THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol 253, No 7, Issue of April 10, pp 2X34-2195, 1978

Printed in U S A

Localization of the a-Chain Cross-link Acceptor Sites of Human Fibrin*

(Received for publication, August 22, 1977)

LARRY J. FRETTO,* EARL W. FERGUSON,§ HOWARD M. STEINMAN,~ AND PATRICK A. McKEE~~

From the Departments of Medicine and Biochemistry, Duke University Medical Center, and Veterans Administration Hospital, Durham, North Carolina 27710

The potential cross-link acceptor sites of fibrin were specifically labeled with the fluorescent, substitute cross-link donor monodansyl cadaverine (MDC). Several fluorescent a-chain peptides generated from enzymatic and cyanogen bromide (CNBr) cleavage of the labeled fibrin were identi- fied by sodium dodecyl sulfate disc gel electrophoresis; they were isolated and then characterized by amino acid analy- sis, NH,-terminal sequence analysis, and chromatographic and electrophoretic analyses of their digestion products. Ancrod cleavage of MDC-labeled fibrin produced a series of six a-chain peptides of molecular weights 34,000 to 12,000, each of which contained an MDC-labeled acceptor site, and an NH,-terminal a-chain derivative of molecular weight 37,500. The latter remains disulfide bound in the residual fibrin and has two MDC-labeled sites which are separable by CNBr cleavage. Mild plasmin digestion of MDC-labeled fibrin generated fluorescent a-chain peptides of molecular weights 45,000, 42,000, 35,000, 23,000, 21,000, and 2,500 in the supernatant and a nonfluorescent NH,-terminal a-chain derivative of molecular weight 25,000 which remained in the insoluble residual fibrin. The alignment of these plasmic supernatant peptides was determined from NH,-terminal sequence analyses which indicated that an MDC acceptor site was located at approximately residue 255 of the Aa-chain. Cleavage of the MDC-labeled a-chain by CNBr, however, localized most of its fluorescence (-80%) to a fragment of molecular weight 29,000 which had the same NH,-terminal sequence as the labeled plasmic peptide of molecular weight 21,000. Both peptides were cleaved by ancrod into two acceptor site-containing peptides of approximately equal fluorescence. The preliminary NH,-terminal sequence anal- yses of these peptides, when combined with the above find- ings, indicated that these two other cross-link acceptor sites

‘i: This work was supported by research grants from the National Heart, Lung and Blood Institute, National Institutes of Health (HL 156151, and the Veterans Administration. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

f Postdoctoral Fellow, National Heart and Lung Institute. 8 Present address, Department of Biochemistry, Uniformed Ser-

vices University of Health Sciences, 6917 Arlington Road, Bethesda, Maryland 20014.

B Present address, Department of Biochemistry, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461.

/I Investigator of the Howard Hughes Medical Institute.

are in a peptide segment which comprises the middle 17% of the AcY-chain.

Human fibrinogen (M,. = 340,000) is composed of three pairs

of polypeptide subunits (AN, B/3, and y) which are joined by disullide bonds located mainly in their NH,-terminal regions; the biochemistry of the conversion of fibrinogen to fibrin has been recently reviewed (1, 2). Cleavage of the Aa- and B/3- subunits by thrombin releases the NH,-terminal fibrinopep- tides A and B and forms the N- and p-chains of fibrin monomer. Fibrin monomers then associate into an ordered matrix of uncross-linked fibrin. In the presence of activated fibrin-stabilizing factor and calcium, two intermolecular t-(7- glutamyl)lysine cross-links are rapidly formed between lysyl donor groups and glutaminyl acceptor groups in antiparallel y-chains to give the y-dimer subunit of partially cross-linked fibrin. Additional cross-linking occurs more slowly between o-chains to produce the o-polymer subunit of stabilized, fully cross-linked fibrin. The p-chains of fibrin do not cross-link. Each fibrin monomer is reported to contain four to six cross- link acceptor-donor pairs; each y-chain has one such pair and each o-chain has either one or two such cross-link pairs. In the y-chain it was shown that both the acceptor and donor

cross-link sites are located within 15 residues of the COOH terminus.

In the case of the a-chain, previous reports have proposed that both the COOH- and NH,-terminal halves may contain cross-link sites (3-5). Earlier studies from this laboratory demonstrated that the o-chain is clearly the first of the subunits to be cleaved by plasmin and that, with the degra- dation of fibrinogen to Fragment X, the cross-linking potential of the o-chain is lost while that of the y-chain remains (6). This suggested that the acceptor and donor cross-link sites probably reside in the COOH-terminal two-thirds of the (Y- chain, which is released to give Fragment X. We previously confirmed this localization for the acceptor sites by utilizing the observations that monodansyl cadaverine is a site-specific fluorescent label for the glutaminyl cross-link sites of fibrin (7, 8). The location of one of these sites, which was labeled with a different substitute donor compound, was proposed to be near the middle of the o-chain (9). However, Doolittle and co-workers have recently indicated that this glutaminyl resi-

due was not cross-linked in fibrin prepared from purified fibrinogen (10). The purpose of this communication is to

2184

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Fibrin a-Chain Cross-link Acceptor Sites 2185

provide evidence for the specific location in the a-chain of three potential cross-link acceptor sites. We report the results of time course digestions of MDU-labeled a-chains of mono- meric fibrin by plasmin, ancrod, CNBr, and trypsin. Fluores- cently labeled peptides from such digestions were identified, isolated, and partially characterized by sodium dodecyl sul- fate-gel electrophoresis, amino acid compositions, and NH,- terminal sequence analysis.

MATERIALS AND METHODS

Reagents -All reagents were analytical grade and were obtained from commercial sources. TrisiNaCl buffer refers to 0.15 M NaCl, 0.05 M Tris/HCl, pH 7.4. Ammonium bicarbonate buffer refers to 0.10 M ammonium bicarbonate, pH 7.9. MDC (Cycle Chemical Company) was made 7.5 rn~ in TrislNaCl buffer and stored at 4”. Iodoacetamide was recrystallized twice from hexane. Pyridine was redistilled prior to use.

Fibrinogen -Human fibrinogen (Grade L, Kabi, Stockholm) was -95% clottable and contained sufficient fibrin-stabilizing factor to catalyze highly cross-linked fibrin formation. The plasmin formed from urokinase activation of the contaminant plasminogen in this fibrinogen was adequate to fully digest the fibrinogen or fibrin formed from this preparation.

Enzymes -Human urokinase (Leo Pharmaceuticals, Copenhagen) was dissolved in TrisiNaCl buffer to a concentration of 5000 Ploug units/ml and stored at -20”. Human thrombin (from Dr. D. L. Aronson, Bureau of Biologics, Food and Drug Administration, Bethesda, MD) was dissolved immediately before use in Tris/NaCl buffer at 90 NIH units/ml. Human plasminogen was isolated as previously described (11) and stored as a precipitate in 40% ammo- nium sulfate at 4”. Ancrod (Abbott Laboratories) was supplied in concentrations of 100 and 900 units/ml and used without further purification. Trypsin treated with L-1-tosylamido-2-phenylethyl chloromethyl ketone and a-chymotrypsin (Worthington Biochemical Corp., Freehold, NJ) were each dissolved in 0.001 N HCl, 0.02 M CaCl, before use.

Polyacrylamide Gel Electrophoresis -Electrophoresis in the pres- ence of SDS was performed on 5, 7.5, and 12.5% polyacrylamide gels which were examined for fluorescence and stained for protein or carbohydrate as previously described (7). Also SDS-20% gels were prepared according to Laemmli (12, 13); SDS, 6 M urea, 10% gels containing 1.2% bisacrylamide were used to enhance resolution of peptides in the range of 8,000 to 20,000 molecular weight. Certain isolated peptides were examined for purity by polyacrylamide gel electrophoresis at pH 3.2 and 8.9 (14, 15).

Thin Layer Chromatography -Ascending chromatography in pyr- idine/butanol/acetic acid/water (10:15:3:12) on plastic-backed cellu- lose (20 x 20 cm, 0.1 mm, MN 300, Brinkmann) was frequently used to assess the purity of peptides separated by column chromatography and to isolate certain small MDC-labeled peptides. These were then eluted from cellulose scrapings with 50% acetic acid.

Preparation of MDC-labeled Fibrin -Incorporation of MDC into the cross-link acceptor sites of fibrin monomer was based on the method of Lorand et al. (8). The fibrin was formed by combining 1 volume of 1% fibrinogen solution, 2 volumes of 7.5 rn~ MDC, 0.1 volume of freshly prepared 0.25 M cysteine (adjusted to pH 7.41, 0.1 volume of 0.8 M CaCl,, and 0.05 volume of thrombin (90 NIH units/ ml). The mixture began to clot at about 20 s; it was covered and left for 4 h at room temperature. Under these conditions only the (Y- and y-chains of fibrin became fluorescently labeled, and this inhibited most of the y-dimerization and virtually all a-polymerization. The clot was then transferred to a Buchner funnel and washed exhaus- tively with 0.2 M NaCl (12 liters/g of fibrin) and then with cold ammonium bicarbonate buffer (3 liters/g of fibrin). This was done to ensure removal of excess free MDC, which migrated on SDS-gels between M, = 11,000 and 6,000 and could thus hamper interpretation of the SDS-gel electrophoretic analyses of the MDC fibrin digestion products.

Amrod Digestion-One gram of MDC-labeled fibrin was incu- bated with 1000 units of ancrod in 50 ml of ammonium bicarbonate buffer at 37” with constant shaking; 1000 units of ancrod were again

-

1 The abbreviations used are: MDC, monodansyl cadaverine; SDS, sodium dodecyl sulfate; CNBr, cyanogen bromide; NPGB, p-nitro- phenyl-p-guanidino benzoate-HCl.

added at 2, 5, and 8 h. Two-milliliter aliquots of supernatant and 2 to 5 mg of the particulate fibrin were removed approximately every 2 h, lyophilized, and analyzed by SDS-gel electrophoresis. After 11 h the remaining supernatant was removed by filtering through What- man No. l-MM paper on a Buchner funnel and was combined with the 250-ml filtrate recovered from washing the residual ancrod- treated fibrin This particulate fibrin was finely minced, resus- pended at 20 mgiml in ammonium bicarbonate buffer, and reincu- bated with 40 units of ancrod/ml for 5 h at 37”. Following this second digestion, the supernatant was removed as before, and the fibrin was washed, lyophilized, and subjected to CNBr cleavage as de- scribed below. The lyophilized supernatant protein from the initial (11 h) and the second (16 h) ancrod digests of MDC-labeled fibrin was combined and dissolved at 20 mg/ml in ammonium bicarbonate buffer and incubated at 37” with 60 units of ancrod/ml which were added at 0 h and again at 3 h. After 5 h the digest was lyophilized, analyzed by SDS-gel electrophoresis and stored for use in other studies as described below.

Ancrod digests of MDC-labeled a-chain and several purified fluorescent a-chain peptides were performed at 37” in ammonium bicarbonate buffer. The substrate and enzyme concentrations were in the range of 2 to 5 mg/ml and 5 to 70 units/ml, respectively.

Plasmin Digestion - MDC-labeled fibrin was minced, suspended at 20 mg/ml in ammonium bicarbonate buffer, and an aliquot of the stock urokinase solution was added to give a final concentration of 10 Ploug units/ml. The mixture was incubated at 37” with constant shaking for 6 h, by which time the fibrin had completely lysed. To obtain larger molecular weight a-chain peptides, several limited digestions were performed at 23” with stirring for 2 h or less and terminated with NPGB. Following centrifugation (3500 x g, 10 min) of the incubation mixture, the supernatant fluid was frozen imme- diately, lyophilized, and stored at -20” until it could be further purified.

Plasmin digests of fibrinogen were prepared by incubating 1 g of fibrinogen at 5.7 mg/ml with urokinase at 5.7 Ploug units/ml in ammonium bicarbonate buffer at 37” with constant shaking. De- pending on the size of the peptides sought, the reaction was termi- nated at 45 min or 3 h by adding 1.2 mg of NPGB and then immediately freeze drying.

The particulate phase of ancrod-digested MDC-labeled fibrin was incubated at 10 mg/ml with 1 mg of plasminogen and 150 Ploug units of urokinase in 15 ml of ammonium bicarbonate buffer at 23” with constant stirring. Digestion was terminated at 90 min by adding 0.3 mg of NPGB; at this time only a smali amount of precipitate remained.

MDC-labeled n-chain peptides, isolated from CNBr or enzymatic cleavages, were each dissolved in an appropriate volume of ammo- nium bicarbonate. Plasmin was prepared by incubating plasminogen at 5 mg/ml in 0.05 M Tris, 0.05 M lysine/HCl, pH 8, with urokinase at 120 Ploug unitslmg for 20 min at room temperature; the plasmin was then diluted to 1 mg/ml with cold ammonium bicarbonate buffer containing 20% glycerol and kept at 4”. About 2 to 5 mg of each peptide were incubated with three to four serial additions of 1 to 3% (w/w) plasmin at 23” or 37” for up to 24 h.

CNBr Cleavage -One gram of MDC-labeled fibrin was dissolved in 100 ml of 70% formic acid by vigorous stirring for 1 h at room temperature and dialyzed against three changes of 75.fold volumes of 70% formic acid at 4”. Solid CNBr was added at room temperature with stirring to the final dialysate to make the solution 3% CNBr, which was a 150-fold molar excess of CNBr relative to the methio- nine content. SDS-gel electrophoretic analysis of aliquots at various times indicated that CNBr cleavage was complete after 2 h. After 7 h the reaction was terminated by diluting the mixture 12-fold with water and lyophilizing. Cleavage of 5 g of fibrinogen was performed as for labeled fibrin, except that fibrinogen was dialyzed against water at 4” and lyophilized before being made 1% in 70% formic acid.

Ancrod-treated MDC-labeled fibrin remaining from the 16-h an- trod digest was dissolved at 10 mg/ml in 70% formic acid (82 ml) and mixed with 2.5 g of solid CNBr for 18 h at 23”; the reaction was terminated as before.

Approximately 45 mg of MDC-labeled S-sulfitolyzed a-chain was reacted with 140 mg of CNBr in 5 ml of 70% formic acid for 8Vz h at 23”. CNBr cleavages of isolated peptides were performed at concen- trations of 5 to 9 mg/ml in 70 to 80% formic acid and molar ratios of CNBrmethionine in the range of 200 to 350.

Digestion by Trypsin and Chymotrypsin -Digestion of certain acceptor site-containing peptides was usually performed in ammo- nium bicarbonate buffer at 37” by three serial additions of 0.02 mg of

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

2186 Human Fibrin a-Chain Cross-link Acceptor Sites

enzyme/mg of peptide. In some cases the substrate solutions, which ranged in concentration from 2 to 20 mg/ml, were placed in a boiling water bath for 40 s before initiating hydrolysis. Tryptic digestion of P21* was also performed in 2 M urea (161.’ The reaction mixtures were maintained between pH 7.5 and 8.0, and the hydrolyses were allowed to continue up to 24 h. Aliquots which were withdrawn from the digests at selected times were lyophilized and analyzed by SDS- gel electrophoresis.

Peptide Isolation -Many of the fluorescent peptides released into the supernatant by either plasmin or ancrod digestion of MDC- labeled fibrin were purified by gel filtration at room temperature on Ultrogel AeA44 (LKB) and on Sephadex G-50 Fine (Pharmacia Fine Chemicals) equilibrated in ammonium bicarbonate buffer at flow rates of 3 to 3.5 ml/h/cm’. The eluant of these and the other chromatography experiments was continuously monitored for ab- sorbance at 280 and 206 nm and for fluroescence (365 nm primary filter and 540 nm secondary filter, Turner model 111 fluorometer). Aliquots from selected fractions of the chromatograms were rou- tinely analyzed by SDS-gel electrophoresis to decide which fractions would be pooled.

The fluorescent CNBr fragment (CN29*) from the cleavage of MDC-labeled fibrin or of isolated MDC-labeled o-chain, or the analogous peptide (CN29) generated by CNBr cleavage of fibrinogen, was isolated by successive gel filtrations. Digests of 1 g or more were dissolved at 60 mg/ml in 10% acetic acid; l&ml samples were applied to Sephadex G-100 (2.5 x 90 cm) and eluted at 18 ml/h in the same solvent. As previously reported, this step produced a broad elution profile (171, and the fluorescence was quenched by this solvent. The pool between 0.63 and 0.92 V, was concentrated to 3 ml and rechromatographed on the same column. The fractions (0.4 to 0.6 V,) containing CN29* and moderate amounts of Hol-DSK” were lyophi- lized, suspended at 9 mgiml in 0.15 M NH,HCO,,, 0.07 M dithiothrei- tol, pH 7.8, and stirred overnight at room temperature in a sealed vial under nitrogen. The supernatant (8 to 10 ml) was filtered on Sephadex G-100 (2.6 x 90 cm) equilibrated with freshly prepared 0.15 M NH,HCO,, 0.07 M dithiothreitol, pH 7.8; to remove the dithiothreitol, the pool containing CN29* was exhaustively dialyzed at 4” against water adjusted to pH 4.

Isolation of the MDC-labeled o-chain peptide, CNAlO*, from the CNBr cleavage of ancrod-treated fibrin was achieved by gel filtration on Ultrogel AcA44 and then on Sephadex G-50 equilibrated in ammonium bicarbonate buffer. Although this was a mediocre sol- vent for virtually all of the CNBr fragments, CNAlO* was very soluble, and this permitted its preparation in large quantities. This method was also used to purify CN29* in small quantities (5 to 10 mg); in this way reduction was unnecessary to achieve its separation from Hol-DSK.

The fluorescent o-chain derivative, A38*N, in ancrod-treated MDC-labeled fibrin was isolated by electrophoresis on SDS-7.5% gels (0.5 x 10 cm). The lyophilized precipitate obtained from the ancrod digest was dissolved at 5 mg/ml in 1% SDS, 0.01 M sodium phosphate, 8 M urea, 5% /3-mercaptoethanol, 0.002% bromphenol blue, pH 7.1; 55 ~1 of this solution was applied to each gel and electrophoresed at 4.5 mA/gel for 12 h. Under these conditions, the brightly fluorescent band which corresponded to the o-chain derivative was well re- solved from the fluorescent y-chain; this band was immediately sliced from the gels while viewing under an ultraviolet hand lamp. The gel slices were kept at 4” until enough had been obtained and then they were ground to a paste in a glass mortar, suspended at -0.7 ml/slice in 0.015% SDS, 0.2% /3-mercaptoethanol and stirred at room temperature for 48 h. This slurry was filtered, and the fluorescent supernatant was lyophilized and redissolved in water at 25 mgiml to which an equal volume of ice-cold 20% trichloroacetic acid was added while stirring. The precipitate which formed (about 0.5 mgiml) was dissolved in 6 M guanidine HCl, 0.55 M Tris/HCl, 0.005 M EDTA, 2% P-mercaptoethanol, and carboxymethylated with l’4Cliodoacetate (18).

’ All peptides referred to by abbreviations in this paper are of a- chain origin. Those labeled with MDC are designated by an asterisk following a number (M, x 10m3) which indicates their approximate molecular weight. The prefix indicates the agent(s) of proteolysis that generated the peptide: A, ancrod; CN, cyanogen bromide; P, plasmin; T, trypsin. The suffix “N” is used for peptides which are derived from the NH,-terminal region of the n-chain of fibrin.

’ The designation Hol-DSK is according to Blomback and refers to the disulfide-bonded CNBr fragment from the Fragment D region of fibrinogen (17).

Tryptic, chymotryptic, and plasmic peptides of molecular weight less than 10,000 were purified by gel filtration followed by either ion exchange or thin layer chromatography. A Sephadex G-25 column (2 x 98 cm) equilibrated in ammonium bicarbonate buffer was generally used for the first fractionation. The ion exchange chroma- tography conditions were (a) DEAE-Sephadex A-25 (0.9 x 25 cm) which was equilibrated with 0.05 M Tris, 0.02 M NaCl, pH 8.3, and developed with a NaCl gradient and (b) CM-Sephadex C-25 (0.9 x 25 cm) which was equilibrated in 0.02 M ammonium acetate and developed with increasing molarity of the same buffer. Thin layer chromatography was performed as described above.

Ultrocelztrifugal Analysis -The weight average molecular weight of the CNBr fragment, CNAlO*, from the ancrod-cleaved o-chain in MDC-labeled fibrin was determined in 0.1 M ammonium bicarbonate by the meniscus depletion sedimentation equilibrium method of Yphantis (19). A Spinco model E analytical ultracentrifuge equipped with absorption optics, a scanner, an An-D rotor, and a 12-mm double-sector centerpiece was used. The sample was analyzed at 18”, at a concentration of 83 pg/ml, and at rotor speeds of 40,000 and 48,000 rpm. The partial specific volume of CNAlO*, as calculated from its amino acid composition (201, was 0.679 ml/g.

MDC Content of Labeled Peptides-The mol fraction of MDC incorporated in certain small o-chain peptides was estimated from the near ultraviolet absorption spectra of the labeled peptides in ammonium bicarbonate buffer. The molar extinction coefficients of free MDC in this solvent were determined (eQBo npl = 4,000 M-I cm-‘, %0 nm = 1,540 M-’ cm-‘, Eqq(inm = 14,750 M-’ cm-‘) and were assumed to be essentially the same as those of the peptide-bound MDC. Independent estimates of MDC incorporation were made from the visible fluorescence emission spectra of the labeled peptides in ammonium bicarbonate buffer with an Aminco-Bowman spectropho- tofluorometer; 345 nm was the excitation wavelength used.

Amino Acid Analyses -Isolated peptides and proteins were hydro- lyzed in cacao at 110” in 1 to 2 ml of 6 N HCl containing 0.1% (w/v) phenol for 22, 72, and occasionally 48 and 120 h. Each hydrolysate was analyzed by standard chromatographic methods (21, 22) on Beckman type AA-15 resin (0.9 x 56 cm) with a Beckman model 119 amino acid analyzer equipped with an Autolab System AA com- puter. Only the 570 nm channel was integrated; proline values were computed manually. Methionine and cysteine were determined as methionine sulfone and cysteic acid from 20-h hydrolyses following performic acid oxidation (23). Tryptophan was determined from separate 22-h hydrolyses at 110” with 0.5 ml of 4 N methane sulfonic acid containing 0.2% 3.(2-aminoethyl)indole (24).

NH,-terminal and Amino Acid Sequence Studies - NH,-terminal analyses of 0.04 to 0.10 pmol of several isolated o-chain peptides containing a MDC-labeled cross-link acceptor site(s) were performed by the qualitative dansyl chloride technique (25, 26) with thin layer chromatography on 7.5~cm square polyamide sheets (25, 27). Certain MDC-labeled a-chain peptides (0.2 to 0.5 pmol) were subjected to amino acid sequence analysis by automated Edman degradations performed with a Beckman 890B Sequencer; a program was used which incorporated modifications of the original method of Edman and Begg (28, 29). Details of these procedures have been previously described from this laboratory (11).

RESULTS

Ancrod Digestion of MDC-labeled Fibrin -Incubation of MDC-labeled fibrin with ancrod results in selective cleavage of the a-chain into two fluorescent regions as evinced from SDS-gel electrophoretic analyses of the digestion products. As shown in Fig. 1, ancrod rapidly released fluorescent peptides of molecular weight 34,000 to 27,000 which could be detected in the supernatant of the digest after 20 min. With time these early products were degraded to a peptide of about M, = 12,000 which contained most or all of the MDC label present in each of the parent peptides. Although the apparent absence of nonfluorescent supernatant peptides in Fig. 1 could be due to their co-migration with labeled peptides, the most likely explanation is that ancrod sequentially removed only small, M, = 3,000 to 5,000, nonfluorescent peptides to produce the observed digestion pattern. The fact that the sum of the molecular weights of the intermediate fluorescent peptides exceeds the molecular weight of the a-chain also supports this

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

NR N R

a-Chain Cross-link Acceptor Sites 2187

NR N R

Mr FIG. 1. SDS-7.5% gel analyses of re-

duced (R) and nonreduced (N) samples of the supernatant removed with t&e from the ancrod digestion of MDC-la- beled fibrin. The gels were examined for protein (P) and fluorescence (F). No fluorescence or protein was detected in the supernatant at “time zero.” M, indicates the apparent molecular weights of the reduced peptides. See text for details of incubation.

Hours of Digestion

explanation. The observed increase in SDS-gel electrophoretic mobility of the nonreduced relative to the reduced M, = 34,000, 31,000, 27,000, 24,000, and 19,000 intermediates indi- cates that these peptides contain one or more intrachain disulfide loops (Fig. 1). This observation becomes useful later in positioning these fragments in the o-chain. Finally, confir- mation of their a-chain origin was obtained from analysis of the residual insoluble ancrod-cleaved fibrin in the digestion mixture.

In Fig. 2 the time course of ancrod digestion of the MDC- labeled fibrin indicates that the p-chain, y-chain, and y-dimer subunits of fibrin were not detectably degraded even after incubation for 24 h; on the other hand, the a-chain was readily cleaved in half. The disappearance of the intact a-chain was noticeable after 1 h and coincided with the appearance of a brightly fluorescent M, = 38,000 peptide observed only in the reduced SDS-gel patterns. The generation of this a-chain fragment was maximal at about 16 h, and it was resistant to further cleavage by ancrod. This peptide is designated A38*N, and it contains the second cross-link acceptor-site region of the a-chain revealed by ancrod cleavage, the first being in the nondisulfide-bound peptides released into the supernatant.

Isolation of A38*N from SDS-polyacrylamide gel slices was achieved with about 95% yield. SDS-lo% gel analysis of the CNBr-cleaved A38*N revealed that (a) most of the fluorescent label was in a peptide of apparent molecular weight 17,000 which stained very poorly with Coomassie blue; (6) the remaining fluorescence appeared in a poorly staining peptide of molecular weight less than 2,000; (c) there were 3 or 4 small (M,. less than 10,000) nonfluorescent peptides detected by Coomassie stain. When examined by thin layer chromatogra- phy, the minor fluorescent peptide had a R,.. of 1.0, but the major fluorescent peptide remained at the origin; the three nonfluorescent ninhydrin-positive peptides were resolved be- tween RF 0.25 and 0.45. Gel filtration of the CNBr digest on Sephadex G-25 gave less than a milligram of the major MDC- labeled peptide; the minor fluorescent fragment was not detected. Treatment with plasmin (7% w/w, 37”, 12 h) had no detectable effect on the peptide of apparent molecular weight 17,000 and thereby indicated that its position in the a-chain should correspond to a plasmin-resistant segment.

Plasmin Digestion of MDC-labeled Fibrin -Very early in the time course of plasmin digestion of labeled fibrin, fluores- cent peptides of molecular weight 45,000, 42,000, and 35,000

FP FP FP FP

--A38*N

0 3 7 16 Hours of Digestion

FIG. 2. SDS-7.5% eel analvses of reduced samules of oarticulate fibrin removed with time from the ancrod digestion of MbC-labeled fibrin. The subunits of MDC-labeled fibrin are identified in the zero time sample. The gels were examined for protein (P) and fluores- cence (F). The band designated A38*N had an apparent molecular weight of 37,500 and was not present in no&educed samples.* Conditions of digestion are the same as in Fig. 1 (see text).

were released from the clot into the supernatant. Contrary to our earlier assessment (7), these large peptides apparently contain one or more intrachain disulfide loops as evidenced by their decreased electrophoretic mobility after reduction when analyzed by SDS-gel electrophoresis. These peptides become degraded fairly quickly to give two brightly fluorescent pep- tides of molecular weight 23,000 and 21,000 (designated P23* and P21*, respectively), the mobilities of which were unaf- fected by reduction (7). Smaller nonfluorescent peptides were also observed, one of which had a molecular weight of 18,000 and appeared to contain the disulfide loop(s), since its mobility was markedly decreased after reduction. With time P23* was degraded to P21* which then remained relatively plasmin- resistant for up to 6 h. This last cleavage generated a weakly fluorescent peptide of apparent molecular weight 2,500 as detected by SDS-12.5% gel electrophoresis (7).

The a-chain origin of these supernatant peptides was estab- lished by SDS-gel analyses of the insoluble portion of the

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

2188 Human Fibrin a-Chain Cross-link Acceptor Sites

digestion mixture which displayed intact p-chain, y-chain, andy-dimer subunits for the first hour as shown in Fig. 3. The disappearance of intact a-chain during this time coincided with the appearance of fluorescent peptides in the supernatant and also with the observation of two nonfluorescent peptides of M, = 25,000 and 23,500 (designated P25N and P24N, respectively) in the residual insoluble fibrin after reduction. Thus, in contrast to the digestion patterns of labeled fibrin produced by ancrod, all of the a-chain cross-link acceptor sites are contained in the plasmic supernatant peptides.

After 2 h, large fluorescent fragments corresponding to fibrinogen Fragments Y and D and fibrin Fragment D dimer appeared in the supernatant of the digestion mixture. How- ever, these derived fluorescence solely from the MDC-labeled -y-chain cross-link acceptor sites. Smaller, nonfluorescent pep- tides which correspond to Fragment E were also observed.

Isolation ofPlasmic a-Chain Peptides which Contain Cross- link Acceptor Sites-Fig. 4 is the elution profile of 2- and 6-h plasmic digests of MDC-labeled fibrin when filtered on Ultro- gel AcA44. The quantity of supernatant peptides recovered from the 2-h digestion was 140 mg per g of labeled fibrin; this represents 50% of the theoretical yield which was computed by assuming that two M, = 45,000 segments per fibrin monomer were released into the supernatant. SDS-gel analysis of the unfractionated supernatant is shown in the upper panel of Fig. 4, gels a and b. The major peak in the chromatogram of the 2-h digest contained approximately equal amounts of P23* and P21* and also smaller nonfluorescent peptides as demon- strated by SDS-7.5% gel analysis of Pool III (Fig. 4, gels c and d). Purification of P42* (Pool I) and P35* (Pool II) was achieved by rechromatography of each pool on the same Ultrogel column; the final yields were 10 to 12 mg of each per g of labeled fibrin. The peptides generated by the digestion of P42* and P35* to P23* were detected in the elution profile by absorption at 206 nm and displayed weak absorption at 280 nm and slight, if any, fluorescence. The weakly fluorescent peaks observed near the column volume (Pool V) probably

Y- -dimer --

FP FP FP FP

P25N

0 IO 40 100 Minutes of Digestion

FIG. 3. SDS-7.5% gel analyses of reduced samples of particulate fibrin removed with time from the plasmin digestion of MDC-labeled fibrin which was suspended at 10 mg/ml in 0.1 M ammonium bicarbonate, pH 7.8, at 37”. Plasmin added was 1% (w/w) of sub- strate. Gels were examined for protein (P) and fluorescence (F). The nonfluorescent band designated P25N is the NH,-terminal M, 25,000 a-chain derivative.2

contained the labeled peptide cleaved from P23* when P21* is formed.

When the 6-h digest was filtered on Ultrogel AcA44 (Fig. 4, lower panel), P21* was well resolved from Fragments D and E which eluted in a broad peak (Pool I) just after the void volume; it should be noted that considerable overlap occurred between P21* and Fragment E when filtered on Sephadex G- 100 whether or not Fragment D was removed by prior heat precipitation. In contrast to the 2-h digest, P23* was not detected in the chromatography of the 6-h digest. Approxi- mately 65% of the maximum theoretical yield of P21* was recovered in Pool II and after rechromatography on Sephadex G-50 (2.7 x 120 cm), the overall yield of P21* was 50%. This preparation was homogeneous by NH&erminal sequence anal- ysis and had an Et’&, (280 nm) of 18.0. Identical chromato- graphic results were obtained when unlabeled P21 was pre- pared from fibrinogen.

CNBr Fragments of MDC-labeled a-Chain -SDS electro- phoresis of the CNBr fragments from MDC-labeled fibrin showed two regions of fluorescence, a discrete protein band of molecular weight 29,000, the mobility of which was slightly decreased after reduction, and a more diffuse band of about M, = 8,000 (Fig. 5). The fluorescent intensity of this smaller fragment on SDS-12.5% polyacrylamide gels appeared to be

Relative Fluorescence - ?. PF PF RI “I 1.0

0.8

0.6

0.4 ‘;. 0 b c d

I

Fraction

ok/k----F+ I H-t-=+=-i- L J

700 900 1100 1300 1500 1700 1900 Elutron Volume (ml)

FIG. 4. Gel filtration on Ultrogel AcA44 (5 x 90 cm) of plasmin digests of MDC-labeled fibrin. The column was equilibrated and developed at 23” in 0.1 M ammonium bicarbonate, pH 7.9; flow rate = 3.0 ml/h/cm*. Upper panel, elution profiles of 2-h digest (75 mg) applied in 7 ml; 12-ml fractions were collected. Absorbance was measured in a 3-mm path flow cell. Inset, SDS-7.5% gel analyses of the reduced whole digest (gels a and b) and of nonreduced Pool III (gels c and d). Peptide designations are defined in text.2P, protein; F, fluorescence. Lower panel, elution profile of 6-h digest (1.6 g) applied in 41 ml; 17-ml fractions were collected. Absorbance values are for a l-cm path length. Pool II was -85% homogeneous. Inset, elution profile of 6-h digest on Sephadex G-100 obtained under equivalent conditions.

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Fibrin a-Chain Cross-link Acceptor Sites 2189

SDS 7.5 % GELS SDS 12.5% GELS

.--- -CN29*

1 Y4 MDC-FBN MDC - a

PPF PF

1 e; ; MDC-FBN MDC - a

FIG. 5. SDS-gel electrophoretic analyses of cyanogen bromide- cleaved MDC-labeled fibrin (MDC-FBN) and isolated MDC-labeled a-chain (MD&Y). Nonreduced (N) and reduced (R) samples were examined. Gels were analyzed for protein (P) and fluorescence (F). Conditions of cleavage and peptide designations’ are given in the text.

greater than that of the upper band. However, the SDS-gel pattern of the CNBr fragments from isolated, MDC-labeled o-chain showed that the major fluorescent product was the M, 29,000 fragment, designated CN29* in Fig. 5. The smaller (Y- chain fragments which stained with Coomassie blue as ob- served on SDS-7.5% and 12.5% polyacrylamide gels exhibited negligible fluorescence.

CN29* was isolated from the other a-chain CNBr fragments by gel filtration on Sephadex G-100 in 10% acetic acid and its homogeneity established by SDS-gel electrophoresis and NH2- terminal sequence analysis. Its extinction coefficient, E:?,,, (280 nm) was 14.4. The molar ratio of bound MDC to CN29*, as estimated spectrofluorometrically, was approximately 2.0, which was the same as the value obtained for P21*. Prepara- tion of pure CN29* from MDC-labeled fibrin (and also nonla- beled CN29 from fibrinogen) required an additional gel flltra- tion step in the presence of dithiothreitol to separate the contaminant nonfluorescent disulfide knot, Hol-DSK, of ap- parent molecular weight 35,000. As seen in Fig. 5 this latter fragment reduces to smaller peptides which are then readily separable from CN29*.

CNBr Cleavage of Ancrod-digested MDC-labeled Fibrin - As demonstrated in Fig. 2, ancrod digestion of MDC-labeled fibrin produced an NH,-terminal fluorescent a-chain deriva- tive which could be isolated under denaturing conditions. It was then cleaved with CNBr to obtain a smaller peptide which contained most (-70%) of the fluorescence in the digest. This acceptor site-containing CNBr fragment was more easily prepared from ancrod-digested MX-labeled fibrin and was purified by gel filtration in 0.1 M ammonium bicarbonate, which permitted the monitoring of fluorescence (Figs. 1s and 2~).~ Although the overall solubility of the CNBr digest in this buffer was less than with acetic acid, the fluorescent fragment

4 Portions of this paper (including Figs. 1s and 2s and Table Is) are presented in miniprint at the end of this paper. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, Maryland 20014. Request Document No. 77111-1358, cite author(s) and include a check or money order for $1.00 per set of photocopies.

of interest was very soluble. This peptide, designated CNAlO*, was recovered in 60% yield; it was homogeneous by SDS-20% gel electrophoresis and by NH,-terminal analysis (glutamic acid) and has a molecular weight of 10,150 by sedimentation equilibrium ultracentrifugation. When analyzed by SDS-gel electrophoresis, it had an aberrantly low mobility and stained faintly reddish-purple with Coomassie blue, which may be related to its unusual amino acid composition (Table I). The mol fraction of MDC bound to CNAlO* was estimated spectro- photometrically to be -0.55, with the experimentally deter- mined E;‘$,,, (280 nm) of 19.2.

Successive Enzymatic Digestions of MDC-labeled Fibrin - MDC-labeled fibrin was digested by ancrod for 12 to 16 h and during this time fluorescent peptides were released into the supernatant. The residual particulate fibrin, now termed AFbn*, was also fluorescent. The supernatant phase was removed by centrifugation and, as indicated in a preceding section, it was found to contain the fluorescent peptides A19* through A12*. The mixture of these fluorescent peptides was then incubated with plasmin for 3 h. By this time, SDS-12.5% gel analyses indicated that the peptides had been digested to a single, brightly fluorescent band of M, - 10,000. This peptide, designated PAlO*, stained well with Coomassie and its mobility was unaltered by reduction or longer incubations with plasmin (12 h). It was purified by gel filtration on Sephadex G-50 and then on G-25; following the last chromatog- raphy step the peptide was homogeneous by SDS-20% gel electrophoresis. Its molar fraction of incorporated MDC was approximately 1.0 as estimated spectrophotometrically, using its experimentally determined E:“&,, 280 nm) of 21.8. The solubility of PAlO* in aqueous buffers was distinctly less than for most of the other isolated plasmic peptides.

The particulate fibrin, AFbn*, which was separated by centrifugation at the termination of MDC-labeled fibrin diges- tion by ancrod, was suspended in buffer and subjected to digestion by plasmin as described under “Materials and Meth- ods.” At 90 min SDS-7.5% gel analysis of the digestion mixture showed that two fluorescent bands of apparent molecular weight 17,000 and 8,000 had been released into the superna- tant by plasmin; neither of these peptides stained well with Coomassie blue. At this stage of digestion, the supernatant also contained MDC-labeled monomeric and dimeric Frag- ment D species. Additional SDS-gel analyses of the particulate fibrin that remained at 90 min showed the p- and y-subunit chains to be essentially intact. However, the NH,-terminal (Y- chain derivative, A38*N, was partially degraded and two nonfluorescent peptides of molecular weight 24,000 (PA24N) and 23,000 were observed after reduction. These latter pep- tides must be analogous to the NH,-terminal o-chain deriva- tives contained in the Fragment X species ordinarily gener- ated during the early phases of fibrinogen digestion by plas- min.

The MDC-labeled peptides released into the supernatant after 90 min of plasmin digestion of AFbn* were isolated by gel filtration and ion exchange chromatography. In addition to Fragment D and trace amounts of P21*, two fluorescent peaks of interest were eluted from Sephadex G-50 at 1.5 V, and 2.8 V,. After rechromatography of the first pool on Sephadex G-25, SDS-20% gel analysis revealed a single band with an electrophoretic mobility and staining properties iden- tical with CNAlO*. Thin layer chromatography of the second pool revealed several nonfluorescent peptides and two MDC- labeled peptides of approximately equal fluorescent intensity. These two latter peptides were purified by Sephadex G-10 gel

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

2190 Human Fibrin a-Chain Cross-link Acceptor Sites

TABLE I

Amino acid compositions of a-chain peptides

Presented as residues/molecule, calculated with the molecular weights indicated in the bottom line. The peptides were derived from MDC-labeled fibrin, except CN29 from fibrinogen and CN29* from isolated, MDC-labeled a-chain. Values were determined from quintuplicate 22-h and triplicate 72-h hydrolyses, unless noted otherwise.

Residue CNAlO* PAlO* P21*” CN2gb CN29* P35’ P42* A19’ A38’N’

S-Cm-cysteine Aspartic acid Threonined Serine” Glutamic acid Proline Glycine Alanine Half-cystinee Valind Methionine” Isoleucind Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Tryptophan’

Total residues Molecular weight”

9.2 9.9 18.9 12.5 7.0 22.2 24.2 23.3 48.6

1.4 8.0 17.2 10.6 9.0 19.3 27.2 20.2 47.9

2.3 3.2 5.1 0.2 0 0.2 0.2 3.2 4.1 0 0 0.9 1.1 0 1.0 1.1 0 2.1 1.1 0 2.1 0 3.0 2.7 0 2.7 3.1 1.2 0 3.3 5.8 5.8 11.8 3.2 4.1 7.0

106 99 218 10,150 10,000 21.500

23.3 34.4 52.3 24.1 22.2 50.0

7.1 2.4

12.5 -0.5”

2.8 4.6 2.8 3.0 5.0

12.3 15.4

(275)’ 29.000

’ Additional duplicate 48-h hydrolyses performed. ’ Additional duplicate 48-h and duplicate 120-h hydrolyses per-

formed. c Determined from 22-h hydrolyses only; destruction of threonine

and serine was assumed to be 4 and 8%, respectively; valine and isoleucine values were increased by 10%.

” Determined from linear extrapolation to O-h hydrolysis, unless noted otherwise.

e Determined as cysteic acid (23). ‘Values from 72-h hydrolyses, unless noted otherwise. g Determined as methionine sulfone (23). ” Present as homoserine and homoserine lactone. i Determined by the method of Liu and Chang (24). ’ Tryptophan not included. ’ Estimated from SDS-gel electrophoresis, except for CNAlO*

which was determined by ultracentrifugation.

filtration followed by CM-Sephadex C-25 chromatography and fected by reduction. The other band had a molecular weight of were designated as PAIVa* and PAIVb*. The amino acid 27,000, was less fluorescent, stained very well with Coomassie compositions and NH,-terminal analyses of these two MDC- blue, and had a significantly lower electrophoretic mobility labeled peptides (Table Is) permitted their placement in the after reduction than before; hence at least one disulfide loop is NH,-terminal sequence of P23*. present in this peptide.

Amino Acid Composition -Most of the a-chain peptides which were purified from digestions of MDC-labeled fibrin were subjected to thorough amino acid analyses as shown in Table I. The peptides CNAlO*, PAlO*, and P21* all have extremely low contents of hydrophobic residues (2 to 6%) and are unusually rich in serine and glycine (45% of the residues); none of these three peptides contain cysteine. The composi- tions of CN29 from fibrinogen and CN29* from MDC-labeled

a-chain were essentially identical and also had low hydropho- bic contents (8%). Two half-cystine residues are present in CN29*, A19*, and P35*, while four were detected in P42*. Threonine, serine, and glycine accounted for approximately 50% of all the residues in CN29*, P35*, and P42*, whereas they comprised only 28% of the large NH,-terminal peptide, A38*N.

Enzymatic Digestions of Certain Isolated a-Chain Pep- tides - Selected peptides purified from digestions of the MDC- labeled a-chain by ancrod, plasmin, or cyanogen bromide were subjected to additional cleavages by ancrod, plasmin, trypsin, or chymotrypsin. CNBr cleavage of P42* produced one major fluorescent fragment which migrated on SDS gels at the position of CN29*. When F42* was incubated with ancrod, two major fluorescent peptides were readily generated; the band of

greatest fluorescent intensity had a molecular weight of -17,000, stained poorly with Coomassie blue, and was unaf-

Cleavage of CN29* by ancrod also generated two major fluorescent peptides as detected by SDS-lo% gel electrophore- sis. One of these had a molecular weight of -19,000 and contained a disulfide loop as evidenced by its slightly de- creased electrophoretic mobility after reduction; the other was less intensely fluorescent, stained faintly reddish-purple with Coomassie and had a molecular weight of -17,000. A band identical with this latter one, and a second brightly fluores- cent band of M, - 11,000 which stained well with Coomassie blue were observed when the products from the digestion of P21* by ancrod were analyzed (Fig. 6). The apparent discrep- ancy in conservation of mass was attributed to the aberrant SDS gel electrophoretic mobility of the M, = “17,000” peptide common to both P21* and CN29*. By comparative SDS gel analyses this M, = “17,000” band is essentially the same as CNAlO*, which has a molecular weight of 10,150 by sedimen- tation equilibrium ultracentrifugation.

Treatment of P21* with CNBr produced no detectable change in its molecular weight nor fragmentation of its fluorescence as demonstrated by SDS gel electrophoresis. Mild plasmin digestion of CN29* produced only one fluorescent peptide which migrated at the position of P21* on SDS-12.5% gels. Continued incubation of CN29* or P21* with plasmin

produced two bands of approximately equal fluorescence with molecular weights of 17,000 and 11,000. The higher molecular

23.4 28.5 31.9 35.1 40.6 47.5 52.4 61.2 70.0 25.3 31.4 37.7 21.7 25.5 28.7 52.6 61.5 67.4

6.4 9.9 12.5 2.1 2.4 3.9

11.8 12.1 14.5 -0.6” 4.3 5.4

2.5 3.5 5.2 5.1 8.6 10.6 3.0 3.1 4.0 3.0 5.9 11.2 4.8 6.5 8.9

12.3 12.9 17.8 15.8 17.7 22.5

6.8 8.0 7.6 285 345 408

29.000 35.000 42,000

2.5 6.4 14.3 46.1 22.9 18.5 29.5 41.7 19.3 40.3

9.5 20.2 26.0 36.8

5.7 12.5

11.5 1.1 2.6 5.6 1.8 4.5 5.1

10.4 8.3

(181Y 19,000

15.0 6.1

15.0 23.2

1.2 6.6 4.6

13.3 22.9

6.1 342

37,500

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Fibrin a-Chain Cross-link Acceptor Sites 2191

weight band did not stain well with Coomassie blue and probably corresponds to the CNAlO* region of the a-chain. When the tryptic digestion products of P21* were compared to those of CNAlO* by chromatography on Sephadex G-25, an additional fluorescent peak was observed in the elution pro- file. The amino acid compositions and NH,-terminal analyses of a MDC-labeled peptide isolated from the tryptic digest of P21* and another from the tryptic digest of CNAlO* are given

PF PF

Mr

Hours of Digestion FIG. 6. SDS, 6 M urea, 10% polyacrylamide gel analyses of an-

trod-digested P21*. Ancrod was dialyzed into ammonium bicarbon- ate buffer and to this solution (70 units/ml) P21* was added to give 5 mg/ml and was incubated at 37”. At 12 h more ancrod was added to give a final E/S ratio of 30 units/mg of P21*. After 2 and 24 h of digestion the gel patterns were essentially the same as that shown for 12 h. M, designates the apparent molecular weight of the peptides. P, protein; F, fluorescence.

in Table Is. From those data and the amino acid composition of PAlO*, it appears that there is a segment in the COOH- terminal half of CNAlO* which contains a cross-link acceptor site. The fluorescent products from the tryptic digestion of P21* and CN29* appeared to be the same on SDS-lo% gels; in each case two equally fluorescent bands with apparent molec- ular weights of 10,000 and 12,000 were observed. Finally it should be noted that the SDS electrophoretic mobilities of the tryptic digestion products of nonlabeled CN29 or P21 were identical with those of MDC-labeled CN29 or P21.

NH,terminal Sequence Analyses-Table II shows the re- sults of NH,-terminal sequence analyses for several of the peptides derived from MDC-labeled fibrin by enzymatic or chemical cleavage. The purified preparations of P21* and CN29* had identical NH,-terminal sequences after the termi- nal methionyl residue in P21*. This is in agreement with and extends previously published sequence data on the analogous peptides from fibrinogen (30, 31). There were no labeled acceptor sites in the first 15 cycles of CN29* nor in the first 24 cycles of P21* which had been purified on Ultrogel AcA44 followed by Sephadex G-50 gel filtration. In the sequence analysis of the P21* preparation obtained by only Ultrogel AcA44 chromatography, definite MDC fluorescence was de- tected at cycle 18. Since serine was the predominant amino acid detected at this position in both P21* preparations, the fluorescence in this cycle was attributed to an MDC-labeled glutaminyl residue in the sequence of a minor component in the AcA44 pool. NH,-terminal sequence analysis of the larger plasmic peptide P42* and P35* revealed a labeled acceptor site at cycle ‘7. These two peptides possessed the same NH,-termi- nal regions; however, each peptide preparation contained a mixture of primarily two peptides which had overlapping sequences that were 12 residues out of phase. Assignment of the two amino acids detected at each cycle to their proper sequence would normally be very difficult since the amounts

TABLE II Partial amino-terminal sequence of isolated a-chain peptides from MDC-labeled fibrin

The yields in nmol as determined by gas chromatography of the amino acid phenylthiohydantoins are given in parentheses. The asterisk (*) denotes fluorescence. Approximately 7% of the ethyl acetate phase of each cycle was spotted and examined for fluorescence.

P3V” p42** Cycle p21* CN29f ~-~

I II I II III

1 Met (350) Glu Ser (71) Ala (43) Ser (59) Ala (39) Met (20) 2 Glu (280) Leu Gln (25) Leu (42) Gln (24) Leu/Ile (48) Glu (17) 3 Leu (260) Glu Leu (84) Pro/Thr (25) Leu (54) Pro/Thr (17) (Leu) 4 Glu (280) Gln (23) ASP (15) Gln (17) ASP (18) Glu (26) 5 Arg Pro LYS (19) Met (26) LYS (25) Met (26) 6 Pro (140) GUY Val (55) Pro (25) Val (57) Pro/Thr (25) 7 Gly (140) GUY Pro/Thr (47) Gin* Pro/Thr (33) Gin* a Gly (120) Pro/Thr (32) Met (19) Pro/Thr (31) Met cm GUY (17)

9 Asn (120) Glu Glu (25) Glu @co 10 Glu (140) Ile ‘b (14) Tw (15) Met (1’3) 11 Ile (190) LYS (8) LYS (10) 12 Thr (114) Ala (2’3 Ala (32) Leu/Ile (26) 13 Arg Leu (31) Leu (32) Glu (14) 14 Gly (112) Pro/Thr (26) 15 Gly (104) Asp (8) Pro/Thr (18) 16 Ser (230) Met (20) 17 Thr (69) ProlThr (22) 18 Ser (230) 19 Tv (99) 20 Gly (105) 21 Thr (72) 22 GUY (73)

’ Discrimination between threonine uersus proline and between isoleucine ucrsus leucine was not obtained for some cycles.

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

2192 Human Fibrin a-Chain Cross-link Acceptor Sites

were nearly equal and since quantitation by gas chromatog- raphy of certain amino acid phenylthiohydantoins is not very reliable (32). However, at the time of this endeavor, the primary structure of a 38residue plasmic peptide from the o(- chain of [“Clglycine ethyl ester-labeled fibrin was published (9). The partial sequence of this peptide, which was termed the “midsection piece,” could be reconstructed from our data and was therefore used as a guide for positioning the two sequences of P35* and P42* (Table II). The yields obtained in each cycle for sequence I are not consistently greater or less than those of sequence II because of the uncertainties in quantitation. With the use of the yields of those residues which could be reliably quantitated, the relative amounts of the different NH, termini generated by plasmin were esti- mated: P42* contained 40% terminal serine, 40% terminal alanine, and 20% terminal methionine; and P35* contained 60% terminal serine and 30% terminal alanine. Evidence for the lesser sequence Met-Glu-Leu-Glu- in P42* established that this peptide contains the smaller plasmic peptide, P21*, in its NH,-terminal half. Furthermore, it is likely that the peptide P23* contains the labeled glutaminyl residue detected at cycle 7 because its NH, terminus was serine and because P21* plus a small MDC-labeled peptide were generated from P23* by

plasmin. It should be noted that this small MDC-labeled pep- tide accounted for only a fraction of the fluorescence in the o(- chain; the majority of the label is contained in P21*, contrary to the finding by Takagi and Doolittle (9) of approximately equal incorporation of [14C]glycine ethyl ester into these two peptides.

Relatively intense fluorescence was detected at cycle 18 in the sequence analysis of the smallest ancrodic supernatant peptide (A12*), but considerable overlap between cycles and possible contamination from a nonfluorescent ancrodic peptide prohibited reliable identification of the residues in the specific site-containing sequence. Further elucidation of the primary structure around the MDC-labeled cross-link acceptor sites within P21* is in the preliminary stages and will be the subject of a separate report.

DISCUSSION

The results of this investigation indicate that there are three distinct regions in the e-chain of fibrin which incorpo- rate the substitute donor MDC. This expands our earlier report (7) and that of Takagi and Doolittle (9) in which only two labeled o-chain acceptor site regions were identified after cleavage by either plasmin or CNBr. The first of these two regions corresponded to either P21* or CN29*, referred to by other authors as either Fragment A or HiB-DSK, respectively (9, 17). The second region was either the nonapeptide (PAIVb”) or the tripeptide (Pro-Gin*-Met) from the midsection piece (9). The presence of at least two acceptor sites within the first region was revealed specifically by ancrod cleavage of the o-chain of MDC-labeled fibrin (Figs. 1 and 2); by comparison, this digestion pattern paralleled that reported for the cleavage of fibrinogen by ancrod (33, 34). This venom enzyme cleaved P21* to give a fluorescent NH,-terminal peptide of molecular weight 10,000 and another fluorescent peptide of molecular weight 11,000. As anticipated from these results with P21*, the CN29* fragment also gave rise to two fluorescent peptides when incubated with ancrod: an NH,-terminal one of molecu- lar weight 10,000 and a disulfide-containing one of molecular weight 19,000. The spectrophotometric, fluorometric, and pre- liminary sequence analyses of these two fluorescent regions (represented by CNAlO* and PAlO*) imply that they each

contain only one specifically labeled cross-link acceptor site, as is the case with the third region (PAIVb*). These data do not exclude the possibility that several (n) glutamine residues per region may be fractionally (-n-l) labeled; to do so would require completion of the primary structures of CNAlO* and PAlO*. However, previous studies with fibrinogen showed little evidence for such nonspecific incorporation of MDC into the CN29 region (8) and therefore lend support to our interpre- tation that each of the three fluorescent regions has only one MDC-labeled site.

All three sites are present in the large plasmic peptides which are released rapidly from the a-chain of MDC-labeled fibrin (e.g. P45*, P42*, and P35*); moreover, they are also contained within a smaller plasmic derivative of these, namely P23*, which is the immediate precursor of P21*. No labeled sites appear in the M, = 25,000 NH,-terminal segment of the o-chain (Fig. 3). The spatial relationship between the peptides P42*, P35*, P21*, and CN29* was established from their overlapping NH,-terminal sequences (Table II) and indicates that the o-chain extends -20,000 daltons beyond the COOH terminus of P21*. Furthermore, there appear to be no cross-link acceptor sites in this region, since the CNBr and plasmic digestion products derived from it had not incorpo- rated MDC. Also the plasmic peptide of molecular weight 3,000 from the very COOH-terminal end of the a-chain was recently shown to contain no glutaminyl residues (35). This peptide appears to be cleaved from a large fraction of a-chains in viuo and must account for some of the observed heteroge- neity in molecular weight (36, 37). This seems pertinent to our observation that even under the mildest conditions of ancrod digestion of MDC-labeled fibrin, the two largest supernatant peptides, A34* and A31*, are generated from the a-chain simultaneously. The difference of 3,000 in molecular weight between these, and similarly between P45* and P42* is therefore tentatively attributed to the COOH-terminal plasmic peptide.

The position in the a-chain of the plasmic supernatant peptides, of CN29*, and ultimately the specific locations of the acceptor sites were determined as follows. The NH,-terminal o-chain derivative A3&3*N resulting from ancrod digestion was cleaved by CNBr to produce the fluorescent peptide that overlaps the first cross-link acceptor site in CN29*. This isolated peptide, CNAlO*, was found to have the same NH, terminus as CN29* and, since it contained no methionine, homoserine, or homoserine lactone (Table I), its COOH ter- minus should be identical with that of A38*N. Using the number of amino acid residues found in A38*N and CNAlO* (Table I) and the fact that ancrod cleaves the Aol-chain after residue 23 (38), the NH, terminus of CN29* (and of CNAlO*) was computed to be residue 259 of the Aa-chain. From this reference point, the locations of the plasmic peptides can be determined and the approximate locations of the three cross- link acceptor sites can be placed at residue positions 255, 310, and 385 from the NH, terminus of the Aa-chain. The primary limitation with these calculations is the uncertainty in the molecular weight of A38*N, which was determined by SDS gel electrophoresis. An independent estimation that places the most NH,-terminal acceptor site at residue 238 in the Ao- chain can be calculated with the partial amino acid sequence of the o-chain remnant in Fragment D and the preliminary composition of the o-chain region linking this sequence with the midsection piece, both of which have been reported re- cently (39, 40). This estimate is in good agreement with our prediction of position 255; the difference in residues corre-

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Fibrin a-Chain Cross-link Acceptor Sites 2193

N

FPA -

PA24N

P25N

* * * n n

* I * n n

n n * n A

n *

*

* x n

3 * x 4

I * * n n

3 * * n n 3 * * n

sponds to only a 5% discrepancy in the molecular weight of A38*N. The spacing between the first and second MDC-la- beled site is based upon (a) fluorescence detected in a minor sequence of a P21* preparation containing lower molecular weight derivatives and (b) preliminary sequence analyses of two overlapping tryptic peptides from CNAlO* and P21* (Table Is). The location of the third site was derived from (a) the number of residues in CNAlO* and (b) the detection of fluorescence in the preliminary sequence analysis of A12*, which had the same NH, terminus (glycine) as A19* and PAlO*.

As shown schematically in Fig. 7, the above findings localize the cross-link acceptor sites to the middle 20% of the Aa-chain in contrast to the near COOH-terminal location of the acceptor site in the y-chain (41). The rationale for most of the peptide alignments shown in this model has been discussed; however, validation of certain features is obtained from examination of the amino acid compositions in Table I. For example, the proposed location of A19* predicts that its amino acid compo- sition should be nearly identical with the difference in compo- sition between CN29* and CNAlO*. The agreement, in fact, is very good, and most amino acid values differ by less than 6%. However, the presence of a methionyl residue and two extra residues each of leucine and phenylalanine in A19* suggests that it extends slightly beyond the COOH terminus of CN29*. In the case of PAlO*, which was derived from A19* and A12* by plasmic cleavage, its position is supported by the equiva- lence of its amino acid composition with the difference in composition between P21* and CNAlO*. Again the agreement is good and indicates that PAlO* must be shorter by only a few residues (lysine, tyrosine, threonine, and glutamic acid). Our amino acid compositions are in very good agreement with published values for peptides corresponding to P21 (31) and CN29 (30). It should be noted that our analyses of CN29*,

either with or without prior performic acid oxidation, showed no evidence for incomplete CNBr cleavage as was observed in one report (42).

When used in conjunction with Fig. 7, the distribution of amino acids within certain peptides can also be calculated from Table I. For example, both halves of the strongly hydrophilic peptide, P21*, have approximately equal numbers of polar residues, while 11 of the 13 hydrophobic residues are localized in its COOH-terminal half. The 7,500-dalton COOH- terminal segment of CN29* is devoid of aromatic amino acids, derives half of its residues from lysine, threonine, and valine, and contains an intrachain disulfide loop. The latter point was

Au-chain A38* N A34*

A31* A27’

A19* PAlO*

CNAlO* CN29* P21X P23*

MSP p45* P42* P35”

FIG. 7. Model for the localization of Aa-chain cross-link acceptor sites and for the relative alignment of the Aol- chain peptides which were charac- terized in this paper. The asterishs denote the locations of MDC-labeled acceptor sites and the arcs represent internal disulfide bonds. As indicated in the text,2 most peptide designations denote the agent(s) of cleavage as well as the approximate molecular weight of the peptide. The model is drawn to scale with respect to molecular weights. The 38-residue midsection piece was previously characterized (9) and is designated MSP; the 16-residue fibrino- peptide A is labeled FPA.

deduced from the slight, but measurable decrease in mobility of reduced CN29* compared to nonreduced CN29* when ana- lyzed by SDS-gel electrophoresis. A much larger decrease in electrophoretic mobility (corresponding to a molecular weight shift of -5,000) after reduction was observed in SDS-gel analyses of A27* and P42*. The magnitude of this difference suggests that the two additional half-cystine residues which were detected in P42* (Table I) are also joined in an internal disulfide bond that is probably also present in A27*. Since P35* contained only one disulfide, the location of this second loop should be within the 7,000-dalton COOH-terminal seg- ment of P42*, as shown in Fig. 7. To our knowledge this is the first indication that the COOH-terminal half of the o-chain may contain another cystine in addition to that of CN29.

The physiological significance of the three potential acceptor

sites reported here could be assessed most directly by compar- ing these findings with the structural analyses of the o-chain cleavage products obtained from cross-linked fibrin formed in vivo. Although such studies have yet to be reported, some relevant deductions can be made from existing data. Pisano and co-workers (43) have detected by chemical and enzymic methods four to six cross-links per molecule of human fibrin,

depending on the source. When whole blood was used the values averaged 4, while plasma and purified fibrinogen produced fibrin with an average of 5.4 cross-links (43). More recent data from our laboratory indicate that this value is closer to 4 when fibrin formed from whole plasma is analyzed (44). Since two cross-links per fibrin molecule are contributed by y-chains (411, each a-chain should contribute from one to a maximum of two cross-links, depending on the clotting condi- tions. This indicates that possibly two of the potential acceptor sites are not required to form cross-linked fibrin. Conse- quently, they might represent either nonspecifically labeled glutamines or acceptor sites which in vivo participate in cross- links with another protein such as cold-insoluble globulin. This latter possibility is supported strongly by the recent reports of Mosher (45, 46). The potential acceptor site in the midsection piece may have this function since it was not involved in cross-linking of fibrin prepared from highly puri- fied fibrinogen under conditions which caused the other two a-chain acceptor sites to become completely cross-linked (10). Complete a-polymer formation was also observed in fibrin prepared from whole plasma under conditions which gave only one cross-link per o-chain in the CN29 region.i The specific

5 L. J. Fretto and P. A. McKee, manuscript in preparation.

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

2194 Human Fibrin a-Chain Cross-link Acceptor Sites

acceptor site involved in this cross-link cannot be reliably predicted from this article because the relative cross-linking reactivities of the two acceptor sites in CN29 appear approxi- mately equal when judged by MDC incorporation under the saturating conditions used in these studies. However, previous

data of Lorand and co-workers (8) can be interpreted to compare the specificity of labeling in the midsection piece with that in CN29. When the CNBr fragments from fibrinogen or fibrin which had both been partially labeled with the substi- tute donor [‘*C]glycine ethyl ester were gel filtered, the only radioactive peak in the chromatograph of the fibrin digest which was significantly larger than the equivalent peak from the fibrinogen digest corresponded to CN29* and it had -5 times more counts than the labeled midsection piece fragment (8). Apparently the release of fibrinopeptide A from fibrinogen increases the accessibility or reactivity of at least one of the

cross-link acceptor sites in CN29, but does not affect pseudo- donor incorporation into the midsection piece. Taken together with the above findings, this suggests that the physiologically important cross-link acceptor site(s) for ol-polymerization is located near residue 310 or 385 of the Am-chain.

1.

2. 3.

4.

5.

6.

7.

8.

9.

10.

11.

12. 13. 14.

REFERENCES

McKee, P. A., Schwartz, M. L., Pizzo, S. V., and Hill, R. L. (1972) Ann. N. Y. Acad. Sci. 202, 127-148

Doolittle, R. F. (1973) Adu. Protein Chem. 27, l-109 Finlayson, J. S., Mosesson, M. W., Bronzert, T. J., and Pisano,

J. J. (1972) J. Biol. Chem. 247, 5220-5222 Pizzo, S. V., Schwartz, M. L., Hill, R. L., and McKee, P. A.

(1973) J. Biol. Chem. 248, 4574-4583 Finlayson, J. S., and Mosesson, M. W. (1973) Thromb. Res. 2,

467-478 Pizzo, S. V., Schwartz, M. L., Hill, R. L., and McKee, P. A.

(1972) J. BioZ. Chem. 247, 636-645 Fereuson. E. W.. Fretto. L. J.. and McKee. P. A. (1975) J. Biol.

&em. 250, 7210-7218’ Lorand. L.. Chenoweth. D., and Gray, A. (1972) Ann. N. Y

Acad. Sci. 202, 155-171 Takagi, T., and Doolittle, R. F. (1975) Biochemistry 14, 5149-

5156 Doolittle, R. F., Cassman, K. G., Cottrell, B. A., and Friezner,

S. J. (1977) Biochemistry 16, 1715-1719 Walther, P. J., Steinman, H. M., Hill, R. L., and McKee, P. A.

(1974) J. Biol. Chem. 249, 1173-1181 Laemmli, U. K. (1970) Nature 227, 680-685 Laemmli, U. K., and Favre, M. (1973) J. Mol. Biol. 80, 575-599 Panyim, S., and Chalkley, R. (1969) Arch. Biochem. Biophys.

15. 16. 17.

18. 19. 20.

21. Moore, S., and Stein, W. H. (1963) Methods Enzymol. 6, 819-831 22. Hubbard. R. W. (1965) Biochem. Biophrs. Res. Commun. 19,

679-685 23. Hirs, C. H. W. (1967) Methods Enzymol. 11, 59-62 24. Liu, T.-Y., and Chang, Y. H. (1971) J. BioZ. Chem. 246, 2842-

25. Hartlev. B. S. (1970) Biochem. J. 119. 805-822 26. Gray, k’. R. (1972) Methods Enzymol. 25, 121-138 27. Woods, K. R., and Wang, K.-T. (1967) Biochim. Biophys. Acta

133,369-370 28.

29 30 31

Hermodson, M. A., Ericsson, L. H., Titani, K., Neurath, H., and Walsh, K. A. (1972) Biochemistry 11, 449334502

Edman, P., and Begg, G. (1967) Eur. J. Biochem. 1, 80-91 Hessel, B. (1975) Thromb. Res. 7, 75-87 Harfenist, E. J., and Canfield, R. E. (1975) Biochemistry 14,

4110-4117 32

33

34 35

36

37.

Evans, H. J., Steinman, H. M., and Hill, R. L. (1974) J. BioZ. Chem. 249, 7315-7325

Pizzo, S. V., Schwartz, M. L., Hill, R. L., and McKee, P. A. (1972) J. CZin. Inuest. 51, 2841-2850

Edgar, W., and Prentice, C. R. M. (1973) Thromb. Res. 2, 85-96 Cottrell, B. A., and Doolittle, R. F. (1976) Biochem. Biophys.

Res. Commun. 71, 754-761 Mills, D., and Karpatkin, S. (1970) Biochem. Biophys. Res.

Commun. 40, 206-211 Semeraro. N.. Collen. D.. and Verstraete. M. (1977) Biochim.

38. 39.

40.

41.

42. 43.

44.

45. 46.

130, 337-346 Davis, B. J. (1964) Ann. N. Y. Acad. Sci. 121, 404-427 Smyth, D. G. (1967) Methods Enzymol. 11, 214-231 Blomback, B., Blomback, M., Finkbeiner, W., Holmgren, A.,

Kowalska-Loth, B., and Olovson, G. (1974) Thromb. Res. 4, 55-75

Him, C. H. W. (1967) Methods Enzymol. 11, 199-203 Yphantis, D. A. (1964) Biochemistry 3, 297-317 Cohn, E. J., and Edsall, J. T. (1943) in Proteins, Amino Acids,

and Peptides, pp. 155-161, Reinhold Publishing Corp., New York

2848

Biophys. A& 492, 204-214 Hessel. B.. and Blomback, M. (1971) FEBS Lett. 18, 318-320 Doolittle, R. F., Cassman,‘K. G., Cottrell, B. A., Friezner, S. J.,

and Takaei. T. (1977) Biochemistry 16, 1710-1715 Doolittle, R-F., Cassman, K. G., Cottrell, B. A., Friezner, S. J.,

Hucko, J. T., and Takapi, T. (1977) Biochemistry 16, 1703-1709 Chen, R., and Doolittle,-R. F. (1970) Proc. N&Z. Acad. Sci.

U. S. A. 66, 472-479 Gollwitzer. R.. and Timnl. R. (1975) FEBS Lett. 58. 269-272 Pisano, J. ‘J., Bronzert,&T: J., Peyton, M. P., and Finlayson, J.

S. (1972)Ann. N. Y. Acad. Ski. 202, 98-113 Ball, A. P., and McKee, P. A. (1977) J. Lab. CZin. Med. 89, 751-

763 Mosher, D. F. (1975) J. BioZ. Chem. 250, 6614-6621 Mosher, D. F. (1976) J. Biol. Chem. 251, 1639-1645

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

;j ii : : j : ; j

ELUTION VOLUME ‘(ml)

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from

L J Fretto, E W Ferguson, H M Steinman and P A McKeeLocalization of the alpha-chain cross-link acceptor sites of human fibrin.

1978, 253:2184-2195.J. Biol. Chem. 

  http://www.jbc.org/content/253/7/2184Access 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/253/7/2184.full.html#ref-list-1

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

by guest on Novem

ber 11, 2018http://w

ww

.jbc.org/D

ownloaded from