a gift of Robert Rosenberg. Neutral peptide-generating protease, obtained from purified human neutrophils, was purified to homogeneity (8) as assessed.
Proc. Nati. Acad. Sci. USA Vol. 77, No. 9, pp. 5448-5452, September 1980 Medical Sciences
Cleavage of fibrinogen by the human neutrophil neutral peptide-generating protease (fibrinolytic/neutrophil enzymes)
BRUCE U. WINTROUB*tI, JONATHAN S. COBLYNt§, CAROL E. KAEMPFERt, AND K. FRANK AUSTENt§ Departments of *Dermatology and §Medicine, Harvard Medical School; and the tDepartment of Rheumatology and Immunology, and the *Division of Dermatology, Department of Medicine, Brigham and Women's Hospital, Boston, Massachusetts 02115
Contributed by K. Frank Austen, June 2, 1980
ABSTRACT The human neutrophil neutral peptide-generating protease, which generates a low molecular weight vasoactive peptide from a plasma protein substrate, is directly fibrinolytic and cleaves human fibrinogen in a manner distinct from plasmin. Fibrinogen was reduced from 340,000 Mr to derivatives of 270,000-325,000 Mr during interaction with the protease at enzyme-to-substrate ratios of 0.3 or 1.0 yg/1.0 mg. The 310,000-325,000 M, cleavage fragments exhibited prolonged thrombin-induced clotting activity but were able to be coagulated, whereas the 270,000-290,000 Mr fragments were not able to be coagulated. Anticoagulants were not generated at either enzyme dose. As analyzed by sodium dodecyl sulfate/polyacrylamide gel electrophoresis in 4-30% gradient gels and 10% gels stained for protein and carbohydrate, the diminution to 310,000-325,000 Mr and the prolongation of thrombin-induced clotting time resulted from cleavage of the fibrinogen Aa chain. The further decrease in size to 270,000-290,000 Mr was associated with Bkchain and y-chain cleavage and an inability to form 'y-'y dimers. The neutral peptide-generating protease, a distinct human neutrophil neutral protease with fibrinolytic and fibrinogenolytic activities comparable to those of plasmin on a weight basis, cleaves fibrinogen in a manner that is distinct from the action of plasmin, leukocyte elastase, and leukocyte granule extracts. It may be that the concerted action of this neutrophil protease to generate a vasoactive peptide and to digest fibrinogen and fibrin facilitates neutrophil movement through vascular and extravascular sites. Neutrophils in mixed leukocyte suspensions were visually observed to degrade fibrin clots (1, 2), and neutrophils have been identified within intravascular and extravascular thrombi (3). Neutrophil-mediated clot lysis may proceed by at least three mechanisms. The finding of degradation products of fibrin within the neutrophil (4, 5) is consistent with fibrin dissolution during phagocytosis. Because phagocytosis is associated with exocytosis of lysosomes, clot lysis may also involve release of lysosomal enzymes such as leukocyte elastase and cathepsin G, which have direct fibrinolytic activity (6). Finally, the neutrophil, when activated by the lectin concanavalin A, synthesizes and secretes an activator of plasminogen that converts plasminogen to the fibrinolytic enzyme plasmin (7). Recently, a human neutrophil neutral protease was isolated and distinguished from leukocyte granule elastase, cathepsin G, and collagenase (8). The protease, designated the neutral peptide-generating protease, cleaves a-N-carbobenzoxy-Llysine-p-nitrophenyl ester (8), a synthetic substrate of plasmin (9); this activity suggested that the protease may also have the capacity to degrade fibrin and fibrinogen directly. The neutral peptide-generating protease (10, 11) has fibrinogenolytic activity comparable to plasmin on a weight basis, but cleaves the chains of fibrinogen at sites different from those cleaved by plasmin so as to generate a different set of reaction products. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
MATERIALS AND METHODS Materials. The following were obtained as noted: aprotinin (Boehringer Mannheim); human fibrinogen, fraction I, B grade (Calbiochem); thrombin, bovine topical (Parke, Davis, St. Louis, MO); streptokinase (Hyland, Costa Mesa, CA); urokinase (Sigma); and trypsin and soybean trypsin inhibitor (Worthington). Human thrombin, purified to homogeneity (12), was a gift of Robert Rosenberg. Neutral peptide-generating protease, obtained from purified human neutrophils, was purified to homogeneity (8) as assessed by NaDodSO4/polyacrylamide gel electrophoresis in 10% (wt/wt) acrylamide gels (13). Plasminogen was purified from normal human plasma by the method of Deutsch and Mertz (14). Fibrinogen was purified from commercial human fibrinogen (15) and rendered plasminogen free as described (14). This preparation was more than 95% coagulable by thrombin [5 National Institutes of Health units/mg of fibrinogen]. The preparation was devoid of plasminogen; it was not digested during 24-hr treatment with urokinase [10 Committee on Thrombolytic Agents (CTA) units/mg of fibrinogen] at 370C. Functional Assay of Fibrinolytic and Fibrinogenolytic Activities. Fibrinolytic activity was assayed on human plasminogen-free fibrin agarose plates (16). Fibrinogen was iodinated and 125I-labeled fibrinogen was prepared in Linbrow plates (Falcon) as described (15). Total available radioactivity was defined as that amount of 125I solubilized when 8 ,ug of trypsin in 200 ,g of 10 mM Tris-HCl, pH 7.4/0.15 M NaCl was added to a well and the plate was incubated for 1 hr at 370C. Radioactivity in a 100-,ul sample from the supernatant of each well was measured in a gamma counter (Searle Model 1185, Chicago, IL) and corrected for radioactivity solubilized by buffer alone. Assessment of Fibrin and Fibrinogen Degradation Products by NaDodSO4/Polyacrylamide Gel Electrophoresis. The electrophoresis was carried out in 5% and 10% acrylamide gels by a modification of a described procedure (13). Electrophoresis in gradient acrylamide gels (gradient electrophoresis) was performed with 4-30% gradient polyacrylamide gels (Pharmacia) in 40 mM Tris-HCl, pH 7.4/20 mM sodium acetate/2 mM EDTA/0.2% NaDodSO4. Preelectrophoresis of gradient gels was carried out at 70 V for 1 hr; after application of the samples, electrophoresis was performed at 300 V for 10 min and then at 150 V for 2.5 hr. Gels were stained for protein with 0.1% Coomassie brilliant blue in 45% (vol/vol) methanol/10% (vol/vol) acetic acid overnight, destained for 18-24 hr at room temperature in 10% acetic acid, and stored in that solution. Alternatively, gels were stained for carbohydrate with the periodic acid-Schiff base reagent (PAS) according to published Abbreviations: DFP, di[iisopropyltluorophosphate; PAS, periodic acid-Schiff base reagent.
Medical Sciences: Wintroub et al. methods (17). The molecular weights of the Aa, BO3, and y chains of fibrinogen, determined by NaDodSO4/gradient polyacrylamide gel electrophoresis, were estimated from the values obtained for reduced protein standards run independently. Functional Assessment of Fibrinogen Degradation Products. Fibrinogen (3.0 mg/ml in 10 mM Tris-HCl, pH 7.4/0.15 M NaCl) was treated with various enzymes or buffer alone and the reaction was stopped by addition of soybean trypsin inhibitor to a final concentration of 10 mM. The functional effects of fibrinogen degradation were studied by determination of the thrombin-induced clotting time, the presence of anticoagulants, and the capacity to form crosslinked fibrin. For determination of thrombin-induced clotting time, 10 MAl of a solution containing purified thrombin (50 National Institutes of Health units/ml) and CaCl2 (10 mM) in 10 mM Tris-HCl, pH 7.4/0.15 M NaCl (thrombin/calcium solution) was added to 100,ul of the reaction mixture, and the time at which fibrin formed was measured visually at room temperature. The reference thrombin-induced clotting time was 25 sec. Based upon dose-response studies with normal fibrinogen, a 50% reduction in functional fibrinogen extended the thrombin-induced clotting time to 35 sec. Failure to form fibrin in less than 180 sec indicated that at most only 10% of the functional fibrinogen remained. To detect anticoagulant activity, we mixed 100-,Ml samples of fibrinogen digests with 100,Ml of fibrinogen (3.0 mg/ml in 10 mM Tris-HCl, pH 7.4/0.15 M NaCl) and added 10 ,ul of thrombin/calcium solution. The thrombin-induced clotting time was measured, and anticoagulant activity was assessed by prolongation of the clotting time. To assess the capacity of fibrinogen digests to form crosslinked fibrin, we added 10 ,l of thrombin/calcium solution to 100-,l samples of fibrinogen digests and incubated the mixtures for 3 hr at room temperature. Fibrin was solubilized by addition of 300,ul of a solution containing 9% urea, 3% NaDodSO4, and 3% 2-mercaptoethanol (all wt/vol). Samples were subjected to gradient electrophoresis and the gels were stained for protein and destained as described above. Under these conditions, protein bands were detected that corresponded in size to fibrin a (66,000-69,000 Mr), (53,000 Mr) and y (45,000 Mr) chains. A reduction in -y-chain staining was associated with the appearance of a 100,000 Mr band representing y-'y dimers. Factor XIII was a trace contaminant of the fibrinogen preparation and was used as the source of crosslinking enzyme (18). RESULTS Fibrinogenolytic and Fibrinolytic Activities of the Neutral Peptide-Generating Protease. The fibrinolytic activity of neutral peptide-generating protease was compared with that of streptokinase-activated human plasminogen by use of plasminogen-free fibrin plates. One- to 4-pig portions of each enzyme were applied in 15 ,l of 10 mM Tris-HCl, pH 7.4/0.15 M NaCl. No lysis was produced by plasminogen, streptokinase, or either of the buffers. Increasing the concentrations of the neutral peptide-generating protease resulted in a linear increase in the zone of lysis (Fig. 1). When compared on a weight basis with streptokinase-activated plasminogen, the neutral peptide-generating protease was approximately half as active. The fibrinogenolytic activity of neutral peptide-generating protease was compared with that of streptokinase-activated human plasminogen by use of l25I-labeled fibrinogen plates. A 1- to 8.0-1ug portion of each enzyme was added to each well in 20 ,ul of 10 mM Tris-HCl, pH 7.4/0.15 M NaCl, and the plates were incubated for 1 hr at 37°C. Trypsin gave a linear dose-response solubilization of radioactivity from 1.0 to 8.0,Mg. Streptokinase-activated plasminogen and neutral peptide-
Proc. Natl. Acad. Sci. USA 77 (1980)
Protein, pg FIG. 1. Fibrinolytic activity of neutral peptide-generating protease (@) and streptokinase-activated human plasminogen (0) assessed on plasminogen-free fibrin plates.
generating protease gave a linear dose-related solubilization of radioactivity from 1.0 to 4.0,Mg, with the highest dose of each enzyme having an effect comparable to 8 ,g of trypsin. Solubilization of radioactivity in streptokinase or buffer alone at 1 hr was less than 10% of that achieved with the enzymes. To assess the kinetics of this reaction, we incubated 0.5 and 1.0 Mg of protease and 2.0 Mg of streptokinase-activated plasminogen in quadruplicate under the same reaction conditions. Single reaction mixtures were assessed at 10, 20, 40, and 60 min. Solubilization of radioactivity increased with time in a linear manner for both enzymes (Fig. 2). Structural Assessment of Degradation of Fibrinogen. In order to determine the ratio of soluble fibrinogen to neutral peptide-generating protease to be used in detailed degradation studies, 1.0-mg portions of fibrinogen in 330 Ml of 10 niM Tris1HCI, pH 7.4/0.15 M NaCl were incubated with 0.3, 1.0, 5.0, and 20.0 Mug of neutral peptide-generating protease at 370C for 1 hr. To stop the reaction, we made each mixture 1.0 mM in diisopropylfluorophosphate (DFP) and held it overnight at 40C. Twenty-microliter samples of each mixture were subjected to gradient electrophoresis without reduction. Fibrinogen incubated with buffer alone gave a major band of 340,000 Mr and a minor, higher molecular weight band. The major band was not apparent in the reaction mixtures incubated with increasing concentrations of neutral peptide-generating protease. Cleavage fragments of 310,000-325,000 Mr, 280,000-290,000 Mr, and 260,000-270,000 Mr were generated by 0.3, 1.0, and 5.0,Mg of protease, respectively. With 20 Mg of protease, derivatives of 300,000-260,000 Mr, 185,000 Mr, 150,000 Mr, 100,000 Mr, 70,000 Mr, 58,0Q0 Mr, 16,000 Mr, and less than 10,000 Mr were present. In order to study the sequential degradation of fibrinogen, we incubated 0.9 ug of neutral peptide-generating protease Protease, 1.0 pgg Plasmin,
800 800 600
~ __Pro~tease, 0.5 p~g
Incubation, min FIG. 2. Time-dependent solubilization of radioactivity from insoluble l25-Ilabeled fibrinogen by neutral peptide-generating protease and streptokinase-activated human plasminogen.
Medical Sciences: Wintroub et al.
with 3.0 mg of fibrinogen (0.3 ,.g of neutral peptide-generating protease per mg of fibrinogen) in a final volume of 1.0 ml of 10 mM Tris-HCI, pH 7.4/0.15 M NaCl for varying periods of time up to 24 hr. Sixty-microliter samples were removed from the reaction mixture at each time point and made 0.1 mM in DFP for subsequent analysis of structural changes. A 5-,1l portion (7.5 Mg of fibrinogen) of the DFP-treated sample from each time point was subjected to gradient electrophoresis without reduction, and a 7.5-,Ml (10.5 ,tg of fibrinogen) DFP-treated sample from each time point was analyzed by gradient electrophoresis after reduction. Analysis of unreduced samples revealed fibrinogen of 340,000 Mr at time zero and the appearance of a 325,000 Mr fibrinogen fragment at 5 min (Fig. 3). Fibrinogen was progressively reduced in size such that the major derivatives were 310,000-325,000 Mr by 60 min and 270,Q00-280,000 Mr by 24 hr (Fig. 3). To detect low molecular weight fragments, we subjected a 2-,ul portion (30 ,g of fibrinogen) of each DFP-treated sample to gradient electrophoresis without reduction. Fragments of less than 10,000 Mr were identified at 5 min and at each subsequent time. Gradient electrophoresis of reduced material from the mixture revealed two proteins in the Aa-chain region (70,000 and 67,000 Mr), the B13-chain region (58,000 Mr), and the y-chain region (45,000 Mr) at time zero (Fig. 4). After 5 min of digestion, the Aa chain was markedly diminished in the 67,000-70,000 Mr region and protein bands of 55,000 Mr and 42,000 Mr appeared (Fig. 4). The 55,000 Mr band is clearly distinguished from the BB chain in the original gels; however, this fragment can be seen in the photographs only as a widening of the stained band in the BB region. At 4 hr, a protein band of 51,000 Mr and protein bands of 32,000-38,000 Mr were detected, and the staining of these fragments increased in intensity between 4 and 24 hr. The BB chain (58,000 Mr) was present during the entire 24-hr digestion, and the y chain (45,000 Mr) was not detectable after 8 hr. Because only the BB and -y chains of fibrinogen are known to stain with PAS (19), a 7.5-,Ml portion (10.5 Mig of fibrinogen) of each reduced DFP-treated sample was subjected to gradient electrophoresis and stained with PAS to determine the most likely source of the 51,000 Mr and the 32,000-38,000 Mr fragments recognized in Coomassie blue-stained gels. The band corresponding to the Bo3 chain stained with PAS and was present
i: -| i I; ~ ~ ~
FIG. 3. Time-dependent degradation of soluble fibrinogen by 0.3 Ag of protease per mg of fibrinogen as assessed by gradient electrophoresis of unreduced substrate stained for protein.
Pioc. Natl. Acad. sci. USA 77 (1980) .a
,, . ;
FIG. 4. Time-dependent degradation of soluble fibrinogen by 0.3
,g of protease per mg of fibrinogen as assessed by gradient electro-
phoresis of reduced substrate stained for protein.
throughout the 24-hr digestion but decreased in intensity at 4 hr with the appearance of a 51,000 Mr PAS-positive protein. The PAS-positive -y chain decreased in intensity at 2 hr and was not apparent after 4 hr. Because the 32,000-38,000 Mr fragments were not defined in the PAS-stained gradient gel, 30,ul (45 Mg of fibrinogen) of each reduced sample was subjected to electrophoresis in 10% acrylamide gels and stained with PAS; the 32,000-38,000 Mr protein bands were PAS positive. Functional and Structural Assessment of Degradation of Fibrinogen. The functional consequences of fibrinogen cleavage by neutral peptide-generating protease were assessed by incubation of 1.8 Mug and 6.0 Mug of neutral peptide-generating protease with 6.0 mg of fibrinogen for various periods of time up to 4 hr. After 0, 5, 10, 30, 60, 120, and 240 min, 250-MI samples of each reaction mixture were removed and made 10 mM in soybean trypsin inhibitor. Each sample was evaluated for the molecular weight of unreduced and reduced fibrinogen degradation products, thrombin-induced clotting activity, capacity to form crosslinked fibrin, and presence of anticoagulant activity. To characterize the molecular weight of fibrinogen degradation products, we subjected duplicate 7.5-,Ml (10.5 Mg of fibrinogen) portions of each sample to gradient electrophoresis with and without reduction. To determine thrombin-induced clotting activity and the capacity to form crosslinked fibrin, we added 10 Ml of thrombin/calcium solution to 100-Ml portions of each sample; each portion was assessed for thrombin-induced clotting time and then incubated for 3 hr at room temperature before analysis for y-'y dimers by gradient electrophoresis of solubilized samples. A separate 100-Ml portion of each sample was analyzed for the presence of anticoagulant activity. As analyzed by gradient electrophoresis of unreduced and reduced samples, neutral peptide-generating protease digestion at a ratio of 0.3 Mug of protease/1.0 mg of fibrinogen reduced fibrinogen to derivatives of 325,000-340,000 Mr, 320,000330,000 Mr, 310,000-325,000 Mr, and 290,000- ,000 Mr by 5, 10, 60, and 240 min, respectively. The Aa chain (67,00070,000 Mr) was progressively cleaved to its 42,000 Mr derivative, and this process was complete at 60 min. Digestion of the BO3 chain (58,000 Mr) to a 51,000 Mr fragment and of the y
Medical Sciences: Wintroub et al.
chain (45,000 Mr) to 32,000-38,000 Mr fragments was detected at 60 min, but the majority of both chains remained intact at 4 hr. During the first 5 min of digestion there was no loss of thrombin-induced clotting activity. After 5 min of fibrinogen digestion there occurred a time-dependent prolongation of the clotting time so that by 30 min the clotting time was greater than 3 min (Fig. 5). Thrombin-induced clotting resulted in formation of fibrin that was normal in appearance through the initial 30 min of digestion but was less well formed after 60 min of digestion. Gradient electrophoresis of solubilized thrombin clots disclosed y-,y dimers at each time point. However, the y-'y dimers progressively diminished in staining intensity in samples obtained at 60 min and after. Anticoagulant activity was not detected at any time during fibrinogen digestion. Neutral peptide-generating protease digestion at a ratio of 1.0 Asg of protease/1.0 mg of fibrinogen reduced fibrinogen to derivatives of 300,000-310,000 Mr and 270,000-280,000 Mr after 10 min and 120 min, respectively. The Aa chain (67,000-70,000 Mr) was completely cleaved to its 42,000 Mr derivative at 10 min. Digestion of the Bf3 chain (58,000 Mr) to a 51,000 Mr fragment was detectable at 10 min and progressed until no intact BB chain was detectable at 240 min. Digestion of the Y chain (45,000 Mr) to 32,000-38,000 Mr derivatives was evident at 30 min and was progressive, although a small amount of intact y chain was detected after 240 min. The thrombininduced clotting time was greater than 3 min after only a 10min incubation. Thrombin treatment of 10-, 30-, and 60-min digests resulted in poorly formed fibrin, and no fibrin was detectable after 120 min of digestion. The 'y-'y dimers dinished in quantity through the first 60 min of digestion and were not detectable after 120 min. Anticoagulant activity was not detected at any time. DISCUSSION The human neutrophil contains a neutral protease, previously designated the neutrophil neutral peptide-generating protease (10, 11), that cleaves fibrin and fibrinogen with an activity comparable on a weight basis to that of plasmin (Figs. 1 and 2). The neutral peptide-generating protease is a single polypeptide chain of 29,000-30,000 Mr with an isoelectric point of pH 7.7-8.3. It is distinguished from previously recognized neutral proteases of the human neutrophil, elastase and cathepsin G, by its failure to cleave their synthetic and natural substrates (8). A third neutrophil neutral protease, collagenase, is at least twice as large as these three neutrophil proteases (19). It seems likely that the activity of neutral peptide-generating protease would
20 30 50 40 Incubation time, min FIG. 5. Time-dependent alterations of thrombin-induced clotting time of fibrinogen during interaction with 0.3 ,gg of neutral peptidegeherating protease per mg of fibrinogen.
Proc. Natl. Acad. Sci. USA 77 (1980)
account for some of the divergent results obtained when the cleavage of soluble fibrinogen by purified leukocyte elastase (20) is compared with the results observed with crude leukocyte extracts (21, 22). Neutral peptide-generating protease cleaved soluble fibrinogen in a time-dependent manner into progressively smaller fragments that differed from those derived during plasmin degradation. After 5 min of digestion with 0.3 ,g of neutral peptide-generating protease per mg of fibrinogen, a major cleavage fragment of 325,000 Mr was recognized by NaDodSO4/polyacrylamide gel electrophoresis of the unreduced protein, and there was a gradual and progressive reduction in size of this fragment to derivatives of 310,000325,000 Mr at 60 min and 270,000-280,000 Mr by 24 hr (Fig. 3). The only other distinct fragments observed at each time point in the unreduced gels had Mr less than 10,000. Anarysis of the degradation products after reduction and alkylation revealed loss of the 67,000-70,000 Mr bands corresponding to the Aa chain and the appearance of bands with 55,000 Mr and 42,000 Mr by 5 min. At this time, the 58,000 Mr B/3 chain and the 45,000 Mr 'Y chain remained intact. Cleavage of the Bf3 chain with the appearance of a PAS-positive 51,000 Mr fragment was seen at 4 hr in the gradient gels (Fig. 4). The y chain disappeared coincidently with the appearance of PAS-positive 32,000-38,000 Mr bands at 4 hr (Fig. 4). Although each of the three chains of fibrinogen is cleaved by the neutral peptidegenerating protease, the order of polypeptide chain digestion is most consistent with Aa before Bf3 and oy. Cleavage of fibrinogen by neutral peptide-generating protease at a ratio of 0.3 and 1.0,Mg of protease/1.0 mg of fibrinogen resulted in 310,000-325,000 Mr derivatives that were slowly clottable and 270,000-280,000 Mr fragments that were not clottable. The prolongation of the thrombin-induced clotting time (Fig. 5) observed with the 310,000-325,000 Mr derivatives resulted from the Aa-chain cleavage (Fig. 4), but the derivatives' capacity to form crosslinked fibrin was retained, as shown by detection of y-'y dimers. The loss of coagulability that characterized the 270,000-280,000 Mr fragments was associated with the appearance of the 51,000 Mr Bf3-chain fragment, the 32,000-38,000 Mr y-chain fragments, and inability to form -y-y dimers. Because anticoagulants were not detected, the initial prolongation of thrombin-induced clotting time is attributed to the extensive Aa-chain damage, whereas the subsequent loss of coagulability is due to Bfl- or y-chain digestion (or both). The functional results and pattern of fibrinogen cleavage by the neutral peptide-generating protease differ from those resulting from digestion by plasmin, leukocyte elastase, and leukocyte extracts. Plasmin-limited proteolytic digestion of fibrinogen is characterized by a number of cleavages in the carboxy-terminal portion of the Aa chain (23, 24) and removal of about 40 residues from the amino-terminal portion of the B3 chain, including fibrinopeptide B (25), to yield fragment X (260,000-300,000 Mr), which exhibits prolonged thrombininduced clotting activity (26). Although the 310,000-325,000 Mr fragment obtained during neutral peptide-generating protease cleavage of fibrinogen was able to be coagulated, the 270,000-280,000 Mr derivative was not. Fragment X is split asymmetrically by y-chain cleavage and by further a- and 3-chain cleavage to yield the anticoagulant fragments, fragment Y (153,000 Mr) and fragment D (85,000 Mr) (27). At no time during degradation by neutral peptide-generating protease at enzyme/substrate ratios of 0.3,Mg or 1.0 Mg of protease/1.0 mg of fibrinogen was anticoagulant activity detected. The a, 3, and y chains of fragment Y are further cleaved to yield another D fragment and an E fragment of 50,000 Mr (23). Core fragments
Medical Sciences: Wintroub et al.
of 50,000 Mr were not detected during neutral peptide-generating protease digestion of fibrinogen at ratios of 0.3 jig and 1.0 ,ug of protease/1.0 mg of fibrinogen. However, 70,000 Mr and 58,000 Mr fragments were recognized at a dose of 20.0 ug of protease. Unlike plasmin-derived D and E fragments, the neutral peptide-generating protease-derived fragments were transient degradation products and did not arise from limited proteolysis. The proteolytic activity of a leukocyte granule extract for fibrinogen was distinguished from that of plasmin by the generation -of a high molecular weight fragment composed of cleaved Aa, B(3, and y chains and by the continued degradation to smaller fragments unlike those seen from the limited proteolytic cleavage by plasmin (21, 22). Because these latter studies were carried out with fractions enriched for the neutrophil granules, the activities observed cannot be attributed to a single protease; it is likely that elastase and cathepsin G are present, both of which have fibrinogenolytic activity (6). Low-dose cleavage of fibrinogen by leukocyte elastase (0.3 ag of elastase per mg of fibrinogen) cleaved the A a chain with little effect "on the By3 and y chains, whereas higher concentrations cleaved all three chains and resulted in the appearance of fragments having the antigenic characteristics of plasmin fragments D and E (20). In addition, digestion with elastase was awsciated with the development of anticoagulant activity (20). Because the purified neutral peptide-generating protease is devoid of elastase activity and cleaves fibrinogen in a manner distinctly different from that of plasmin, it represents a neutrophil peutral protease with fibrinolytic and fibrinogenolytic activities. It is tempting to speculate that the function of this protease is to facilitate neutrophil movement through vascular and extravascular sites by generating a vasoactive peptide and controlling local fibrin deposition. This work was supported by Grants AI-07722, AI-10356, AM-05577, HL-17382, HL-19777, and RR-05669 from the National Institutes of Health. J.S.C. was a postdoctoral trainee supported by Training Grant AM.07031 from the National Institutes of Health. B.U.W. is the recipient of a Clinical Investigator Award (AM-00430) from the National Institutes of Health. 1. Rulot, H. (1904) Arch. Int. Physiol. 1, 152-158. 2. Opie, E. L. (1907) J. Exp. Med. 9,391-413. 3. Welch, W. H. (1887) Trans. Pathol. Soc. (Philadelphia) 13, 281-300.
Proc. Nati. Acad. Sci. USA 77 (1980) 4. Barnhart, M. I. (1965) Fed. Proc. Fed. Am. Soc. Exp. Biol. 24, 846-853. 5. Riddle, J. M. & Barnhart, M. I. (1964) Am. J. Pathol. 45,805823. 6. Schmidt, W. & Havemann, K. (1974) Hoppe-Seyler's Z. Physiol. Chem. 355, 1077-1082. 7. Granelli-Piperno, A., Vassali, J. & Reich, E. (1977) J. Exp. Med. 146, 1693-1706. 8. Coblyn, J. S., Austen, K. F. & Wintroub, B. U. (1979) J. Clin. Invest. 63, 998-1005. 9. Silverstein, R. M. (1973) Anal. Biochem. 65,500-506. 10. Wintroub, B. U., Goetzl, E. J. & Austen, K. F. (1974) J. Exp. Med.
140,812-824. 11. Wintroub, B. U., Goetzl, E. J. & Austen, K. F. (1977) Immunology 33,41-49. 12. Rosenberg, R. D. & Damus, P. S. (1973) J. Biol. Chem. 248, 6490-6505. 13. Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244, 44064412. 14. Deutsch, D. & Mertz, E. (1970) Science 170, 1095-1096. 15. Gordon, S., Werb, Z. & Cohn, Z. A. (1976) in In Vitro Methods in Cell-Mediated and Tumor Immunity, eds. Bloom, B. & David, J. (Academic, New York), pp. 343-352. 16. Binder, B. R., Spragg, J. & Austen, K. F. (1979) J. Biol. Chem. 254, 1998-2003. 17. Zacharius, R. M., Zell, T. E., Morrison, H. & Woodlock, J. J. (1969) Anal. Biochem. 30,148-152. 18. Doolittle, R. (1975) in The Plasma Proteins: Structure, Function and Genetic Control, ed. Putnam, F. (Academic, New York), pp. 110-162. 19. Ohlsson, K. & Olsson, I. (1973) Eur. J. Biochem. 36,473-481. 20. Gramse, M., Bingenheimer, C., Schmidt, W., Egbring, R. & Havemann, K. (1978) J. Clin. Invest. 61,1027-1033. 21. Plow, E. F. & Edgington, T. S. (1975) J. Clin. Invest. 56,3038. 22. Bilezikian, S. & Nossel, H. (1977) Blood 5,21-28. 23. Pizzo, S., Schwartz, M., Hill, R. & McKee, P. (1972) J. Biol. Chem.
247,636-645. 24. Fullan, M. & Beck, E. A. (1972) Biochim. Blophys. Acta 263, 631-644. 25. Lahiri, B. & Shainoff, J. R. (1973) Biochim. Biophys. Acta 303, 161-170. 26. Marder, V. J., Shulman, N. R. & Caroll, W. R. (1969) J. Blol.
Chem. 244,2111-2119. 27. Marder, V. J. & Shulman, R. (1969) J. Biol. Chem. 244,21202124.