Studies on the Acute Release of Tissue-Type Plasminogen Activator From Human Endothelial Cells In Vitro and in Rats In Vivo:

Evidence for a Dynamic Storage Pool By Y. van den Eijnden-Schrauwen, T. Kooistra, R.E.M. de Vries, and J.J. Emeis

The process of acute release of tissue-type plasminogen acti- vator (tPA) is important in locally speeding up fibrinolysis. Using a sensitive enzyme-linked immunosorbent assay for tPA, we investigated the acute release of tPA from cultured human umbilical vein endothelial cells. The addition of thrombin (0.003 to 3 NIH UlmL) caused the dose-dependent release of noncomplexed, enzymatically active tPA into the medium. The amount of tPA released into the medium by thrombin was similar to the difference in the amounts of tPA present in extracts from thrombin-treated cells and control cells. The process of acute release of tPA was complete in 1 minute, whereas the concomitant release of von Willebrand factor into the medium was slightly slower (maximum after 3 minutes). By increasing (c.q. decreasing) tPA synthesis, it

ISSUE-TYPE PLASMINOGEN activator (tPA) is a key enzyme in the onset of fibrinolysis. It converts plas-

minogen to plasmin, which degrades the fibrin component of a thrombus. tPA incorporated into a clot during coagula- tion is, in vitro, more efficient in dissolving a clot than tPA added at a later time.’.* Therefore, the speed and extent of local delivery of tPA during thrombus formation is important in speeding up thromboly~is.~ In humans, plasma levels of tPA increase rapidly after venous occlusion or after the ad- ministration of, for example, desmopressin or adrenalin? There is general agreement that this tPA is derived from the vascular endothelial cells.5 Secretion of tPA can occur by a constitutive and a regulated p a t h ~ a y . ~ This latter pathway, also called the acute release of tPA, only occurs once recep- tors on endothelial cells have been triggered.6 Until now, the acute release process has been studied mainly in animal models, whereas studies on the acute release of tPA from human endothelial cells were limited,’,’ presumably because of the very low tPA concentrations involved. In contrast, the acute release of von Willebrand factor (vWF), another pro- tein which is secreted from the endothelium by both a consti- tutive and a regulated pathway, has been intensively studied in vitro, probably because of the relatively large amounts that are secreted in culture.’ However, the recent development of a sensitive enzyme-linked immunosorbent assay (ELISA) for tPA antigen has made it possible to study in vitro the acute release of tPA as well.” The present study was de- signed to characterize the secretion of tPA from human endo- thelial cells in vitro, especially with regard to the distinction

T

From the Division of Vascular and Connective Tissue Research, Gaubius Laboratory TNO-PG, Leiden, The Netherlands.

Submitted July 28, 1994; accepted January 30, 1995. Supported by a grant from the Netherlands Heart Foundation. Address reprint requests to Y. van den Eijnden-Schrauwen, PhD,

Gaubius Laboratory TNO-PG, P.O. Box 430, 2300 A K Leiden, The Netherlands.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advenisement” in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 1995 by The American Society of Hematology. 0006-4971/95/8512-0026$3.00/0

3510

was found that the amount of tPA constitutively secreted, the amount acutely released, and the amount in cell extracts were increased k.q. decreased) to the same extent. The same relation was found in vivo. When rats were pretreated with cholera toxin or retinoic acid to increase tPA synthesis, plasma levels of tPA were increased, whereas acute release of tPA, as induced by bradykinin, was increased to the same extent. Acutely released tPA and constitutively secreted tPA were liberated from different pathways in human umbilical vein endothelial cells; tPA had, relative to the in vivo situa- tion, a short residence time in the acutely releasable path- way. 0 1995 by The American Society of Hematology.

between constitutive secretion and regulated secretion of tPA. Also, we studied the relationship between synthesis and acute release of tPA.

MATERIALS AND METHODS

Materials. Cell culture reagents were from Flow Laboratories (Irvine, Scotland). Sterile, pyrogen-free, human serum albumin (HSA; 20% wt/vol) was from the Central Laboratory of the Nether- lands Red Cross Blood Transfusion Service (Amsterdam, The Neth- erlands). L-methionine was from Life Technologies (Breda, The Netherlands). 35S-Methionine, labeling grade, was from Amersham (’s-Hertogenbosch, The Netherlands). Human cy-thrombin, calcium ionophore-free acid (A23187), sodium butyrate, cholera toxin, bra- dykinin, and all-trans retinoic acid (RA) were from Sigma (St Louis, MO). Materials used in the human and rat tPA ELISA and in the vWF ELISA have been described.’””’ Casein was from Merck (Darmstadt, Germany), and sodium pentobarbital (Nembutal) from Sanofi (Paris, France). Mouse monoclonal antibody against fluores- cein isothiocyanate (FITC) was kindly provided by Dr R. Bos of our institute.

Dejnitions. Constitutive secretion of tPA (or vWF) is defined as the amount of tPA antigen (or vWF antigen) secreted during the 30-minute preincubation period in Medium 199 (M199)MSA. The tPA (vWF) present in the cell extract is defined as the amount of tPA (vWF) found in control extracts at t = 3 minutes. The thrombin- induced acute released tPA (vWF) is defined as the amount of tPA (vWF) in medium of thrombin-treated cells minus the amount of tPA in medium of control cells at t = 3 minutes.

Cell culture. Human umbilical vein endothelial cells (HUVECs) were isolated using the method of Jaffe et al” and were cultured as previously describedL4 in M199, containing 10% (vol/vol) human serum (HS), 10% (voUvol) heat-inactivated newborn calf serum, 100 IU/mL penicillin, 100 pg/mL streptomycin, 200 pg/mL endothelial cell growth factor (ECGF), 2.5 U/mL heparin, and 2 mmoUL L- glutamine (henceforth called total M199 medium) under 5% CO2 at 37°C. Cells were passaged once by trypsin/EDTA treatment at a split ratio of 1:3 and were cultured to confluency in 2-cmZ wells (final density 2 X lo5 cells/well).

Induction of the acute release of tPA in vitro. To exclude my interference of HS, ECGF, or heparin with the acute release mecha- nism, HUVECs were cultured in M199 medium, containing 20% (vol/vol) newborn calf serum, 1 0 0 IU/mL penicillin, 100 pg/mL streptomycin, and 2 mmoK L-glutamine (“deficient” medium) for 2 hours before an experiment. The 2-hour conditioned media were collected.” The cells were then washed twice with sterile phosphate- buffered saline (PBS; pH 7.4) and 0.3 mL M199iHSA medium

Blood, Vol 85, No 12 (June 15), 1995: pp 3510-3517

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

STUDIES ON THE ACUTE RELEASE OF TPA 351 1

(M199, containing 0.03% (wt/vol) HSA, 100 IU/mL penicillin, 100 pg/mL streptomycin, and 2 mmom L-glutamine) was added to the wells. Thirty minutes later, at t = 0, 15 pL human a-thrombin in M199iHSA was added to the indicated final concentration (generally 1 NIH U/mL). The same volume of M199iHSA medium was added to control cells. The media were then collected after 3 minutes, unless otherwise stated. The cells were washed once with ice-cold PBS and immediately put on ice. Cell lysates were made by scraping the cells with a rubber policeman into 0.3 mL PBS containing 0.05% (vol/vol) Tween 20, 0.01 m o m EDTA, and 0.5% (vol/vol) Triton X-100. All incubations were performed at 37°C.

Decreasing the constitutive secretion of tPA. To decrease the constitutive P A secretion, HUVECs were cultured for 0, 2, 4, 8, and 24 hours in the deficient medium described above.

Increasing the constitutive secretion of IPA. HUVECs were in- cubated for 24 hours with 0, 0.3, 1, or 3 mmom sodium butyrate" or with 1 pmoVL RA16 in total M199 medium to increase the consti- tutive secretion of tPA.

Methionine-labeling of the constitutive and regulated tPA path- ways. HUVECs were cultured for 24 hours in total M199 medium, containing 3 mmoVL sodium butyrate. The cells were then incubated in total M199 medium that had been made free of methionine by overnight dialysis against methionine-free M199 and that contained radiolabeled '%-methionine (20 pCilwel1) and 3 mmoVL sodium butyrate. A 100-fold excess of unlabeled methionine was added after 0, 1,2, 3, or 4 hours. After a total incubation time of 4 hours, release was induced; media and cell extracts were prepared, and casein and EDTA were added as described." The collected media and cell extracts were incubated overnight at 4°C in microtiter plate wells, which had been coated with antibodies against tPA as described for the ELBA procedure." For controls, we used wells that had been coated with the same concentration of a monoclonal mouse antibody against FITC. After washing, the bound protein was dissolved with 0.3 moVL NaOH, neutralized with 1.5 m o m HCI, and counted.

Inhibition of the constitutive and regulated tPA pathways by inhib- iting protein synthesis. To inhibit protein synthesis, HUVECs were incubated with 5 pg/mL cycloheximide in total M199 for 0, 0.5, or 1 hour. The acute release was then induced in the presence of 5 pg/ mL cycloheximide (without culturing the cells for 2 hours in defi- cient medium). In this experiment, 2 pmoVL calcium ionophore A23187 was used to induce acute release of tPA, because calcium ionophore A23187 is a receptor-independent stimulant of tPA and vWF release, which is not dependent on protein synthesis. Also, A23187 has been shown to induce the acute release of tPA."

Measuring overall protein synthesis. Overall protein synthesis was determined by incorporating "S-methionine into endothelial cell proteins and measuring the amount of 3'S-methionine in the trichloroacetic acid-precipitable fraction, as described."

Induction of the acute release of tPA in vivo. A control group of eight male Wistar rats (Iffa-Credo, Someren, the Netherlands), weighing 200 to 300 g, were fed a standard laboratory diet. To a second group of eight animals, RA (8 mgkg body weight) was orally administered daily for 5 days to increase tPA synthesis.16 A third group was intravenously injected with cholera toxin (500 pg/ kg body weight) 48 hours before an experiment. Rats were anesthe- tized with Nembutal (60 mgkg intraperitoneally) and were cannu- lated, and bradykinin (50 pgkg body weight) was injected as a bolus into the vein of the penis. Blood was collected through a carotid artery cannula, and citrated plasma was prepared. Rat tPA antigen concentrations were determined in citrated plasma by ELISA.I7 Ani- mal experiments had been approved by the Animal Experiments Committee of the Netherlands Organization for Applied Scientific Research TNO and were in accordance with the guidelines on animal experimentation presented to the International Committee of Throm- bosis and Haemostasis.'*

Assays. Human tPA antigen was measured by ELISA as de-

scribed." Recombinant human one-chain tPA (Activase; Genetech, San Francisco, CA) was used for making calibration curves in a range of 30 to 500 pg/mL. The detection limit in this assay is 10 pg/mL. Both tPA and tPA-PAL1 complexes are detected with equal efficiency by this ELISA." Rat tPA antigen was measured by ELISA using rabbit-antirat tPA IgG, as described.17 vWF antigen was mea- sured by ELISA, essentially as described," with the following minor modifications: the rabbit antihuman vWF IgG was diluted 1:2,000, and 0.1% (wt/vol) of casein was used in all buffers instead of 0.05% (wt/vol) bovine serum albumin. Human pooled plasma in a range of 0.078% to 1.25% was used for calibration. Lactate dehydrogenase activity was determined using a kit, according to the manufacturer's (Sigma) instructions.

Fibrin zymography." Samples were put immediately in sample buffer containing 2% sodium dodecyl sulfate (SDS) and were frozen at -20°C. SDS/9% polyacrylamide slab gels were prepared ac- cording to Laemmli.zo Gels were then washed, placed on top of a plasminogen-rich fibrin layer, and incubated for 36 hours at 37°C as we described in detail elsewhere.''

Datu presentation and units used. The data are presented as mean 2 SD of three measurements. Human tPA antigen is given as picograms or nanograms per milliliter, using Activase as standard. Rat tPA antigen is given as nanograms per milliliter, using rat L, tPA" as standard. vWF antigen is given as units per milliliter, 100 U being defined as the amount of vWF antigen present in 1 mL of pooled human plasma.

RESULTS

The regulated (acute) release of tPA in vitro. The consti- tutive and regulated secretion of tPA was studied in vitro using first-passage human endothelial cells. In a representa- tive experiment (Fig lA), control cells constitutively se- creted 13 pg of tPA per 2 x lo5 cells into the medium over 10 minutes (1.3 pg tPA/min/2 X lo5 cells). This rate of constitutive tPA secretion resembled the rate of constitutive tPA secretion during the 30-minute preincubation period (1.6 pg tPA/min/2 X l Os cells) and during the 2 hours preceeding the experimental period (2.2 pg tPA/min/2 X lo5 cells). These data suggested that no major changes in constitutive secretion occurred because of manipulation of the cells dur- ing an experiment. On addition of human thrombin (1 NIH U/mL), the cells very rapidly released tPA into the medium. The highest increase of tPA occurred during the first minute, resulting in an additional 25 pg tPA/min/2 X lo5 cells in the medium. The tPA content of the extracts from thrombin- treated cells was always lower than the content of the extracts from control cells (see Fig IA). This difference in tPA con- tent between extracts from thrombin-treated cells and control cells was of the same magnitude as the amount released into the medium. In five experimental series, the difference averaged 94% ? 30% (mean ? SD) of the amount of tPA acutely released by thrombin into the medium at 1 minute.

Because, in vivo, acute release of vWF and tPA are often observed together, the secretion of v W F was measured in this experiment as well. vWF was, similar to tPA, constitutively secreted by control cells, 0.033 U110 min/2 X lo5 cells (see Fig 1B). On addition of thrombin (1 NIH U/mL), vWF was released slightly slower into the medium than tPA. The in- crease in vWF was maximal at 3 minutes, at which time an additional 0.22 U/mL of vWF had been released into the medium. Figure 1B also shows that the increase of vWF in the medium resulted in a corresponding decrease of vWF in

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

250

200

150

100

50

0

I ,

0 2 4 6 8 10

time (minutes)

2.5 1 I r

l . - l 1 .o

0.0 0 2 4 6 8 10

t ime (minutes)

Fig 1. Time course of the induction of the acute release of tPA (A) and vWF (B) by thrombin. First-passage HUVECs were treated with 1 NIH U/mL of human a-thrombin, or with control M199/HSA me- dium after a preincubation period on M199/HSA medium in which 48 pg tPA/3O min/2 x l o 5 cells and 143 U vWFl30 min/2 x lo5 cells were secreted constitutively. Media and cell extracts of control cells and thrombin-treated cells were collected. In (A), the tPA concentra- tion (pg/2 x l o 5 cells) in the conditioned media is shown for control (A) and thrombin-treated cells (0 ) . Also shown are tPA concentra- tions in the corresponding cell extracts of control (A) and thrombin- treated cells (0). The difference in tPA concentrations between con- trol and thrombin-treated cell extracts is approximately equal to the amount of tPA acutely released. In this experiment, the tPA concen- trations in the cell extracts at t = 0 are low as compared with the other time points. Similarly low values have been observed more often at random time points and are not understood. In (B), the vWF concentration (expressed as U/2 x l o5 cells) is shown for the same experiment using the same symbols as in (A). All data shown are mean k SD In = 31.

VAN DEN EIJNDEN-SCHRAUWEN ET AL

l 2 3 4 5

Fig 2. Zymography of plasminogen activator (PA)-activity in endo- thelial cells and conditioned medium after thrombin-induced release of tPA. The acute release of tPA was induced by 1 NIH U/mL of human thrombin for 3 minutes. The plasminogen activator activity in the conditioned media and cell extracts was visualized by fibrin zymography after SDS/polyacrylamide-gel electrophoresis. Lane 1, recombinant tPA (Activase); lane 2, conditioned medium of control cells; lane 3, conditioned medium of thrombin-treated cells; lane 4, cell extract of control cells (diluted 1:7); and lane 5, cell extract of thrombin-treated cells (diluted 1:4). The lower bands in lanes 2 to 5 correspond with the band of free, uncomplexed recombinant tPA (lane l) , which is denoted by the arrow (+). The upper bands in lanes 2 to 5 represent tPA complexed to PAI-1, which is denoted by the arrowhead (F). All PA-activity could be quenched by antibodies against human tPA (data not shown).

cell extracts. Because the acute release of both tPA and vWF has reached maximal values at 3 minutes, in subsequent experiments, the acute release of tPA and vWF was always determined at 3 minutes. The release of tPA and vWF was not caused by cell lysis, because the lactate dehydrogenase content of media was not increased by thrombin and always remained below 3% of the amount present in the cells.

To study whether free tPA or tPA-plasminogen activator inhibitor (PAI) complex was acutely released, samples of media and cell extracts were analyzed by fibrin zymography. Despite the presence of an approximately fivefold excess of PAI-I in these media, complex formation is prevented by a 1: 1 dilution in an SDS-containing buffer of the media, immediately after a 3-minute stimulation. Figure 2 shows that only free tPA, but not tPA:inhibitor complex, was acutely released into the medium on addition of thrombin (Fig 2, lanes 2 and 3), and that free tPA had disappeared from the cell extracts (Fig 2, lanes 4 and 5). Please note that tPA in cell extracts has the same mobility (ie, the same M,) as constitutively secreted and acutely released tPA in media (see Fig 2, lanes 2 through 5) .

Dose-dependence of tPA release by thrombin. Figure 3 shows that thrombin increased the amount of tPA (and vWF) released in a dose-dependent manner. In this experiment, the constitutive secretion of tPA was 60 pg/2 X I O 5 cells in 30 minutes. In the presence of I NIH U/mL of thrombin, maxi- mal acute release of 40 pg/2 X IOs cells was observed at 3 minutes. From 0.01 NIH U/mL upwards, the acute release exceeded the constitutive secretion by at least twice the stan- dard deviation. Similar data were obtained for vWF (Fig 3).

To see if cells could release still more tPA on further

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

STUDIES ON THE ACUTE RELEASE OF TPA 351 3

125 n

Fig 3. Thrombin dose-depen- dently induces the acute release of tPA and vWF. Concentrations of 0 to 3 NIH U/mL of thrombin were used to induce acute re- lease of tPA and vWF. The amount of tPA (0; pg/2 x lo5 cells) that was acutely released into the medium by the indi- cated concentration of thrombin and the amount of acutely re- leased vWF (A; U/2 x I O 5 cells) are shown.

40 -

30 -

20 -

10 -

0 0.01 0.1 1 10

thrombin (NIH U/mL)

25 9 0

0

stimulation after maximal thrombin stimulation, cells treated with thrombin (1 NIH U/mL for 3 minutes) were subse- quently incubated with calcium ionophore A23 187 (1 pmoV L) for another 7 minutes. Cells treated with thrombin fol- lowed by ionophore did not release more tPA than did cells treated only with thrombin. The doubly stimulated cells re- leased 114% ? 44% (n = 6; n.s.) of tPA, similar to the amount released by cells treated only with thrombin. Doubly stimulated cells released in 10 minutes 148% ? 73% (n = 6) of vWF, as compared with cells treated for 10 minutes with thrombin only.

Constitutive tPA secretion and acute release of tPA in various cell isolates. To investigate whether constitutive tPA secretion (representing tPA synthesis) influenced acute tPA release in HUVECs, a correlation study was performed with data from 17 isolates of HUVECs. These 17 isolates provided naturally occuring variable levels of constitutive tPA secretion (from 2 to 10 ng/mL/24 h). By linear regres- sion analysis, a correlation between constitutive tPA secre- tion and acute release of tPA of 0.68 was found (Fig 4). The percentage of tPA in the cell extract that could be released averaged 33% with an SD of 17%.

Constitutive tPA secretion and acute release of tPA: De- creased constitutive secretion of tPA. When HUVECs were cultured in deficient medium, the levels of the three tPA parameters constitutive secretion, acute release, and cellular concentration (expressed as percentages of the level at t = 0) significantly decreased in parallel in time to about 20% after 24 hours (Fig 5; the absolute values of these parameters are given in the legend to the figure). These parallel changes suggest that constitutive secretion, acute release, and cellular content of tPA are closely linked by some, as yet undeter- mined, physiological process(es). At 24 hours, the rate of overall protein synthesis did not differ from that in control cells (127% +- 18%, n = 6), showing that the viability of cells cultured in deficient medium had not diminished. Fur- thermore, vWF release did not differ between cells cultured in deficient medium and cells cultured in total medium (data not shown). These data suggest that cells in deficient medium

were as sensitive to thrombin stimulation as were cells cul- tured in total medium.

Constitutive tPA secretion and acute release of tPA: In- creased constitutive secretion of tPA. HUVECs were treated with 0 to 3 mmoK sodium butyrate or with 1 pmoV L RA to increase tPA synthesi~.'~.'' After 24 hours, acute release of tPA was induced. The constitutive secretion of tPA was enhanced dose-dependently by sodium butyrate, showing a 20-fold stimulation at 3 mmol/L (Fig 6). These

250 I T

W

0 0 100 200 300

constitutive tPA secretion (pg/30 min/2.1 O5 cells)

Fig 4. tPA synthesis and thrombin-induced acute release of tPA in 17 isolates of HUVECs. In 17 isolates of HUVECs, the acute release of tPA was induced by 1 NIH U/mL of human a-thrombin, as de- scribed in Materials and Methods. The mean constitutive tPA secre- tion (during 30 minutes) and the mean tPA release 13 minutea after thrombin-addition) are plotted; the horizontal and vertical bars repre- sent the corresponding SDI In = 3). After linear regression, a correla- tion of 0.68 was found (n = 17).

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

361 4 VAN DEN EIJNDEN-SCHRAUWEN ET AL

l00

80

60

40

20

increases in constitutive tPA secretion were parallelled by increased concentrations of tPA in the cell extracts and by increased acute release of tPA (Fig 6), again suggesting that the processes involved are closely linked. RA enhanced the constitutive secretion twofold, which resulted in a l . 1 -fold increase of tPA in the cell extracts and in a 1.6-fold increase of acute release of tPA. Increasing the constitutive tPA secre- tion did not influence the time profile of acute release, whereas, in most experiments using thrombin stimulation followed by ionophore, all available tPA was released by thrombin (data not shown). Sodium butyrate and RA did not change constitutive vWF secretion, the amount of vWF in cell extracts, or the acute release of vWF.

An acutely releasable tPA pool in HUVECs. The close correlation between constitutive tPA secretion and acutely

0 I , , , , , , releasable tPA, described above, might suggest that acutely

releasable tPA is derived from the same pathway as constitu- 0 4 8 12 16 20 24 tively secreted tPA. If so, the induction of acute release of

tPA should cause a decrease in constitutive secretion over the subsequent time period. This was studied as follows: at

hours "deficient" medium t = 0, acute release was induced with thrombin; after 3

Fig 5. The effects of a decreased tPA synthesis on acute release minutes, hirudin (0.67 pg/mL) was added to the cells to

of tPA. First-passage HUVECs were cultured for 0, 2, 4, 8, and 24 inactivate thrombin; and the media were collected at 0,3, 10, hours in deficient medium (without ECGF, HS, or heparin). Acute 60, and 120 minutes after the addition of thrombin. Figure 7 release was induced with 1 NIH U/mL of human thrombin. The abso- shows that thrombin-treated cells had released 17 pg tPN2 lute amounts of tPA constitutively secreted, acutely released, and x 105 cells at t = 3 minutes, and that this difference persisted present in cell at = Or shown mean ' = 3)' were for the next 2 hours. These data suggested that the releasable 47 f 4, 59 k 9, and 240 i: 0 (pg/2 x 10' cells), respectively. The constitutive tPA secretion (m), the acute release of tPA (AI, and the pool was not part Of the constitutive tPA pathway. In a first amount of tPA in the cell extracts (e) are expressed as percentages attempt to discriminate between tPA following the constitu- of the concentrations at t = 0 hours.

4.50

3.00

1 S O

0.00 0 0.3 1 3

180

120

60

0

T

i

0 30 60 90 120

time (minutes)

sodium butyrate (mmol/L) Fig 7. The secretion of tPA over a longer period. At t = 0. acute release was induced as described above, and the reaction was

Fig 6. Sodium butyrate increases tPA synthesis and tPA release. stopped at t = 3 minutes with hirudin. The concentration of tPA First-passage HUVECs were cultured for 24 hours with various con- was maasured after 0,3,10,60, and 120 minutes. tPA is continously centrations of sodium butyrate. Acute release of tPA was then in- secreted into the medium (A), and, on addkion of thrombin, addi- duced as described in Materlab and Methods. The constitutive tPA tional tPA is secreted into the medium (e), which is still found in the Secretion (A), the acute release of tPA (e), and the amount of tPA in medium after 80 and 120 mlnutes. Data shown are mean f SD (n = the call extracts (m) are shown as mean f SD (n = 3). 3).

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

STUDIES ON THE ACUTE RELEASE OF TPA 3515

n 0 II

.W

E .-

100

75

50

25

0

a T

II-0

0 75

2 25 Q ) o

A

Q 0.0 0.5 1 .O

hours - cycloheximide

0 1 2 3 4 hours unlabeled methionine

Fig 8. The residence time of tPA in the tPA storage pool. HUVECs that had been cultured for 24 hours in total M199 medium containing 3 mmol/L sodium butyrate were incubated for 4 hours in total methi- onine-free M199 medium containing 20 p C i radiolabeled "S-methio- nine and 3 mmol/L sodium butyrate. Meanwhile, the cells were chased for 4, 3, 2, 1, or 0 hours with a 100-fold excess of unlabeled methionine. Release was then induced with 1 NIH UlmL of human a-thrombin, end the media were collected. After overnight incuba- tion of the media at 4°C in anti-tPA-coated wells of a microtiter plate, the wells were washed, and the bound protein was dissolved, neutralized, and counted. The data of three independent experiments are expressed as percentages o f t = 0 (shown as mean t SD), con- cerning constitutive tPA secretion (A) and acute release of tPA (0). In three independent experiments, the absolute counts found in media containing constitutively secreted tPA at t = 0 were 143, 297, and 196 disintegrations per minute (dpm)/3O mid2 x lo6 cells. In three independent experiments, the absolute counts found in media con- taining acutely released tPA at t = 0 were 194, 157, and 47 dpm/3 m id2 x lo5 cells. Aspecific binding was determined by adding media or cell extracts to microtiter plates that were coated with anti-FITC antibodies. This aspecific binding was about 6%. as compared with that for anti-tPA-coated microtiter plates. The inset shows the con- stitutive secretion of tPA (A) and acute release of tPA (0) after inhibi- tion of the protein synthesis with 5 pg/mL cycloheximide in total M199 for 0, 0.5, or 1 hours. Acute release of tPA was induced by 2 pmol/L calcium ionophore A23187 in the presence of 5 pg/mL cycloheximide, because calcium ionophore is a receptor-independent stimulant that is as effective as thrombin but is not dependent on protein synthesis. Data shown are mean f SD (n = 3).

tive or regulated pathway, we analyzed the retention time of tPA. To this end, HUVECs that had been stimulated for 24 hours with butyrate were metabolically labeled with 35S- methionine, chased for 0 to 4 hours with an excess of unla- beled methionine, and then induced to release tPA. Con- stitutive secretion and acute release of labeled tPA were determined by overnight binding of radiolabeled material in the media to anti-tPA-coated microtiter wells. These data (Fig 8) are expressed as percentages of the amount of radiola- bel bound at t = 0 (ie, unchased). Background binding to anti-FITC coated wells was about 6% for unchased media. Binding of radiolabeled tPA could not be determined in cell extracts, because background binding on anti-FITC-coated plates was too high to permit reliable determination of spe-

cific binding (data not shown). As shown in Fig 8, after a l-hour chase with unlabeled methionine, the amount of radioactivity bound from media of unstimulated cells was less than 50% relative to the amount bound from unchased medium of unstimulated cells. In media from thrombin-stim- ulated cells, the amount of radiolabel additionally released by thrombin was still 100% relative to the amount addition- ally released from unchased cells. After 2 hours, both path- ways contained about 25% of the amount of label present in unchased media. Therefore, these data suggest that the releasable tPA pool is separate from the constitutive secre- tion pathway. The inset of Fig 8 shows that comparable data are obtained when protein synthesis is inhibited with cycloheximide (overall protein synthesis, as measured by incorporation of radiolabeled methionine, was blocked for 90% after treatment of the cells for 1 hour with cyclohexi- mide). The constitutive secretion of tPA antigen was de- creased by 57% after l hour. In contrast, the acutely released amount of tPA antigen was still 114%, supporting the evi- dence for a separate intracellular source for tPA. Blocking of protein synthesis for more than 1 hour severely reduced the cellular adenosine triphosphate content (data not shown); thus, experiments were not continued beyond 1 hour.

Acute release of tPA in vivo. To study the relation be- tween constitutive secretion and acute release of tPA in vivo, rats were pretreated for 2 days with cholera toxin or for 5 days with RA to increase tPA synthesis. After 2 days of cholera toxin treatment, P A mRNA levels in heart and lung were significantly increased as were the tPA antigen concen- trations, showing increased tPA synthesis in these animals (data not shown). Treatment of rats with RA also resulted in increased levels of tPA activity and tPA antigen in several tissues.'6 Treatment with cholera toxin increased the steady- state plasma levels of tPA 5.4-fold as compared with that for control animals (ie, t = 0 in Fig 9); treatment with RA increased tPA plasma levels 1.4-fold (Fig 9). The acute re- lease of tPA was then induced in these animals by bradyki- nin. As shown in Fig 9, the acute release of tPA had, at 1 minute, increased 1 .%fold in RA-treated rats and 3.7-fold in cholera toxin-treated animals, as compared with that for con- trol animals. These data indicate that an increased tPA syn- thesis gives rise to an increased acute release of tPA in vivo as well.

DISCUSSION

In this study, we addressed the questions of how acute release of tPA from human endothelial cells proceeds and how this release is related to the constitutive secretion of tPA. Using first-passage HUVECs, we observed, as have many other^,^ constitutive secretion of tPA into the medium at a steady rate over many hours. On the addition of throm- bin, acute release occurred, as evidenced by the rapid appear- ance of uncomplexed, enzymatically-active tPA in the me- dium. The acute release of tPA was very rapid, showing a maximal increase of tPA within 1 minute. In both these respects, the acute release reaction in vitro resembled acute release of tPA as observed in humans in vivo: Release in- duced by thrombin was dose-dependent. At maximal stimu- lation by thrombin (2 l NIH U/mL), most if not all endothe- lial tPA was released into the medium, because a second

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

3516 VAN DEN EIJNDEN-SCHRAUWEN ET AL

plexed to an inhibitor (Fig 2) . Because tPA in the subcellular matrix and tPA on the plasma membrane are complexed to PAL 1 ,23 the tPA released must be derived from an intracellu- lar compartment (see also Fig 2, lane 4), in line with pre- viously published data,” which showed that plasma mem- brane-bound tPA is not released. Moreover, on subcellular fra~tionation,’~ the tPA content of an intracellular high-den- sity fraction, free of plasma membrane markers, is dimin- ished after stimulation with thrombin (Schrauwen et al“’ and Emeis and van den Eijnden-Schrauwen, manuscript in prepa- T ration). Therefore, we propose that, after synthesis, tPA is

-M temporarily stored in an intracellular compartment, presum- ably consisting of high-density granules. In all these respects, 1 the pathway followed by tPA destined for acute secretion

storage pool is less permanent than pools of other stored proteins.

thesis of tPA and the amount of acutely releasable tPA. This suggests that a fixed percentage of newly synthethized tPA

-*- *“f resembles that of other secretory proteins,’.25 although the l

0 2 4 6 8 10 We observed a close correlation between the rate of syn-

time (minutes) goes into the tPA storage compartment, regardless of the

Fig 9. Acute release of tPA in vivo. Rets were pretreated for 2 days with cholera toxin (m) or daily for 5 days with RA (0) to increase their tPA synthesis. Acute release of tPA was then induced in pre- treated rata and in control rata (AI as described in Materials and Methods. The concentration of tPA was measured in citrated plasma at 0, 1, 2, 5, 7, and 10 minutes after the injection of bradykinin. The highest concentration of tPA is found at 1 minute. After 1 minute, the tPA Concentration decreases because of clearance by the liver. The vertical bars represent SDs of the mean (n = 8).

stimulation by calcium-ionophore did not result in signifi- cantly more tPA release. Concomitantly with tPA, vWF was also released, as has often, but not always, been observed in vivo.”,” Compared with tPA, vWF was released slightly more slowly, showing a maximum at 3 minutes. The time course of the acute release of tPA and vWF in vitro also resembled the time course of acute release in the perfused rat hindleg, which also showed very rapid release of tPA and a slightly slower release of vWF (as shown by Tranquille and Erneis’’).

The amount of thrombin-releasable tPA strongly corre- lated with its rate of constitutive secretion (ie, tPA synthesis) in HUVECs (Figs 4 through 6) . A similar correlation was found in vivo in rats (Fig 9). Because the constitutive secre- tion of tPA (tPA synthesis) was coupled, at least quantita- tively, so closely with the release of tPA, the question arose how closely the two pathways are coupled functionally. The following evidence suggests that both pathways are not iden- tical. Thrombin-induced release of tPA does not diminish the subsequent constitutive secretion (Fig 7), indicating that the former is not a precursor of the latter. After metabolic labeling (Fig 8), released tPA retains label about 1 hour longer than constitutively secreted tPA. Also, after inhibition of protein synthesis by cycloheximide, release of tPA is maintained for at least 60 minutes at control level, whereas the level of constitutively secreted tPA decreases strongly. These observations suggest that releasable tPA was delayed in an additional compartment. This conclusion is supported by the observation that acutely released P A is not com-

rate of synthesis. This has also been described for vWF

tPA that is not acutely released from this pool into the medium still disappears from the pool, because the time tPA stays in the pool is only about 1 hour. What happens to this tPA is not known. We call this tPA that has disappeared “lost tPA,” because it is not secreted in the constitutive pathway (Fig 7) or in the acute release pathway. One possi- bility is that this lost tPA is deposited in the subcellular matrix.’* However, we have been unable to detect the re- quired amounts of tPA in the matrices of our cell cultures. Another possibility is that the lost tPA is degraded intracellu- larly, possibly in a lysosomal compartment, because label leaves the storage compartment but is not recovered in the constitutive pathway (Fig 8; compare also Fig 7). In an at- tempt to find further evidence for this hypothesis, we incu- bated cells for 4 hours in conditions that diminish lysosomal enzyme activity (50 pmol/L chloroquine). Under these con- ditions, the constitutive secretion did not change, whereas the amount of tPA in cell extracts did indeed increase. How- ever, this increase could account for only some 10% of the lost tPA (our unpublished observations).

The storage pool in HUVECs (residence time about 1 hour) was less stable than the tPA storage pool in vivo in rats, the latter pool being hardly diminished by inhibiting protein synthesis for 5 hours.” This might explain why endo- thelial cells in vitro possess a relatively small storage pool for tPA as compared with that for rats. In cell extracts from HUVECs about 0.1 ng of tPA/cm* is present, whereas 0.5 to 2 ng of tPA/cm2 has been estimated to be present in vivo (calculations in Emeis‘).

Acutely released tPA may play an important role in speed- ing up thrombolysis. Not only is the tPA thus released itself enzymatically active, but acute release will also result in a local tPA concentration much larger than the basal level of tPA and also than that of its inhibitor PAI-1. Thrombin, which is an effective stimulus for the induction of tPA re- lease (Fig 3), may be the major stimulus for tPA release, because it will reach high local concentrations near forming

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

STUDIES ON THE ACUTE RELEASE OF TPA 3517

or growing thrombi. Indeed, in vivo in primates, thrombin formation has been shown to induce acute release of large amounts of tPA.30 In quantitative terms, acute secretion of 50 pg tPA/cmz (as in Fig 6) will result, in a capillary with a diameter of 6 pm, in a local concentration of 340 ng tPA/ mL. If P A synthesis is increased (Fig 6), up to 900 pg tPA/ cm2 can be released, resulting in a local concentration of 6 pg tPA/mL in that same capillary (compare with Schrauwen et aI3). This concentration is even higher than the plasma tPA concentration reached during thrombolytic therapy. Moreover, tPA present before or during coagulation will be much more potent in dissolving a thrombus than tPA present only after thrombus formation has occurred,‘,’ as is the case during thrombolytic therapy.

The results of this study show that in human endothelial cells tPA can be acutely released from an intracellular pool, and that it should be possible to enhance this capacity for acute tPA release by increasing tPA synthesis. Therefore, in combination with the possibility” to induce the release of tPA without inducing the release of vWF, it should be possi- ble to increase local tPA concentrations specifically and to speed up fibrinolysis/thrombolysis when required.

REFERENCES 1. Brommer EJP: The level of extrinsic plasminogen activator (t-

PA) during clotting as a determinant of the rate of fibrinolysis; inefficiency of activators added afterwards. Thromb Res 34:109, I984

2. Fox KAA, Robison AK, Knabb RM, Rosamond TL, Sobel BE, Bergmann SR: Prevention of coronary thrombosis with subthrom- bolytic doses of tissue-type plasminogen activator. Circulation 72:1346, 1985

3. Schrauwen Y, de Vries REM, Kooistra T, Emeis JJ: Acute release of tissue-type plasminogen activator (PA) from the endothe- lium; regulatory mechanisms and therapeutic target. Fibrinolysis 8:8,

4. Chandler WL, Levy WC, Veith RC, Stratton JR: A kinetic model of the circulatory regulation of tissue plasminogen activator during exercise, epinephrine infusion, and endurance training. Blood 81:3293, 1993

5. Kooistra T, Schrauwen Y, Arts J, Emeis JJ: Regulation of endothelial cell t-PA synthesis and release. Int J Hematol 59:233, 1994

6. Emeis JJ: Regulation of the acute release of tissue-type plas- minogen activator from the endothelium by coagulation activation products. Ann NY Acad Sci 667:249, 1992

7. Booth F, Marshall JM, Cederholm-Williams SA: Effect of bra- dykinin on the exposure of tissue plasminogen activator from human endothelial cells. Fibrinolysis 2: 107, 1988

8. Booyse FM, Bruce R, Dolenak D, Grover M, Casey LC: Rapid release and deactivation of plasminogen activators in human endo- thelial cell cultures in the presence of thrombin and ionophore A23187. Semin Thromb Hemost 12:228, 1986

9. Wagner DD: Cell biology of von Willebrand factor. Annu Rev Cell Biol 6:217, 1990

10. Schrauwen Y , Emeis JJ, Kooistra T: A sensitive ELISA for human tissue-type plasminogen activator applicable to the study of acute release from cultured human endothelial cells. Thromb Haemost 7 1 :225, 1994

1 1 . Padr6 T, van den Hoogen CM, Emeis JJ: Distribution of

1994 (suppl 2)

tissue-type plasminogen activator (antigen and activity) in rat tissues. Blood Coagul Fibrinolysis l:601, 1990

12. Tranquille N, Emeis JJ: The simultaneous acute release of tissue-type plasminogen activator and von Willebrand factor in the perfused rat hindleg region. Thromb Haemost 63:454, 1990

13. Jaffe EA, Nachman RL, Becker CG, Minick CR: Culture of human endothelial cells derived from umbilical cord veins: Identifi- cation by morphological and immunologic criteria. J Clin Invest 52:2745, 1973

14. Van Hinsbergh VWM, Bertina RM, Van Wijngaarden A, Van Tilburg NH, Emeis JJ, Haverkate F: Activated protein C decreases plasminogen activator-inhibitor activity in endothelial cell-condi- tioned medium. Blood 65:444, 1985

15. Kooistra T, Berg J van den, Tons A, Platenburg G, Rijken DC, Berg E van den: Butyrate stimulates tissue-type plasminogen- activator synthesis in cultured human endothelial cells. Biochem J 247:605, 1987

16. Bennekum A van, Emeis JJ, Kooistra T, Hendriks HFJ: Mod- ulation of tissue-type plasminogen activator by retinoids in rat plasma and tissues. Am J Physiol 264:R931, 1993

17. Emeis JJ, Hoekzema R, de Vos A F Inhibiting interleukin- 1 and tumour necrosis factor-a does not reduce the induction of plasminogen activator inhibitor type- 1 by endotoxin in rats in vivo. Blood 85:115, 1995

18. Giles AR: Guidelines for the use of animals in biomedical research. Thromb Haemost 58:1078, 1987

19. Granelli-Piperno A, Reich E: A study of proteases and prote- ase-inhibitor complexes in biological fluid. J Exp Med 148:223, 1978

20. Laemmli UK: Cleavage of stmctural proteins during the as- sembly of the head of the bacteriophage T4. Nature 227:680, 1970

21. Medh RD. Santell L, Levin EG: Stimulation of tissue plasmin- ogen activator production by retinoic acid: Synergistic effect on protein kinase C-mediated activation. Blood 80:981, 1992

22. Smalley DM, Fitzgerald JE, O’Rourke J: Adenosine diphos- phate stimulates the endothelial release of tissue-type plasminogen activator but not von Willebrand factor from isolated-perfused rat hind limbs. Thromb Haemost 70:1043, 1993

23. Kooistra T, Sprengers ED, Van Hinsbergh VWM: Rapid inac- tivation of the plasminogen-activator inhibitor upon secretion from cultured human endothelial cells. Biochem J 239:497, 1986

24. Emeis JJ, van den Hoogen CM, Schrauwen Y: An endothelial storage granule for tissue-type plasminogen activator. Thromb Haemost 69:609, 1993 (abstr)

25. Kelly FU3: Pathways of protein secretion in eukaryotes. Sci- ence 230:25, 1985

26. Reinders JH, Vervoom RC, Verweij CL, van Mourik JA, de Groot PG: Perturbation of cultured human vascular endothelial cells by phorbol ester or thrombin alters the cellular von Willebrand factor distribution. J Cell Physiol 133:79, 1987

27. Mayadas T, Wagner DD, Simpson PJ: von Willebrand factor biosynthesis and partitioning between constitutive and regulated pathways of secretion after thrombin stimulation. Blood 73:706, 1989

28. Korner G, Bjornsson TD, Vlodavsky I: Extracellular matrix produced by cultured corneal and aortic endothelial cells contains active tissue-type and urokinase-type plasminogen activators. J Cell Physiol 154456, 1993

29. Tranquille N, Emeis JJ: Protein synthesis inhibition by cyclo- heximide does not affect the acute release of tissue-type plasminogen activator. Thromb Haemost 61:442, 1989

30. Giles AR, Nesheim ME, Herring SW, Hoogendoom H, Stump DC, Heldebrant CM: The fibrinolytic potential of the normal primate following the generation of thrombin in vivo. Thromb Haemost 63:476, 1990

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom

1995 85: 3510-3517  

Y van den Eijnden-Schrauwen, T Kooistra, RE de Vries and JJ Emeis a dynamic storage poolfrom human endothelial cells in vitro and in rats in vivo: evidence for Studies on the acute release of tissue-type plasminogen activator 

http://www.bloodjournal.org/content/85/12/3510.full.htmlUpdated information and services can be found at:

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

  Copyright 2011 by The American Society of Hematology; all rights reserved.Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American

For personal use only.on December 31, 2018. by guest www.bloodjournal.orgFrom