A specific binding protein from Manduca sexta for the insecticidal toxin of Bacillus thuringiensis subsp. berliner.

Biopesticides based on the bacterium Bacillus thuringiensis have attracted wide attention as safe alternatives to chemical pesticides. In this paper, we report, for the first time, the identification and purification of a single binding protein from a lepidopteran insect, Manduca sexta, that is specific for a cryIA toxin of B. thuringiensis. The purified protein appeared as a single band of 210 kDa on a two-dimensional gel, had a pI of approximately 5.5, and stained with Schiff's reagent. The band material was sensitive to proteolytic digestion and was rich with acidic amino acids, indicating its protein nature. Radiolabeled toxin bound to the protein with a Kd value of 708 pM and could be specifically blocked by unlabeled toxin but not by toxins from other subspecies of B. thuringiensis. This study lays the groundwork to clone the toxin binding protein and to determine the molecular mechanism(s) of toxin action.

function, and cessation of feeding followed by death (4, 5 ) . Inhibition of K+-dependent amino acid transport has been observed when brush-border membrane vesicles (BBMV),' isolated from insect midgut, were incubated with activated toxins (6, 7). Formation of pores and colloid osmotic lysis of midgut epithelial cells also has been proposed as a possible mechanism of toxin action (8). If the toxin forms pores, it is not known whether the toxin does so by interacting with a protein present in the BBMV or by insertion into the BBMV. More recent studies have shown the presence of high affinity binding sites in BBMV proteins prepared from susceptible insects (9, 10). van Rie et al. (11) using radioligand binding assays demonstrated a correlation between toxicity and toxin binding to insect BBMV. Using ligand blotting, Oddou et al. (12) demonstrated binding of different cryZ toxins to specific binding proteins (170,140,and 120 kDa) in Heliothis uirescens BBMV proteins. Binding of cryZA(c) toxin to a 120-kDa protein in Manduca sexta BBMV also was reported (13,14).
Although there are several preliminary investigations on the mechanism of action of various B. thuringiensis toxins and the identification of their putative receptors, one of the biggest impediments to our understanding of both the mode of toxicity and the specificity of these toxins is the almost complete lack of information about the nature of the insect target receptor(s). In the present study, we have identified, purified, and partially characterized a protein of 210 kDa to which the cryZA(b) toxin of B. thuringiensis subsp. berliner binds in BBMV of M. sexta. Previously, we first reported the sequence (15) of the cryZA(b) gene of B. thuringiensis subsp. berliner, which is considered the holotype sequence for the gene (3). The 210-kDa protein, present in BBMV of M . sexta (tobacco hornworm), was absent in BBMV of Leptinotarsa decemlineata (Colorado potato beetle, a coleopteran) and Aedes aegypti (mosquito, a dipteran) BBMV. Significantly, the latter two insects lacking the 210-kDa protein in BBMV were not susceptible to the cryIA(b) toxin.

Bacterial Strains and Growth Conditions-B. thuringiensis subsp.
berliner was the gift of M.-M. Lecadet, Institut Pasteur, Paris. Growth, sporulation, and parasporal crystal purification were as described by Tyrell et al. (16). B. thuringiensis subsp. tenebrionis and israelensis were obtained from R. M. Faust, Beltsville Agricultural Research Center, Beltsville, MD. Subspecies tenebrionis was grown in liquid broth medium, and crystals were purified using buoyant density centrifugation in 68% Renografin as described previously (17). Crystal protein dissolved in 68% Renografin, whereas spores and debris sedimented at the bottom of the gradient. The dissolved crystal protein was recrystallized by dialyzing against water containing 1 mM EDTA and 1 mM phenylmethylsulfonyl fluoride (PMSF) at 4 "C. The recrystallized protein was washed and stored in water at 4 "C. Growth of subspecies israelensis and purification of its parasporal crystals were as described by Hurley et al. (18).
berliner were solubilized at 37 "C for 2 h in 50 mM Na2C03 buffer (pH 10) containing 5 mM dithlothreitol. Solubilized crystals were dialyzed against 100 mM NH,HCO, and treated with trypsin as described by Lilly et al. (19) to render a 60-kDa toxin. The toxin fragment was precipitated with 40% ammonium sulfate, and the precipitate was solubilized in 50 mM Tris-HC1 buffer containing 100 mM NaCl, 5% glycerol and dialyzed against the same buffer. The 60-kDa toxin was further purified using fast protein liquid chromatography as described by Hofmann et al. (9). Parasporal crystals from B. thuringiensis subsp. tenebrionis were solubilized in 3.3 M NaBr solution, treated with papain, and the resulting 67-kDa toxin was purified according to the method of Li et al. (20). The 65-kDa protein toxin from B. thuringiensis subsp. isruelensis crystals was selectively solubilized using the Chilcott and Ellar procedure (21), and the resulting 65-kDa polypeptide was further purified on fast protein liquid chromatography using an anion exchange column (9). Activated and purified toxins from 8. thuringiensis subsp. berliner, tenebrionis, and israelensis are referred to hereafter as crylA(b), crylllA, and cryIVD toxins (3), respectively. The amount of toxic protein was determined using the Bradford method (22).
Radio-iodination-The crylA(b) toxin was radio-iodinated using the chloramine-T method (23). Ten pg of cryIA(b) was mixed with 100 pg of chloramine T and 5 pl of NaIZ5I (0.5 mCi) in 100 pl of phosphate-buffered saline (PBS). The reaction mixture was shaken gently at room temperature for 20 s, and the reaction was stopped by the addition of 200 pg of Na&&. Free iodine was removed by gelfiltration on a Sephadex G-50 column equilibrated with PBS. Specific activities of the labeled toxin varied from 10 to 15 mCi/mg.
Preparation of BBMV-M. serta eggs were purchased from Carolina Biological Supply Co., and the larvae were reared on an artificial diet also obtained from Carolina Biological Supply Co. L. decemlineata eggs were kindly provided by J. Kemp, New Mexico State University, and the larvae were reared on potato plants. Aedes aegypti larvae were provided by G . Hunt, Arthropod-Borne Disease Research Laboratory, Agricultural Research Service, United States Department of Agriculture, Laramie, WY. BBMV from third-and fourthinstar larvae of M. sexta, L. decemfineata, and A. aegypti were prepared according to the method described by Wolfersberger et al. (7) except that 1 mM PMSF was added to the buffers. The final pellet was resuspended in 150 mM NaC1,2.7 mM KCl, and 9.5 mM NaP04 buffer containing 0.1 mM PMSF (buffer A) using a 23-gauge hypodermic needle fitted to a plastic syringe (12). The BBMV were used either immediately or frozen in liquid nitrogen at -80 "C. The amount of BBMV proteins was determined by the Bradford method (22).
Ligand Binding Assays-Binding assays were performed as described by Hofmann (9). For competition binding assays, duplicate samples of BBMV (10 pg) were incubated with 0.31 nM 1251-crylA(b) toxin in the presence of varying concentrations of either unlabeled crylA(b), crylllA, or crylVD toxins. Labeled and unlabeled toxins were mixed together before adding them to the BBMV. Radioactive counts due to nonspecific binding (binding observed in the presence of M unlabeled cryIA(b) toxin) was subtracted from the total amount of radioactivity measured. Binding data were analyzed by Scatchard plot analysis (24). To further determine the specificity of 1261-crylA(b) toxin binding to BBMV, labeled toxin (3.1 nM) was incubated with increasing amounts of BBMV in the presence of 100 nM unlabeled cryIA(b) toxin.
Radioligand Blotting-Gel electrophoresis was carried out according to the procedure of Laemmli (25). Fifty pg of BBMV proteins were solubilized in sodium dodecyl sulfate (SDS)-solubilization buffer and electrophoresis was carried out at 4 "C. BBMV proteins then were transferred to Immobilon-P membranes (PVDF) in 25 mM Tris-HCl, 192 mM glycine, and 20% methanol buffer at 100 mA for 16-18 h at 4 "C in a Bio-Rad Trans-Blot apparatus. Lanes containing molecular weight markers were stained with Amido Black, and lanes containing BBMV proteins were incubated at room temperature for 4 h in TBS (10 mM Tris, 0.9% NaCl) containing 5% non-fat dry milk powder, 0.5% Tween 20, and 0.02% sodium azide (pH 8.0) (blocking buffer). The strips were incubated for 90 min at room temperature in 5 ml of blocking buffer containing 1261-crylA(b) toxin (1 X lo6 cpm).
The strips then were washed four times with blocking buffer (20 min/ wash) and finally rinsed with TBS, dried, and exposed to Kodak X-Omat AR x-ray film at -70 "C for 1-2 days. For competition experiments, strips were incubated in 5 ml of the blocking buffer containing 1 x IO6 cpm of 1251-cryZA(b) toxin along with an appropriate concentration of unlabeled toxin as stated in the appropriate figure legends. The position of radioactive bands was determined from the autoradiogram and the amount of bound '261-crylA(b) toxin was determined by excising the radioactive bands and measuring the radioactivity in a Beckman y counter.
Immunodetection of Ligand Binding-BBMV proteins transferred and blocked as described above were incubated overnight in 5 ml of blocking buffer containing unlabeled crylA(b) toxin (1 pglml). Unbound toxin was removed by washing with PBS. Bound toxin was immunochemically visualized as described previously (26). The primary antibody was prepared in rabbits by injecting the purified activated 60-kDa toxin of B. thuringiensis subsp. berliner.
Immunoprecipitation of crylA(b) Binding Protein-Immunoprecipitation was carried out according to Oddou et al. (12). Three pl of crylA(b) antiserum were added to 100 p1 of protein A-Sepharose CL-4B equilibrated in dilution buffer (TBS containing 0.1% Nonidet P-40, 5 mM EDTA, and 0.02% sodium azide) and mixed for 1 h at 4 "C. After washing three times with dilution buffer, cryIA(b) toxin (10 pg in 100 pl) was added, and the mixture was incubated for an additional hour at 4 'C and then washed again three times with dilution buffer. BBMV proteins (100 pg) were solubilized in 100 pl of TBS containing 1 mM PMSF and 1% Nonidet P-40. Solubilized proteins were diluted to 500 p1 with dilution buffer, added to 100 pl of protein A-Sepharose beads linked to crylA(b) toxin, and the sample was mixed for 2 h at 4 "C. Sepharose beads were pelleted and washed four times with dilution buffer and finally rinsed with TBS. The binding complex was dissociated from the beads by heating in SDS-solubilization buffer, and the binding protein was visualized by ligand blotting with crylA(b) toxin as described in the legend to Fig. 6.
Purification of crylA(b) Toxin Binding Protein-A combination of immunoprecipitation and two-dimensional gel electrophoresis, as described by Wada et al. (28), was used to further purify the crylA(b) toxin binding protein. Brush-border membrane vesicles (5 mg/ml) in buffer A containing 1 mg/ml of ovalbumin were incubated with 20 pg of unlabeled crylA(b) toxin for 20 min at 37 "C to saturate the toxin binding sites. After incubation, the suspension was chilled on ice, washed with a buffer containing 50 mM Tris-HC1, 250 mM NaC1, 1 mM PMSF, 6 mM EDTA, 10 p M aprotinin, 2 p M leupeptin, 2 p M pepstatin A, and 2 p~ antipain (buffer B). Washed crylA(b) saturated BBMV then were solubilized by stirring in buffer B containing 1% Nonidet P-40 at 4 "C for 1 h, and the insoluble material was removed by centrifugation at 13,000 rpm for 30 min at 4 "C. The supernatant was precleared with cryZIIA antisera and protein A-Sepharose beads. Precleared supernatant was incubated with 2 p1 of crylA(b) antiserum for 10 min. Fifty pl of protein A-Sepharose beads were then added, and the suspension was mixed for 1 h at 4 "C. The protein A beads were pelleted and washed four times with buffer B containing 0.5% Nonidet P-40 and 0.02% SDS. The binding protein-crylA(b) complex was eluted from the beads by heating the sample in 2% sample buffer for 2 min at 100 "C. The extract was mixed with 5 volumes of 9 M urea, 4% Nonidet P-40, 2 % 2-mercaptoethanol, and 2% 3.5-10 ampholytes and then analyzed by two-dimensional gel electrophoresis according to O'Farrell et al. (29). Isoelectric focusing was done in the first dimension in a 4% acrylamide gel containing 8 M urea, 2% Nonidet P-40, and 2% 3.5-10 ampholytes. Electrophoresis was carried out in the second dimension in an SDS-polyacrylamide (7%) slab gel. Binding protein present in the two-dimensional gel was identified by ligand blotting with '251-crylA(b) toxin after transferring the protein in the gel to PVDF.
Amino Acid Analysis-Multiple two-dimensional gels were run and the 210-kDa protein was transferred to PVDF membranes (30), stained with Coomassie Blue, and the protein spots were excised from the membranes. The 210-kDa binding protein spots (0.3 pg) were hydrolyzed with 6 N HCl for 35 min at 155 "C under vacuum. Amino acid composition was determined after high performance liquid chromatography by the 0-phthaladehyde method (31) using a C18 reversephase column. Blank spots from the same membrane were excised and hydrolyzed under similar conditions, and the amino acid values from the blank spots were used to correct the amino acid composition of the 210-kDa binding protein.
Protease Digestion-M. serta BBMV proteins were incubated with four different proteolytic enzymes (trypsin, proteinase K, endoproteinase Glu-C, and papain, all from Sigma) at a final concentration of 0.5 mg/ml at room temperature for 60 min. Buffers used for trypsin and proteinase K digestion were prepared according to Hofmann et al. (27). Papain digestion was carried out in 2.5 mM sodium phosphate buffer containing 0.5 mM cysteine, 0.5 mM Na2EDTA, and 0.5 mM

B. thuringiensis Toxin Binding Protein
MgSO,. Sodium phosphate buffer (0.2 M (pH 7.4)) was used for endoproteinase Glu-C digestion. The reactions were stopped by adding SDS-solubilization buffer and heating the mixture at 98 "C for 3 min. The digested proteins were electrophoresed as described above and analyzed by ligand blotting with 1251-cryIA(b) toxin.
Insect Toxicity Assays-Toxicity assays using cry toxins on neonate larvae of M. sexta were performed as described by Wabiko et al. (26).
Toxicity assay of purified cry toxins on neonate larvae of L. decemlineata was performed by painting the potato leaves with the purified cry toxins and feeding the treated leaves to the larvae (32). Toxicity of purified cry toxins on early fourth-instar larvae of A. aegypti was measured as described by Hurley et al. (18).

RESULTS
Specific Toxicity of B. thuringiensis Toxins-To demonstrate the specific toxicity of the three different toxins of B. thuringiensis, we bioassayed M. sexta, L. decemlineata, and A. aegypti. As can be seen in Table I, each toxin was specific for only one insect. No toxicity was exhibited by cryZA(b) toxin against L. decemlineata and A. aegypti at the highest concentration used in the bioassay (2.5 x lo4 ng/ml) for both insects.
The cryZIZA toxin was not toxic to M. sexta and A. aegypti at the highest concentration tested (500 ng/cm' and 2.5 x IO4 ng/ml, respectively). The cryIVD toxin was not toxic to M. sexta and L. decemlineata at the highest concentrations used (500 ng/cm* and 2.5 X lo4 ng/ml, respectively). Because of these extreme specificities, we used the cryZIIA and cryZVD toxins as controls in our ligand binding and blotting experiments as well as BBMV proteins from L. decemlineata and A. aegypti in the ligand blotting experiments. toxin. Consistent with these data, binding of the 1251-cryIA~) toxin increased linearly with increasing concentrations of BBMV proteins from 100 to 1000 @g/ml (Fig. 1B). The LC50 value for 1251-cryZA(b) was similar to that of the natural toxin (Table I), demonstrating that the radio-iodinated toxin is fully active.

Identification of 1251-cryZA(b) Binding Protein by Radioli-
gand Blotting-BBMV proteins of M. sexta were blotted onto PVDF, and a binding protein was identified by incubating the membrane with 1251-cryIA(b). Under both reducing (Fig. 2, lane 1 ) and nonreducing (Fig. 2, lune 2 ) conditions, cryIA(b) toxin exclusively bound to a protein band of 210 kDa. The profile of stained BBMV proteins of M. sexta, L. decemlineata, and A. aegypti was quite broad (Fig. 3A, lunes  2-4). Molecular sizes ranged from greater than 200 kDa for   (Fig. 3B, lane 2 ) but not in BBMV proteins prepared from L. decemlineata and A. aegypti (Fig. 3B, lanes 3 and 4 ) . The major protein that recognized the cryIA(b) toxin is approximately 210 kDa (see arrow in Fig. 3B, lane 2). The smaller bands that were faintly labeled with the toxin probably are degradation products resulting from endogenous proteases in the BBMV. This particular preparation was several days old, whereas the preparation used in Fig. 2 was only a few minutes old.

Specificity of '25Z-cryIA(b) Toxin Binding in Ligand
Blots-T o determine the specificity of binding of cryZA(b) toxin to the 210-kDa protein, blots of M. sexta BBMV proteins were incubated with 'z51-cryZA(b) toxin in the presence of increasing concentrations of unlabeled cryIA(b) toxin. Autoradiography of these blots revealed a gradual reduction in the intensity of the 210-kDa band in parallel to increasing amounts of unlabeled cryIA(b) toxin (Fig. 4, lanes 2-6). Bound '251-cryIA(b) toxin was quantitated by excising the appropriate bands and measuring their radioactivity in a y counter. At a concentration of 83 nM (Fig. 4, lane 6), unlabeled cryIA(b) toxin reduced the binding of 1251-cryIA(b) toxin to 3% of the binding observed in the absence of unlabeled cryZA(b) toxin (Fig. 4

, lane I ) .
Scatchard analysis using this binding data revealed a K d value of 708 PM. Binding of cryIA(b) toxin to the 210-kDa binding protein was not affected by cryIIZA and cryIVD toxins a t a concentration of 83 nM (Fig. 4, lanes 7 and 8 ) . These results suggest that binding of the 1251-cryZA(b) toxin to the 210-kDa protein on the blots is very specific with high affinity.
Zmmunodetection of cryIA(b) Binding Protein-BBMV proteins of M. sexta were blotted onto nylon membrane and incubated with unlabeled cryZA(b) toxin. The toxin-binding protein complexes were visualized using polyclonal antisera to the purified 60-kDa toxin of B. thuringiensis subsp. berliner.
A single protein band of 210 kDa was observed (Fig. 5). No bands were observed on blots when M. sexta BBMV proteins were replaced with A. aegypti and L. decemlineata BBMV or when M. sexta BBMV blots were incubated without cryIA(b) toxin (data not shown).
Immunoprecipitation of cryIA(b) Binding Protein-Immunoprecipitation experiments were performed to further confirm the specificity of the binding of cryIA@) toxin to the 210-kDa protein. M. sexta BBMV proteins were solubilized in 1% Nonidet P-40 and immunoprecipitated with toxin-antitoxinprotein A-Sepharose beads. The precipitate was solubilized in SDS, electrophoresed, and blotted (Fig. 6). Immunoprecipitation and ligand blotting with '251-cryIA(b) showed a band of 210 kDa (Fig. 6, lane Z), indicating precipitation of the same binding protein identified in earlier experiments. Upon immunoprecipitation and subsequent ligand blotting with unlabeled cryZA(b) toxin, Western blotting showed a 210-kDa band (Fig. 6, lane 1 ) as well as a major band of 60 kDa (cryIA(b) toxin) and other minor bands of 50-55 kDa (immunoglobulin fragments).
Purification of the Binding Protein-The binding protein was purified by immunoprecipitation and two-dimensional gel electrophoresis. Immunoprecipitates first were analyzed by SDS-polyacrylamide gel electrophoresis (Fig. 7). The COOmassie-stained gel revealed three protein bands of 210, 60, and 55 kDa as shown above, showing selective precipitation of the 210-kDa cryZA(b) toxin binding protein with the 60-kDa crylA(b) toxin. The cryIA(b) binding protein present in the two-dimensional gel was verified by Coomassie Blue staining (Fig. 8 A ) and ligand blotting with Iz51-cryZA(b) toxin (Fig.  8B). The binding protein migrated to the acidic side of the pH gradient and has an estimated PI of 5.5 f 0.1 and an apparent molecular mass of 210 kDa. The purified 210-kDa binding protein stained with Schiffs reagent ((data not shown), suggesting glycosylation of the binding protein.
Amino Acid Analysis- Table I1 shows the amino acid composition of 210-kDa spots blotted from two-dimensional gels to PVDF membranes. Because of the possibility of contami-  lanes 7 and 8). A Scatchard plot of the binding observed in lanes 1-5 is shown in the inset. Molecular size markers are the same as those in Fig. 2. Protein markers on the left are the same as in Fig. 2. composition of the sample, and therefore, the blank spot values were used to correct the amino acid values of the 210-kDa protein spots. The binding protein had a high content of acidic amino acids (Asp, Glu) and hydrophobic amino acids (Ala, Val, Leu, Ile). No attempts were made to estimate the amounts of cysteine and tryptophan.

B. thuringiensis Toxin Binding Protein
Protease Treatment of M. sexta BBMV Proteins-BBMV proteins from M. sexta were treated with trypsin, proteinase K, papain, and endoproteinase Glu-C, subjected to SDSpolyacrylamide gel electrophoresis, and then ligand-blotted with '251-cryIA(b) (Fig. 9). Lane 1 shows the control band pattern when BBMV proteins were kept a t room temperature for 1 h without any enzyme treatment. Lane 1 contained several protein bands, including the 210-kDa band. As was indicated above in the BBMV competition experiments (Fig.  3), we believe that the lower molecular mass components probably are proteolytic products of the 210-kDa protein resulting from the action of endogenous proteases present in the BBMV. All four enzymes destroyed the ability of the 210-kDa protein to bind the '2sI-cryIA(b) toxin. Trypsin and papain digestion (lunes 2 and 4 ) resulted in binding the cryIA(b) toxin to three lower molecular mass polypeptides (50-65 kDa). Upon digestion with endoproteinase Glu-C, three polypeptides of mass ranging from 60 to 120 kDa bound to the 12RI-~ryIA(b) toxin (lune 5 ) . Proteinase K completely destroyed the binding of '2sI-cryIA(b) toxin to M. sexta BBMV proteins (lane 3 ) .

TABLE I1
Amino acid composition of thecryIA(b) binding protein Protein spots containing the binding protein and three blank spots of similar size from the same membrane were hydrolyzed and analyzed as described under "Materials and Methods." Values shown are the mean of two analyses corrected relative to the blank values.

Amino acid
Mol   Binding protein-cryIA@) toxin complex was eluted from protein A-Sepharose beads with 2% SDS sample buffer, heated for 2 min a t 100 "C, and mixed with urea solution as described under "Materials and Methods." The first dimension was isoelectric focusing in a 3.5-10 pH gradient 4% acrylamide gel containing 8 M urea, 2% ampholytes, and 2% Nonidet P-40. The second dimension was polyacrylamide gel electrophoresis in a 7% slab gel. A, the gel was stained with Coomassie Blue, and the corresponding 210-kDa band was cut from the gel, eluted, and subjected to amino acid analysis. R, a duplicate gel was blotted to PVDF and probed with '2sI-cryIA(b) toxin. The membrane was washed, dried, and autoradiographed for 24 h. Protein markers are the same as in Fig. 2. cally recognizes and binds the cryIA(b) toxin of B. thuringiensis subsp. berliner. Trypsin-activated cryIA(b) toxin bound to M. sexta BBMV saturably, and the binding was displaceable only with unlabeled cryIA(b) toxin and not with cryIIIA or cryIVD toxins. The equilibrium dissociation constants, calculated using ligand binding ( Fig. 1) and ligand blotting assays (Fig. 4), were 1,075 and 708 PM, respectively. These results indicate the high-affinity saturable binding of the cryIA(b) toxin to a single site in M. sexta BBMV. Apparently, this high-affinity binding accounts for, in part, the specificity of the toxin for a lepidopteran insect rather than for a coleopteran or dipteran insect (Table I).

DISCUSSION
Because the 210-kDa protein recognized and bound the cryIA(b) toxin under both reducing and nonreducing conditions, it is not a disulfide-linked oligomer. The 210-kDa toxin binding molecule was susceptible to proteolysis and, therefore, is likely to be a protein. Degradation occurred in preparations that were allowed to stand for only 1 h a t room temperature. Similar degradation occurred in membrane vesicles that were stored a t -70 "C for several days (results not shown).
Binding of the cryIA(b) toxin to the 210-kDa protein was observed (Fig. 5) when BBMV proteins were ligand blotted with unlabeled cryIA(b) toxin, and the binding complex was visualized using anti-cryIAfb) antiserum. Immunoprecipitation with the same antiserum and subsequent ligand blotting with cryIA(b) toxin also showed binding of the toxin to the 210-kDa protein (Fig. 6). The radioligand binding and blotting experiments, taken together with immunochemical detection, unequivocally demonstrates that the 210-kDa protein is an exclusive binding protein for the cryIA(b) toxin. This binding is toxin-specific because it could be blocked by unlabeled cvIA(b) toxin but not by other cry toxins. Furthermore, the binding is species-specific for M. sexta but not for A. uegypti and L. decemlineata.
Protease digestion of the cryIA(b) binding protein produced interesting results (Fig. 9). Trypsin and papain treatment yielded polypeptides of 50-65 kDa in parallel to disappearance of the 210-kDa protein, suggesting a precursor-product relationship. Interestingly, the putative 50-65-kDa fragment was capable of recognizing the toxin. Therefore, the toxin binding site is likely to be located in this fragment. It has been reported that treatment of Pieris brassicae (cabbage white butterfly) and H. virescens (tobacco budworm) BBMV proteins with trypsin did not affect the ability of cryIA(b) toxin to bind to BBMV or to a protein in protein blots (12, 28). Endoproteinase Glu-C treatment produced binding fragments of 120, 100, and 60 kDa with concurrent disappearance of the 210-kDa protein, indicating the presence of susceptible aspartic acid or glutamic acid residues. The amino acid composition of the protein purified by two-dimensional gel electrophoresis revealed an abundance of these two amino acids (Table 11). Protease K treatment resulted in the total loss of toxin binding, probably due to the exhaustive digestion of the cryZA(b) binding protein.
We have demonstrated that M. sextu contains within its midgut a 210-kDa protein which is capable of binding the cryIA ( there are different binding proteins, depending on the insect species, for cryZ toxins of B. thuringiensis, and in some instances, cryI toxins can compete for the same binding site. Such molecules appear to have binding sites for both toxins. Also, it is possible that distinct binding proteins form noncovalent complexes in BBMV and become capable of recognizing various toxins competitively. Consequently, any lack of competition in blotting experiments (12) could be due to separation of these proteins into individual subunits.
To our knowledge, we are the first to report the purification 14. Obviously, the cryZIIA and the cryIVD toxins do not bind the 210-kDa protein of M. sextu, and therefore, we are investigating further the biochemical properties of this protein to gain a better understanding of its binding specificity and, eventually, to determine the molecular basis of insecticidal action of the cryIA(b) toxin of B. thuringiensis subsp. berliner.