Structural Requirements for Ligand Binding by a Probable Plant Vacuolar Sorting Receptor

Ligand Binding Assay with Fluorescent Proaleurain Peptide

The Superdex 200 column was calibrated with markers of different molecular weights, and the elution positions of those proteins were compared with that of tBP-80 (Figure 2A). BSA, with a molecular weight of ∼66,000, eluted with a peak positioned between fractions 31 and 32. tBP-80, as detected by protein gel blot analysis (Figure 2A, bottom), eluted with a peak at fraction 31. This result, a size of ⩽66 kD for tBP-80, is consistent with that shown in Figure 1B and indicates that tBP-80 chromatographed as a monomer in the Superdex 200 system. If tBP-80 were a dimer under these conditions, it would have eluted at a position close to the peak obtained for alcohol dehydrogenase (150 kD) at fraction 27 (data not shown). The sensitivity of the assay system would have readily permitted detection of a shift of three or four fractions in its elution position.

The BODIPY-FL–labeled peptide was then used to characterize ligand binding by tBP-80; the results are presented in Figure 2B. We used medium from Drosophila S2 cells expressing the protein, which had been concentrated approximately fivefold for the assays. When an aliquot was incubated with 10⁻⁷ M labeled peptide and then chromatographed on the Superdex 200 column, two fluorescent peaks were detected (open circles). The one at approximately fraction 50 corresponded to free peptide, whereas the second peak, at fraction 31, corresponded to the elution position of tBP-80 (Figure 2A, bottom). Similar results were obtained when the proaleurain peptide was labeled with iodine-125, and the presence of 10⁴ M excess unlabeled peptide completely abolished the peak centered on fraction 31 (data not presented). Thus, the peak centered at fraction 31 was not dependent on any single peptide-labeling strategy. We hypothesized that it represented peptide bound by tBP-80, and used assays of sequence specificity requirements for binding and peak shifts induced by anti–BP-80 MAbs to confirm the hypothesis.

Sequence Specificity of Binding

To determine whether this binding depended on the presence of a functional vacuolar targeting motif in the peptide, we used different synthetic peptides to compete with the labeled peptide for binding (Kirsch et al., 1994). As shown in Figure 2B, when a competitor peptide, sequence SRFNPgRLPT (designated Spo-G), was included in the incubation at a concentration of 1 mM, no alteration of the fluorescent peaks was observed (Figure 2B, squares). In contrast, when a competitor peptide, sequence SRFNPIRLPT (designated Spo), was included, the peak representing bound fluorescent peptide (Figure 2B, X symbols, Bound) was diminished, whereas the peak representing free fluorescent peptide (X symbols, Free) was proportionately increased. The Spo-G peptide represents an Ile-to-Gly mutation of the central NPIR motif of the prosporamin vacuolar targeting determinant, a mutation that abolished targeting in vivo (Nakamura et al., 1993), whereas the Spo peptide represents the intact targeting determinant. The mutated motif did not compete with labeled proaleurain peptide for binding to full-length BP-80, whereas the intact motif did, albeit incompletely, in a manner similar to that observed here for tBP-80 (Kirsch et al., 1994). We conclude that sequence specificity of binding to tBP-80 is similar to that determined for full-length BP-80.

Effects of MAbs on Binding of Labeled Peptide to tBP-80

We used the anti–BP-80 MAbs to confirm that binding of the fluorescent peptide was attributable to tBP-80 and not to some other protein in the Drosophila cell culture medium. In each instance, 10 μg of purified MAb was incubated in 200 μL of concentrated medium containing tBP-80 plus 10⁻⁷ M labeled peptide. Results are shown in Figure 3, in which elution profiles obtained in the presence of different MAbs are compared with the superimposed profile of fluorescent peptide binding obtained in the absence of antibodies (open circles). MAb 14G7 chromatographed in a peak within fractions 19 to 24 and caused a shift of ∼40% of tBP-80 to the same position as detected by protein gel blot (Figure 3A, bottom); the remaining ∼60% of tBP-80 eluted at its normal position (Figure 3A, bottom, peak centered on fraction 31). At the same positions were peaks of fluorescence (Figure 3A) distributed in similar proportions, corresponding to a complex of labeled peptide plus tBP-80 plus 14G7 in fractions 20 to 24 and the labeled peptide plus tBP-80 in a peak centered on fraction 31 (Figure 3A, Bound + MAb and Bound, respectively). This result indicated that MAb14G7 bound to tBP-80 with a relatively weak affinity such that only ∼40% of the tBP-80 remained complexed to the antibody during chromatography. In contrast, MAbs 17F9 and 18E7 shifted essentially all of the fluorescent peptide binding activity to a peak centered on fraction 21 (Figure 3B), a position that corresponded to tBP-80 detected by protein gel blot analysis (Figure 3B, bottom, shown for 17F9; similar results obtained for 18E7 are not shown). Thus, 17F9 and 18E7 appeared to have higher affinities for tBP-80 such that most tBP-80 remained complexed with the antibodies during chromatography. The patterns shown in Figure 3 were consistent regardless of whether the antibodies were preincubated with tBP-80 before ligand was added or were added after the tBP-80–ligand interaction had been allowed to occur. Thus, interaction with the antibodies did not detectably affect ligand binding by tBP-80. MAb 19F2 appeared to have the least affinity for tBP-80 in a native conformation, causing neither a detectable shift in the elution position of tBP-80 nor a detectable effect on ligand binding by tBP-80 (data not shown).

Ligand Binding by Δ3EGFR Is Enhanced by Antibody Interactions

The preceding experiments served as controls against which the results obtained by utilizing Δ3EGFR could be compared. In similar experiments, we used dialyzed culture medium containing Δ3EGFR that had been concentrated approximately fourfold. The concentration of Δ3EGFR in this medium was approximately half that of tBP-80 in medium used in the preceding experiments, as determined by comparing serial dilutions of the two media on protein gel blots (data not shown).

We first determined the ability of Δ3EGFR to bind the fluorescent ligand in the presence of MAb 14G7 as a control, given that 14G7 does not interact with that protein (Figure 1B). The fluorescence profile from that assay is presented in Figure 4, which indicates the presence of a very small peak of bound peptide in fractions 31 to 35 (indicated by Δ3EGFR above the bracket); these fractions corresponded to the position of Δ3EGFR protein detected by protein gel blot (Figure 4, bottom). This result was the best we obtained; in other experiments, a peak for fluorescence from the bound peptide could not be reliably detected (not shown). These results indicated that Δ3EGFR could bind the fluorescent peptide but at a much lower affinity than that demonstrated by tBP-80. In striking contrast, however, was the ability of Δ3EGFR to bind ligand in the presence of either 17F9 or 18E7 MAbs (Figure 4); both antibodies caused a substantial increase in the amount of ligand bound by Δ3EGFR, as demonstrated by the fluorescent peaks centering on fraction 21. Consistent with previous results, these fluorescent peaks corresponded to the presence of antibody plus Δ3EGFR in those fractions (Figure 4, bottom, shown for 17F9; similar results for 18E7 are not shown) and represented a complex of antibody plus Δ3EGFR plus bound ligand. Given that approximately half as much Δ3EGFR as tBP-80 was used in these assays and that peak area for bound peptide relative to free peptide for Δ3EGFR + 18E7 or 17F9 was approximately one-third of that obtained with tBP-80, we concluded that the ligand binding affinity of Δ3EGFR was similar to or only slightly less than that of tBP-80.

The specificity of peptide binding by Δ3EGFR was tested with competition assays utilizing Spo-G (Figure 4) and Spo peptides. The presence of 1 mM Spo-G peptide had no effect on ligand binding by Δ3EGFR in the presence of 17F9 MAb, but the same concentration of Spo completely abolished the fluorescent peak centered on fraction 21. This result is substantially different from the partial competition obtained with tBP-80 (above) or with full-length BP-80 (Kirsch et al., 1994); its implications are discussed below. When 17F9 was incubated with labeled peptide in the absence of Δ3EGFR, all of the fluorescence eluted in a peak corresponding to unbound peptide (Figure 4), demonstrating that the MAb itself did not have any ligand binding activity.

Structural Domains of tBP-80 Defined by Protease Digestion

We isolated tBP-80 from Drosophila S2 cell medium, as described in Methods. An extensive purification protocol was necessary because even though tBP-80 in medium bound the proaleurain peptide ligand with high affinity (as judged from the fluorescent peptide binding assays), it interacted poorly with the proaleurain peptide–agarose affinity column used to purify intact BP-80 (Kirsch et al., 1994; see Figure 6). This indicates that the loss of attachment of BP-80 to the membrane altered the access of the ligand binding site to the peptide on a rigid matrix but not to the peptide in solution. The final tBP-80 preparation was estimated to be ∼95% pure (Figure 5A, lane 1), and its N-terminal amino acid sequence was identical to that of BP-80 (Paris et al., 1997; data not shown).

Digestion with endoproteinase Asp-N, which cuts to the N-terminal side of Asp residues, was utilized to probe tBP-80 structure; resistant fragments were identified on Coomassie blue–stained SDS–polyacrylamide gels and by probing protein gel blot transfers from those gels on membranes with different antibodies. We initially tested different conditions for digestion of the protein, as shown in Figure 5A. When a ratio of protease to tBP-80 of 1:1000 was used and the mixture was incubated at 22°C for 15 min, ∼75% was digested to yield two major products of ∼43 kD (Figure 5A, lane 2, arrow) and ∼38 kD (solid triangle), as assessed by electrophoresis in the absence of disulfide-reducing agents. The bands at ∼43 and ∼38 kD could represent collections of smaller fragments held together by multiple intramolecular disulfide bonds, or they could represent intact fragments resistant to protease.

To determine the size of protease-resistant polypeptides more stringently, we then performed SDS-PAGE on digests after disulfide reduction. Results are presented in Figure 5A, lanes 3 to 5. When the protein was digested at pH 7.5 for 1 hr at 22°C followed by 1 hr at 37°C (lane 3), resistant fragments presented bands at ∼43 and ∼38 kD that stained with approximately similar intensity. In contrast, when the protein was digested at pH 5.0 for 1 hr at 22°C (Figure 5A, lane 4) or for 1 hr at 22°C followed by 1 hr at 37°C (lane 5), the staining intensity of the ∼43-kD bands relative to that of the ∼38-kD bands appeared to be diminished at each time point. The intensity of staining for the 38-kD band in the 2-hr incubation samples for each pH was similar (Figure 5A, cf. lanes 3 and 5), whereas the intensity of staining of the ∼43-kD band in the pH 5.0 sample appeared to be less than that in the pH 7.5 sample (cf. lanes 3 and 5).

These results led to two conclusions. First, SDS-PAGE gave similar results whether or not the proteins were first treated with a disulfide-reducing agent. Thus, most protease-resistant fragments detected by electrophoresis of unreduced samples, as was necessary for mapping with the MAbs (see below), could be interpreted to represent intact polypeptide chains. Second, the decrease in pH may have affected the conformation of fragments in the ∼43-kD band such that they became more susceptible to protease digestion; however, this possibility requires confirmation by alternative experimental strategies before it can be accepted.

We then mapped the protease-resistant fragments by means of structural data presented in the diagram of tBP-80 in Figure 5B. The open rectangle at the left indicates the epitope for RA3 rabbit polyclonal antibodies raised against a 15–amino acid synthetic peptide representing the N terminus of BP-80 after cleavage of the signal peptide (Paris et al., 1997). Fragments recognized by RA3 therefore map from the N terminus of the protein. The oval to the right indicates the EGF-CB motif necessary for binding MAb 14G7; fragments recognized by 14G7 map from the C terminus of the protein. Lines terminated with open circles indicate the positions of all Asp residues outside of the EGF repeats; closed circles indicate the positions of Cys residues.

Fifteen micrograms of tBP-80 was digested at pH 7.5, protease activity was destroyed by addition of EDTA, the fragments were fractionated by incubation with 5 μg of affinity-purified RA3 antibodies, and the immune complexes were removed by treatment with protein A–Sepharose. This strategy was designed to produce two fractions, one with predominantly N-terminal fragments (selected by RA3) and one of fragments that predominantly lacked the N terminus. Proteins in the RA3 bound (Figures 5C and 5B) and unbound fractions were separated by SDS-PAGE after denaturation at 100°C in the absence of disulfide-reducing agents and then transferred to membranes; the resulting protein gel blots were probed with different antibodies. Digestion with the protease for 1 hr at 22°C, or for 1 hr at 22°C and then 1 hr at 37°C, gave results that were essentially indistinguishable for all antibodies (data not shown); therefore, only the samples digested for 2 hr are shown. Two patterns for N-terminal fragments were obtained. One constituted a species of ∼43 kD that remained unbound by RA3 in solution but displayed RA3 epitopes when the protein was in denatured form on the blots (Figure 5C, RA3, black arrow at right). This is similar in size to that of Δ3EGFR (Figure 1) and must represent Asp-N cleavage near residue 410 (Figure 5B). The fact that this fragment remained unbound by RA3 in solution, coupled with the fact that we were unable to demonstrate any binding of RA3 by tBP-80 in the fluorescent ligand assay (data not shown), argues that the N terminus of tBP-80 is ordinarily inaccessible. In contrast, a band of slightly faster mobility was in the bound fraction selected by RA3 (Figure 5C, open arrow at left); we hypothesize that the slight increase in mobility resulted from another proteolytic event within the molecule that caused a change in conformation and allowed access by the antibodies to the N terminus. The third most abundant fragment, of ∼27 kD (Figure 5C, open triangle), most likely represents a cleavage near residue 260 (Figure 5B).

In contrast, all four MAbs predominantly recognized a different fragment of ∼38 kD in the RA3 unbound fraction. Because a fragment this size did not react with RA3 (Figure 5C, lane U, right) and because a band this size was strongly labeled by 14G7 (Figure 5C, lane U, marked with filled arrowhead), it must represent the C-terminal portion of the protein with its N terminus near residue 210 (Figure 5B). We hypothesize that the identically sized (∼38 kD) fragments recognized by the other three MAbs also represent C-terminal fragments; however, the validity of this hypothesis is not required for interpretation of any of the subsequent experiments. These results indicate that the epitopes for both 17F9 and 18E7 reside within the C-terminal half of the unique region but differ because 18E7, but not 17F9, reproducibly labeled a fragment of ∼27 kD in the bound fraction (Figure 5C). Additionally, faint bands of ∼43 kD were recognized by 14G7, 19F2, and 18E7. The significance of the 14G7 band is investigated, as shown in Figure 6; bands recognized by the other two MAbs were not studied further.

We wanted to determine whether the polypeptide fragments making up the ∼43-kD band (Figure 5C, RA3) in the unbound fraction were intact, as the results shown in Figure 5A indicate. Protein gel blots carrying proteins from the RA3 unbound fraction from both 1- and 2-hr digestions were therefore prepared from proteins either not treated (Figure 5D, gels at left; RA3 and 2D9) or treated with β-mercaptoethanol (βME) (Figure 5D, gels at right) and probed with two antibodies that recognized N-terminal epitopes: the RA3 polyclonals and MAb 2D9. Both antibodies recognized the 43-kD band in samples not treated with βME (Figure 5D, gels at left panel), and both antibodies recognized 43-kD bands that appeared unchanged after reduction in the presence of βME (Figure 5D, gels at right panel). Therefore, most 43-kD proteolytic fragments were fully protease resistant. We could not repeat the experiment with the other MAbs because their epitopes required the presence of intact disulfide bonds.

Localization of the Ligand Binding Site

We combined protease digestion with assays for ligand binding. Fifteen micrograms of tBP-80 was digested with endoproteinase Asp-N for 1 hr at 22°C at pH 7.5; EDTA was then added, and the digest was incubated in three separate reaction mixtures overnight with proaleurain peptide–agarose. (In the fluorescent peptide binding assay, the absence of divalent cations plus the presence of 5 mM EDTA did not interfere with ligand binding by tBP-80 [data not shown].) In one mixture, no competitor peptide was added; in the second mixture, 0.5 mM Spo peptide was added; and in the third mixture, 0.5 mM proaleurain peptide was added. For each mixture, the resin was collected by centrifugation, and the buffer containing unbound proteins was removed and saved; the resin remaining was then washed, and the bound proteins were eluted at 100°C in SDS-PAGE sample buffer without βME. Bound and unbound proteins were then fractionated without βME treatment on replicate SDS–polyacrylamide gels, transferred to membranes, and probed with MAb 14G7 to map fragments from the C terminus of tBP-80. Results are presented in Figure 6.

Fragments of three different size classes were recognized by 14G7. One was slightly smaller than the 68-kD marker and represented full-length, or nearly full-length, tBP-80 that had not been digested by the protease (Figure 6). A second, less abundant fragment was ∼43 kD (solid arrow); the third was ∼38 kD (open triangle). Most of the full-length and ∼38-kD molecules were recovered in the unbound fractions in each set (Figure 6, lanes 2, 4, and 6); in the absence of competing peptide, however, ∼10% of each was retained on the affinity resin (lane 1). Retention of the ∼38-kD fragment on the resin was the result of interactions that did not depend on the covalently attached proaleurain peptide because the fragment's abundance in the bound fraction was not appreciably diminished by the presence of a vast excess of either Spo (Figure 6, lane 3) or proaleurain (lane 5) peptides. In contrast, retention of the full-length molecules was prevented by both Spo and proaleurain peptides (Figure 6, cf. lane 1 with lanes 3 and 5). Because NPIR is the only sequence motif shared by the two peptides, high-affinity binding by the full-length protein may be determined by that sequence. Interestingly, the ∼43-kD fragment showed a different pattern. Its binding was not prevented by excess Spo peptide (Figure 6, lane 3) but was prevented by excess proaleurain peptide (lane 5). Thus, the binding specificity of a fragment lacking the N-terminal ∼150 amino acids of tBP-80 (the ∼43-kD fragment recognized by 14G7) depended on sequences in the proaleurain peptide that differed from NPIR. Sequences other than NPIR within the proaleurain vacuolar targeting determinant (represented by the proaleurain peptide) that contribute to vacuolar sorting are known to exist (Holwerda et al., 1992; see Discussion).

Expression of Truncated Proteins

Truncation of the BP-80 coding sequence to yield a construct expressing a 562–amino acid protein minus transmembrane domain and cytoplasmic tail sequences has been described (construct NP473; Paris et al., 1997). NP473 was removed from its plasmid by XbaI and SacI digestion and was ligated into pBlueBac4 (Invitrogen, Carlsbad, CA). This intermediate was excised as an EcoRI-XhoI fragment and inserted between those restriction sites in pMT/V5–His A (Invitrogen) to yield CX580; this construct encodes tBP-80 protein. Using a polymerase chain reaction (PCR) 5′ primer positioned just upstream from the BamHI site (Paris et al., 1997), we generated further truncations of the coding sequence with the following primers, which placed a translation stop codon after residues 414, 468, and 517, respectively: CX622, 5′-CCTCGAGCTCACTCATTTGTTTCCACATCATTG-3′; CX623, 5′-CCCCTCGAGCTCAACGCCCATGTCCACTAACTT-3′; and CX624, 5′-CCTCGAGCTCATTCATCAATGT-CTTCACAATTTT-3′. These PCR products were inserted into the BamHI-SacI interval of CX580 to generate constructs encoding Δ3EGFR, Δ2EGFR, and Δ1EGFR, respectively. The fidelity of the constructs was confirmed by sequencing the PCR-generated regions.

Plasmids were transfected into Drosophila melanogaster S2 cells by using protocols supplied with the Drosophila Expression System kit (Invitrogen). We found it important to use culture medium from JRH Biosciences (Lenexa, KS) or from Hyclone (Logan, UT) for optimal results. Stably transformed cell lines were selected with hygromycin and gradually transferred to serum-free medium over a month's time, according to kit protocols. When cells were at a density of 4 × 10⁶ to 6 × 10⁶ per mL, expression of the recombinant proteins was induced by treatment with 500 μM copper sulfate for 36 to 40 hr. The medium was harvested, and protease inhibitors (Sigma) were added to give the following final concentrations: 5 μg/mL leupeptin; 5 μg/mL pepstatin; 5 μg/mL antipain; 1 μg/mL trans-epoxysuccinyl-l-leucylamido-(4-guanidino)butane; and 0.2 mM 4-(2-aminoethyl)benzenesulfonyl fluoride. When necessary, the medium was concentrated under nitrogen gas pressure with an Amicon concentrator (Amicon Inc., Beverly, MA) using a membrane with a 10-kD limit. Glycerol was added to a final concentration of 10%, and the medium was stored at −80°C.

Protein Gel Blots

Methods for SDS-PAGE, transfer of proteins to nitrocellulose membranes, incubation with primary antibodies, and detection with alkaline phosphatase–linked secondary antibodies have been described elsewhere (Rogers et al., 1997). For detection of antibody complexes by enhanced chemiluminescence (Pierce Chemical Co., Rockford, IL), the secondary antibodies were conjugated to horseradish peroxidase, and the treated membranes were exposed to x-ray film according to the manufacturer's instructions.

Purification of tBP-80 from S2 Cell Medium

At each step, tBP-80 was detected on protein gel blots by using enhanced chemiluminescence. Ammonium sulfate was added to 3.6 liters of medium to a final concentration of 2.4 M, and the resulting precipitate was discarded. The supernate was applied with use of a peristaltic pump to a 1.2 × 8-cm column of phenyl–Sepharose (Amersham Pharmacia Biotechnology, Arlington Heights, IL) at 22°C, after which the column was washed sequentially with 2.4 M ammonium sulfate and 1.3 M ammonium sulfate, both in PB (0.05 M sodium phosphate, pH 7.0). All of the following steps were performed at 4°C, and a fast-performance liquid chromatography system (Amersham Pharmacia) was used. The protein retained was batch-eluted with 15 mL of PB and dialyzed extensively against 0.05 M sodium acetate solution, pH 4.5, containing 0.05 M NaCl. After centrifugation at 10,000g for 20 min, the supernate was applied to a Mono S HR 5/5 column (Amersham Pharmacia) equilibrated with the same buffer, and the protein retained was eluted with a 10-mL linear gradient from 0.05 M sodium acetate solution, pH 4.5, containing 0.05 M NaCl to 0.05 M sodium Hepes solution, pH 7.4, containing 0.05 M NaCl; tBP-80 eluted in a narrow peak as the gradient reached completion. The pooled 0.5-mL fractions were made to 3.3 M with ammonium sulfate and applied to a phenyl–Superose HR 5/5 column (Amersham Pharmacia); the column was washed with 1.3 M ammonium sulfate in PB, and the proteins retained were eluted with a 10-mL linear gradient from 1.3 M ammonium sulfate in PB to PB and collected in 0.5-mL fractions; tBP-80 eluted in the fractions between 0.9 and 0.7 M ammonium sulfate. Pooled fractions were dialyzed against 1 liter of 0.05 M Hepes-NaOH, pH 7.9, and applied to a 0.46 × 5-cm Poros 20 HQ column (Perseptive Biosystems, Cambridge, MA) equilibrated with the same buffer. Retained protein was eluted with a 10-mL linear gradient from 0 to 0.5 M NaCl in 0.05 M Hepes-NaOH, pH 7.9, and collected in 0.5-mL fractions; tBP-80 eluting in fractions 9 and 10 represented a total of 0.49 mg of protein, which was pooled and stored at 4°C. The concentrations of the purified protein of the proaleurain peptide were calculated from their A₂₈₀ after dilution into a solution of 6 M guanidine HCl, 0.05 M phosphate buffer, pH 7.0, and 1 mM DTT (Gill and von Hippel, 1989).

Binding Assays

Proaleurain peptide, sequence SSSFADSNPIRPVTDRAASTYC, 200 nmol in 2 mL of 0.05 M Tris-HCl solution, pH 8.0, containing 5 mM EDTA, was mixed with 400 nmol of N-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-yl)methyl) iodoacetamide (BODIPY-FL; Molecular Probes, Eugene, OR) in 0.1 mL of dimethylformamide and 0.04 mL of 0.5 M tris-(2-carboxyethyl)phosphine hydrochloride and incubated at 22°C in the dark for 90 min. Peptide coupled to BODIPY-FL through the Cys residue was separated from free BODIPY-FL by chromatography on a C₁₈ reversed-phase HPLC column by using a gradient from 0.1% trifluoroacetic acid in water to 0.1% trifluoroacetic acid in acetonitrile. The fractions containing peptide coupled to BODIPY-FL were pooled and lyophilized; the peptide was resuspended in sterile water, and its concentration was determined by amino acid analysis.

For the fluorescent peptide binding assay, the labeled peptide was added to 0.3 mL of S2 cell culture medium (pH 6.5 to 6.7) to a final concentration of 10⁻⁷ M and incubated at 22°C for 1 hr; 0.2 mL was then applied to a 30 × 1-cm column of Superdex 200 (Amersham Pharmacia) and eluted with column buffer (0.05 M Hepes-NaOH, pH 7.0, 0.05 M NaCl, 1 mM CaCl₂, and 1 mM MgCl₂) at a flow rate of 1 mL/min. The fluorescence of 0.2-mL aliquots of each fraction was measured with a Luminescence Spectrometer LS50B (Perkin-Elmer Corp., Norwalk, CT), using and and expressed in units of fluorescence intensity.

Anti–BP-80 Monoclonal Antibodies

Monoclonal antibody (MAb) 14G7 has been described previously (Paris et al., 1997); 2D9 was isolated in the same screen. MAbs 19F2, 18E7, and 17F9 were isolated as follows. Membranes from the less dense membrane fraction were prepared from developing pea cotyledons, as previously described (Kirsch et al., 1994). A suspension of the intact membranes was used to immunize mice, and hybridoma cells were prepared at the Cell and Immunobiology Core Facility (University of Missouri, Columbia), as previously described (Rogers et al., 1997). Positive clones were identified by using an ELISA, and the identity of the proteins within the less dense membranes that reacted with the antibodies in the hybridoma media samples was determined by SDS-PAGE and protein gel blot analysis.

Digestion of Purified tBP-80 with Endoproteinase Asp-N

Endoproteinase Asp-N, sequencing grade, was purchased from Roche Molecular Biochemicals (Indianapolis, IN); 2 μg was taken up in 50 μL of sterile water, according to the supplier's directions, and incubated with purified tBP-80 at a ratio of 1:1000 (w/w, protease:tBP-80). The incubation buffer was 10 mM Tris-HCl, pH 7.5, except for the experiment shown in Figure 5A, lanes 3 to 5, for which the buffers were 20 mM Tris-HCl, pH 7.5, or 20 mM sodium acetate, pH 5.0. Protease activity was terminated by bringing the reaction mixture to a final concentration of 10 mM EDTA.