THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 258, No. 23, Issue of December 10. pp. 14106-14115,1963 Printed in U. S. A.

A Proenzyme from Chicken Plasma Similar to Human Plasma Prekallikrein’

(Received for publication, August 3, 1982)

Wolf-Dieter SchleuningS, Marius SudolQ, and Edward Reichll From The Rockefeller University, New York, New York 10021

We report the isolation of a specific protease zymo- gen from chicken plasma. The purification procedure involves barium citrate precipitation, ammonium sul- fate fractionation, removal of plasminogen and plas- min on lysine-Sepharose, followed by anion and cation exchange, and gel permeation chromatography. Based on quantitative radioimmunoassay the zymogen is present in plasma at a concentration of 160 mgfliter, and it is obtained by our procedure in highly purified form with a yield of 1.4%. The single polypeptide chain contains an NHz-terminal alanine residue.

The native molecule migrates in sodium dodecyl sul- fate-polyacrylamide gel electrophoresis with an appar- ent molecular weight of 84,000 under reducing con- ditions. It can be identified as an inactive proenzyme because it has very low amidolytic activity, does not react with the fluorescent active site titrant 4-methy- lumbelliferylp-guanidinobenzoate, and does not incor- porate radioactive [sH]diisopropylfluorophosphate. It is very susceptible to limited proteolysis which con- verts it to an active enzyme with trypsin-like specific- ity. The active enzyme, likewise a single polypeptide chain, migrates as a doublet with apparent molecular weights of 39,000 and 40,000. Its amidolytic activity with synthetic peptide substrates is at least 40-fold higher than that of the proenzyme, it reacts efficiently with 4-methylumbelliferyl p-guanidinobenzoate, and incorporates [sHldiisopropylfluorophosphate while undergoing irreversible inactivation.

The enzyme appears to be a reasonably efficient plasminogen activator in zymographic gels, but not in solution. With human high molecular weight kininogen as substrate the enzyme was about 26% as efficient as human plasma kallikrein. It lacks any plasminogen- independent proteolytic activity with other protein substrates, and it hydrolyzes small peptide substrates

* This work was supported by Grant CA 08290 from the National Cancer Institute and Grant BC 316 from the American Cancer Society. A preliminary account of some of this work has been pre- sented as part of a lecture at the 30th Colloquium of the Gesellschaft fur Biologische Chemie, April 26-28, 1979, in Mosbach, West Ger- many and has been published as part of a review article, Plasminogen Activator from Cultured Cells and from Blood Plasma (1). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduer- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Recipient of a fellowship from the Deutsche Forschungsgemein- schaft Schl 171/2 and NATO Research Grant 1339. Present address, Laboratoire Central d’Hematologie 1011 CHUV, Lausanne, Switzer- land.

J Partial support for this research was derived from the Merinoff Cancer Research Fund (to M.S.).

lf Present address, Friedrich Miescher Institut, Postfach 2543, CH- 4002 Basel, Switzerland.

designed for both human kallikrein and urinary uro- kinase, respectively. Inhibition studies with peptide chloromethyl ketones indicate enzymatic properties closer to human plasma kallikrein than to the human plasminogen activator urokinase (EC 3.4.21.31). The chicken plasma enzyme and the plasminogen activator from the conditioned media of Rous sarcoma virus- transformed chick embryo fibroblasts treated with tu- mor promoter are different by criteria of tryptic pep- tide maps, and amino acid composition and enzymatic specificity. The designations chicken plasma prekalli- krein plasminogen proactivator and chicken plasma kallikrein plasminogen activator are proposed for the zymogen and enzyme forms, respectively.

Using rabbit antibodies against the proenzyme we developed a solid phase immunoadsorption procedure that allowed us to isolate the protein with an overall yield of 11.4%.

Our interest in the fibrinolytic systems of the chicken has been prompted by a number of observations. These include the findings that cultured chick embryo fibroblasts, which normally produce little or no detectable PA,’ can be stimu- lated to form large amounts of enzyme by viral transformation (Z), by the tumor-promoting phorbol esters (3), and by tera- togenically active retinoids (4). Further, high rates of PA production appear to be correlated with tissue remodeling and cell migration at various stages of the chicken embryo life cycle (5). With these phenomena in mind it seemed desirable to characterize some of the plasminogen activators present in chicken tissue to obtain basic information that could be used to study the role of PA, and the regulation of its synthesis, in the context of tumorigenesis, tumor promotion, and embry- onic development in the chicken.

During a survey of chicken material based on the use of a zymographic procedure (6)) the presence of PA activity was observed in chicken plasma. We describe here the isolation and characterization of this factor that exists in plasma as a proenzyme which, on apparent autocatalytic activation, is converted to a serine protease with amidolytic, kinin-liberat- ing, and plasminogen-activating activity.

The abbreviations used are: PA, plasminogen activator; iPr,P-F, diisopropylfluorophosphate; MES, 4-morpholineethanesulfonic acid; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electropho- resis; Cbz, benzyloxycarbonyl-; MUGB, 4-methylumbelliferyl p-gua- nidinobenzoate hydrochloride; CK/PA, chicken plasma kallikrein plasminogen activator; CPK/PPA, chicken plasma prekallikrein plas- minogen proactivator; dansyl, 5-dimethylaminonaphthalene-1-sul- fonyl; PBS, phosphate-buffered saline; TPA/RSV/CEF/PA, plasmin- ogen activator from conditioned media of Rous sarcoma virus-trans- formed chick embryo fibroblasts treated with tumor promotor.

14106

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

Chicken Prekallikrein 14107

MATERIALS AND METHODS

Chemicals Chicken plasma containing 10% (v/v) of a 3.8% solution of triso-

dium citrate was obtained from Pel-Freeze Biologicals. Sepharose GB-CL, CM-Sepharose, and Sephacryl S-200 were pur-

chased from Pharmacia. Microgranular DEAE-cellulose (Whatman DE52) was from Whatman. All buffer reagents, benzamidine hydro- chloride, L-lysine, cyanogen bromide, and t-aminocaproic acid were obtained from Sigma, [3H]iPr$-F, Omnifluor, and 1,4-bis[2-(5-phen- yloxazolyl)]benzene were purchased from New England Nuclear. Synthetic bradykinin triacetate was purchased from Sigma. Plasmin- ogen was prepared according to the method of Deutsch and Mertz (7).

All reagents for polyacrylamide gel electrophoresis were obtained from Eastman Kodak or Bio-Rad Laboratories. All other chemicals were of the best analytical grade available.

Sera and powdered culture medium were from Grand Island Bio- logical Co.; all disposable plastic Petri dishes were from Falcon Plastics.

MUGB and 4-methylumbelliferone were purchased from Sigma. D- Pro-Phe-Arg-p-nitroanilide (S-2302, KABI) was from Ortho Diag- nostics. D-Glu-Gly-Arg-p-nitroanilide (S-2227, U B I ) was a generous gift of Dr. Goran Claesson, KABI Peptide Research Molndal, Sweden. Urokinase reference standard (2240 Ploug units/vial) was obtained from Leo Pharmaceutical Products, Denmark. 12-0-Tetradecanoyl phorbol 13-acetate was obtained from Dr. P. Borchert, University of Minnesota, Minneapolis. 7-(N-Cbz-glycyl-glycyl-argininamido)-4- methylcoumarin trifluoroacetate was from Bachem, Switzerland. Hu- man high molecular weight kininogen was generously provided by Drs. Brigitte Dittmann and Hans Fritz, University of Munich, West Germany.

Column Chromatographic Procedures All column chromatographic procedures were carried out in silicon-

ized glassware and with tubing and connectors of silicone rubber. Elution rates were maintained constant by using peristaltic pumps. Fractions were collected in 10-ml polystyrene tubes if not otherwise specified.

Measurement of Protein Concentration Protein concentrations were determined by the method of Lowry

et al. (8) or by absorbance measurements a t 280 nm if extinction coefficients were available. The extinction coefficient E% for plas- minogen was assumed to be 17.0 (9) and for IgG 11.8.

Electrophoretic Procedures Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was per-

formed on 10% gels employing the conditions described by Laemmli (10). Plasminogen activator activity was detected qualitatively by the zymographic method of Granelli-Piperno and Reich (6). Preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis was con- ducted in a Savant apparatus based on the specifications given by Hagen and Young (111, with a modification of the buffer combination of Laemmli (10); the electrode buffer 0.1 M Tris/glycine, pH 8.6, was replaced by 0.1 M Trismrate, pH 8.2. A 1-ml solution containing up to 25 mg of protein was fractionated in a gel (1.5 X 10 cm) at 50 mA (constant current). Cooling was provided by a continuous flow of tap water through the cooling jacket. The proteins emerging at the bottom of the gel were continuously removed with buffer (electrode buffer) at a rate of 12 ml/h and collected in 4-ml fractions. The active fractions were pooled and lyophilized. In another approach the active fractions were pooled and precipitated by adjusting them to 20% trichloroacetic acid in an ice bath; the precipitate was recovered by centrifugation and stored at -20 "C. Quantitative immunoelectropho- resis was performed according to the method of Laurel1 (12) using a concentration of 5% anti-CPK/PPA (purified IgG) in the agar, cor- responding to 5 pl of purified IgG/cm2 of gel area. Electrophoresis was performed in barbital buffer, p = 0.02, with a gradient of 2 V/cm in the gel at 15 "C for 18 h.

Preparation of Antibodies and Immunochemical Procedures New Zealand white rabbits received an initial injection of 250 pg

of highly purified antigen suspended in 50 pl of complete Freund's adjuvant in the popliteal nodes. After 5 weeks they were restimulated by 3 subsequent weekly intradermal injections of 100 pg of antigen

suspended in incomplete Freund's adjuvant. During the period 5-8 weeks after the initial injection the rabbits were bled weekly, and antibody titers were determined by double immunodiffusion in 1% agarose, using a buffer system composed of 0.05 M sodium veronal and 0.0073 M veronal a t pH 8.6.

Purification of the ZgG Fraction Solid ammonium sulfate was added to 100 ml of pooled antisera to

35% of saturation and the precipitate recovered by centrifugation. The pellet was redissolved in 10 ml of 10 mM Tris/HCl, 10 mM benzamidine, pH 8.6 ("DEAE-buffer"), and dialyzed overnight against the same buffer. The dialyzed material was applied to a DEAE- cellulose column (2.5 X 40 cm) equilibrated in DEAE-buffer. The proteins were eluted with a linear salt gradient with the reservoir containing 1 liter of DEAE-buffer which was 0.5 M in NaCl and the mixing chamber 1 liter of DEAE-buffer. The IgG-containing fractions were identified by SDS-polyacrylamide gel electrophoresis, pooled, and concentrated to 10 ml by ultrafiltration using an Amicon ultra- filtration cell equipped with a PM-30 membrane. When tested by immunodiffusion against native chicken plasma these antibodies yielded a single precipitin band and a line of identity with authentic highly purified proenzyme CPK/PPA.

Quantitative Activity Determinations Activity of plasminogen activator was determined in the following

three ways. I . By the Lysis of '251-Fibrin in the Presence of Plasminogen-This

was described by Unkeless et al. (2). Units of plasminogen activating activity were determined as follows: 4 wells of the '=I-fibrin plate were reserved for urokinase (EC 3.4.21.31) references standards (2,4, 6, and 8 X Ploug units/well). The reaction time of the assay was 1 h. From the radioactivity released by urokinase a standard curve was constructed, which was used as a reference to determine the activity of the unknown samples.

2. By the Hydrolysis of the Peptides D-Pro-Phe-Arg-p-nitroanilide (S-2302, KABI) or D-Glu-Gly-Arg-p-nitroanilide (S-2227, KABI) (I3)"The increase of absorbance was recorded at 405 nm in a Gilford spectrophotometer in thermostabilized (25 "C) plastic cuvettes.

The Michaelis constants were determined in 0.2 M Tris/HCl, pH 8.6, in a total volume of 1 ml. The amount of enzyme was 0.4 pg and the substrate concentration was varied in the range 10-3-10-6 M. K,,, was graphically derived from Lineweaver-Burk plots. For routine determinations of enzyme activity the same conditions were used except that the substrate concentration was set a t 2 X K,,,. One unit of enzyme activity was defined as the increase of absorbance in units a t 405 nm/mg of enzyme in 1 min at 37 "C under the conditions of substrate saturation.

3. A Second Rate Assay of Plasminogen Activator Was Based on Hydrolysis of Synthetic Fluorogenic Peptide Substrate 7-(N-Cbz-gly- cyl-glycyl-argininoamido)-4-methylcoumarin Trifluoroacetate-Con- ditions were as given by Zimmerman et al. (14). Fluorescence of the released 7-amino-4-methylcoumarin was monitored continuously with a Hitachi Perkin-Elmer MPF-2a spectrofluorometer equipped with a chart recorder. Activation and emission wavelengths were 382 and 455 nm, respectively.

Active Site Titration (15) Stock solutions were: 0.1 M sodium barbital adjusted to pH 8.6 by

the addition of 1 N HCL; 10 mM MUGB in dimethyl formamide diluted to 0.1 mM with 1 mM HCl just prior to use. The reaction mixture consisted of 480 pl of sodium barbital, 10 pl of MUGB stock, and 10 pl of sample. 1 mg of 4-methylumbelliferone was dissolved in 1 ml of dimethyl formamide and then diluted to 1 liter by addition of double distilled water (to serve as standard). Fluorescence was meas- ured in a Hitachi Perkin-Elmer MPF-2A fluorescence spectrometer using an excitation wavelength of 365 nm; emission was recorded at 445 nm.

Amino Acid Analysis Amino acid analysis was performed according to Moore and Stein

(16) with a Durrum D-500 amino acid analyzer equipped with a Digital Equipment PDP-8 computer. Methionine, serine, and threo- nine values were obtained by extrapolation to zero time from the 24-, 48-, and 72-h hydrolysis times. Values for valine, leucine, and isoleucine were obtained from the 72-h hydrolysis. Cysteine was determined as cysteic acid according to Hirs (17) or to Moore (18).

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

14108 Chicken Prehllikrein

Tryptophan content was estimated spectrophotometrically (19). A Cary spectrophotometer model 219 was employed to determine ab- sorption spectra.

NHz-terminal Amino Acid Determination The NHZ-terminal amino acid was determined according to the

procedure of Hartley (20) following the modifications of Gros and Labouesse (21). Acid hydrolysis was performed for 4 h at 110 "C. Dansyl-amino acids were identified by thin layer chromatography on polyamide layer sheets (5 X 5 cm) according to the method of Woods and Wang (22).

Carbohydrate Analysis Quantitative sugar analyses were performed by gas chromato-

graphic determinations of hexoses and 2,5-anhydrohemses deriva- tized as the alditol acetates (23, 24). A Beckman model GC 65 gas chromatograph, equipped with a 6 foot column (2 mm, inner diame- ter), was used in these studies. The column packing material consisted of 1% ECNSS-M on 60/80 mesh Gas-Chromosorb Q. The amount of sialic acid was determined by thiobarbituric acid assay using N-acetyl neuraminic acid as a standard (25).

Preparation of the Affinity Reagents Sepharose 4B CL was activated by the cyanogen bromide procedure

according to Porath and Kristiansen (26), CNBr being first dissolved in 1 ml of acetonitrile per g of CNBr, as described by March et al. (27). Lysine was coupled in 0.1 M NaHC03, pH 8.0, using 50 mg of lysine per ml of activated Sepharose. Purified IgG was coupled to activated Sepharose in 0.1 M MES, pH 6.5, for 48 h, at a ratio of 1 mg of IgG to 1 ml of swollen Sepharose.

Labeling with PH]iPr,P-F One hundred pg of proenzyme or enzyme dissolved in 200 p1 of 0.1

M ammonium acetate were mixed with 300 p1 of 0.5 M potassium phosphate, pH 6.8, and 100 pl of [3H]iPrzP-F stock solution (0.2 mg/ ml; specific activity, 1 Ci/mmol). The pH was adjusted by careful addition of 0.1 N NaOH and the mixture incubated in an ice bath for 30 min. The pH was again determined and, if necessary, readjusted to 6.8, a further 100 p1 of [3H]iPr~P-F stock solution were added, and the mixture was kept on ice for another 30 min. Finally, a third portion of [3H]iPrJ"F was added and the incubation was continued for 2 h. The mixture was then dialyzed against 0.01% sodium dodecyl sulfate in double distilled water and lyophilized.

Determination of Kininogenuse Activity Kininogenase activity was determined by incubating the enzyme

with highly purified human high molecular weight kininogen. The kinins formed were measured by monitoring the contractions of an isolated rat uterus preparation using a polygraph recorder by a variation of the method of Trautschold (28); 1 pg of CK/PA and 150 pg of human high molecular weight kininogen in 1 ml of 0.1 M Tris/ HC1, pH 8.5, M phenanthroline were incubated for 30 min at 25 'C. The reaction was stopped by heating the sample for 10 min at 95 "C. The kinins generated were quantified by comparing the activity with a standard curve, constructed by measuring the effects induced by 10-200 pg of synthetic bradykinin triacetate.

Iodination Procedure, Radioimmunoassay, and Tryptic Peptide Mapping

Protein was iodinated according to Krohn et al. (29). To 10 pl of (10 mCi, carrier-free) 100 p1 of a solution containing 100 pg of

protein were added; 25 pl of chloramine-T (6.6 X M in 0.4 M sodium phosphate buffer, pH 7.8) were added, the mixture mixed for 5 s , and then immediately loaded onto an 18-ml column of Sephadex G-25 (previously washed with PBS containing 5 mg/ml of bovine serum albumin and then washed and equilibrated with PBS). The specific radioactivity of 1Z51-protein obtained in this way was 6-8 pCi/ pg; it was diluted and stored at a concentration of 2000 cpmlpl.

A competitive radioimmunoassay was developed in order to deter- mine the absolute amount of CPK/PPA in fresh plasma. Purified rabbit anti-CPK/PPA IgG with an antigen binding capacity of 0.3 mg/ml (0.025 mg/mg of IgG) was used. All dilutions were in PBS supplemented with 0.5% of bovine serum albumin. The total assay volume was 1 ml. Incubations were carried out at 4 "C. The iodinated antigens were diluted to concentrations that yielded 20,000 cpm/100

pl. An immunoglobulin concentration was chosen such that 50% of the labeled antigen was bound in absence of unlabeled antigen during an incubation period of 24 h. The assay was started by adding, to 500 pl of PBS/O.5% bovine serum albumin in an Eppendorf tube, 200 pl of appropriately diluted antibody, 100 pl of iodinated tracer antigen (containing 20,000 cpm) and either 100 p1 of standard dilutions of nonradioactive CPK/PPA or 100 p1 of serial dilutions of chicken plasma, respectively. After incubation for 24 h immune complexes were precipitated by adding 100 pl of heat-killed and formalin-treated Staphylococcus aureus strain Cowan I, as described by Kessler (30). Bound and unbound radioactivity were separated by centrifugation at 13,000 X g for 20 min at 4 "C.

Bound and unbound radioactivity were measured in a Packard Auto-Gamma scintillation spectrometer. The ratio of bound/unbound antigen was expressed as logit values as suggested by Rodbard et al. (31) and plotted as a function of the logarithm of the concentration of the standard dilutions of the nonradioactive antigen. A standard curve obtained in this way is linear in this plot. The concentrations of CPK/PPA in chicken plasma were derived by comparing the logits of the bound/unbound radioactivity ratios of the samples containing serial dilutions of chicken plasma with the corresponding logarithm of the concentration on the standard curve.

The immunoaffinity isolation procedure was monitored by means of isotope dilution assay using 'l-labeled proenzyme. The radioiodi- nation of proteins in single polyacrylamide gel slices, tryptic digestion, and peptide mapping were performed as described by Elder et al. (32)

Reactivity with Natural Inhibitors Employing D-Pro-Phe-Arg-p-nitroanilide (S-2302 M I ) as a sub-

strate we have screened a number of natural inhibitors of plant and microbial origin to determine their possible interaction with CK/PA. The amidolytic activity of CK/PA was measured by monitoring the increase of adsorbance at 405 nm in a thermostabilized (25 "C) Gilford spectrophotometer. Concentration of the substrate was one-fourth, one-half, and three-fourths of K,,,. Inhibitor concentrations were varied between 10"' and M. Dissociation constants of the com- plexes were derived from Dixon plots (32). The inhibitors were: soybean trypsin inhibitor, Trasylol, leupeptin, and antipain.

Reactivity with Synthetic Peptide Chloromethyl Ketones CK/PA dissolved in 0.1 M sodium barbital, pH 8.6, was incubated

at 25 "C with the respective chloromethyl ketones. Samples were withdrawn at 2-min intervals and immediately subjected to active site titration with MUGB as described above in order to determine free enzyme. Ki values were derived from Dixon plots (33) and second order rate constants determined by the method of Kitz and Wilson (34).

Purification of a Plasminogen Activator from 12-0-Tetradecanqvl Phorboll3-Acetate-treated Rous Sarcoma Virus-transformed Chick

Embryo Fibroblasts The conditioned medium from TPA/RSV/CEF was prepared as

described by Wilson et al. (4). Ammonium sulfate precipitation and SP-Sephadex chromatography were performed as described by Unke- less et al. (2). The active fractions from the SP-Sephadex eluate were subjected to benzamidine-Sepharose chromatography using exactly the conditions described by Schleuning and Fritz (35). The final purification step was preparative SDS-polyacrylamide gel electropho- resis as described above. Under the conditions employed a 1000-fold purification and an overall yield of 8% was achieved. Recently Gold- farb and Quigley have reported a purification protocol which is based on very similar steps (36).

When analyzed by SDS-PAGE (10) the purified preparation mi- grated as a closely spaced doublet (apparent molecular weights, 43,000 and 44,000). The specific activity was 70,000 Ploug units/mg com- pared with urokinase and human plasminogen as substrate. Different assay conditions may account for somewhat different results of Gold- farb and Quigley (36).

Reactivity with Peptide Chloromethyl Ketones In order to characterize the enzymatic properties of CK/PA in

more detail, a study was undertaken to determine the kinetics of inactivation of the enzyme by peptide chloromethyl ketones. Free enzyme was determined by active site titration using MUGB. K; values were derived from Dixon plots (33).

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

Chicken Prekallikrein 14109

RESULTS

Purification of Chicken Prekallikrein Plasminogen Proactivator

Starting Material-800 ml of frozen chicken plasma (con- taining 10% of a 3.8% solution of trisodium citrate) were thawed in a water bath at 37 "C and filtered through gauze. The filtrate was centrifuged at 4 "C for 20 min at 6000 X g to remove cellular debris.

Barium Citrate Adsorption-To remove all of the vitamin K-dependent proteases and proenzymes the clear supernatant was adjusted to 0.1 M BaC12 by addition of the appropriate amount of a 1 M stock solution. The whitish precipitate was sedimented at 4000 X g at 4 "C for 20 min and discarded.

Ammonium Sulfate Precipitation-The clear supernatant (Fraction I) was poured into a stirred vessel in an ice bath and adjusted to 35% of saturation by addition of solid am- monium sulfate. After 20 min of stirring at 4 "C the precipitate was removed by centrifugation (4000 x g, 20 min) and the pellet discarded. The supernatant was adjusted to 55% of saturation by further addition of ammonium sulfate; the precipitate that accumulated after 20 min of stirring was collected by centrifugation and the supernatant discarded.

Plasminogen Removal by Lysine-Sepharose Affinity Chro- matography (7)"The pellet (Fraction 11) was dissolved in 100 ml of 0.1 M phosphate, pH 7.4. The clear reddish solution was passed over a column (2.5 X 80 cm) of lysine covalently coupled to Sepharose GB-CL at a flow rate of 40 ml/h. The breakthrough (Fraction 111) was collected and passed over a Sephadex G-25 column (12 X 30 cm) equilibrated in 10 mM Tris/HCl, 10 mM benzamidine, pH 8.6, flow rate, 400 ml/h.

DEAE-cellulose Chromatography-The protein excluded from the G-25 column was collected and pH and ionic strength were adjusted, if necessary, to those required for DEAE- cellulose chromatography (10 mM Tris/HCl, 10 mM benzam- idine, pH 8.6, "DEAE-buffer"). A DEAE-cellulose column (4 X 100 cm) equilibrated to the same conditions was then loaded with the protein solution and subsequently washed with 2000 ml of the starting buffer containing 0.022 M NaC1. The proactivator was eluted by washing the column with 2000 ml of DEAE-buffer, 0.035 M NaCl. Elution was at 230 ml/h and 10-ml fractions were collected. Each collecting tube contained 180 pl of 0.5 M MES to bring the eluate to approximately pH 6.0 (Fig. 1).

CM-Sepharose Chromatography-The active fractions from the DEAE-chromatography (Fraction IV) were pooled and, if needed, the pH was adjusted to exactly pH 6.0 by addition of 0.5 M MES. The resulting solution was loaded onto a column (2.5 x 40 cm) of CM-Sepharose GB-CL equilibrated with 5 mM MES/NaOH, pH 6.0, 10 mM benzamidine HCl, 50 mM NaCl ("CM buffer"). The column was washed with 3 volumes (600 ml) of CM buffer and subsequently eluted using a linear gradient of the same buffer containing NaCl with 50 and 250 mM as limiting concentrations. The total volume of the gra- dient was 2000 ml, the flow rate was 100 ml/h, and 10-ml fractions were collected (Fig. 2).

Sephacryl S-200 Chromatography-The active fractions from the CM-Sepharose step (Fraction V) were pooled and precipitated by dialysis against saturated ammonium sulfate. The precipitate was dissolved in 3 ml of 0.005 M MES/NaOH, 10 mM benzamidine, 0.5 M NaCl, pH 6.0, and passed over a Sephacryl S-200 column (1 x 100 cm) equilibrated in the same buffer at a rate of 25 ml/h. The fractions were monitored by SDS-PAGE and those containing apparently pure CPK/ PPA were pooled, precipitated by dialysis against saturated ammonium sulfate, and collected by centrifugation at 10,000 x g, 4 "C, for 20 min. The supernatant was carefully removed

I I I I I I

0 40 80 120 160 200 FRACTION NUMBER

"

FIG. 1. DEAE-cellulose chromatography of C P K P P A . De- tails of the procedure are described in the text. Protein was deter- mined by the method of Lowry et al. (8). Every tenth fraction was analyzed by SDS-polyacrylamide gel electrophoresis and the plasmin- ogen-dependent fibrinolytic activity was identified as in Ref. 6. Frac- tions 80-170 were pooled, adjusted to pH 6.0, and subsequently applied to a column of CM-Sepharose.

0 -40 80 120 160 200 FRACTION NUMBER

FIG. 2. CM-Sepharose chromatography of CPKPPA. De- tails of the procedure are described in the text. Protein was deter- mined by the method of Lowry et d . (8). Every tenth fraction was analyzed by SDS-polyacrylamide gel electrophoresis and the plasmin- ogen-dependent fibrinolytic activity was identified as in Ref. 6. Frac- tions 146-195 were pooled and precipitated by dialysis against satu- rated ammonium sulfate, redissolved in 3 ml of CM-buffer, and subsequently fractionated by gel filtration.

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

14110 Chicken Prekallikrein

and discarded, and the pellet (Fraction VI) stored a t -20 "C (Fig. 3).

Table I presents a summary of the purification procedure. Yields have been determined by monitoring the individual

I I

0 20 40 FRACTION NUMBER

r

FIG. 3. Sephacryl S-200 gel filtration. Details of the procedure are described in the text. Protein was determined by the method of Lowry et al. (8). Alternate fractions were analyzed by SDS-polyacryl- amide gel electrophoresis and the plasminogen-dependent fibrinolytic activity was identified as in Ref. 6. Fractions 14-20 were pooled and precipitated by dialysis against saturated ammonium sulfate. The precipitates were recovered by centrifugation and stored a t -20 'C.

TABLE I Purification of CPKIPPA

Fraction Total pro- Protein '?:/ recov- cation Total Purifi-

erv factor ml tein

Plasma ( N H I ) P S O ~

fractiona- tion

Lysine Se- pharose

DEAE-cellu- lose

CM-Sephar- ose

Sephacryl S- 200

mg jrnl 96 800 30,400 38 X lo3 160" 100 100 3,200 32 X lo3 300* 23.4 2.3

130 2,900 22 X lo3

1.600 800 500 6.2

140 14 40 120

60 1.8 30 30 1.4 260

Determined by radioimmunoassay. Determined by quantitative immunoelectrophoresis according to

Laurell (12).

fractions with a quantitative radioimmunoassay. The high initial concentration of the antigen in chicken plasma is noteworthy.

Zmmunoaffinity Chromatography-Once the preceding pu- rification procedure had been developed the highly purified protein was subjected to preparative SDS-PAGE and used to immunize rabbits. An immunoaffinity column, based on affin- ity-purified specific IgG molecules, was constructed (see under "Materials and Methods") and the proenzyme was thereafter isolated by the following abbreviated protocol.

We note here that the coupling of IgG to activated Sephar- ose was carried out at relatively low pH in order to minimize multipoint attachment of the protein. Columns prepared in this way have been used repeatedly without detectable loss of binding capacity for at least 1 year.

800 ml of chicken plasma were fractionated by ammonium sulfate precipitation as described above (Fraction I). The precipitate was dissolved in 100 ml of CM buffer containing 0.5 M NaCl and passed over a 10-ml anti-CPK/PPA Sephar- ose column a t a flow rate of 15 ml/h. The column was washed with 20 ml of the same buffer and subsequently eluted with 0.1 M glycine HCI, pH 2.2. A summary of the immunoaffinity purification procedure is given in Table 11.

Conversion of the High Molecular Weight Form to a Lower Molecular Weight Form-Identification of the high molecular weight form as a proenzyme and the low molecular weight form as a serine protease.

High molecular weight activator stored as an ammonium sulfate precipitate a t -20 "C was dissolved in 1 ml of 0.1 M (NH,)HCO,. After determination of the optical density at 280 nm the sample was diluted to a concentration of 0.5 absorb- ance unit (280 nm) per ml. The sample was incubated at 37 "C and aliquots of 20 pl were withdrawn after 20-min intervals, frozen immediately a t -60 "C, and lyophilized. One set of samples was analyzed by sodium dodecyl sulfate-gel electro- phoresis, while another was used for active site titration with 4-methylumbelliferylp-guanidinobenzoate, and the respective results are presented in Fig. 4. It can be seen that the forma- tion of catalytically active sites (as determined by MUGB titration) paralleled the appearance of the lower molecular weight forms.

The starting material appeared as a single band migrating as a protein with an apparent molecular weight of 87,000. After the incubation times indicated in Fig. 4 new bands appeared; these migrated more rapidly, their molecular weights corresponding to values of 37,000,38,000,39,000, and 40,000, of which only the 39,000 and 40,000 species were active (data not shown); their appearance coincided with the reduction in intensity of the high molecular weight bands. After complete activation the number of active sites (0.1 nmol) determined by active site titration corresponded approxi- mately to that expected from a proenzyme solution of the

TABLE I1 Purification of CPKIPPA by immunoaffinity chromatography

steps Volume concentra- CPKIPPA Recovery tELzr Protein

tion

rnl mglml % (NH4)PS04 frac- 100 32 0.300"

Anti-CPK/PPA 10 1.5 0.4 13.5 41 tion

Sepharose eluate

' Determined by quantitative immunoelectrophoresis according to Laurell (12).

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

Chicken Prekallikrein 14111

1.0

P I

0 * 0.5 d I

w f 0 60 I20

TIME (MIN)

a b c d s f g h

FIG. 4. Conversion of CPK/PPA into CK/PA monitored by active site titration (upper) and SDS-polyacrylamide gel elec- trophoresis (lower). CPK/PPA was dissolved in 0.1 M ammonium bicarbonate buffer, pH 7.4, and incubated a t 37 "C. Aliquots were withdrawn a t 20-min intervals, frozen a t -60 'C and lyophilized. The lyophilized samples were dissolved in SDS-containing sample buffer and electrophoresed in a SDS-polyacrylamide gel. Protein was stained with Coomassie brilliant blue. Time intervals were: a, 0 time; b, 20 min; c, 40 min; d, 60 min; e, 80 min; f, 100 min; g, 120 min. CPK/ PPA gives rise to a lytic zone on the plasminogen-containing fibrin agar, but does not react with [3H]iPrJ'-F or MUGB. The two bands representing apparent M, = 40,000 and 39,000 reacted with ['HI iPr,P-F and MUGB. The two bands with apparent M, = 37,000 and 38,000 were enzymatically inactive and presumably represent the activation peptides. Active site titration of CK/PA with MUGB during "autoactivation" of CPK/PPA was performed as follows. CPK/PPA (1.2 nmol or 100 pg) was autoactivated in a volume of 100 pl under the conditions described in the text. 10-pl samples were withdrawn at 20-min intervals after starting the reaction, frozen a t -60 "C, and lyophilized. Lyophilized samples were dissolved in 0.1 M sodium barbital, pH 8.6, and immediately subjected to active site titration. The appearance of active sites closely paralleled the ap- pearance of CK/PA.

same concentration, the molarity of which was calculated from the protein concentration assuming E;$ = 10 and mo- lecular weight of 87,000. Therefore, it is unlikely that the catalytically inactive low molecular weight bands represent degradation products of contaminating proteins comigrating with the proenzyme.

By resorting to the inhibitor iPrzP-F we established that CK/PA is a serine enzyme. Hence, a further indication that these events represented the activation of a proenzyme was obtained by incubating companion sets of aliquots with ['HI iPr2P-F to label the serine residue at the active site. When such reactions were analyzed by SDS-PAGE (10) and subse- quent autoradiography, radioactivity was detected only in the

species ( M , = 40,000) that represented CK/PA (Fig. 5). This demonstrated (a) that CPK/PA, the M , = 87,000 species, did not incorporate ['HH]iPrZP-F and (b ) that CPK/PPA was converted to CK/PA during the course of incubation.

Separation of CK/PA and the Actiuation Peptides by Chro- matography-CK/PA could be readily separated from the activation peptides by CM-Sepharose chromatography. The column (0.6 X 8 cm) was equilibrated at 4 "C with 50 mM MES/Na+, 25 mM NaCl, pH 6.0; the sample was applied and the column was washed with the same buffer. Subsequently, CK/PA was eluted by stepwise increments of sodium chloride (0.050, 0.100, 0.150, 0.250, and 0.500) buffered with 50 mM MES/Na+, pH 6.0. 4-ml fractions were collected. SDS-PAGE analysis (10) of the purified preparation is presented in Fig. 5 (left). The elution profile of the column is shown in Fig. 6. No experiments have been performed to recover or to char-

. M W ~ ~ O - ~

origin -

94 -

67 -

43 -

30 -

20.1 -

14.4 -

dye -

FIG. 5. Labeling of CPK/PPA and CK/PA with ['HliPrz- F. CPK/PPA was incubated a t neutral pH in the presence of [3H] iPr2P-F under the conditions described in the text. 5 pg of the ['HI iPr2P-F-treated sample were subjected to SDS-polyacrylamide elec- trophoresis followed by autoradiography. No radioactivity could be detected in the gel area corresponding to the molecular weight of CPK/PPA. Two fused radioactive bands (right), however, migrated exactly corresponding to the molecular weights of CK/PA (left). The dark spots represent markers that have been applied in order to establish the orientation of the gel.

FRACTION NUMBER FIG. 6. Separation of the active enzyme from its activation

peptides by ion exchange chromatography on CM-Sepharose. The left and right vertical axes are absorbance a t 280 nm (0) and sodium chloride concentration (X), respectively. The units on the additional right vertical axis indicate activity of CK/PA expressed in PIoug units (0). Fractions of 4 ml were collected, and the ionic strength was determined by conductance. Experimental details are described in text.

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

14112 Chicken Prekallikrein

acterize the activation peptides in greater detail. The recovery of active enzyme during this step was 60%.

Characterization of Chicken Prekallikrein Plasminogen Proactivator and Chicken Kallikrein Plasminogen Activator Polypeptide Chain Composition and NHz-terminal Amino

Acid-The apparent molecular weight of the highly purified protein, based on the rate of migration in SDS-PAGE, was 87,000 and 84,000, respectively, under nonreducing and re- ducing conditions (Fig. 7). It has been shown above that the molecule obtained in this way is apparently a catalytically inactive proenzyme, with the active enzyme being formed by proteolytic conversion to a lower molecular weight form. The proactivator consists of a single polypeptide chain and, using the dansyl procedure a single NHz-terminal residue (alanine) was found. The active species of the enzyme have apparent molecular weights of 40,000 and 39,000, respectively (Fig. 7).

Amino Acid Analysis-The amino acid analysis of CPK/ PPA, CK/PA, and TPA/RSV/CEF/PA is presented in Table 111. Values were calculated on the basis of the molecular weight obtained by dodecyl sulfate-electrophoresis of the re- duced form. There are no unusual features in the amino acid compositions of these proteins.

Carbohydrate Analysis-The carbohydrate composition of CPK/PPA and CK/PA is presented in Table IV. The presence of arabinose in CPK/PPA is an unusual feature of this sial- oglycoprotein.

Reaction with Peptide Substrates and Natural Inhibitors- Both D-Pro-Phe-Arg-p-nitroanilide (S-2302, KABI) and D- Glu-Gly-Arg-p-nitroanilide (S-2227, KABI) originally devel- oped, respectively, for human serum kallikrein and urokinase, proved to be excellent substrates for CK/PA; their respective K,,, values were 0.8 X and 2.5 X M. It is significant that CK/PA hydrolyzed both, the specific activity being 108 units/mg for D-Pro-Phe-Arg-p-nitroanilide and 39 units/mg for D-Glu-Gly-Arg-p-nitroanilide. Leupeptin and antipain strongly inhibited CK/PA; the K; values derived from Dixon plots were 1.4 X and 2 X lo-' M, respectively. Trasylol and soybean trypsin inhibitor inhibited CK/PA only rela- tively weakly (Ki value = 1.1 x M for trasylol and 1.8 x

M for soybean trypsin inhibitor). The fluorescent sub- strate 7-(N-Cbz-glycyl-glycyl-argininamido)-4-methylcou- marin trifluoroacetate designed for plasminogen activators by Zimmerman et al. (14) proved also to be an excellent substrate

' I 7 I-"

B C D a b

i

CPK/PPA and (B , b, C, c) the two forms of CK/PA. D, d , FIG. 7. SDS-polyacrylamide gel electrophoresis of (A, a)

molecular weight ( M W ) markers. A , B, C, D, nonreduced proteins; a, b, c, d, reduced proteins. Experimental details are described in the

TABLE 111 Amino acid Commsition of CPKIPPA and CKIPA

Amino acid CPKIPPA" CK/PA" TPA'RSVI CEF/PAb residue/IOO residues

Lysine 6.4 5.8 3.1 Histidine 4.5 2.9 1.2 Arginine 5.4 5.0 4.6 Aspartic' 10.0 8.9 10.3 Threonine 5.2d 5.od 5.0 Serine 5.8d 8.6d 5.9 Glutamic 12.7 11.7 9.4 Proline 2.8 2.2 1.7 Alanine 3.1 6.0 10.8 Half-cysteine 6.6' 5.4' 1.7' Valine 4 .8R 5.1 6.3 Methionine 1 .Od 0.8d 1.3 Leucine 6.4" 4.9 6.7 Isoleucine 4.Y 6.6 3.8 Tyrosine 6.0 3.4 2.3 Phenylalanine 6.4 3.7 2.5 Tmtophan NDh 4.9 ND

a Average of values from 24-, 48-, and 72-h hydrolysates. ' Average value of 24-h hydrolysate. 'Aspartic acid and glutamic acid values include asparagine and

Corrected for decomposition by extrapolation of values obtained

Determined as cysteic acid by the method of Hirs (17). 'Determined as cysteic acid by the method of Moore (18).

Values obtained from the 72-h hydrolysis. * Estimated spectrophotometrically by the method of Bencze and

glutamine, respectively.

from 24-, 48-, and 72-h hydrolysates.

Schmid (19). ND, not determined.

TABLE IV Carbohydrate composition of CPKIPPA and CKIPA

Carbohydrate CPK/PPA CK/PA

Fucose Arabinose Glucosamine Galactosamine Mannose Galactose Sialic acids' Total

9% weight 1.7" (9)' 0.15" (<<1)' 1.4 (8) 0.15 ( ~ 1 ) 3.2 (16) 4.0 (9) 0.6 (3) 1.7 (4) 2.0 (10) 2.5 (5) 1.9 (9) 2.9 (6) 2 .p (7) 2.5d (4)

13.0 (62) 13.9 (29) ~

Values represent averaged duplicate determinations. 'The numbers in parentheses are the number of residues per

molecule assuming molecular weights 87,000 and 39,500 for CPK/ PPA and CK/PA, respectively.

Determined by the method of Warren (25). Values represent averaged triplicate determinations. Standard

error values are below 0.1%.

for CK/PA (K,,, = 3.25 X M). Experiments with pyro- Glu-Gly-Arg-p-nitroanilide (S-2444, KABI) have not yet been performed.

Kininogeme Activity-The kininogenase activity of 1 pg of CK/PA was equal to that of 0.25 pg of highly purified human plasma kallikrein* when tested under the conditions described under "Materials and Methods."

Plasminogen-activating Activity-The plasminogen-acti- vating activity of CK/PA was 280 Ploug units/mg or 11.2 X IO9 Ploug units/mol as measured by the '2sII-fibrinolysis assay (2) employing human urokinase as a standard.

Inactivation with Peptide Chloromethyl Ketones-The re- versible dissociation constants of the substrate-like complex Ki, determined as described under "Materials and Methods" for CK/PA and the chloromethyl ketones of Ala-Phe-Arg, Glu-Gly-Arg, Ile-Glu-Arg, and Val-Ile-Pro-Arg were in the

text. P. S. Aiyappa, A. Guha, and E. b i c h , manuscript in preparation.

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

Chicken Prekallikrein 14113

respective order 0.17, 2.45,0.49, and 0.23 mM. This indicates a higher affinity of Ala-Phe-Arg-CH2C1 than Glu-Gly-Arg- CH,Cl for the active site of CK/PA and suggests, taking into account the findings of other authors (37,38), a specificity of the enzyme more related to human plasma kallikrein than to urokinase.

Tryptic Peptide Mapping-In Fig. 8 the tyrosine-containing tryptic peptides of CK/PA and TPA/RSV/CEF/PA as well as peptide maps of the two components of CK/PA which differ in molecular weight are compared in Fig. 8. The com- parison shows that the two forms of the plasma enzyme are similar. The ratio of the number of peptides with the same indices to the number of peptides with different indices is equal to 1.36 (41:30). The difference in molecular weight of the two components is most likely due to limited proteolytic degradation of the lesser molecular weight component. CK/ PA and TPA/RSV/CEF/PA are with regard to their peptide

a$, i-

' e

- ELECTROPHORESIS

+ FIG. 8. The comparison shows that the two forms of the

plasma enzyme are similar. The ratio of the number of peptides with the same indices to the number of peptides with different indices is equal to 1.36 (41:30). The difference in molecular weight of the two components is most likely due to limited proteolytic degradation of the lesser molecular weight component. CK/PA and TPA/RSV/CEF/ PA are with regard to their peptide maps heterogeneous; the ratio as defined above is 0.42 (24:57), indicating that a close structural rela- tionship is unlikely, even though poor background exposure may contribute to the differences observed.

I I I 6.25 125 25 20 loo 2'00 ' anti-CPK/PPA IgG(pgg/rnl)

FIG. 9. Inhibition of TPA/RSV/CEF activity by anti-CPK/ PPA IgC. 100-pl aliquots of TPA/RSV/CEF-conditioned medium (containing 0.4 Ploug units/ml of PA activity) were mixed with 800 pl of 50 mM Tris/HCI, 0.2 M NaC1, 0.1% bovine serum albumin and 100 pl of anti-CPK/PPA IgG diluted in the same buffer, just to the point to achieve the indicated concentrations. After incubation of the sample overnight a t 4 "C, 10-pl aliquots were withdrawn and assayed by the '251-fibrin plate method (2) (A-A). Control experiments were performed with nonimmune rabbit I& (U).

maps heterogeneous; the ratio as defined above is 0.42 (24:57), indicating that a close structural relationship is unlikely, even though poor background exposure may contribute to the dif- ferences observed.

Immunological Relationship between CPKIPPA and TPAI RSVICEFIPA-TPA/RSV/CEF/PA was bound to anti- CPK/PPA IgG-Sepharose but the capacity of the immunoad- sorbent was very low (20 Ploug units/ml of settled Sepharose). 50 pg of purified anti-CPK/PPA IgG inhibited 50% of the activity of 0.4 Ploug unit of TPA/RSV/CEF/PA (Fig. 9).

DISCUSSION AND CONCLUSIONS

Several aspects of the present work deserve brief comment. These include 1) the purification procedure; 2) the purity of the protein; 3) the conversion of the high to low molecular weight form; 4) the functional nature of the enzyme and its homology and analogy, if any, to mammalian plasma proteins; and 5) its relationship to the plasminogen activator produced by cultured chick fibroblasts in response to sarcoma virus transformation and/or tumor promoters.

1. The purification scheme described above reproducibly yields in a few steps a highly purified protein. At an early stage in the development of this procedure we observed that the M, = 84,000 component often disappeared, particularly during or after column chromatography, and it was apparent that the native molecule was unusually susceptible to limited proteolysis. The procedure was, therefore, designed to mini- mize proteolysis; the removal of vitamin K-dependent factors by barium citrate precipitation and of plasminogen and plas- min by lysine-Sepharose, as well as the maintenance of low- ered pH in the presence of benzamidine, all reduced the level of potential proteolysis during purification. The satisfactory yield of highly purified material demonstrates the effective- ness of these measures, but it is also dependent on the elimination of delays at any stage of the workup.

2. The high purity of the final product appears to be well established by the following criteria: (a) the protein showed only a single Coomassie blue-stainable band following SDS- PAGE of large amounts of protein under both reducing and nonreducing conditions; (b) a single precipitin line was ob- tained in Ouchterlony double immunodiffusion using high titer rabbit antisera; (c) a single NHderminal amino acid (alanine) was found by the dansyl method. Based on Coomas- sie blue staining of heavily loaded gels we estimate that the material obtained according to the procedure summarized in Table I is approximately 95% pure, and this can be improved (to perhaps 98%) by a second passage over Sephacryl S-200. Preparative SDS-PAGE yields material suitable for immuniz- ing rabbits and is apparently free of detectable impurities on analytical SDS-PAGE.

We considered the possibility that the two catalytic activi- ties, plasminogen activator and kininogenase, might be due to separate enzymes but this appears unlikely for several rasons. The proenzyme was isolated using its PA activity in zymographic gels as an assay, and the kallikrein activity would have had to co-purify throughout the procedure summarized in Table I. The high kallikrein activity would then suggest that a substantial fraction of the final product necessarily consists of kallikrein. To conform with the observations de- scribed above, the putatively separate PA and kallikrein com- ponents would then have to be identical in molecular weight under both reducing and nonreducing conditions, in NHZ- terminal amino acid residues, in double immunodiffusion; further, both would have been isolated as proenzymes requir- ing activation and, as seen in Fig. 4, both would be converted to active enzymes of identical molecular weight. The proba-

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

14114 Chicken Prekullikrein

bility that all these requirements could fortuitously be met by two unrelated proteins appears to us to be very small.

3. Several observations provide a coherent indication that the M , = 84,000 form, as isolated, is a catalytically inactive zymogen. Thus, this component failed to react with the active site titrant MUGB, it was neither inactivated by iPr2P-F, nor did it incorporate this radioactive inhibitor, and its amidolytic activity was low. In contrast, the M, = 39,000 species, whose appearance during incubation parallels the disappearance of the starting material, reacts rapidly with MUGB, is irrevers- ibly inactivated by incorporated [3H]iPr2P-F, and has a sub- stantial amidolytic activity, Although it seems reasonable to conclude that the active enzyme is generated by limited pro- teolysis of the larger precursor, the mechanism of this con- version is at present obscure. Prolonged exposure of proen- zyme to high concentrations of [3H]iPr2P-F did not reduce the rate of activation at 37 "C after the inhibitor was removed by dialysis, nor was activation blocked by the presence of Trasylol. While these results might be taken to suggest that activation was autocatalytic, such a conclusion seems pre- mature because the presence of traces of contaminating pro- tease cannot easily be ruled out.

If it is accepted that the M, = 84,000 species is indeed a zymogen it might seem paradoxical that it behaves as an efficient plasminogen activator in the zymographic gel assay. With the high susceptibility to spontaneous activation in mind, it is likely that this component is activated during the incubation period either in the running gel after removal of SDS or by trace proteolytic impurities in the indicator gel.

4. Concerning the identity of the (pro)enzyme, there is not yet sufficient evidence to resolve this issue definitively. From the results of zymographic analysis of either native chicken plasma or plasma fractions it appears that the component we have isolated is the major and probably the only detectable PA. However, its catalytic efficiency as a plasminogen acti- vator in solution is very low, 280 Ploug units/mg, whereas the comparable value for the 53,000 component of human uroki- nase is 150,000 Ploug units/mg. 1 unit of CK/PA (defined by amidolysis S-2302) had a specific PA activity of 2.6 Ploug units whereas 20 Ploug units of urokinase did not hydrolyze S-2302 under otherwise identical assay conditions. Further, the kininogenase activity of CK/PA was high, even with nonhomologous human kininogen as substrate, and the inhib- itor spectrum of the enzyme differed significantly from that of the major and efficient human PA urokinase. The combi- nation of high kininogenase and low PA activity is character- istic of human plasma kallikrein (39, 40) and, hence, the plasma (pro)enzyme we have isolated resembles' human plasma (pre)kallikrein, whereas TPA/RSV/CEF/PA seems to be more related to human urokinase and/or tissue plasmino- gen activator.

5. Although both are serine enzymes and similar in their molecular weight, several lines of evidence show conclusively that the enzymes from chicken plasma (CK/PA) and chicken fibroblast culture (TPA/RSV/CEF/PA) differ clearly in structural, immunochemical, and catalytic properties. The two enzymes can be distinguished by amino acid composition and tryptic peptide maps, by the fact that the ratio PA/ amidolytic activity is very high for the cell culture enzyme whereas the converse holds for the plasma enzyme, and by differences in sensitivity to a range of inhibitors. Immuno- chemical data indicate a structural relationship between CPK/PPA and TPA/RSV/CEF/PA. This is not surprising if one takes into account the common evolutionary origin of the serine enzymes. A close structural homology was found be- tween urokinase and human urinary kallikrein (41). By com- paring the capacity of anti-CPK/PPA Sepharose for CPK/

PPA and TPA/RSV/CEF/PA we estimate that the popula- tion of cross-reacting antibodies is lower than 1%. Thus CPK/ PPA and TPA/RSV/CEF/PA share some antigenic determi- nants and are perhaps closely related in evolution. It appears, however, that the classification of CK/PA as a new PA (6) or a PA possibly identical with TPA/RSV/CEF/PA (1, 36) is not warranted. Instead CPK/PPA seems to be the avian analogue of human serum prekallikrein. Its high concentra- tion in blood indicates a significant function; whether this function is analogous to the human kallikrein/kininogen/ kinin system remains to be established.

Acknowledgments-We thank Richard Hajdu and Kart Grizzuti for excellent technical assistance, Drs. Peter Blackburn and Stanford Moore for the amino acid analysis, Drs. Hans Fritz, Brigitte Ditt- mann, and Bruni Forg-Brey for their generous help with the kinino- genase assay and the kinetic measurements, Dr. Reinhard Mailham- mer for carrying out the sugar analysis and Dr. Elliot Shaw for gifts of the peptide chloromethyl ketones. We thank Dr. H. Hanafusa for a generous gift of the ts-68 mutant of Rous sarcoma virus.

REFERENCES 1. Schleuning, W.-D. and Piperno, A. (1979) in Biological Functions

of Proteinases (Holzer, H., and Tschesche, H. eds) 30th Collo- quium of the Gesellschaft fur Biologische Chemie, Mosbach, pp. 171-185, Springer-Verlag, Heidelburg

2. Unkeless, J. C., Tobia, A., Ossowski, L., Quigley, J. P., Rifkin, D. B., and Reich, E. (1973) J. Exp. Med. 137,85-111

3. Wigler, M., and Weinstein, I. B. (1976) Nature (bnd.) 269,

4. Wilson, E. L., and Reich, E. (1978) Cell 15,385-392 5. Valinsky, J. E., Reich, E., and Le Douarin, N. M. (1981) Cell 26,

6. Granelli-Piperno, A., and Reich, E. (1978) J. Ezp. Med. 148,

7. Deutsch, D. G., and Mertz, E. T. (1970) Science 170,1095-1096 8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

9. Barlow, G . H., Summaria, L., and Robbins, K. C. (1969) J. Biol.

232-233

471-476

223-233

(1951) J. Biol. Chem. 193, 265-275

Chem. 244, 1138-11 10. Laemmli, U. K. (1970) Nature (bnd.) 227,680-685 11. Hagen, F. S., and Young E. T. (1974) Biochemistry 13, 3394-

12. Laurell, C.-B. (1966) Anal. Biochem. 15,45-52 13. Latallo, Z. S., Teisseyre, E., and Raczka (1979) in Chromogenic

Peptide Substrates (Scully, M. F., and Kakkar, V. V., eds) pp. 154-164, Churchill Livingstone, Edinburgh

14. Zimmerman, M., Quigley, J. P., Ashe, B., Dorn, C., Goldfarb, R., and Troll, W. (1978) Proc. Natl. Acad. Sci. U. S. A. 76, 750- 753

15. Jameson, G. W., Roberts, D. V., Adams, R. W., Kyle, W. S. A., and Elmore, D. T. (1973) Biochem. J. 131,107-117

16. Moore, S., and Stein, W. H. (1963) Methods Enzymol. 6, 819- 83 1

17. Hirs, C. H. W. (1956) J. Bwl. Chem. 219,611-621 18. Moore, S. (1963) J . Biol. Chem. 238,235-237 19. Bencze, W. L., and Schmid, K. (1957) Anal. Chem. 29, 1193-

20. Hartley, B. S. (1970) Biochem. J. 119,805-822 21. Gros, C., and Labouesse, B. (1969) Eur. J. Biochem. 7,464-470 22. Woods, K. R., and Wang, K. T. (1967) Biochim. Biophys. Acta

23. Porter, W. H. (1975) Anal. Biochem. 03,27-43 24. Glick, M. C. (1974) Methods Membr. Biol. 157-204 25. Warren, L. (1959) J. Biol. Chem. 234,1971-1975 26. Porath, J., and Kristiansen, T. (1975) in The Proteins (Neurath,

H., and Hill, R. L., e&) pp. 115-121, Academic Press, New York

27. March, S. C., Parikh, I., and Cuatrecasas, P. (1974) Anal. Biochem. 60, 149-152

28. Trautschold, I. (1970) in Handbook of Experimental Pharmacol- ogy. Vol. XXV. Bradykinin, Kallidin und Kallikrein (Erdos, E.,

29. Krohn, K., Sherman, L., and Welch, M. (1972) Bioehim. Biophys. ed) p. 55, Springer-Verlag

30. Kessler, S. W. (1975) J. Zmmunol. 116.1617-1624

3400

1196

133,369-370

Acta 286,404-413

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

Chicken Prehllikrein 14115

31. Rodbard, D., Rayford, P. L., Cooper, J. A., and Roos, G. T. (1968)

32. Elder, J. H., Pickett, R. A., 11, Hampton, J., and Lerner, R. A.

33. Dixon, M. (1953) Biochem. J. 55, 170-171 34. Kitz, B., and Wilson, I. B. (1962) J. Bwl. Chem. 237,3245-3249 35. Schleuning, W. D., and Fritz, H. (1974) Hoppe-Seyler’s 2. Physiol.

36. Goldfarb, R. H., and Quigley, J. P. (1980) Biochemistry 19,5463-

J. Clin Endocrinol. Metab. 28, 1412-1418

(1977) J. Biol. Chem. 252,6510-6515

Chem. 355, 125-130

5471

37. Kettner, C., and Shaw, E. (1978) Biochemistry 17,4778-4784 38. Kettner, C., and Shaw, E. (1979) Biochim. Biophys. Acta 569,

39. Gallimore, M. J., Fareid, E., and Stormorken, H. (1978) Thromb. Res. 12,409-420

40. Bouma, B. N., Miles, L. A., Beretta, G., and Griffin, J. H. (1980) Biochemistry 13,1446-1454

41. Henschen, A., Reich, E., Sauser, D., and Lottspeich, F. (1982) in Urokinase: Basic and Clinical Aspects. Serono Symposia, Vol. 48, pp. 27-31, Academic Press, New York

31-40

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from

W D Schleuning, M Sudol and E ReichA proenzyme from chicken plasma similar to human plasma prekallikrein.

1983, 258:14106-14115.J. Biol. Chem. 

  http://www.jbc.org/content/258/23/14106Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/258/23/14106.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on July 3, 2020http://w

ww

.jbc.org/D

ownloaded from