and SDS-PAGE. Approximately 30 pg of protein were electrophoresed in a 10% acryl- amide gel; the 30 kD region was then excised and subjected to the protease V-8 (8 pg) and electrophoresed in a 15% acrylamide gel. The polypeptides were revealed by Coomassie Blue staining. (8) Peptide map analysis of radioiodinated peptides of 60 kD isolated from seed SBA (lane 1) and SB-1 lectin (lane 2). The labeled peptides (10,000 cpm) were subjected to limited V-8 protease (32 pg/ml) hydrolysis in a test tube; the digestion mixtures were electrophoresed in a 15% acryl- amide gel and revealed by autoradiography. The numbers indicate the molecular weights of the peptide fragments. The individual peptides were assigned with letters from a to g. 98 Figure 2 99 These results indicate that, at least at the one-dimensional peptide mapping level, there is no detectable difference iritnue polypeptides of seed SBA and SB-1 lectin. Seed SBA (30 kD) and S8-1 lectin (30 kD) were further compared by two-dimensional peptide mapping after tryptic hydrolysis. 1251- labeled seed SBA and SB-1 lectin were subjected to SDS-PAGE. Radioactive seed SBA (30 kD) and SB-1 (30 kD) were extracted owxn‘the gel and exhaustively digested with trypsin. The tryptic peptides were separated on thin layer chromatography plates by electrophoresis in the first dimension and chromatography in the second dimension. The separated radioactive peptides were detected by autoradiography (Figure 3, A and B). A detailed analysis of the films of the peptide maps, after long and short autoradiographic exposures, indicated that the number euui position of radioactive spots produced from both samples were identical. These results strongly suggest that seed SBA (30 kD) and SB-1 lectin (30 kD) were, in fact, identical. Interconversion of the 30 kD and 60 kD Forms of the Lectin In previous studies (1), we had shown that rabbit anti-seed SBA antibodies, affinity purified on the basis of binding to seed SBA (30 kD), also recognized seed SBA (60 kD). In the present study, limited digestion of seed SBA (60 kD) with V-8 protease yielded pep- tide fragments similar to those derived from similar treatment of seed SBA (30 kD) (lane 1 in Figure 2A and 28). Together, these data strong- ly suggested that seed SBA (60 kD) was a dimeric form of seed SBA (30 kD). The question is now raised as to whether this represents a ' 100 Figure 3. Two-dimensional peptide map of radioiodinated 30 kD poly- peptides from seed SBA and SB-1 lectin. Radioiodinated proteins were separated in a 10% SDS-PAGE and the 30 kD region was sliced from the gel. After removal of SDS from each gel slice by 50% methanol, the polypeptides in the gel slice were digested with trypsin (0.25 ml, 1 mg/ml). The digestion mixtures (- 105 cpm) were subjected to separa- tion by electrophoresis in the first dimension and chromatography in the second dimension as described in Experimental Procedures. The radioactive peptides were visualized by autoradiography. (A) seed SBA, (8) $341 lectin. Am .1, 101 4—meifioiemoiqo Figure 3 <— electrophoresis [+] [-1 J + [ CC 06 .3 Eu 102 covalent dimerization or non-covalent associatnnlcfi‘the 30 kD subunit. The frequent occurrence of the 60 kD material, even under strong denaturing conditions such as heating in buffers containing SDS, B-mercaptoethanol, and urea, suggested a possible covalent dimeriza- tion. The fact that V-8 digestion products of seed SBA (60 kD) yielded a band with a Mr - A8,000 (band f, Figure 28, lane 1) is consistent with the covalent dimer hypothesis. In more detailed analysis, however, we observed interconversion of the two forms of the lectin. For example, seed SBA was subjected to SDS-PAGE and seed SBA (30 kD) and seed SBA (60 kD) were separately isolated from their corresmmunng regions of the gel. When seed SBA (30 kD) was main- tained iml().1% SDS (2A h at room temperature) and subjected to re-electrophoresis, immunoblotting analysis revealed only a single band (M1. 30,000) (Figure A, lane 2). In contrast, when seed SBA (30 kD) was hunmwted in PBS (A-5 days at A°C), re-electrophoresis and immunoblotting yielded bands at both Mr 30,000 and 60,000 (Figure A, lane 1). Conversely, seed SBA (60 kD) yielded predominantly a Mr 60,000 band after incubation in PBS (A-5 days at room temperature) followed by SDS-PAGE and immunoblotting (Figure A, lane 3). After prolonged incubation in 0.1% SDS (A-S days at room temperature), however, there was a significant amount of conversion into the Mr 30,000 polypepthfla(F&gure A, lane A). These results indicate that seed SBA (60 kD) most probably represent non-covalently associated dimers of the 30 kD subunit. 1 . .2 fl. 2. A: «G .1 $ C n m. a . Ml U U be Cu Ho 4%. Pd 3 a J 3. ,J .a . «<4 U OJ. ..u 3 e . C: H. I n. all. 0 n1... \ i F T. «O .. . r. .3 L 0 .. H 2 r. to ..u. .1 A a .. . 2* «D .1 no r, ... .1 . : . an n.» 5 Q» “l 1" 40318 c 10““ u 5r- 103 Figure A. thterconversion of 60 and 30 kD polypeptides from seed SBA and purified SB-1 lectin analyzed by SDS-PAGE and immunoblotting. The samples were electrophoresed, transferred to nitrocellulose paper, and immunoblotted with rabbit anti-seed SBA and alkaline phosphatase- conjugated goat anti-rabbit immunoglobulin. Lane 1, seed SBA (30 kD) (= 3 pg) stored in PBS at A°C. Lane 2, seed SBA (30 kD) (= 3 pg) incubated in 0.1% SDS at room temperature. Lane 3, seed SBA (60 kD) (= 0.5 pg) stored in PBS at A°C. Lane A, seed SBA (60 kD) (2 2 pg) incubated in 0.1% SDS at room temperature. Lane 5, SB-1 lectin (100 ng) derived from the cell wall of SB-1 cells purified by affinity chromatography on Gal-Sepharose column and stored in PBS. Lane 6, the same material as in lane 5 (100 ng) after prolonged incubation in 0.1% (SDS at room temperature. The numbers on the left indicate the posi- tions of migration of polypeptides of Mr. 30,000 and 60,000 relative to known molecular weight markers. 10A 60K- 9 e...- 1 .1 30K- ”i «~— ' 11;; 123456 Figure A 105 In a similar fashion, SB-1 lectin also appears to exhibit non-covalent association of its subunits, even under denaturing conditions of SDS-PAGE. SB-1 lectin stored in PBS yielded predomi- nantly a Mr 60,000 band upon subsequent SDS-PAGE and immunoblotting (Figure A, lane 5). The same preparation of SB-1 lectin, after prolonged incubation (A-5 days at room temperature) in 0.1% SDS, yielded a Mr 30,000 polypeptide (Figure A, lane 6). SB-1 lectin isolated from the medium, from the supernatant fraction after cell wall digestion, and from detergent extracts of SB-1 cells all yielded similar results. Characterization of SB-1 Lectin When purified SB-1 lectin was subjected to gel filtration on Sephadex G-200 equilibrated with 10 mM phosphate buffer containing 0.1 M NaCl and 0.2 M Gal, a single component was observed using a radioimmunoassay for seed SBA. The position of migration of the component corresponded to the tetrameric form of seed SBA (Mr - 110,000) observed previously under non-denaturing conditions (12). These results suggest that SB-1 lectin also forms tetrameric structures of the basic subunit. In order to probe the saccharide-binding speciflnflty of SB-1 lectin, various sugars were tested for their capacity to elute the lectin bound in the Gal-Sepharose affinity column. When the column was developed sequentially with mannose, glucose, and GalNAc, the lectin was observed only upon the addition of GalNAc (Figure 58, lane 1-3), as revealed by SDS-PAGE and immunoblotting analysis. After the elution with GalNAc, no additional lectin was eluted upon :u In.» C» A \y n . 9. 61V .4‘ u ~ ~ s 5‘. QV 61¢ mu u. s 2.. :1.‘ '9‘». ~61" we '106 Figure 5. SDS-PAGE and immunoblotting analysis of seed SBA and SB-1 lectin obtained from Gal-Sepharose column upon sequential elution with various saccharides. Samples were electrophoresed, transferred to nitrocellulose paper, and immunoblotted with rabbit anti-seed SBA and alkaline phosphatase-conjugated goat anti-rabbit immunoglobulin. (A) Seed SBA: lane 1, material eluted with 0.2 M mannose; lane 2, material eluted witti().2 M glucose; lane 3, material eluted with 0.2 M GalNAc; lane A, material eluted with 0.2 M Gal. Equal volume from each sample was loaded in the gel. (B) SB-1 lectin derived from the cell wall of SB-1 cells (- 15 8) purified on Gal-Sepharose column: Lane 1, material eluted with 0.2 M mannose; lane 2, material eluted with 0.2 M glucose; lane 3, material eluted with 0.2 M GalNAc; lane A, material eluted with 0.2 M Gal. The numbers on the left indicate the positions of migration of polypeptides relative to known molecular weight markers. 107 Figure 5 108 further development of the column with Gal (Figure 58, lane A). Inasmuch as Gal was used in the original isolation of SB-1 lectin, these results suggest that the material eluted with GalNAc was the same as that eluted with Gal. More importantly, the results of using different saccharides to elute the lectin bound on the Gal-Sepharose column were identical to those obtained when seed SBA was subjected to parallel analysis (Figure 5A, lanes 1-A). These results indicate, therefore, that the carbohydrate-binding specificity of seed SBA and SB-1 lectin were the same. Identification of the Lectin in Soybean Roots Although the SB-1 cell line was derived originally owMn roots, the expression of the lectin molecule that appears to be identical to seed SBA may only reflect the dedifferentiated state of the cell rather than the state of true lectin expression in root tissues. Therefore, we have carried out a parallel analysis for the presence of the lectin in soybean roots. Homogenates of room tissue were subjected to SDS-PAGE and immunoblotting analysis with rabbit anti-seed SBA. The immunoblots revealed two bands (Mr 30,000 and 60,000),vnuch co-migrated with seed SBA (30 kD) and seed SBA (60 kD), respectively (Figure 1C). In contrast, preimmune serum failed to blot either the 30 kD or the 60 kD bands in root extracts. “These resultxs suggest that root tissues contain a lectin whose polypeptide composition corresponds to that of seed SBA and SB-1 lectin. This conclusion was supported by affinity chromatography studies. Root extracts fractionated on Gal-Sepharose columns yielded, upon 109 specific elution with Gal, polypeptides of Mr 30,000 and 60,000 corresponding to seed SBA (30 kD) and seed SBA (60 kD). .FU 5L DISCUSSION Soybean agglutinin is a well-characterized glycoprotein lectin from seeds of the cultivated soybean, Glycine max (L.) Merr. The lectin (Mr = 120,000) is a tetramer with identical subunits (Mr = 30,000). It displays carbohydrate-binding specificity for GalNAc and, to a lesser extent, for Cal. The concentration of SBA in the seed can be quite substantial (13-15), but the distribution and origin of the lectin in other organs and tissues, particularly the roots, is a subject of much controversy (15,16). There are several reports that crude extracts of soybean roots contain hemagglutinating activity (1A,17,18); such activity is insufficient reason, however, to conclude that the roots contain SBA. Gade g_t_ _t. (19) isolated and characterized a lectin from Chippewa soybean roots. This lectin was shown to be similar to seed SBA in terms of structural characteristics, immunological reactivity, and carbohydrate-binding specificity. Although there was some question concerning the original source of the root lectin, Gade gt gt. (20) have provided additional evidence that the root lectin is endogenous to the tissue of analysis. These results are consistent with observations that a soybean lectin can be found in root exudates _i_rt m (21), that callus cultures derived from soybean roots express SBA-like material on the cell surface (22) and callus cultures enriched in root hair cells elaborate appreciable amounts of soybean 110 111 lectin (23). Attempts to detect SBA on the surfaces of soybean roots and root hairs by immunochemical techniques have yielded conflicting results; some groups have found cross-reactive polypep- tide(s) (2A-26), while other have failed to detect it or any immunologically reactive molecule (13,16). The results of our present experiments document the properties of a lectin produced endogenously by the cultured SB-1cxfll.line, which was originally derived from soybean roots (8). We had shown previously that fluorescently labeled rabbit anti-seed SBA yielded specific immunofluorescent staining on the cell wall and plasma membrane of the SB-1 cells, implicating the presence of a SBA-like molecule (1). We had also shown that the same rabbit anti-seed SBA can block tflua specific adhesion of Rhizobium japonicum to the cultured SB-1 cells. It was important, therefore, to characterize the structure and activities of the SB-1 lectin, one likely candi- date that may mediate the recognition between Rhizobium and the SB-1 cells. Because lectins in soybeans belong to multi-gene families (27), it was particularly important to establish the relationship between SB-1 lectin and seed SBA, whose gene structure and eXpression has been analyzed in detail (27,28). SB-1 lectin.appears to be identi- cal tn) seed SBA on the basis of the following criteria: (a) peptide mapping of the polypeptide subunit (Mr. 30,000); (b) subunit struc- ture and oligomerization; (c) carbohydrate-binding specificity: and (d) immunological cross-reactivity. 112 In the course of our studies, we found that both seed SBA and SB-1 lectin yielded two bands (Mr's 30,000 and 60,000) in immunoblotting experiments after SDS-PAGE. More strikingly, the Mr 30,000 and Mr 60,000 bands were interconvertible. The conditicnus of incubation were important in determining the extent of interconversion. Thus, seed SBA (60 kD) maintained in PBS yielded predominantly a Mr 60,000 band upon SDS-PAGE and immunoblotting. In contrast, the same sample of seed SBA (60 kD), after prolonged incubation in SDS, yielded both Mr 30,000 and Mr 60,000 bands. These results suggest that seed SBA (60 kD) represents a dimer consisting of the non-covalently associated M, 30,000 subunits. Aggregation and dimerization of seed SBA has also been observed previously (29). Similar results were obtained for SB-1 lectin. The stability of the 60 kD dimer, even under strong denaturing conditions such as heating in buffers containing SDS, B-mercaptoethanol, and urea, was surprising. These observations on the SB-1 lectin of a cultured soybean cell line and seed SBA need to be put in perspective in the context of other studies on comparing lectins from seeds and roots. In Pisum sativum, the root lectin yielded polypeptides that had similarities (molecular weights and immunological reactivity) and differences (isoelectric points, saccharide specificity, and hemagghndnation activity) with the seed lectin (30). A similar conclusion was obtained when lectins of Phaseolus vulgaris were compared for seed and non-seed tissues including roots (31). By contrast, tun) distinct lectins in terms of polypeptide composition, native molecular weight, carbohydrate-binding specificity, and 113 reactivity with antibodies were isolated from the roots and seeds of Lotononis bainesii (32). There was no evidence of any relatedness between the lectins from the two different organs/tissues of the same plant. Finally, Etzler and co-workers (33-35) have carried out an extensive analysis of the lectins derived from the Dolichos biflorus at the molecular, subcellular, and systemic levels. This plant contains at least two lectins which are structurally related. Callus cultures from the epicotyl and leaves, hypocotyl, and roots showed no seed lectin; in contrast, lectins that were isolated from these cultures had molecular properties similar to those of the lectin isolated originally from stems and leaves (36). In the soybean system, a lectin has been isolated from roots on the basis of its carbohydrate-binding activity and has been shown to be similar to the well-characterized seed SBA (19,20). More recently, Vodkin and Raikhel (26) have demonstrated that a protein (Mr. 33,000), reactive with antibodies directed against seed SBA, can be found in the roots of four soybean varieties, regardless of whether SBA is present (Le+) or absent (Le') in the seed. The 30 kD seed SBA subunit, however, was not detectable in the roots. These results should be contrasted with our present studies, in which we do detect polypeptides corresponding to seed SBA (30 kD) and seed SBA (60 kD) in extracts of roots. In this connection, it should be noted that in our SDS-PAGE analysis of root extracts, the Mr 33,000 region represented a section of the gel that contained large amounts of protein (detected by silver or Coomassie Blue staining). Under these conditions, a 33 kD band can be observed in certain immuno- blots, including those with preimmune serum. This calls for caution in 51: in 33 Le 2.. Co K (‘1 a .‘lv 11A in interpreting the significance of the Mr 33,000 band in immuno- blots, as it may represent non-specific detection of a polypeptide in root extracts. In any case, it does not appear that the Mr 33,000 polypeptide in soybean roots represents a gene product of the Le 1 gene, which encodes the Mr 30,000 seed SBA subunit (37). The relationship of the Mr 33,000 polypeptide with the Mr 30,000 seed SBA subunit remains, therefore, to be established. In our previous studies (1), we had noted that most saccharides known to bind to seed SBA also inhibited the binding of Rhizobium japonicum to SB-1 cells. The lone exception was GalNAc, which bound to seed SBA but did not inhibit Rhizobium-SB-1 cell adhesion. One possible explanation was that SB-1 lectin was different from seed SBA, despite their immunological cross-reactivity. The results of our present paper indicate, however, that the lectins from the two sources were the same, even in terms of their carbohydrate-binding specificity. In light of these observations, we are now forced to face alternative hypotheses to account for the failure of GalNAc to inhibit Rhizobium adhesion to SB-1 cells: (a) lectin anchored on the cell wall may bind saccharides with different specificity than solubilized lectin; (b) the saccharide specificity in inhibition of Rhizobium-SB-1 cell binding may reflect additional recognition components which may exhibit Gal-binding characteristics. The latter notion of dual recognition, involving other sets of comple- mentary molecules, would be greatly advanced if a protein molecule mediating such interactions could be identified, isolated, and characterized. h! 10. 11. REFERENCES Ho. S.-C., Malek-Hedayat, 8., Wang, J.L., and Schindler, M. (1986) J. Cell Biol. 103, 10A3-105A Allen, A.K., and Neuberger, A. (1975) FEBS Lett. 50, 362-36A Ho. S.-C., Izumi, H., and Michelakis, A.M. (1982) Biochim. Biophys. Acta 717, A05-A13 Laemmli, U.K. (1970) Nature (Lond.) 227, 680-685 Bonner, W.M., and Laskey, R.A. (197A) Eur. J. Biochem. A6, 83-88 Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, A350-A35A Blake, M.S., Johnston, K.H., Russell-Jones, G.J., and Gotschlich, E.C. (198A) Anal. Biochem. 136, 175-179 Gamborg, 0.L., Miller, R.A., and Ojima. K; (1968) Exp. Cell. Res. 50, 151-158 Constabel, F. (1975) in Plant Tissue Culture Methods (0.L. Gamborg and L.R. Wetter, eds.) pp. 11-21. National Research Council of Canada, Saskatoon, Saskatchewan, Canada Cleveland, D.W., Fisher, S.G., Kirschner, M.W.,:nuiiaemmli, U.K. (1977) J. Biol. Chem. 252, 1102-1106 Elder, J.H., Pickett, R.A., Hampton, J., and Lerner, R.A. (1977) J. Biol. Chem. 252, 6510-6515 115 12. 13. 1A. 15. 16. 17. 18. 19. 20. 21. 22. 23. 2A. 25. 116 Lotan, R., Siegelman, H.w., Lis, H., and Sharon, N. (197A) J. Biol. Chem. 2A9, 1219-122A Liener, I.E. (1976) Ann. Rev. Plant Physiol. 27, 291-319 Pueppke, S.G., Bauer, W.D., Keegstra, K., and Ferguson, A.L. (1978) Plant Physiol. 61, 779"78A Pull, S.P., Pueppke, S.G., Hymowitz,'L” and Orf,IiuL (1978) Science 200, 1277-1279 Su, L.C., Pueppke, 8.0., and Friedman, H.P. (1980) Biochim. Biophys. Acta 629, 292-30A Bohlool, 8.8., and Schmidt, E.L. (197A) Science 185, 269-271 Bowles, ILAJ., Lis, H., and Sharon, N. (1979) Planta 1A5, 193-198 Gade, W., Jack, M.A., Dahl, J.B., Schmidt,ILJ“, and Wold, F. (1981) J. Biol. Chem. 256, 12905-12910 Gade, W., Schmidt, E.L., and Wold,1h.(1983) Planta 158, 108-110 Sengupta-Gopalan, C., Pitas, J.H., and Hall, T.C. (198A) in Advances huthrogen Fixation (Veeger, C., and Newton, W.E., eds.) PP. A27, Nyhoff Publishers, The Hague, Holland Del Campillo, E., Howard, J., and Shannon, L.M. (1981) Z. Pfanzenphysiol. 10AS, 97-102 Hermina, N., and Reporter, M. (1977) Plant Physiol. 59, 97-102 Staceyy (L., Paan, A.S., and Brill, W.F. (1980) Plant Physiol. 66, 609-61A Tsien, H.C., Jack, M.A., Schmidt, E.L., and Wold, F.(1983) Planta 158, 128-133 27. .1.) 26. 27. 28. 29. 30. 31. 32. 33. 3A. 35. 36. 37. 117 Vodkin, L.0., and Raikhel, N.V. (1986) Plant Physiol. 81, 558-565 Goldberg, R.B., Hoschek, G., and Vodkin, LJl.(1983) Cell 33, A65-A75 Vodkin, L.O. (1981) Plant Physiol. 68, 766-771 Lotan, R., Lis, H., and Sharon, N. (1975) Biochem. Biophys. Res. Comm. 62, 1AA-150 Gatehouse, J.A., and Boulter, D. (1980) Physnfl” Plant. A9, A37-AA2 Borrebaeck, C.A.K. (198A) Planta 161, 223-228 Law, I.J., and Strijdom, B.W. (198A) Plant Physiol. 7A, 773‘778 Etzler, M.E. (1981) Phytopathology 71, 7AA-7A6 Etzler, M.E., MacMilJJni, S., Scates, 8., Gibson, D.M., James, D.W.,.kn, Cole, D.,anuiThayer, S. (198A) Plant Physiol. 76, 871-878 Roberts, D.M., and Etzler, M.E. (198A) Plant Physiol. 76, 879-88A James, D.W., Ghosh, PL, and Etzler, M.E. (1985) Plant Physiol. 77, 630-63A Vodkin, lull. (1983) in Chemical Taxonomy, Molecular Biology, and Function of Plant Lectins (Goldstein, I.J., and Etzler, M.E., eds.) pp. 87-98. Alan R. Liss, Inc., New York, New York. Chapter IV ENDOGENOUS LECTIN FROM CULTURED SOYBEAN CELLS: SB-1 LECTIN ON THE CELL WALL 118 yi re SC In (A ABSTRACT Digestion of seed soybean agglutinin (SBA) with V-8 protease yielded seven distinct peptides (Mr. 10,000 - 20,000) that were well- resolved by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Each individual peptide (F1 through F7) was isolated; determination of the amino acid sequence at the NH2-terminal portion of each peptide established its position in the sequence of SBA. The isolated peptides were used as affinity adsorbents to obtain anti-peptide antibodies (anti-F1 through anti-F7). These anti-peptide antibodies were used in a comparative study to explore: a) potential conformational differences between SBA in solution and the structur- ally identical soybean cell lectin SB-1 attached to the cell wall surface, and b) examine the polypeptide presentation of cell surface SB-1 lectin to the environment. Results of fluorescence analysis of whole cell labeling and affinity chromatography methods suggest that the peptide sites available for antibody binding in non-denatured SBA are exposed and demonstrate similar antibody reactivity for cell sur- face anchored SB-1 lectin. In addition, the pattern of anti-peptide reactivity demonstrated for SB-1 lectin in the cell wall suggest that the C-terminal half of the molecule is involved in cell surface anchoring. 119 ar .8. 1i INTRODUCTION The cultured soybean cell line, SB-1, synthesizes a lectin desig- nated the SB-1 lectin (9). In previous studies (11), we demonstrated that SB-1 lectin was identical to seed soybean agglutimn1(SBA) hi terms of: (a) molecular weight of the polypeptide (Mr 30,000), detwnxnined by sodium dodecyl sulfate (SDS) polyacrylamide gel electro- phoresis (PAGE); (b) molecular weight of the native protein (Mr 120,000 tetramer); (c) carbohydrate-binding specificity; and (d) immunological reactivity with a rabbit antiserum raised against seed SBA. This lectin was detected on the cell wall of intact SB-1 cells by the binding of fluorescently-labeled or radioactively-labeled rabbit anti-seed SBA. Moreover, quantitative studies showed that the amount of SB-1 lectin detectable at the cell surface was not affected ‘By the addition of specific saccharide ligands (9). These data indicated that the molecule was not anchored on the cell surface via its carbohydrate binding properties, but by some other mechanism. A question raised at this point was one of integration and presentation of the SB-1 polypeptides on the cell surface and, in particular, what regions of the polypeptide chains serve to hook the lectiJi‘to the cell wall. One approach to this problem is to probe the cell surface with SBA sequence specific antibodies and examine extents of reactivity. Since, in the course or our previous studies (11), we had observed that V-8 protease digestion of SBA and SB-1 lectin 120 re fr US 121 resulted in a number of peptides (Mr 10,000 - 20,000), we initiated an effort to prepare anti-peptide antibodies purified by immunoaffinity from polyclonal anti-denatured SBA antibodies. These antibodies were used to examine the exposed sequences of SB-1 lectin on soybean cells. Ne MATERIALS AND METHODS Preparation of Seed SBA and Anti-SBA Antibodies Seed SBA was isolated and purified by the procedure of Allen and Neuberger (1) on galactosamine-Sepharose affinity column. The details of this procedure, as well as the characterization of the purified lectin, have been reported previously (9). Seed SBA was first denatured by boiling in 0.1% SDS for 5 min. Antiserum to the denatured SBA were raised in rabbits (New Zealand, white, female). The primary injection consisted of 0.2 mg of protein in Freund's complete adjuvant (Gibco Laboratories). Booster injec- tions of 0.2 mg denatured protein in Freund's incomplete adiuvant were administered at weekly intervals. Antiserum was collected three days after the last booster, and the immunoglobulin fraction was purified by affinity chromatography on a protein A-Sepharose (Pharmacia) column. This is designated as anti-seed SBA. Polyacrylamide Gel Electrophoresis and Immunoblotting Polyacrylamide gel electrOphoresis in the presence of sodium dodecyl sulfate (SDS-PAGE) was performed by the method of Laemmli (10) using 12.5 and A1 (w/v) acrylamide in the running and stacking gels, respectively. Following electrophoresis, the proteins were revealed either by staining with Coomassie brilliant blue, with the silver reagent (6,17), or by immunoblotting. For immunoblotting, the 122 prot and mat phos 100 usi Pr! Th ([7 ("W (1, 123 proteins were electrotransferred to nitrocellulose paper (Schleicher and Schuell) by the method of Towbin gt a. (15). The immunoreactive material on nitrocellulose membrane were revealed via alkaline phosphatase-conjugated goat anti-rabbit immunoglobulin (Sigma, 1 to 1000 dilution) following the procedure described by Blake fl al. (3) using 5-bromo-A-chloro-3-indolyl phosphate and nitro blue tetrazolium (Sigma) as substrates. Enzymatic Digestion of Seed SBA Purified seed SBA was digested with Staphylococcus aureus V-8 protease (Miles Laboratories) as described by Cleveland e_t _l. (A). The digestion was carried out with 32 ug/ml enzyme for 2 hr and then reaction was terminated. The digests were kept at -80°C. Peptides generated via V-8 hydrolysis were separated by SDS-PAGE. The gel was sliced into 1-2 mm sections, which were then rinsed for 10 min with water. The peptides were eluted from the gel slices by soaking in 0.05 M Tris-H01, pH 7.9. containing 0.1% SDS, 0.1 mM EDTA, 0.15 M NaCl, 5.0 mM dithiothreitol (0.A ml/slice) for at least A-6 h at 25°C as described by Hager and Burgess (8). Proteins were precipi- tated by acetone. Four volumes of cold acetone (-20°C) were added to one volume of the gel eluate in a silanized 30 ml Corex tube and allowed to precipitate for 30 min in a dry ice-methanol bath. The samples were next centrifuged for 10 min (9,000 x g, A°C). This procedure, almost completely removed Coomassie blue, as well as SDS. The supernatant was discarded and the tubes were tilted to allow the evaporation of residual acetone and water. Dried samples were resuspended in 0.1 ml of 0.05% SDS and sonicated for 2 min. These sampl Instr Faci amin: nl 12A samples were analyzed by re-electrophoresis and by amino acid sequence determination. Amino acid sequence analyses were performed on a Beckman Instrument 890M (Michigan State University, Macromolecular Structure Facility) by the Edman degradation method (5). Phenylthiohydantoin amino acids were identified by high pressure liquid chromatography (13). Preparation of Anti-Peptide Antibodies by Specific Adsorption to Poly- peptides Immobilized on Nitrocellulose Membrane To prepare anti-peptide antibodies, seed SBA was digested with V-8 protease and then subjected to SDS-PAGE to separate the resulting peptides. They were then electrotransferred to nitrocellulose paper. The transferred proteins on nitrocellulose were stained with amido black for 2 minutes. This revealed the exact position of each peptide. Nitrocellulose strips containing individual peptides were carefully cut out and incubated for A h with Tris-buffered saline (20 mM Tris, 500 mM NaCl, pH 7.5) containing 0.05% Tween-20 and 5% bovine serum albumin (BSA) in order to block nonspecific sites. This procedure also removed most of the amido black stain from the membrane. Each of the peptide-containing nitrocellulose strips was then incubated overnight with 1.0 ml of Tris buffered saline containing anti-seed SBA (0.5 mg/ml) overnight. Unbound antibodies were removed by three washes with Tris buffered saline containing 0.05% Tween-20. The bound antibodies were then eluted from each peptide containing nitrocellu- lose by the method described by Smith and Fisher (1A), with the modification that only 5 mM glycine, pH 2.3 was used for elution. The elu1 of: The pho 125 eluted antibody preparations were immediately neutralized by addition of 2 M Tris-base, pH 7.5, and dialyzed against water extensively. These samples were then lyophilized and redissolved in 50 pl of phosphate buffered saline, pH 7.A. Culture and Immunolabeling of SB-1 Cells The SB-1 cell line, derived from soybean roots (Glycine max (L.) Merr.