Vol. 171, No. 2, 1990 September

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Pages729-737

14, 1990

PRESENCE AND FUNCTION OF CHONDROllTN-4-SULFATE ON RECOMBINANT HUMAN SOLUBLE TEROMBOMODDLIN Katsuhlko Nawa,* Katsu-lchi Sakano, Hiroyukl Fu jlwara, Yoshlnarl Sato, Norlfuml Suglyama, Toshlyuki Teruuchl, Masahlro Iwamoto, and Yasumasa Marumoto Research Institute, Dallchl Pharmaceutical 16-13. Kltakasal 1-chome. Edogawa-ku, Tokyo Received

July

30,

Co., Ltd. 134, Japan

1990 -

We constructed a human soluble thrombomodulln (sTM) expression vector using the RSV promoter. Recombinant sTM (rsTM) was expressed in CHO cells and was recovered from culture medium by ion exchange chromatography. Two active fractions, designated as rsTMa (low salt elutlon) and rsTMB (high salt elutlon). were detected and further purified by lmmunoafflnity chromatography. Purified rsTMB contained bound chondroltin-4-sulfate as judged by HPLC detection of the chondroltlnase ABC and AC I digestion product, 2-acetamldo-2-deoxy-3-O-(8 -D-gluco-4-enepyranosyluronlc acid)-4-0sulfo-D-galactose. The apparent Kd values for thrombln of a and B were 7.4 and 1.4 nM respectively. RsTMB was more effective at inhibition of thrombln clotting activity and had antlthrombln III-dependent anticoagulant activity which was not possessed by rsTMa . Both anticoagulant activities were lost after chondroltinase treatment of rsTMB . 0 1990Acxdemic Press,Inc.

Thrombomodulln binds

thrombln

tion

of protein

lant

by

human

forming

domain,

glycosylatlon domains.

site-rich

*

coagulation

revealed

amino-terminal

To whom correspondence

on the

a 1:l stoichiometrlc

C (PC) (1,2). Activated

inactivating TM (6-8)

(TM), a glycoproteln

that

six

complex

PC (APC) acts

factors TM consists

epldermal region,

endothelial

growth

and

the

for

(3-5).

factor(EGF)-like

activaantlcoagu-

The cDNA of

5 structural

transmembrane

surface,

accelerated

as a potent

Va and VIIIa of

ceil

domains:

the

structures, and

O-

cytoplasmlc

should be addressed.

Abbreviations used in the text are: A Di-OS, 2-acetamido-2-deoxy-3-0-(fi -D--gluco-4-enepyranosyluronlc acid)-D-galactose; A Dl-4S, 2-acetamldo-2-deoxy3-O-( /3 -D-gluco-4-enepyranosyluronlc acid)-4-O-sulfo-D-galactose: A Dl-GS, 2acetamldo-2-deoxy-3-O-@ -D-gluco-4-enepyranosyluronlc acid)-6-O-sulfo-Dgalactose; A Di-diSD, 2-acetamldo-2-deoxy-3-O-(2-0-sulfo-B -D-gluco-4-enepyranosyluronlc acid)-6-0-sulfo-D-galactose; A Dl-dlSE, 2-acetamldo-2-deoxy-3-0(B -D-gluco-4-enepyranosyluronlc acid)-4.6-bls-O-sulfo-D-galactose; A Di-trlS, 2-acetamido-2-deoxy-3-O-(2-0-sulfo-/3 -D-gluco-4-enepyranosyluronlc acld)4,6-bls-0-sulfo-D-galactose. 0006-291X/90

129

$1.50

Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

Vol.

AND BIOPHYSICAL

TM functions to a) accelerate thrombin’s substrate specificity

alter

c) accelerate (10,ll).

inactivation

An acidic

15) in which that

BIOCHEMICAL

171, No. 2, 1990

was associated

an analogous

form

form contained

thrombin of rabbit

human TM although

it

has not

has been

detection in the human molecule In this study we expressed of

the

three

by antithrombin lung

ABC-sensitive

The presence

proposed

extracellular

that

clearly

vitronectin

recombinant domains

(12moiety

and structure

been demonstrated

(15). and purified

III (AT III)

TM were reported

a chondroitinase

with b) and c) activities.

glycosaminoglycan

(rsTM) consisting

thrombin-catalyzed PC activation, b) to reduce fibrinogen cleavage (9) and

of the bound

and non-acidic

the acidic

RESEARCH COMMUNICATIONS

yet

of for

may mask its human soluble

of native

TM.

TM Two

species were found in the cell culture medium, an acidic (rsTMj?) and a nonacidic form (rsTMa), both of which functioned as cofactors for PC activation. The rsTMB and antithrombin rsTMa

was found

to possess

III (AT III)-dependent

chondroitin-4-sulfate activities

differed

and its

direct

from the non-acidic

form. MATERIALSANDMEl'DODS

Materials. The materials and their suppliers were as follows: enzymes for constructing the expression vector, Takara Shuzo, Kyoto, Nippon Gene, Tokyo, Toyobo, Osaka and International Biotechnologies Inc., New Haven; bovine thrombin (2500 NIH units/m&, heparin and bovine serum albumin, Sigma Chemical, St. Louis: human protein C, American Diagnostica Inc., Greenwich; S-2366, Kabi-Vitrum, Stockholm: bovine fibrinogen, Daiichi Pure Chemicals, Tokyo; antithrombin III, Green Cross Inc., Osaka; chondroitinase AC I (chondroitin AC lyase, EC 4.2.2.5) derived from Flavobacterium heparium, A Di-OS, A Di-4S, A Di-6S, A Di-diSD, ADi-diSE and ADi-triS, Seikagaku Kogyo, Tokyo. A DNA fragment of the human TM gene. which was Expression of rsTM. cloned as an XhoI-NcoI fragment from a gene library (a gift from Dr. D.V. Goeddel, Genentech Inc., South San Francisco), was inserted into pUC119 and the 5’- and 3’-flanking regions were removed. Plasmid p7TM19 was constructed with the TM gene fragment containing 30bp of the 5’-flanking region, the SV40 terminator and pUC119. It was cleaved with NruI, treated with exonuclease BAL 31 (slow) followed by digestion with XbaI, and filled-in. After ligation, plasmid pTMsO7 was obtained which contained the DNA fragment encoding for soluble TM (Ala’ - Ala4’l). A fragment of the Neo’ gene with the SV40 early promoter and terminator was inserted downstream of the SV40 terminator on pTMsO7 for a selectable marker. The RSV-LTR was placed at a Hind111 site as a promoter to generate the expression plasmid pRS7TMneo (Fig. 1). Plasmid pRS7TM-neo was introduced into CHO-Kl cells using the calcium phosphate-DNA precipitation procedure. The cells were subcultured in G418 selective medium. The amount of rsTM secreted into the medium was measured by a PC activation assay (see below). The cell line CHO-RS7TM No.29 was established by the limiting dilution method and produced rsTM with high efficiency. This cell line was cultured in 36 ml of GIT medium (Wake Pure Chemical Industries) in 150 cm2 cuture bottles at 37 ‘C. 5% C02/air. After confluency culture medium was removed every day, replaced with fresh medium, and the cultures were continued for one week. from

Enzyme linked immunosorbent assay (ELISA). Hybridomas were obtained Balb/c mice immunized with purified human placenta TM (16). Hybridoma 730

Vol.

171,

No.

2,

1990

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

A

Fig. 1.

A. Restriction map of human TM gene and structural domains of TM. S: Signal DeDtide, A: Amino-terminal domain, E: EGF-like structures domain. G: 0-Glycosylation site-rich domain, T: Transmembrane domain, C: Cytoplasmic The rsTM gene lacks the sequence which encodes the transmembrane domain. and cytoplasmic domains. B. Structure of expression plasmid pRSTl’M-neo that encodes sTM. The rsTM gene is placed between the raus sarcoma virus long terminal repeat (RSVLTR) and the SV40 terminator. The neomycin resistance gene transcription unit (SV40 early promoter and terminator) is placed downstream.

were assayed for antibody production by a solid-phase enzyme immunoassay. A sandwich-type ELISA was constructed using a monoclonal antibody conjugated with horse radish peroxidase (17). Weighed rsTMa was used as standard protein for estimation of rsTM concentration. supernatants

Coagulation and PC activation assay. Fibrinogen clotting time was measured at 37’C in a coagulometer (CLOTEK, Travenol Laboratories). Bovine fibrinogen was present at a final concentration of 1.0 mg/ml. Activity of rsTM was assayed by its ability to accelerate thrombin-catalyzed activation of protein C. Protein C (0.5 BM) was incubated with rsTM at 37*C in 20 mM Tris-HCl (PH 7.5), 0.15 M NaCl, 5 mM CaCl and 0.5% bovine serum albumin in 96-well microtiter plates. Bovine thromb 1n was added to a final concentration of 10 nM. The 60 p 1 reaction mixture was incubated at 37’C for 15 min and the reaction stopped by addition of 20 p 1 of AT III (1 mg/ml) and 20 ~1 of heparin (400 u/ml). The APC formed was assayed by the addition of 100 fi 1 of 0.4 mM S-2366 and monitoring the absorbance at 405 nm using a Vmax kinetic microplate reader (Molecular Devices). Kd values were determined by analysis of the data using Lineweaver-Burk plots. The data was not corrected for the free thrombin concentration. Preparation of anti-TM-MoAb conjugated cellulofine. Anti-TM-MoAb conjugated cellulofine was prepared by coupling 54 mg of anti-TM-MoAb (1.2 mg/ml) in 0.2 M phosphate buffer, pH 8.2, to 10 ml of formyl-cellulofine (Seikagaku kogyo) for 90 min at room temperature. Schiff’s bases were reduced by addition of 50 mg solid cyanoborohydride for 3 hr and residual active formyl residues were blocked by incubation with 0.2 M Tris-HCl (pH 7.2). 0.1 M monoethanolamine, 50 mg cyanoborohydride for 2 hr. The gel was washed with 20 volumes of 20 mM Tris-HCl (pH 7.5). 0.15 M NaCl. The final concentration

of

anti-TM-MoAb

conjugated

was 5.0 mg/ml

gel.

Purification of rsTMa and & Culture medium was centrifuged to remove cells and the-as adjusted to 7.5 with 10 N NaOH. The supernatant (3.6 L) was applied to a coiumn ( 2.5 x 20 cm) of Q-Sepharose Fast Flow (Pharmacia-LKB) equilibrated at 4°C with 0.15 M NaCl, 0.02 M Tris-HCl, pH 7.5. The column was washed with 700 ml of the same buffer and eluted with a 2,000 ml I.inear gradient from 0.15 M to 1.20 M NaCl in 0.02 M Tris-HCl, pH 7.5.

Fractions

(20 ml)

were

collected

and

731

elution

was

monitored

by

absorb-

Vol.

171,

No.

2,

1990

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

ante at 280 nm. The concentration of rsTM in each fraction was measured by ELISA. RsTMa -containing fractions (Fr.l9-27) were pooled and applied to a column of anti-TM-MoAb cellulofine (1.5 x 5.7 cm) equilibrated with 0.15 M NaCl, 0.02 M Tris-HCl, pH 7.5. The column was washed with 80 ml of 0.35 M NaCl, 0.02 M Tris-HCl, pH 7.5, and eluted with 150 ml of 3.0 M NaSCN in 0.02 M Tris-HCl. pH 7.5. The fractions containing rsTMB (Fr.44-80) were pooled rsTMs were further purified by gel and purified in the same manner. The filtration on a Sephacryl S-300 column ( 2.6 x 60 cm) equilibrated with 0.15 M NaCl, 0.02 M Tris-HCl, pH 7.5. Chondroitinase & 1 digestion. RsTMB (220 or g) was incubated for 5 hr at 37“C with 0.02 U/ml chondroitinase AC I in 0.5 ml of 40 mM Tris-HCl, pH 7.5, 40 mM sodium acetate and 0.01% bovine serum albumin. The reaction mixture was applied to a 4.6 x 75 mm reverse-phase TSKgel Pheny-5PW RP column (Tosoh Co., Tokyo) equilibrated with 5% acetonitrile/l mM NH OH at The desired unsaturated disaccharides were recovered 4n the 1.0 ml/min. fraction of pass through the column and lyophilized. The chondroitinase treated r&M/3 was recovered from the column by elution with a linear acetonitrile gradient from 5 to 65% in 1 mM NH40H and was lyophilized. Analysis of unsaturated disaccharides derived from rsTMB. Analysis of unsaturated disaccharides were performed on a 4.0 x 300 mm Nucleosil equipped with a Shimadzu LC-GAD delivery system and equili5 NH2 column the column was washed brated with 0.01 M NaH2PO4 After sample injection, at 1 ml/min for 5 min with the same solution and then eluted with a 40 min linear gradient from 0.01 to 0.5 M NaH P04. Chromatography was carried out at 45’C and elution was monitored at l 32 nm. RESULTS

We expressed extracellular was cultured antigen

rsTM in CHO-KI cells

domains of native and the secreted

by ion exchange

from a vector

coding

human TM. The established rsTM was separated into

chromatography

(Fig.

2). A sharp

for

the

three

cell line No. 29 two peaks of TM peak of TM anti-

gen eluted at 0.35 M NaCl (rsTMa) followed by a broad elution of TM antigen at higher NaCl concentration (rsTMB). Both peaks of rsTM antigen were separately

pooled

chromatography

and

further

purified

to

homogeneity

by

immunoaffinity

and gel filtration.

The acidic chromatographic describing chondroitin/dermatan

behavior of rsTM/ sulfate on native

and previous reports rabbit lung TM (13.15)

suggested a similar post-translational modification of rsTMB . Purified rsTMB migrated as a broad band of Mr 75,000-95,000 under non-reducing conditions on SDS-PAGE (Fig. 3, lane d). After treatment of rsTMB with chondroitinase AC I. the treated protein migrated as a sharp band at Mr 67,000 (Fig. 3, lane c) similar to that of rsTMa (Fig. 3, lane b). The migration of rsTMa with chondroitinase AC I was not changed (data not shown). RsTMB was digested with chondroitinase AC I and the released unsaturated disaccharides were analyzed by HPLC (Fig. 4). The major peak elution position coincided with the ADi-4S standard ADi-4S and the samples derived from rsTM,8 732

and a 1:l mixture eluted as a single

of standard peak at the

Vol.

171,

No.

2,

1990

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

2

50

40 ” E 2 -

30

F

0 r : a - 1.0

-1

z 20 E

2

2 5

3

10

z”

20

40

Fraction

-0

10 100

80

60

No.

Fig. 2. Ion-exchange chromatography of rsTMs. Cell free culture medium was applied to a Q-Sepharose Fast Flow column equilibrated in 0.15 M NaCl, 20 mM Tris-HCl, pH 7.5 and eluted with a linear gradient of 0.15 to 1.2 M NaCl in 20 mM Tris-HCl. pH 7.5. Fractions were monitored for absorbance at 280 nm (+) and rsTM antigen ( 0 ) by ELISA as described in “Materials and Methods”.

d

1’0

2’0 Time,

3’0

4’0

min

Fig. 3. SDS-PAGE analysis of chondroitinase AC I-digested rsTM. HsTMB was digested with chondroitinase AC I as described in “Materials and Methods” and analyzed by SDS-PAGE using the buffer system of Laemmli (18) with an 8% gel (lane c). All rsTM samples were non-reduced and proteins were stained with Coomassie Brilliant Blue. Lane a, reduced molecular weight standards (Pharmacia); lane b. rsTMa: lane d, rsTMB. HPLC analysis of unsaturated disaccharides derived from chondroiFig. 4. tinase AC I-treated rsTM,9. Elution was performed with 0.01-0.5 M NaH2P04, as Arrows indicate the elution positions described in “Materials and Methods”. of commercial standards.

733

5’0

Vol.

171,

No.

2,

1990

05

rsTM

BIOCHEMICAL

/ Thrornbin,

AND

BIOPHYSICAL

RESEARCH

06

M/M

ATIll

COMMUNICATIONS

/ rsTM,

M/M

Fig. 5. The presence of chondroitin-4-sulfate on rsTMB affects thrombin activity. Fibrinogen (1 mg/ml) was incubated at 37 C with different amounts of the rsTMs in 20 mM Tris-HCl, pH 7.4, 0.15 M NaCl. Thrombin was added to a final concentration of 18 nM and the clotting time was measured. 0 ,rsTMa ; l ,rsTM,9 ; n ,rsTMB treated with chondroitinase AC I. Fig. 6. Effect of rsTMa, fi, and B treated with chondroitinase AC I on AT III-dependent anticoagulant activity. Fibrinogen was incubated at 37 C with 17 nM rsTMs and various concentration of AT III in 0.15 M NaCl, 20 mM TrisHCl, pH 7.4. Thrombin was added (18 nM) and the clotting time was measured. 0 ,rsTMa; 0 .rsTMB; l ,rsTMfl treated with chondroitinase AC I.

same position ADi-4S ment

(data

not

was confirmed with

the

chondro-4-sulfatase

chondroitin-4-sulfate fate

as a major Under

(data

the

does not conclusion

glycosaminoglycan

similar

altering thrombin’s the 50% inhibition

to the not

disaccharide

ADi-OS position shown).

after

Identical

conditions,

act on dermatan that

rsTMfi

sulfate

contains

as

treat-

data

of rsTMB with chondroitinase ABC (data AC I cleaves the N-acetylhexosaminide but

with

of the unsaturated

by the peak shift

obtained by treatment Since chondroitinase are consistent

shown). Identity

were

not shown). linkage in

(19.20) the

data

chondroitin-4-sul-

component. rsTMB

was more effective

clotting activity (Fig. concentration of rsTMa

5). With

than

rsTMn

18 nM bovine

for thrombin,

were 76 and 20 nM, reAC I resulted in activi-

and B Treatment of rsTMB with chondroitinase to that of rsTMa. Rs’IMB also accelerated AT III inhibition of thrombin activity, although high concentrations of AT III were required (Fig. 6). Rsl’Ma did not affect AT III inhibition under these conditions and chondroitinase AC I-treated rs’i’MB effects were similar to rsl?la.

spectively. ty similar

The effects of rsTMa and B function of thrombin concentration

on PC activation (Fig. 7). RsTMa 734

rates were studied as a and B showed different

Vol.

No. 2, 1990

171,

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

60 -

0

10

20

30

Thrombin,

40

50

nM

Fig. 7. Thrombin concentration dependence of PC activation. Human PC (0.5 LLM) was activated with 2 nM rsl?J and various concentration of bovine thrombin and assayed as described in “Materials and Methods”. Values are means for duplicate wells. 0 ,rsTMa ; 0 ,rsTMB. affinities for thrombin with apparent Kd values of 7.4 and 1.4 nM, respectively. The maximum rate of PC activation was higher for rsTMa at saturating thrombtn

concentration.

DISCUSSION The current human recombinant glycan

was present

data describe the presence of chondroitin-4-sulfate soluble TM expressed in CHO-Kl cells. The glycosminoon the acidic

form,

rsTMB, and enzymatic

removal

carbohydrate resulted in TM functional activity similar to rsTMa. Whi.le the functional role of the chondroitin/dermatan fication

has been studied

for

rabbit

lung

TM (12-19,

modification

of human TM. The ability

to

recombinant

protein

detailed

permitted

more

express

is known

substantial

analysis o-f the chondroitin-4-sulfate modification. Recently, Parkinson et al. described decreased

of the

the unmodified sulfate modi-

little

structural

in

about

quantities and

thrombin

of

functional

affinity

for

TM

in endothelial cells treated with an inhibitor of glycosaminoglycan attachment to core proteins (21). Our results are in agreement in that rsTMB had increased affinity and their represents

thrombin parallels

affinity differences

relative

to

rsTMa.

in the bound

This

thrombin

difference interaction

ability to inhibit thrombin clotting activity. the soluble form of native endothelial cell

in thrombin with

AT III

The rsTMB probably TM since the Kd for

thrombin is similar (21) whereas the behavior of rsTMa is more similar to the soluble proteolytic fragment of rabbit TM (22) which does not contain the glycosaminoglycan modification (23). It is possible that rsTMa is an & vitro phenomenon of the transformed cells and is incompletely processed 735

Vol.

171,

No.

2,

1990

BIOCHEMICAL

for post-translational resolved.

carbohydrate

The rsTMs include

the

shown to be minimally 26). Additional sulfate,

although

neous

expression

determine the of post-translational

for

binding

our data

RESEARCH

but

domain

of native

thrombin

binding

affinity

cannot

or a distinct

BIOPHYSICAL

attachment,

EGF-like

essential

thrombin

al contributions

AND

is provided

distinguish

between

this

COMMUNICATIONS

remains

TM which

to

has been

and PC activation by the distant

be

(23-

chondroitin-C conformation-

secondary thrombin binding site. The simultaand rsTM,9 may prove useful in studies to

of rsTMa site of chondroitin-4-sulfate glycosylation

attachment

and the

regulation

of TM.

ACKNOWLEDGMENTS We thank K.N. and K.S. should be considered as equal first authors. Dr. D.J. Stearns-Kurosawa for many helpful discussions and critical reading of this manuscript, Drs. T. Horiuchi and T. Itani for kind valuable advice, and Drs. T. Suzuki and M. Furusawa for his support and encouragement. We are grateful to K. Takeda, N. Honda, K. Yamashita and M. Tanaka for their excellent technical assistance.

REFERENCES

1 Esmon. C.T. (1989) J. Biol. Chem. 264, 4743-4746. 2 .Dittman. W.A.. & Majerus, P.W. (1990) Blood 75. 329-336. 3 Kisiel, W.. Canfield. W.M., Ericsson, LX, & Davie E.W. (1977) Biochemistry 16. 5824-5831. 4 Vehar, G.A., 81 Davie, E.W. (1980) Biochemistry 19. 401-410. 5 Marlar, R.A., Kleiss, A.J., 81 Griffin, J.H. (1982) Blood 59, 1067-1072. 6 Suzuki. K., Kusumoto, H., Deyashiki. Y., Nishioka, J., Maruyama, I., Zushi, M., Kawahara. S., Honda, G.. Yamamoto, S., & Horiguchi, S., (1987) EMBO J. 6, 1891-1897. 7 Jackman, R.W., Beeler, D.L., Fritze. L.. Soff, G., & Rosenberg, R.D. (1987) Proc. Natl. Acad. Sci. USA 84, 6425-6429. 8 Wen, D.. Dittman, W.A., Ye, R.D., Deaven, L.L., Majerus, P.W.. & Sadler, J.E. (1987) Biochemistry 26, 4350-4357. 9 Esmon, C.T., Esmon. N.L.. 81 Harris, K.W. (1982) J. Biol. Chem. 257, 7944-7947. 10 Hofsteenge. J., Taguchi, H., 81 Stone, S.R. (1986) Biochem. J. 237. 243-251. 11 Preissner, K.T., Delvos, U., & Muller-Berghaus, G. (1987) Biochemistry 26. 2521-2528. 12 Bourin, M.-C., Boffa. M.-C.. Bjork, I., 81 Lindahl, U. (1986) Proc. Natl. Acad. Sci. USA 83, 5924-5928. 13 Bourin, M.-C., Ohlin. A.-K., Lane, D.A., Stenflo, J., 81 Lindahl, U. (1988) J. Biol. Chem. 263, 8044-8052. 14 Bourin, M.-C. (1989) Thromb. Res. 54, 27-39. 15 Preissner. K.T., Koyama, T., Muller, D., Tschopp, J., & Muller-Berghaus, G. (1990) J. Biol. Chem. 265, 4915-4922. 16 Salem. H.H.. Maruyama, I., Ishii. H., & Majerus, P.W. (1984) J. Biol. Chem. 259, 12246-12251. 17 Nakane, P.K., 81 Kawaoi, A. (1974) J. Histochem. Cytochem. 22, 1084-1091. 18 Laemmli. U.K. (1970) Nature 227, 680-685. 19 Yamagata, T., Saito, H., Habuchi, O., & Suzuki, S. (1968) J. ~i01. Chem. 243. 1523-1535. 736

Vol.

171, No. 2, 1990

20

Saito, IL, Yamagata, T., & Suzuki, S. (1968) J. Biol. Chem. 243, 1536-1542. Parkinson, J.F., Garcia, J.G.N., & Bang, N.U. (1990) Biochem. Blophys. Res. Commun. 169. 177-183. Kurosawa. S.. Galvin. J.B., Esmon, N.L., & Esmon C.T. (1987) J. Biol. Chem. 262, 2206-2212. Stearns, D.J., Kurosawa. S., & Esmon, C.T. (1989) J. Biol. Chem. 264, 3352-3356. Kurosawa, S., Stearns, D.J., Jackson, K.W., & Esmon, C.T. (1988) J. Biol. Chem. 263, 5993-5996. Suzuki, K., Hayashi, T., Nishioka. J., Kosaka, Y.. Zushi, M., Honda, G., & Yamamoto, S. (1989) J. Biol. Chem. 264, 4872-4876. Zushi, M.. Gomi, K., Yamamoto, S.. Maruyama, I., Hayashi, T., & Suzuki, K. (1989) J. Biol. Chem. 264, 10351-10353.

21 22 23 24 25 26

BIOCHEMICAL

AND BIOPHYSICAL

737

RESEARCH COMMUNICATIONS