JOURNAL

OF INVERTEBRATE

PATHOLOGY

5%

75-80 (1992)

Bacillus thuringiensis Crystal Protein: Effect of Chemical Modification of the Cysteine and Lysine Residues CHRISTIN

T.

CHOMA

AND

HARVEY

KAPLAN

Department of Chemistry, University of Ottawa, Ottawa, Ontario, Canada KIN 6N5 Received November 13. 1990; accepted April 3, 1991

1988; Haider and Ellar, 1987). Due to the disruption of osmotic balance, the cells swell and lyse, resulting in larval death. At present, the molecular mechanism for this activity is not known, but there is evidence that the toxin is composed of a binding domain which determines host specificity and a lytic domain which is responsible for disruption of the membrane (Ge et al., 1989; Hiifte and Whiteley, 1989). The lepidopteran-active protoxin molecule derived from the cryIA(c) gene type contains 16 cysteine residues in the form of disulfide linkages (Hofte and Whiteley, 1989). The cysteine residues stabilize the crystal by forming symmetrical interchain bonds; i.e., each cysteine residue forms a bridge to the corresponding cysteine in an adjoining protoxin molecule (Bietlot et al., 1990). In the reduced form, the sulfhydryl groups in the protoxin lie on the surface of the molecule, and under the alkaline conditions (pH 9-10) of the larval gut (Narayanan et al., 1976), the sulthydryl groups are present in the S- ionized state (Bietlot et al., 1990). The disulfide and sulfhydryl structures of the protein crystal are responsible for the unusual solubility properties of this protein and appear to play an important Press, In.% role in the generation of toxin in the larval gut (ArvidKEY WORDS: Bacillus thuringienais; Choristoneura fu- son et al., 1989). ninifera; chemical modification; protoxin; solubility; cysThe toxin is derived from the N-terminal half of the teine; sulfhydryl; lysine; amino. protoxin molecule. It is generated by an unusual activation process involving sequential proteolytic cleavages beginning at the C-terminus and proceeding toINTRODUCTION ward the N-terminus of the molecule until the proThe major component of the Bacillus thuringiensis teinase-resistant toxin is released (Choma et al., 1990). subsp. kurstuki HD-73 protein crystal is a 130-kDa pro- All the cysteine residues and 31 of the 34 lysine resitein, protoxin. The crystal can be readily solubilized dues are removed during activation (Bietlot et al., under alkaline conditions or with small amounts of 1989). Although the C-terminal half of the protoxin thiol reagents (Nickerson, 1980). On ingestion by sus- molecule is not required for toxic activity, it is noneceptible insect larvae, the alkaline conditions of the theless highly conserved among protoxins produced by larval gut release the protoxin, which is then acted on various B. thuringiensis strains (Hofte and Whiteley, by proteinases to produce a 58-70 kDa toxic fragment, 1989). This suggests that the unusual distribution of toxin (Andrews et al., 1987). Current evidence indicysteine and lysine residues has some functional sigcates that Bacillus thuringiensis toxins bind to recep- nificance. It appears that the sulfhydryl groups play a tors in the midgut epithelium and generate pores in crucial role in crystal formation (Hiifte and Whiteley, the cell membrane (Jaquet et al., 1987; Hofmann et al., 1989; Nickerson, 19801, but the role of the lysine resiThe 16 cysteine residuesof reduced protoxin from Bacillus thutingiensis subsp. kurstaki I-ID-73 can be quantitatively reacted with: (a) iodoacetic acid, to give carboxymethyl protoxin; (b) iodoacetamide, giving carbaminomethyl protoxin and (c) N-(@iodoethyl)trifluoroacetamide to give aminoethyl protoxin. The carboxymethyl derivative was found to be significantly more soluble at neutral pH values where both the native protoxin and the carbaminomethyl derivative exhibit low solubilities. At the alkaline pH values (pH 9.5-10.5) normally usedto solubilize the crystal protein, the native protein was slightly more soluble than either the carboxymethyl or the carbaminomethyl derivatives. The aminoethyl derivative had an extremely low solubility at all pH values. Succinic anhydride reacted with only 35% of the lysine residues in both the carboxymethyl and the carbaminomethyl protoxin derivatives. Nonetheless, these succinylated protoxins exhibited significantly increasedsolubilities at neutral pH values. All the derivatives were found to retain full insecticidal activity toward spruce budworm (Choristeneura fufimerana) larvae. It is concluded that all the cysteine residuesand modified lysine residuesare on the surface of the protein and that derivatization doesnot alter Academic the conformation of the solubilized protoxin. o 1992

75 0022-2011192

$1.50

Copyright 0 1992 by Academic Press, Inc. All rights

of reproduction

in any form reserved.

76

CHOMA AND KAPLAN

dues is less clear. The present communication examines the roles of the sulfhydryl and amino groups in the protoxin and describes the effects of chemical modification of these residues on the physical and biological properties of the protein. MATERIALS

AND METHODS

Reagents

Aminoethylreagent was purchased from Pierce (Rockford, IL), and 2,4-dinitrofluorobenzene (DNFB)l was obtained from Aldrich (Milwaukee, WI). All biochemicals were acquired from Sigma (St. Louis, MO); other chemicals used were high purity preparations from commercial sources. Crystal Preparation Bacillus thuringiensis subsp. kurstaki HD-73 was grown in half-strength trypticase soy broth. The cells were lysed in double distilled water and the crystals were purified by Renografin (Squibb, Montreal, Quebec) gradients as described previously (Carey et al., 1986). Purified crystals were stored in distilled water at 4°C. Protein Quantification

Protoxin was quantified in KOH solution, pH 13, by absorbance at 280 nm. Amino acid analysis showed that protoxin at a concentration of 1 mg/ml gives an absorbance of 1.37 for a l-cm pathlength; the corresponding absorbance for toxin is 1.61 (Bietlot et al., 1989). Preparation

of Protoxin Derivatives

Carboxymethylcysteine cysteine (CAM) protoxin.

(CM) and carbaminomethylB. thuringiensis HD-73 pro-

Succinyl-carboxymethylcysteine (SCM) and succinylcarbaminomethylcysteine (SCAM) protoxin. CM or

CAM protoxin (10 mg) was dissolved in 10 ml 0.1 M carbonate/bicarbonate buffer, pH 9.0. At 1-min intervals, succinic anhydride was added in four, 25-mg aliquots to the rapidly stirring sample, and the pH was maintained at 9.0 by the addition of base. The samples were thoroughly dialyzed at 4°C against water. Aminoethylcysteine (AE) protoxin. Aminoethylation of native protoxin was achieved by modifying the procedure of Schwartz et al. (1980). Protoxin (10 mg) was dissolved in 10 ml of 0.1 M CAPS buffer, pH 10.0, containing 16 ~1 @mercaptoethanol. The temperature of the sample was raised to 40°C and maintained there for the duration of the reaction. The sample was stirred rapidly and kept under nitrogen to prevent oxidation of the cysteine residues. N-(@Iodoethyl)trifluoroacetamide (also called aminoethylreagent) was added in a 25fold molar excess over the concentration of -SH in the sample. The reagent (1.1 g), dissolved in 1 ml of methanol, was added in two aliquots over a 30-min period. The pH of the sample was maintained at pH 10.0 by the addition of base. The reaction was complete after 2 hr, and the sample was then extensively dialyzed against water at 4°C. Quantification

of Succinylated Lysine

Rationale. DNFB reacts with free amino groups in proteins to form acid-stable derivatives. If a succinylated protein is treated with DNFB and then hydrolyzed in acid, only those lysine residues which were originally succinylated will appear as free lysine upon amino acid analysis. Therefore, the amount of lysine detected will be a direct measure of the extent of succinylation in the original protein.

toxin (10 mg) was dissolved at 20°C in 10 ml of 0.1 M carbonate/bicarbonate buffer containing 100 ~1 B-mercaptoethanol. To help solubilize the protoxin, 2 M NaOH was added to bring the solution to pH 10.0. After 5 min, HCl was added to lower the pH to 9.0. The solution was thereafter maintained at pH 9.0 by the addition of base. Carboxymethylation of the cysteine residues was achieved by adding 270 mg of iodoacetic acid, dissolved in 1 ml of 1 M NaOH, to the stirring protoxin solution. To carbaminomethylate the sulfhydry1 groups, 270 mg of iodoacetamide was added directly to the reduced protein. Each reaction was stopped after 5 min by the addition of 100 ~1 of 8-mercaptoethanol. The samples were extensively dialyzed against water at 4°C.

Amino Acid Analysis

1 Abbreviations used: CM, carboxymethyl protoxin; CAM, carbaminomethyl protoxin; AE, aminoethyl protoxin; SCM, succinylcarboxymethyl protoxin; SCAM, succinyl-carbaminomethyl protoxin; LD,,, dose resulting in 50% lethality; DNFB, 2,4-dinitrofluorobenxene; CAPS, 3-(cyclohexylamino)-1-propane-sulfonic acid.

Samples were hydrolyzed with 6 M HCl in vacua for 24 hr with norleucine added as an internal standard. Amino acids were quantified on a TSM Technicon amino acid analyzer using a ninhydrin detection sys-

Method. Succinylated protein (1 mg) was dissolved in 1 ml of 6 M guanidinium chloride. The sample was saturated with sodium bicarbonate and 25 ~1 of DNFB (50% in acetonitrile) was added. The sample was protected from light and was shaken at 20°C for 4 hr, then a further 25 ~1 of DNFB was added and the sample was shaken for 16 hr. One drop of octanol was then added to prevent foaming and the sample was acidified to pH 2 by the addition of concentrated HCl. Unreacted DNFB was extracted with ether and residual salt was removed by washing the protein pellet several times with water. The sample was dried, acid hydrolyzed, and the lysine content was quantified by amino acid analysis.

MODIFICATION

tern. Carboxymethyl acid, and aminoethyl lysine peak.

OF THE B. thuringiensis

cysteine eluted before aspartic cysteine eluted in front of the

Toxin Preparation

Toxin was prepared from native protoxin as previously described (Bietlot et al., 1989). Toxins were generated from the chemically modified protoxins in a similar manner. pH-Dependent Solubility

of Derivatives

Approximately 8 mg of protein was added to 0.5 ml of a solution containing 0.1 M boric acid, 0.1 M citric acid, and 0.1 M disodium phosphate buffers. For the crystal protein sample, the buffer also contained 0.1% P-mercaptoethanol. The solubilities of the protoxin derivatives at pH values ranging from pH 5 to 11 were determined by adjusting the pH of the samples, allowing the samples to stir vigorously for 30 min at 2O”C, and then measuring the pH of the suspensions. Undissolved protein was removed by centrifugation and the absorbance of the supernatants at 280 nm was measured. The absorption coefficient noted above was used to calculate the concentration of protoxin in each solution. The supernatants were returned to the corresponding pellets and vortexed, the pH was adjusted by an increment of approximately 0.5 pH unit, and the procedure was repeated. Proteolysis of Protoxin

The proteolytic degradation of the C-terminal half of both native and chemically modified protoxins was conducted using the procedure described previously (Choma et al., 1990). Gel Electrophoresis

Homogeneous SDS-polyacrylamide gels (12%) were run on a Pharmacia Phast electrophoresis system with preformed gels and other materials supplied by Pharmacia. Gels were stained with Coomassie blue. Cell Bioassays

The biological activities of toxins generated by tryptic digestion of the chemically modified protoxins were quantified using CF-1 insect cells in a lawn assay (Gringorten et al., 1990). Samples were diluted to 50 ng/$ in 0.1 M CAPS buffer, pH 10.5, and twofold serial dilutions were applied to the assay plates. Threshold levels were estimated to lie between the last visible spot colored by trypan blue, indicating injured cells, and the following unaffected sample spot.

77

PROTOXIN

Insect Bioassays

The toxicities of the protoxin derivatives toward eastern spruce budworm (Cholistoneura fuminiferana) were determined by force-feeding larvae with a homogeneous suspension of protoxin in water. Day-l sixthinstar larvae were force-fed with 2 ~1 of either native crystal protoxin or derivatized protoxin dissolved in 0.2 M CAPS, pH 10.5. Tenfold dilutions of protoxin starting at 1000 p,g/pl were prepared; the concentration of protein in each original sample was determined by absorbance at 280 nm, and the samples were diluted accordingly. The toxicity of each dilution was determined by force-feeding 15 larvae. After treatment, the larvae were held at 25°C on a 16-hr photoperiod and mortality was scored after 10 days. Once the approximate concentration (within an order of magnitude) of each protoxin derivative required to kill spruce budworm larvae had been determined by this method, the procedure was repeated using twofold serial dilutions spanning the appropriate concentration range. On average, six twofold serial dilutions were each fed to 35 larvae and mortality (LD& was determined after 10 days. The statistical reliability of the data was analyzed by the Polo-PC probit analysis program, obtained from LeOra Software, Berkeley, California. RESULTSANDDISCUSSION

The extent of chemical modification of cysteine residues in the CM, CAM, and AE protoxin derivatives, and of the lysine residues in the SCM and SCAM derivatives, is shown in Table 1. Amino acid analysis was conducted on at least three preparations of each derivative; the mean value for the extent of derivatization is reported. It is clear from Table 1 that, within the limits of experimental error for amino acid analysis, the cysteines are quantitatively derivatized in CM, CAM, and AE protoxin. The ease with which the sulfhydryl groups in protoxin can be derivatized in the absence of denaturants supports the conclusion of Bietlot et al. (1990) that all the cysteine residues are on the surface of the protoxin molecule, since a surface location for these residues is the most straightforward explanation TABLE 1 Extent of Chemical Modification of Cysteine or Lysine Residues in B. thuringiensis HD-73 Protoxin Derivatives Derivative CM CAM AE SCM SCAM

No. modified residue9 15.7 16.4 15.7 11.6 12.6

-+ 1.2 f 0.8 2 1.1 ? 1.4 2 1.7

cysteines cysteines cysteines lysines lysines

Percentage modified residues 98 102 98 34 37

a Average from amino acid analysis of three preparations derivative.

of each

78

CHOMA AND KAPLAN

for their ready accessibility to modifying reagents. In contrast, only about 35% of the lysine residues are succinylated in SCM and SCAM protoxin. Although the relative difficulty of modifying the lysine residues may be attributed in part to electrostatic and steric effects, the lack of extensive modification probably reflects the inaccessibility of the lysine residues within the protoxin molecular structure. Only 70% of the lysines could be converted to uncharged residues with acetic anhydride, suggesting that a significant number of the lysine residues in protoxin are buried within the molecule as salt bridges. The effect of derivatization on the solubility of protoxin was determined for the pH range 5-11 (Fig. 1). All the derivatives except AE protoxin are more soluble below pH 8 than native protoxin. Above pH 9, native protoxin (in the presence of 0.1% P-mercaptoethanol) is slightly more soluble than any of the derivatized protoxins. The introduction of negatively charged groups into the cysteine residues in CM protoxin renders the protein more soluble at neutral pH than its electrically neutral counterpart, CAM protoxin. The addition of succinyl groups to the lysines of the CM derivative significantly increases its solubility; a similar effect is observed with SCAM protoxin. The insolubility of AE protoxin at all pH values may be due to the overall electrical neutrality of the molecule, as the introduction of 16 extra basic groups into the protein effectively balances the number of acidic residues. It has been reported previously (Choma et al., 1990) that the C-terminal half of protoxin is digested by trypsin in an unusual fashion, with fragments of ap-

J”

........w

crystal

--t

CM protoxin

-.-‘9-.-- o-

----Q---,

SCM

I

protoxin

protoxin

CAM -

SCAM

proximately 10 kDa being removed in a sequential manner, starting from the C-terminus and proceeding toward the N-terminal region. The sequential proteolysis ends with a 67-kDa proteinase-resistant toxin being produced. Upon exposure to trypsin, CM, CAM, AE, SCM, and SCAM protoxins were each found to give rise to a sequential fragmentation pattern similar to that obtained from native protoxin (Fig. 2). Proteolysis of all the derivatives ended in the production of a proteinase-resistant 67-kDa fragment. The fragmentation patterns of SCM and SCAM appear more complex than those obtained from the nonsuccinylated derivatives, suggesting that succinylation may expose more arginine cleavage sites. The retention of the unusual sequential proteolysis process by all the protoxin derivatives indicates that no major change in the tertiary structure of the C-terminal half of the protoxin mole-

A 12345678 -100

- 25

-25

protoxin protoxin

AE protoxin

I

I

- 25

5

6

7

0

9

10

11

12

PH

1. The solubility of native crystal and derivatized protoxins as a function of pH. The native crystal sample contained 0.1% p-mercaptoethanol. FIG.

FIG. 2. SDS-polyacrylamide gel electrophoresis of the tryptic proteolysis products of native crystal and derivatized protoxins. The protoxin derivatives were: (A) native crystal; (B) AE; (C) CM; (D) SCM; (E) CAM; (F) SCAM. Proteolysis was conducted at 2O”C, pH 10.5, in 0.1 M CAPS with a protoxin:trypsin ratio of 5OOO:l by mass. Incubation times are designated by the numbered lanes as follows: 1, 0 min; 2,6 min; 3,10 min; 4,20 min; 5,40 min; 6,60 min; 7,120 min. For all samples, the proteinase-resistant toxin is the principal band after 120 min of proteolysis. The molecular mass scale on the left is in kilodaltons and was determined from the mobility of the molecular mass standards in lane 8: phosphorylase (94 kDa); bovine serum albumin (67 kDa); ovalbumin (43 kDa); carbonic anhydrase (30 kDa); soybean trypsin inhibitor (20 kDa) and a-lactalbumin (14 kDa).

MODIFICATION

OF THE B. thuringiensis

cule has resulted from the chemical modification of the cysteine and lysine residues. The toxicity of the proteinase-resistant 67-kDa fragment from each of the derivatives was tested using a CF-1 cell lawn assay (Gringorten et al., 1990). The results presented in Table 2 show that except for the AE derivative, modification of the cysteine and lysine residues in protoxin has no effect on the biological activity of the toxin molecule derived from the N-terminal half of protoxin. Similarly, in force-feeding assays using spruce budworm (Choristeneuru fufimeruna) larvae, all the derivatives exhibited the same LD,, (Table 3). It is not clear why toxin prepared from AE protoxin displayed a lower toxicity in the cell assay than in the feeding assay. Nonetheless, these results indicate that the structure of the N-terminal toxic half of the protoxin molecule remains essentially unaffected by chemical modification of the C-terminal half. The retention of toxicity further supports the view that the C-terminal half of the molecule functions only in crystal formation (Choma and Kaplan, 1990; Hofte and Whiteley, 1989). The results presented here conflict with the observations of Schesser et al. (1977) and Bulla et al. (1977), who reported that treatment of protoxin from B. thuringiensis strain HD-1 with iodoacetic acid resulted in a total loss of toxicity toward the tobacco hornworm, Manduca se&u. However, Schesser et al. (1977) did not report the reaction conditions used for the carboxymethylation of protoxin, nor was the derivative characterized. Bulla et al. (1977) treated reduced protoxin with a large excess of iodoacetic acid for 12 hr at pH 9, but found that only four of the cyteine residues were derivatized. The carboxymethylation of protoxin described in the present work was allowed to proceed for only 5 min. It is likely that under prolonged reaction times other residues such as histidine and tyrosine are derivatized, resulting in inactivation of the protein. Arvidson et al. (1989) noted a significant alteration in host specificity toward Trichoplusia ni and Heliothis uirescens larvae of crystal protein from B. thuringiensis subsp. kurstuki HD-73, depending on whether the cysteine residues of the protoxin were oxidized or reduced. It is unclear why the oxidation state of these TABLE2 Toxicity to CF-1 Cells of Toxins Prepared Modified Parent derivative Native protoxin CM protoxin CAM protoxin AE protoxin SCM protoxin SCAM protoxin a Threshold

from Chemically

Protoxins Threshold

level (nginl)”

0.38-0.19 0.38-6.19 0.38-0.19 2.14-1.07 0.38-6.19 0.38-0.19

level lies between the given values

79

PROTOXIN

Toxicity

TABLE3 Protoxins to Spruce Budworm Larvae

of Derivatized

Derivative Crystal protein CM protoxin CAM protoxin AE protoxin SCM protoxin SCAM protoxin a Value determined

Lb” ( ug/larva) 31 51 26 51 76 47

95% confidence limits Lower 25 41 16 31 81 31

Upper 39 65 42 88 125 12

on 200 larvae.

residues should affect host specificity, since the toxic moiety of the protoxin molecule would apparently remain unaltered. However, this observation by Arvidson et al. suggests that the oxidation state of the sulfhydryl groups may affect processing of the protoxin in the insect gut. In summary, the present results show that extensive chemical modification of the sulfhydryl and amino groups of the protoxin from Bacillus thuringiensis subsp. kurstuki HD-73 has little effect on the overall structure or function of the molecule, although the solubility properties of the protein are greatly affected. As the protein crystal is essentially insoluble at neutral pH, and as there is evidence that dissolution of the crystal in the larval gut is a requirement for proteolysis to the active toxic (Tojo and Aizawa, 1983; Jaquet et al., 1987), the possibility arises that more soluble and potent protoxins can be developed by site-directed mutagenesis of the cysteine and lysine residues. ACKNOWLEDGMENTS This work was supported by the Natural Sciences and Engineering Research Council of Canada. The authors thank Dr. Kees van Frankenhuesen, Mr. Ross Milne, and Ms. Christine Badeau of the Forest Pest Management Institute, Sault Ste. Marie, Ontario, for the insect toxicity assays, and Dr. Luke Masson and Dr. Rolland Brousseau of the Biotechnology Research Institute, Montreal, Quebec, for the insect cell bioassays. REFERENCES Andrews, R. E., Faust, R. M., Wabiko, H., Raymond, K. C., and Bulla, L. A. 1987. The biotechnology of Bacillus thuringiensis. CRC Critical Rev. Biotechnol. 6, 163-232. Arvidson, H., Dunn, P. E., Strnad, S., and Aronson, A. I. 1989. Specificity of Bacillus thuringiensis for lepidopteran larvae: Factors involved in vivo and in the structure of a purified protoxin. Mol. Microbial. 3, 1533-1543. Bietlot, H., Carey, P. R., Choma, C. T., Kaplan, H., Lessard, T., and Pozsgay, M. 1989. Facile preparation and characterization of the toxin from Bacillus thuringiensis var. kurstaki. B&hem. J. 260, 87-91. Bietlot, H., Vishnubhatla, I., Carey, P. R., Pozsgay, M., and Kaplan, H. 1990. Characterization of the cysteine residues and disulphide linkages in the protein crystal of Bacillus thuringiensis subsp. kurstaki and entomocidus. Biochem. J. 267, 309315.

80

CHOMA

AND KAPLAN

Bulla, L. A., Kramer, K. L., and Davidson, L. I. 1977. Characterization of the entomocidal parasporal crystal of Bacillus thuringiensis. J. Bacterial. 130, 375-383. Carey, P. R., Fast, P., Kaplan, H., and Pozsgay, M. 1986. Molecular structure of the protein crystal from Bacillus thuringiensist A Raman spectroscopic study. Biochim. Biophys. Actu 872, 169176. Choma, C. T., and Kaplan, H. 1990. Folding and unfolding of the protoxin from Bacillus thuringiensis: Evidence that the toxic moiety is present in an active conformation. Biochemistry 29, 10,97110,977. Choma, C. T., Surewicz, W. K., Carey, P. R., Pozsgay, M., Raynor, T., and Kaplan, H. 1990. Unusual proteolysis of the protoxin and toxin from Bacillus thurinpiensis: Structural imnlications. Eur. J. Biochem. 189, 523-527. Ge, A. Z., Shivarova, N. I., and Dean, D. H. 1989. Location of the Bombyx mori specificity domain on a Bacillus thuringiensis deltaendotoxin. Proc. NatZ. Acad. Sci. USA 86, 40374041. Gringorten, J. L., Witt, D. P., Mime, R. E., van Frankenhuyzen, K., Fast, P. G., and Sohi, S. S. 1990. An in vitro system for testing Bacillus thuringiensis toxins: The lawn assay. J. Insect. Pathol. 56, 237-242. Haider, M. Z., and Ellar, D. J. 1987. Analysis of the molecular basis of insecticidal specificity of Bacillus thuringiensis delta-endotoxin. B&hem. J. 248, 197-201.

Hofmann, C., Vanderbruggen, H., H&e, H., Van Rie, J., Jansens, S., and Van Mellaert, H. 1988. Specificity of Bacillus thuringiensis delta-endotoxins is correlated with the presence of high-affinity binding sites in the brush border membrane of target insect midguts. Proc. NatZ. Acad. Sci. USA 85, 7844-7848. Hbfte, H., and Whiteley, Baciltus

thuringiensis.

H. R. 1989. Insecticidal MicrobioZ.

Rev.

crystal proteins of

53, 242-255.

Jaquet, F., Hiitter, R., and Liithy, P. 1987. Specificity of Bacillus thuringiensis delta-endotoxin. AppZ. Environ. Microbial. 53, 500504. Nickerson, K. W. 1980. Structure and function of the Bacillus thuringiensis protein crystal. Biotechnol. Bioenp. I 12. 1305-1335. Schesser, J. H., Kramer, K. J., and Bulla, L. A. 1977. Bioassay for homogeneous parasporal crystal of BaeiZZus thuringiensis using the tobacco hornworm Munduca se&a. AppZ. Environ. Microbial. 33, 787-880. Schwartz, W. E., Smith, P. K., and Royer, G. P. 1980. N-(6IodoethyB-trifluoroacetamide: A new reagent for the aminoethylation of thiol groups in proteins. Anal. Biochem. 106, 4348. Tojo, A., and Aizawa, cillus

thuringiensis silkworm Bombyx

K. 1983. Dissolution and degradation of Badelta-endotoxin by gut juice protease of the mori.

AppZ.

Environ.

MicrobioZ.

45, 57G580.