1. No category

From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
1996 87: 5061-5073
Differential mechanisms targeting type 1 plasminogen activator
inhibitor and vitronectin into the storage granules of a human
megakaryocytic cell line
SA Hill, SG Shaughnessy, P Joshua, J Ribau, RC Austin and TJ Podor
Updated information and services can be found at:
http://www.bloodjournal.org/content/87/12/5061.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
Differential Mechanisms Targeting Type 1 Plasminogen Activator Inhibitor
and Vitronectin Into the Storage Granules of a Human Megakaryocytic
Cell Line
By Stephen A. Hill, Stephen G. Shaughnessy, Penny Joshua, Jose Ribau, Richard C. Austin, and Thomas J. Podor
Type 1 plasminogen activator inhibitor (PAI-1) and itscofactor vitronectin(Vn) are stored within the
a-granules of platelets. The two possible sources for their biosynthetic origin
are endogenous synthesis in megakaryocytes or endocytosis from plasma. Using ultrastructural and confocal laser
scanning microscopic (CLSM) image analysis, we observed
that treatment of Dami cells, a human megakaryocytic cell
line, with phorbol myristateacetate (PMA) induces the accumulation of PAL1 and Vn in intracellular storage vacuoles
that contain other
a-granule proteins such as von Willebrand
factor. To examine evidence for biosynthesis of PAL1 and
Vn by Dami cells, we immunoprecipitated PAL1 and Vn from
the conditionedmedia of cells biosynthetically radiolabeled
with ’%-methionine in the presence or absence of PMA. In
contrast t o Hep G2 cells, which synthesize both PAL1 and
Vn, only ’5S-PAI-1 was recovered from PMA-treated Dami
cells. Reverse transcription-PCR analysis of RNA extracted
from resting and
PMA-treated Damicells confirmed thatPAI1 mRNA expression was detectable a t l o wlevels in resting
cells and induced by PMA treatment. In contrast, Vn mRNA
was notdetected. We examined binding and internalization
(endocytosis) of PAL1 and Vn by Dami cells using biotinylated analogs (b-PAL1 and b-Vn). Flow cytometry analysis
indicated that the bindingof b-Vn t o Dami cells was dosedependent, saturable, and specific for multimeric forms of
Vn. Cells were incubated at 4°C or 37°C and endocytosis of
b-Vn was shown by probing
electrophoretically fractionated
cell lysates with ‘251-labeled streptavidin. Only cells incubated at 37°C internalized b-Vn. CLSM image analysis confirmed that the b-Vn was
internalized and thatit colocalized
with PAL1 in storage granules. The binding of b-Vn t o cells
was inhibited by the presence of PAI-1, and there was no
evidence of specific b-PAI-1 binding or uptake t o resting or
PMA-treated cells. Thesedata suggest that accumulation of
PAI-1 in Dami cell storage granules is due t o endogenous
synthesis and that theaccumulation of Vn is
due t o endocytosis of serum-derived Vn.
0 1996 by The American Societyof Hematology.
P
ticware, RPMI medium, a-minimal essential medium (a-MEM),
methionine-free RPMI, calf and goat serum, Hank’s Buffered Salt
Solution (HBSS), trypsin, penicillin, streptomycin, streptavidin-conjugated Texas Red, and goat antimouse IgG-conjugated fluorescein
isothiocyanate (FITC) were obtained from GIBCO BRLLife Technologies Inc (Burlington, Ontario, Canada). Bovine skin gelatin,
PMA, leupeptin, soybean trypsin inhibitor, phenylmethylsulfonyl
fluoride, N-ethyl maleimide, benzamidine, aprotinin, casein, Triton
X-100, caprylic acid, normal goat IgG, normal rabbit IgG, and isotype-matched, nonspecific mouse IgG were obtained from Sigma
Chemical CO (St Louis MO). The monoclonal antibody (MoAb)
MAI-12 IgG was obtained from Biopool (Burlington, Ontario, Canada). Purified IgG fractions of monoclonal and polyclonal antibodies
were prepared by caprylic acid precipitation.” Antisera to human
PAL1 or human Vn were raised in rabbits, and the IgG fractions
further purified by affinity chromatography on immobilized PAI-I
or Vn, respectively.” The polyclonal antibodies to PAI-I andVn
were shown to be monospecific by immunoblot analysis of human
LATELETSMAYinfluencethrombolysisduetothe
presence of type 1 plasminogen activatorinhibitor (PAI1) and vitronectin (Vn)
in their a-granules.’ PAI-1 is the major
physiologic regulatorof fibrinolysis. It binds to and inactivates
tissue- and urokinase-type plasminogen activator and thus inhibits the generation of plasmin. PAI-1 circulates bound to
Vn and is stabilized by this interaction.’.’ Moreover, recent
studies indicate that platelet-associated PAI-1 is largely responsible for inhibiting the lysis of platelet-rich thrombi.6”’
There is also evidence that PAL1 is released from activated
platelets in associationwithVn.’ Thus, Vn associated with
platelet a-granules may be an important mechanismfor modulating platelet-associated PAI-1 activity.
Although there is evidence that PAI-1 is synthesized by
megakaryocytes and megakaryocyte-type cell lines,l3.l4the
biosynthetic origin of Vn in a-granules is unknown. Three
possible mechanisms exist for the targeting of proteins to
the platelet a-granule, including (1) endogenous synthesis
within the precursor megakaryocyte, ( 2 ) receptor-dependent
endocytosis, or (3) receptor-independent uptake by platelets
or megakaryocytes.” The purpose of this study wasto examine the biosynthetic origin of PAI-I and Vn in Dami cells,
a permanent line of cultured megakaryocytes isolated from
a human megakaryoblastic leukemia.I6 These cells can be
induced to differentiate down the megakaryocytic lineage by
treatment with the tumor promoter phorbol-myristate-acetate
(PMA) and thus can be used to examine the biochemical
events of megakaryocyte differentiation.
In this report, we present evidence consistent with the
hypothesis that platelet Vn arises from transport into the agranule from the surrounding medium, whereas PAI-l is
synthesized by the precursor megakaryocyte and deposited
in the a-granule before the platelet is shed.
MATERIALS AND METHODS
Chemicals,proteins, and reagents. All chemicals were of the
highest analytic grade commercially available. Tissue culture plas-
Blood, Vol87, No 12 (June 15). 1996: pp 5061-5073
From the Department of Pathology, McMaster University, and
the Hamilton Civic Hospitals Research Centre, Hamilton, Ontario,
Canada.
Submitted June 21, 1995; accepted February 9, 1996.
Supported by an operating grant to T.J.P. from the Heart and
Stroke Foundation of Ontario and Medical Research Council and
to S.A.H.from the Hamilton Civic Hospitals Research Fund. T.J.P.
is a Career Investigator, S.G.S. a Postdoctoral Research Fellow of
the Heart and Stroke Foundation of Ontario, andR. C.A. aResearch
Scholar of the R.K. Fraser Foundation.
Address reprint requests to Thomas J. Podor,PhD, Hamilton
Civic Hospitals Research Centre,71I Concession St, Hamilton, Ontario, Canada, L8V 1C3.
The publication costsof this article were defrayedin part by page
chargepayment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 sorely to
indicate this fact.
0 1996 by The American Society of Hematology.
0006-4971/96/8712-000$3.00/0
5061
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
5062
HEP G2 hepatoma cell lysates. These antibodies against human Vn
cross-react with similaraviditytobovine
Vn, as determined by
Western blot andenzyme-linkedimmunosorbentassay
(ELBA)
analysis. Recombinant human PAI-I , purified from Escherichia coli
lysates,was kindly provided by DrDanLawrence(University
of
Michigan,
Ann
Arbor,
Antibodies
to
von Willebrand
factor
(vWF) wereobtainedfromSigmaImmunochemicals.
Protein A
sepharose beads were obtained from Pharmacia Biotech (Montreal,
Quebec, Canada). Permafluor was obtained from Biomeda (Foster,
CA). Polymerase chain reaction (PCR) primers were synthesized by
the Institute for Molecular Biology, McMaster University (Hamilton,
Ontario, Canada). Reverse transcription-PCR (RT-PCR)reagent kits
were obtained from Perkin Elmer Cetus (Rexdale, Ontario, Canada).
of
PAI-I cDNA'" was a gift of DrDavidGinsburg(University
Michigan). GRGDSP and GRGESP peptides and Vn cDNA" were
obtained from Telios Pharmaceuticals (San Diego, CA).
Isolation of Vn from human plusma. Vn was isolated from human plasma by one of two methods, including (1) denaturing urea
and heparin-Sepharose affinity chromatography22 and (2) nondenaturing heparin Sepharose and immunoaffinity chromatography. For
immunoaffinity chromatography, human plasma(50 mL) was equilibrated with 20 mmol/L Tris, 0.15 m o m NaCI, pH 7.4, and passed
over a 15-mL heparin Sepharose precolumn tandemly coupled to an
affinity-purified sheep anti-Vn IgG Affi-gel affinity column. Vn was
eluted under nondenaturing conditions using Gentle Antibody Elution buffer (Pierce Biochemical, Rockford, IL). Native Vn was homogeneous for the 75-kD, single-chain and a two-chain form.
To
prepare multimeric Vn, native Vn preparations were treated with 6
mol/L urea for 1 hourfollowed by extensive dialysis." Vn was
analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)24 and native-PAGE and quantitated using
both
Micro BCA protein assays (Pierce Chemical) andVn-specific competitive
ELISAs
using
polyclonal
and conformation-sensitive
MoAbs to human Vn. MoAb 8E6 was obtained from Boehringer
Mannheim (Montreal, Quebec, Canada). MoAbs 153 and 1244 were
kindly provided by Dr D. Seiffert.*'
Biotinylation of puri$ed proteins. Purified proteins (PAI- 1 and
Vn) were biotinylated with biotinyl-t-amino caproic-N-hydroxysuccinimide ester (Boehringer Mannheim) following the manufacturer's
instructions. Briefly, 100 pg of protein in phosphate-buffered saline
(PBS)wasadjustedtopH
7.7 with bicarbonatebufferand
then
incubated with 10 p g biotinylation reagent in dimethylformamide
(25mg/mLjfor2hoursatroomtemperature.Labelingmixtures
were dialyzed against several volumes of PBS, pH 7.4. and aliquots
werestoredat-70°C
until used. Thebioactivity of biotinylated
PAI-I(b-PAI-I)wasdetermined
by measuringitsabilityto
bind
to Vn or tissue-type plasminogen activator (t-PA)-coated 96-well
plates.IxBoundb-PAI-l
was detecteddirectly with streptavidinconjugated alkaline phosphatase or indirectly with rabbit anti-PAII and goat antirabbit IgG-conjugated alkaline phosphatase followed
by pNPP substrate, and the rate of change in the absorbance at 405
nm was quantitated using a microtiter plate reader (EL 340; Bio-tek
Instruments Inc. Highland Park, VT). Biotinylation of PAI-I did not
or PAI-l-mediated inhibition of tblock PA[-I binding to t-PAPA activity, as assessedusingthechromogenicsubstrateS-2251
(Chromogenix; Helana Laboratories, Mississauga, Ontario, Canada).
Biotinylationhadnoeffect
on thebindingofPAI-1to
affinityIgG. Biotinylation had
purified anti-PAI-l IgG or to MoAb MAI-12
no observable effect on Vn-dependent adhesion of cultured human
umbilical vein endothelial cellsto coated microtiterwells'" or trypsin
sensitivity. Biotinylation of Vn did not interfere with its binding to
or MoAbs 153
immobilized PAI-I, polyclonal antibodies against Vn,
and 1244; however, binding to MoAb 8E6 was relatively decreased.
Similarly. relative to unlabeled multimeric Vn, binding of b-Vn to
HILL ET AL
heparin-sepharose was reduced in proportion to the decrease in the
MoAb 8E6 binding. Analysis of the multimeric forms of b-Vn by
denaturing and nondenaturing gel electrophoresis indicated
that it
containedgreateramounts
of lowermolecular weight native V n
forms as compared with unlabeled controls.
Cell culture. Dami cells,I6ahumanmegakaryocytic
cell line
(courtesy of Dr Sheryl Greenberg, Boston, MA), were obtained from
AmericanTypeCultureCollection(CRL9792).
The cells were
maintained in RPM1 mediumcontaining 5% fetalbovineserum
(FBS) and 50 pg/mL penicillin and streptomycin. For some experiments, the cells were incubated
with 5 nmol/L PMA for up to 24
hours before harvesting. The
human hepatoma cell line, Hep G2,
was maintained in a-MEM containing 10%FBS." Human umbilical
vein endothelial cells were isolated and cultured as previously described."
Flow cytometry and ligand binding studies. Cells were harvested
in PBS, pH 7.4, containing 2.5 mmol/L EDTA and resuspended at
2 X IO6 per mL in HBSS containing Ca2+andMg'+. Cells ( 5 X IO4
cells in 50 pL) were exposed to various concentrations of b-PAI-I
or b-Vn for I hour at 4°C. At the end of the incubation period. the
cellswerewashed, fixed in 1%formaldehyde in HBSS,and then
incubated with FITC-conjugated streptavidin (final concentration, 20
&mL) for 15 minutes. Cell-bound fluorescence (l0,OOO cells) was
assessed using a Coulter Epics I1 flow cytometer (Coulter, Hialeah,
FL) with the following settings: sample volume, 150 pL; flow rate,
50 pL/min; sheath pressure, 7.50 psi; argon laser power, 15 mW;
fluorescence, 1 PMT voltage, 9.25 V; fluorescence l gain, 7.5.
Vn incorporation into Dumi cells. Cells(2.5 X 10' in 50 pL)
were harvested as described above and incubated for various lengths
of time at either4"C
or 37°C with 4 pg/mL h-Vn and then centrifuged
to separate the bound from the free
b-Vn. The pellets were resuspended in 500 pg/mL trypsin for 2 minutes at room temperature to
digest any surface-associated b-Vn and the
trypsin inactivated by
the addition of 750 FL of ice-cold HBSS containing 20% FBS. Flow
cytometry analysis,as described above, confirmedthat this treatment
removed greater than 95% of the b-Vn from cells incubated in the
presence of the ligand at 4°C. The cells were again pelleted,
the
supernatant was removed, and the cell pellet was
frozen at -70°C
until analyzed.Thepelletswerethawed
in 1 X Laemmlisample
buffercontainingP-mercaptoethanolandfractionated
by SDSPAGE" under reducing conditions and the proteins were electroblotted onto nitroceliulose paper. The filters were blocked with PBS
containing 5% fat-free milk and 0.5% Triton X-100, probed with
0.25 pCi/mL "'I-streptavidin, washed, and exposed to x-ray film at
-70°C. The resulting autoradiographic bands at 65 kD were examined using a scanning laser densitometer. Band density is expressed
as absorbance units times millimeter. To assess specific uptake, the
density of the band from cells incubatedat 4°C was subtracted from
that of the cells incubated at 37°C.
Immunoc)ltochemi,~tryand image unulysis. Cells were grown on
gelatin-coated coverslips in the presence5 n m o m PMA for thefinal
24 hours. Cells grown in theabsence of PMAdid not adhere to
coverslips and were therefore harvested and deposited on slides by
cytocentrifugation (Shandon Instruments, Artmoor, UK). The coverslips were washed in PBS, pH 7.4, fixed with cold 3% formaldehyde in PBS for 5 minutes, washed with PBS, neutralized for 1 0
minutes in 0.1 nmol/L glycine, permeabilized with 0.025% Triton
X-100 in PBS for 5 minutes, and blocked for 30 minutes with PBS/
0.5% bovine serumalbumin(BSA)containing
50 pg/mL normal
goat globulin (NGG). Primary antibodies used were an MoAb antihuman PAI-I (MAI-I2), a rabbit and sheep polyclonal antihuman
Vn, and anaffinity-purified rabbit polyclonal anti-vWF IgG. Controls
included staining cells with each primary antibody separately, staining without primary antibodies, or staining with nonspecific m o w
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
5063
PAI, VITRONECTIN, AND MEGAKARYOCYTES
or rabbit IgG. Primary antibodies were diluted in PBS/BSA/NGG
and incubated with the fixed cells for 1 hour at 37°C. The coverslips
were washed and stained with Texas Red rhodamine (TxR)-conjugated goat antimouse or FITC-conjugated goat antirabbit IgG diluted
1:20 in PBS/BSA/NGG for 1 hour at 37°C. After washing, the coverslips were mounted on glass slides using Permafluor mounting
medium.
For some studies, cells were viewed by video enhanced fluorescence microscopy using an epifluorescence microscope (Leica Laborlux S; Willowdale, Ontario, Canada) fitted with a 63X planapo
oil immersion lens (na = 1.4), a 100 W mercury lamp, and transmission filters for FITC (525 nm) and TxR (600 nm long pass filter).
Images were captured using a CCD video camera module electronically linked to a computer imaging software system (Northern Exposure; Empix Inc, Mississauga, Ontario, Canada).
To discriminate the relative spatial distribution of intracellular and
extracellular PAL1 and Vn, cultured Dami cells were analyzed by
dual-labeling immunofluorescence confocal laser scanning microscopy (CLSM) and subjected to 2-plane optical sectioning (200 nm/
section). Optical sectioning of cells was performed using a Zeiss
LSM 10 equipped with a 63X planapo oil immersion lens (numerical
aperture, 1.4). Dual-wavelength images were acquired using an
argon ion laser (488 nm excitation), a helium/neon ion laser (543
nm excitation), and two matched long pass barrier filters for FITC
(515 to 525 nm emission) and TxR (575 to 640 nm emission) images.
Image processing and three-dimensional volume rendering and rotations were performed on Metamorph software (Empix Imaging Inc).
Before analysis, all images were Z-plane line averaged and corrected
for dark current and instrument background. Cells stained with nonspecific primary antibodies or streptavidin-FITC alone were used to
establish background intensity.
To examine the uptake and intracellular location of b-Vn, cells
were grown on coverslips and were exposed to 4 pg/mL b-Vn for
6 hours at 37”C, washed, and not only immunologically stained
for PAI-1 as above, but also probed for internalized b-Vn with
streptavidin-FITC, and then analyzed by dual-labeling immunofluorescence confocal microscopy and 2-plane optical sectioning as described above.
For postembedding immunogold electron microscopy, resting or
PMA-treated (24 hours) Dami cells were processed inLR White
medium without postfixation by osmium tetr~xide.’~
Thin sections
were mounted on Formvar-coated, nickel-plated grides and incubated overnight with affinity-purified rabbit anti-PAI-l, anti-Vn, or
preimmune rabbit IgG. The sections were washedwith PBS and
incubated with gold-conjugated (10 nm) goat antirabbit IgG.All
sections were double-stained with uranyl acetate and lead citrate.
Specimens were examined with a Joel Biosystems (Montreal, Quebec, Canada) 1200 electron microscope at 75 kV.
Metabolic labeling and immunoprecipitation. Dami cells (confluent @-mm dishes) were metabolically labeled for 24 hours with
1 0 0 pCi/mL 35S methionine (NENmuPont Canada, Mississauga,
Ontario, Canada) in 3 mL of 90% methionine-reduced medium (4
mmolk), as previously des~ribed.~”,~’
To show the secretion of radiolabeled PAI-I or Vn, conditioned media were centrifuged to remove cell debris, precleared with protein A-Sepharose, andthen
incubated with affinity-purified rabbit anti-PAI-l or anti-Vn IgG
bound to protein A-Sepharose beads in immunoprecipitation buffer
containing 0.1 mmoliL iodoacetamide, 0.2 mmolfL phenylmethyl
sulfonyl fluoride (PMSF), and 5 nmol/L N-ethyl maleimide.**The
beads were washed to remove unbound proteins and the bound proteins were subjected to SDS-PAGE.Z4The gels were dried and exposed to x-ray film (Kodak XR; Eastman Kodak, Rochester, NY)
for autoradiography at -70°C.
RT-PCR and Southern blotanalysis. Total RNA was isolated
from either control or PMA-treated Dami cells and human hepatoma
Hep G2 cells using the method of Chomzynski and Sacchi.” Oligonucleotide primers were as follows. The PAI-I forward 5”CGGAGCACGGTCAAGCAAGTG-3’ and PAL 1 reverse S-GGTGAGGGCAGAGAGAGGCAC-3’ primers were used to amplify a 400-bp
fragment near the 3’ end of the PAI-l cDNA. Vn forward 5’-GAGCGGGACAGCTGGGAGGAC-3‘ and reverse 5”AGGAGCTGGGCAGCCCAGCC-3’ primers were used to amplify a 501-bp fragment
also in the 3’ end of the Vn cDNA. Primers for glyceraldehyde-3phosphate dehydrogenase (GAPDH; forward, 5‘-CCCATGGCAAA’ITCCATGGCA-3‘; reverse, 5”TCTAGACGGCAGGTCAGGTCGACC-3’) amplify a 600-bp fragment.33The reverse transcription
reaction to synthesize cDNA was performed using 1 pg of total
RNA, 1 mmolk reverse primer, 50 mmoVL KCl, IO mmolk Tris
HCI (pH 8.3), 5 mmol/L MgCI2, 1 mmolk deoxynucleotides, l U/
RNAse inhibitor, and 2.5 U/pL reverse transcriptase in a final volume of 20 pL.One cycle of 42°C for 15 minutes, 99°C for 5 minutes,
and5°C for 5 minutes was performed. PCR amplification, using
cDNA from the RT reaction, was performed in 50 mmol/L KCl, 10
mmol/L Tris HCI (pH 8.3), 2 mmol/L MgC12, 0.2 mmoVL deoxynucleotides, 0.15 mmol/L primers, and 2.5 U Taq polymerase in a
final volume of 100 pL. All samples were subjected to amplification
in a Perkin-Elmer Cetus Thermal Cycler (model 480) to 95°C for 5
minutes, then to 35 cycles of94°C for 1 minute and60°C for 1
minute, and then to a final incubation of 60°C for 7 minutes. Amplified products were separated on 0.8% agarose TBE gels stained with
ethidium bromide and photographed. To confirm the identity of the
amplicons, RT-PCR products from the control and PMA-treated
Dami cells and the Hep G2 cells amplified with PAL1 and Vn
primers were transferred to Zeta-Probe GT blotting membranes (BioRad, Mississauga, Ontario, Canada) and incubated with 32P-labeled
cDNA for PAL1 or Vn that had been radiolabeled using the random
primer method.34After overnight hybridization at 42°C. the membranes were washed twice in 2 x SSC/O.I% SDS, followed by two
washes at 55°C for 15 minutes in 0.1 x SSC/O.l% SDS. The membranes were then wrapped in plastic wrap and exposed to Kodak Xmat AP film at -70°C for I to 3 days.
RESULTS
Immunostaining of PAI-I and Vn. We have previously
shown that PMA-induced differentiation of Dami cells results inthe de novo synthesis and colocalization of vWF
and multimerin within the extensive network of precursor
platelet a-granule-like storage vacuoles.30 To determine
whether PMA also influences the expression of PAL1 and
Vn within the vWF-positive storage granules, we performed
dual-labeling immunofluorescence CLSM image analysis on
fixed, permeabilized Dami cells that were preincubated for
24 hours in the presence or absence of PMA (Fig 1). Dami
cells were cultured on coverslips and incubated in the final
24 hours in the presence or absence of PMA. The cells were
fixed, permeabilized and stained for PAX-l, Vn, and vWF
(Fig l). Untreated cells did not adhere to the coverslips and
were therefore prepared by cytocentrifugation before fixation
and staining (see Fig 3D). Figure 1A illustrates the differential contrast of adherent, PMA-treated Dami cells that were
stained for Vn (Fig 1B) and PAI-1 (Fig 1C). These pseudocolored images represent the composite of varying fluorescence from all optical planes and show that the greatest
intensity of stain (red > yellow > green > blue) for both
antigens is located in the large vacuolar structures that radiate
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
p-
=
Fig 1.
I
.c
\
Fig 7.
"
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
PAI, VITRONECTIN, AND MEGAKARYOCMES
5065
Fig 1. PIwdocolor enhanced images of PAL1 and Vn dMribution in Dami cells. Dami cells were grown on e o v d i p s in the presence or
absence of PMA (24 hours) and then fixed, permeabilized, andstained with an MoAb anti-PAL1 and polyclonal anti-Vn IgG, followed by FITCPAI-l andVn fluoresgoat-antimouseandTxR-goat-antirabbitIgG.Imageswerepseudocolored to enhance the relative intensities of
cence (white > red > yallow > green > blue > purple). (AI Digitally enhanced interferencecontrast image of adherent PMA-treatedcells (24
hours). Note the elevated centerof the two cells indicating the position of the nudeus. (B)Vn is present in a large vacuoleoverlying the nuclei
(arrow) and to a lesser extent within smaller vesicles extending towards the periphery of the cell. Note the extensive staining of Vn on the
culture substratum (small arrow). (C) PAI-l is also present predominantly in the large central vacuole (arrow) and in tho smaller vesicles. (D)
Single optical section (200 nm) of an untreated Dami cell stained for PAL1 illustrating the absence of PAL1 in undifferentiatedcells. (E and F)
Single optical sections from a stack of images from PMA-treated Dami cells stained for vWF (E) and PAI-l (F) showing that PAI-l is also
colowlized within thevWF-positive storage vacuoles (arrows). (G and
H) PMA-treated cells werestained for Vn (G and G') and PAL1 (H and
H') and subjected to CLSM optical sectioning. The images displayedin G and H were single planes taken from the middle of the optical stack.
From these stacks,cross-sections (200 nm wide) were taken through the 2-plane (blueline) and rotated 90" to view the localization of Vn and
H', respectively). The three-dimensional colocalization of Vn and isPAL1
particularly
PAI-1 within intracellular storage vacuoles (arrows, G' and
evident within theupper (small arrows) and lower portion of the convoluted storage granulesin thecell. Scale bars = 25 pm.
from the central perinuclear region (large arrows) as well as
in discreet vesicles in the periphery of the cell. Bovine Vn
in the culture medium binds to the coverslips and coats the
substratum (purple; Fig lB, small arrow). To better define
the three-dimensional distribution of Vn, PAI-1, and vWF
within the cells, we used CLSM image analysis to generate
separate high-resolution, 200-nm thick single plane images
(N = 18) through the Z-plane of the cells. The results are
displayed as pseudocolor images from a single representative
plane (plane 11 of 18) midway through the perinuclear vacuolar structures seen in Fig 1A through C . Staining for PAI1 in untreated D& cells (Fig 1D) indicates only a faint
background fluorescence. Single-plane images of PMAtreated cells dual-labeled for vWF (Fig 1E) and PAI-1 (Fig
1F) show distinct, yet clearly overlapping images (arrows)
for the areas of most intense staining, thus identifying these
structures as a-granule-like vacuoles. Figure 1G and H illustrate the higher magnification of a pseudocolor image
through the perinuclear region from two PMA-treated cells
stained for Vn (Fig 1G) and PAI-1 (Fig 1H). Distinct yet
overlapping distributions of these two proteins can be seen
in the large storage granules. Ninety degree rotation of the
Z-plane cross-sections of the entire stack of optical sections
taken through the vertical line in Fig 1G and H are presented
in G' and H', respectively. Again, distinct, yet overlapping
distributions of Vn and PAI-1 can be seen (small arrows).
Cells stained with preimmune IgG were similar to nonPMA-treated cells stained for PM-1 (Fig ID, data not
shown). Staining for Vn in non-PMA-treated cells was localized in similar structures and equally as intense as that
for PMA-treated cells.
To further examine the subcellular location of PAI-1 and
Vn, Dami cells were treated in the presence or absence of
PMA for 24 hours, the cells were harvested and pelleted,
and the cell pellets prepared for postembedding immunogold
electron microscopy. Figure 2A shows a low-power view of
a PMA-treated D& cell stained for PAI-1 and shows the
extensive network of convoluted vacuoles in Dami cells, as
seen in Fig 1. The PAI-1 immunogold labeling in these
vacuoles (arrow) is localized to characteristic branching filamentous matrix structures. These can be seen in high magnificationin Fig 2D.As in Fig lD, PAI-l labeling is not
detected in untreated, control cells (Fig 2C). Vn is also present among the fibrillar matrix and was also among the amorphous proteinaceous aggregates localized within these
storage granules in both untreated (Fig 2E, arrows) and
PMA-treated cells (Fig 2F, arrows). PMA treatment of the
cells had no apparent effect on the pattern of the Vn staining.
Staining of PMA-treated Dami cells with preimmune rabbit
IgG did not show labeling in these vacuoles (Fig 2B).
Immunoprecipitationof biosynthetically radiolabeled proteins. To examine the metabolic origin of PAI-1 and Vn,
we next examined the evidence for endogenous synthesis of
these molecules. Dami cells, in the presence or absence of
5 nmol/L PMA, were incubated for 24 hours in the presence
of 35S-methionineand methionine-reduced medium. The culture supernatant was collected and immunoprecipitated with
either normal rabbit IgG, affinity-purified rabbit antihuman
PAI-1, or antihuman Vn IgG. Precipitated radiolabeled proteins were separated by SDS-PAGE and visualized by autoradiography (Fig 3). Even after prolonged exposure, radiolabeled Vn was not detected in the conditioned medium from
the basal or PMA-treated cells (lanes 2 and 5 ) . The results
from immunoprecipitation analysis of the Triton X-100 soh-
<
Fig 7. CLSM visualization of biotinylated Vn uptake by Dami cells. Dami cells
grown on coverslips werepretreated with PMA for 24 hours
and then incubated with 4 pglmL b-Vn for 6 hours at 37°C. At the end of the incubation period, the slips were washed, fixed, permeabilized
with Triton X-100. and stained with anti-PAl-l IgG followed by TxRgoat-antirabbit lgG (A andC) and with FITC-streptavidinto visualize the
distribution of b-Vn (B). Pseudocolor imagesindicatedthat PAI-l is localized in the storage vacuoles as in Fig 1. The red vertical line indicates
the position of the 2 line scan through 20 optical sections of two cells, which were then rotated 75" and viewed in V-2 plane cross-section
with the
culture substratum surface orientated down and the dome-shaped apicalcell surface upwards.The lower cell in (A; arrow) is transected
through the storage granuleswithin the attenuated portion of the cytoplesm adjacent to the nucleus andindicated b-Vn (B; whke arrow) is
localized within the storage vacuolescontaining PAL1 (C; black arrow). Similarly, in the upper cell, b-Vn is also localized within the transected
storage vacuoles lying above the nucleus. Scale bar = 25 pm.
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
5066
HILL ET AL
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
PAI, VITRONECTIN, AND MEGAKARYOCYTES
Control
Dami
l 2 3
5067
PMA Treated
Dami
Hep G2
4 5 6
7 0 9
205
116
W
B
Y
49
32
26
Fig 3. Immunoprecipitation of biosynthetically labeled proteins from Dami cells. Dami cells grown in the absence (lanes 1, 2, and 3) or
presence (lanes 4, 5, and 6) of PMA and Hep G2 cells (lanes 7, 8, and 9) were biosynthetically radiolabeled for 24 hours with %hnethionine
and the conditionedmedia was collected. Radiolabeled proteins were isolatedby immunoprecipitation with anti-Vn (lanes 2, 5, and 81, antiPAI-1 (lanes 3, 6, and 91, or preimmune (lanes 1, 4, and 7) IgG and visualized by SDS-PAGE and autoradiography. No radiolabeled Vn was
observed from either untreated orPMA-treated cells. Radiolabeled PAI-1 (arrowhead) wasdetected from PMA-treated Dami cells (lane 4) and
detected only in Hep G2 cells (lane 8). No bands were visiblein the lanes immunoprecipitated
Hep G2 cells (lane 9). Radiolabeled Vn (arrow) was
with preimmune lgG (lanes 1, 4, and 7).
ble cell lysates and Triton X-l00 insoluble extracellular matrix extracts showed no evidence of "S-labeled Vn in these
cell compartments (data not shown). "S-labeled PAI-I is
apparent in the culture supernatant recovered from PMAtreated cells (lane 6 ) . but not from the untreated cells (lane
3). No bands were visible in the culture supernatant immunoprecipitated with normal rabbit IgG (lanes 1, 4, and 7). Culture supernatant from Hep G2 cells contained both radiolabeled Vn (lane 8) and PAI-I (lane 9) and serves as a positive
control for the experiment.
RT-PCR analysis. A second approach was used to examine the evidence for endogenous synthesis of PAI-I and Vn
at the RNA level. Total RNA was isolated from cells that
were incubated in the presence or absence of PMA for 24
hours, and the RNA was subjected to reverse transcription
and PCR amplification using primers specific for PAI- I and
Vn. Vn mRNA was not present in either control or PMAtreated Dami cell extracts, but was present in the Hep G2
cell extracts (Fig 4A, lanes 6 , 7, and 8). However, PAI-I
transcripts were present in untreated cell extracts. Treatment
of the cells with PMA for 24 hours before harvest resulted
in an increase in PAI-I transcripts, as evidenced by an increase in the band intensity of the amplicon (Fig 4A, lanes
2 and 3). Equal loading of RNA was confirmed by the equal
signal of all three cell extracts for GAPDH (Fig 4A, lanes
IO. 11, and 12). Southern blot analysis usingradiolabeled
PAI-I and Vn cDNA probes confirmed the identity of the
amplicons (Fig 4B).
Ligand binding and uptake studies. Because endogenous
expressions and synthesis ofVn could notbe shown, we
examined the evidence for an exogenous origin of Vn (and
PAI-I) from plasma. Dami cell suspensions were incubated
for 1 hour at 4°C in the presence of various concentrations
of native or multimeric b-Vn followed by RTC-conjugated
streptavidin and then subjected to flow cytometry analysis.
As shown in Fig 5, multimeric b-Vn bound to the Dami cells
in a dose-dependent, saturable manner, withhalf-maximal
binding observed at approximately 1 pg/mL. Figure 5 (inset)
shows the effect of increasing concentrations of unlabeled
Vn on the binding of saturating amounts of b-Vn to Dami
cells. Approximately 58% of the b-Vn binding can be inhibited by excess unlabeled ligand. Treatment of the cells with
5 nmol/L PMA for 24 hours before harvesting did not significantly increase the binding of b-Vn to the cells. Table 1
illustrates the effects of various agents on the binding of bVn to Dami cells. Binding was inhibited approximately 90%
by the addition of 10% FBS. RGE-or RGD-containing peptides (up to 1,000-fold excess) had no effect on theb-Vn
<
Fig 2. lmmunogold labeling of PAI-1 and Vn in basal and PMA-treated Damicells. Untreated (C and E) Or PMA-treated (A, D, and F) cells
Were stained for PAI-1 (A and D) or Vn(Eand F) using affinity-purified polyclonal antibodies or preimmuneW (B) and goat-antirabbt gold10. (A) LOW power image (scale bar = 1 pm). PAL1 is present in PMA-treated Dami cells (N marks the nucleus). (B through F) Higher
magnification (scale bar = 200 nm). (B) No staining is evident with preimmune IgG. PAI-1 is not Present in untreated cells (c), but Can be
visualized after the induction of differentiation with
PMA (D; arrow). Vn ispresent in approximately equal amounts in both untreated(E) and
PMA-treated (F) Dami cells.
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
HILL ET AL
5068
A
PAM
Vn GAPDH
112 3 4 51 6 7 8 9110111213114
8
8
2
0
2
L
cn
0
-J
c
m
2
Fig 5. Flow cytometry measurement of biotinylated Vn binding
t o Dami cells. Cells were harvested and incubated at 4°C in thepresence of various concentrations of b-Vn in the presence or abscence
of 10-fold excess unlabeled Vn. The cells were fixed and incubated
with excess FITC-streptavidin and the cell-associated fluorescence
was measured by flow cytometry. Each point represents the mean
specific fluorescence (total fluorescence minus the fluorescence in
the presence of 10-fold excess unlabeled ligand) of 10,000 cells 122
SD). (Inset) Cells were incubated with 4 p g l m L b-Vn and various
concentrations ofunlabeled Vn and the cell-associated fluorescence
was measured as described above.
Fig 4.
RT-PCR amplification
and Southern blot analysis of Dami
cell RNA. (A)Total RNA fromuntreatedDami cells (lanes 2, 6, and
101, PMA-treated Dami cells (lanes 3, 7, and l l ) , Hep G2 cells (lanes
4.8. and 12). or negative controls containing no RNA (lanes 5,9, and
13) were reverse-transcribed and PCR-amplified using primers for a
401-bpfragment of PAL1cDNA llanes2through 5). a 501-bpfragment
of vn c~~~ (lanes
6 through g), or a 6 0 0 - b ~fragment of GAPDH
cDNA (lanes 10 through 13). Molecular weight markers are in lanes
1 (123-bp ladder) and 14 (100-bp ladder). (B1 To confirm the identity
of the amplicons produced using the PAL1 and Vn primers, RT-PCR
products from lanes 2, 3, and 4 (PAI-1 primers) andlanes 6, 7, and 8
(Vn primers) were transferred t o nitrocellulose paper and probed
with radiolabeled cDNA for PAI-1 (left) and Vn (right), respectively.
Untreated Dami cells (lane 1); PMA-treated Dami cells (lane 2); and
Hep G2 cells (lane 3).
binding to thecells.Incubation
of the b-Vn with excess
unlabeled PAI-l beforetheiraddition
to the cells resulted
in a 90% reduction Of b-Vn binding to the ce'1 surface'
Experiments using native b-Vn Or b-PA1-l showed no significant binding to Dami cells.
TO determinewhether fluid-phase b-Vn is internalized,
Dami cells were incubated with b-Vn (4 pg/mL) for various
lengths of time at either 4°C or 37°C. After trypsin treatment
to remove cell surface-associated b-Vn. the cell lysates were
fractionated by SDS-PAGE under reducing conditions. transferred to nitrocellulose membranes, and probed with '"Istreptavidin. Incubation of cells with b-Vn at 37°C resulted
in a time-dependent internalization, as shown by the increase
in the 65- and 75-kD Vn bands (Fig 6A). Cells incubated at
4°C incorporated significantly less b-Vn. To better illustrate
Table 1. Inhibition of Bindingand Uptake
of b-Vn byVarious Agents
Treatment
96 Binding
% Uptake
Control
10% serum
200 pmol/L RGD peptide
200 pmol/L RGE peptide
20 pglmL heparin
100
10
100
100
100
10
100
l?
100
100
46
100
300 n m o lPA!
/~
Dami cells
were incubated in the presence of 4 ,,g/mL
b-Vn as in
the competitive agent. Thecell surface binding was determined using
flow cytornetry as described and the uptake of b-Vn was assessed by
determining the net band intensity of '251-streptavidin-probed transfers of trypsin-treated Dami cell homogenates as described in the
Materials and Methods.
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
5069
PAI, VITRONECTIN, AND MEGAKARYOCYTES
u u u u
B
2h
lh
8h
4h
level of the PAI-Istaining vacuolewithintheattenuated
cytoplasm of the lower cell (whitearrow) andthrougha
PAI-l staining vacuole that overlays the nucleus of the upper
transected cell. Rotation of the resulting images through 75"
shows both the endogenous PAI-I (Fig 7A and C. arrows)
and the exogenously added b-Vn (Fig 7B. arrow) to colocalize within the storage granules.
DISCUSSION
8 -
6-
4-
2-
0
1
2
4
8
Time (hrs)
Fig 6. Uptake of biotinylated Vn by Dami cells. (A) Cells were
harvested and incubated at 4°C or 37°C in the presence of 4 pg/mL
b-Vn for 1.2, 4, or 8 hours. At the end of the incubation period, the
cells were treated with trypsin, homogenized, separated by electrophoresis, transfered to nitrocellulosepaper,and probed with "'lstreptavidin. The filters were exposed to x-ray film at -70°C. (B) Net
uptake of b-Vn. The 65-kD band of the b-Vn doublet was scanned
using a laser densitometer and the band intensity (absorbance units
x mm) of the 4°C lane was subtracted from that of the 37°C lane to
obtain net band intensity.
the specific time-dependent uptake of b-Vn. the autoradiograph bands of the 65-kD Vn were quantitated at each time
point by laserdensiometry and theintensity of the corresponding band at 4°C was then subtracted from that at 37°C
(Fig 6B). Table 1 shows the effect of numerous competitive
agents on the uptake of b-Vn. As with the binding of b-Vn.
excess unlabeledVn (data not shown) and FBS inhibited
uptake. Similarily, neither the RGD nor its inactive analog
RGE peptides inhibited uptake. Moreover, a IO-fold excess
of standard heparin was unable to inhibit the binding of Vn
to the cell surface, but did inhibit the uptake. Conversely, a
IO-fold excess of PAI-I was able to inhibit the binding of
Vn to the cell surface, but had little effect on the uptake of
Vn by these cells.
The uptake and intracellular targeting
of b-Vn by Dami
cells was visualized by immunofluorescence CLSM image
analysis(Fig7). PMA-treatedDami cells wereincubated
with b-Vn for 6 hours, fixed, and not only immunostained
for PAI-I (Fig 7A), as in Fig I , but also probed with streptavidin-FITC to localize internalized b-Vn. To better visualize
intracellular PAI- I and b-Vn, we performed a Z-plane crosssectionthroughthe image stacks (Fig 7A, red line) at the
Active PAI- I circulates in human plasma as a binary complex with Vn.'"' Under normal conditions and durin,0 acutephase states in humans and rats, plasma Vn levels arein 100to 1,000-fold molar excess relative to PAI-I levels.'x~25~26
Moreover, we have shown that increases in plasma PAI-I
and Vn after various acute-phase inflammatory stimuli in
rats are associated with increased PAI-I and Vn mRNA
expressionwithindifferenthepatocellularsources
that are
not coordinatelyregulated by interleukin-6 (1L-6).'x.25.2fiIn
the present report, we again show divergent origins of PAI1 and Vn found within the storage granules of PMA-treated
Dami cells. We have shown that PAI-I accumulates only in
cells that have been induced to differentiate by treatment
with the tumor promoter PMA. whereas Vn ispresent in
both treated and untreated cells. PAI-l is endogenously synthesized, whereas Vnisinternalized from thesurrounding
medium.
Some molecules arise in the a-granule as a result of endogenoussynthesis within the precursormegakaryocyte.
Proteins known to arise in this manner include multimerin,2"
vWF," platelet factor 4." and transforming growth factorKonkle et all3 have shown that PAI-I mRNA is synthesized upon activation in CHRF-288 cells. but not in HEL
cells. Both ofthesecelllinesaremegakaryoblasticlines
derivedfrom human tumors. The state of maturity of the
megakaryocytic phenotype and the potential for differentiation of various cell lines may affect their ability
to synthesize
PAI- I . Allessi et al'' used in situ hybridization to show PAI1 mRNA in MEG-01cells, a megakaryocyticline at an
earlystage of development, as well as in normalhuman
megakaryocytes. Recent studies4" have shown that a mouse
pituitary cell line capable of sorting proteins for secretion
into secretory granules. when transfected with human PAI1 cDNAconstructs,can
synthesize.package.andsecrete
PAI-I through secretory granules.The experiments presented
here suggest that PAI-I also arises as a result of endogenous
synthesis. PAL l mRNA expression,protein translation, sorting to storage granules, and secretion into the medium can
be induced by treatment with PMA. This suggests that the
synthesis and storageof PAI-I does not occur in early megakaryocyteprecursors but begins as the cells differentiate
down the lineage to mature megakaryocytes.
Greater than 90% ofthecirculating PAI-I is contained
within platelets." This suggests that, in addition to perturbed
endothelium,25the bone marrow is a majorsiteofPAI-I
gene expression and that biosynthesis of PAI-I is a normal
component of the hemopoietic differentiation of megakaryocytes. There is evidence that certain a-granule proteins arise
from at least two sources. Megakaryocyte-associated factor
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
5070
HILL ET AL
V is both endogenously synthesized and taken up from the
uptake appears to be largely independent of RGD-binding
surrounding medium.42 PAL1 accumulationin Dami cells
sites on the cells. RGD-independent binding of multimeric
appear to occur predominantly
by endogenous synthesis. The
Vn by culturedhumanfibroblastshas
been previously
lack of b-PAL1 binding to the Dami cell surfaceitsor
uptake
shown; however, the uptakeof multimeric Vn was sensitive
by these cells suggests that there is
little or no receptorto inhibition by RGD-containing peptides andby antibodies
mediated endocytosis of PAI-l, but does not rule it out comdirected againsta&, but notby those againsta& .'l' These
pletely. Furthermore, our inability to show a significant upinvestigatorshavealsoshownthat
both thenativeand
take of the fluid-phase PAI-l into Dami cells supports the
multimeric form of Vn is bound to the cells, but only the
hypothesis that there are at least two major sources of intramultimeric form was internalized." Moreover, this uptake
vascular PAI- 1 within the body, the platelet-associated storwas inhibited by heparin and thrombospondin. Volker
et
age pool and de novo synthesis by endothelialcells. The
alh2 have shown
that porcine vascular endothelial cells interproportion that each
of these poolscontributes to the
circulatnalize multimel-ic (denatured) Vn andthatthis uptake is
ing fluid-phase PAL1 is unclear. Simpson et a14' report that
inhibited by heparin but not by RGD peptides. Thus, our
the plasma pool and the platelet-associated pool of PAI-l
experimental findings appear to be largely consistent with
vary independently in a number of disease states. This supthese findings in cultured porcine endothelium and human
ports thehypothesisthatPAI-I
circulatingintheplasma
fibroblasts.
may arise from the vascular endothelial or smooth muscle
There is indirect evidence from other cell sources to supcells, whereas platelet-associated PAI- 1may arise from megaport the contention that Vn is synthesized in megakaryocytes
karyocytes, and that this pool serves a hemostatic function
and deposited in platelets before they are released into the
at sites of hemostatic plug and platelet accumulation.
circulation. Platelets of patients with Glanzmann thrombasProteins can also arise in the a-granule as
aresultof
theniathatlackboth
the GPII&II, and the a,P3 integrin
transport from the surrounding plasma. Albumin, IgG, IgM, receptors containincreased amounts of Vn within the aandIgAenterthea-granulesas
a result of fluid-phase
granules, whereasthose that lackthe GPIIJIII, but have
endocyt~sis.~~
fibr~nectin,".~~
the
and on
normal
normal or increased amounts of a& contain
amounts of Vn." These findings cannot differentiate whether
other hand, are bound at the cell surface to specific recepmegakaryocytelplatelet Vn is synthesized endogenously and
tors and are subsequently translocated to the cell interior
transported out of the cell via an a&-mediated process or
via receptor mediated-endocytosis. The source of plateletinternalized by an RGD-independent process. We are unable
associated Vn is not clear. However, the integrin receptor
on the
to showthepresence of metabolicallylabeledVn
in the
molecules GPII&II, and, to a lesser extent, (Y,PY.~('
surface of platelets are capable of binding plasma ligands,
culture supernatant or cell extracts from either untreated or
PMA-treated cells. RT-PCRexperiments to identify Vn
suchasfibrinogen,fibronectin,and
Vn.'5.",52 Moreover,
there is evidence on the intracellular movement
of these
mRNA in untreated and PMA-treated cells were undertaken
integrin receptors and their ligands, both to and from the
to strengthen the data obtained from the metabolic labeling
a-granuleandthecell
surface.s3,s4 Wencel-Drake et als5
experiments. This lack of an identifiable mRNA or a "Slabeled protein in our experiments strongly suggests that Vn
haveshownthat a syntheticligandcontainingtheRGD
is not synthesized by Dami cells. Thus, our data are consisrecognition site is bound to and internalized by the GPIId
tent with divergent pathways of origin for Dami cell Vn and
111, receptor proteins on resting platelets. We have shown
PAI- 1 and that the hepatocyte is the most significant source
saturable and specific binding of b-Vn to the surface
of
of circulating Vn and the Vn expression in extrahepatic cell
Dami cells. We have also shown that b-Vn is internalized
types is
by these cells and is targeted to the a-granule-like storage
vacuoles. The fact that the RGD peptide is unable to inhibit The fact that Vn is known to bind and maintain PAI-1 in
its active conformation brings up the question as to whether
either binding or uptake suggests that RGD-dependent rea-granule. Recent
ceptors are not involved in either process
in these cells. The PAI-1/Vn complexes existwithinthe
studies' indicate concommitant release of PAL1 andVn from
urokinasereceptor(uPAR)hasrecentlybeenimplicated
aplatelets upon thrombin stimulation, suggesting that the
as a receptor for Vn.56 The fate of cell surface Vn/uPAR
granule is a common storage site. PAI-1/Vn complexes can
complexes remains unclear, but endocytosis is a distinct
be shown in the supernatant of activated platelet releasates.
of uPAR on Dami cells and its
po~siblity.~~ T
presence
he
possible rolein Vn uptake are currently under investigation. Our data indicate three-dimensional colocalization of these
two molecules within Dami cells,as evidencedby the intense
Cell surface binding of b-Vn is inhibited by the addition
dual-fluorescence staining of the large storage vacuoles
in
of PAI- I , but not by heparin. Conversely, the uptake of bthe center of the cell (Fig 1). The ultrastructural morphology
Vn is inhibited by heparin, but not by PAI-1. The findings
of Vn-labeled filaments in Dami cells resembles the 5- to
are suggestive of two processes. The cell adhesion domain
IO-nm diameter filamentsassociatedwithfibronectin-rich
of VnislocatedwithintheRGDsequence
motifin the
extracellularmatrix
of culturedbovine
aortic endotheamino terminus of the r n ~ l e c u l e . ~
One
~ ~ of
~ *two possible
Rotary shadow electron microscopic evaluation of
PAI-1bindingsequencesinVnisalsolocalized
in the
purified multimeric Vn showed it to be composed
of a heteramino-terminalsomatomedin B domainadjacenttothe
ogenous population of globular aggregates with an average
RGD domain." Preincubation of Vn with PAI-l inhibits
diameter between15 and 28 nm.23 Dual-labeledimmunogold
Vn binding to the Dami cell surface, but Vn binding and
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
PAI, VITRONECTIN, AND MEGAKARYOCYTES
electron microscopic examination of ultrathin sections (approximately 70 to 80 nm thickness) of bovine endothelial
cell extracellular matrices65.66and purified proteins showed
PAL1 immune complexes composed of gold (4nm) conjugated goat antimouse IgG bound to murine antibovine PAI1 IgG, colocalized within a maximal 60 nm radius ofVn
(10 nm gold) immune complexes. By comparison, our
CLSM image analysis showed a high degree of three-dimensional colocalization of Vn and PAL 1 within Dami cell storage granules at a 200 nm limit of resolutuion. However,
these observations do not prove whether PAI-1 and Vn exist
as a complex within these storage granules. Despite colocalization within the same organelle (ie, a-granules in platelets
and the storage vacuoles in Dami cells), a large proportion
of platelet-associated PAL1 is believed to be inactive.67Our
preliminary evidence indicates that a significant portion of
the PAI-1 secreted by activated Dami cells is active, because
it inhibits t-PA-mediated plasmin generation. Detergent extracts from platelets do not release PAI-1Nn complexes.'
This evidence suggests that PAL 1 N n complexes do not exist
to any great extent within the a-granule but may form in the
microenvironment of the platelet plug after release of the a granule contents from the activated platelet. Hep G2 cells
synthesize and secrete both PAL1 and Vn, but pulse-chase
studies6*indicate that they interact to form complexes only
after secretion from the cell. Similarly, cultured endothelium
and fibrosarcoma cells synthesize PAI-1 but not Vn, yet the
secreted PAL1 forms complexes with serum-derived Vn in
the culture substratum and extracellular m a t r i ~ . ~ 'If. ~PAI~.~~
1 N n complexes exist in the a-granule in situ, our data would
suggest that the complex does not form until the molecules
are brought together from distinct sources in the a-granule.
Further studies are currently ongoing to determine the role
of Dami cell-associated Vn with respect to PAL1 activity.
ACKNOWLEDGMENT
The authors thank Drs Jeff Weitz and Jack Hirsh for their helpful
discussions and Susan Luscombe for her preparation of the manuscript.
REFERENCES
1. Preissner KT, Holzhuter S, Justus C, Muller-Berghaus G: Iden-
tification and partial characterization of platelet vitronectin: Evidence
of complex formation with platelet-derived plasminogen activator
inhibitor-l. Blood 74:1989, 1989
2. Declerck PJ, De Mol M, Alessi M-C, Baudner S, Paques EP, Preissner KT, Muller-Berghaus G, Collen D: Purification and
characterization of a plasminogen activator inhibitor-l binding protein from human plasma. Identification as a multimeric form of S
protein (vitronectin). J Biol Chem 263:15454, 1988
3. Wiman B, Almquist A, Sigurdardottir 0, Lindahl, T: Plasminogen activator inhibitor-l (PAI) is boundto vitronectin in plasma.
FEBS Lett 242:125, 1988
4. Mimuro J, Loskutoff DJ: Purification of a protein from bovine
plasma that hinds to type-l plasminogen activator inhibitor and prevents its interaction with extracellular matrix-Evidence that the
protein is vitronectin. J Biol Chem 264:936, 1989
5. Salonen E-M, Vaheri A, Poll8nen J, Stephens R, Andreasen P,
Mayer M, Dano K, Gailit J, Ruoslahti E: Interaction of plasminogen
5071
activator inhibitor (PAI-l) with vitronectin. J Biol Chem 264:6339,
1989
6. Levi M, Biemond BJ, van Zonneveld A-J,ten Cate JW,
Pannekoek J: Inhibition of plasminogen activator inhibitor-l activity
results in promotion of endogenous thrombolysis and inhibition of
thrombus extension in models of experimental thrombosis. Circulation 85:305, 1992
7. Serizawa K, Urano T, Kozima Y, Takada Y, Takada A: The
potential role of platelet PAI- 1 in t-PA-mediated clot lysis of plateletrich plasma. Thromb Res 71:289, 1993
8. Braaten JV, Handt S, Jerome WG, Kirkpatrick J, Lewis JC,
Hantgan RR: Regulation of fibrinolysis by platelet-released plasminogen activator inhibitor -1: Light scattering and ultrastructural examination of lysis of a model platelet-fibrin thrombus. Blood 81: 1290,
1993
9. Robbie LA, Booth NA, Croll AM, Bennett B: The roles of aZantiplasmin and plasminogen activator inhibitor-l (PAI-l) inthe
inhibition of clot lysis. Thromb Haemost 70:301, 1993
10. Fay WP, Eitzman DT, Shapiro AD, Madison EL, Ginsburg
D: Platelets inhibit fibrinolysis in vitro by both plasminogen activator
inhibitor- 1-dependent and independent mechanisms. Blood 83:35 1,
1994
11. Shohet RV, Spitzer S, Madison EL, Bassel-Duby R, Gething
M-J, Sambrook IF: Inhibitor-resistant tissue-type plasminogen activator: An improved thrombolytic agent in vitro. Thromb Haemost
71:124, 1994
12. Stringer HAR, van Swieten P, Heijnen, HFG, Sixma JJ,
Pannekoek H: Plasminogen activator inhibitor-l released from activated platelets plays a key role in thrombolysis resistance. Arterioscler Thromb 14:1452, 1994
13. Konkle BA, Schick PK, He X, Lui R J , Mazur EM: Plasminogen activator inhibitor-l mRNA is expressed in platelets and megakaryocytes andthe megakaryoblastic cell line CHRF-288. Arterioscler Thromb 13:669, 1993
14. Alessi ML, Chomiki N, Berthier R, Schwietzer A, Fossat C,
Juhan-Vague I: Detection of plasminogen activator inhibitor-l (PAI1) mRNA in human megakaryocytes by
in
situ hybridization.
Thromb Haemost 72:93 1, 1994
15. George JN: Platelet immunoglobulin G: Its significance for
the evaluation of thrombocytopenia and for understanding the origin
of a-granule proteins. Blood 76:859, 1990
16. Greenberg SM, Rosenthal DS, Greeley TA, Tantravahi R,
Handin RI: Characterization of a new megakaryocytic cell line: The
Dami cell. Blood 72:1968, 1988
17. Hum BAL, Chantler SM: Production of reagent antibodies.
Methods Enzymol 70:104, 1980
18. Seiffert D, Podor TJ: Immunological detection of conformational changes of type-l plasminogen activator inhibitor associated
with activation. Fibrinolysis 5:225, 1991
19. Lawrence D, Strandberg L, Grundstrom T, Ny T: Purification
of active human plasminogen activator inhibitor-l from E. coli:
Comparison with natural and recombinant forms purified from eucaryotic cells. Eur J Biochem 186:523, 1989
20. Ginsburg D, Zeheb R, YangAY, Rafferty UM, Andreasen
PA, Nielsen L, Dan0 K, Lebo RV, Gelehrter TD: cDNA cloning of
human plasminogen activator inhibitor from endothelial cells. J Clin
Invest 78:1673, 1986
21. Suzuki S, Oldberg A, Hayman EG, Pierschbacker MD, Ruoslahti E: Complete amino acid sequence of human vitronectin decluced from cDNA. Similarity of cell attachment sites in vitronectin
and fibronectin. EMBO J 4:2519, 1985
22. Yatohgo T, Izumi M, Kashiwagi H, Hayashi M: Novel purification of vitronectin from human plasma by heparin affinity chromatography. Cell Struct Function 13:281, 1988
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
5072
23. Stockmann A, Hess S, Declerck P, Timpl R, Preissner KT:
Multimeric vitronectin. Identification and characterization of conformation-dependent self-association ofthe adhesive protein. J Biol
Chem 268:22874, 1993
24. Laemlli UK: Cleavage of structural proteins during assembly
of the head of bacteriophage T4. Nature 277:680, 1970
25. Seiffert D, Ciambroni G, Wagner NV, Binder BR, Loskutoff
DJ: The somatomedin B domain of vitronectin. J BiolChem
269:2659, 1994
26. Lafrenie RM, Podor TJ, Buchanan MR, Orr FW: Up-regulated
biosynthesis and expression of endothelial cell vitronectin receptor
enhances cancer cell adhesion. Cancer Res 52:2202, 1992
27. Aden DP, Fogel A, Plotkin S, Damjanov I, Knowles BB:
Controlled synthesis of HbsAg in a differentiated human liver carcinoma-derived cell line. Nature 282:615, 1979
28. Kassis J, Hirsh J, Podor TJ: Evidence that postoperative fibrinolytic shutdown is mediated by plasma factors that stimulate
endothelial cell type 1 plasminogen activator inhibitor biosynthesis.
Blood 80:1758, 1992
29. Newman GR: LR white embedding medium for colloidal gold
methods, in Hayat MA (ed): Colloidal Gold: Principles, Methods,
and Application, v01 2. London, UK, Academic, 1989, p 47
30. Hayward CPM, Bainton DF, Smith JW, Horsewood P, Stead
RH, Podor TJ, Warkentin TE, Kelton JG: Multimerin is found in
the a-granules of resting platelets and is synthesized by a megakaryocytic cell line. J Clin Invest 91:2630, 1993
3 l . Bonifacio JS: Metabolic labelling of cells, in Ausubel FM,
Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl
K (eds): Current Protocols in Molecular Biology, v01 2. New York,
NY, John Wiley and Sons, 1991, p 10.18.1
32. Chomzynski P, Sacchi N: Single-step method of RNA isolation by guanidinium thiocyanate-phenol-chloroformextraction. Anal
Biochem 162:156, 1987
33. Allen RW, Trach KA, Hoch JA: Identification of the 37 kDa
protein displaying a variable interaction with the erythroid membrane
Chem
as glyceraldehyde 3-phosphate dehydrogenase. J Biol
262:649, 1987
34. Feinberg AP, Vogelstein BV: A technique for radiolabelling
DNA restriction endonuclease fragments to high specific activity.
Anal Biochem 132:6, 1983
35. Podor TJ, Hirsh J, Gelehrter TD, Zeheb R, Torrey D, Guigoz
Y, Gauldie J: Type-l plasminogen activator inhibitor (PAI-l) is not
an acute phase reactant in rats: Lack of interleukin-6 and hepatocyte
dependent synthesis. J Irnmunol 150:225, 1993
36. Seiffert D, Geisterfer M, Gauldie J, Young E, Podor TJ:
Interleukin-6 stimulates vitronectin gene expression in vivo. J Immuno1 155:3180,1995
37. Ginsburg D, Konkle BA, Gill JC, Montgomery RR, Bockenstedt PL, Johnson TA, Yang AY: Molecular basis ofhumanvon
Willebrand disease: Analysis of platelet von Willebrand mRNA.
Proc Natl Acad Sci USA 86:3723, 1989
38. Sottile J, Mosher DF, Fullenweider J, George JN: Human
platelets contain mRNA transcripts for platelet factor IV and actin.
Thromb Haemost 62:1100, 1989
39. Fava RA, Casey TT, Wilcox J, Pelton RW, Moses HL, Nanney LB: Synthesis of transforming growth factor-p1 by megakaryocytes and its localization to megakaryocytes and platelet a-granules.
Blood 76: 1946, 1990
40. Gambau L, Schleef RR: Processing of type-l plasminogen
activator inhibitor (PAI-I) into the regulated secretory pathway. J
Biol Chem 269:3875, 1994
41. Booth NA, Simpson A J , Croll A, Bennett B, MacGregor IR:
Plasminogen activator inhibitor (PAI-1) in plasma and platelets. Br
J Haematol 70:327, 1988
HILL ET AL
42. Gewirtz AM, Shapiro C, Shen YM, Boyd R, Coleman RW:
Cellular and molecular regulation of factor V expression in human
megakaryocytes. J Cell Physiol 153:277, 1992
43. Simpson AJ, Booth NA, Moore NR, Bennett B: The platelet
and plasma pools of plasminogen activator inhibitor (PAI-I) vary
independently in disease. Br J Haematol 75:543, 1990
44. George JN, Saucerman S: Platelet IgG, IgA, IgM and albumin:
Correlation ofplateletandplasma
concentrations in normal subjects
and in patients with ITP or dysproteinemia. Blood 72:362, 1988
45. Gerrard JM, Phillips DR. Rao GHR.Plow EF, Wulz DA,
Ross R, Harker LA, White SG: Biochemical studies of two patients
with grey platelet syndrome: Selective deficiency of platelet alpha
granules. J Clin Invest 66:102, 1980
46. Mosher DF, Presciotta DM, Loftus JC, Albrect RM: Secreted
alpha proteins: The race for receptors, in George JN, Nurden AT,
Phillips DR (eds): Platelet Membrane Glycoproteins. New York,
NY, Plenum, 1988, p 171
47. Ginsberg MH, Forsyth J, Lightsey A, Chediak J, Plow EF:
Reduced surface expression and binding of fibronectin by thrombinstimulated thrombasthenia platelets. J Clin Invest 71:619,1983
48. Mosher DF: Fibronectin. Prog Hemost Thromb 5: 1 1 I , 1980
49. Coller BS, Cheresh DA, Asch E, Seligsohn U: Platelet
vitronectin receptor expression differentiates Iraqi-Jewish from
Arab patients with glanzmann thrombasthenia in Israel. Blood
77:75, 1991
50. Tomiyama Y, Tsubakio T, Piotrowicz RS, Kurata Y, Loftus
JC, Kunicki TJ: The Arg-Gly-Asp (RGD) recognition site of platelet
glycoprotein IIb-IIIa on nonactivated platelets is accessible for highaffinity macromolecules. Blood 79:2303, 1992
51. Thiagarajan P, KellyKL:
Exposure of binding sites for
vitronectin on platelets following stimulation. J Biol Chem263:3035,
1988
52. Asch E, Podack E: Vitronectin binds toactivated human platelets and plays a role in platelet aggregation. J Clin Invest 85:1372,
1990
53. Morgenstern E, Ruf A, Patscheke H: Transport of anti-glycoprotein IIbDIIa-antibodies into the alpha-granules of unstimulated
human blood platelets. Thromb Haemost 67: 121, 1992
54. Suzuki M, Kawakatsu T, Nagata H, Hamamoto K, Iwata K,
Ohga S, Nomura S, Kokawa T, Yasunaga K: Effects of injected
antibody against the platelet glycoprotein IIbDIIa complex on monkey platelet fibronectin. Thromb Haemost 67:578, 1992
55. Wencel-Drake JD, Frelinger AD In, Dieter MG, Lam SC-T:
Arg-Gly-Asp dependent occupancy of GP IibDIIa by applaggin: Evidence for internalization and cycling of a platelet integrin. Blood
81:62, 1993
W ,Rosenberg S,
56.Ying W, Waltz DA,RoaN,Drurnmond
Chapman HA: Identification of the urokinase receptor as anadhesion
receptor for vitronectin. J Biol Chem 269:32380, 1994
57. Seiffert D,Loskutoff D: Evidence that type 1 plasminogen
activator inhibitor binds to the Somatomedin B domain of vitronectin. J Biol Chem 266:2824, 1991
58. Morris CA, Underwood
PA,
Bean
PA,
Sheehan M,
Charlesworth JA: Relative topographyof biologically active domains of human vitronectin. J Biol Chem 269:23845, 1994
59. Sigurdardottir 0, Wiman B: Identification of a PM-I binding
site on vitronectin. Biochim Biophys Acta 1208:104, 1994
60. Panetti TS, McKeown-Longo PJ: The a& integrin receptor
regulates receptor-mediated endocytosis of vitronectin. J Biol Chem
268:11492, 1993
61, Panetti TS, McKeown-Longo PJ: Receptor-mediated endocytosis of vitronectin is regulated by its conformational state. J Biol
Chem 268:11988, 1993
62. Volker W, Hess S, Visher P, Preissner KT: Binding and pro-
From www.bloodjournal.org by guest on November 20, 2014. For personal use only.
PAI,VITRONECTIN, AND MEGAKARYOCYTES
cessing of multimeric vitronectin by vascular endothelial cells. J
Histochem Cytochem 41:1823, 1993
63. Coller BS, Seligsohn U, West SM, Scudder LE, Norton W:
Platelet fibrinogen and vitronectin in glanzmann thrombasthenia:
Evidence consistent with specific roles for glycoprotein IIbhIIa and
a,P3 integrins in platelet protein trafficking. Blood 78:2603, 1991
64. Seiffert D, Keeton M, Eguchi Y,Sawdey M, Loskutoff DJ:
Detection of vitronectin mRNA in tissues and cells of the mouse.
Roc Natl Acad Sci USA 88:9402, 1991
65. Podor TJ,Loskutoff DJ: Immunoelectron microscopic localization of type 1 plasminogen activator inhibitor in the extracellular
matrix of transforming growth factor-p activated endothelial cells.
Ann NY Acad Sci 667:46, 1992
5073
66. Schleef RR,Loskutoff DJ, Podor TJ: Immunoelectron microscopic localization of type l plasminogen activator inhibitor
on the surfaceof activated endothelial cells.J Cell Biol 113: 1413,
1991
67. Booth NA, Croll A, Bennett B: The activity of plasminogen
activator inhibitor-l (PAI-I) of human platelets. Fibrinolysis 4: 138,
1990
68. Owensby DA, Morton PA, Wun T-C, S c h w a AL: Binding
of plasminogen activator inhibitor type-l to extracellular matrix of
Hep G2 cells. J Biol Chem 266:4334, 1991
69. Seiffert D, Wagner NN, Loskutoff DJ: Serum-derived
vitronectin influences the pericellular distribution of type 1 plasminogen activator inhibitor. J Cell Biol 111: 1283, 1990