[Agr. Biol. Chem., Vol. 30, No. 12, p. 1261 `1268, 1966] Studies on Bacterial Protease Part XVI. Purification, Crystallization and Some Enzymatic Properties of Alkaline Protease of Bacillus subtilis var. amylosacchariticus By Daisuke TSURU, Heizo KIRA, Takehiko YAMAMOTO and Juichiro FUKIIMOTO Faculty of Science, Osaka City University, Sumiyoshi-ku, Osaka Received June 18, 1966 A crystalline alkaline protease was prepared from B. amylosacchariticus, which was isolated as a strain of saccharogenic ƒ -amylase-producing Bacillus subtilis. The enzyme was most active at ph values between 10.3 and 10.7 towards casein and was stable at ph values from 6 to 11 on twenty hour incubation at 30 Ž. Calcium ions were effective to stabilize the enzyme especially at higher temperatures. The enzyme was markedly inactivated by DFP as well as protease inhibitor from potato and slightly by surface active agents, but not affected by sulfhydryl reagents and divalent metal ions except Hg++. Hemoglobin was the best substrate for the enzyme and more than 20% of the peptide bonds were hydrolyzed. Of numerous synthetic peptides tested, only the two compounds, glycyl-l-prolyl-l-alanine and Cbz-L-tyrosyl-glycinamide, were found to be hydrolyzed. A cyclic peptide, gramicidin S, was split by the enzyme only at the peptide bond of -L-valyl-L-ornithyl-. Methyl n-butyrate and tributyrin were also good substrates for the alkaline protease obtained here. INTRODUCTION In previous papers, the authors reported the purification procedures1) of the neutral protease of B. amylosacchariticus that was isolated as a saccharogenic a-amylase2,3) producing Bacillus subtilis, and characterized some enzymatic1) and the physicochemical properties,4) compar ing with those of B. subtilis neutral proteases Abbreviation: DFP, diisopropyl fluorophosphate; EDTA, ethylenediamine-tetraacetic acid; SLS, sodium lauryl sulfate; BKCI, benzalkonium chloride; PCMB, p-chloromercuri benzoate; MIA, monoiodoacetic acid. 1) D. Tsuru, T. Yamamoto and J. Fukumoto, This journal, 30, 651, 856 (1966). 2) J. Fukumoto, T. Yamamoto and K. Ichikawa, Proc. Japan Acad., 27, 352 (1951). 3) J. Fukumoto and S. Okada, J. Ferm. Tech., 41, 427 (1963). 4) D. Tsuru, H. Kira, T. Yamamoto and J. Fukumoto, This journal, 30, 1164 (1966). so far reported.5 `9) During the purification procedure of the neutral protease, an alkaline protease was also obtained in a crystalline form of needle. Ottesen and his colleagues10 `12) have isolated two kinds of B. subtilis alkaline proteases in crystalline forms and named subtilopeptidase A and B (previously called the latter novo enzyme), respectively. 5) J. Fukumoto and H. Negoro, Proc. Japan Acad., 27, 441 (1951). 6) J. Fukumoto, T. Yamamoto and K. Ichikawa, J. Agr. Chem. Soc. Japan, 32, 230, 233 (1958). 7) D. Tsuru, J. D. McConn and K. T. Yasunobu, Biochem. Biophys. Res. Comm., 15, 367 (1964). 8) J. D. McConn, D. Tsuru and K. T. Yasunobu, J. Biol. Chem., 239, 3706 (1964). 9) D. Tsuru, J. D. McConn and K. T. Yasunobu, ibid., 240, 2415 (1965). 10) A. V. Guntelburg and M. Ottesen, Compt. Rend. Trav. Lab. Carlsberg, 29, 36 (1954). 11) M. Ottesen and A. Spector, ibid., 33, 63 (1960). 12) G. Johansen and M. Ottesen, ibid., 34 199 (1964).
1262 Daisuke TSURU, Heizo KIRA, Takehiko YAMAMOTO and Juichiro FUKUMOTO Recently, Johansen and Ottesen12) have shown that amino acid composition of sub. tilopeptidase B is in good agreement with that of the alkaline protease, BPN' (nagarse),13,14) prepared from the culture broth of liquefying ƒ -amylase-producing Bacillus subtilis. Fukumoto et al., on the other hand, have reported the purification procedure and some enzymatic properties of the alkaline protease obtained from a strain of B. subtilis that it deficient in ƒ -amylase-producing ability.15) Thus, it is of interest, from the genetical and taxonomic points of view, to ascertain whether the alkaline protease of B. amylosacchariticus is distinct from the other B. subtilis alkaline was measured according to the method described by Matsubara et al.17) The initial rate of hydrolysis of various proteins by enzymes was measured by the standard assay method and also by formol titration method.17) B. subtilis neutral protease I and II1,4) were purified as described by McConn et al.8) and the authors,1) respectively, and chymotrypsin and pronase were generous gifts from Dr. F. Itho of Department of Internal Medicine, Osaka University School of Medicine. A commercial crystalline nagarse was also used for the comparison of enzymatic properties with those of the alkaline protease presented here. A cyclic peptide antibiotic, gramicidin S, was kindly supplied by Dr. S. Otani of Department of Biochemistry, Osaka City University School of Medicine. proteases so far reported.10 `15) RESULTS AND DISCUSSION The present paper describes the purification I. Purification and Crystallization of Alkaline procedure and some enzymatic properties of Protease alkaline protease of B. amylosacchariticus, comparing with those of proteases obtained from The purification was carried out as described for the neutral protease1) of this strain (neutral the other strains of B. subtilis. protease II) with a slight modification. The MATERIALS AND METHODS commercial powder of the protease preparation was dissolved in 10 mm calcium acetate, fol Protease activity was assayed as described pre viously,1) except that ph 10.5 was chosen for the lowed by the ammonium sulfate precipitation enzyme reaction, and the specific activity was defined and then subjected to sequential chromato as the protease activity units per mg protein. Protein graphies on Duolite A-2 and CM-cellulose concentration was determined spectrophotometrically columns according to the method described at 280mƒÊ using the factor E_??_=11.9 that was previously.1) The protease adsorbed onto the estimated by the dry weight measurement of the CM-cellulose was eluted with a buffer con crystalline preparation. taining 0.3N sodium chloride and precipitated Moving boundary electrophoresis was kindly operated by the addition of two volumes of acetone. by Dr. Y. Tsujisaka of Department of Applied Bio The precipitate was collected by a centrifuge chemistry, the Osaka Municipal Technical Research, dissolved in a small volume of 10mM Tris Institute, with a Hitachi Tiselius Electrophoresis HCl-2mM calcium acetate buffer solution Apparatus HT-B. Ultracentrifugal analysis was kindly, ph 8.0, and dialyzed against the same buffer performed by Dr. K. Kakiuchi of Institute for Protein Research, Osaka University, using a Hitachi Analyti overnight. During the dialysis, almost all of cal Centrifuge Type UCA-1. alkaline protease were precipitated out as Peptide bond-cleaving activity of the enzyme on needle crystals leaving the neutral protease several proteins was measured by ninhydrin method16) in the mother liquid (Fig. 1-A). The crystals and cleavage of synthetic peptides by the enzyme was were collected by a centrifuge, washed twice checked by paper chromatography.4) Esterase activity with the buffer mentioned above and dissolved 13) B. Hagihara et al., J. Biochem., 45, 185 (1958). by the dropwise addition of 1 N ammonium 14) B. Hagihara, "The Enzyme," Vol. 4, Academic Press Inc., New York, 1958, p. 193. hydroxide solution. After neutralization to 15) J. Fukumoto T. Yamamoto and K. Ichikawa, J. Agr. Chem. Soc. Japan, 33, 9 (1959). ph 8.5 with 0.5 N acetic acid, the solution 16) E. W. Yemm and E. C. Cocking, Analyst, 80, 209 (1955). 17) H. Matsubara et al., j Biochem., 45, 251 (1958).
Studies on Bacterial Protease. Part XVI 1263 neutral protease and the other impurities (Fig. 2). To the effluent were added three volumes of acetone and the mixture was allowed to stand for three hours. The result ing precipitate was collected by a centrifuge, dissolved in a minimum quantity of 10mM Tris HCl-2mM calcium acetate buffer solution, FIG. 1. Micrographs of Crystalline Alkaline Protease. A: Rapidly crystallized at ph 8.0. B: Recrystallized at ph 8.5. was passed through a DEAE sephadex A-50 column equilibrated with the same buffer as described above to adsorb the contaminating FIG. 2. Chromatography of Alkaline Protease Purified by CM-Cellulose on DEAE Sephadex A-50 Column. About 1 gram protein was loaded on column (3.5 ~70cm) equilibrated with 10mM Tris-HCl buffer containing 2mM calcium ph and chromatographed at acetate, 8.0, was a flow of 35ml/hr. The mixing chamber 1.51 rate contained of buffer mentioned above and the reservoir consisted the of 1.51 of 50mM Tris-HCl-2mM calcium acetate buffer, ph 7.2, with 0.5N NaCl. \ absorbancy protease activity : ---: TABLE I. PURIFICATION OF ALKALINE PROTEASE * Decrease in specific activity was due to the removal of the neutral protease that was 5 times more active than the alkaline protease per mg protein. The mother liquid of the first crystals of the alkaline protease and the washings were combined and used for the preparation of the neutral protease.
1264 Daisuke TSURU, Heizo KIRA, Takehiko YAMAMOTO and Juichiro FUKUMOTO FIG. 3. Sedimentation Pattern of Alkaline Protease. The sedimentation run was operated at 55,100r.p.m. at 18.4 Ž. The enzyme con centration was in 10mM Tris-HC1 buffer solution, ph 9.2, where sodium chloride 0.68% was added to adjust the ionic strength to The photograghs were from 0.1. taken right of to left B 7, 33, 53 and minutes after reaching full speed. A of at 18, 43, 60 ph 8.5, and was dialyzed against the same buffer for two days, during which the alkaline protease was crystallized out as shown in Fig. 1-B. After washing twice with 1mm calcium acetate solution, ph 7.5, the enzyme was freeze-dried without any appreciable loss in the activity. All of the procedures were carried out in the cold. The flow sheet of purification is summarized in Table I. To check the purity of the alkaline protease, the freeze-dried preparation was dissolved in 1mM Tris maleate-2mm calcium acetate buffer, ph 6.4, and applied to CM-cellulose and SE sephadex C-50 column chromatographies. A single protein peak of the same specific ac tivity was always observed on a gradient elu tion with sodium chloride. Ultracentrifugal and electrophoretic analyses also confirmed that the preparation was monodisperse (Fig. 3 and 4). Both experi ments were carried out at low enzyme con centrations, since a part of enzyme was precipitated down as crystals during dialysis owing to the low solubility at ph valuses be tween 6 and 9 in the cold. FIG. 4. Electrophoretic Pattern of Alkaline Protease. The electrophoresis carried out at 100 volts, 9.5mA was 18.2 Ž. The concentration was 0.36% 25m and enzyme in Tris HCl-2mM calcium acetate solution, ph M buffer 7.4, containing chloride which was added adjust the sodium to ionic to 0.1. strength
Studies on Bacterial Protease. Part XVI 1265 II. Enzymatic Properties Effect of ph on activity and stability of enzyme. Fig. 5-A illustrates the effect of ph on the activity of the alkaline protease, indicating that the enzyme is most active at ph values between 10.3 and 10.7 towards casein. In Fig. 5-B, ph-stability relationship is shown. After twenty hour incubation at 30 Ž about 90% of the original activity were found to remain at ph values from 5.5 to 10.5, if the enzyme concentrations were low. FIG. 6. Effect of Temperature on Enzyme Activity. The reaction was carried out at the temperatures indi cated and at ph 10.5 for 10 minutes. FIG. 5. Effect of ph on Activity and Stability of Alkaline Protease. A: ph-activity curve; The ph of casein substrate was adjusted with m/50 Veronal (ph 5.9 `9.0) or Borate (9.4 `12.0) buffer. The other conditions were the same as those the standard assay method. of PH-Stability curve; used were B: Buffers 10-2M acetate (ph 4.0 `6.0), phosphate 5.5 `8.0) Borate (ph (ph and 8.7 `12.0). After incubation at 30 Ž for 20hrs with buffers various ph values, remaining activity of the of the mixture assayed using 1ml aliquot by was the standard method. FIG. 7. Effect of Temperature on Stability of Enzyme. The enzyme was dissolved in 10mM Tris-HCl buffer solution, ph 7.4, and incubated at the temperatures indi cated for 15 minutes. Temperature optimum of the enzyme reaction and thermal stability of the enzyme. The en zyme reaction was carried out at various tem peratures and ph 10.5 for ten minutes. The temperature optimum was at 55 Ž but, in the presence of 2mM calcium acetate, the optimum was shifted to 62 Ž, being accompanied by about 70% increase in the reaction rate (Fig. 6). As shown in Fig. 7, the enzyme lost its activity after fifteen minute incubation at 56 Ž and ph 7.4. In the presence of 5mM calcium acetate, however, about 50% of ac tivity were found to remain under the same condition. Effect of various protease inhibitors, surface active agents and sulfhydryl reagents. The enzyme was markedly inactivated by DFP as well as protease inhibitor prepared from potato, but slightly by surface active agents such as SLS and BKCl as shown in Table II. EDTA
1266 Daisuke TSURU, Heizo KIRA, Takehiko YAMAMOTO and Juichiro FUKUMOTO TABLE II. EFFECT OF VARIOUS REAGENTS ON ENZYME ACTIVITY The enzyme was incubated with the reagents in the presence of 20mM Tris-HCl buffer, ph 7.4, at 25 Ž for 30 minutes. TABLE III. COMPARATIVE PROTEOLYTIC ACTIVITY OF VARIOUS ENZYMES TOWARD CABIN The enzyme reaction was carried out at optimum ph for each protease for 10 minutes at 30 Ž. The protein concentration of chymotrypsin and pronase was determined spectrophotometrically assuming that E1%1cm at 280mƒÊ was 20 and 11, respectively. TABLE IV. RELATIVE INITIAL RATE OF HYDROLY SIS OF VARIOUS PROTEINS BY B. amylosacchariticus PROTEASES * A partially purified preparation. ** Results of 1 hour incubation at 25 Ž. Abbreviation: DFP, diisopropyl fluorophosphate; EDTA, ethylenediamine-tetraacetic SLS, acid; so dium lauryl sulfate; benzalkonium BKCI, chloride; PCMB, p-chloromercuribenzoate; MIA, monoiodoacetic acid. was also slightly inhibitory for the enzyme at ph values higher than 7 and sulfhydryl re agents showed no inhibitory effect on the enzyme. Among the various divalent metal ions tested, only Hg++ was shown to inactivate the enzyme. Comparative activity. The specific activity of the alkaline protease obtained here was estimated to be 2400 on casein substrate at ph 10.5. In Table III, the specific activity of purified proteases obtained from various sources is compared in their initial rates of hydrolysis on casein. III. Hydrolysis of Various Proteins, Peptides and Fatty Acid Esters by Alkaline Protease In Table IV, the initial rates of hydrolysis of various proteins catalyzed by the alkaline protease are summarized. Hemoglobin was most rapidly hydrolyzed by the enzyme. Degree of peptide bond cleavage of various TABLE V. PEPTIDE BOND HYDROLYSIS DEGREE OF VARIOUS PROTEINS BY B. subtilis PROTEASES The enzyme reaction mixture consisted of 50 mg of protein, 1mg of enzyme and 20mM Tris-HCl buffer, ph 10.0, in total volume of 10ml. After incubation at 37 Ž for 40 hours, aliquots were withdrawn and subjected to ninhydrin reaction. Hydrolysis per cent was calculated by comparing the ninhydrin value of the test sample with that of acid hydrolyzate of proteins. * Examined at ph 7.3.
Studies on Bacterial Protease. Part XVI 1267 TABLE VI. HYDROLYSIS OF PEPTIDES, AMINO ACID- AND FATTY ACID ESTERS BY B. subtilis PROTEASES The esterase activity was assayed at ph 8.5 and 30 Ž. After incubation for 20 or 60 minutes, an equal volume of 40% formol solution (ph 8.0) was added to the mixture and the carboxyl, liberated was titrated electrometrically to ph 8.3 with 10-2M NaOH solution. * Tested at ph 7.5. protein was also checked by ninhydrin method (Table V), and more than 20% of peptide bond of hemoglobin was shown to be split on a long incubation with the enzyme. On the other hand, egg-white albumin was hy drolyzed very slowly by the alkaline protease but the hydrolysis degree was found to increase with time, and on a long incubation the final per cent of the hydrolysis reached a value of 14%, that was greater than that of milk casein. The peptidolytic activity of the alkaline protease against peptides was also examined by paper chromatography (Table VI). Among 43 peptides tested, only the three compounds, gramicidin S*, glycyl-l-prolyl-l-alanine and Cbz-L-tyrosyl-glycinamide were hydrolyzed. Towards the former two compounds, B. subtilis neutral proteases were completely inert. On the other hand, several amino acid- and fatty * A cyclic peptide antibiotic: The alkaline protease splits only one peptide bond, -L-valyl-L-ornithyl-, of the peptide, as has been reported by Yukioka et al.18) on alkaline protease, BPN'. 18) M. Yukioka, Y. Saito and S. Otani, J. Biochem., 60, 295 (1966). acid esters, especially butyrate compounds, were good substrate for the alkaline protease while the neutral protease of this strain was completely inert on these esters.4,8) In conclusion little difference was observed in the enzymatic properties between the alka line protease presented here and B. subtilis alkaline proteases so far reported. There may need, however, further investigation on the physical and protein chemical natures to ascertain whether or not the alkaline protease of B. amylosacchariticus is distinct from that of the other strains of B. subtilis. Preliminary analysis of amino acid composition and mo lecular weight measurement suggested the diversity of alkaline proteases of B. subtilis in their protein chemical properties. The experi mental results and the enzyme specificity on polypeptides will be reported elsewhere in detail. Acknowledgement. The authors wish to ex press their gratitude to Dr. S. Otani of Osaka City University School of Medicine and Dr.
1268 Daisuke TSURU, Heizo KIRA, Takehiko YAMAMOTO and Juichiro FUKUMOTO F. Itho of Osaka University School of Medicine for their generous supply of gramicidin S and some enzyme preparations used in this experi ment. Thanks are also due to Dr. K. Kakiuchi of Institute for Protein Research, Osaka Uni versity, and to Dr. Y. Tsujisaka of Depart ment of Applied Biochemistry, the Osaka Municipal Technical Research Institute, for their courtesy on ultracentrifugal and elec trophoretic analyses.