Eur. J. Biochem. 210, 1007-1014 (1992) 0FEBS 1992
Purification and characterization of proteinase In, a trypsin-like proteinase, in Escherichia coli Munetoshi KATO, Takaji IRISAWA, Makiko OHTANI and Mutumi MURAMATU Faculty of Pharmacy, Tokushima Bunri University, Japan (Received July 6iSeptember 7, 1992)
~
EJB 92 0946
We previously found a trypsin-like proteinase which momentarily appears immediately before DNA synthesis in the cell cycle of Escherichia coli synchronized by phosphate starvation and which is closely related to the initiation of DNA replication (Kato, M., Irisawa, T., Morimoto, Y. and Muramatu, M., unpublished results). The proteinase was named proteinase In. It was purified approximately 2880-fold with a recovery of 15%. The isolated enzyme appeared homogeneous by gel filtration and electrophoresis. Its molecular mass was estimated by analytical gel filtration and SDS/ PAGE as approximately 66 kDa. The isoelectric point of proteinase In is 4.9 and its optimal pH is approximately 9. Although protein In hydrolyzes fluorogenic substrate for trypsin, its hydrolytic activity seems markedly affected by amino-acid sequence lying towards the N-terminal from the PI (lysine, arginine) residue. The proteinase does not hydrolyze NZ-benzoyl-~,~-arginine-4-nitronanilide and fluorogenic substrates for chymotrypsin and elastase. The proteinase activity is inhibited by leupeptin, antipain and 4-nitrophenyl 4-guanidinobenzoate, but the effects of tosyl-r-lysine chloromethane, diisopropylfluorophosphate, benzamidine and pentamidine isethionate on the proteinase activity are weak or not inhibitory. Its activity is strongly affected in the presence of NaCl and KCl, and at a concentration of 1.5 M, these increase the activity 14-fold and 13-fold, respectively, above that without salt. Proteinase In was strongly inhibited by various esters of trans-Cguanidinomethylcyclohexanecarboxylic acid, and their inhibitory effects werc roughly correlated with those on growth of E. coli. Proteinase activity was found in the cytoplasmic fraction.
sible proteasomes) [13], and identity of protease Do and htrA-gene product [14]. Many workers reported the effects of various protease inhibitors on E. coli, such as suppression of the degradation of protein by various inhibitors [15],inhibition of biosyntheses of protein and RNA by various chloromethylketones and recovery of these inhibitions by reduced glutathione [16], inhibition of synthesis of protein I in membranes and degradation of a protein by tosyl-L-lysine chloromethane (TosLysCH2Cl) [17] and by antipain and leupeptin [IS]. We previously reported that various esters of trans-4guanidinomethylcyclohexanecarboxylic acid (GMCH A), synCorrespondence to M. Muramatu, Faculty of Pharmacy, Toku- thetic trypsin inhibitors, strongly inhibited the growth of E. shima Bunri University, Yamashiro-cho, Tokushima-shi, Tokushima, coli. These effects may be due to suppression of DNA repliJapan 770 cation, and a trypsin-like proteinase may possibly participate Abbreviations. GMCHA, trans-4-guanidinomethylcyclohexanecarboxylic acid; GMCHA-OPheBu, 4-ferf-butylphenyl ester of in DNA synthesis (Kato, M., Irisawa, T. and Muramatu, M., GMCHA; Boc, tert-butoxycarbonyl; Bz, N2-benzoyl; SUC,succinyl; unpublished results). Moreover, using E. coli synchronized Glt, glutaryl; Boc-Val-Pro-Arg-NH-Mec, 4-methylcoumaryl-7-amide by a phosphate-starvation method, the suppression effect of of tert-butoxycarbonyl-L-valyl-L-prolyl-L-arginine; NAn, 4-nitro- GMCHA esters on the growth of E. coli was found to be due anilide; TosLysCH,Cl, l-chloro-3-tosylamido-7-amino-2-heptanoneto inhibition of a trypsin-like proteinase which momentarily hydrochloride, tosyl-1.-lysine chloromethane; TosPheCHZC1,L-l-toappeared immediately before DNA synthesis in the cell cycle sylamide-2-phenylethyl chloromethane, tosyl-L-phenylalanine chloof E. coli. The proteinase may be closely related to the initiaromethane; NPGB, 4-nitrophenyl 4-guanidinobcnzoate: STI, soybean trypsin inhibitor; ICso,inhibitor concentration required for tion of DNA replication in synchronized E. coli cells. The proteinase was named proteinase In (Kato, M., Irisawa, T., 50% inhibition. Morimoto, Y. and Muramatu, M., unpublished results). Enzymes. Trypsin (EC 3.4.21.4); lysozyme (EC 3.2.1.17); alkaline phosphatase (EC 3.1.3.1): B-galactosidasc (EC 3.2.1.23); lactate deThis paper reports the purification of proteinase In and hydrogenase (EC 1.1.1.27). its properties are discussed.
Proteolytic processes play an important role in various functions of Escherichiu coli: degradation of normal and abnormal proteins [l, 21; maturation of secretory and membrane proteins [ 3 , 41; breakdown of colicins [5, 61; inactivation of certain regulatory proteins [7, 81. A comprehensive and pertinent review of these matters has been presented by Lazdunski [9]. Recent studies have reported, identity of-outer-membraneassociated protease VII and OmpT which specifically cleaves ferric enterobactin receptor protein [lo], in analysis of a novel outer-membrane-associated protease [l 11the induction of protease La under stress [12],high-molecular-mass proteases (pos-
1008 MATERIALS AND METHODS Materials Various fluorogenic substrates, N’-benzoyl-D,L-arginine4-nitroanilide (Bz-Arg-NAn), leupeptin, antipain and chymostatin were purchased from Peptide Institute Inc. l-Chloro-3tosylamido-7-amino-2-heptanonehydrochloride (tosyl-L-lysine chloromethane, TosLysCHzC1) and ~-1-tosylamide-2phenylethyl chloromethane (tosyl-L-phcnylalanine chloromethane, TosPheCH’Cl) were obtained from Nakarai Chemical Ltd. Diisopropylfluorophosphate, phenylmethylsulfonyl fluoride, 4-nitrophenyl 4-guanidinobenzoate (NPGB), soybean trypsin inhibitor (STI) and aprotinin were from Sigma Chemical Co. DEAE-cellulose (Brown) and egg-white lysozyme were from Seikagaku Kogyo Co. Benzamidine was from Tokyo Chemical Ind. L-Arginine-Sepharose 4B, phenyl-Sepharose CL-4B and Sephadex G-200 were from Pharmacia. Hydroxyapatite (Bio-GEL . HTP) was from Bio-Rad Laboratories. 4-Nitrophenylphosphate and 2-nitrophenyl-P-~-galactopyranoside were from Boehringer Mannheim. Various esters of trans-4-guanidinomethylcyclohexanecarboxylicacid (GMCHA) were obtained as hydrochlorides from Nippon Chcmiphar Co. Ltd.
E. coli and growth condition E. coli strain K-12 IAM1264 was kindly provided by Dr. M. Takahashi, Godo Shusei Co. Ltd., Toyko. E. coli cells were grown in a glucose minimal medium [19] at 37°C. When absorbance at 600 nm reached 0.55 (approximatcly 6 x 10’ cells/ml), the cells were harvested by centrifugation at 7000g for 20 min at 4°C. From l l culture medium approximately 1 g cells was obtained. The cells (1 g) were resuspended in 10 ml0.1 M sodium borate, pH 8.0, containing 0.1 5 M NaCl and frozen at 20°C.
albumin, 45 kDa; chymotrypsinogen, 25 kDa; cytochrome c, 12.5 kDa. Electrophoresis and proteinase activity For electrophoresis the Phast System and Phast Gel media (Pharmacia) were used for IEF pH 3 - 9. native PAGE and SDS/PAGE. The Phast System was also used for automated Coomassie-blue staining. All procedures were conducted according to the manufacturer’s recommendations. Detection of proteinase activity in the gel was performed according to Fujimoto et al. [24] with minor modification. After electrophoresis, the gel was immersed in 15 ml 100 1 M Boc-Val-Pro-Arg-NH-Mec dissolved in 0.1 M sodium borate, pH 8.0, containing 1 mM CaCI2 and 0.08 M NaCl. It was kept in the dark at 37°C. After 3 h, the gel was irradiated at 365 nm with Ultra Violet Lamp VL-6LC (Vilber Lourmat, France). Proteinase activity was detected in the gel as a fluorescent band. Buffer for measuring the pH/activity profile The pH/activity profile of purified proteinase was examined with the following buffers. Britton and Robinson’s universal buffer: a mixture of 50 mM citric acid, KH2P04,boric acid and sodium diethylbarbiturate was adjustcd to a L.w e n pH with 3.5 M NaOH. Borate buffer for measuring activity in the pH range 7.5 - 9.0 the pH of the buffer was adjusted by mixing 60mM boric acid and sodium borate. For the pH range 9.5- 10.0 the pH was adjusted by mixing 60 mM boric acid and 3.5 M NaOH. Inhibition of proteinase activity
Mixtures of 1.0 m10.05 M sodium borate, pH 9.0, containing various amounts of an inhibitor and proteinase were incubated with 0.5 ml Boc-Val-Pro-Arg-NH-Mec dissolved in the Assay of proteinase activity same buffer, and proteinase activity was determined as above. When thc high-molecular-mass inhibitors, STI and aprotinin, The hydrolytic activity of proteinase on 4-methyl- were used, the proteinasc was incubated with various amounts coumaryl-7-amide of tert-butoxycarbonyl-L-valyl-L-prolyl-Lof the inhibitor in 1 ml of the above buffer at 25 C. After arginine (Boc-Val-Pro-Arg-NH-Mec) was determined in 10 min, 0.5 ml substrate solution was added and incubation 0.1 M sodium borate, pH 8.0, containing 1 mM CaC12 and was continued at 37°C for 1 h. The final substrate concen0.08 M NaC1. Mixtures of 0.5 ml of the above buffer containtration was 20 pM. ing various amounts of proteinase and 1 ml substrate solution were incubated at 37’C. After 1 h, 1 ml 30% acetic acid was added and 7-amino-4-methylcoumarin released was deter- Cellular localization mined as described by Kanaoka et al. [20]. Final substrate Spheroplasts were prepared according to the method of concentration was 6.67 pM. Yamato et al. [25]. Those obtained from approximately 1.6 g Bz-Arg-NAn hydrolytic activity was measured at 410 nm E. coli cells were lysed by osmotic shock as described by in the above buffer as described by Erlanger et al. [21]. Osborn et al. [26]. The subcellular fractions obtained were ~
Protein assay Protein concentration was determined by the method of Lowry [22] using bovine serum albumin as the standard.
dialyzed against 0.1 M sodium borate, pH 8.0, containing 1 mM CaCI, and 0.08 M NaCI. The fractions were monitored by measuring p-gdlactosidase [27]. alkaline phosphatase [28] and L-lactate dehydrogenasc which was assayed by diaphorase using lactate dehydrogenase C II-Test Wako (Wako Pure Chemical Industries Ltd.).
Determination of molecular mass by gel filtration The molecular mass of the purified proteinase was determined by gel filtration on a Sephadex (3-200 [23]. The column (2.5 cm x 101 cm) was equilibrated and eluted with 0.05 M sodium borate, pH 8.0, containing 1 mM CaCI, and 0.15 M NaCl, at a flow rate of 15 ml/h. Protein standards were y-globulin, 158 kDa; bovinc serum albumin, 66 kDa; ov-
RESULTS Purification of proteinase In Unless otherwise stated, all procedures were conducted at 4°C using 0.1 M sodium borate, pH 8.0, containing 1 mM CaCI, as the standard buffer (buffer A) for purifying protein-
1009 ase In. Proteinase activity was assayed with the fluorogenic substrate Boc-Val-Pro-Arg-NH-Mec. Preparation of crude extract Approximately 100 g E . coli K-12 cells suspended in 1 1 0.1 M sodium borate, pH 8.0, containing 0.15 M NaCl, frozen at - 20 'C, were allowed to thaw and each 50 ml of suspension was sonicated six times at 100 W for 30 s in an ice-bath. The homogenate was centrifuged at 13000 x g for 30 min. Thc precipitate was washed twice with 500 ml buffer A by sonication and centriruging as described above. The supernatant fluid was collected and designated the crude extract. DEAE-cellulose column chromatography (step 1) The crude extract (2 1) was dialyzed against buffer A containing 0.08 M NaCI. The dialysate was applied onto a DEAEcellulose column (3.2 cm x 30 cmj equilibrated with the above buffer and washed with 700ml of the same buffer. The adsorbed materials were eluted with 1.2 1 of a linear gradient of 0.08 M to 0.5 M NaCl in buffer A. 8.7-ml fractions were collected at a flow rate of 36ml/h. Protein and proteinase activity of the effluent were assayed and the activity fractions (fraction numbers 52 - 85) were pooled. Total activity of proteinase In then increased approximately 237% above that of the crude extract, possibly as a result of removing endogcnous protease inhibitors or competing substances.
0.'
0 10
z
50
0
100
Fraction number (4.2 mlltube)
Fig. 1. Second gel filtration of proteinase In on Sephadex G-200. Proteins obtained from L-arginine-Sepharose 4B (pH 8.2) through the column. Details of the procedure are indicated in the text (step 8). Proteinase-In activity was assaycd with Boc-Val-Pro-Arg-NH-Mec as descrtbcd. Thc bar indicates the fraction pooled. (0 El), Protein.), absorbance at 750 nm. ase activity; -.( ~
-
Ammonium-sulfate fractionation (step 2) The active fraction (280 mlj obtained from DEAE-cellulose column Chromatography was precipitated by 45% saturation with ammonium sulfate. After 2 11, the suspension was centrifuged at 13000 ,q for 30 min. The pellet thus obtained was discarded. To the supernatant fluid, solid ammonium sulfate was added to 80% saturation and the mixture was allowed to settle for 4 h. The precipitates were collected by centrifugation as described above. Sephadex G-200 gel filtration (step 3) The precipitates at 80% ammonium-sulfate saturation were dissolved in 16 ml 0.05 M sodium borate, pH 8.0, containing 1 mM CaCl, and 0.15 M NaCl and dialyzed against the same buffer. The dialysate (23 11) was separated into two parts, followed by gel filtration of each on a Sephadex G-200 column (2.5 cm x 101 cmj cquilibrated with the same buffer. The flow rate was 15 ml/h and 4.2-ml fractions were collected. The proteinase fraction (fraction numbers 63 - 81j was combined (146 ml). Phenyl-Sepharose CL-4B chromatography (step 4)
To the combincd fractions from the Sephadex G-200 column NaCl was added at 3 M NaCI, and the solution was applied to phenyl-Sepharose C L d B (2.2 cm x 13.5 cm) equilibrated with buffer A, pH 7.5, containing 3 M NaCl. The column was washed with 340ml of the same buffer and adsorbed materials were eluted with 1 1 of a linear gradient of 3.0 M to 0 M NaCl in buffer A, pH 7.5. 4.8-ml fractions were collccted at a flow rate of 15 ml/h. The active fractions (fraction numbers 161 -220) were combined (280 ml).
Hydroxyapatite chromatography (step 5) The active fraction from the above column chromatography was dialyzed against 10 mM potassium phosphate, pH 6.8, and applied onto a column of hydroxyapatite (1.8 cm x 10 cm) equilibrated with the same buffer. The column was washed with 50ml of the same buffer and adsorbed materials were eluted with 600 ml of a linear gradient of 10 mM to 150 mM potassium phosphate, pH 6.8. The flow rate was 18 ml/h and 4.1-ml fractions were collected. The active fractions (fraction numbers 27 - SO) were combined (90 ml). L-Arginine-Sepharose 4B chromatography (step 6) The active fraction from hydroxyapatite column chromatography was dialyzed against buffcr A and applied onto a column of L-arginine-Sepharose 4B (1.6 cm x 7 cm) equilibrated with the same buffer. The column was washed with 90 ml of the 5ame buffer. The adsorbed materials were eluted with 600 ml of a linear gradient of 0 M to 0.15 M NaCl in buffer A. 4.1-ml fractions were collectcd at a flow rate of 18 ml/h. The active fractions (fraction numbers 91 - 115) were combined (92 ml). Further chromatography on L-Arginine-Sepharose 4B column (step 7) The active fraction from the L-arginine-Sepharose 4B column was dialyzed against 0.1 M sodium borate, pH 8.2, containing 1 mM CaCI, and applied onto a column of Larginine Sepharose 4B (1 cm x 4.5cmj equilibrated with the
1010 Table 1. Purification of proteinase In from E. cofi. Proteinase In activity was examined with Boc-Val-Pro-Arg-NH-Mcc as the substrate.
Purification step
Activity
Protein
Specific activity
Yield
Purification
pmol/min 18333 43371 41 831 36933 18713 11446 5888 4089 2788
mg 4144 1482 660 316 44.1 8.7 0.90 0.41 0.22
pmol . min-' . mg-' 4.4 29 63 117 424 1314 6 542 9 973 12673
Yo
-fold
Crude cxtract DBA R-cellul ose Ammonium sulfatc (45- 80%) Sephadex G-200 Phenyl-Sepharose CL-4B Hydroxyapatite (Bio-GEL HTP) L-Arginine Sepharosc 4B at pH 8.0 L-Arginine Sepharose 4B at pH 8.2 2nd Sephadex G-200
100 237 228 202 102 62 32 22 15
1 7 14 27 96 299 1487 2267 2880
same buffer. The column was washed with 30 ml of the same buffer. The adsorbed materials were eluted with 300 ml of a linear gradient of 0 M to 0.15 M NaCl in the above buffer. 2.6-ml fractions were collected at a flow rate of 15 ml/h. The fractions with proteinase activity (fraction numbers 41 - 61) were combined (46 ml). Second gel filtration on Sephadex G-200 column (step 8) The active fraction from the above column chromatography was concentrated to 6 ml in an Amicon YM-5 membrane and dialyzed against 0.05 M sodium borate, pH 8.0, containing 1 mM CaC12 and 0.15 M NaCl. The dialysate was subjected to gel filtration on a Sephadex (3-200 column (2.5 cm x 101 cm) equilibrated with the same buffer. 4.2-ml fractions were collected at a flow rate of 15 ml/h (Fig. 1). Fractions with proteinase activity (fraction numbers 70 - 80) were combined (42 ml), dialyzed against 0.05 M sodium borate, pH 9.0, and stored at 0°C. A Summary of the purification is shown in Table 1. The yield of enzyme activity was 15% with 2880-fold purification. Criteria of homogeneity Rechromatography of proteinase In on Sephadex G-200 gave a single peak with constant specific activity (Fig. 1). From analytical gel electrophoresis, only a single band with non minor contaminants was detected by protein staining. The band showed hydrolytic activity for Boc-Val-Pro-ArgNH-Mec (Figs 2A, B). Molecular-mass determination and isoelectric point Prom the results of step 8 and method of Andrews 1231, molecular mass was calculated as 66 kDa (Fig. 3A). SDS/ PAGE gave an average value of 66 kDa (Fig. 3B). Proteinase In appeared to be composed of a single subunit. Proteinase In had an jsoelectric point of approximately 4.9 as determined on a Phast Gel IEF 3-9 (Fig. 2C). pH/activity profile The pH dependence of proteinase In was determined in Britton and Robinson's universal buffer and sodium borate buffer. A sharp activity profile was observed with maximal activity at pH 9.0 in both buffers (Fig. 4). Substrate specificity Hydrolytic activity of proteinase In on various fluorogenic substrates is shown in Table 2. Proteinase In clearly hy-
drolyzed trypsin-specific substrates, particularly Boc-ValLeu-Lys-NH-Mec and Boc-Val-Pro-Arg-NH-Mec. K, for Boc-Val-Pro-Arg-NH-Mec was 62.5 pM. However, hydrolytic activity on Boc-Leu-Gly-Arg-NH-Mec, Boc-Glu-Lys-LysNH-Mec, Boc-Phe-Ser-Arg-NH-Mec. Boc-Leu-Ser-Thr-ArgNH-Mec, Boc-Ile-Wu-Gly-Arg-NH-Mec, Glt-Gly-Arg-NHMec (Glt, glutaryl) and Bz-Arg-NH-Mcc was very low or not detectable. Arg-NH-Mec was not hydrolyzed. Proteinase In would thus appear to require hydrophobic amino acid sequence lying towards the N-terminal from the P1 (lysine, arginine) residue for maximal activity. Fluorogenic substrates for chymotrypsin and elastase were not hydrolyzed by proteinase In. No Bz-Arg-NAn hydrolytic activity was observed. Effects of monovalent salts Fig. 5 shows the effects of monovalent salts. Proteinase In activity increased with the concentrations of NaC1 and KC1, and at 1.5 M, activity reached 14-fold and 2 3-fold, respectively, above that without these salts. At higher salt concentrations, activity decreased. LiCl had a diphasic effect on activity as shown in Fig. 5. Effects of various inhibitors and bivalent cations The effects of proteasc inhibitors, EDTA and bivalent cations are shown in Table 3. Proteinasc In was strongly inhibited by leupeptin, antipain and NPGB. Chymostatin showed SO -60% inhibition at 200 pM. Aprotinin had a strong inhibitory effect, but STI had none. The effects of phenylmethylsulfonyl fluoride, TosPheCH2C1, diisopropylfluorophosphate, benzamidine and pentamidine isethionate were not inhibitory. TosLysCH,CI, EDTA and CaClz were considerably inhibitory and MnClz particularly so. The effect of MgC12 was slight. Effects of various GMCHA esters As described by Kato et al. (Kato, M., Irisawa, T., and Muramatu, M., unpublished results), various GMCHA esters were shown to strongly inhibit the growth of E. coli. Table 4 shows the inhibitory effects of GMCHA esters on proteinaseIn activity, growth of E. coli and Ki for trypsin. Thc order of inhibitory GMCHA ester effect on proteinase activity was roughly correlated with that on E. coli growth, but not with Ki for trypsin. GMCHA-OPheBu, a representative GMCHA ester, competitively inhibited proteinase In and its Ki was 41 pM.
1011
h
3-
-
i
i 0
I
I
I
10
20
30
40
Distance from cathode (mm)
Fig.2. Electrophoretic pattern of the purified enzyme in native PAGE (Phast Gel Gradient 8-25) and determination of its isoelectric point. (A) Preparation obtained from Sephadex G-200 stained with Coomassie blue. Details of the procedure are indicated in Materials and Methods. (B) Hydrolytic activity on Boc-Val-Pro-Arg-NH-Mec. Details of the procedure were given in Materials and Methods. (C) Determination of the isoelectric point. The isoelectric point of proteinase In was determined as described in Materials and Methods. Coomassie-stained proteinase In and marker proteins ( P I range 3 -9, Pharmacia) bands are shown on the left. The pH-gradient profile was obtained by the broad piCalibration Kit running on Phast Gel IEF 3 -9. The arrow is at thc value of proteinase In. The corresponding pIare: a, lentil lectin (basic), 8.65; b, lentil lectin (middle), 8.45; c, lentil lectin (acidic), 8.15; d, horse myoglobin (basic, 7.35; e, horse myoglobin (acidic), 6.85; f, human carbonic anhydrase R, 6.55; g, bovine carbonic anhydrase R, 5.85; h, /?-lactoglobulin A, 5.20; i, soybean trypsin inhibitor, 4.55; j, amyloglucosidasc, 3.50.
Elution volume (ml)
Relative mobility
Fig. 3. Molecular-mass determinationfor proteinase In. (A) Estimation of the molecular mass of proteinase In by gcl filtration on Sephadex G200. Conditions are indicated in Materials and Methods. The arrow is at the value of elution volume of the proteinase In. Marker proteins (Sigma) used were a, ?-globulin (bovine); b, bovine serum albumin; c, ovalbumin; d, chymotrypsinogen; e, cytochrome c. (B) Calibration of SDSjPAGE on Phast Gel Gradient 10- 15 with marker proteins (LMW kit E, Pharmacia). Protein molecular mass is plotted against rclativc mobility determined from bromophenol-blue migration. The arrow marks the position of proteinase In. The corresponding molecular masses are f, phosphorylase b, 94 kDa; g, albumin, 67 kDa; h, ovalbumin, 43 kDa; i, carbonic anhydrase, 30 kDa; j, trypsin inhibitor, 20.1 kDa; k, cc-lactalbumin, 14.4 kDa.
Cellular localization
DISCUSSION
Alkalinc phosphatase was found in the periplasmic fraction and proteinase In and jl-galactosidase, in the cytoplasmic fraction (Table 5 ) , for cells in the exponential phase.
Proteinase In was purified as described in this study and appeared to be homogeneous by the elution profile from gel filtration on Sephadex G-200 column and from electro-
1012
1 h
aQ Y
a c ._ > ._ c 0
m W
> .c
-6W a
6
I
I
I
I
7
8
9
10
11
PH
Fig. 4. Effects of pF1 on proteinase In. The aclivity of protcinase In was assaycd with 60 mM sodium borate (0)and Britton and Robinson’s universal buffer ( 0 ) Boc-Val-Pro-Arg-NH-Mec . was used as the substrate (final concentration of 20 pM). Values represent percentages of activity of protcinase In in 60 mM sodium borate, pH 9.0.
Table 2. Hydrolytic activity of proteinase In on various fluorogenic substrates. Hydrolytic activity was examined in 0.05 M sodium borate, pH 9.0 at 3 7 T , containing 20 FM of each substrate. Values represent percentages of activity of protcinase In on Boc-Val-Pro-Arg-NHMec. Boc, rert-butoxycarbonyl ; Glt, glutaryl; Bz, N2-benzoyl; Suc, succinyl; n.d., not detected. Fluorogenic substrate
Relative activity
Yo Boc-Val-Pro-Arg-NH-Mec Boc-Val-Leu-Lys-NH-Mec Boc-Leu-Gly-Arg-HN-Mec Boc-Glu-Lys-Lys-NH-Mec Boc-Phe-Ser-Arg-NH-Mcc Boc-Leu-Ser-Thr-Arg-NH-Mec Boc-Ile-Glu-Gly-Arg-NH-Mec Glt-Gly- Arg-NH-Mec Rz-Arg-NH-Mec Suc-Leu-Leu-Val-Tyr-NH-Mec SUC-Ah-Pro-Ala-NH-Mec Arg-NH-Mec
100 216.8 1.7 1.3 3.0 1.4 5.0 n. d. 6.3 n. d. n. d. n. d.
phoresis. Additional purity tests have not been made owing to the small amount of enzyme available. The present purification procedure consists of seven steps and is quite tedious, because crude extract obtained from 100 g wet cells contained much Arg-NH-Mec-hydrolytic activity (1.56 pmol/min), which was detected over the cell cycle of synchronized E. coli (unpublished results), and the aminopeptidase activity was quite difficult to separate from proteinIn activity. However, this aminopeptidase activity could be
0
1
Concentration of monovalent salts (MI
Fig. 5. Effects of monovalent salts on proteinase In. Effects of NaCI. KCI and LiCl on proteinase In were examined with Boc-Val-ProArg-NH-Mec in 0.05 M sodium borate, pH 9.0, at 37°C. The final substrate concentration was 20 pM. Each activity is shown as the activity (-fold) relative to NaCI, KCI and LiCI-free control. (0-O), NaCl; (0-0), KCI; (A-A), LiCl.
Table 3. Effects of various inhibitors and bivalent cations on proteinase In. Inhibitory effects on proteinase In were examined using Boc-ValPro-Arg-NH-Mec as the substrate. All inhibitors were previously incubated with proteinasc In at 25°C for 10 min, except for leupeptin. antipain, chymostatin and NPGB. Inhibitor
Inhibitory effects IC50
inhibition
Yo Leupeptin Antipain Chymostatin NPGB Aprotinin 100 Fg/ml STI 1 mM Phenylmethylsulfonyl fluoride 1 mM Benzamidine 1 mM TosLysCHzCl 1 mM TosPheCHzCl 1 mM Diisopropylfluorophosphate 1 mM EDTA 0.2 mM Pentamidine isethionate 1 mM CaCI, 5 mM CaCI2 1 mM MgClz 5 mM MgClz 1 mM MnCI, 5 mM MnClz
84 FM 14 pM 56 pM 102 pM 8.5 pg/ml 0 3.6 3.2 33.5 7.6 0.6 51.9 0 8.6 46.1 0 15.4 76.0 98.2
1013 Table 4. Inhibitory effects of various GMCHA esters on proteinase In, growth of E. coli and Kifor trypsin. Inhibitory effects on protcinase In were examined with Boc-Val-Pro-Arg-NH-Mec as a substrate. * n = 15, P < 0.01.
GMCHA ester
Inhibitory effect in E. coli on Proteinase In
Growth 1C5*
Trypsin Ki
21 26 36 42 43 44 45 64 68 74 87 92 105 135 142 153 167
35 54 29 45 45 273 38 103 325 187 56 46 51 113 85 56 48 78 110
IC50
4-Chloro-2-isopropyl-5-methylphenyl 20 4-Diphenyl 4-Benzyloxycarbonylethylphen yl 4-Bcnzyloxycarbonylphenyl 3-Benzylox ycarbon ylphenyl 2,4,6-TnchlorophenyI 4-tert-Butylphenyl 4-Phenyloxycarbonylphenyl 2-Isopropyl-5-methylphenyl 2-Diphenyl 2-Benqlox ycarbonylphen yl 2,4-Dichlorophenyl 4-Iodophenyl 4-Bromophenyl 4-Cyanophenyl 2-Phenyloxycarbonylphenyl 4- Ethylphen y1 4-Methylphenyl Phenyl
17 16 22 24 35 38 24 76 83 80 62 80 115 155 > 200 >200 >200 1200
>200 > 200
Table 5. Subcellular distribution of proteinase In activity in E. coli. E. coli K-12 IAM1264 was grown as far as the exponential phase (absorbance at 600 nm of 0.55). Each fraction was assayed for proteinase In and marker en7ymes, alkaline phosphatase, 8-galactosidase and lactatc dehydrogenase. Proteinase activity was examined with Boc-Val-Pro-Arg-NH-Mec (final concentration of 6.67 pM) as the substrate in 0.1 M borate buffer, pH 8.0, containing 1 mM CaClz and 0.08 M NaCl at 37'C. n. d., not dctcctcd. Subcellular fraction
Total activity of alkaline phosphatase
Periplasm 91.2 Cytoplasm 7.4 Crude mem1.4 brane
8-galactosidase
lactate dehydrogenase
proteinase In
4.4 90.2
n.d. 2.5
5.8 92.4
5.4
97.5
1.8
iin) (23) (366) (7)
completely removed by repeated L-arginine-Sepharose 4B chromatography. The apparent molecular mass was estimated as 66 kDa by gel filtration on Sephadex G-200 and SDS/PAGE. Purified proteinase In hydrolyzes fluorogenic substrates for trypsin. A similar en~yme,protease 11, whose molecular mass is about 58 kDa and hydrolyzes a trypsin a substrate Bz-Arg-NAn, was purified from E. coli by Pacaud and Richaud [29]. In the present study, the crude extract obtained from 100 g wet cells contained 2.915 pmol/min of Bz-Arg-NAn-hydrolytic activity. However, almost all of this activity (approximately 97%) was removed by DEAE-cellulose column chromatography (step 1) and precipitation by 45% saturation of am-
monium sulfate (step 2), mainly at step 2. Purified proteinase In contained no Bz-Arg-NAn-hydrolytic activity Protease I1 was completely inhibited by incubation with TosLysCH,CI and diisopropylfluorophosphate [29]. Proteinase In was weakly inhibited by the former but not by the latter. Protease 11 was competitively inhibited by benzamidine (Ki, 0.6 mM) [29], but not by proteinase In at 1 mM. Proteinase In thus differs from protcase TI. Although proteinase In hydrolyzes fluorogenic substrates for trypsin, its activity appears markedly affected by an adjacent amino-acid sequence lying N-terminal of the arginine or lysine residue. Hydrophobic amino-acid residues such as proline and leucine may be required for maximal activity, and hydrophilic residues such as glycine, lysine, serine and threonine apparently markedly reduce activity. Fluorogenic substrates for chymotrypsin and elastase were not hydrolyzed. Proteinase In may thus belong to the trypsin-like proteinases, but differ from serine enzyme since diisopropylfluorophosphate and phenylmethylsulfonyl fluoride had no effect. Leupeptin, antipain and aprotinin strongly inhibited proteinase In, while STI, benzamidine and pentamidine isethionate did not. It thus follows that proteinase In may be a unique trypsin-like proteinase. Kato et al. (Kato, M., Irisawa, T. and Muramatu, M., unpublished results) found that various substituted phenyl esters of GMCHA strongly inhibit the growth of E. coli, this apparently being correlated with a hydrophobic or steric factor of the substituent, and from inhibition experiments on DNA synthesis they suggested the possible involvement of a trypsin-like proteinase in cell growth of E. coli. A Boc-ValPro-Arg-NH-Mec-hydrolytic activity, proteinase In: has been found just prior to the initiation of DNA replication in the cell cycle of E. coli synchronized by phosphate starvation. Inhibition of proteinase activity by GMCHA-OPheBu, a representative GMCHA ester, has been noted to prolong the DNA-replication period, thus retarding the course of the cell cycle (Kato, M., Irisawa, T., Morimoto, Y. and Muramatu, M., unpublished results). As shown in Table 4, proteinase In was strongly inhibited by various substituted phenyl esters of GMCHA, it was competitively inhibited by GMCHAOPheRu. The order of inhibitory effect on proteinase In was roughly correlated with that of their effects on E. coli growth. These effects on proteinase appear correlated with a hydrophobic or steric factor of substituent. These results appear to provide additional evidence for the above view that proteinase In is involved in DNA replication. Prouty and Goldberg rcported that pentamidine isethionate inhibited the growth of E. coli and macromolecular synthesis in E. coli [15]. It was also noted that pentamidine isethionate dose-dependently inhibit this growth and the uptake of [3H]thymidine into DNA. However, the effects were transient and reparable and differed from those of GMCHAOPheRu (Kato, M., Irisawa, T. and Muramatu, M., unpublished results). That proteinase In is strongly inhibited by GMCHA esters, but not by pentamidine isethionate may be the reason for this difference. The inhibitory effects of the latter may arise from some other mechanism. The specific intracellular substrate for proteinase In, and its role in the initiation of DNA replication in E. coli cells are points that remain to be determined.
REFERENCES 1. Goldberg, A. L. & John, A. C. St. (1976) Annu. Rev. Biochhem. 45. 747 - 803.
1014 2. 3. 4. 5. 6. 7. 8. 9. 10.
11. 22. 13. 14. 15.
Mount, D. W. (1980) Annu. Rev. Genet. 14, 279 -319. Wickner, W. (1979) Annu. Rev. Biochem. 48, 23-45. Davis, B. D. & Tai, P. C. (1980) Nature 283,433-438. Cavard, D. & Lazdunski, C. (1979) Eur. J . Biochem. 96, 529533. Watsen, D. H. & Sherratt, D. J. (1979) Nature 278, 362-364. Roberts, J. W., Roberts. C. W. & Craig, N. L. (1978) Proc. Nut/ Acad. Sci. U S A 75,4714-4718. Little, J . W., Edmiston, S. H., Pacelli, L. Z. & Mount, D. W. (1980) Proc. Nut1 Acud. Sci. USA 77,3225 - 3229. Lazdunski, A. M. (1989) FEMSMicrobiol. Rev. 63, 265-276. Sugimura, K. & Nishihara, T. (1988) 1. Baeteriof. 170, 56255632. Sugimura, K. & Higashi, N. (1988) J. Bacteriol. 170,3650- 3654. Lee, C. S., Park, W. J . , Jung, E. M., Choi, K. H., Ha, D. B. & Chung, C. H. (1991) Biochem. Znt. 23, 1155-1163. Vaithilingam, I. & Robert, A. (1989) Biochem. Int. 19, 12971307. Seol, J. H., Woo, S. K., Jung, E. M., Yoo, S. J., Lee, C. S., Kim, K., Tanaka, K. & Ichihara, A. (1991) Biochem. Biophys. Res. Commun. 176,730-736. Prouty, W. F. & Goldberg, A. C. (1972) J. Biol. Chem. 247, 3341 - 3352.
16. Rossmam, T., Norris, G. & Troll, W. (1974) J . Biol. Chem. 249, 3412- 3417. 17. Ito, K. (1977) J . Bacteriol. 132, 1021 -1023. 18. Ito, K. (1978) Biochem. Biophys. Res. Commun. 82, 99 - 107. 19. Davis, B. D. & Mingioli, E. S. (1950) J . Bacteriol. 60, 17-28. 20. Kanaoka, Y., Takahashi, T., Nakayama, H., Takada, K., Kimura, T. & Sakakibara, S. (1977) Chem. Phurrn. Bull. (Tokyo) 25, 3126-3128. 21. Erlanger, B. F., Kokowsky, N. & Cohen, W . (1961) Arch. Biochem. Biophys. 95,271 -278. 22. Lowry, 0. H., Rosebrough, N . J., Farr, A . L. & Randall, R. J. (1951) J . Biol. Chem. 193, 265-275. 23. Andrews, P. (1964) Biochem. J. 91, 222-223. 24. Pujimoto, Y., Moriya, H., Yamaguchi, K . & Moriwaki, C. (1972) J . Biochem. (Tokyo) 71,751 -754. 25. Yamato, I., Anraku, Y. & Hirosawa, K. (1975) J . Biochem. (Tokyo) 77,705-738. 26. Osborn, M. J.; Gander, J. E.; Parisi, E. & Carson, J. (1972) J . Biol. Chem. 247, 3962 - 3972. 27. Zipser. D. (1963) J. Mol. Biol. 7, 113-121. 28. Malamy, M. &Horeker, B. L. (1964)Biochem.J . 3, 1893-1897. 29. Pacaud, M. & Richaud, C. (1975) .I. Biol. Chern. 250, 77717779.
Purification and characterization of proteinase In, a trypsin-like proteinase, in Escherichia coli Munetoshi KATO, Takaji IRISAWA, Makiko OHTANI and Mutumi MURAMATU Faculty of Pharmacy, Tokushima Bunri University, Japan (Received July 6iSeptember 7, 1992)
~
EJB 92 0946
We previously found a trypsin-like proteinase which momentarily appears immediately before DNA synthesis in the cell cycle of Escherichia coli synchronized by phosphate starvation and which is closely related to the initiation of DNA replication (Kato, M., Irisawa, T., Morimoto, Y. and Muramatu, M., unpublished results). The proteinase was named proteinase In. It was purified approximately 2880-fold with a recovery of 15%. The isolated enzyme appeared homogeneous by gel filtration and electrophoresis. Its molecular mass was estimated by analytical gel filtration and SDS/ PAGE as approximately 66 kDa. The isoelectric point of proteinase In is 4.9 and its optimal pH is approximately 9. Although protein In hydrolyzes fluorogenic substrate for trypsin, its hydrolytic activity seems markedly affected by amino-acid sequence lying towards the N-terminal from the PI (lysine, arginine) residue. The proteinase does not hydrolyze NZ-benzoyl-~,~-arginine-4-nitronanilide and fluorogenic substrates for chymotrypsin and elastase. The proteinase activity is inhibited by leupeptin, antipain and 4-nitrophenyl 4-guanidinobenzoate, but the effects of tosyl-r-lysine chloromethane, diisopropylfluorophosphate, benzamidine and pentamidine isethionate on the proteinase activity are weak or not inhibitory. Its activity is strongly affected in the presence of NaCl and KCl, and at a concentration of 1.5 M, these increase the activity 14-fold and 13-fold, respectively, above that without salt. Proteinase In was strongly inhibited by various esters of trans-Cguanidinomethylcyclohexanecarboxylic acid, and their inhibitory effects werc roughly correlated with those on growth of E. coli. Proteinase activity was found in the cytoplasmic fraction.
sible proteasomes) [13], and identity of protease Do and htrA-gene product [14]. Many workers reported the effects of various protease inhibitors on E. coli, such as suppression of the degradation of protein by various inhibitors [15],inhibition of biosyntheses of protein and RNA by various chloromethylketones and recovery of these inhibitions by reduced glutathione [16], inhibition of synthesis of protein I in membranes and degradation of a protein by tosyl-L-lysine chloromethane (TosLysCH2Cl) [17] and by antipain and leupeptin [IS]. We previously reported that various esters of trans-4guanidinomethylcyclohexanecarboxylic acid (GMCH A), synCorrespondence to M. Muramatu, Faculty of Pharmacy, Toku- thetic trypsin inhibitors, strongly inhibited the growth of E. shima Bunri University, Yamashiro-cho, Tokushima-shi, Tokushima, coli. These effects may be due to suppression of DNA repliJapan 770 cation, and a trypsin-like proteinase may possibly participate Abbreviations. GMCHA, trans-4-guanidinomethylcyclohexanecarboxylic acid; GMCHA-OPheBu, 4-ferf-butylphenyl ester of in DNA synthesis (Kato, M., Irisawa, T. and Muramatu, M., GMCHA; Boc, tert-butoxycarbonyl; Bz, N2-benzoyl; SUC,succinyl; unpublished results). Moreover, using E. coli synchronized Glt, glutaryl; Boc-Val-Pro-Arg-NH-Mec, 4-methylcoumaryl-7-amide by a phosphate-starvation method, the suppression effect of of tert-butoxycarbonyl-L-valyl-L-prolyl-L-arginine; NAn, 4-nitro- GMCHA esters on the growth of E. coli was found to be due anilide; TosLysCH,Cl, l-chloro-3-tosylamido-7-amino-2-heptanoneto inhibition of a trypsin-like proteinase which momentarily hydrochloride, tosyl-1.-lysine chloromethane; TosPheCHZC1,L-l-toappeared immediately before DNA synthesis in the cell cycle sylamide-2-phenylethyl chloromethane, tosyl-L-phenylalanine chloof E. coli. The proteinase may be closely related to the initiaromethane; NPGB, 4-nitrophenyl 4-guanidinobcnzoate: STI, soybean trypsin inhibitor; ICso,inhibitor concentration required for tion of DNA replication in synchronized E. coli cells. The proteinase was named proteinase In (Kato, M., Irisawa, T., 50% inhibition. Morimoto, Y. and Muramatu, M., unpublished results). Enzymes. Trypsin (EC 3.4.21.4); lysozyme (EC 3.2.1.17); alkaline phosphatase (EC 3.1.3.1): B-galactosidasc (EC 3.2.1.23); lactate deThis paper reports the purification of proteinase In and hydrogenase (EC 1.1.1.27). its properties are discussed.
Proteolytic processes play an important role in various functions of Escherichiu coli: degradation of normal and abnormal proteins [l, 21; maturation of secretory and membrane proteins [ 3 , 41; breakdown of colicins [5, 61; inactivation of certain regulatory proteins [7, 81. A comprehensive and pertinent review of these matters has been presented by Lazdunski [9]. Recent studies have reported, identity of-outer-membraneassociated protease VII and OmpT which specifically cleaves ferric enterobactin receptor protein [lo], in analysis of a novel outer-membrane-associated protease [l 11the induction of protease La under stress [12],high-molecular-mass proteases (pos-
1008 MATERIALS AND METHODS Materials Various fluorogenic substrates, N’-benzoyl-D,L-arginine4-nitroanilide (Bz-Arg-NAn), leupeptin, antipain and chymostatin were purchased from Peptide Institute Inc. l-Chloro-3tosylamido-7-amino-2-heptanonehydrochloride (tosyl-L-lysine chloromethane, TosLysCHzC1) and ~-1-tosylamide-2phenylethyl chloromethane (tosyl-L-phcnylalanine chloromethane, TosPheCH’Cl) were obtained from Nakarai Chemical Ltd. Diisopropylfluorophosphate, phenylmethylsulfonyl fluoride, 4-nitrophenyl 4-guanidinobenzoate (NPGB), soybean trypsin inhibitor (STI) and aprotinin were from Sigma Chemical Co. DEAE-cellulose (Brown) and egg-white lysozyme were from Seikagaku Kogyo Co. Benzamidine was from Tokyo Chemical Ind. L-Arginine-Sepharose 4B, phenyl-Sepharose CL-4B and Sephadex G-200 were from Pharmacia. Hydroxyapatite (Bio-GEL . HTP) was from Bio-Rad Laboratories. 4-Nitrophenylphosphate and 2-nitrophenyl-P-~-galactopyranoside were from Boehringer Mannheim. Various esters of trans-4-guanidinomethylcyclohexanecarboxylicacid (GMCHA) were obtained as hydrochlorides from Nippon Chcmiphar Co. Ltd.
E. coli and growth condition E. coli strain K-12 IAM1264 was kindly provided by Dr. M. Takahashi, Godo Shusei Co. Ltd., Toyko. E. coli cells were grown in a glucose minimal medium [19] at 37°C. When absorbance at 600 nm reached 0.55 (approximatcly 6 x 10’ cells/ml), the cells were harvested by centrifugation at 7000g for 20 min at 4°C. From l l culture medium approximately 1 g cells was obtained. The cells (1 g) were resuspended in 10 ml0.1 M sodium borate, pH 8.0, containing 0.1 5 M NaCl and frozen at 20°C.
albumin, 45 kDa; chymotrypsinogen, 25 kDa; cytochrome c, 12.5 kDa. Electrophoresis and proteinase activity For electrophoresis the Phast System and Phast Gel media (Pharmacia) were used for IEF pH 3 - 9. native PAGE and SDS/PAGE. The Phast System was also used for automated Coomassie-blue staining. All procedures were conducted according to the manufacturer’s recommendations. Detection of proteinase activity in the gel was performed according to Fujimoto et al. [24] with minor modification. After electrophoresis, the gel was immersed in 15 ml 100 1 M Boc-Val-Pro-Arg-NH-Mec dissolved in 0.1 M sodium borate, pH 8.0, containing 1 mM CaCI2 and 0.08 M NaCl. It was kept in the dark at 37°C. After 3 h, the gel was irradiated at 365 nm with Ultra Violet Lamp VL-6LC (Vilber Lourmat, France). Proteinase activity was detected in the gel as a fluorescent band. Buffer for measuring the pH/activity profile The pH/activity profile of purified proteinase was examined with the following buffers. Britton and Robinson’s universal buffer: a mixture of 50 mM citric acid, KH2P04,boric acid and sodium diethylbarbiturate was adjustcd to a L.w e n pH with 3.5 M NaOH. Borate buffer for measuring activity in the pH range 7.5 - 9.0 the pH of the buffer was adjusted by mixing 60mM boric acid and sodium borate. For the pH range 9.5- 10.0 the pH was adjusted by mixing 60 mM boric acid and 3.5 M NaOH. Inhibition of proteinase activity
Mixtures of 1.0 m10.05 M sodium borate, pH 9.0, containing various amounts of an inhibitor and proteinase were incubated with 0.5 ml Boc-Val-Pro-Arg-NH-Mec dissolved in the Assay of proteinase activity same buffer, and proteinase activity was determined as above. When thc high-molecular-mass inhibitors, STI and aprotinin, The hydrolytic activity of proteinase on 4-methyl- were used, the proteinasc was incubated with various amounts coumaryl-7-amide of tert-butoxycarbonyl-L-valyl-L-prolyl-Lof the inhibitor in 1 ml of the above buffer at 25 C. After arginine (Boc-Val-Pro-Arg-NH-Mec) was determined in 10 min, 0.5 ml substrate solution was added and incubation 0.1 M sodium borate, pH 8.0, containing 1 mM CaC12 and was continued at 37°C for 1 h. The final substrate concen0.08 M NaC1. Mixtures of 0.5 ml of the above buffer containtration was 20 pM. ing various amounts of proteinase and 1 ml substrate solution were incubated at 37’C. After 1 h, 1 ml 30% acetic acid was added and 7-amino-4-methylcoumarin released was deter- Cellular localization mined as described by Kanaoka et al. [20]. Final substrate Spheroplasts were prepared according to the method of concentration was 6.67 pM. Yamato et al. [25]. Those obtained from approximately 1.6 g Bz-Arg-NAn hydrolytic activity was measured at 410 nm E. coli cells were lysed by osmotic shock as described by in the above buffer as described by Erlanger et al. [21]. Osborn et al. [26]. The subcellular fractions obtained were ~
Protein assay Protein concentration was determined by the method of Lowry [22] using bovine serum albumin as the standard.
dialyzed against 0.1 M sodium borate, pH 8.0, containing 1 mM CaCI, and 0.08 M NaCI. The fractions were monitored by measuring p-gdlactosidase [27]. alkaline phosphatase [28] and L-lactate dehydrogenasc which was assayed by diaphorase using lactate dehydrogenase C II-Test Wako (Wako Pure Chemical Industries Ltd.).
Determination of molecular mass by gel filtration The molecular mass of the purified proteinase was determined by gel filtration on a Sephadex (3-200 [23]. The column (2.5 cm x 101 cm) was equilibrated and eluted with 0.05 M sodium borate, pH 8.0, containing 1 mM CaCI, and 0.15 M NaCl, at a flow rate of 15 ml/h. Protein standards were y-globulin, 158 kDa; bovinc serum albumin, 66 kDa; ov-
RESULTS Purification of proteinase In Unless otherwise stated, all procedures were conducted at 4°C using 0.1 M sodium borate, pH 8.0, containing 1 mM CaCI, as the standard buffer (buffer A) for purifying protein-
1009 ase In. Proteinase activity was assayed with the fluorogenic substrate Boc-Val-Pro-Arg-NH-Mec. Preparation of crude extract Approximately 100 g E . coli K-12 cells suspended in 1 1 0.1 M sodium borate, pH 8.0, containing 0.15 M NaCl, frozen at - 20 'C, were allowed to thaw and each 50 ml of suspension was sonicated six times at 100 W for 30 s in an ice-bath. The homogenate was centrifuged at 13000 x g for 30 min. Thc precipitate was washed twice with 500 ml buffer A by sonication and centriruging as described above. The supernatant fluid was collected and designated the crude extract. DEAE-cellulose column chromatography (step 1) The crude extract (2 1) was dialyzed against buffer A containing 0.08 M NaCI. The dialysate was applied onto a DEAEcellulose column (3.2 cm x 30 cmj equilibrated with the above buffer and washed with 700ml of the same buffer. The adsorbed materials were eluted with 1.2 1 of a linear gradient of 0.08 M to 0.5 M NaCl in buffer A. 8.7-ml fractions were collected at a flow rate of 36ml/h. Protein and proteinase activity of the effluent were assayed and the activity fractions (fraction numbers 52 - 85) were pooled. Total activity of proteinase In then increased approximately 237% above that of the crude extract, possibly as a result of removing endogcnous protease inhibitors or competing substances.
0.'
0 10
z
50
0
100
Fraction number (4.2 mlltube)
Fig. 1. Second gel filtration of proteinase In on Sephadex G-200. Proteins obtained from L-arginine-Sepharose 4B (pH 8.2) through the column. Details of the procedure are indicated in the text (step 8). Proteinase-In activity was assaycd with Boc-Val-Pro-Arg-NH-Mec as descrtbcd. Thc bar indicates the fraction pooled. (0 El), Protein.), absorbance at 750 nm. ase activity; -.( ~
-
Ammonium-sulfate fractionation (step 2) The active fraction (280 mlj obtained from DEAE-cellulose column Chromatography was precipitated by 45% saturation with ammonium sulfate. After 2 11, the suspension was centrifuged at 13000 ,q for 30 min. The pellet thus obtained was discarded. To the supernatant fluid, solid ammonium sulfate was added to 80% saturation and the mixture was allowed to settle for 4 h. The precipitates were collected by centrifugation as described above. Sephadex G-200 gel filtration (step 3) The precipitates at 80% ammonium-sulfate saturation were dissolved in 16 ml 0.05 M sodium borate, pH 8.0, containing 1 mM CaCl, and 0.15 M NaCl and dialyzed against the same buffer. The dialysate (23 11) was separated into two parts, followed by gel filtration of each on a Sephadex G-200 column (2.5 cm x 101 cmj cquilibrated with the same buffer. The flow rate was 15 ml/h and 4.2-ml fractions were collected. The proteinase fraction (fraction numbers 63 - 81j was combined (146 ml). Phenyl-Sepharose CL-4B chromatography (step 4)
To the combincd fractions from the Sephadex G-200 column NaCl was added at 3 M NaCI, and the solution was applied to phenyl-Sepharose C L d B (2.2 cm x 13.5 cm) equilibrated with buffer A, pH 7.5, containing 3 M NaCl. The column was washed with 340ml of the same buffer and adsorbed materials were eluted with 1 1 of a linear gradient of 3.0 M to 0 M NaCl in buffer A, pH 7.5. 4.8-ml fractions were collccted at a flow rate of 15 ml/h. The active fractions (fraction numbers 161 -220) were combined (280 ml).
Hydroxyapatite chromatography (step 5) The active fraction from the above column chromatography was dialyzed against 10 mM potassium phosphate, pH 6.8, and applied onto a column of hydroxyapatite (1.8 cm x 10 cm) equilibrated with the same buffer. The column was washed with 50ml of the same buffer and adsorbed materials were eluted with 600 ml of a linear gradient of 10 mM to 150 mM potassium phosphate, pH 6.8. The flow rate was 18 ml/h and 4.1-ml fractions were collected. The active fractions (fraction numbers 27 - SO) were combined (90 ml). L-Arginine-Sepharose 4B chromatography (step 6) The active fraction from hydroxyapatite column chromatography was dialyzed against buffcr A and applied onto a column of L-arginine-Sepharose 4B (1.6 cm x 7 cm) equilibrated with the same buffer. The column was washed with 90 ml of the 5ame buffer. The adsorbed materials were eluted with 600 ml of a linear gradient of 0 M to 0.15 M NaCl in buffer A. 4.1-ml fractions were collectcd at a flow rate of 18 ml/h. The active fractions (fraction numbers 91 - 115) were combined (92 ml). Further chromatography on L-Arginine-Sepharose 4B column (step 7) The active fraction from the L-arginine-Sepharose 4B column was dialyzed against 0.1 M sodium borate, pH 8.2, containing 1 mM CaCI, and applied onto a column of Larginine Sepharose 4B (1 cm x 4.5cmj equilibrated with the
1010 Table 1. Purification of proteinase In from E. cofi. Proteinase In activity was examined with Boc-Val-Pro-Arg-NH-Mcc as the substrate.
Purification step
Activity
Protein
Specific activity
Yield
Purification
pmol/min 18333 43371 41 831 36933 18713 11446 5888 4089 2788
mg 4144 1482 660 316 44.1 8.7 0.90 0.41 0.22
pmol . min-' . mg-' 4.4 29 63 117 424 1314 6 542 9 973 12673
Yo
-fold
Crude cxtract DBA R-cellul ose Ammonium sulfatc (45- 80%) Sephadex G-200 Phenyl-Sepharose CL-4B Hydroxyapatite (Bio-GEL HTP) L-Arginine Sepharosc 4B at pH 8.0 L-Arginine Sepharose 4B at pH 8.2 2nd Sephadex G-200
100 237 228 202 102 62 32 22 15
1 7 14 27 96 299 1487 2267 2880
same buffer. The column was washed with 30 ml of the same buffer. The adsorbed materials were eluted with 300 ml of a linear gradient of 0 M to 0.15 M NaCl in the above buffer. 2.6-ml fractions were collected at a flow rate of 15 ml/h. The fractions with proteinase activity (fraction numbers 41 - 61) were combined (46 ml). Second gel filtration on Sephadex G-200 column (step 8) The active fraction from the above column chromatography was concentrated to 6 ml in an Amicon YM-5 membrane and dialyzed against 0.05 M sodium borate, pH 8.0, containing 1 mM CaC12 and 0.15 M NaCl. The dialysate was subjected to gel filtration on a Sephadex (3-200 column (2.5 cm x 101 cm) equilibrated with the same buffer. 4.2-ml fractions were collected at a flow rate of 15 ml/h (Fig. 1). Fractions with proteinase activity (fraction numbers 70 - 80) were combined (42 ml), dialyzed against 0.05 M sodium borate, pH 9.0, and stored at 0°C. A Summary of the purification is shown in Table 1. The yield of enzyme activity was 15% with 2880-fold purification. Criteria of homogeneity Rechromatography of proteinase In on Sephadex G-200 gave a single peak with constant specific activity (Fig. 1). From analytical gel electrophoresis, only a single band with non minor contaminants was detected by protein staining. The band showed hydrolytic activity for Boc-Val-Pro-ArgNH-Mec (Figs 2A, B). Molecular-mass determination and isoelectric point Prom the results of step 8 and method of Andrews 1231, molecular mass was calculated as 66 kDa (Fig. 3A). SDS/ PAGE gave an average value of 66 kDa (Fig. 3B). Proteinase In appeared to be composed of a single subunit. Proteinase In had an jsoelectric point of approximately 4.9 as determined on a Phast Gel IEF 3-9 (Fig. 2C). pH/activity profile The pH dependence of proteinase In was determined in Britton and Robinson's universal buffer and sodium borate buffer. A sharp activity profile was observed with maximal activity at pH 9.0 in both buffers (Fig. 4). Substrate specificity Hydrolytic activity of proteinase In on various fluorogenic substrates is shown in Table 2. Proteinase In clearly hy-
drolyzed trypsin-specific substrates, particularly Boc-ValLeu-Lys-NH-Mec and Boc-Val-Pro-Arg-NH-Mec. K, for Boc-Val-Pro-Arg-NH-Mec was 62.5 pM. However, hydrolytic activity on Boc-Leu-Gly-Arg-NH-Mec, Boc-Glu-Lys-LysNH-Mec, Boc-Phe-Ser-Arg-NH-Mec. Boc-Leu-Ser-Thr-ArgNH-Mec, Boc-Ile-Wu-Gly-Arg-NH-Mec, Glt-Gly-Arg-NHMec (Glt, glutaryl) and Bz-Arg-NH-Mcc was very low or not detectable. Arg-NH-Mec was not hydrolyzed. Proteinase In would thus appear to require hydrophobic amino acid sequence lying towards the N-terminal from the P1 (lysine, arginine) residue for maximal activity. Fluorogenic substrates for chymotrypsin and elastase were not hydrolyzed by proteinase In. No Bz-Arg-NAn hydrolytic activity was observed. Effects of monovalent salts Fig. 5 shows the effects of monovalent salts. Proteinase In activity increased with the concentrations of NaC1 and KC1, and at 1.5 M, activity reached 14-fold and 2 3-fold, respectively, above that without these salts. At higher salt concentrations, activity decreased. LiCl had a diphasic effect on activity as shown in Fig. 5. Effects of various inhibitors and bivalent cations The effects of proteasc inhibitors, EDTA and bivalent cations are shown in Table 3. Proteinasc In was strongly inhibited by leupeptin, antipain and NPGB. Chymostatin showed SO -60% inhibition at 200 pM. Aprotinin had a strong inhibitory effect, but STI had none. The effects of phenylmethylsulfonyl fluoride, TosPheCH2C1, diisopropylfluorophosphate, benzamidine and pentamidine isethionate were not inhibitory. TosLysCH,CI, EDTA and CaClz were considerably inhibitory and MnClz particularly so. The effect of MgC12 was slight. Effects of various GMCHA esters As described by Kato et al. (Kato, M., Irisawa, T., and Muramatu, M., unpublished results), various GMCHA esters were shown to strongly inhibit the growth of E. coli. Table 4 shows the inhibitory effects of GMCHA esters on proteinaseIn activity, growth of E. coli and Ki for trypsin. Thc order of inhibitory GMCHA ester effect on proteinase activity was roughly correlated with that on E. coli growth, but not with Ki for trypsin. GMCHA-OPheBu, a representative GMCHA ester, competitively inhibited proteinase In and its Ki was 41 pM.
1011
h
3-
-
i
i 0
I
I
I
10
20
30
40
Distance from cathode (mm)
Fig.2. Electrophoretic pattern of the purified enzyme in native PAGE (Phast Gel Gradient 8-25) and determination of its isoelectric point. (A) Preparation obtained from Sephadex G-200 stained with Coomassie blue. Details of the procedure are indicated in Materials and Methods. (B) Hydrolytic activity on Boc-Val-Pro-Arg-NH-Mec. Details of the procedure were given in Materials and Methods. (C) Determination of the isoelectric point. The isoelectric point of proteinase In was determined as described in Materials and Methods. Coomassie-stained proteinase In and marker proteins ( P I range 3 -9, Pharmacia) bands are shown on the left. The pH-gradient profile was obtained by the broad piCalibration Kit running on Phast Gel IEF 3 -9. The arrow is at thc value of proteinase In. The corresponding pIare: a, lentil lectin (basic), 8.65; b, lentil lectin (middle), 8.45; c, lentil lectin (acidic), 8.15; d, horse myoglobin (basic, 7.35; e, horse myoglobin (acidic), 6.85; f, human carbonic anhydrase R, 6.55; g, bovine carbonic anhydrase R, 5.85; h, /?-lactoglobulin A, 5.20; i, soybean trypsin inhibitor, 4.55; j, amyloglucosidasc, 3.50.
Elution volume (ml)
Relative mobility
Fig. 3. Molecular-mass determinationfor proteinase In. (A) Estimation of the molecular mass of proteinase In by gcl filtration on Sephadex G200. Conditions are indicated in Materials and Methods. The arrow is at the value of elution volume of the proteinase In. Marker proteins (Sigma) used were a, ?-globulin (bovine); b, bovine serum albumin; c, ovalbumin; d, chymotrypsinogen; e, cytochrome c. (B) Calibration of SDSjPAGE on Phast Gel Gradient 10- 15 with marker proteins (LMW kit E, Pharmacia). Protein molecular mass is plotted against rclativc mobility determined from bromophenol-blue migration. The arrow marks the position of proteinase In. The corresponding molecular masses are f, phosphorylase b, 94 kDa; g, albumin, 67 kDa; h, ovalbumin, 43 kDa; i, carbonic anhydrase, 30 kDa; j, trypsin inhibitor, 20.1 kDa; k, cc-lactalbumin, 14.4 kDa.
Cellular localization
DISCUSSION
Alkalinc phosphatase was found in the periplasmic fraction and proteinase In and jl-galactosidase, in the cytoplasmic fraction (Table 5 ) , for cells in the exponential phase.
Proteinase In was purified as described in this study and appeared to be homogeneous by the elution profile from gel filtration on Sephadex G-200 column and from electro-
1012
1 h
aQ Y
a c ._ > ._ c 0
m W
> .c
-6W a
6
I
I
I
I
7
8
9
10
11
PH
Fig. 4. Effects of pF1 on proteinase In. The aclivity of protcinase In was assaycd with 60 mM sodium borate (0)and Britton and Robinson’s universal buffer ( 0 ) Boc-Val-Pro-Arg-NH-Mec . was used as the substrate (final concentration of 20 pM). Values represent percentages of activity of protcinase In in 60 mM sodium borate, pH 9.0.
Table 2. Hydrolytic activity of proteinase In on various fluorogenic substrates. Hydrolytic activity was examined in 0.05 M sodium borate, pH 9.0 at 3 7 T , containing 20 FM of each substrate. Values represent percentages of activity of protcinase In on Boc-Val-Pro-Arg-NHMec. Boc, rert-butoxycarbonyl ; Glt, glutaryl; Bz, N2-benzoyl; Suc, succinyl; n.d., not detected. Fluorogenic substrate
Relative activity
Yo Boc-Val-Pro-Arg-NH-Mec Boc-Val-Leu-Lys-NH-Mec Boc-Leu-Gly-Arg-HN-Mec Boc-Glu-Lys-Lys-NH-Mec Boc-Phe-Ser-Arg-NH-Mcc Boc-Leu-Ser-Thr-Arg-NH-Mec Boc-Ile-Glu-Gly-Arg-NH-Mec Glt-Gly- Arg-NH-Mec Rz-Arg-NH-Mec Suc-Leu-Leu-Val-Tyr-NH-Mec SUC-Ah-Pro-Ala-NH-Mec Arg-NH-Mec
100 216.8 1.7 1.3 3.0 1.4 5.0 n. d. 6.3 n. d. n. d. n. d.
phoresis. Additional purity tests have not been made owing to the small amount of enzyme available. The present purification procedure consists of seven steps and is quite tedious, because crude extract obtained from 100 g wet cells contained much Arg-NH-Mec-hydrolytic activity (1.56 pmol/min), which was detected over the cell cycle of synchronized E. coli (unpublished results), and the aminopeptidase activity was quite difficult to separate from proteinIn activity. However, this aminopeptidase activity could be
0
1
Concentration of monovalent salts (MI
Fig. 5. Effects of monovalent salts on proteinase In. Effects of NaCI. KCI and LiCl on proteinase In were examined with Boc-Val-ProArg-NH-Mec in 0.05 M sodium borate, pH 9.0, at 37°C. The final substrate concentration was 20 pM. Each activity is shown as the activity (-fold) relative to NaCI, KCI and LiCI-free control. (0-O), NaCl; (0-0), KCI; (A-A), LiCl.
Table 3. Effects of various inhibitors and bivalent cations on proteinase In. Inhibitory effects on proteinase In were examined using Boc-ValPro-Arg-NH-Mec as the substrate. All inhibitors were previously incubated with proteinasc In at 25°C for 10 min, except for leupeptin. antipain, chymostatin and NPGB. Inhibitor
Inhibitory effects IC50
inhibition
Yo Leupeptin Antipain Chymostatin NPGB Aprotinin 100 Fg/ml STI 1 mM Phenylmethylsulfonyl fluoride 1 mM Benzamidine 1 mM TosLysCHzCl 1 mM TosPheCHzCl 1 mM Diisopropylfluorophosphate 1 mM EDTA 0.2 mM Pentamidine isethionate 1 mM CaCI, 5 mM CaCI2 1 mM MgClz 5 mM MgClz 1 mM MnCI, 5 mM MnClz
84 FM 14 pM 56 pM 102 pM 8.5 pg/ml 0 3.6 3.2 33.5 7.6 0.6 51.9 0 8.6 46.1 0 15.4 76.0 98.2
1013 Table 4. Inhibitory effects of various GMCHA esters on proteinase In, growth of E. coli and Kifor trypsin. Inhibitory effects on protcinase In were examined with Boc-Val-Pro-Arg-NH-Mec as a substrate. * n = 15, P < 0.01.
GMCHA ester
Inhibitory effect in E. coli on Proteinase In
Growth 1C5*
Trypsin Ki
21 26 36 42 43 44 45 64 68 74 87 92 105 135 142 153 167
35 54 29 45 45 273 38 103 325 187 56 46 51 113 85 56 48 78 110
IC50
4-Chloro-2-isopropyl-5-methylphenyl 20 4-Diphenyl 4-Benzyloxycarbonylethylphen yl 4-Bcnzyloxycarbonylphenyl 3-Benzylox ycarbon ylphenyl 2,4,6-TnchlorophenyI 4-tert-Butylphenyl 4-Phenyloxycarbonylphenyl 2-Isopropyl-5-methylphenyl 2-Diphenyl 2-Benqlox ycarbonylphen yl 2,4-Dichlorophenyl 4-Iodophenyl 4-Bromophenyl 4-Cyanophenyl 2-Phenyloxycarbonylphenyl 4- Ethylphen y1 4-Methylphenyl Phenyl
17 16 22 24 35 38 24 76 83 80 62 80 115 155 > 200 >200 >200 1200
>200 > 200
Table 5. Subcellular distribution of proteinase In activity in E. coli. E. coli K-12 IAM1264 was grown as far as the exponential phase (absorbance at 600 nm of 0.55). Each fraction was assayed for proteinase In and marker en7ymes, alkaline phosphatase, 8-galactosidase and lactatc dehydrogenase. Proteinase activity was examined with Boc-Val-Pro-Arg-NH-Mec (final concentration of 6.67 pM) as the substrate in 0.1 M borate buffer, pH 8.0, containing 1 mM CaClz and 0.08 M NaCl at 37'C. n. d., not dctcctcd. Subcellular fraction
Total activity of alkaline phosphatase
Periplasm 91.2 Cytoplasm 7.4 Crude mem1.4 brane
8-galactosidase
lactate dehydrogenase
proteinase In
4.4 90.2
n.d. 2.5
5.8 92.4
5.4
97.5
1.8
iin) (23) (366) (7)
completely removed by repeated L-arginine-Sepharose 4B chromatography. The apparent molecular mass was estimated as 66 kDa by gel filtration on Sephadex G-200 and SDS/PAGE. Purified proteinase In hydrolyzes fluorogenic substrates for trypsin. A similar en~yme,protease 11, whose molecular mass is about 58 kDa and hydrolyzes a trypsin a substrate Bz-Arg-NAn, was purified from E. coli by Pacaud and Richaud [29]. In the present study, the crude extract obtained from 100 g wet cells contained 2.915 pmol/min of Bz-Arg-NAn-hydrolytic activity. However, almost all of this activity (approximately 97%) was removed by DEAE-cellulose column chromatography (step 1) and precipitation by 45% saturation of am-
monium sulfate (step 2), mainly at step 2. Purified proteinase In contained no Bz-Arg-NAn-hydrolytic activity Protease I1 was completely inhibited by incubation with TosLysCH,CI and diisopropylfluorophosphate [29]. Proteinase In was weakly inhibited by the former but not by the latter. Protease 11 was competitively inhibited by benzamidine (Ki, 0.6 mM) [29], but not by proteinase In at 1 mM. Proteinase In thus differs from protcase TI. Although proteinase In hydrolyzes fluorogenic substrates for trypsin, its activity appears markedly affected by an adjacent amino-acid sequence lying N-terminal of the arginine or lysine residue. Hydrophobic amino-acid residues such as proline and leucine may be required for maximal activity, and hydrophilic residues such as glycine, lysine, serine and threonine apparently markedly reduce activity. Fluorogenic substrates for chymotrypsin and elastase were not hydrolyzed. Proteinase In may thus belong to the trypsin-like proteinases, but differ from serine enzyme since diisopropylfluorophosphate and phenylmethylsulfonyl fluoride had no effect. Leupeptin, antipain and aprotinin strongly inhibited proteinase In, while STI, benzamidine and pentamidine isethionate did not. It thus follows that proteinase In may be a unique trypsin-like proteinase. Kato et al. (Kato, M., Irisawa, T. and Muramatu, M., unpublished results) found that various substituted phenyl esters of GMCHA strongly inhibit the growth of E. coli, this apparently being correlated with a hydrophobic or steric factor of the substituent, and from inhibition experiments on DNA synthesis they suggested the possible involvement of a trypsin-like proteinase in cell growth of E. coli. A Boc-ValPro-Arg-NH-Mec-hydrolytic activity, proteinase In: has been found just prior to the initiation of DNA replication in the cell cycle of E. coli synchronized by phosphate starvation. Inhibition of proteinase activity by GMCHA-OPheBu, a representative GMCHA ester, has been noted to prolong the DNA-replication period, thus retarding the course of the cell cycle (Kato, M., Irisawa, T., Morimoto, Y. and Muramatu, M., unpublished results). As shown in Table 4, proteinase In was strongly inhibited by various substituted phenyl esters of GMCHA, it was competitively inhibited by GMCHAOPheRu. The order of inhibitory effect on proteinase In was roughly correlated with that of their effects on E. coli growth. These effects on proteinase appear correlated with a hydrophobic or steric factor of substituent. These results appear to provide additional evidence for the above view that proteinase In is involved in DNA replication. Prouty and Goldberg rcported that pentamidine isethionate inhibited the growth of E. coli and macromolecular synthesis in E. coli [15]. It was also noted that pentamidine isethionate dose-dependently inhibit this growth and the uptake of [3H]thymidine into DNA. However, the effects were transient and reparable and differed from those of GMCHAOPheRu (Kato, M., Irisawa, T. and Muramatu, M., unpublished results). That proteinase In is strongly inhibited by GMCHA esters, but not by pentamidine isethionate may be the reason for this difference. The inhibitory effects of the latter may arise from some other mechanism. The specific intracellular substrate for proteinase In, and its role in the initiation of DNA replication in E. coli cells are points that remain to be determined.
REFERENCES 1. Goldberg, A. L. & John, A. C. St. (1976) Annu. Rev. Biochhem. 45. 747 - 803.
1014 2. 3. 4. 5. 6. 7. 8. 9. 10.
11. 22. 13. 14. 15.
Mount, D. W. (1980) Annu. Rev. Genet. 14, 279 -319. Wickner, W. (1979) Annu. Rev. Biochem. 48, 23-45. Davis, B. D. & Tai, P. C. (1980) Nature 283,433-438. Cavard, D. & Lazdunski, C. (1979) Eur. J . Biochem. 96, 529533. Watsen, D. H. & Sherratt, D. J. (1979) Nature 278, 362-364. Roberts, J. W., Roberts. C. W. & Craig, N. L. (1978) Proc. Nut/ Acad. Sci. U S A 75,4714-4718. Little, J . W., Edmiston, S. H., Pacelli, L. Z. & Mount, D. W. (1980) Proc. Nut1 Acud. Sci. USA 77,3225 - 3229. Lazdunski, A. M. (1989) FEMSMicrobiol. Rev. 63, 265-276. Sugimura, K. & Nishihara, T. (1988) 1. Baeteriof. 170, 56255632. Sugimura, K. & Higashi, N. (1988) J. Bacteriol. 170,3650- 3654. Lee, C. S., Park, W. J . , Jung, E. M., Choi, K. H., Ha, D. B. & Chung, C. H. (1991) Biochem. Znt. 23, 1155-1163. Vaithilingam, I. & Robert, A. (1989) Biochem. Int. 19, 12971307. Seol, J. H., Woo, S. K., Jung, E. M., Yoo, S. J., Lee, C. S., Kim, K., Tanaka, K. & Ichihara, A. (1991) Biochem. Biophys. Res. Commun. 176,730-736. Prouty, W. F. & Goldberg, A. C. (1972) J. Biol. Chem. 247, 3341 - 3352.
16. Rossmam, T., Norris, G. & Troll, W. (1974) J . Biol. Chem. 249, 3412- 3417. 17. Ito, K. (1977) J . Bacteriol. 132, 1021 -1023. 18. Ito, K. (1978) Biochem. Biophys. Res. Commun. 82, 99 - 107. 19. Davis, B. D. & Mingioli, E. S. (1950) J . Bacteriol. 60, 17-28. 20. Kanaoka, Y., Takahashi, T., Nakayama, H., Takada, K., Kimura, T. & Sakakibara, S. (1977) Chem. Phurrn. Bull. (Tokyo) 25, 3126-3128. 21. Erlanger, B. F., Kokowsky, N. & Cohen, W . (1961) Arch. Biochem. Biophys. 95,271 -278. 22. Lowry, 0. H., Rosebrough, N . J., Farr, A . L. & Randall, R. J. (1951) J . Biol. Chem. 193, 265-275. 23. Andrews, P. (1964) Biochem. J. 91, 222-223. 24. Pujimoto, Y., Moriya, H., Yamaguchi, K . & Moriwaki, C. (1972) J . Biochem. (Tokyo) 71,751 -754. 25. Yamato, I., Anraku, Y. & Hirosawa, K. (1975) J . Biochem. (Tokyo) 77,705-738. 26. Osborn, M. J.; Gander, J. E.; Parisi, E. & Carson, J. (1972) J . Biol. Chem. 247, 3962 - 3972. 27. Zipser. D. (1963) J. Mol. Biol. 7, 113-121. 28. Malamy, M. &Horeker, B. L. (1964)Biochem.J . 3, 1893-1897. 29. Pacaud, M. & Richaud, C. (1975) .I. Biol. Chern. 250, 77717779.
