Processing of Type 1 Plasminogen Activator Inhibitor (PAI-1) into the

THE JOURNAL OF BIOmlCAL C m S T R Y Vol. 269, No. 5, Issue of February 4, pp. 3675-3880, 1994 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in USA.

Processing of Type 1 Plasminogen Activator Inhibitor (PAI-1) into the Regulated Secretory Pathway*

(Received for publication, August 4, 1993, and in revised form, October 5, 1993)

Lourdes Gombau and Raymond R. SchleefS From The Scripps Research Institute, Department of Vascular Biology, La Jolla, California 92037

To understand the processing of type 1 plasminogen activator inhibitor (PAI-1) into the storage granules of platelets, we utilized a eukaryotic expression vector ( P R O to transfer the human cDNA for PAI-1 into AtT-20 cells, a mouse pituitary cell line known to sort proteins in a regulated fashion. Immunofluorescence staining of PAI-1-transfected AtT-20 clones revealed co- localization of PAI-1 with an endogenously produced and stored hormone ( i e . adrenocorticotropic hormone, ACTH). Stimulation of PAI-1-transfected AtT-20 cells with a secretagogue resulted in the release of both active PAI-1 and the latent form. In comparison, PAI-1- transfected Chinese hamster ovary cells (Le. a nonpack- aging cell line) did not release PAI-1 in response to a secretagogue and exhibited immunoreactivity for PAI-1 primarily confined to the Golgi region. Percoll density gradient fractionation of AtT-20 cells revealed a codis- tribution of PAI-1 and ACTH in cellular compartments of the same density.The half-life of PAI-1 activity at 37 “C was prolonged in intact granules (tin, 6 h) in comparison with its half-life in lysed granules (tllz, 2 h). These studies demonstrate the presence of a new functional property associated with the PAI-1 molecule that directs this inhibitor into the storage secretory pathway.

Type 1 plasminogen activator inhibitor (PAI-1)’ is the primary physiological inhibitor of vascular tissue-type plasminogen activator (t-PA) (1,2). The ability of PAI-1 to neutralize the activity of t-PA and other serine proteases (e.g. urokinase, factor XIa, activated protein C) has led to the classification of this inhibitor in the serine protease inhibitor (serpins) superfamily (1, 2). This inhibitor is produced as an M, 50,000 glycoprotein by a wide variety of cells and is present in blood either at low concentrations in plasma or in a large storage pool within platelets (3-11). Agonist-induced platelet activation is known to cause the release of active PAI-1 which has led to the concept that PAI-1 is stored in conjunction with other hemostatic proteins within platelet a-granules (3,4, 7,9-11). The presence of PAI-1 antigen in megakaryocytes (12,131, the hemopoietic pre- cursor of platelets, suggests that PAI-1 may be deposited into storage organelles ( i e . a-granules) during the maturation of

Fondo de Investigaciones Sanitarias (FIS), Spain, and by Grant HL * This work was supported in part by a fellowship (to L. G.) from El

45954 from the National Institutes of Health (to R. R. S.). The costs of

charges. This article must therefore be hereby marked “aduertisement” publication of this article were defrayed in part by the payment of page

in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $To whom correspondence should be addressed: The Scripps Re-

search Inst., Department of Vascular Biology (CVB-3), 10666 North Torrey Pines Rd., La Jolla, CA 92037.

The abbreviations used are: PAI-1, type 1 plasminogen activator inhibitor; t-PA, tissue-type plasminogen activator; ACTH, adrenocorticotropic hormone; 8-Br-CAMP, 8-bromo-W, CHO, Chinese hamster ovary; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis.

these cells. However, no information exists on the ability of PAI-1 to be specifically targeted into storage granules or if the presence of PAI-1 in a-granules is a nonspecific entrapment event that occurs during megakaryocytopoiesis.

Research from a number of laboratories indicates that PAI-1 is synthesized in an active form, but it is a relatively labile inhibitor that is rapidly converted into an inactive form at 37 “C with a half-life of approximately 1 h (1). The conforma- tion of PAI-1 resulting from inactivation at 37 “C is commonly referred’to as latent PAI-1, because inhibitory activity can be revealed by treatment with denaturants or negatively charged phospholipids (1). In light of the observation that platelets pos- sess low biosynthetic capabilities (141, it is not unexpected that the majority of PAI-1 is present within platelets in a latent form. Although vitronectin is known to be capable of increasing by 2-fold the half-life of PAI-1 activity in solution (37 “C) (151, preliminary data of Pressnier et al. (16) suggest that complexes between vitronectin and PAI-1 are not present in nonactivated platelets. Therefore, little information exists on the mechanisms that account for the presence of active PAI-1 stored within platelets that have a half-life span of 9-12 days in the circulation (14).

Two distinct pathways are known to be responsible in eukaryotic cells for the secretion of proteins from cells (17, 18). The “constitutive” pathway externalizes proteins rapidly using post-Golgi vesicles and does not require an external stimu- lus for release of a compound into the extracellular milieu (17, 18). In the “regulated” pathway, proteins are stored in secretory granules until the cells are stimulated to secrete in response to the appropriate stimuli (17, 18). A number of tumor- derived cell lines exhibit both a constitutive and a storage secretory pathway, and these cell lines have been used as in vitro model cell systems for analyzing the processing of proteins into these two pathways (17, 18). A classical system is the mouse pituitary tumor cell line, AtT-20, that has been shown to divert a majority of the endogenously synthesized adrenocorticotropic hormone (ACTH) into the regulated storage pathway (19, 20). Treatment of AtT-20 cells with the appropriate secretagogue (e.g. 8-Br-CAMP) results in release of the contents of the secretory granule (19,211. These cells have been shown to have the capacity, after transfection with the appropriate DNA, to package heterologous peptide hormones and enzymes into the regulated secretory pathway. For ex- ample, proinsulin (211, trypsinogen (20, 221, human growth hormone (231, and peptidylglycine a-amidating monooxygen- ase (24) are transported to the regulated pathway with a similar efficiency as the endogenous hormone ACTH. This cell line has also proven useful in studies investigating the sorting of two platelet a-granule proteins into the regulated secretory pathway, including P-selectin (25, 26) and von Willebrand Factor (27). In light of this information, this study was de- signed to investigate the applicability of utilizing the AtT-20 cell line as a model system for analyzing the processing of PAI-1 into storage granules.

3875

3876 Plasminogen Activator Inhibitor in Storage Granules

MATERIALS AND METHODS Plasmid Construction and Pansfection of Eukaryotic Cells-The

cDNA encoding the full-length molecule of human PAI-1, including the signal peptide, was kindly provided by Dr. David Ginsburg (University of Michigan) (28). The entire coding region for PAI-1 was excised using the EcoRI site at the 5' end of the cDNA and NsiI site at nucleotide 1328 in the 3"nontranslated region and transferred into an EcoRVPstI-digested pBluescript I1 KS plasmid vector (Stratagene, La Jolla, CA). Two unique restriction sites in the multiple cloning region of pBluescript I1 KS (i.e. Hind111 at the 5' end andXbaI a t the 3' end) were then used to transfer the PAI-1 cDNA in an orientation-specific manner into a Hin-

CA). dIIIIXba1-digested pRC/CMV plasmid vector (Invitrogen, San Diego,

Growth and Pansfection of Eukaryotic Cells-AtT-20 cells and Chi- nese hamster ovary (CHO) cells were obtained from the American Tis- sue Culture Collection (Rockville, MD). Both cell lines were grown in Dulbecco's modified eagle's medium supplemented with 10% fetal calf serum (Life Technologies, Inc.). AtT-20 cells were cultured at 37 "C in the presence of a 15% C02 atmosphere, whereas a 5% C02 atmosphere was used for CHO cells. AtT-20 and CHO cells were transfected with 2.5 pg of the DNA construct (pRC/CMV-PAI-1) by utilizing Lipofectin ac- cording to the manufacturer's instructions (Life Technologies, Inc.). Af- ter 48 h, stable cells were selected by treatment with the antibiotic Geneticin (G-418; Life Technologies, Inc.) at 0.5 mg/ml for AtT-20 cells and 2 mg/ml for CHO cells, and individual clones were isolated by ring cloning. Colonies were subsequently screened for the expression of soluble human PAI-1 antigen as described below.

Immunofluorescence Analysis-AtT-20 and CHO cells were grown for 4 days on polylysine-coated glass coverslips (12-mm diameter; Fisher) in serum-containing media supplemented with 2 nm 8-Br-CAMP (Sigma) as described by Burgess et al. (20) to induce the extension of cellular processes in AtT-20 cells. Both cell lines were fixed with 4% paraformaldehyde for 20 min at 4 "C, washed with PBS, and incubated (22 "C, 20 min) with 0.2 M glycine in PBS to neutralize any residue fixative. Washed cells were permeabilized by incubation (22 "C, 20 min) with 1% Triton X-100 diluted in PBS and incubated (22 "C, 20 min) in PBS containing 1% heat-inactivated goat serum (Life Technologies, Inc.). After blocking, the cells were incubated (22 "C, 1 h) with a primary antibody (i.e. rabbit anti-ACTH or rabbit nonimmune serum, 12300 dilution, Sigma; monoclonal antibody 380 against PAI-1; or mouse nonimmune IgG, 10 pg/ml, American Diagnostica, Inc., Greenwich, CT) diluted in PBS supplemented with 1% heat-inactivated goat serum. Bound antibodies were detected by incubation (22 "C, 1 h) with a com- bination of rhodamine-conjugated goat anti-mouse IgG (1:25 dilution; Calbiochem) and fluorescein-conjugated goat anti-rabbit IgG (1:25 dilution) in 1% heat-inactivated goat serum. After washing, the samples were examined using an Olympus BH-RFL microscope using a mercury lamp for excitation. Fluorescence of fluorescein was detected using a DM-500 dichroic mirror and a M 1 5 bamer filter. Fluorescence of rhodamine was detected using a DM-580 dichroic mirror and an R-610 bamer filter. Samples were photographed using a 10-s exposure with Kodak 3200 ASA.

Secretagogue-induced Secretion of PAI-I-AtT-20 and CHO cells (lo6 celleldish) were plated in 60-mm diameter tissue culture dishes, and the cells were allowed to equilibrate by incubation (37 "C, 24 h) in growth medium. For experimental analysis, the cells were washed twice with PBS and incubated in 2.5 ml of serum-free medium for three consecutive 1-h periods. During the last period, the cells were incubated in serum-free medium supplemented either with or without 5 m 8-Br- CAMP (20, 25). Conditioned media were collected during each 1-h period. At the end of the third incubation period, the cells were washed twice with PBS, detached by incubation in PBS supplemented with 20 m EDTA, centrifuged, and the resuspended cells counted on a hemo- cytometer.

Quantitation ofPAI-1 Activity andAntigen-PAI-1 activity was quan- titated as described previously (29-31) by using immobilized t-PA (American Diagnostica) to bind active PAI-1 in a sample and the bound PAI-1 immunologically detected by incubation with affinity-purified rabbit anti-PAI-1 (10 pg/ml; Ref. 32) followed by 12sI-goat anti-rabbit IgG (50,000 cpdwell; Amersham Corp.). To activate latent PAI-1, samples were incubated with 0.1% SDS (1 h, 37 "C) followed by neu- tralization of the SDS with a 50-fold excess of Triton X-100.

Quantitation of PAI-1 antigen was performed as described previously (30) by using an immobilized monoclonal (2D2) antibody against human PAI-1 to bind PAI-1 antigen in a sample and bound PAI-1 detected as described above.

Subcellular Fractionation of AtT-20 Cells and Dot Blotting

Assays-AtT-20 cells were fractionated on Percoll density gradients (Pharmacia LKB Biotechnology Inc.) utilizing protocols described by Koedam et al. (25). Briefly, AtT-20 cells (five 75-cm2 tissue culture flasks) were washed twice with PBS and detached by scraping into 10 ml of homogenization buffer (0.2 M sucrose, 1 m EDTA, 20 nm Tris- HCI, pH 7.2). The cells were centrifuged, washed with PBS, and resuspended in 2 ml of homogenization buffer. The suspension was subjected to 20 strokes in a Dounce homogenizer (size 19; Kontes Glass Co., Vineland, NJ) at 4 "C. Nuclei and cellular debris were removed by centrifugation (600 x g, 10 mid, and the supernatant was centrifuged (10,000 x g, 5 min) to collect cell membranes and granules. The pellet was resuspended in 200 pl of homogenization buffer, layered on 13 ml of 20% Percoll, 0.25 M sucrose, 20 n m Tris-HC1, pH 7.2, in a Beckman quick-seal centrifuge tube (16 x 76 mm), and centrifuged (20,000 xg, 30 min) in a Beckman L7-65 ultracentrifuge using a Vti65 rotor. Fractions (800 pl) were collected manually from the top and frozen at 80 "C until assayed.

The distribution of ACTH, PAI-1, and laminin in the fractions from the Percoll gradient were determined by dot blotting. For this proce- dure, aliquots (100 pl) were applied on a nitrocellulose m'embrane (0.2 pm; Schleicher and Schuell) using a Millipore vacuum dot blotting apparatus. After blocking of the nitrocellulose with Blotto (5% non-fat dry milk in 10 m Tris-HC1, pH 7.5), the membranes were incubated (1 h, 22 "C) in the appropriate primary rabbit antibody diluted 1:250 in Blotto (i.e. anti-ACTH, Sigma; anti-laminin, Telios Pharmaceutical, Inc., San Diego, CA anti-PAI-1, Ref. 32). Washed membranes were incubated (1 h, 22 "C) with lZ6I-goat anti-rabbit I& (los cpdml diluted in Blotto), washed, and the radioactivity associated with each dot determined in a y-counter.

Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS- PAGE) and Reverse Fibrin Autography-SDS-PAGE and reverse fibrin autography were performed as described previously (29, 31). Briefly, samples were subjected to SDS-PAGE, and the electrophoresed gel was subsequently soaked for 1.5 h in 2.5% Triton X-100. Fibrin-agar indi- cator films were prepared containing 1% agarose, 2.4 mg/ml fibrinogen, 25 pg/ml human plasminogen, 0.5 unit/ml a-thrombin, and 0.05 unit/ml urokinase (all h a 1 concentrations). The solutions were mixed and poured onto prewarmed glass plates. After solidification of the fibrin- agar film, the Triton X-100-washed SDS-gel was layered onto the film and incubated in a humid chamber. PAI activity is revealed by the development of opaque lysis-resistant zones in the otherwise clear in- dicator film.

RESULTS Production and Deposition of PAI-1 within PAI-1-transfected

AtT-20 Cells-From three transfection experiments with the vector pRC/CMV/PAI-1, 10 colonies were isolated that were neomycin-resistant and produced human PAI-1 antigen. Based upon the production of PAI-1 antigen over a 24-h period, one high PAI-1 producer (i.e. clone AtT-20-pRC-CMV/PAI-1-3F, 201.5 &mu24 h/105 cells) was selected for further studies. Because our initial screening utilized a murine monoclonal antibody capable of reaction against total PAI-1 (i.e. free PAI-1, t-PAPAI-1-complexes, etc.), we decided to employ reverse fibrin autography to gain an insight into the levels of free PAI-1 produced and associated with the AtT-20 cells (Fig. 1). Condi- tioned media and cell lysates of both nontransfected and PAI- 1-transfected AtT-20 cells were fractionated by SDS-PAGE and placed onto a fibrin-agarose film that was supplemented with both plasminogen and urokinase to initiate fibrin degradation. After 2 h at 37 "C, lysis-resistant zones were evident in both the cell lysates (Fig. 1, lanes 1 and 2 ) and conditioned media samples (lanes 3 and 4 ) that co-migrated with purified PAI-1. The size of the lysis-resistant zones were significantly larger in the samples from the PAI-1-transfected AtT-20 (lanes 2 and 4 ) than those from nontransfected cells (lanes 1 and 3 1. Immuno- precipitation experiments revealed that the majority of the inhibitor activity present in lanes 2 and 4 could be depleted utilizing Sepharose beads conjugated to a monoclonal antibody against human PAI-1, whereas no decrease in the lysis-resistant zones was observed utilizing this Sepharose-antibody com- plex to immunoprecipitate the samples of nontransfected cells (data not shown).

1 2 3

Plasminogen Activator Inhibitor in Storage Granules

-97.4 kd

-69 kd

-46 kd

FIG. 1. PAI-1 activity profile of control and PAI-1-transfected AtT-20 cells by reverse fibrin autography. Samples from nontrans-

2 and 4 ) were subjected to SDS-PAGE and analyzed for PAI-1 activity fected (lanes 1 and 3 ) and PAI-1-transfected AtT-20 cells (clone 3F, lanes

by reverse fibrin autography as described under “Materials and Meth- ods.” Samples included cell lysates (lanes 1 and 2, lo5 cellsflane) and the corresponding media conditioned by lo5 cells (5100 plflane, lanes 3 and 4) . Lane 5 contained nonconditioned medium (100 pl), and lane 6 contained 100 ng of purified human PAI-1. Position of M, markers are indicated.

To investigate if human PAI-1 produced in the transfected cells was contained within storage granules, experiments were directed at co-localizing this inhibitor with an endogenously produced and stored hormone (ie. ACTH) (20). Double label immunofluorescence experiments were employed initially to ascertain the distribution of these two molecules. Fig. 2 indicates that the staining for PAI-1 (Fig. 2 A ) and for ACTH (Fig. 2 B ) could be co-localized in AtT-20 cell processes. No staining for ACTH (Fig. 2 0 ) or PAI-1 (Fig. 2 E ) was observed if the primary antibody was replaced with rabbit IgG or nonimmune mouse. As a control in these experiments, CHO cells were transfected with the pRC/CMVPAI-l construct, and stable colonies were selected based upon their resistance to G-418 and production of human PAI-1. A similar analysis for human PAI-1 and ACTH revealed that PAI-1 was primarily confined to the Golgi region in CHO cells with little staining in the cell periph- ery (Fig. 2G). No staining for ACTH was observed in the PAI- 1-transfected CHO cells (Fig. W).

Characterization of PAI-1 Released and Stored by PAI-1- transfected AtT-20-It is well known that the secretory products contained with dense core secretory granules can be released following stimulation with certain secretagogues (e.g. 8-Br-CAMP) (19, 20, 22). Further evidence for the presence of PAI-1 within storage granules could be obtained if the release of this inhibitor was increased in the presence of an appropriate secretagogue. Therefore, PAI-1-transfected AtT-20 and CHO cells were incubated for three separate 1-h incubation periods in the presence or absence of the secretagogue 8-Br-CAMP and the conditioned media assayed for PAI-1 antigen. The first two successive 1-h incubation periods shown in Fig. 3 were employed to determine the constitutive expression of PAI-1 in both cell lines. Fig. 3A indicates that the presence of 8-Br-cAMP during the third incubation period with the PAI-1-transfected AtT-20 cells resulted in a 60-fold increase in the PAI-1 antigen content in the conditioned media, which is equivalent to 18% of the total cellular PAI-1. In contrast, no increase in the PAI-1 content of the media conditioned by CHO cells was observed upon their incubation with 8-Br-CAMP (Fig. 3B).

3877

PAI-1-transfected cells. PAI-1-transfected AtT-20 cells (A-F) and FIG. 2. Immunofluorescence analyses of PAI-1 and ACTH in

PAI-1-transfected CHO cells (G and H ) were fixed, permeabilized, and incubated with the following primary antibodies: A, C, and G, monoclonal antibody 380 against human PAI-1; B, F, andH, rabbit antiserum to human ACTH; D and E , rabbit nonimmune serum or mouse nonimmune I&, respectively. Bound antibody was detected by incubation with rhodamine-labeled goat anti-mouse IgG (A, C, E , and G ) and fluorescein-labeled goat anti-rabbit IgG (B, D, F, and H). Fluorescence of rhodamine or fluorescein was photographed at x 1000 magnification as described under “Materials and Methods.”

Although current information indicates that PAI-1 is synthesized in an active form (l), both the active and latent forms of PAI-1 have been detected stored in platelet a-granules (7, 11, 33,34). To determine the applicability of this model cell system for understanding the PAT-1 molecule released from platelet storage organelles (i.e. a-granules), it was relevant to charac- terize the AtT-20 cell-associated and secretagogue-released forms of PAI-1. Fig. 4A indicates that both the active (specific activity: 4,158 unitdmg) and latent form (specific activity following SDS treatment: 8,800 unitdmg) of PAI-1 could be detected in the media conditioned by the AtT-20 cells incubated in the presence of the secretagogue 8-Br-CAMP. In comparison, neither form of PAI-1 could be detected in the media conditioned by these cells incubated in the absence of a secretagogue (Fig. 4A). Analysis of the cell lysates revealed that both the active (specific activity: 976 unitdmg) and latent (specific activity following SDS treatment 3,771 units/mg) forms were present in the PAT-1-transfected AtT-20 cells. Incubation of these cells with a secretagogue resulted in a decrease in both forms of PAI-1 (Fig. 4B). In contrast, only the active form could be detected in the samples of the CHO cells, which was not

3878 Plasminogen Activator Inhibitor in Storage Granules

550

500

450

Z 400 W h

350

z 10 300 - > 250

I - z 4 0

_I > 200 a, a 150

100

50

0

A. L

B. ,tT-20-pRC/CUV/PAI-l-3F CHO-pRC/CYV/PAI-l-ZC

HRS 1 2 3 3 CAMP - - - +

1 2 3 3 - + "

by PAI-1-transfected At"-20 cells and CHO cells. Clone AtT-20- FIG. 3. Kinetics of PAI-1 release into the medium conditioned

pRCICMVk'AI-1-3F (A) and clone CHO-pRC/CMV/PAI-l-BC ( E ) were plated into 60-mm dishes (n = Wclone) in complete media. The next day, the cells were washed twice with PBS and incubated in serum-free medium for three consecutive 1-h periods. The numbers 1,2, and 3 refer to the consecutive incubation periods. During the final incubation pe-

(n = 4) of 5 m~ 8-Br-CAMP. The conditioned medium was harvested, nod, the cells were maintained either in the presence (n = 4) or absence

centrifuged, and assayed for PAI-1 antigen as described under "Mate- rials and Methods." Data are represented as mean * standard devia- tion.

8-Br-CAMP ~ - + + - - + + " + + " + + A t l - 2 0 CHO All-20 CHO

FIG. 4. Activity status of PAI-1 associaWreleased from PAI-1- transfected cells under basal and secretagogue-stimulated con- ditions. PAI-1-transfected AtT-20 and CHO cells were washed twice over a 2-h period and then incubated for 1 h in the presence (n = 4) or absence (n = 4) of 8-Br-CAMP as described in the legend to Fig. 3. The conditioned media were harvested and the cells were collected with 20 m~ EDTA, washed, resuspended in PBS, and counted. Conditioned media (A ) and cell lysates ( E ) were assayed for PAI-1 activity using the functional immunoassays as described under "Materials and Methods." Active PAI-1 is represented by open bars, whereas inhibitor activity detected following treatment with a denaturant (e.g. SDS) is indicated by the closed bars.

affected by addition of 8-Br-CAMP (Fig. 4, A and B ) . The secretagogue-induced release of the active and latent

form of PAI-1 (Fig. 4) would suggest that both forms of this inhibitor are stored within the transfected AtT-20 cells. Evi- dence for this concept would be obtained by the isolation and analysis of PAI-1 containing dense core secretory granules. For this reason, PAI-1-transfected AtT-20 cells were subfraction- ated on a Percoll density gradient and the isolated fractions analyzed for the presence of three secretory proteins by dot blotting. PAI-1 antigen (Fig. 5 A ) was recovered as two well separated bands: (i) a low density band that has been observed to correspond to rough endoplasmic reticulum, Golgi apparatus, cellular membranes (20,25) and (ii) a high density, storage

1000 A.

0 ~ " " " " " " " 1200

1000

100

100

400

200

=\ 750,

500 - =\ - =\ ., =\

=\ 250 - =\ =\

, . , , , , , a n n , ' * , , - 1 2 3 4 5 8 7 I 0 101112131415 ' I

Fraction

gradient fractionated PAI-1-transfected AtT-20. Clone AtT-20- FIG. 5. Analysis of PAI-1, ACTH, and laminin in Percoll density

pRC/CMV/PAI-1-3F (five T-75-cm2 tissue culture flasks) were harvested, washed, homogenized, and the 10,000 x g pellet was layered on 20% Percoll as described under "Materials and Methods."After centrifugation for 30 min at 20,000 x g, 800-pl fractions were collected from the top and analyzed by dot blotting using rabbit anti-PAI-1 (A), rabbit anti-ACTH ( E ) , or rabbit anti-laminin (C) as described under "Materi- als and Methods."

granule-containing region (20,25). In agreement with previous published reports (20, 251, Fig. 5B indicates that the endogenously synthesized and stored hormone, ACTH, was associated with both the low density fractions (i.e. endoplasmic reticulum, ek.) and the high density fractions. In contrast, laminin, a secreted extracellular matrix protein that is known to be transported exclusively via the constitutive secretory pathway (201, was detected only in the region with low density (Fig. 5C). Dense granules present within fractions 14 and 15 of the Percoll gradient were pooled and analyzed for PAI-1 activity. Fig. 6 (inset) demonstrates that both the active and latent form of this inhibitor were associated with these dense granules. To determine if the active form of PAT-1 is stabilized

Plasminogen Activator Inhibitor in Storage Granules 3879

- E \ 3

1.5

1 .o

0.5

0.0

ACTIVE SDS-ACTIVATED

0 4 0 12 I 6 20 24 28 32 36 40 44 40

TIME FIG. 6. Activity status and stability at 37 "C of PAL1 in isolated

dense core secretory granules. PAI-1-transfected AtT-20 cells were fractionated on 20% Percoll as described in the legend to Fig. 5 , and the dense secretory granule-containing fractions (i.e. fractions 14 and 15; density 1.062-1.070 a t e r ) were pooled. Dense granules (0, 0) and purified PAI-1 (0, D) were incubated at 37 "C in the absence (closed symbols) or presence (open symbols) of 1% Triton X-100. At the indicated times, the samples were analyzed for residual PAI-1 activity in the functional immunoassay as described under "Materials and Meth- ods." Inset, dense secretory granules were assayed for active PAI-1 (open bars) or for PAI-1 activity following treatment with SDS (closed bars).

within these storage organelles, the Percoll-isolated dense granules were incubated at 37 "C and the activity of PAI-1 monitored over a 48-h period. Fig. 6 indicates that the half-life of PAI-1 activity in the isolated granules was 5.5 h in comparison with purified active human PAI-1, which decayed with a half-life of 1 h. Lysis of the Percoll-isolated granules with Triton X-100 prior to their incubation at 37 "C reduced the half-life of PAT-1 activity to 2 h (Fig. 6).

DISCUSSION

Platelet PAT-1 has been established to play a key role in regulating the fibrinolytic system (35,36). Although PAI-1 was first detected in platelets in 1984 (3,7), little is known about its origin (e.g. uptake from the plasma or synthesis in megakaryocytes) or the mechanisms that stabilize this relatively labile inhibitor within platelets, which have a life span in the circulation of 9-12 days and are formed over a period of 3-5 days during megakaryocytopoiesis (14). This study provides morpho- logical and biochemical evidence indicating that PAI-1 contains a functional region or domain that enables this inhibitor to be directed into the regulated secretory pathway. First, double label immunofluorescence experiments utilizing PAI-l-transfected AtT-20 cells indicate that PAI-1 co-distributes with the endogenously stored hormone ACTH (Fig. 2). Second, fractionation of PAI-1-transfected AtT-20 cells over Percoll gradients resulted in the co-distribution of PAI-1 and ACTH in a cellular compartment (Fig. 5) that is known to contain dense core secretory granules (20). Third, the release of PAT-1 was observed to be increased following stimulation with a classical secretagogue (i.e. 8Br-CAMP; Fig. 3). The latter property being char- acteristic of a component that is routed into the regulated secretory pathway (17).

Current information concerning the sorting of proteins into the regulated secretory pathway using AtT-20 cells and a number of other in vitro model cell systems indicate that two distinct mechanisms may participate in this process: protein-pro-

tein aggregation and specific sorting signals (17, 18). Supportive evidence for a specific sorting signal has been re- cently provided by transfection experiments utilizing the AtT-20 cell line and the cDNA for P-selectin, a platelet a-granule protein (26). More specifically, chimeric constructs of the cytoplasmic region of P-selectin and the luminal domain from another plasma membrane protein (2.e. tissue factor) were observed to be routed into secretory granules, thus demonstrating that the 35-amino acid cytoplasmic domain of P-selectin contains the necessary information for its storage (26). The presence of this domain on P-selectin's cytoplasmic region has been speculated to play a direct role in sorting by permitting its direct interaction with the underlying submembrane cytoskel- eton (181, whereas prevailing theories (17, 37) would require that a soluble protein (e.g. PAI-1) contain a signal that interacts with a membrane-associated receptor dedicated to facilitate sorting. Because a highly conserved sequence of amino acids has not been identified within the diverse group of proteins that are routed into the regulated secretory pathway, Kizer and Tropsha (38) have proposed that a degenerate motif occupying one side of an amphipathic a-helix may function as a targeting signal for this pathway. Analysis of the amino acid sequence of PAI-1(28) reveals that the degenerate motif proposed by Kizer and Tropsha (38) is present in several regions of PAI-1 (i.e.

). As an alternative to a sorting signal, the ability of certain secretory products to form molecular aggregates with each other and condense into electron dense material within the Golgi has been proposed as a mechanism to initiate sorting by excluding nonaggregating molecules from the forming dense core secretory granule (18,371. It is known that several factors appear to play a role in the aggregation or condensation of molecules within the trans Golgi, including a high calcium concen- tration and a low pH. Experiments with agents that disrupt the internal pH within the trans Golgi have provided biochemical evidence, indicating that mildly acidic pH plays a role in protein condensation and sorting of several molecules (17, 18, 39). In light of these data, it is relevant to note that PAI-1 also exhibits a low isoelectric point (i.e. 4.5-5) (1) and that the activity of this molecule is prolonged at mildly acidic pH (40).

Our study also provides data indicating that the activity status of PAI-1 stored in the transfected AtT-20 cells is similar to its activity status within porcine (34), canine (331, and human platelets (4,6,9,33,34). Present data indicate that platelet PAI-1 is composed of two distinct forms: the majority in a latent but denaturant activatable form and a second population comprising its active form (4,6,9,33,34). Thus, the active and latent denaturant-activatable form of PAI-1 are present in a 1:3 ratio both in platelets and the transfected AtT-20 cells, and these two forms of PAI-1 can be released from both platelets and the transfected AtT-20 cells following agonist-induced secretion (Fig. 4). In addition, stabilization experiments revealed that the activity of PAI-1 within the isolated dense core secretory granules exhibit a half-life ( t lI2) of 5 h at 37 "C in comparison with the relative lability of this inhibitor in solution (tv2 = 1 h, Fig. 6). Preliminary data from our group2 indicate that activity of PAI-1 within isolated platelet a-granules also ap- pears to be stabilized ( t m , 5 h). Therefore, the data in this report indicate that the transfection experiments with AtT-20 cells and PAT-1 are applicable for dissecting the mechanisms involved in both the targeting and stabilization of PAI-1 into storage granules.

Ser172-bu175 Thr184-hu192 hu295-hu298 Th$06-bu313

Markillie, Scott Curriden, and Karen Roegner for technical assistance Acknowledgments-We express our appreciation to Lye-Meng

I. M. Lang, L. Gombau, and R. Schleef, personal observations.

3880 Plasminogen Activator Inhibitor in Storage Granules and Greg Reynolds for assistance in the preparation of the final manu- script.

REFERENCES 1. Loskutoff, D. J., Sawdey, M., and Mimum, J. (1989) in Progress in Hemostasis

and Thrombosis (Coller, B., ed) pp. 87-115, W. B. Saunders Company, Phila- delphia

2. Sprengers, E. D., and Kluft, C. (1987) Blood 69,381-387 3. Erickson, L. A,, Ginsberg, M. H., and Loskutoff, D. J. (1984) J. Clin. Znuest. 74,

4. Booth, N. A,, Cmll, A., and Bennett, B. (1990) Fibrinolysis 4, Suppl. 2,13.+140 5. Declerck, P. J., Alessi, M.-C., Verstreken, M., Kruithof, E. K. O., Juhan-Vague,

6. Kruithof, E. K. O., Nicolosa, G., andBachmann, F. (1987)Blood 70,1645-1653 7. Booth, N. A,, Anderson, J. A,, and Bennett, B. (1985) J. Clin. Pathol. 38,

8. Simpson, A. J., Booth, N. A., Moore, N. R., and Bennett, B. (1990) Br. J.

9. Sprengers, E. D., Akkerman, J. W. N., and Jansen, B. G . (1986) Thromb.

10. Kruithof, E. K. O., Tran-Thang, C., and Bachmann, F. W. (1986) Thromb.

11. Booth, N. A,, Simpson, A. J., Croll, A., Bennett, B., and Macgregor, I. R. (1988)

12. Simpson, A. J., Booth, N. A., Moore, N. R., and Bennett, B. (1991) J. Clin.

13. Konkle, B. A,, Schick, P. K., He, X., Liu, R. J., and Mazur, E. M. (1993)

14. Bithell, T. C. (1993) in Clinical Hematology (Bithell, T. C., ed) pp. 511-539, Lea

15. Preissner, K. T., and Jenne, D. (1991) Thromb. Huemostasis 66, 123-132 16. Preissner, K. T., Holzhuter, S., Justus, C., and Muller-Berghaus, G. (1989)

17. Burgess, T. L., and Kelly, R. B. (1987) Annu. Reu. Cell Biol. 3, 243-293 18. Rindler, M. J. (1992) Curr: Opin. Cell Biol. 4,616-622

20. Burgess, T. L., Craik, C. S., and Kelly, R. B. (1985) J. Cell Biol. 101, 639-645 19. Gumbiner, B., and Kelly, R. B. (1982) Cell 28, 51-59

1465-1472

I., and Collen, D. (1988) Blood 71,220-225

825-830

Huematol. 76, 543-548

Huemostusis 65,325-329

Huemostusis 5S.201-205

Br. J . Hqematol. 70, 327-333

Pathol. 44, 139-143

Arterioscler: Thromb. 13, 669474

& Febiger, Philadelphia

Blood 74, 1989-1996

21. Moore, H.-P. H., Walker, M. D., Lee, F., and Kelly, R. B. (1983) Cell 35.531638 22. Burgess, T. L., Craik. C. S., Matsuuchi, L., and Kelly, R. B. (1987) J. Cell Biol.

23. Moore, H. H., and Kelly, R. B. (1986) Nature 321,443446 24. Milgram, S. L., Johnson, R. C., and Mains, R. E. (1992) J. Cell Biol. 117,

717-728 25. Koedam, J. A., Cramer, E. M., Briend, E., Furie, B., and Furie, B. C. (1992) J.

Cell Biol. 116, 617-625 26. Dirrdier, M., Momssey, J. H., Fugate, R. D., Bainton, D. F., and McEver, R. P.

(1992) Mol. Biol. Cell 3,309321 27. Wagner, D. D., Saffaripour, S., Bonfanti, R., Sadler, J. E., Cramer, E. M.,

Chapman, B., and Mayadas, T. N. (1991) Cell 64,403-413 28. Ginsburg, D., Zeheb, R., Yang, A. Y., RaITerty, U. M., Andreasen, P. A., Nielsen,

L., Dano, K, Lebo, R. Y., and Gelehrter, T. D. (1986) J. Clin. Invest. 78,

29. Schleef, R. R., Sinha, M., and Loskutoff, D. J. (1985) J. Lub. Clin. Med. 106, 1673-1680

30. Schleef, R. R., Higgins, D. L., Pillemer, E., and Levitt, J. J. (1989) J. Clin. 406415

31. Schleef, R. R., Bevilacqua, M. P., Sawdey, M., Gimbrone, M. A,, Jr., and Loskut- Inuest. 83, 1747-1752

32. Schleef, R. R., Loskutoff, D. J., and Podor, T. J. (1991) J. Cell Biol. 113, off, D. J . (1988) J. Biol. Chem. 263,5797-5803

33. Lang, I. M., Marsh, J. J., Konopka, R. G.. Olman, M. A., Binder, B. R., Moser, 1413-1423

34. Lang, I. M., Marsh, J. J., Moser, K. M., and Schleef, R. R. (1992) Blood 80, K M., and Schleef, R. R. (1993) Circulation 87, 1990-2OOO

2269-2274 35. Levi, M., Biemond, B. J., van Zonneveld, A.J., ten Cate, J. W., and Pannekoek,

105,659468

36. Fay, W. P., Shapiro, A. D., Shih, J. L., Schleef, R. R., and Ginsburg, D. (1992) H. (1992) Circulation 85, 305-312

37. Kelly, R.B. (1991) Curr: Opin. Cell Biol. 3, 654-660 38. Kizer, J. S., and Tropsha, A. (1991) Biochem. Biophys. Res. Commun. 174,

39. Mayadas, T. N., and Wagner, D. D. (1989) J. Biol. Chem. 264, 13497-13503 40. Lindahl, T. L., Sigurdardottir, O., and Wiman, B. (1989) Thromb. Huemostasis

N. Engl. J. Med. 327, 1729-1733

586-592

62, 748-751

Documents

Processing of Type 1 Plasminogen Activator Inhibitor (PAI-1) into the