Characterization of a New Lectin of Soybean Vegetat ive ... · column with buffer B plus 500 mM a-methyl-D-mannopy- ranoside. Fractions containing bound protein were pooled, concentrated,

Plant Physiol. (1996) 110: 825-834

Characterization of a New Lectin of Soybean Vegetat ive Tissues'

Steven R. Spilatro*, Graham R. Cochran, Ruth E. Walker, Khari 1. Cablish, and Christine C. Bittner

Department of Biology, Marietta College, Marietta, Ohio 45750-401 6

Lectins are carbohydrate-binding proteins that occur widely among plants. Lectins of plant vegetative tissues are less well characterized than those of seeds. Previously, a protein of soybean (Clycine max [L.] Merr.) leaves was shown to possess properties similar to the seed lectin. Here we show that the N-terminal amino acid sequence of this protein shares 63% identity with the seed lectin. lmmunoblot analysis indicated that the protein occurs in leaves, petioles, stems, and cotyledons of seedlings but not in seeds. These observations prompted designation of the protein as a soybean vegetative lectin (SVL). lmmunohistochemical localization in leaves indicated that SVL was localized to the vacuoles of bundle- sheath and paraveinal mesophyll cells. Removal of sink tissues or exposure to atmospheric methyl jasmonate caused increased levels of SVL in leaves and cotyledons. Co-precipitation of SVL and the soybean vegetative storage protein (VSP) during purification suggested an interaction between these proteins. SVL-horseradish peroxidase conjugate bound to dot blots of VSP or SVL, and binding was inhibited by porcine stomach mucin and heparin but not simple carbohydrates. Binding between SVL and VSP and similarities in localization and regulation support a possible in vivo interaction between these proteins.

Lectins are abundant carbohydrate-binding proteins that are widely distributed but of uncertain function in plants (Etzler, 1986; Goldstein and Poretz, 1986; Kijne et al., 1986; Pusztai, 1991). Lectins of leguminous seeds are among the most extensively studied of the plant lectins. The lectins found in legume seeds make up a family of related proteins sharing structural and amino acid sequence homologies (Strosberg et al., 1986; Sharon and Lis, 1990). Typically, these lectins are glycoproteins consisting of subunits with molecular masses in a range of 25 to 35 kD arranged as dimers or tetramers. Lectins are particularly abundant in the seeds of legumes, making up as much as 10% of the total seed protein (Etzler, 1986). The seed lectins accumulate in protein storage vacuoles of cotyledons and are de- graded during seed germination and maturation of the seedling (Pusztai, 1991). Lectins also occur in vegetative organs including leaves, stem, roots, and bark (Borrebaeck, 1984; Etzler, 1986; Van Damme and Peumans, 1990; Ho Pak et al., 1992). In general, vegetative lectins are present at much lower levels than those of seeds and have not been as

This material is based on work supported by the National Science Foundation under grant No. DCB-9104525. The govern- ment has certain rights to this material.

* Corresponding author; e-mail spilatrs8mcnet.marietta.edu; fax 1- 614 -376 - 4896.

825

well characterized. Although some vegetative lectins possess strong homologies with seed lectins, others display unusual carbohydrate-binding characteristics and an in- ability to agglutinate erythrocytes (Etzler, 1986, 1994a; Pusztai, 1991).

A protein found in leaves of soybean (Glycine max [L.] Merr.) plants was shown to possess structural and immunological similarities to SBA, the major lectin of soybean seeds (Spilatro and Anderson, 1989). Like SBA, the soybean leaf protein is a 120-kD glycoprotein with a tetrameric arrangement of subunits. The subunits of the leaf protein have molecular masses of 28 and 33 kD and react with antibodies raised against the seed lectin (Spilatro and Anderson, 1989). In this investigation we provide further evidence that the leaf protein, henceforth referred to as a SVL, is widely distributed in soybean plants and is a member of the legume lectin family.

Despite their ubiquity, the specific functions of vegetative lectins are unresolved. A variety of studies have supported the hypothesis that lectins found in roots of some plants may be involved in symbiotic interactions between the plants and microorganisms (Pueppke et al., 1978; Pusz- tai, 1991). However, functions suggested for lectins in other vegetative tissues are based largely on conjecture. Roles suggested for these lectins relate to plant defense and stress physiology, carbohydrate metabolism, and packaging of storage proteins (Rudiger, 1984; Etzler, 1986; Kijne et al., 1986).

The physiological functions of plant lectins will not be fully understood until the identity and biological function of endogenous receptors are elucidated (Kijne et al., 1986). No such receptors have been unequivocally identified for plant lectins. Previously, we reported that SVL and the soybean VSP co-precipitate during purification of SVL from leaf homogenate and suggested a possible interaction between these two proteins (Spilatro and Anderson, 1989). VSP is a dimeric glycoprotein comprising subunits with molecular masses of approximately 27 and 29 kD (Witten- bach, 1983) arranged as homo- and heterodimers (Spilatro and Anderson, 1989). VSP is widely distributed among plant tissues, including stems, flowers, seed pods, and cotyledons (Mason et al., 1988; Staswick, 1989), where it is believed to provide temporary storage for nitrogen and photoassimilates (Wittenbach, 1983; Staswick, 1990). In

Abbreviations: HRP, horseradish peroxidase; SBA, soybean agglutinin; SVL, soybean vegetative lectin; VSP, soybean vegetative storage protein.

https://plantphysiol.orgDownloaded on December 31, 2020. - Published by Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.

https://plantphysiol.org

826 Spilatro et al. Plant Physiol. Vol. 110, 1996

leaves VSP is stored in the vacuoles of bundle-sheath and paraveinal mesophyll cells (Franceschi et al., 1983), a tissue specialized for synthesis, storage, and remobilization of photoassimilates (Franceschi and Giaquinta, 1983). Re- mova1 of seed pods, phloem girdling, and other treatments that disrupt the transport of photoassimilates to sink organs induce an accumulation of VSP (Wittenbach, 1983; Staswick, 1990). SVL in leaves also increases in sink- deprived plants (Spilatro and Anderson, 1989), although other factors that may regulate SVL have not been investigated. One such factor is methyl jasmonate, a natural substance that occurs widely among plants. Methyl jasmonate has been shown to induce gene expression and accumulation of VSP and other proteins in soybean plants (Anderson et al., 1989; Franceschi and Grimes, 1991; Staswick et al., 1991; for review, see Staswick, 1992).

In this paper we report further characterization of SVL. We show that SVL and VSP are co-localized at the organ and cellular levels and display similar responses to sink tissue deprivation and methyl jasmonate. Furthermore, we have used SVL conjugated with HSP to demonstrate binding between SVL and VSP.

MATERIALS A N D METHODS

Plant Materials

Soybean (Glycine max [L.] Merr. var FFR-343) plants were grown in a controlled environment chamber under a 14-h photoperiod in Sunshine I11 potting mix (Sungrow Horti- culture, Bellevue, WA). Plants were fertilized weekly with a nutrient solution (Streeter, 1978) containing 20 mM NH,NO,. In some experiments leaves were harvested from plants at anthesis. Other plants were allowed to set seed normally (control) or their seed pods were removed daily (depodded plants), and tissues were harvested 30 d postanthesis. In other experiments, seedlings were allowed to grow normally (control) or, beginning 4 d postemergence, their epicotyl and axillary shoots were removed (sink deprived). Cotyledons were harvested from seedlings 10 d postemergence. Seed pods and seeds designated “immature” were harvested 20 d postanthesis from plants allowed otherwise to set seed normally. Dry seeds were designated ’“ature.”

General Procedures and Tissue Homogenization

A11 procedures were carried out at 4”C, and a11 centrifu- gations were performed at 12,OOOg. A11 protein chromatography procedures were performed using a Pharmacia Fast Protein Liquid Chromatograph. Protein fractions were concentrated using an ultrafiltration cell with a PM-30 membrane (Amicon, Beverly, MA). A11 percentage concentrations given below are weights/volume.

Plant tissues were homogenized in buffer A (50 mM Tris-HC1, pH 7.5, 1 mM EDTA) containing 30 mM ascorbic acid, 28 mM 2-mercaptoethanol, and 0.03 g/mL insoluble PVP (20 g tissue/100 mL buffer) with a mortar and pestle or an Ultra-Turrax T-25 homogenizer (Janke & Kunkel, Staufen, Germany) for large sample volumes. Crude ho-

mogenates were passed through two layers of cheesecloth and then centrifuged.

Purification of SVL

SVL and VSP were purified from leaves of depodded plants. Ammonium sulfate was added sequentially to centrifuged leaf homogenate to final concentration:, of 25, 45, 55, and 75% saturation, with each precipitate removed by centrifugation. The 55% ammonium sulfate precipitate was resolubilized in buffer A plus 14 mM 2-mercaptoethanol and dialyzed overnight against 2 X 500 mL of this buffer. The precipitate that formed during dialysis wa:j collected by centrifugation, washed twice with buffer A, and redissolved in buffer A plus 1 M NaC1. This fraction is referred to as the ”dialysis precipitate.” Dialysis precipiteite (200-pL volumes) was chromatographed through a Superose HR-12 10/30 column equilibrated in buffer A plus 1 M NaCl and eluted with the same buffer at a flow rate of 0.3 mL min-’. Fractions containing SVL were pooled, concentrated, and stored at -80°C. When necessary, SVL samples were re- chromatographed to remove remaining traces of VSP.

Purification of VSP

The 75% ammonium sulfate fraction was redissolved in buffer B (100 mM sodium acetate, pH 6.0, 500 II~M NaC1, 1 mM CaCI,, 1 mM MgCl,, 1 mM MnC1,) and dialyzed overnight against 2 X 500 mL of the same buffer. The dialyzed sample was applied to a column containing Con-A Sepha- rose 48 (25-mL bed volume) pre-equilibrated in buffer B. Nonbinding protein was eluted by washing the column with approximately 5 column volumes of buffer B, and bound protein was subsequently eluted by washing the column with buffer B plus 500 mM a-methyl-D-mannopyranoside. Fractions containing bound protein were pooled, concentrated, and dialyzed against 2 X 500 mL of buffer A. The dialysate was applied to a Mono-Q HR 5/5 column pre-equilibrated in buffer A, and the column was washed with buffer A until the A,,, approached zero. EIound protein was eluted with a linear salt gradient (15 mM NaCl mL-’, 0.5 mL min-’). Fractions containing the VSP peak eluting at approximately 100 mM NaCl were pooled, concentrated, and stored at -80°C.

Gel-Permeation Chromatography of Leaf Homogenate

Centrifuged leaf homogenate (100 pL) was applied to a Superose 12-HR 10/30 column pre-equilibrated in buffer A plus 50 mM NaCl and eluted with the same buffer at a flow rate of 0.3 mL/min. The column fractions were subjected to SDS-PAGE as described below. The column was calibrated using the following protein standards: ferritin (450 kD), catalase (240 kD), aldolase (158 kD), BSA (68 kD), ovalbumin (45 kD), chymotrypsino- gen-A (25 kD), and Cyt c (12.5 kD). The void volume was determined using blue dextran.

Antibody Preparation

Rabbit antibodies to SVL and VSP were prepared com- mercially (Lampire Biologicals, Inc., Pipersville, PA). Dial-



Vegetative Lectin of Sovbean Plants 827

ysis precipitate (100 pg) in Freund's incomplete adjuvant was injected into New Zealand White rabbits at O, 1,2, and 4 weeks, and antiserum was collected 3 weeks after the final injection.

Antibodies specific to the peptide portion of SVL or VSp were affinity purified as follows. Dialysis precipitate (10 pg) was subjected to SDS-PAGE and electroblotted to a nitrocellulose membrane as described below. The regions of the nitrocellulose membrane containing VSP or SVL peptides were excised and treated to remove carbohydrate epitopes (Woodward et al., 1985). The membrane strip wa's incubated for 24 h in 20 mL of a solution containing 50 mM sodium acetate and 20 mM p-eriodic acid, pH 4.5. It was then washed three times with 20 mL of 50 mM sodium acetate (pH 4.5) and incubated for 30 min in 20 mL of PBS containing 50 mM sodium borohydride. The membrane strip was washed three times with 20 mL of PBS, blocked in PBS plus 1% BSA overnight, and stored at -20°C. The membrane strip was used to affinity purify antibody as previously described (Smith and Fisher, 1984).

SDS-PAGE and lmmunoblot Analysis

Protein samples were subjected to SDS-PAGE (Laemmli, 1970) using acrylamide concentrations of 11.5% in the re- solving gel and 4.5% in the stacking gel. Protein molecular mass standards used in SDS-PAGE were BSA (66 kD), ovalbumin (45 kD), glyceraldehyde-3-P dehydrogenase (36 kD), trypsinogen (24 kD), trypsin inhibitor (20.1 kD), and a-lactalbumin (14.2 kD). Proteins were electroblotted from gels to nitrocellulose according to the method of Towbin et al. (1979) using a TE-50 electroblotting apparatus (Hoefer Scientific Instruments, San Francisco, CA) for 1 h. Blotted membranes were either stained with amido black or ana- lyzed as immunoblots (Burnette, 1981) by probing with anti-VSP or anti-SVL antibodies. Locations of antigenic proteins were visualized by incubating the membranes with goat anti-rabbit IgG-peroxidase conjugate and staining with diaminobenzidine reagent.

Amino Acid Sequence Analysis

N-terminal amino acid of SVL subunits was determined by automated Edman degradation. Sample preparation, electrophoresis, and electroblotting to a polyvinylidene di- fluoride membrane were performed according to the methods of Tranbarger et al. (1991). Microsequencing was performed on an Applied Biosystems 470 protein sequencer by the Laboratory for Biotechnology and Bioanalysis, Bio- chemistry, and Biophysics Department (Washington State University, Pullman).

lmmunocytology

Immunohistochemistry was performed as previously described by Tranbarger et al. (1991) with the following mod- ifications. The tissue was incubated overnight with gentle shaking in fixative containing 1.25% (v/v) glutaraldehyde, 2% (v/v) formaldehyde, 25 mM Hepes, pH 7.3, and 2 mM CaCI, and then washed three times with 25 mM Hepes, pH

7.3. Tissue was embedded in LR White resin and sections (1.5 pm) were cut using a 2065 microtome (Richert Jung, Heidelberg, Germany). After mounting on microscope slides, the sections were probed with either anti-VSP antibody (diluted 1:100) or anti-SVL antibody (diluted 1:10) for 1 or 16 h, respectively, and then with colloidal gold-conjugated goat anti-rabbit IgG (Auroprobe-LM GAR, Amer- sham). Sections were silver enhanced using lntense-M (Amersham) and counterstained with safranin.

Methyl Jasmonate Treatment

Seedlings were exposed to gaseous methyl jasmonate essentially as previously described (Franceschi and Grimes, 1991). Seedlings at 10 d postgermination were exposed to atmospheric methyl jasmonate (Bedoukian Re- search, Danbury, CT) in sealed 10-gallon glass aquaria. After seedlings were exposed for 5 d, they were harvested, homogenized, and subjected to SDS-PAGE, and SVL was determined on immunoblots.

Preparation of SVL-HSP Conjugate and Dot-Blot Analysis

SVL-HSP was prepared by incubating 0.5 mg of purified SVL with 1.25 mg of HRP (Sigma catalog No. P-6782) for 15 h at 4°C in a solution containing 10 mM NaH,PO,, 500 mM NaCI, and 0.2% glutaraldehyde. The conjugate was dialyzed against 50 mM Tris-HCI plus 500 mM NaCl and stored at -80°C. As a control, fetuin-HRP was prepared using the same procedure.

The conjugates were used to probe protein samples (1.0 pg) dot blotted to nitrocellulose membranes. Dot blots were either allowed to first air dry or immediately blocked in 50 mM NaH,PO, plus 1% BSA, pH 6.0, for 1 h. The membranes were incubated for 3 h with conjugates diluted 1:500 in 50 mM NaH2P0, plus 0.1% BSA, pH 6.0. In inhibition experiments, carbohydrates or glycoproteins were included as indicated in "Results." After the incubation period, the membranes were then washed five times with 50 mM NaH,PO, containing 0.1% BSA and 0.05% Tween 20, pH 6.0, and stained with diaminobenzidine reagent.

RESULTS

N-Terminal Amino Acid Sequence of SVL

Structural similarities had suggested that SVL was a member of the legume lectin family (Spilatro and Ander- son, 1989). We sought further evidence for this relationship by determining the N-terminal amino acid sequence of SVL. The sequences for the 21 N-terminal amino acids of the 29- and 33-kD subunits were the same, and a comparison of this sequence with those of other lectins is shown in Figure 1. Highest sequence identity occurred with SBA (63%) and the pea seed lectin (48%). Lesser homologies occurred between SVL and other legume lectins and gen- erally involved residues highly conserved among most leguminous lectins: Ser5, Phe6, Phe', Phe", Leu2', and GlnZ1. As previously noted for other lectins (Strosberg et al., 1986), greatest homology is found when the N terminus




LECTINSOURCE

SVL

SEEDSSBAPEACON-A ,DBPHASJA

(DT

AI

a> 'A

A

D T

E 1I IN AN I

E I

V S F T F

i iT 1L HQ i

iL I

S WL IM iS iS iS I

N K

i iT ii QK NQ RP i

(H)F N

I VI SI S1 11 1i A

P V Q

1 K Ii D IK D iS - -- - ES N i

P N

I IQ iK Di ST 1E D

*IDENTITY

(21)I H L Q

M I iL I FL IF IL IL L

VEGETATIVEDB58BLECSJ-LLI

AIA

i IE AE I

Q i i S iLi N iL i i S i

K NP iP i

i iI TI V

S - -i G NS N I

S ST AE D

F II TL V L

624838383838

383835

Figure 1. N-terminal amino acid sequence of SVL and other legu-minous lectins. -, A deletion introduced to maximize homology; |,amino acid identity with SVL. CON-A is aligned from residue 123.SBA, Soybean; PEA and BLEC (Ho Pak et al., 1992), Pisum sativum;DB, D. biflorus; CON-A, Canavalia ensiformis; PHA, P. vulgaris; SJA,S. japonica; DB and DB58, D. biflorus; SJ-LLI (Hankins et al., 1988),S. Japonica. Where not cited, references for sequences can be foundin Sharon and Lis (1990).

of SVL is aligned with position 123 of Con-A. Homologiesbetween SVL and other vegetative lectins were similar tothose observed between SVL and most seed lectins.

Specificity of Affinity-Purified Antibodies

Our preliminary experiments confirmed that preimmuneserum did not react with any soybean proteins but that thecrude antiserum reacted extensively with many leaf glyco-proteins (data not shown). Nonspecific reaction with gly-coproteins was eliminated by affinity purifying antibodiesagainst the peptide portions of either VSP or SVL, and thereaction of these antibodies is shown in Figure 2. Stainingwith amido black for total protein shows the two SVLpeptides and 27-kD VSP in the dialysis precipitate (lane 3).

Anti-SVL antibodies reacted with SVL in the dialysisprecipitate and homogenate from depodded leaves (Fig. 2,lanes 5 and 6). We also saw a weak reaction with two minorbands that may be proteolytic products of SVL. We rou-tinely used antibody that was affinity purified against bothSVL subunits; however, if only the 33-kD SVL subunit wasused for affinity purification, the antibodies still reactedwith both SVL subunits (data not shown), indicating a closestructural relationship between these subunits. Antigenicsimilarity between SVL and SBA had been previouslynoted (Spilatro and Anderson, 1989), and we see furtherevidence for this relationship in the reaction of anti-SVLantibody with SBA (lane 7).

Anti-VSP antibodies identified the 27-kD VSP peptide inthe dialysis precipitate (lane 1) and both VSP peptides inhomogenate from depodded leaves (lane 2). The anti-VSPantibody tended to react more strongly with the smallerVSP peptide, although in many samples this peptide oc-curred in greater abundance.

Distribution of SVL in Soybean Plants

Since the presence of SVL in plant organs other thanleaves has not been previously examined, we compared

29-Kd VSP-'27-Kd VSP-

Kd-66.0

-14.2

1 5 6 7Figure 2. Reaction of affinity-purified anti-VSP and anti-SVL antibod-ies. Protein samples were subjected to SDS-PAGE and electroblottedto nitrocellulose membranes, which were stained with amido blackfor total protein (lanes 3 and 4) or probed with either anti-SVLantibody (lanes 5, 6, and 7) or anti-VSP antibody (lanes 1 and 2).Lanes 1, 3, and 5 contain dialysis precipitate (10 /xg); lanes 2, 4, and6 contain leaf homogenate from depodded plants (100 jig of protein).Lane 7 contains SBA (10 jxg).

distributions of SVL and VSP. Tissue homogenates weresubjected to SDS-PAGE, electroblotted to a nitrocellulosemembrane, and probed with either anti-SVL (Fig. 3, top) oranti-VSP (Fig. 3, bottom) antibodies. The relative levels ofSVL approximately paralleled those of VSP in the differenttissues, although VSP was consistently present at muchhigher levels than SVL. In depodded plants, VSP and SVLwere present in stems and petioles and at lower levels inroots. Both proteins were present in relatively low levels incotyledons and seed pods, but neither SVL nor VSP wasfound in immature or mature seeds. A weak reaction ofanti-SVL antibody with SBA was observed in seed ho-mogenates.

-24.0

-36.029-Kd VSP -27-Kd

-24.0

oo —o o

co•o

£•oo

T3 TJ0 0 )o- a>

I•o<D

Figure 3. Distributions of SVL and VSP in different soybean tissues.Stem, root, and petiole samples were obtained from depoddedplants. Homogenates (100 ^g of protein) were subjected to SDS-PAGE and immunoblot analysis. The presence of SVL and VSP wasdetermined by probing nitrocellulose membranes with either anti-SVL antibody (top) or anti-VSP antibody (bottom). Imm., Immature;Mat., mature.



Vegetative Lectin of Soybean Plants 829

Immunohistochemical Localization

Previously, VSP was shown to accumulate principally inthe bundle-sheath and paraveinal mesophyll cells of soy-bean leaves (Franceschi et al., 1983). Thus, it was of interestto examine the cellular location of SVL in leaves. Sectionsof leaf tissue from plants at anthesis were embedded in LRWhite resin and probed with anti-SVL or anti-VSP antibod-ies. The location of bound antibodies was visualized withcolloidal gold-conjugated secondary antibody and silverenhancement. No staining occurred on leaf sections probedwith preimmune serum (Fig. 4A). For leaf sections probedwith anti-SVL (Fig. 4B) or anti-VSP (Fig. 4C), silver stainingoccurred extensively in the vacuole of bundle-sheath andparaveinal mesophyll cells. Neither VSP nor SVL was de-tected in the palisade mesophyll or spongy mesophyll.Both antibodies reacted sporadically with cells of the epi-dermis, an observation previously reported for VSP(Huang et al., 1991).

Effects of Sink Deprivation

Several types of plant treatments that disrupt the normaltransport of photoassimilates to sink tissues have beenshown to cause an accumulation of VSP in leaves (Witten-bach, 1983). We compared the responses of SVL (Fig. 5, top)and VSP (Fig. 5, bottom) in leaves and cotyledons to theremoval of sink tissues. Both proteins were present inleaves at anthesis. The amount of SVL was lower in plantsallowed to set seed normally and higher in plants subjectedto pod removal. This response pattern was very similar tothat observed for VSP; the level of VSP in leaves washighest in depodded plants and lowest in plants allowed toset seed normally, as previous observed (Wittenbach,1983).

Figure 5 also shows the effects of sink deprivation onlevels of SVL and VSP in cotyledons. VSP and SVL werepresent in the cotyledons of seedings allowed to developnormally for 10 d postemergence. In seedlings, the devel-oping shoot and leaflets represent major sink tissues fornutrient reserves stored in the cotyledons. When the de-veloping shoot and axial buds of seedlings were removedshortly after seed germination, the levels of VSP and SVL inthe cotyledons were higher than in controls after the samegrowth period.

Methyl Jasmonate Induction of SVL

Previously it was shown that the plant growth regulatormethyl jasmonate causes induction of VSP gene expressionand accumulation of VSP proteins (Anderson et al., 1989;Franceschi and Grimes, 1991; Staswick et al., 1991). Figure6 shows extracts of leaves, cotyledons, and stems of seed-lings exposed to different concentrations of atmosphericmethyl jasmonate. SVL levels were elevated in seedlingsexposed to methyl jasmonate as low as 4 ju,L L"1, althoughin stems a slightly stronger response occurred at 8 /nL L"1.The amount of SVL began to diminish in plants exposed to12 ju,L L"1 methyl jasmonate, which may be due to the

Figure 4. Immunohistochemical localization of SVL and VSP in soy-bean leaf. LR White-embedded leaf tissue was sectioned at a thick-ness of 2 /urn and probed with preimmune serum (A), anti-SVLantibody (B), or anti-VSP antibody (C). Bound antibodies were lo-cated by probing with gold-conjugated secondary antibody and sil-ver enhancement. Cell types are bundle sheath (B), spongy meso-phyll (SM), palisade mesophyll (PM), paraveinal mesophyll (P), lowerepidermis (LE), and upper epidermis (UE). Arrows indicate silverstaining in bundle-sheath and paraveinal mesophyll cells.Bars = 30 p.m.




33-Kd SVL-28-Kd SVL-

.

29-Kd VSP -27-Kd VSP^

Kd'-36.0

-24.0

S -36.0

-24.0

Leaf Cotyl

Figure 5. Effects of sink deprivation (Sink Depr) on accumulation ofSVL and VSP. SVL and VSP were detected by subjecting tissuehomogenates (100 jug of protein) to SDS-PAGE and immunoblotanalysis. The presence of SVL and VSP was determined by probingnitrocellulose membranes with either anti-SVL antibody (top) oranti-VSP antibody (bottom). Cotyl, Cotyledon.

onset of senescence observed for these seedlings at the timeof harvest.

Gel-Permeation Chromatography of Leaf Homogenates

As previously reported (Spilatro and Anderson, 1989),SVL and VSP occur in a precipitate that forms duringdialysis of ammonium-sulfate-fractionated leaf homoge-nate. We examined whether high-molecular-weight aggre-gates of SVL or VSP were present in homogenates prior toammonium sulfate treatment. Crude leaf homogenateswere subjected to gel-permeation chromatography, and thecomposition of the column fractions was determined bySDS-PAGE and immunoblot analysis. The proteins elutedfrom the column as a series of broad peaks (Fig. 7A). As

METHYL JASMONATE (|»UL)

0 4 8 12

LEAF

STEM

COTYL.

Figure 6. Methyl jasmonate induction of SVL in seedlings tissues.Seedlings were exposed to atmospheric methyl jasmonate at theindicated concentrations for 5 d in a sealed aquarium. SVL wasdetected by subjecting tissue homogenates (100 ng of protein) toSDS-PAGE and immunoblot analysis using anti-SVL antibodies.COTYL., Cotyledon.

native proteins, VSP and SVL would elute at volumescorresponding to molecular masses of 60 and 120 kD, re-spectively. However, SVL was found only in the voidvolume fractions (Fig. 7B, lanes 1 and 2). The void volumealso contained the 27-kD VSP (lanes 1 and 3) and peptidesthat correlated in molecular mass with the large (55 kD)and small (14 kD) subunits of Rubisco. Since the largestmolecular mass resolved on a Superose 12-HR resin isapproximately 300 kD, proteins in the void volume havemolecular masses equal to or greater than 300 kD.

The results presented above suggest that SVL and VSPexist in higher-molecular-weight aggregates before treat-ment of the proteins with ammonium sulfate but do notresolve whether they form homogenous or heterogenousaggregates. We found that purified SVL rapidly precipi-tates out of solution in the absence of VSP, apparentlythrough self-association. However, we did not observehigh-molecular-weight aggregates of VSP when purifiedVSP was subjected to gel-permeation chromatography(data not shown). Furthermore, we did not find that pre-cipitates formed in solutions of purified VSP at concentra-

0.3

0.2

QO

0.1

0.0

10'

8 10 12 14 16Elution volume, ml

18

• 27-Kd VSP

-20.1

-14.2

1 2 3

Figure 7. Gel-permeation chromatography of leaf homogenate. Leafhomogenate was chromatographed through a Superose 12-HR 10/30column and fractions were collected and examined for the presenceof SVL and VSP. A, Elution profile (continuous line) of proteinmeasured at 280 nm. The calibration curve (dashed line) was deter-mined with molecular mass standards listed in "Materials and Meth-ods." B, Protein composition of fractions eluting at the void volume.The fractions from the void volume were pooled, and a sample (20jitg of protein) was subjected to SDS-PAGE and immunoblot analysis.The electroblotted samples were stained for total protein with amidoblack (lane 1) or probed for antigenic proteins with either anti-SVLantibody (lane 2) or anti-VSP antibody (lane 3).




tions as high as 20 mg/mL. These observations suggest thatoccurrence of VSP in the dialysis precipitate is dependenton the presence of SVL.

Reaction of SVL-HSP Conjugate

To test for direct binding between SVL and VSP, SVL-HRP was prepared and used to probe VSP and otherproteins dot blotted on nitrocellulose. As shown in Figure8A, SVL-HRP reacted with both VSP and SVL. When dotblots were not allowed to dry during the probing proce-dure, the conjugate reacted more strongly with SVL thanVSP. Binding to VSP was greatly enhanced, however, ifthe protein blots were allowed to air dry prior to theblocking step.

A number of controls were run to examine the specificityof the SVL-HRP reactions. The absence of backgroundstaining on the immunoblots showed that SVL-HRP didnot react with nondried BSA, a nonglycosylated proteinused as the blocking agent (Figs. 8 and 9). SVL-HRP alsodid not react with the glycoproteins fetuin and ovalbuminon dried or nondried blots, and unconjugated HRP did notreact with SVL or VSP (data not shown). Fetuin-HRP didnot react detectably with nondried VSP or SVL, although aslight reaction with dried VSP blots was observed (Fig. 8A).

Drying of the blots probably caused protein denatur-ation, possibly exposing hydrophobic sites to which theconjugate bound. The role of hydrophobic sites was exam-ined by probing heat-denatured BSA, a protein that doesnot possess bound carbohydrates. As noted above, SVL-HRP did not bind to untreated BSA used as a blockingagent. However, when BSA was heat treated and thenblotted, binding of SVL-HRP was observed (Fig. 8B).

Inhibition of the reaction of SVL-HRP with VSP and SVLwas tested using a range of carbohydrates and glycopro-teins. Figure 9A shows that binding was not affected by thepresence of Glc, Man, a-methyl-D-mannopyranoside, Gal,N-acetylgalactose-amine, N-acetylglucose-amine, or Fuc at10 mM. These carbohydrates had no effect whether or notVSP blots were dried. Other carbohydrates tested andfound to have no effect at 10 mM include sialic acid, glu-

BLOTS BLOTSDRIED NOT DRIED

SVL-HRP • 0

FETUIN-HRP

BLOTS BLOTSDRIED NOT DRIED

B DENATUREDBSA

SVL-HRP

Figure 8. Reaction of SVL-HRP with proteins dot blotted to nitrocel-lulose. After blotting, protein (1 jug) samples were either allowed tofirst air dry and blocked or immediately blocked in 1% BSA. Mem-branes were probed with SVL-HRP or fetuin-HRP (A) and boundconjugate was detected using diaminobenzidine reagent. B, Reactionof SVL-HRP with heat-denatured BSA.

B

NO INHIBITOR

GLUCOSE

MANNOSE

GALACTOSE

N-AC-GAL-AM

N-AC-GLC-AM

METHYL-MAN

FUCOSE

mg /ml MUCIN HEPARINVSP SVL VSP SVL

0 • • • •

o.ooi • • • ®0.01 • • ^ •

0.1 - @1.0

Figure 9. Inhibition of binding of SVL-HRP to protein blots. Protein(1 |u,g) dot blots were incubated with SVL-HRP in the presence of theindicated carbohydrates at 10 mM (A) or in the presence of porcinestomach mucin or heparin at the indicated concentrations (B). N-AC-GAL-AM, N-Acetylgalactose-amine; N-AC-GLC-AM, N-acetylglu-cose-amine; METHYL-MAN, a-methyl-D-mannopyranoside.

curonic acid, lactose, N-acetyllactose-amine, chitobiose,chitotriose, gentiobiose, melibiose, stachyose, and raffinose(data not shown).

Binding of SVL-HRP to protein dot blots was inhibitedwhen porcine stomach mucin (Sigma catalog No. M-1778)or heparin (Sigma catalog No. H-2149) was included in thereaction mixture (Fig. 9B). In the experiment shown, pro-tein blots were allowed to dry before blocking, althoughsimilar results were obtained when blots were not allowedto dry. Partial inhibition of binding to VSP was detected atconcentrations of mucin and heparin as low as 1 jiig/mL.Higher concentrations of these substances were required toinhibit binding of SVL-HRP to SVL, possibly reflecting astronger affinity of SVL for itself than for the inhibitors.The effect of N-desulfated heparin (Sigma catalog No.D-4776) was similar to that of heparin.

A number of other proteins were tested for their abilityto inhibit binding of SVL-HRP to blots. Binding was notinhibited when BSA, ovalbumin, fetuin, or bovine submax-illary gland mucin were present at 1 mg/mL (data notshown).

DISCUSSION

A comparison of the N-terminal sequences of SVL andSBA indicates that these proteins are products of distinct,yet closely related, genes. Along with previously reportedstructural and immunological relationships (Spilatro and




Anderson, 1989), the sequence data strongly suggest that SVL is a member of the legume family of plant lectins. Sequence similarity varies widely between seed and vegetative lectins within legume species; for example, the seed and vegetative lectins of Dolickos biflorus and Sopkora japonica are nearly identical in primary structure (Hankins et al., 1988; Schnell and Etzler, 1988; Harada et al., 1990), whereas the pea lectins are only 38% identical (Ho Pak et al., 1992). No significant sequence homology with lectins from nonleguminous plants was observed, suggesting that SVL is not related to lectins in other plant families. Identity in N-terminal sequence between SVL subunits is similar to that reported for Dolickos lectins, which differ only at the C terminus (Etzler, 1994b).

The results of this investigation show that SVL occurs widely throughout the plant. Wide distribution among plant organs has been observed for vegetative lectins in certain other plants, such as D. biflorus and Pkaseolus vulgaris (Borrebaeck, 1984; Etzler, 1986; Pusztai, 1991). Lectins with diverse properties have been isolated from various organs of soybean plants (Pusztai, 1991), although SVL is more widely distributed than other soybean lectins. A vegetative protein of soybean plants that shares some properties with SVL and forms a complex with Rubisco has re- cently been described (Crafts-Brandner and Salvucci, 1994). This protein is widely distributed in the vegetative tissues, occurs as a tetramer of 30-kD subunits, and increases in amount when seed pods are removed. However, unlike SVL, the Rubsico-binding protein reportedly occurs in the cytosol (Crafts-Brandner and Salvucci, 1994).

Many properties of SVL suggest that this protein is a lectin; however, neither specific carbohydrate-binding nor hemagglutination has been demonstrated. Results reported in this paper could explain why hemagglutination activity has not been observed for soluble extracts of soybean vegetative tissues or purified SVL (Pueppke et al., 1978; Bowles and Marcus, 1981; Spilatro and Anderson, 1989). For example, since SVL appears to exist in a high-molecular-mass aggregate in crude homogenates, hemagglutination activity of SVL may be masked through binding to VSP or self-association. Alternatively, SVL may be mono- valent and unable to cross-link blood cells.

The ability of heparin and porcine mucin, but not simple carbohydrates, to inhibit binding of SVL-HRP to VSP and SVL suggests that SVL binds to an oligosaccharide. Many legume lectins display a higher affinity for oligosaccharides than for monosaccharides, and some seed and vegetative lectins are only known to bind to oligosaccharides (Goldstein and Poretz, 1986; Sharon and Lis, 1990). Lectins, such as those of P. vulgaris, that display a high affinity for a complex oligosaccharide structure are not readily inhibited by simple carbohydrates (Goldstein and Poretz, 1986). The vegetative lectin of D. biforus, DB58, binds to the blood group A + H substance, but not to simple carbohydrates, and does not agglutinate erythrocytes (Etzler, 1994a). SVL may not bind to the oligosaccharides exposed on the erythrocytes used in agglutination tests. Lectins from some plants agglutinate only specific types of erythrocytes (Goldstein and Poretz, 1986). This includes a lectin from

tulip bulbs, which agglutinates only mouse and rat erythrocytes and cannot be inhibited by simple sugars (Oda et al., 1987) and many legume lectins (Goldstein and Poretz, 1986).

Enhanced binding of SVL-HRP to dried protein blots suggests that binding of SVL to proteins involves hydrophobic interactions. Other lectins also appear to possess hydrophobic sites, as evidenced by stronger binding to hydrophobic glycosides than simple carbohydrates (Goldstein and Poretz, 1986; Strosberg et al., 1986; Sha- ron and Lis, 1990). These hydrophobic sites are likely to be important to the function of lectins (Goldstein and Poretz, 1986); however, little is known about the role of hydrophobic sites in the binding of plant lectins to glycoprotein receptors or the conditions that inhibit these interactions. Involvement of hydrophobic sites in the binding of SVL-HRP to VSP could explain why simple carbohydrates and oligosaccharides alone are not inhib- itory. Properties unique to each of the VSP peptides might explain why only the 27-kD VSP occurs in the dialysis precipitate. VSP peptides differ in pI (Mason et al., 1988), enzymatic activity (DeWald et al., 1992), and possibly in bound carbohydrates or exposed hydrophobic sites. Binding characteristics of lectins are affected by many other factors, including the presence of ions, ionic strength, and pH (Etzler, 1986; Goldstein and Poretz, 1986), and such factors would be important to the physicochemical properties of SVL in vivo.

Severa1 lines of evidence suggest that SVL and VSP may interact in vivo. Both proteins are widely distributed among the vegetative organs of the plant and accumulate in the vacuolar compartment. SVL responds to signals that are similar to those that regulate VSP. The response of VSP to sink deprivation is believed to reflect its role as a storage site for nitrogen and photosynthates (Wittenbach, 1983; Staswick, 1990). Levels of SVL are much lower than those of VSP under a11 conditions investigated, arguing against a similar function for SVL. SVL is the first lectin shown to be regulated by methyl jasmonate, which has previously been shown to regulate expression of VSP genes (Franceschi and Grimes, 1991; Mason et al., 1992). Methyl jasrnonate has been implicated in plant responses to physiological stresses such as wounding, herbivory, and water deficit (Staswick, 1992), suggesting that SVL may serve a function relating to these responses.

However, immunohistochemical data that show that SVL occurs principally in the bundle-sheath and paraveinal mesophyll cells of leaves argue that SVL relates to the specialized function of these cells. The paraveinal mesophyll and associated bundle sheath make up a specialized tissue believed to play an important role in the storage and transfer of photoassimilates (Franceschi and Giaquinta, 1983). It has been suggested that lectins may function in the packaging or remobilization of storage proteinis, and seed lectins have been shown to form complexes with some seed storage proteins (Clarke et al., 1975; Bowles and Marcus, 1981; Einhoff et al., 1986). Conceivably, SVL could contrib- ute to the packaging of VSP in vacuoles. It will be necessary to further characterize the binding properties of SVL and




covalently bound carbohydrates of VSP to evaluate this possibility .

ACKNOWLEDCMENTS

The research presented in this paper was performed as a series of undergraduate research projects. We would like to thank John Sweeney, RoAnn Stanley, and Jeff DeLong who also participated in this project. Much gratitude is also extended to Vincent France- schi for help with the immunocytology and microsequencing experiments and to Marilyn Etzler for continuous encouragement and helpful suggestions.

Received September 6, 1995; accepted November 28, 1995. Copyright Clearance Center: 0032-0889/96/ 110/0825/ 10.

LITERATURE CITED

Anderson JM, Spilatro SR, Klauer SF, Franceschi VR (1989) Jasmonic-acid dependent increases in the leve1 of vegetative storage protein in soybean suspension cultures and seedlings. Plant Sci 62 45-52

Borrebaeck CAK (1984) Detection and characterization of a lectin from non-seed tissue of Phaseolus vulgaris. Planta 161: 223-228

Bowles DJ, Marcus S (1981) Characterization of receptors for the endogenous lectins of soybean and jackbean seeds. FEBS Lett

Bumette WN (1981) ”Western blotting”: electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Ana1 Biochem 112 195-203

Clarke AE, Knox RB, Jermyn MA (1975) Localization of lectins in legume cotyledons. J Cell Sci 19: 157-167

Crafts-Brandner SJ, Salvucci ME (1994) The Rubisco complex protein: a protein induced by fruit remova1 that forms a complex with ribulose-1,5-bisphosphate carboxylase/oxygenase. Planta 194 110-116

Dewald DB, Mason HS, Mullet JE (1992) The soybean vegetative storage proteins VSP-(Y and VSP-p are acid phosphatases active on polyphosphates. J Biol Chem 267: 15958-15964

Einhoff W, Fleischmann G, Freier T, Kummer H, Rudiger H (1986) Interactions of leguminous seed lectins with seed proteins-lectins as packing aids of storage proteins. In TC B0g- Hanson, E van Driessche, eds, Lectins: Biology, Biochemistry, Clinical Biochemistry, Vol5. Walter de Gruyter, Berlin, pp 45-52

Etzler ME (1986) Distribution and function of plant lectins. In IE Liener, N Sharon, eds, The Lectins: Properties, Functions, and Applications in Biology and Medicine. Academic Press, New York, pp 371-435

Etzler ME (1994a) A comparison of the carbohydrate binding properties of two Dolichos biflorus lectins. Glycoconjugate J 11: 395-399

Etzler ME (1994b) Isolation and characterization of subunits of DB58, a lectin from the stems and leaves of Dolichos biforus. Biochemistry 33: 9778-9783

Franceschi VR, Giaquinta RT (1983) The paraveinal mesophyll of soybean leaves in relation to assimilate transfer and compartmentalization. 11. Structural, metabolic, and compartmental changes during reproductive growth. Planta 157: 422-431

Franceschi VR, Grimes HD (1991) Induction of soybean vegetative storage proteins and anthocyanins by low-leve1 atmospheric methyl jasmonate. Proc Natl Acad Sci USA 88: 6745-6749

Franceschi VR, Wittenbach VA, Giaquinta RT (1983) The paraveinal mesophyll of soybean leaves in relation to assimilate transfer and compartmentalization. 111. Immunohistochemical localization of specific glycopeptides in the vacuole after depod- ding. Plant Physiol 72 586-589

Goldstein IJ, Poretz RD (1986) Isolation, physicochemical characterization, and carbohydrate-binding specificity of lectins. In IE Liener, N Sharon, IJ Goldstein, eds, The Lectins: Properties,

129: 135-138

Functions, and Applications in Biology and Medicine. Academic Press, New York, pp 33-245

Hankins CN, Kindinger JI, Shannon LM (1988) The lectins of Sophora japonica. 11. Purification, properties and N-terminal amino acid sequences of 5 lectins from bark. Plant Physiol 86:

Harada JJ, Spadora-Tank J, Maxwell JC, Schnell DJ, Etzler ME (1990) Two lectin genes differentially expressed in Dolichos bi- Jorus differ primarily by a 116 bp sequence in their 5’ flanking regions. J Biol Chem 265: 4997-5001

Ho Pak J, Hendrickson T, Dobres MS (1992) Predicted sequence and structure of a vegetative lectin in Pisum sativum. Plant Mo1 Biol 18: 857-863

Huang J-F, Bantroch DJ, Greenwood JS, Staswick PE (1991) Methyl jasmonate treatment eliminates cell-specific expression of vegetative storage protein gene in soybean leaves. Plant Physiol97: 1512-1520

Kijne JW, Díaz CL, Bakhuizen R (1986) The physiological function of lectins. In TC Bsg-Hanson, E van Driessche, eds, Lectins: Biology, Biochemistry, Clinical Biochemistry, Vol 5. Walter de Gruyter, Berlin, pp 45-52

Laemmli UK (1970) Cleavage of structural proteins during assem- bly of the head of bacteriophage T4. Nature 227: 680-685

Mason HS, DeWald DB, Creelman RA, Mullet JE (1992) Coregu- lation of soybean vegetative storage protein gene expression by methyl jasmonate and soluble sugars. Plant Physiol98: 850-867

Mason HS, Guerrero FD, Creelman RA, Mullet JE (1988) Proteins homologous to leaf glycoproteins are abundant in stems of dark grown soybean seedlings. Analysis of proteins and cDNAs. Plant Mo1 Biol 11: 845-856

Oda Y, Minami K, Ichida S, Aonuma S (1987) A new agglutinin from Tulipa gesneriana bulbs. Eur J Biochem 165 297-302

Pueppke SG, Bauer WD, Keegstra K, Ferguson AL (1978) Role of lectins in plant-microorganism interactions. 11. Distribution of soybean lectin in tissues of Glycine max (L.) Merr. Plant Physiol

Pusztai A (1991) Plant Lectins. Cambridge University Press, New York

Rudiger H (1984) On the physiological role of plant lectins. Bio- science 34 95-99

Schnell DJ, Etzler ME (1988) cDNA cloning, primary structure, and in vitro biosynthesis of the DB58 lectin from Dolichos biflorus. J Biol Chem 263 14648-14653

Sharon N, Lis H (1990) Legume lectins-a large family of homologous proteins. FASEB J 4: 3198-3208

Smith DE, Fisher PA (1984) Identification, developmental regulation, and response to heat shock of two antigenetically related forms of a nuclear envelope protein in Drosophila embryos: application of an improved method for affinity purification of antibodies using polypeptides immobilized in nitrocellulose blots. J Cell Biol 99: 20-28

Spilatro SR, Anderson JM (1989) Characterization of a soybean leaf protein that is related to the seed lectin and is increased with pod removal. Plant Physiol 90 1387-1393

Staswick PE (1989) Preferential loss of an abundant storage protein from soybean pods during seed development. Plant Physiol

Staswick PE (1990) Nove1 regulation of vegetative storage protein genes. Plant Cell 2: 1-6

Staswick PE (1992) Jasmonate, genes, and fragrant signals. Plant Physiol 99: 804-807

Staswick PE, Huang J-F, Rhee Y (1991) Nitrogen and methyl jasmonate induction of soybean vegetative storage protein genes. Plant Physiol 96 130-136

Streeter JG (1978) Effect of N starvation of soybean plants at various stages of growth on seed yield and N concentration of plant parts at maturity. Agron J 70 74-76

Strosberg AD, Buffard D, Lauwereys M, Foriers A (1986) Legume lectins: a large family of homologous proteins. In IE Liener, N Sharon, IJ Goldstein, eds, The Lectins: Properties, Functions, and Applications in Biology and Medicine. Academic Press, New York, pp 249-264

67-70

61: 779-784

90: 1252-1255




Towbin H, Staehlin T, Gordon J (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedures and some applications. Proc Natl Acad Sci USA 76:

Tranbarger TJ, Franceschi VR, Hildebrand DF, Grimes HD (1991) The soybean 94-kilodalton vegetative storage protein is a lipoxygenase that is localized in paraveinal mesophyll cell vacuoles. Plant Cell 3: 973-987

4350-4354

Van Damme EJM, Peumans WJ (1990) Developmentajl changes in the tissue distribution of lectin in Galantkus nivalis L. and Nar- cissus cv. Calton. Planta 182 605-609

Wittenbach VA (1983) Purification and characterization of a soybean leaf storage glycoprotein. Plant Physiol 73: 125-129

Woodward MP, Young WW, Bloodgood RA (1985) Detection of monoclonal antibodies specific for carbohydrate epitopes using periodate oxidation. J Immunol Methods 7 8 143-153



Documents

Characterization of a New Lectin of Soybean Vegetat ive ... · column with buffer B plus 500 mM a-methyl-D-mannopy- ranoside. Fractions containing bound protein were pooled, concentrated,