Download CD63 Is a Component of Weibel

From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CD63 Is a Component of Weibel-Palade Bodies of Human Endothelial Cells
By Ulrich M. Vischer and Denisa D. Wagner
Weibel-Palade bodies are secretory granules of vascular
endothelial cells specialized in the storage of von Willebrand factor (vWF) and P-selectin, two adhesion proteins
that can be rapidly mobilizedto the cell surface by exocytosis in responseto thrombin or other agonists. In this study,
w e attempted to identify additional components of WeibelPalade bodies by raising monoclonal antibodies to these
granules, purified by cell fractionation. One antibody, 2C6,
was found to be specific for CD63, a membrane glycoprotein previously described in the lysosomes of platelets and
other cell types. The immunopurified2C6 antigen was recognized by an anti-CD63 reference antibody, 2.28, by
Western blotting. Also, the biosynthetic profile of the 2C6
antigen in endothelial cells showed a nascent molecular
mass and a glycosylation pattern identical to that of CD63.
Immunofluorescencestaining with 2C6 showed the lysosomes, and also elongated structures identified as WeibelPalade bodies by their shape, distribution, and positive
staining with anti-vWF antibodies. CD63 was also found
by Western blotting of subcellular fractions highly enriched
in Weibel-Palade bodies. Our results indicate that CD63
colocalizes with vWF and P-selectin in the Weibel-Palade
bodies of endothelial cells, and together with these adhesion proteins it could be rapidly expressed on the cell surface in areas of vascular injury and inflammation.
0 1993 by The American Society of Hematology.
T
Although only vWF and P-selectin have been identified
in Weibel-Palade bodies, many additional proteins must be
contained in (or be associated with) the granule membrane.
Several of those can be expected to play a critical role in the
process of storage and secretion. Furthermore, it is possible
that yet undescribed molecules involved in cell adhesion or
other inducible cell-surface proteins are stored in WeibelPalade bodies. Since Weibel-Palade bodies are a minor component of endothelial cells, and given the difficulty in obtaining large numbers of cells from primary cultures, direct
biochemical analysis of granule membrane proteins is not
feasible. We therefore attempted to generate monoclonal
antibodies to purified granules, to be used as reagents specific to individual proteins.
Our work has led to the finding that CD63 (lysosome-associated membrane protein 3 [Lamp 3]), a lysosomal membrane protein,”,” is also localized to Weibel-Palade bodies.
CD63 is thus an agonist-inducible cell-surface antigen in
endothelial cells.
HE VASCULAR ENDOTHELIUM is critically involved in blood coagulation and in the inflammatory
process. Upon activation, endothelial cells acquire a procoagulant state and also become adhesive for circulating leukocytes, which allows for their binding and subsequent
transfer to the extravascular space. One important activation mechanism is the fusion with the plasma membrane of
specialized secretory granules, called Weibel-Palade bodies.’ These granules are known to store and secrete von Willebrand factor (vWF), an adhesive glycoprotein involved in
the formation of the platelet
vWF is stored in a
highly multimerized form. The linear multimers are
thought to assemble into bundles, giving the granules a
characteristic rod-like shape, with a diameter of 0.1 pm and
The granules also contain P-seleca length of up to 4 p~m.’,~
tin,5*6a 140-Kd membrane glycoprotein thought to be a
receptor for circulating leukocytes. P-selectin is a member
of a family of adhesion proteins that binds specific carbohydrate ligands via their calcium-dependent lectin domain.
This family also includes endothelial E-selectin (ELAM- I )
and leukocyte L-selectin (MEL 14 Ag, LAM-I, Leu-8).’
When endothelial cells are stimulated by agonists such as
thrombin, histamine, and complement components C5b-9,
Weibel-Palade bodies undergo exocytosis: vWF is released
and P-selectin is translocated to the cell
WeibelPalade bodies may thus be critical players in primary hemostasis, as well as in the early steps of inflammation.
From the Centerfor Hemostasis and Thrombosis Research, Division of Hematology-Oncology, New England Medical Center, Boston; and the Department of Anatomy and Cellular Biology, Tufts
University School of Medicine, Boston MA.
Submitted March I I , 1993; accepted April 21, 1993.
Supported by National Institutes of Health Grants No. ROI
HL41002 and POI HL42443.
Address reprint requests to Denisa D. Wagner, PhD, Division of
Hematology-Oncology, Box 832, New England Medical Center,
750 Washington St, Boston, MA 02111,
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 I993 by The American Society of Hematology.
0006-4971/93/8204-00I6$3.00/0
1184
MATERIALS AND METHODS
Materials. All chemicals were from Sigma (St Louis, MO) unless specified. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) supplies were from BioRad (Richmond,
CA). Secondary antibodies were from Cappel (Durham, NC). Lcysteine (1,300 Ci/mmol) was from Amersham (Arlington
Heights, IL).
Cell culture. Endothelial cells were obtained from human umbilical veins as previously described: and grown in McCoy’s 5a
medium supplemented with 20% fetal calf serum (FCS), 50 pg/mL
endothelial cell mitogen (Biomedical Technologies, Stoughton,
MA), and 100 pg/mL porcine heparin. Cells were used in the first or
second passage. For the inhibition of N-linked glycosylation, subconfluent cells were grown overnight in the same medium supplemented with tunicamycin 1 pg/mL or vehicle alone (0.I % dimethyl
sulfoxide [DMSO]).
Cell fractionation. The procedure was adapted from that reported by Ewenstein et al.” All steps were performed at 4°C.
Twelve dishes of endothelial cells ( 1 5-cm diameter) grown to confluency were harvested in two batches by scraping in homogenization buffer ([HB], 250 mmol/L sucrose, I mmol/L EDTA, 20
mmol/L Tris/HCI, pH 7.2). Each batch was resuspended in 0.8 mL
HB and homogenized with 50 strokes of a Dounce homogenizer
(no. 19, Kontes Glass, Vineland, NJ). The homogenate was precleared by centrifugation at 6008 for 10 minutes. The supernatant
was saved, and the pellets were resuspended in 0.4 mL HB and
Blood, VOI 82, NO4 (August 15). 1993: pp 1 1 84-1 191
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CD63 IS A COMPONENT OF WEIBEL-PALADE BODIES
subjected to a second round of homogenization and preclearing.
The two postnuclear supernatants were pooled and layered on top
of 2 X 8.4 mL of a solution made up of 35% Percoll (Sigma, St
Louis, MO), 250 mmol/L sucrose, pH 7.2, in I O mL OakRidge
tubes (Nalge, Rochester, NY). A gradient was generated by centrifugation at 40,000gm, in a Sorvall RC-SB centrifuge (SM24 rotor).
To recover purified Weibel-Palade bodies free of Percoll, the lower
third of the two gradientswas transferred to a fresh tube and mixed
with Nycodenz (Sigma; from an 80% stock made up in HB) to a
final concentration of 30%.This preparation was overlaid with 1
mL Nycodenz 30%and 1 mL HB, and floated by centrifugation for
45 minutes at lOO,OOOg,, in a Beckman ultracentrifuge (model
L8-80M) using an SW41 swing-out rotor. The Weibel-Palade bodyenriched fraction was recovered from the upper interface and pelleted by dilution and recentrifugation at 50,000gm,,for 20 minutes.
Markers studies. vWF was determined by enzyme-linked immunosorbent assay (ELISA) as previously de~cribed.’~
The following marker enzyme activities were measured (7’1 -4galactosyl transferase14for the Golgi apparatus, succinate dehydrogenase” for the
mitochondria, N-acetyl glucosaminidase16for the lysosomes, and
alkaline phosphatase” for the plasma membrane. Total proteins
were measured by the BCA assay (Pierce, Rockford IL).
Generation of monoclonal antibodies. Purified Weibel-Palade
bodies were further fractionated by sonication in a buffer containing 10 mmol/L (NH,),C03 and 5 mmol/L EDTA, pH 8.0, and
for I hour. Somewhat unexpectcentrifugation at 1OO,OOOg,,
edly, most of the vWF was found to be associated with the pellet.
Anticipating that vWF might be an immunodominant antigen, we
used the vWF-depleted supernatant for immunization. Female
BALB/c mice were immunized on days 0,2 I , and 35 with 20 to 30
pg protein mixed with RIBI adjuvant (RIBI Immunochem Research, Hamilton, MT). Splenocyteswere fused to NSI cells in the
presence of polyethylene g l ~ c o l .The
’ ~ hybridomas were first tested
by ELISA for reactivity to purified plasma vWF (kindly provided by
Dr P. Fay, Rochester, NY). Negative wells were further tested by
immunofluorescenceon endothelialcells for the stainingof WeibelPalade bodies.
The hybridoma 2C6 was subcloned by limiting dilution three
times to ensure monoclonality, and ascites fluid was prepared in
pristane-primed BALB/c mice.
Immunofluorescence. Cells were grown on glass coverslips,
fixed in 3.7% formaldehyde, and permeabilized in 0.5% Triton
X- 100 in phosphate-buffered saline. For hybridoma screening, the
cells were incubated sequentially with undiluted tissue culture supernatants and fluorescein-conjugatedantibody to mouse immunoglobulins. For double-label staining, cells were incubated sequentially with ascites fluid I :200, rhodamine-conjugated goat antibody
to mouse IgG 1: 100, rabbit antiserum to human vWF (Diagnostica
Stago, Asnibres, France) 1:250, and fluorescein-conjugatedgoat antibody to rabbit IgG 1500. Each incubation was for 30 minutes.
With antibody 2C6, we obtained an enhanced staining sensitivity
by using the purified antibody directly coupled to rhodamine (prepared by overnight incubation at 4°C with I50 mg/mL tetramethylrhodamine isothiocyanate in 10 mmol/L Tris/HCI, pH 7.4, followed by removal of the unbound label by gel filtration over a G25
column).
Protein electrophoresis and Western blot. Proteins were analyzed by SDS-PAGE as described by Laemmli.” The separatedproteins were transferred to nitrocellulose (Schleicher & Schuell,
Keene, NH) for 1 hour at 100 V in a buffer containing 48 mmol/L
Tris, 39 mmol/L glycine, 20% methanol, and 0.0375% SDS, pH
9.0. The filters were blocked in Blotto (50 mmol/L Tris/HCl, pH
8.0,5% Carnation [LosAngeles, CAI dry skim milk), and incubated
sequentially with the test antibody (ascites fluid I: 1,OOO), affinity-
1185
purified rabbit antimouse immunoglobulins I :500, and horseradish
peroxidase-conjugated goat antirabbit immunoglobulins I :3,000
(BioRad), all diluted in Blotto.
Immunoprecipitation. The monoclonal antibody 2C6 and the
isotype-matched control antibody MOPC-2 1 (Sigma) were purified
from ascites fluid on protein A Sepharose (Pharmacia, Piscataway,
NJ) using the MAPSII kit buffer system (Biorad). The purified antibodies were coupled to CNBr-activated Sepharose at a concentration of 3 mg/mL wet gel. A crude membrane preparation from
endothelial cells was obtained by high-speed centrifugation of a
postnuclear supernatant (lOO,OOOg,, for 1 hour). Aliquots of the
resuspended pellet were gently sonicated in lysis buffer (1% Triton
X-100, 100 mmol/L Tris/HCl, 150 mmol/L NaCl, 5 mmol/L
EDTA, pH 8.0) supplemented with protease inhibitors (1 mmol/L
phenylmethylsulfonyl fluoride [PMSF], 10 pmol/L pepstatin A, 10
pmol/L E64, and I O pmol/L leupeptin), and incubated with Sepharose-coupled antibody overnight at 4°C. The beads were washed
four times with lysis buffer, and the antigens were eluted by boiling
in electrophoresis sample buffer for 2 minutes and analyzed by
Western blot.
For biosynthetic studies, subconfluent endothelial cells (grown in
IO-cm diameter dishes) were labeled with 0.2 mCi/mL L-[’~S]cysteine in cysteine-freeRPMI medium supplemented with 5% dialyzed FCS for 30 minutes. The medium was then replaced by
McCoy’s 5A medium supplemented with 20% FCS and 2 mmol/L
L-cysteine. At various time points, the cells were rinsed with phosphate-buffered saline, harvested by scraping, and resuspended in
0.1 mL lysis buffer (0.5%SDS, 50 mmol/L Tris/HCl, 2.5 mmol/L
EDTA, pH 8.0) supplemented with protease inhibitors. After 10
minutes, 1 mL of washing buffer (100 mmol/L Tris/HCl, pH 8.0,
0.1% SDS, 0.5% NP40, 0.5% deoxycholate, 2 mmol/L EDTA, 5
mg/mL ovalbumin) was added and the lysate was precleared by
centrifugationat 16,000gfor 5 minutes. The supernatant was transferred to a fresh tube and the test antibody ( 1 pL ascites fluid) was
added, together with 5 pg affinity-purified rabbit anti-mouse antibody. After an overnight incubation at 4”C, the tubes were supplemented with 40 pL protein A Sepharose, and further incubated for
2 hours at room temperature. The Sepharose beads were then
washed five times with washing buffer and the immunoprecipitated
proteins were recovered by boiling in electrophoresissample buffer
containing 50 mmol/L dithiothreitol.
RESULTS
Isolation of Weibel-Palade bodies and preparation of
monoclonal antibodies. A postnuclear supernatant of homogenized endothelial cells was fractionated over a Percoll
gradient (Fig 1). vWF, as measured by ELISA, was found i n
two peaks. T h e fractions corresponding t o the higher-density peak (fractions 8 through 12) were found to be essentially free of mitochondria, Golgi components, and plasma
membranes, and to be only minimally contaminated with
lysosomes, suggesting that they contain fairly pure WeibelPalade bodies. The high density of these fractions (1.08 t o
1.12) is typical of dense-core secretory granules.” It should
be noted that Weibel-Palade bodies are a minor cell constituent: the enriched fraction obtained from one procedure contains approximately 50 pg protein, which represents approximately 0. I % of proteins in the starting homogenate.
Mice were immunized with a vWF-depleted subfraction
of Weibel-Palade bodies (see Materials and Methods) to generate monoclonal antibodies. Hybridomas negative for reactivity toward v W F were screened by immunofluorescence.
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
VISCHER AND WAGNER
1186
1
3
5
7
9
11
r l5
200 1
1
3
5
7
9
11
n
molecular weight of this heterogeneous band, as well as the
perinuclear aspect of the immunofluorescence staining,
suggested that the 2C6 antigen was the lysosomal membrane protein CD63. Western blotting of endothelial-cell
membrane proteins with the monoclonal antibody 2.28, an
IgG2b specific for CD63 (kindly provided by Drs M.J. Metzelaar and J.J. Sixma, Utrecht, The Netherlands) visualized
a band similar to that ofthe 2C6 antigen, although the staining was slightly weaker (Fig 2, lane 2). Next, the 2C6 antigen
was purified by immunoaffinity using the antibody coupled
to Sepharose. Western blotting of the purified antigen with
antibodies 2C6 and 2.28 showed an identical pattern, confirming that both antibodies recognize the 2C6 antigen and
therefore that the 2C6 antigen was immunologically related
to CD63 (Fig 2, lanes 3 through 5).
We also analyzed extracts from platelets, human erythroleukemia (HEL) cells, neutrophils, and HL60 cells, as well
200
97
0
Y
r
n
69
u)
0
c
n
Y
46
m
1
3
5
fraction
7
9
11
number
Fig 1. Percoll density gradient fractionation of an endothelialcell postnuclear supematant. Approximately 1.2 mL starting material was loaded on top of 8.4 mL 35% Percoll before centrifugation.
Fractions of 0.8 mL were collected from the top. The values for
vWF are given in micrograms per fraction, succinate dehydrogenase (succDH), alkaline phosphatase (alk. phosph.) and N-acetylglucosaminidase (NAcGlnase) are given in optical density (OD)
units per hour per fraction, and galactosyl transferase (galact.-T) in
nanomols per hour per fraction. Data shown are from a single experiment; however, the profiles of each marker were found to be
highly reproducible.
One hybridoma, 2C6, was selected for further study, as it
was found to stain Weibel-Palade body-like structures, in
addition to abundant coarse perinuclear granules. The isotype of this antibody was determined to be IgGl-K.
Identification of the 2C6 antigen as CD63. Endothelialcell membrane proteins were obtained from a postnuclear
supernatant by centrifugation at 1 OO,OOOg,,,,,, separated on
an 1 1% SDS gel, and analyzed by Western blotting with the
2C6 antibody. This procedure showed a disperse band with
a molecular mass ranging from 30 to 60 Kd (Fig 2, lane 1).
This band could be detected only when the protein gel was
analyzed in unreduced form. An isotype-matched IgG I -K
antibody, MOPC-2 1, failed to detect this band. The range in
30
21
1
2
3
4
5
Fig 2. Detection of the 2C6 antigen in endothelial cells by Westem blot. The proteins were resolved unreduced on an 11% SDSPAGE. Lanes 1 and 2, 20 pg protein from total endothelial cell
membranes was immunostained with antibody 2C6 (lane 1) and
2.28 (lane 2). Lanes 3 to 5 contain the 2C6 antigen, purified by
immunoaffinity from total endothelial-cell membranes with the
2C6 antibody coupled to Sepharose. Aliquots equivalent to 60 pg
protein were resolved and immunostained with antibodies 2C6
(lane 3). 2.28 (lane 4). and the control antibody MOPC-21 (lane 5).
The bands seen in lane 5 represent contaminating immunoglobulin
fragments released from the immunoaffinity column. These fragments are also visible in lanes 3 and 4, in part superimposedon the
2C6 antigen. The numbers on the left indicate the migration of molecular weight markers (Kd).
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CD63
1187
IS A COMPONENT OF WEIBEL-PALADE BODIES
as human foreskin fibroblasts. The 2C6 antigen could be
detected in all these cell types, showing a wide distribution
that is also observed for CD63 (not shown).
In addition, we performed studies on the biosynthesis of
the 2C6 antigen. Endothelial cells were pulse-labeled for 30
minutes with ~-[’~S]-cysteine,and, after a chase period of 0
to 3.5 hours, the antigens were immunoprecipitated from
the cell lysate with the 2C6 antibody (Fig 3). A prominent
32-Kd band was already present at the end ofthe 30-minute
labeling period (lane I), and was followed at I hour by the
appearance of a diffuse band centered at 50 Kd, whose
prominence increased during the chase period (lanes 2
through 4). When the cells were grown in the presence of
tunicamycin, a single band of 23 Kd was immunoprecipitated at all time points (lanes 5 through 8). The molecular
weight of this band is compatible with that predicted by the
CD63 cDNA sequence.’* A small increase in the molecular
weight of this band during the chase period was consistently
observed. This most likely represents the addition of 0linked carbohydrates, which have been described in platelet
CD63.” The biosynthetic studies were also performed with
the 2.28 antibody, and identical results were obtained.
Taken together, these results establish that the 2C6 antigen
is the lysosomal membrane protein CD63 (Lamp 3).
Localization of CD63 to Weihel-Paladebodies by immunojluorescence. Fixed, permeabilized endothelial cells
were stained by double-label immunofluorescence to visualize both CD63 and vWF in the same cell (Fig 4). vWF antibodies stained the elongated, rod-shaped structures previously identified as Weibel-Palade bodies. These structures
were also shown by antibody 2C6, with an even staining
intensity. There was a one-to-one con-espondence of the
granules with the two stains, showing that CD63 is a component of all Weibel-Palade bodies. The 2C6 antibody also
showed the lysosomes as a population of coarse, brightly
stained granules clustered in the perinuclear area. These
granules were not visible with the vWF stain. No staining
was seen with a rhodamine filter when 2C6 was omitted,
ruling out the possibility of an interference from the bright
vWF fluorescein stain (Fig 4E and F). Next, the cells were
stimulated to exocytose by incubation with the calcium ionophore A23 187 ( 10 pmol/L) for 15 minutes, and fixed, permeabilized, and stained as before. The stain with anti-vWF
antibodies showed the virtual disappearance of the WeibelPalade bodies, with the vWF visible in large patches on the
cell membrane (Fig 4D).* Staining with 2C6 confirmed the
disappearance of Weibel-Palade bodies, whereas the lysosomes appeared to be essentially unaffected (Fig 4C). The
exocytosis would be expected to increase the cell-surface
expression of CD63. However, this was not observed, due
either to the dilution of the antigen in the plane ofthe membrane or to masking of the surface proteins with released
vWF. These experiments were repeated using the monoclonal antibody 2.28, with qualitatively identical results (not
shown).
Localization ofCD63 to Weibel-Paladebodies by cellfractionation. An endothelial-cell postnuclear supernatant
was fractionated over a Percoll gradient as shown in Fig 1,
and each fraction was tested by Western blot with the 2C6
antibody. Using this procedure, immunoreactivity peaked
in fractions 4 and 5 , and colocalized with the lysosomal
marker activity, N-acetyl glucosaminidase (not shown),
Chase time (min.)
0 30 90 210 0 90 0 30 90 210 0 90
200
97
-
46
-
30
-
21
-
14
-
69
Fig 3. Biosynthesis of CD63 in the presence or
absenceof an inhibitorof N-linkedglycosylation. Endothelial cells were grown ovemight in the absence
(lanes 1-6) or presence (lanes 7-1 2) of tunicamycin. Cells were labeled with L-[%] cysteine for 30
minutes, and lysed after a chase period indicated
above each lane. Immunoprecipitation with antibody 2C6 (lanes 1 through 4 and 7 through 10) and
the control antibody MOPC-21 (lanes 5 , 6 , 11 and
12) was then performed and the antigens analyzed
unreduced on an 1 1% polyacvlamide gel, the autoradiograph of which is shown. The numbers on the
left indicate the migration of molecular weight
markers (Kd).
-
1
2
3
4
5 6 7
8
9 101112
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
VISCHER AND WAGNER
1188
Fig 4. CD63 d i s t r i b u t e s with vWF in endothelial-cell Weibel-Palade bodies. Fixed, permeabilized endothelial cells were stained by
sequential incubations with purified, rhodamine-conjugated2C6 antibodies, rhodamine-conjugated goat anti-mouse antibodies, anti-vWF
rabbit serum, and fluorescein-conjugated goat anti-rabbit antibodies (except in E and F, where the 2C6 antibodies were omitted). An
identical field is shown in A and B, C and D, and E and F. Panels A, C, and E show the rhodamine stain; B, D, and F show the fluorescein stain.
(A and B) The 2C6 stain (A) shows rod-shapedstructures (arrowheads)that correspondto the Weibel-Palade bodies identified by the vWF
stain (B). Arrows indicate lysosomes, clustered in the pen-nuclear area. (C and D) Cells were stimulated with 10 pmol/L A23187 for 1 5
minutes before fixation. vWF now appears as patches associated with the cell membrane (D, arrowhead), whereas the Weibel-Palade
bodies have disappeared; the 2C6 stain (C) also shows the virtual disappearance of the rod-shapedstructures, confirming their identity as
Weibel-Palade bodies. (E and F) No staining is seen with a rhodamine filter when the 2C6 antibodies were omitted (E), confirming that the
vWF stain (F) does not interfere with the CD63 stain. Bar = 20 pm.
confirming that the bulk of the cellular pool of the protein is
localized to this compartment. To visualize CD63 associated with Weibel-Palade bodies, we performed a second
fractionation step to further separate these granules from
lysosomes. Fractions 8 through I 1 from three Percoll gradients were pooled and rediluted with HB to an approximate 40% concentration of Percoll, and this preparation
was subjected to a second centrifugation for 1 hour at
40,OOOg, . Individual fractions were analyzed by Western
blot (Fig 5). Under these conditions, the 2C6 immunoreactivity was found in fractions spanning both the 8-N-acetyl
glucosaminidase peak (fractions 5 and 6), and the vWFpeak
(fractions 8 and 9). Thus, CD63 can be detected in both
Weibel-Palade bodies and in residual lysosomes. From
these experiments, the fraction of the CD63 pool localized
to the Weibel-Palade bodies can be crudely estimated. The
amounts of CD63 immunoreactivity attributable to Weibel-Palade bodies and lysosomes are roughly equal. This
second gradient contains approximately 5% of the lysosomes present in the starting material (judging from the
activity of the marker, P-N-acetyl glucosaminidase), and,
therefore, also 5% of lysosomal CD63. Assuming complete
recovery of Weibel-Palade bodies, we can estimate that
CD63 in Weibel-Palade bodies also represents approximately 5% of the cellular pool of this protein.
DISCUSSION
CD63 is a membrane sialoglycoprotein initially described
in platelets, where it was located to the lysosomes and
shown to be translocated to the cell surface after thrombin
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CD63 IS A COMPONENT OF WEIBEL-PALADE BODIES
1189
*
Fig 5. Localization of CD63 to Weibel-Palade
bodies by cell fractionation. Endothelial-cell postnuclear supematants were fractionated as described in Fig 1. Fractions 8-1 1 from three gradients were pooled and subjected to a second
Percoll gradient centrifugation. Twelve fractions
were analyzed by Western blot with the 2C6 antibody. The values for vWF are given in micrograms
per fraction; for N-acetyl glucosaminidase in OD
units per hour per fraction. Note that fraction 9
corresponding to the vWF peak contains significantly more CD63 than fraction 3, although they
have a similar content of the lysosomal marker. The
arrowheads indicate the molecular weight markers
(from the top: 69,46, and 30 Kd).
L
40
.
--
rn
-20
3
'
'
,-,
-
ta
z
r 0
I
0
0
aJ
1
"
--
2
4
6
fraction
8
10
12
number
activation." Subsequently it was found to have a wide tissue
been studied, other lysosomal membrane proteins have
distribution and to be identical with the tumor-associated
been shown to reach the lysosomes either by a direct intracellular route, or by first reaching the cell surface, followed
antigen ME49 1 F2 It has also been classified as Lamp 3." A
closely related rat protein has been described under the
by internalization via en do some^.^^"' In steady-state, the
name of L i m ~ 1 CD63
. ~ ~ is a member of a family of lysoproportion ofthe protein pool expressed at the cell surface is
somal membrane proteins that also includes the 120-Kd
very low (-2% in the case of LEP100).30We estimate the
fraction of CD63 located to Weibel-Palade bodies to be a p
proteins Lamp 1 (Igp120) and Lamp 2 (lgpl IO), and the
80-Kd protein Limp 11." These proteins share extensive glyproximately 5%. Exocytosis could therefore cause quantitatively significant increases in CD63 cell-surface expression.
cosylation, with multiple sialylated polylactosaminoglycan
Our observations raise the question of the targeting of
residues, and have significant structural similarities in their
proteins to the secretory granule membrane. It should be
cytoplasmic tails." However, unlike the other Lamps,
noted that few proteins have been shown to be associated
CD63 has the structure ofa type I11 membrane protein, with
four transmembrane domains.22
specifically with granule membra ne^.^' One exception is PThe main finding ofthis study is that CD63 is localized in
selectin, which is expressed in platelets and endothelial cells
Weibel-Palade bodies of endothelial cells. This conclusion is
only, where it is found in a-granules and Weibel-Palade
based on the colocalization of CD63 with vWF in characterbodies, respectively. Transfection studies of P - s e l e ~ t i n , ' ~ . ~ ~
istic rod-shaped structures spread throughout the cytoplasm
as well as of deletion mutants into heterologous secretory
typical of Weibel-Palade bodies, and by their virtual disap
cells (AtT-20), have shown that the targeting is critically
pearance after agonist-induced exocytosis. This conclusion
dependent on the cytoplasmic tail. When the cytoplasmic
is strengthened by cell-fractionation results, which showed
tail of tissue factor is replaced by the P-selectin tail, the
CD63 immunoreactivity in Weibel-Palade body fractions,
chimeric protein is targeted to secretory granules, rather
defined by their high density and their enrichment in vWF.
than to the plasma membrane.32 It is possible that CD63
These observations were made using both our newly generand P-selectin share a targeting signal in their cytoplasmic
ated antibody 2C6 and the reference anti-CD63 antibody
tail. The comparison of the two domains shows no evident
2.28 with superimposable results, validating 2C6 as an antisimilarities. However, such a signal could well fail to be
CD63 monoclonal antibody.
apparent from the primary structure. CD63 has a well-charAlthough we have not directly demonstrated cell-surface
acterized gly-tyr motif shared with other lysosomal proteins,
expression, localization to the Weibel-Palade bodies implies
which was found to be critical for the lysosomal targeting of
that CD63 surface expression can be induced by endothelial
Lamp-l .28*29 We think it is unlikely that this motif contains
cell activation with secretagogues such as thrombin, fibrin,
an ambiguous dual signal for targeting to both secretory
complement components, vascular permeability factor, and
granules and lysozomes. Indeed, this hypothesis would prehi~tamin."~'Inducible surface expression of CD63, as well
dict that Lamp I and Lamp 2 are also targeted to Weibel-Paas L a m p I and Lamp2, has been reported in p I a t e l e t ~ . ~ ~ * ~lade
~ * ~bodies.
'
Immunofluorescent staining of endothelial
However, this represents an unusual case where lysosomes
cells using the monoclonal antibodies BB6 (anti-Lamp 1)
fuse directly with the plasma membrane after thrombin-inand CD3 (anti-Lamp 2) (kindly provided by Dr M. Fuduced activation. Although the trafficking of CD63 has not
kuda, La Jolla, CA) showed a typical lysosomal pattern, but
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
VISCHER AND WAGNER
1190
failed to show Weibel-Palade bodies (data not shown). We
do not believe that the targeting of CD63 is a passive, missorting phenomenon. Barriocanal et a124have suggested that
the targeting of LIMP I (rat CD63) to lysosomes is a relatively slow process, compared with the other lysosomal
membrane proteins lgp120 (LIMP 111) and LIMP 11. This
results in an unusually prolonged residence time-and presumably a high concentration-of the protein in the distal
Golgi apparatus. CD63 could thus be passively incorporated into the membranes of nascent Weibel-Palade bodies
emerging from the distal Golgi. However, in that case, we
would expect it to be retrieved from immature secretory
granules, as they undergo remodeling and trimming of the
membrane.31This appears not to be the case, since by immunofluorescent staining we found CD63 in all Weibel-Palade
bodies, rather than only in the immature, more centrally
located subset.
The functional role of CD63 remains unknown. In its
lysosomal location, it has been suggested that, due to its high
density and its extensive glycosylation, it creates a carbohydrate lining on the membrane lumen that protects the membrane from the action of intragranular proteases. The inducible expression on the endothelial cell surface suggested by
our study and that observed in plateletsl0 could imply additional functions for this protein. A similar protective effect
from proteases released by leukocytes during an inflammatory response could be envisioned. In addition, Toothill et
a133have shown that neutrophil adhesion to thrombin-activated endothelial cells is largely inhibited by an anti-CD63
antibody. This raises the possibility that CD63 on the cell
surface is involved in adhesive interactions with circulating
leukocytes. It has been hypothesized that lysosomal membrane proteins are ligands for the selectins." Indeed, the
amount of Lamp 1 expressed on cancer cells was shown to
correlate with the cells' binding to E-selectin; the binding
was inhibited by sialyl Lewis,-containing glycolipid^.^^ Similarly, CD63 could be a ligand for the selectins, in particular
for L-selectin, and thus mediate leukocyte adhesion to the
vascular endothelium. Arguing against the hypothesis of
such a highly cell-specific function is the wide tissue distribution of CD63. However, it should be noted that CD63 is
extensively glycosylated. In our tissue distribution studies,
we have observed substantial differences in the apparent
size of CD63 from different cell types, most likely reflecting
specific glycosylation patterns. It is conceivable that a particular carbohydrate structure gives the protein cell type-specific functional properties.
In summary, from our studies we can conclude that
CD63 colocalizes with vWF and P-selectin in the WeibelPalade bodies of endothelial cells, and that together with
these adhesion proteins it could be rapidly expressed in
areas of vascular injury or inflammation.
ACKNOWLEDGMENT
We thank Kathy Reebenacker for excellent technical assistance
with endothelial cell culture. We also thank Dr P. Fay for the gift of
purified vWF, and Drs M.M. Metzelaar, J.J. Sixma, and M. Fukuda
for the gift of monoclonal antibodies; Drs L.I. Hecker and C.J.
Rhodes, who gave helpful advice throughout the course of this
study; and Drs R. Bonfanti and B.M. Ewenstein, who helped to set
up the endothelial cell-fractionation procedure.
REFERENCES
1. Weibel ER, Palade GE: New cytoplasmic components in the
cytoplasm of endothelia. J Cell Biol 23: 10 1, 1964
2. Wagner DD, Olmsted JB, Marder VJ: Immunolocalization of
von Willebrand protein in Weibel-Palade bodies of human endothelial cells. J Cell Biol 99355, 1982
3. Wagner DD: Cell biology of von Willebrand factor. Ann Rev
Cell Biol6:2 17, 1990
4. Wagner DD, Saffaripour S, Bonfanti R, Sadler JE, Cramer
EM, Chapman B, Mayadas TN: Induction ofspecific storage organelles by von Willebrand factor propolypeptide. Cell 64:403, 1991
5. Bonfanti R, Furie BC, Furie B, Wagner DD: PADGEM
(GMPI 40) is a component of Weibel-Palade bodies of human endothelial cells. Blood 73:1109, 1989
6. McEver RP, Beckstead JH, Moore KL, Marshall-Carlson L,
Bainton DF: GMP 140, a platelet alpha-granule membrane protein,
is also synthesized by vascular endothelial cells and is localized in
Weibel-Palade bodies. J Clin Invest 84:92, 1989
7. Lasky LA: Selectins: Interpreters of cell-specificcarbohydrate
information during inflammation. Science 258:964, 1992
8. Sporn LA, Marder VJ, Wagner DD: Inducible secretion of
large, biologically potent von Willebrand factor multimers. Cell
46:185, 1986
9. Hattori R. Hamilton KK, Fugate RD, McEver RP, Sims PJ:
Stimulated secretion of endothelial von Willebrand factor is accompanied by rapid redistribution to the cell surface of the intracellular
granule membrane protein GMP 140. J Biol Chem 264:7768, 1989
10. Nieuwenhuis HK, van Oosterhout JJG, Rozenmuller E, van
Iwaarden F, Sixma JJ: Studies with a monoclonal antibody against
activated platelets: evidence that a secreted 53,000 molecular
weight lysosome-like granule protein is exposed on the surface of
activated platelets in the circulation. Blood 70:838, 1987
11. Fukuda M: Lysosomal membrane glycoproteins. Structure,
biosynthesis and intracellular trafficking. J Biol Chem 266:2 1327,
1991
12. Ewenstein BM, Warhol MJ, Handin RI, Pober JS: Composition of the von Willebrand factor storage organelle (Weibel-Palade
body) isolated from cultured human umbilical vein endothelial
cells. J Cell Biol 104:1423, 1987
13. Koedam JA, Cramer EM, Briend E, Furie B, Furie BC,
Wagner DD: P-selectin, a granule membrane protein of platelets
and endothelial cells, follows the regulated secretory pathway in
AtT-20 cells. J Cell Biol 116:617, 1992
14. Chaney W, Sundaram S, Friedman N, Stanley P The Lec4A
CHO glycosylation mutant arises from miscompartmentalization
of a Golgi glycosyltransferase. J Cell Biol 109:2089, 1989
15. David GA, Bloom FE: Subcellular particles separated
through a histochemical reaction. Anal Biochem 5 1:429, 1973
16. Sellinger OZ, Beaufay H, Jacques P, Doyen A, deDuve C:
Tissue fractionation studies. 15. Intracellular distribution and properties of @-N-acetylglucosaminidase and @-galactosidasein rat liver.
Biochem J 74:450, 1960
17. Stonie B, Madden EA: Isolation of subcellular organelles.
Methods Enzymol 182:203, 1990
18. Grimaldi K, Hutton JC, Siddle K: Production and characterization of monoclonal antibodies to insulin secretory granule membranes. Biochem J 245557, 1987
19. Kohler G, Milstein C: Continuous culture of fused cells secreting antibody of predefined specificity. Nature 256:495, 1975
20. Laemmli U K Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680, 1970
2 1. Hutton JC, Peshavaria M: Proton-translocating Mgz+-dependent ATPase activity in insulin secretory granules. Biochem J
204:161, 1982
22. Metzelaar MJ, Wijngaard P U , Peters PJ, Sixma JJ, Nieu-
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CD63 IS A COMPONENT OF WEIBEL-PALADE BODIES
wenhuis HK, Clevers HC: CD63 Antigen: A novel lysosomal membrane glycoprotein cloned by a screening procedure for intracellular antigens in eukaryotic cells. J Biol Chem 266:3239, 1991
23. Hildreth JEK, Derr D, Azorsa D O Characterization of a
novel self-associating M, 40,000 platelet glycoprotein. Blood
77:121, 1991
24. Barriocanal JG, Bonifacino JS, Yuan L, Sandoval IV: Biosynthesis, glycosylation, movement through the Golgi system, and
transport to lysosomes by an N-linked carbohydrate-independent
mechanism of three lysosomal integral membrane proteins. J Biol
Chem 261:16755, 1986
25. Brock T, Dvorak HF, Senger DR: Tumor-secreted vascular
permeability factor increases cytosolic Ca” and von Willebrand
factor release in human endothelial cells. Am J Pathol 138:2 13,
1991
26. Febbraio M, Silverstein RL: Identification and characterization of Lamp 1 as an activation-dependent platelet surface glycoprotein. J Biol Chem 265:18531, 1990
27. Silverstein RL, Febbraio M: Identification of lysosome-associated membrane protein 2 as an activation-dependent platelet surface glycoprotein. Blood 80: 1470, I992
28. Williams MA, Fukuda M: Accumulation of membrane gly-
1191
coproteins in lysosomes requires a tyrosine residue at a particular
position in the cytoplasmic tail. J Cell Biol 1 1 1:955, 1990
29. Harter C, Mellman 1: Transport of the lysosomal membrane
glycoprotein lgp 120 (lgp-A) to lysosomes does not require appearance on the plasma membrane. J Cell Biol 1 17:31 I, 1992
30. Lippincott-Schwartz J, Fambrough DM: Cycling of the integral membrane glycoprotein, LEP 100, between plasma membrane
and lysosomes: Kinetic and morphological analysis. Cell 49:669,
1987
3 1. Arvan P, Castle D: Protein sorting and secretion granule formation in regulated secretory cells. Trends Cell Biol2:327, 1992
32. Disdier M, Morrissey JH, Fugate RD, Bainton DF, McEver
R P Cytoplasmic domain of P-selectin contains the signal for sorting into the regulated secretory pathway. Mol Biol Cell 3:309, 1992
33. Toothill VJ, van Mourik JA, Nieuwenhuis HK, Metzelaar
MJ, Pearson JD: Characterization ofthe enhanced adhesion of neutrophil leukocytes to thrombin-stimulated endothelial cells. J Immunol 145:283, 1990
34. Sawada R, Lowe JB, Fukuda M: Increased expression of a
lysosomal membrane glycoprotein in the cell surface of poorly metastatic colonic cells results in stronger adhesion to E-selectin expressing cells. J Cell Biochem 17S:374a, 1993 (abstr)
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
1993 82: 1184-1191
CD63 is a component of Weibel-Palade bodies of human endothelial
cells
UM Vischer and DD Wagner
Updated information and services can be found at:
http://www.bloodjournal.org/content/82/4/1184.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.