Agric. Biol. Chem., 53 (12), 3251 -3256, 1989 3251

Purification and Characterization of a Thermostable from Bacillus thuringiensis Akane Kunitate, Masaji Okamotoand Iwao Ohmori Biochemical Research Department, Toagosei Chemical Industry Co., Ltd., Minato-ku, Nagoya 455, Japan Received June 21, 1989

Bacillus thuringiensis var. kurstaki HD-255was found to produce an extracellular, thermostable protease after the end of the vegetative growth phase. The purified had a molecular weight of 34,000 according to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and an isoelectric point of9.0. Its proteolytic activity was inhibited by an active-site inhibitor of serine protease, phenylmethyl- sulfonyl fluoride, and also by an SH-modifying reagent, /?-chloromercuribenzoic acid, suggesting that the enzymeis one of a subfamily of SH-containing serine proteases. The enzyme showed maximal proteolytic activity at 70°C and pH 8.5~9. The most interesting characteristic was its thermostability; it retained 88.4 %of its initial activity at pH 8.7 and 60°C after more than 7hr incubation in the presence of 2mMCaCl2.

Bacteria in the genus Bacilli generally pro- because it was inhibited by p- duce large amounts of extracellular proteases. chloromercuribenzoic acid and had higher Of them, alkaline proteases represented by thermostability. subtilisins have well been characterized, such The purification and properties of serine as Carlsberg produced by B. lichenif- proteases from B. thuringiensis var. galleriae, ormis^3) subtilisin BPN' by B. amyloli- var. finitirnus, and var. israelensis have been quefaciens,1 ~3) subtilisin amylosacchariticus reported by Epremyan et al.6) and Chestukhina by B. subtilis var. amylosacchariticusA) and et al.,8) but there has not been any report those from B. cereus,5) B. purnilus,l) and B. on those of var. kurstaki. This protease dif- thuringiensis.6 ~8) These have similar fers clearly from the other proteases men- properties: they are serine proteases with mo- tioned above in the point of thermostability. lecular weights of 25,000- 30,000 and optimal In this paper, we report on the purification pHs of9-ll, and they are stabilized at high and characterization of the thermostable pro- temperature by Ca2+.1 ~3'9) tease produced by B. thuringiensis var. kurstaki Bacillus thuringiensis forms (5-endotoxin, an HD-255, named protease TH. insecticidal protein toxic to larvae of Lepi- doptera or Diptera, which is produced dur- Materials and Methods ing sporulation. The insecticidal activities Materials. CM-cellulose CM52 was obtained from sometimes vary with fermentation or decrease Whatman.Sephadex G-75 super fine was obtained from during storage. As we suspected the involve- Pharmacia Fine Chemicals AB (Sweden). Protein Assay ment of proteases produced by B. thuringiensis Kit, carrier ampholyte (ByoLyte 3/10), and sodium do- in these phenomena, we examined the activity decyl sulfate-polyacrylamide gel electrophoresis (SDS- of protease and the ability of protease pro- PAGE)molecular markers were purchased from Bio-Rad duction by the bacterium. In our studies, we Laboratories (Richmond, Calif.). B. amyloliquefaciens al- kaline protease (subtilisin BPN), azocasein, phenyl- found that B. thuringiensis var. kurstaki HD- methylsulfonyl fluoride (PMSF), pepstatin, soybean tryp- 255 secreted a rather thermostable protease, sin inhibitor (STI), and molecular markers for gel fil- which was distinguishable from other Bacillus tration were products of Sigma. /7-Chloromercuribenzoic 3252 A. Kunitate, M. Okamoto and I. Ohmori

acid (PCMB). jVa-/?-tosyl-L-lysine chloromethyl ketone anode one was 0.1 MH3PO4. Protein was measured using (TLCK), and ethylene glycol-bis(£-aminoethyl ether)- the Bio-Rad Protein Assay Kit with bovine plasma gamma N,Ar,W,JV'-tetraacetic acid (EGTA) were from Wako globulin as a standard. The NH2-terminal sequence was Pure Chemical Industries, Ltd. All other chemicals and identified using an automated gas-phase protein sequencer reagents were of the best grade commercially available. of Applied Biosystems model 470Awhich was equipped with an on-line HPLCmodel 120A. Organismand culture conditions. Isolates were routinely cultivated on a modified 2xSG agar plate which con- tained 2 x SG,10) except for glucose. For fermentation, B. Results thuringiensis var. kurstaki HD-255 was cultured on this plate at 30°C overnight, a single colony was picked up and Fermentation and production ofprotease TH suspended in 50ml of modified 2 x SG medium and in- B. thuringiensis var. kurstaki HD-255 was cubated at 30°C for 10hr. This was used as the source of grown in modified 2xSG medium (Fig. 1). inoculum, and bacterial growth was monitored by measur- Protease activity was detected after 4~ 5 hr of ing the absorbance at 660nmwith a Klett-Summerson fermentation at the end of the logarithmic photoelectric colorimeter. growth phase and reached the maximumlevel Purification ofprotease TH. All procedures were done at 14~ 15hr after the start of fermentation. The 0~5°C. pH of the culture after 15hr of fermentation Step 1. (NH^SO^ precipitation. The culture super- was about 8.4. When0.1% glucose was added natant (300 ml) was obtained by centrifugation at 9000 x g to the medium, protease activity was not deV for 30 min. Protease activity in the supernatant, assayed by tected after 15hr of cultivation (data not the azocasein digestion method of Millet,1!) was 89,000 shown). units/1 and the total protease activity was precipitated by 50~70% saturation with (NH4)2SO4. The precipitated proteins were dissolved in buffer B (5mM potassium- Purification ofprotease TH phosphate buffer containing 2mMCaCl2, pH 6.0), con- centrated with PEG6000, and dialyzed against the same Protease THwas purified as described in buffer. Materials and Methods. The specific activity Step 2. CM-cellulose column chromatography. The di- of the purified protease THwas 12.28 units///g alysate was put on a column (2 x 34cm) of CM-cellulose by the azocasein digestion method. This was (CM52) pre-equilibrated with buffer B. The column was about 16.8-fold higher than that of the culture washed with 200ml of buffer B, and protease was eluted supernatant, with 29.3% recovery. The purify using a linear gradient of0 to 0.3m KC1in a total volume of200ml of buffer B. Samples of the effluent, 2.5 ml, were cation scheme is summarized in Table I. The collected at a flow rate of 25ml/hr and monitored for protein by measuring the absorbance 280nm. The major active fraction was eluted at about 0.23 m KC1. The active fraction was pooled, concentrated with PEG6000, and dialyzed against buffer B. Step 3. Sephadex G-75 superfine gel filtration. The dialysate was put on a column (2.5 x 75cm) of Sephadex G-75 super fine pre-equilibrated with buffer B. Samples of the effluent, 3.4ml, were collected at a flow rate of 10.3 ml/hr. The active fraction was pooled, concentrated with PEG6000, and dialyzed against buffer B.

Protease assay. Protease was assayed by the azocasein digestion method,11] except the reactions were done at 60°C. One unit of the enzyme activity was defined as the amount of enzyme which hydrolyzed 1 mg of azocasein in 15min at 60°C. Fig. 1. Growth Curve and Extracellular Protease Ac- tivity of B. thuringiensis var. kurstaki HD-255. Other procedures. SDS-PAGEswere done by the meth- Fermentation was done in 50ml of modified 2xSG od of Laemmli.12) Analytical isoelectric focusing was medium.Protease activity was measured as described in done using a 7.5% polyacrylamide slab gel containing 12% Materials and Methods. (#), growth curve; (O), protease BioLyte 3/10. The cathode buffer was 0.1 mNaOHand the activity. B. thuringiensis Thermostable Serine Protease 3253

Table I. Purification of Protease TH from B. thuringiensis var. kurstaki HD-255 The protease activity was measured by the azocasein digestion method as described in Materials and Methods.

Purificationstep Total. activity. , Total , protein. Specific, . activity, . Recovery,0/. Purificationtc... F (units) (mg) (units/mg) ( %) (fold)

Culture supernatant 26,900 36.85 7.30 x 102 100 1 Ammonium sulfate fractionation 25,600 1 5.61 1.64 x 103 95.2 2.25 CM-Cellulose 13,800 2.45 5.63 x 103 51.3 7.71 Sephadex G-75 super fine 7,870 0.64 1.23 x 104 29.3 16.82

Fig. 2. SDS-PAGEofFractions at Various Steps in the Purification of Protease TH. Gel electrophoresis was done by the method of Laemmli12) using a 15% polyacrylamide slab gel. The molecular Fig. 3. Effects of pH (A) and Temperature (B) on the weight of protease THwas calculated using phosphorylase Activity of Protease TH. B (92,500), bovine serum albumin (66,200), ovalbumin A: Optimal pH and pH stability of protease TH. The (45,000), carbonic anhydrase (31,000), soybean optimal pH ofprotease THwas identified by the azocasein inhibitor (21,500), and lysozyme (14,400) as molecular digestion method using 5 mMBritton-Robinson buffer, a weight standards. Lanes 1 and 6, molecular weight stan- wide area buffer, containing 2mMCaCl2 (pH 4~ 12). To dards; Lane 2, the culture supernatant; Lane 3, the frac- examine the pHstability, the residual activity was assayed tion precipitated with ammoniumsulfate; Lane 4, the ac- after incubation in the same buffer of different pHs at 25°C tive fraction after CM-cellulose chromatography; Lane for 24hr. 5, the active fraction after Sephadex G-75 chromatog- B: Optimal temperature and thermal stability of protease raphy. TH. To identify the optimal temperature, protease THwas dissolved in buffer A (10mMTris-HCl buffer containing 2mMCaCl2, pH 8.0) and assayed by the azocasein diges- purity of protease THwas 92%measured by tion method at 10~90°C for 15min. For the thermal densitometoric scanning of the SDS-PAGE stability, the residual activity was assayed after incubation pattern. at 10~90°C for 15min in buffer A. Properties ofprotease TH The molecular weight of protease THwas reagent, PCMB, an inhibitor of thiol pro- 34,000 (Fig. 2) and the isoelectric point was teases, but not by the inhibitors of metallo- or 9.0. Protease TH was almost completely in- carboxy-proteases (Table II). The maximal hibited by an active-site inhibitor of serine activity was observed at pH 8~9 like a sub- proteases, PMSF, and by an SH-modifying tilisin (Fig. 3A). Protease THwas stable over a 3254 A. Kunitate, M. Okamoto and I. Ohmori

Table II. Effects of Inhibitors on the Activity of Protease TH Protease TH was incubated with inhibitors for 40 min at 37°C. Residual activities were measured by the azoc- asein digestion method as described in Materials and Methods.

T ..,.. _ . Residual Inhibitor Concentration .. activity/O/A (%)

none (control) 100

P MSF "mM ° 0.3mM 0 EDTA 10 mM 87.7 EGTA 10 mM 102. 1 PCMB 2 mM ° eM 0.2 mM 2.8 Iodoacetate lO mM 1 5.9 STI l mg/ml 91.9 TLCK 10mM 107.8 Pepstatin 0. 1 mM 101.5 Fig. 5. SDS-PAGE Pattern of <5-Endotoxin Degraded by Protease TH. Nine jug of <5-endotoxin was incubated at 37°C for 1 hr in lOmM Tris-HCl buffer (pH 8.0). The weight ratio of enzyme to substrate was 1 : 10. Gel electrophoresis was done by the method of Laemmli12) using a 12%polyacry- lamide slab gel. The molecular weight of protease THwas estimated using the samemarkers described in Fig. 2. Lane 1, (S-endotoxin; Lane 2, <5-endotoxin digested by protease TH;Lane 3, molecular weight markers. broad pH range, retaining full activity after incubation at pH 7-ll for 24hr at 25°C. The optimal temperature of the enzyme was 70°C (Fig. 3B). This protease was very stable up to 60°C but the residual activity at 70°C was under 20% of the initial activity after 15 min of incubation. The thermostability of protease THwas much greater than that of subtilisin BPN/ (Fig. 4A), and 88.4% of the initial activity remained even after 7 hr of incubation at 60°C (Fig. 4B). To examine whether or not this enzyme digests <5-endotoxin, 9 /ig of <5-endotoxin was Fig. 4. Thermal Stability of Protease TH. incubated at 37°C for 1 hr in lOmMTris-HCl A: Comparison of thermal stability of protease THwith subtilisin BPN\Thermal inactivation kinetics was exam- buffer (pH 8.0). The weight ratio ofenzyme to ined by measuring the residual proteolytic activity after substrate was 1 : 10. (5-Endotoxin (130kD) was incubation at 60°C: (#), protease TH; (O), subtilisin partially degraded to less than 60kD (Fig. 5). bpn'. Thus, protease THcould digest <5-endotoxin B: Thermal stability of protease TH was measured by the under certain conditions. residual activity after incubation at 60°C for 7 hr in buffer A in the presence (#) and the absence (O) of2him Ca2+. NH2-terminal amino acid sequence The amino acid sequence of protease TH B. thuringiensis Thermostable Serine Protease 3255

Fig. 6. Comparison of NH2-Terminal Sequences. The amino acid sequences of B. thuringiensis var. israelensis, var. finitimus, var. galleriae, and B. cereus proteases are from Chestukhina et al.,8) that of thermitase is from Meloun et al.,l4) and those of subtilisin BPN' and subtilisin Carlsberg are from Markland et al.2) Gaps (-) were introduced to increase the homology, and identical amino acids are boxed. The numbering above the sequences corresponds to thermitase. å¡, not identified; -, gaps. wascomparedwith those of other extracellular ditions (Fig. 4A). It is interesting that protease serine proteases (Fig. 6), and was found to be TH is much more thermostable than sub- highly homologous with those of serine pro- tilisins, though both enzymes are products of teases from B. thuringiensis, reported by mesophiles. Chestukhina et al.8) The amino acid sequence Chestukhina et al.8) reported that the Non- of protease THwas identical to that of the terminal amino acid sequence of an extracel- israelensis protease at positions 6 to 14. As lular serine protease from B. thuringiensis var. compared with those of the galleriae and fi~ israelensis was homologous with that of ther- nitimus protease, the residues, 9Lys, 16Gln, and mitase, which is an extracellular serine pro- 18Val of the two enzymes were replaced by tease produced by a thermophile, Thermo- 9Asn, 16Gly, and 18Ile in protease TH. actinomyces vulgarisP] Protease TH is also homologous with thermitase in the NH2-ter- Discussion minal sequence and less homologous with subtilisins. Thermitase and proteinase K are The thermostable serine protease, protease classified into the super family of subtilisins,140 TH, produced extracellularly by B. thur- but they are distinguished from subtilisin ingiensis var. kurstaki HD-255 had a molecular BPN' and subtilisin Carlsberg by the presence weight of34,000 and an isoelectric point of9.0. of the Cys residue which is involved in their The enzyme was most active at pH 8~9 and activity. Since protease THis also inhibited by stable at pH 7~11 after 24hr incubation at an SH-modifying reagent, PCMB, this pro- 25°C (Fig. 3A). The enzyme activity was op- tease seems to be an enzymeof the thermitase timal at 70°C (Fig. 3B), and it was stabilized at subtype. high temperatures by Ca2+ (Fig. 4B). Such Epremyan et al. and Chestukhina et al. stabilization is seen with other Bacillus extra- previously reported on the purification and cellular serine proteases.1 ~3'9) In the presence of properties of the serine proteases from B. 2mMCaCl2, the proteolytic activity of pro- thuringiensis var. galleriae^8) var. finitimus 8) tease THretained 88.4% of its initial level after and var. israelensis.8) However, no report has 7hr of incubation at pH 8 and 60°C (Fig. 4B). yet appeared on the purification of B. thur- The thermostability of protease THwas much ingiensis var. kurstaki extracellular serine pro- higher than that of subtilisin BPN', whose teases. Protease TH showed homology in half-life was only lOmin under the same con- NH2-terminal amino acid sequence with these 3256 A. Kunitate, M. Okamoto and I. Ohmori proteases, especially with the israelensis pro- Acknowledgment. The authors wish to thank Professor tease. Considering the homology in their H. Kagamiyamaof Osaka Medical College for analyzing amino acid sequences, protease THis probably the NH2-terminal sequence of protease TH. subspeciesa counterpartdescribedof the proteasesabove.7'8) fromAlthoughthe threethe References three proteases from israelensis, galleriae, and 1) L. Keay, P. W. Moser and B. S. Wildi, Biotechnol finitimus showed a similar level of thermosta- Bioeng., 12, 213 (1970). 2) F. S. Markland, Jr. and E. L. Smith, "The enzyme," bility to subtilisin BPN',6'8) protease TH had Vol. 3, Academic Press Inc., New York, 1971, pp. much higher stability. These facts suggest that replacement of the amino acid(s) involved in 3) M. Ottesen561and I. Svendsen, -608.Methods Enzymol., 19, the thermostability may occur in protease TH. 199 (1971). Weare interested in the fact that a thermo- 4) D. Tsuru, H. Kira, T. Yamamoto and J. Fukumoto, Agric. Biol. Chem., 30, 1261 (1966). stable enzymeis produced by a mesophile. The 5) Y. Furukawa, Y. Fujii and H. Takahashi, Agric. growth of B. thuringiensis is maximal at 30°C Biol. Chem., 32, 822 (1968). to 37°C (data not shown). Although protease 6) A. S. Epremyan, G. G. Chestukhina, R. R. TH is produced by a mesophile, B. thur- Azizbekyan, E. M. Netyksa, G. N. Rudenskaya and ingiensis var. kurstaki HD-255, its thermosta- V. M. Stepanov, Biokhimiya, 46, 920 (1981). bility is comparable to that of thermolysin 7) V. M. Stepanov, G. G. Chestukhina, G. N. Rudenskaya, A. S. Epremyan, A. L. Osterman, O. produced by a thermophile B. stearother- M. Khodova and L. P. Belyanova, Biochem. Biophys. mophillus. In this respect, the fact that a meso- Res. Commun., 100, 1680 (1981). phile produces a thermostable protease may 8) G. G. Chestukhina, O. P. Zagnitko, L. P. Revina, F. be economically useful since it can be culti- S. Klepikova and V. M. Stepanov, Biokhimiya, 50, vated at 25°C to 37°C. Still more we are 1724 (1985). 9) F. G. Priest, Bacteriol. Rev., 41, 711 (1977). interested that protease THparticipates in the 10) T. J. Leighton and R. H. Doi, /. Biol. Chem., 246, degradation of <5-endotoxin. Protease TH may 3189 (1971). take part in the variation of the insecticidal ll) J. Millet, J. Appl. Bad., 33, 207 (1970). activities in someway. Further studies on the 12) U. K. Laemmli, Nature (London), 277, 680 (1970). primary structure of protease TH are prere- 13) R. Kleine, Ada Biol. Med. Germ., 41, 89 (1982). 14) B. Meloun, M. Baudys, V. Kostka, G. Hausdorf, C. quisite to clarify the cause of the high thermo- Frommel and W. E. Hohne, FEBS Lett., 183, 195 stability of the enzyme. (1985).