A New Blood-coagulating Protease in Mitochondrial Membranes of

0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. THE JOURNAL OF BIOLCGICAL CHEMISTRY Vol. 269, No. 23, Issue of June 10, pp. 16229-16235, 1994

Printed in U.S.A.

A New Blood-coagulating Protease in Mitochondrial Membranes of Rat Submaxillary Glands PURIFICATION AND CHARACTERIZATION OF PROTEASE AND ITS BLOOD-COAGULATING ACTIVITY*

(Received for publication, November 8, 1993, and in revised form, February 26, 1994)

Dwaipayan BharadwajS, Mousumi Sinha Roy$, Dipak Bose, and Ratha N. HatiP From the Indian Institute of Chemical Biology, 4, Raja S.C. Mullick Road, Jadaupur, Calcutta 700032, West Bengal, India

An integral membrane protease was solubilized and purified to homogeneity from rat submaxillary mitochondria. The purified enzyme could coagulate rabbit plasma. The molecular mass of the enzyme is 22 kDa on SDS-polyacrylamide gel electrophoresis under reducing conditions and 24 kDa on gel filtration on a Sephadex 6-100 column. Its isoelectric point is 4.2-4.25. Enzyme activity is strongly inhibited by diisopropyl fluorophosphate, soybean trypsin inhibitor, benzamidine, aprotinin, and antipain, suggesting the enzyme as a serine protease. Its pH optimum for activity is 8.5. Zn2+ is strongly inhibitory; at 1 m~ concentration it produced 72% inhibition. The enzyme is active toward different synthetic substrates (p-nitroanilide derivatives) containing Arg at the PI position with blocked N H 2 terminus. KJK, was highest with the substrate N-Bz-Pro- Arg-pNa (where Bz is benzoyl and pNA is paranitroanilide).

The purified enzyme coagulates rabbit plasma in a dose-dependent manner. Plasma coagulation by the enzyme is completely blocked in the presence of aprotinin or soybean trypsin inhibitor, suggesting that protease activity is required for this coagulation reaction. Anti- body raised against the purified enzyme inhibits the plasma coagulation initiated by the enzyme. The enzyme can correct the prolonged clotting time of factor X-deficient human plasma but is unable to convert purified fibrinogen to fibrin clots, indicating factor &,-like activity of the enzyme. The enzyme has the ability to activate prothrombin. Several properties of the enzyme distin- guish it from other reported submaxillary proteases.

Submaxillary tissue is a rich source of several soluble proteases along with some physiologically active peptides (1-8). Some proteases have been correlated with the processing of naturally occurring precursor proteins (9, 10). Kallikrein (2), tonin (31, thiol-activated T-kininogenase (4), RSP-V (5), and esterase B (6) are examples of some well studied soluble proteases in submaxillary tissue. Despite several investigations on a large number of soluble proteases, work on intracellular membrane proteases from this tissue has not been reported. Growing appreciation in the field of intracellular membrane proteases has emerged when it is known that these proteases are involved in several cellular events like peptide processing, posttranslational modifications in the release of biologically

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertzsement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Scientific and Industrial Research. $ Recipient of senior research fellowship awarded by the Council of

8 To whom correspondence should be addressed.

active peptides from precursors, removal of signal peptides, and hydrolysis of proteins associated with membrane.

We observed that mitochondrial membrane of rat submaxillary gland contained significant protease activity (11, 12). The enzyme strongly associated with the mitochondrial membrane was solubilized by deoxycholate treatment and purified to homogeneity. Interestingly, it was found that the purified enzyme could coagulate rabbit plasma very effectively. Most of the factors involved in blood coagulation, which comprises a series of cascade reaction, have high molecular masses whereas the molecular mass of the protease tested here was 22 or 24 ma. This enzyme was also different from the established tissue factors (13) that are known to participate in the extrinsic pathway of blood coagulation. The plasma-coagulating activity of a submaxillary protease has not been reported earlier.

In this communication, we describe the purification of a mitochondrial membrane protease from rat submaxillary gland, characterization of the purified protein, and its ability to coagulate rabbit plasma.

EXPERIMENTAL. PROCEDURES Materials

DEAE-cellulose, sodium deoxycholate, TEMED,’ acrylamide, Phar- malyte 3-10, molecular weight marker proteins, isoelectric point marker proteins, enzyme inhibitors, synthetic substrates, NJV-meth- ylenebisacrylamide, bovine fibrinogen, and bovine thrombin were purchased from Sigma. Sephadex G-100 was from Pharmacia (Uppsala, Sweden). BAPNA was purchased from Boehringer Mannheim. Human plasmas with deficiencies of coagulation factors were obtained from George King Biomedical. Plasma was supplied in dry ice.

Enzyme Preparation

Preparation of Mitochondrial Fraction from Rat Submaxillary Gland

Sprague-Dawley rats (weighing 170-225 g) of either sex were used. A 10% homogenate of rat submaxillary tissues was made with 10 mM phosphate buffer (pH 7.4) containing 300 mM sucrose and 1 m~ EDTAin a motor-driven, Teflon-pestled Potter-Elvehjem homogenizer. The homogenate filtered through a gauze was centrifuged in a cold centrifuge at 1000 x g for 10 min. The pellet suspended in the same buffer was recentrifuged at 1000 x g for 10 min. The pooled supernatant was centrifuged at 12,000 x g for 30 min, and the pellet was washed three times with homogenizing buffer.

Purification of Mitochondria

gradient of sucrose between 1 and 2 M (14). Mitochondria were purified from a 12,000 x g pellet on a continuous

~~~~~ ~~~ ~ ~

The abbreviations used are: TEMED, N,NJV,iV’-tetramethylethyl- enediamine; BAPNA, N-benzoyl-DL-arginine paranitroanilide; BAEE, N-benzoyl-L-arginine ethyl ester; TAME, p-tosyl-L-arginine methyl ester; Bz, benzoyl; t-Boc, t-butoxycarbonyl; SBTI, soybean trypsin inhibitor; TLCK, Ne-p-tosyl-L-lysine chloromethyl ketone; TPCK, L-l-tosyl- amido-2-phenylethyl chloromethyl ketone; pNA, paranitroanilide; HPLC, high performance liquid chromatography; PAGE, polyacrylamide gel electrophoresis.

16229

16230 Blood-coagulating Protease in Mitochondrial Membrane

KC1 Deatment of Mitochondria The purified mitochondria were treated with 10 mM phosphate buffer

(pH 7.4) containing 1 M KC1 for 3 h with occasional stirring and then centrifuged at 105,000 x g for 1 h in an ultracentrifuge (Beckman L5-5OB). The pellet was resuspended, and the process was repeated.

Solubilization of Protease The pellet obtained after KC1 extraction was treated with 0.5% (w/v)

deoxycholate in 10 mM phosphate buffer (pH 7.4) containing 0.15 M KC1 and kept overnight at 4 “C. The supernatant obtained by centrifuging the mixture at 105,000 x g for 1 h in the ultracentrifuge was used for purification.

Purification Steps The enzyme at different steps in preparation was concentrated using

an Amicon, Inc., PM-10 filter. All steps of enzyme preparation were carried out at 4 “C.

Step 1: Ammonium Sulfate Precipitation-Solid ammonium sulfate was added to make 45% ammonium sulfate saturation of the solubilized enzyme which, after 15 min, was centrifuged at 15,000 x g for 30 min. The ammonium sulfate saturation of the supernatant was raised to 75%. The precipitate appearing at 45-75% saturation was dissolved in 50 ITLM Tris-HC1 (pH 7.0) and dialyzed for 5 h against 50 m~ Tris-HC1 buffer (pH 7.0) containing 0.05% deoxycholate with two changes.

Step 2: Sephadex G-100 Gel Chromatography-The dialyzed material centrifuged at 10,000 x g for 5 min was loaded on a Sephadex G-100 column (1.6 x 80 cm), which was equilibrated and eluted with 50 mM Tris-HC1 buffer (pH 7.0) containing 100 mM NaCl and 0.1% deoxycholate.

Step 3: DEAE-cellulose Chromatography-The protease fraction was concentrated and dialyzed against 50 mM Tris-HC1 buffer containing 0.001% deoxycholate (pH 7.5) for 5 h with two changes. The sample was put on a DEAEkellulose column (1.2 x 15 cm) equilibrated with the dialyzing buffer. The column was eluted initially with the equilibrating buffer followed by a gradient between 50 mM Tris-HC1 buffer and 500 m~ Tris-HC1 buffer (pH 7.5).

Step 4: Native PAGE-The concentrated protease fraction from the ion-exchange column was dialyzed for 4 h against distilled H,O with two changes containing 0.001% deoxycholate. The sample was then put into gel tubes for electrophoresis (15) in non-denaturing conditions. After electrophoresis, each gel was cut into 2-mm segments. Each segment was taken in 0.5 ml of 50 mM Tris-HC1 buffer, pH 7.0, disrupted, and kept for 3 h at 4 “C with occasional stirring. After centrifugation, the activity was assayed in the supernatant. SBTI-sensitive (>80% inhibition) protease-containing fractions were concentrated.

Step 5: HPLC-The material from step 4 was passed through a gel filtration HPLC column (TSK G2000 SW, 30 x 0.75 cm), which was equilibrated and eluted with 250 mM Tris-HC1 buffer, pH 6.7, containing 0.001% deoxycholate.

SDS-PAGE SDS-PAGE was carried out according to Laemmli (15) under reduc-

ing conditions in 10% gel. Gels were stained with Coomassie Brilliant Blue R-250.

Isoelectric Focusing This was carried out in an LKB 2117 Multiphor electrophoresis sys-

tem with a pH gradient of 3-10. The protein sample and PI markers were applied directly to the gel and focused at 2000 V until complete. The gel was stained with Coomassie Brilliant Blue G-250. The PI of the purified protease was calculated by comparison with the electrophoretic mobility of the PI markers.

Assay of Enzymes

Protease Activity Hydrolysis of BAPNA-The enzyme activity in all experiments (un-

less stated) was assayed using BAPNA substrate by the method of Erlanger et al. (16) with slight modifications. The reaction mixture (1.0 ml) contained 90 mM Tris-HC1 buffer (pH 8.0), 1 mM BAPNA, and a suitable volume of enzyme preparation. The production of p-nitroaniline was measured at 410 nm in a Pye Unicam spectrophotometer. The enzyme activity was expressed as nanomoles ofp-nitroaniline liberated per 30 min. The activity against p-nitroanilide derivatives of different synthetic substrates was assayed similarly.

BAEE Hydrolysis-The enzyme activity in column eluates during purification was monitored by measuring BAEE hydrolysis (17) a t pH 8.0. The activity was measured spectrophotometrically at 253 nm.

Succinic Dehydrogenase

Kearney (18) with slight modifications.

Plasma Coagulation Coagulating activity was measured essentially by following the

method of Kamiguti et al. (19). Plasma was prepared from citrated blood collected from healthy adult white Belgian rabbits (2.0 -c 0.3 kg). Clot- ting was studied in the reaction mixture (0.3 ml) containing 0.1 ml of plasma, 8 mM CaCl,, and a suitable volume of enzyme or saline. Clotting time was recorded after the addition of CaCl,. In controls, clotting time was measured in the absence of enzyme. Clotting times of human plasmas deficient in clotting factors were determined similarly. Fibrinogen clotting was assayed similarly except that purified fibrinogen (600 pg) was used in place of plasma.

Kinetic Studies Kinetic parameters, K,,,, V,,, and Kc,, were calculated from Eadie

and Hofstee plots, which were done in a computer using “Sigma Plot.”

Preparation of Polyclonal Antibody Antiserum against the purified protease was raised in two rabbits by

four successive injections to each. The immunization schedule comprised four subcutaneous injections at days 0, 7, 15, and 35. The first and second injections were given in the thumbpads using 300 and 150 pg of protein, respectively, in Freund’s complete adjuvant. The last two injections comprised 100 pg of protein each in incomplete Freund’s adjuvant. At the 10th day after the last injection, blood was drawn by cardiac puncture under aseptic conditions. Serum prepared from pooled blood was stored at -70 “C until use. Non-immune serum was prepared from the rabbits before immunization started.

Partial Purification of Antiserum The antiserum and non-immune serum were partially purified by

ammonium sulfate (045% saturation) precipitation. The precipitate dissolved in 50 ITLM Tris-HC1, pH 7.5, was dialyzed overnight against the same buffer. Protein was determined by the Bradford method (20).

Electron Microscopic Study The sample was negatively stained with 2% phosphotungstic acid

(pH 7.01, dried in air, and observed in a Jeol, Ltd. lOOCX transmission electron microscope at an operating voltage of 60 kV.

The activity was assayed by following the method of Singer and

RESULTS

We observed significant protease activity as evidenced by BAPNA hydrolysis in the mitochondrial fraction (12,000 x g pellet) of rat submaxillary glands. Mitochondria were purified (14) from the pellet on a continuous gradient of sucrose between 1.0 and 2.0 M. The fraction (at about 1.18 g/ml) enriched in succinic dehydrogenase activity (70%) also contained most of the protease activity. The fraction when checked by negative staining under the electron microscope showed 100% purity of the mitochondria.’ Glucose-6-phosphatase and acid protease activities were not observed in this fraction. The protease associated with the membrane after extraction of this fraction with 1.0 M KC1 was solubilized by 0.5% deoxycholate and purified in several steps as mentioned under “Experimental Proce- dures.”

Purification-The results of purification are summarized in Table I. Solubilization of the enzyme by deoxycholate produced a 65% increase in activity, indicating that part of the enzyme was buried inside the membrane. Deoxycholate itself could not alter the activity of the purified enzyme. On DEAE-cellulose chromatography (Fig. l), a small activity appeared just after the start of the gradient whereas major activity eluted later on was further purified. This fraction when electrophoresed on polyacrylamide gel under non-denaturing conditions was resolved into two active fractions: a slower migrating enzyme (A) sensitive to SBTI and a faster migrating enzyme ( B 1 resistant to SBTI (Fig. 2). The enzyme fraction ( A ) was resolved into six

D. Bharadwaj, M. S. Roy, D. Bose, and R. N. Hati, unpublished observation.

Blood-coagulating Protease in Mitochondrial Membrane

TABLE I Purification ofprotease from mitochondrial membrane of rat

submaxillary gland Activity toward BAPNA at pH 8.0 is shown below.

steps T o t a l T o t a l Specific Purifica- Recovery protein activity activity tion of activity

mg KC1-extracted pellet 372.6 5168 13.87 Deoxycholate-soluble 208.6 8515 40.8

Ammonium sulfate 18.4 4212 228.9 extract

SephadexG-100 precipitation

8.0 3877 484.6 DEAE-cellulose 3.3 2680 812.1 Native PAGE HPLC

0.07 184.4 2634.3 0.02 151.4 7570.0

-fold %

1 100 2.94 164

16.5 81.5

34.9 75.0 58.55 51.8

189.9 3.5 545.8 2.9

0.3 1

FRACTION NUMBERS

FIG. 1. DEAE-cellulose chromatography of the pooled fractions of the protease from Sephadex G-100 chromatography. The experimental procedures are mentioned under “Experimental Proce- dures.” Fractions (1-29) containing equilibrating buffer were followed by gradient.

Segment m

FIG. 2. Non-denaturing PAGE elution profile of the enzyme. The methods are described under “Experimental Procedures.” 100-150 pg of enzyme preparation from DEAE-cellulose chromatography was put in each gel for electrophoresis. The protease activities extracted from gel slices were assayed by using BAPNA substrate in the presence and absence of 50 pg of SBTI. Segments 21 and 22 (strongly inhibited by SBTI) were pooled together.

peaks on gel filtration HPLC (Fig. 3), among which onlypeak B possessed protease activity. The final preparation had a specific activity of 7570 with a 545.8-fold purification over the KCl- extracted pellet (Table I). The results of purification represent- ing the data of a typical experiment were reproduced with different batches of the preparation.

0.3(

5 N o.1: 0 OD

0

0 d

0.0

3

i

16231

ic Time (rnin)

FIG. 3. HPLC of the enzyme. The methods followed were as mentioned under “Experimental Procedures.” The enzyme activity of each peak was measured by studying BAPNA hydrolysis.

“” A -B-. C D E F G K D a 5

5

6

9 4

0. I

3.2

FIG. 4. SDS-PAGE of the enzyme. Lane A, 7 pg of purified protease; lane B, 20 pg ofpeak B (Fig. 2); lanes C-G, molecular mass standards. Molecular mass standards used were: bovine serum albumin, 66 kDa; egg albumin, 45 kDa; glyceraldehyde-3-phosphate dehydrogenase, 36 kDa; carbonic anhydrase, 29 kDa; trypsinogen, 24 kDa; SBTI, 20.1 kDa; and p-lactalbumin, 14.2 kDa. Lane C, SBTI; lane D, carbonic anhydrase; lane E, egg albumin; lane F , bovine serum albumin.

Homogeneity-The final preparation of the enzyme showed a single protein band on SDS-PAGE under reducing conditions (Fig. 4). We found that high molarity of phosphate buffer inhibited the activity to a great extent. The purified enzyme (peak B in Fig. 3) when chromatographed on the same gel filtration column in HPLC in 250 mM phosphate buffer, pH 6.7, eluted as a single peak. A single precipitation line produced by cross-reaction of the enzyme with the antisera against the purified enzyme also suggested the apparent homogeneity of the purified protein.

Characterization of Purified Protease-The molecular mass of the enzyme was found to be 22 kDa when determined from its electrophoretic mobility in SDS-gel electrophoresis (using the data from Fig. 4) and 24 kDa as determined from the elution position of the enzyme on a Sephadex G-100 column. On isoelectric focusing, the PI of the enzyme was calculated to be 4.2-4.25. The enzyme is not a glycoprotein as evidenced by negative periodic acid-Schiff staining. The enzyme shows significant activity from pH 7.0 to 9.5 with maximum activity a t pH 8.5 (Fig. 5) . Table I1 illustrates that diisopropyl fluorophos-


phate, benzamidine, aprotinin, antipain, and SBTI are strong inhibitors. Iodoacetamide, N-ethylmaleimide, and TPCK inhibit the enzyme activity moderately. Other reagents tested (Table 11) were without any significant effect. TLCK, an inhibitor of trypsin-like proteases, could not alter the protease activity. Among the metal ions tested (data not shown) Zn2+ (1 mM) produces 72% inhibition of the activity. K', Na+, Cu2+, M P , and Ca2+ have no significant effect. Co2+ and Mn2+ (2 mM each) inhibit the enzyme activity by 42 and 44%, respectively. Tonin from submaxillary glands is inhibited by Cu2+ (21).

The substrate specificity of the enzyme is shown in Table 111. The enzyme is active toward different synthetic substrates containing At-g at the P, position with a blocked COOH-terminal end and also with a blocked NH,-terminal end. No other amino acids including another basic residue, Lys at P,, is cleaved by the enzyme.

The esterase activity of the enzyme was studied using BAEE. TAME, and benzoyl-arginine methyl ester. The enzyme hydro- lyzes BAEE but fails to hydrolyze TAME or benzoyl-arginine methyl ester. In controls, trypsin could hydrolyze BAEE and

m E loor

(pH 5.0-6.0); A, phosphate (pH 6.0-7.5); 0, Tris-HC1 (pH 7.5-9.5); 0, FIG. 5. pH profile of the enzyme. The buffers used were: 0, acetate

NaOH glycine (pH 9.5-11.5). Other details are described under "Experi- mental Procedures."

Inhibitors Inhibition

TAME effectively. The insensitivity toward TAME singles out the enzyme from many reported submaxillary proteases that can hydrolyze TAME at varying rates (6).

Table IV describes kinetic analyses of some synthetic substrates that were properly hydrolyzed by the enzyme. KJK,,, was highest in N-Bz-Pro-Phe-Arg-pNA; the next highest value was obtained in the hydrolysis of N-t-Boc-L-Leu-Gly-Arg-pNA. The enzyme shows similar KJK,,, values with all other substrates except L-BAPNA and DL-BAPNA. Kc, of the enzyme is also maximal in N-Bz-Pro-Phe-Arg-pNA hydrolysis. Other kinetic parameters like V,, and K,,, are also displayed in Table IV

Plasma-coagulating Activity of Enzyme-The purified enzyme coagulates plasma in a dose-dependent manner (Fig. 6). Significant plasma clotting is observed even with 1 pg of enzyme. We observed that plasma coagulation by the enzyme is completely blocked in the presence of 5 kallikrein-inactivating units of aprotinin or 10 pg of SBTI. But aprotinin (up to 50 kallikrein-inactivating units) or SBTI (up to 50 pg) had no influence on the control plasma-clotting time. To locate the site of enzyme action, studies were carried out with human plasmas deficient in clotting factors. The enzyme corrects the clotting time of plasma deficient in factors X, E, VII, XI, and XII. The coagulation of factor X-deficient plasma by the enzyme (Fig. 6) shows that the prolonged clotting time of the deficient plasma is shortened to a great extent by the enzyme in a dose-dependent manner.

Immunological Characterization-Cross-reaction of the antiserum raised against the purified protease with the enzyme was observed by immunodiffusion. Precipitation was observed up to 0.5 pg of enzyme. The antiserum showed specificity with the enzyme only: it revealed no cross-reaction with trypsin, chymotrypsin, or thrombin. Fig. 7 demonstrates that the plasma-clotting time is lengthened when the enzyme is preincubated with varying amounts of partially purified antibody, illustrating that the antiserum strongly inhibits the plasma coagulation elicited by the enzyme. But plasma coagulation remains unaffected on preincubation of the enzyme with partially purified non-immune serum.

TABLE I1 Effect of various inhibitors on protease activity

Enzyme was preincubated with different inhibitors for 10 min at 37 "C. Inhibition (%) was calculated from the activity in the absence of any inhibitor. 0.4 pg of enzyme was used in each assay. Activity represented the value of 60 nmol of p-nitroaniline liberated per 30 min. D m , dithiothreitol; p-ME, P-mercaptoethanol; pCMB, p-chloromercuribenzoate; LBTI, lima bean trypsin inhibitor; DFP, diisopropyl fluorophosphate; PMSF, phenylmethylsulfonyl fluoride; KIU, kallikrein-inactivating units.

f DTT

0.5 m~ 1.0 mM

90

4.4 2.7

P-ME 0.5 mM 2.1 1.0 mM 2.0

1.0 mM 54.3 5.0 mM 51.2

1.0 mM 54.0 5.0 mM 55.1

0.5 mM 0.9 1.0 mM 0.4

1,lO-Phenanthroline monohydrate (1.0 m ~ ) 12.0 pCMB (1.0 m ~ ) 18.6 Puromycin

25.0 pg 15.7 50.0 pg 14.9

Pepstatin A (50.0 pg) 0.0 Chymostatin (50.0 pg) 18.0

Iodoacetamide

N-Ethylmaleimide

EDTA

Inhibitors Inhibition

%

SBTI 5.0 pg 10.0 10.0 pg 40.2 50.0 UP 85.6

LBTI , -

50.0 pg

Aprotinin 5.0 KKJ 10.0 KIU

Benzamidine 50.0 pg 100.0 pg

0.5 mM

100.0 pg

DFP

27.8 29.4

89.4 95.0

81.6 89.1

70.1 1.0 mM 81.2

Bestatin (50.0 pg) 6.7 PMSF (1.0 m ~ ) 1.1 Antipain (50.0 pg) 76.5 Bacitracin (50.0 pg) 9.1

TPCK (1.0 mM) TLCK (1.0 mM)

48.7 0.1

Blood-coagulating Protease in Mitochondrial Membrane 16233

T,w.E I11 Hydrolysis of different svnthetic suhstrates

Final concentration of each substrate used was 0.2 mM. The activity toward RAPNA (I)I.-BAPNA), which was used in most assays, was de- tined as 100%. 0.4 pg of enzyme was used in each assay. SUC, succinyl.

Position

NH, P, P, P, COOH Activity

c$

-

N Bz-nl. Arg pNA 100.00

N N Rz-I. Arg pNA 163.80

N Bz Pro Phe Arg pNA 215.10

N-t Rz Phe Val Arg pNA 215.90

N I3oc Leu Gly Arg pNA 214.40 Hz Val Gly Arg pNA 214.70

N-t-Roc-0 132 Ser Gly Arg pNA 214.70 I. Arg pNA 7.20

N SUC S

Phe pNA 2.16 Bzl Cys pNA 14.20

t Roc Ala Ala pNA 8.90 1) Val Lcu Lys pNA 5.50

N

~ ~

Tan[x IV Kinetic constants for hydrolvsis of diffrrent synthrtic suhstrntes

Purified protease (0.46 pg) was assayed with each of the indicated substrates using peptide concentrations of 10-100 p ~ . Experimental details and kinetic measurements are as mentioned under "Experimcn- tal Procedures."

Suhstrnte Vm.. K" K t Y , , K

yrnoll XIO-' I! s" x l V s" . t ~ - '

min IrnR

N-Bz-Pro-Phe-Arg-pNA 5.32 0.083 1.95 23.5 N-Hz-Phe-Val-Arg-pNA 4.12 0.080 1.50 18.6 N-Rz-Val-Gly-Arg-pNA 3.13 0.072 1.15 16.0 N-t-Roc-0-Rz-Ser-Gly-Arg-pNA 2.80 0.07 1.03 14.70 N-t-Roc-Leu-Gly-Arg-pNA 1.35 0.024 0.50 20.46 L-BAPNA 4.68 0.22 1.71 7.89 IL-BAPNA 1.99 0.25 0.71 2.86

I I I I I I

0 1 2 3 4 5 Enzyme (119)

600

50 0

- u

700 2 E

01

al

. - c

OI

600 0

V -

500

4 0 0

FIG. 6. Plasma coagulation by the protease. Clotting time was measured in the presence of varying amounts of the enzyme. 0, rabbit plasma (n = 4); 0, normal human plasma (n = 3); A., human plasma deficient in factor X In = 3) . Mean S.D. are presented in the data.

We observed that this enzyme (2.5-10 pg) cannot coagulate purified fibrinogen even up to 720 s whereas bovine thrombin (Sigma) clots fibrinogen efficiently (Fig. 8). But the enzyme can produce fibrin clots from the purified fibrinogen in the presence of isolated prothrombin (Fig. 8). The cleavage of prothrombin by the enzyme (Fig. 8, inset ) reveals that on preincubation (4 h )

500 r

Part lo l ly p u r l f l e d IqG ( p q l

Flc;. 7. Immunoinhihition study on the plasma coagulation by the enzyme. Varying amounts of partially pur~fird immunt. sera or non-immune sera ( a s mrntionrd~ wrrr usrd. 4 pg ofenzymr was usrd in each assay. Mean 2 S.1). In = 4 ) arr prrxented.

4 5 -

I I 1

2.5 5 0 IO 0

( p 9 ) FIG. 8. Fibrinogen clotting by the enzyme in the presence of

purified prothrombin and SDS-gel analysis on the cleavage of prothromhin hy the enzyme. In fihnnnprn clottinp. purified pru- thrombin 1 x 0 p g ~ was incuhated with thr v:ll?;ing amount?; of rnzymr (as stated) for 12 min a t 37 C heforr addition of fihrinoprn. In the absence of the enzymr, fihrinogen clotting was not ohsrmrd study carried up to 720 s ) by prothrombin. 0, enzyme + prothromhln; . thrombin. Data represent the average vnlur of thrre drtrrminnnts. For SDS-PAGE study, purifird prothrombin I 100 p g ~ was Incuhatrd wlth enz-me (8 pg). SDS-PAGE was carried out hy thr mpthod of Larmmli (15) in non-reducing conditions. 14 pg of protrin was put in grI iX.5'; I

for electrophorrsis. A, 0-h incuhation: R . 4-h incuhntion; f*, 2s-h Incu- bation. Prothromhin was purifird from hnvlnc. pl:wm;t I .I7 I

of prothrombin with the enzyme. four prominent hands are observed on SDS-PAGE. The hand corresponding to the position of prothrombin (72 kDa) decreases as compared with 0-h preincubation with the appearance of bands corresponding to nearly 57, 46, and 36 kDa. Preincuhation of the zymogen for a longer period (28 h ) produced a protein hand a t 36 kDa and two smaller cleavage products.

DISC'C'SSIOS

We report here the purification and characterization of a mitochondrial memhrane protease from rat submaxilla? gland, which shows strong blood-coagulating activity. Require- ment of a detergent for solubilization of the enzyme suggests its nature as an integral memhrane protein. The enzyme activity


TABLE V Comparison of submaxillary mitochondrial membrane (SMM) protease with different known submaxillary proteases

Subman-

kalli- kreinsf

Property Esterase B" T h i n b SalivainC Glandulind Rsp-vo dibular Protease Protease SMM A8 B8 protease

Molecular 25-27 (dimer) 28 30 23 25 (dimer) 34 28.2 30.4 22

Isoelectric 4.45 6.0-6.2 6.0 8.0-9.0 5.3 3.87-4.16

Optimum pH 9.5 6.8 9.2 NDh 10.0 DFP' Inhibition No inhibition Inhibition No inhibition ND PMSFJ Inhibition No inhibition Inhibition No inhibition Inhibition SBTI Inhibition Inhibition No inhibition ND Inhibition LBTI Inhibition ND No inhibition Inhibition Inhibition Aprotinin Inhibition No inhibition Inhibition Inhibition Inhibition

mass ( m a )

point 4.2-4.25

8.5

No inhibition

Slight inhibition

Inhibition Inhibition Inhibition

Inhibition Inhibition Inhibition

Inhibition Inhibition Inhibition No inhibition No inhibition

No inhibition Sensitive Sensitive Sensitive Not sensitive

TLCK Inhibition ND ND ND CU2' No inhibition Inhibition ND ND Hydrolysis Sensitive Sensitive ND ND

of TAME

a Ref. 6. Ref. 21.

e Ref. 1. Ref. 22.

e Ref. 5. f Ref. 45. 8 Ref. 46. h ND, not detected. ' Diisopropyl fluorophosphate. j Phenylmethylsulfonyl fluoride. ' Lima bean trypsin inhibitor.

is strongly inhibited by diisopropyl fluorophosphate, SBTI, benzamidine, aprotinin, and antipain (Table 11), suggesting that the enzyme is a serine protease. Lima bean trypsin inhibitor, which strongly inhibits many submaxillary proteases (5,6,22) has little effect on the enzyme. The inhibition produced by N-ethylmaleimide and iodoacetamide (1 m each), two thiol blocking agents, was not progressive up to 5 m, which is not characteristic of the behavior of an enzyme that depends directly on a thiol group for its catalytic activity (23). Thus, it appears that this enzyme is not a cysteine protease having active thiol groups, but thiol group(s) in the protein structure are required for its full catalytic activity. Though a significant inhibition by the chymotrypsin inhibitor TPCK was observed, the enzyme could neither hydrolyze the chymotrypsin substrates N-Suc-Phe-pNA (where Suc is succinyl) (Table 111) and N-acetyl-L-tyrosine ethyl ester nor was it affected by chymostatin (Table 111, a specific inhibitor of chymotrypsinlike and cysteine proteases (24). Lack of inhibition by pepstatin A and by EDTA and 1,lO-phenanthroline suggests that the enzyme is neither an acid protease nor a metalloprotease. Also, aminopep- tidase inhibitors like puromycin and bestatin (24) did not alter the activity significantly.

Kinetic studies (Table IV) show that all substrates tested, except L-BAPNA and DL-BAPNA, displayed high affinity with K,,, values lower than 0.1 nm. K, was lowest with N-t-Boc-Leu- Gly-Arg-pNA, a substrate conventionally used for the conver- tase assay of the complement system (25). But, both turnover rate constant (Kcat) and specificity constant (K,,JK,) were highest with N-Bz-Pro-Phe-Arg-pNA which is, incidently, a specific substrate for thrombin-like enzymes and plasma kallikrein (26, 27).

Peptides with Gly-Arg at the COOH terminus are hydrolyzed by factor Xa but are essentially resistant to plasmin, tissue, and plasma kallikreins. t-Boc-Ser-Gly-Arg-methylcoumarin- amide is one of the most specific substrates for factor Xa (28). Our enzyme that has high Kcat and KJK, values for the N-t- Boc-0-Bz-Ser-Gly-Arg-pNA substrate differs, in this respect, from plasmin, tissue, and plasma kallikreins. The enzyme does not show any clot lysis activity similar to plasmin-like enzymes.2

The substrate N-Bz-Phe-Val-Arg-pNA, which is efficiently hydrolyzed by thrombin, is insensitive to factor Xa and tissue kallikrein (29). Our enzyme has effective specificity toward the substrate (Table IV). Esnouf and Macfarlane (30) showed that factor Xa, unlike our enzyme, is resistant to SBTI, inhibited strongly by phenylmethylsulfonyl fluoride, and can hydrolyze TAME and is aprotinin-insensitive. All kallikreins hydrolyze various synthetic amino acid esters, but BAPNA is hydrolyzed by plasma kallikrein only (31,321. Moreover, it is reported that salivary kallikrein does not hydrolyze BAPNA (33), while our enzyme can hydrolyze BAPNA efficiently. Also, the molecular mass of kallikreins is different from that of our enzyme (Table VI.

A comparison of the properties of this enzyme with those reported for other known submaxillary proteases (Table V) demonstrates that the protease reported is a new one, hitherto unreported.

The preference of the protease toward some thrombin and factor Xa substrates (Table IV) induced us to study whether the enzyme possesses blood-coagulating activity. Fig. 6 clearly demonstrates the potency of this protease in coagulating plasma effectively. Inability of the enzyme to produce fibrin clots from purified fibrinogen suggests that it is not a thrombin- like enzyme. Tissue factor that initiates blood coagulation via the extrinsic pathway (13) differs from our enzyme in the following points. This factor is localized to plasma membranes of cells of many tissues (341, is a lipoprotein (131, has a larger molecular mass (13), and is also a glycoprotein (35). The pep- tidase activity associated with tissue factor is not required for its clotting action (36, 371, whereas plasma clotting by this enzyme is completely blocked by the protease inhibitors SBTI and aprotinin.

Our study with plasmas deficient in different clotting factors suggests a factor Xa-like activity for this enzyme. Factor Xa is a serine protease with the specific function of activating prothrombin and plays a pivotal role in the blood coagulation cascade system. Prothrombin activation by the enzyme has been demonstrated in our study (Fig. 8).

Prothrombin activation by factor Xa in the presence of factor V, phospholipid, calcium ion (38-401, and by some snake ven-

Blood-coagulating Protease

oms (41, 42) has been investigated by several workers who demonstrated that conversion of prothrombin to thrombin is a multistep reaction. Formation of intermediates has been docu- mented in prothrombin activation by all activating agents. In our system, cleavage of prothrombin led to the formation of several cleaved products (Fig. 8); one (36 kDa) corresponds to the molecular mass of thrombin. The formation of the product (about 46 kDa) in the activation process has not been reported. Further studies are necessary to elucidate the mechanism of prothrombin activation by our enzyme and its proper comparison with other systems.

The ability of this enzyme to activate prothrombin suggests a processing role. It may be that the enzyme functions as a processing enzyme in rat submaxillary mitochondria. The physiological importance of the extramitochondrial role of this enzyme in the blood coagulation process is yet to be elucidated. Different patterns of leakage of mitochondrial enzymes into blood in different pathogenic and abnormal physiological states have been reported (43,44). It may be that in abnormal physiological states when fibrin formation is enhanced, this enzyme leaches out of mitochondrial membranes into the blood to participate in the coagulation process.

Acknowledgments-We thank Prof. A. N. Bhaduri. Director of the Indian Institute of Chemical Biology, and Prof. J. J. Ghosh of Calcutta Science College for valuable suggestions in this work. The interest of Dr. K. Beaman of the University of Health Sciences, the Chicago Medical Center, in this work is also especially acknowledged. Finally, we thank H. N. Datta and S. Sahu for art work.

REFERENCES 1. Riekkinen, P. J., Ekfors, T. O., and Hopsu, V. K. (1966) Biochim. Biophys. Acta

2. Schachter, M. (1979) Pharmacol. Reu. 31, 1-18

4. Barlas, A,, Gao, X., and Greenbaum, L. M. (1987) FEES Lett. 218, 266-270 3. Ikeda, M., and Arakawa, K. (1984) Hypertension (Dallas) 6, 222-228

5. Ikeno, K., Ikeno, T., Kuzuya, H., and Ishiguro, I. (1986) J. Biochem. (Tokyo) 99,

6. Khullar, M., Scicli, G., Carretero, 0. A,, and Scicli, A. G. (1986) Biochemistry

8. Turkington, R. W., Males, J. L., and Cohen, S. (1971) Cancer Res. 31,252-256 7. Cohen, S. (1960)Proc. Natl. Acad. Sci. U. S. A. 46, 302-311

9. Greene, L. A,, Shooter, E. M., and Varon, S. (1969) Biochemistry 8,37353741 10. Server, A. C., and Shooter, E. M. (1976) J. Biol. Chem. 261, 165-173 11. Sinha Roy, M., Bharadwaj, D., and Hati, R. N. (1991) Cum Sci. (Bangalore) 60,

12. Sinha Roy, M., Bharadwaj, D., and Hati, R. N. (1991) Biochem. Int. 26,1035-

116,604-620

1219-1226

25,1851-1857

325-326

1041

in Mitochondrial Membrane 16235 13. Esnouf, M. P. (1984) in Human Blood Cougulation, Haemostasis and Throm-

bosis (Biggs, R., and Rizza, C. R., eds) 3rd Ed., pp. 46-56, Blackwell Scien-

14. Rickwood, D., Wilson, M. T., and Darley-Usmar, V. M. (1987) in Mitochondria: tific Publications Ltd., Oxford

A Practical Approach (Darley-Usmar, V. M., Rickwood, D., and Wilson, M. T., eds) pp. 3-5, IRL Press, Oxford

15. Laemmli, U. K. (1970) Nature 227,68&685 16. Erlanger, B. F., Kokowsky, N., and Cohen, W. (1961) Arch. Biochem. Biophys.

17. Schweart, G. W., and Takenaka, Y. (1955) Biochim. Biophys. Acta 16,570-575 18. Singer, T. P., and Keamey, E. B. (1957) Methods Biochem. Anal. 4,307-333 19. Kamiguti, A. S., Sousa, E., Silva, M. C. C., Morena, P., and Nahas, L. (1985)

20. Bradford, M. (1976) Anal. Biochem. 72,24%254 21. Schiller, P. W., Demassieux, S., and Boucher, R. (1976) Circ. Res. 39,629432 22. Riekkinen, P. J., Ekfors, T. O., Hollmen, T., and Hopsu-Havu, V. K. (1967)

23. Barrett, A. J., and Brown, M. A. (1990) Biochem. J . 271,701-706 24. Beynon, R. J., and Salvesen, G. (1989) in Proteolytic Enzymes, a Practical

Approach (Beynon, R. J., and Band, J. S., eds) pp. 241-249, IRL Press a t

25. Caporale, L. H., Gaber, S.-S., Kell, W., and Gotze, 0. (1981) J. Zmmunol. 126, Oxford University Press, Oxford

26. Powers, C . A,, and Nasjletti, A. (1982) J. Biol. Chem. 257, 5594-5600 1963-1965

27. Lottenberg, R., Christensen, U., Jackson, C. M., and Coleman, P. L. (1981)

28. Morita, T., Kato, H., Iwanaga, S., Takada, K., Kimura, T., and Sakakibara, S.

29. Orlowski, M., Flick, M. R., Rand, J., and Lesser, M. (1987) Arch. Biochem.

30. Esnouf, M. P., and Macfarlane, R. G. (1968)Adu. Enzymol. 30, 255315 31. Schachter, M. (1969) Physiol. Reu. 49, 50%547 32. Schachter, M. (1980) Pharrnacol. Rev. 31, 1-17 33. Kurosawa, N., and Ogita, Z. (1989) Electrophoresis 10, 189-194 34. Zeldis, S. M., Nemerson, Y., Pitlick, F. A,, and Lentz, T. L. (1972) Science 176,

36. Nemerson, Y., and Esnouf, M. P. (1973) Proc. Natl. Acad. Sci. U. S. A. 70, 35. Pitlick, E A. (1975) J. Clin. Inuest. 55, 175-179

95,271-278

Toxicon 23, 383391

Enzymologia 32, 97-109

Methods Enzymol. 80, 341361

(1977) J. Biochem. (Tokyo) 82,1495-1498

Biophys. 254, 156-169

766-768

310-314 37. Bjorklid, E., Storm, E., and Prydz, H. (1973) Biochem. Biophys. Res. Commun.

~~~ .~~

55.969-976 38. Aronson, D. L., and Menache, D. (1966) Biochemistry 6, 2635-2640 39. Mann, K G., Heldebrant, C. M., and Fass, D. N. (1971) J. Biol. Chem. 246,

40. Owen, W. G., Esmon, C. T., and Jackson, C . M. (1974) J. Biol. Chem. 249,

41. Jobin, F., and Esnouff, U. P. (1966) Nature 211, 873-878 42. Morita, T., Iwanaga, S., and Suzuki, T. (1976) J. Biochem. (Tokyo) 79, 1089-

43. Schmidt, F. N., and Schmidt, E. (1989) Clin. Chim. Acta 186, 253-264 44. Kamiike, W., Fujikawa, M., Koseki, M., Sumimura, J., Miyata, M., Ka-

washima, y., Wada, H., and Tagawa, K. (1989) Clin. Chim. Acta 185, 265- 270

45. Brandtzaeg, P., Cautvik, K. M., Nustad, K., and Pierce, J. V. (1976) BI: J. Pharrnacol. 56, 153-167

46. Kato, H., Nakanishi, E., Enjyoji, K., Hayashi, I., Oh-ishi, S., and Iwanaga, S. (1987) J. Biochem. (Tokyo) 102, 1389-1404

47. Mann, K G. (1976) Methods Enzymol. 46, 127-131

6106-6114

594-605

1108

Documents

A New Blood-coagulating Protease in Mitochondrial Membranes of