402
Studies of the Release from Human Platelets of the
Growth Factor for Cultured Human Arterial
Smooth Muscle Cells
LARRY D. WITTE, KAREN L. KAPLAN, HYMIE L. NOSSEL, BRUCE A. LAGES,
HARVEY J. WEISS, AND DEWITT S. GOODMAN
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
SUMMARY Platelets contain a growth-promoting factor for arterial smooth muscle cells (SMC) that may play a
major role in atberogenesis. We have studied some of the effects of the platelet-derived growth factor (PDGF) on
human arterial SMC in culture and the release of PDGF from human platelets in relationship to other released
substances. Material released from platelets was highly potent in stimulating human SMC to proliferate. A substantial
portion of the growth-promoting activity of human serum could be attributed to a factor(s) released from platelets.
Similar dose-response patterns to PDGF were observed with human SMC and with mouse 3T3 cells. The time-course
of release of PDGF and its concentration dependence on human thrombin were determined in comparison with
serotonin, ADP, ATP, an acid bydrolase, platelet factor 4 (PF4), and /3-thromboglobulln (/JTG). PDGF activity
was assayed by stimulation of the incorporation of 'H-thymidine into DNA of 3T3 cells; PF4 and /JTG were
measured by newly developed radioimmunoassays. PDGF, PF4, and /JTG were released from platelets by lower
concentrations of thrombin than those required for release of the other components. The results suggest that PDGF,
PF4, and /3TG are localized in the platelet in granules different from either the dense bodies (that contain serotonin,
ADP, ATP) or the acid hydrolase-containing granules, possibly in a-granules. The contents of these PDGFcontainlng a-granules are actively released during the release reaction and are particularly sensitive to release by low
doses of thrombin.
IT IS well established that whole blood serum is essential
for the growth of most mammalian cells in culture in
vitro.1"3 Recent studies have, moreover, demonstrated
that a substantial proportion of the growth-promoting
activity of serum is derived from blood platelets.4"8 Thus,
material from platelets has been shown to stimulate the
growth of monkey arterial smooth muscle cells (SMC)145
of monkey and mouse (3T3) fibroblasts,5-6 and of human
glial cells.7-8
Ross and his colleagues have suggested that the growth
factor derived from platelets may play a critical role in the
pathogenesis of atherosclerosis.9-10 Proliferation of intimal
SMC in affected arteries appears to be a key element in
the development of the early atherosclerotic lesion9- "• vl
Such SMC proliferation is dramatically evident in vivo
after endothelial cell injury by mechanical, chemical, or
other means (reviewed in Ref. 9). It has been suggested
that factors released by platelets are largely responsible
for promoting the intimal SMC proliferation associated
with the development of experimental or human arteriosclerotic disease.8-l0 Support for this hypothesis was obtained recently from studies in which intimal thickening
and arteriosclerotic lesion formation in rabbits were prevented by rendering the animals thrombocytopenic with
From the Department of Medicine, Columbia University College of
Physicians and Surgeons, New York, New York.
This work was supported in part by National Institutes of Health Grants
HL 14236, HL 21006, HL 15486, and HL 14595.
Address for reprints: Dr DeWitt S. Goodman, Columbia University,
College of Physicians and Surgeons, 630 West 168th Street, New York,
New York 10032.
Received June 20, 1977; accepted for publication September 13, 1977.
antiplatelet serum; n similar results were obtained in homocystinemic baboons treated with the antiplatelet drug,
dipyridamole.14
Limited information is available about the chemical or
biological characteristics of the platelet-derived growth
factor. The growth factor has been reported to have the
properties of a relatively heat-stable, basic protein11'8 and
to be released from platelets during the release reaction
induced by several different agents.15
We now report the growth-promoting effects of material
released from human platelets upon human arterial
smooth muscle cells in culture and describe the characteristics of the release of the growth factor from platelets in
comparison to other known platelet-released products.
Methods
Human Smooth Muscle Cells
The human arterial smooth muscle cells were grown
from explants of segments of media from normal-appearing pieces of lower abdominal aorta obtained within
minutes of death from two young male kidney transplant
donors. Informed consent for the procedure was obtained
in conformity with institutional policy. The method used
was similar to that of Ross,16 as applied by Albers and
Bierman17 for the culture of human arterial SMC. One- to
2-mm explants of the superficial layer of media were
prepared as described by Coltoff-Schiller et al.1* Ten to
15 explants were placed into 25-cm2 Falcon flasks (Fisher
Scientific Co.) containing 1 ml of Dulbecco's modified
Eagle's medium (DME), supplemented with penicillin
(100 U/ml); streptomycin (100 Mg/ml); 1% (wt/vol) non-
RELEASE OF GROWTH FACTOR FROM HUMAN PLATELETS/Witfe et al.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
essential amino acids; 2mM L-glutamine (Grand Island
Biological Co.); and 20% (vol/vol) normal human whole
serum (freshly prepared). The flasks were incubated undisturbed for 4 days in a humidified Wedco model E2178, 37°C COj incubator (Wedco, Inc.), after which the
medium was changed three times per week. Following 34 weeks of growth, the cells (outgrown from the explants)
were dissociated for subculturing with 0.05% trypsin0.02% EDTA (Grand Island Biological Co.). Inocula of
0.75-1.5 x 106 cells were plated into 75-cm2 Falcon flasks
(Fisher Scientific Co.), using 10 ml of the above-described
medium. The cells were allowed to grow to confluency
(approximately 1-2 weeks), during which time the human
serum concentration was gradually reduced from 20% to
10% (vol/vol). After the second or third passage, a
portion of the cells was frozen" The cells used in the
experiments were between the third and seventh passage.
The SMC were distinguished from other cell types by
their characteristics, as described previously.16> 17'20 Compared to fibroblasts, the SMC had a longer lag period
before outgrowth from the explant and a slower rate of
growth after being established in culture. The pattern of
growth was unlike that of fibroblasts in that the SMC
tended to pile up into many cell layers in a multicentric
proliferative pattern.20 By electron microscopy, the SMC
demonstrated myofilaments and dense bodies, characteristics of SMC in culture.16
3T3 Mouse Cells
3T3-Swiss albino (contact-inhibited fibroblasts, no.
ATCC CCL 92) mouse cells were purchased from the
American Type Culture Collection (ATCC), Rockville,
Md. The cells were stored frozen in vials and were grown
in stock 75-cm2 Falcon flasks according to the instructions
provided by the ATCC. The medium was DME supplemented with 10% (vol/vol) fetal calf serum. The medium
was changed every other day, and the cells were passed
every 3-4 days, with care taken not to allow the cells to
reach confluency before passage.
403
sion of Biologies, Bethesda, Md.). After aggregation, the
mixture was centrifuged and the supernatant fluid was
brought to 56°C for 10 minutes to inactivate complement
and defibrinogenate the plasma, as described by KaserGlanzmann et al." After cooling, the supernatant fluid
was centrifuged at 10,000 g for 30 minutes at 4°C to
remove any precipitated protein.
Preparation of Platelet-Rich Plasma, and Platelet-Poor
Plasma, Heat-Treated (PPPA)
Platelet-rich plasma was prepared, as described by
Holmsen et al.,23 from venous blood drawn from healthy
donors who had not taken aspirin within the preceding 2
weeks. The PPPA was prepared from the remainder of the
blood (after the removal of the platelet-rich plasma) by
two consecutive centrifugations for 30 minutes at 4°C, at
1200 g-max and at 10,000 g-max, to remove any
remaining platelets. The platelet-free plasma obtained
after the second centrifugation was dialyzed against phosphate-buffered saline, heated at 56°C for 30 minutes (in
order to remove the fibrinogen by precipitation and to
inactivate complement), cooled, and centrifuged at
10,000g- max for 30 minutes at 4°C to remove the formed
precipitate. The resulting plasma was designated PPPA.
Preparation of Labeled Gel-Filtered Platelets (GFP)
Platelet-rich plasma (12-15 ml) was incubated at 37°C
for 15 minutes with 5-hydroxytryptamine-l-14C, 0.04 ^Ci/
ml (48 mCi/mmol, from Schwarz/Mann). The labeled,
platelet-rich plasma was then immediately gel filtered, as
described by Lages et al.,24 on a 4-cm diameter column
packed with Sepharose 2B (Pharmacia Fine Chemicals) to
a bed height of 20 cm, and eluted with Ca*+-free Tyrode's
buffer containing 0.1% glucose and 0.2% crystalline
human albumin (Sigma Chemical Co.). Approximately 30
ml of GFP were collected, containing 1-3 x 108 platelets/
ml.
Aggregation of GFP and Preparation of Samples for
Assay
Preparation of Lipoproteins
Human very low-, low-, and high-density lipoproteins
(VLDL, LDL, and HDL) and the lipoprotein-deficient
(density > 1.21) fraction of plasma or serum, were
isolated by sequential ultracentrifugation at densities
1.006, 1.063, and 1.21, according to the method of Havel
et al.21 The lipoprotein and lipoprotein-deficient fractions
were dialyzed exhaustively against phosphate-buffered
saline (pH 7.4) and then against DME before being added
to cells.
Preparation of Supernatant Fluid (PS) from Aggregated
Platelets
Platelets from one unit of blood, concentrated into 2025 ml of plasma, were obtained from the New York Blood
Center in fresh 20-unit batches. The platelets were washed
free of plasma by centrifugation at 1,200 g for 15 minutes
at 22°C, adjusted to a concentration of approximately 1010
platelets/ml, and aggregated with purified human thrombin, 1 U/ml, (kindly provided by Dr. D. Aronson, Divi-
The GFP were aggregated with purified human thrombin (dissolved in phosphate-buffered saline) at the concentrations and for the times described in the figures. In all
experiments, buffer blank controls (samples containing
only the buffer in which the platelets were suspended, plus
the thrombin solvent, phosphate-buffered saline) and
freeze-thawed controls (samples of GFP containing the
thrombin solvent, frozen and thawed four times) were
prepared at the same time. In the time-course experiments, the entire experimental sample of GFP was aggregated as a single unit in a siliconized beaker at 37°C with
stirring, and the individual samples were removed at the
times indicated. For the dose-response studies, each sample was aggregated separately at 37°C with the indicated
dose of human thrombin, using a Payton Dual Channel
Aggregation Module with a Riken-Denshi Vertical TwoChannel Recorder. In each study, the samples were
removed at the times indicated in the Results section and
placed immediately into an ice-water bath. The samples,
plus controls, were centrifuged at 10,000 g-max for 10
404
CIRCULATION RESEARCH
minutes at 4°C, and the supernatant fluids were decanted
and used for the various assays described below.
Assay for Serotonin Release
Fifty microliters of each GFP supernatant sample were
removed, placed into a scintillation vial containing 10 ml
of Insta-Bray (York Town Research) and 0.5 ml of 0.2 M
NaOH and assayed for MC in a Packard model 3003 liquid
scintillation spectrometer (Packard Instrument Co.).
Assays for Cell Growth Stimulation
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Growth stimulation was assayed in SMC and 3T3 cells
by both cell counting, with a ZB-1 Coulter Counter," and
by the incorporation of 3H-labeled thymidine into cellular
DNA, by the method described by Gospodarowicz*6 and
Rubin,27 with modifications. 3T3 cells or SMC were
dissociated from stock flasks with 1-2 ml of trypsin, and
an inoculum of 4 x 104 cells was seeded into plastic dishes
(35 mm diameter) (Fisher Scientific Co.) containing 2 ml
of the appropriate 10% serum-containing medium previously described for each cell line. Six to 8 hours later for
the 3T3 cells, and 24 hours later for the SMC, the medium
was removed and replaced with medium containing 1%
(vol/vol) PPPA in place of the 10% serum, plus penicillin
(100 U/ml), streptomycin (100 /xg/ml), and 15 mM N-2hydroxyethylpiperazine-N'-2-ethanesulfonic acid (Grand
Island Biological Co.). Twenty-four hours (3T3) or at 48
and 96 hours (SMC) after the initial medium change,
identical changes were made. The cells entered a resting
stage within 48 hours after the last medium change; the
samples to be tested for growth-promoting activity were
added at this time (48 hours). Sixteen hours (3T3) or 32
hours (SMC) after the sample addition, 50 ftl of DME
containing 2 /xCi of thymidine-methyl-3H (2 Ci/mmol,
from New England Nuclear Corp.) and 2 ^tg of unlabeled
thymidine (Sigma Chemical Co.) were added to each dish.
Four hours (3T3) or 16 hours (SMC) later, the medium
was removed from each plate, and the monolayer of cells
was washed at ice temperature two times with phosphatebuffered saline, once with 5% trichloroacetic acid for 5
minutes and once with water. The cells were dissolved in
0.5 ml of 0.2 N NaOH at 37°C, after which the dish
contents were quantitatively transferred (with a further
0.5 ml wash of 0.2 N NaOH) to scintillation vials containing 12 ml of ScintiVerse (Fisher Scientific Co.) and 100
/il of trichloroacetic acid. The vials were assayed for 3H
(representing the incorporation of 3H-thymidine into
DNA) in a Packard liquid scintillation spectrometer. All
samples were assayed for growth factor activity in duplicate or triplicate.
Assays for Adenine Nucleotides and Acid Hydrolase
For analysis of ADP, ATP, and /S-N-acetylglucosaminidase (/3-N) from the GFP supernatant and the appropriate control supernatant fluids, samples were taken and
treated as follows: adenine nucleotides: 0.1 ml into 0.1 ml
of an ice-cold mixture of ethanol-0.1 M EDTA (pH 7.4),
9:1 (vol/vol); acid hydrolases: 0.8 ml into 0.1 ml 1.8%
Triton X-100 in water.
ATP and ADP were analyzed by the firefly-luciferase
VOL. 42, No. 3, MARCH
1978
method of Holmsen et al.,28 using an Aminco microphotometer (American Instruments Co.), as described by
Weiss etal. 28
/3-N was determined by the liberation of p-nitrophenol
from the appropriate substrates, as described by Holmsen
et al.30 The results were expressed as units of acid hydrolase activity, where 1 U is defined as the hydrolysis of 1
/imol of p-nitrophenol per minute from the appropriate
substrate.
Assays for Platelet Factor 4 (PF4) and /3-ThromboglobuIin(/3TG)
PF4 and /3TG were measured by radioimmunoassay.31
The PF4 assay used a goat antiserum to PF4 and I2S1-PF4
labeled by the chloramine-T method; 50% inhibition of
binding was obtained with 26 fmol purified PF4. The /3TG
assay was performed with a rabbit antiserum to /3TG and
125
I-/3TG labeled by the chloramine-T method; 50% inhibition of binding occurred with 1.8 fmol purified /3TG.
Results
Response of Human Arterial Smooth Muscle Cells to
Various Serum and Platelet-Poor Plasma Fractions
Initial studies were carried out with human arterial
SMC to investigate the growth-promoting activity of various human serum components. Some of the results obtained are shown in Figure 1. SMC growth and proliferation were maximal when the cells were grown in the
presence of 10% whole human serum. Most (approximately two-thirds) of the growth-promoting activity of
whole serum was found to be present in the lipoproteindeficient (density > 1.21) fraction of the serum. In
contrast to serum, 10% heat-defibrinogenated plateletpoor plasma (PPPA) showed only slight growth-promoting
activity, and the lipoprotein-deficient fraction of the PPPA
appeared able to maintain the cells in a resting stage but
unable to stimulate new cell growth.
Lipoproteins alone were not able to maintain the cells
in a resting stage, and the cell count decreased when the
SMC were incubated with LDL (see Fig. 1), VLDL, or
HDL (data not shown) prepared from either whole human
serum or PPPA. When lipoproteins were added to the
lipoprotein-deficient fraction prepared from PPPA, some
improvement in cell growth was apparent (data not shown
in Figure 1), especially with addition of VLDL and LDL.
However, the improvement was not large, and the addition of lipoproteins could not compensate for the loss in
growth-promoting activity seen in the PPPA, compared to
whole serum.
Response of Human Arterial Smooth Muscle Cells and
Mouse 3T3 Fibroblasts to Platelet Supernatant (PS)
Material
Since serum contained growth-promoting activity not
present in platelet-poor plasma (Fig. 1), experiments were
carried out to investigate directly whether this activity
resided in material released from human platelets during
their aggregation. The studies employed PS preparations,
comprising substances released from washed platelets
RELEASE OF GROWTH FACTOR FROM HUMAN PLATELETS/Wi'we et al.
J«IOV
K3% WHS
LLJ
_)
Q.
17.
LU
O
LDL(WHS)
0 % SERUM
10-
DAYS IN CULTURE
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIGURE 1 The effect of whole human serum (WHS), plateletpoor plasma heat-treated (PPP&), the lipoprotein deficient (density
> 1.21) fraction of WHS and ofPPPb, the LDL fraction from
WHS, and incubation medium alone (0% serum) on the proliferation of human arterial smooth muscle cells. All of the fractions
used in this study were prepared from a sample of blood from a
normal donor. One-half the blood was clotted in glass centrifuge
tubes and the serum separated (WHS). The remaining blood was
used for the preparation of PPPL (see Methods). Portions of the
WHS and PPP& were separated into various lipoprotein and
lipoprotein-deficient fractions by ultracentrifugation. The various
fractions were further prepared for addition to the cells, as
described under Methods for the preparation of PPPA, except that
the lipoprotein fractions were not heat inactivated. The cells were
plated at 3 x 10* cells 135-mm diameter dish in 2 ml of standard
growth medium containing 1 % WHS, with medium changes every
48 hours throughout the experiment. On day 7, the medium was
changed to the appropriate test medium, with the PPPA and
lipoprotein-deficient (density > 1.21) samples adjusted to the same
protein content as that of the 10% WHS, and the LDL concentration equal to that found for LDL in the 10% WHS. Cell counts
were made on duplicate plates on the days indicated. Duplicate
samples agreed closely with each other, and the data points shown
represent the mean values for each duplicate pair.
during the thrombin-induced platelet release reaction.
The protein concentration of the PS preparations averaged
approximately 1 mg/10'° platelets.
Before this work was developed, experiments were
conducted to ensure that the purified human thrombin
used did not itself provide significant growth stimulation
in the system under study, since Chen and Buchanan have
reported the stimulation of chick embryo fibroblasts by
thrombin.32 In two experiments with 3T3 cells, purified
human thrombin, tested at concentrations varying from
0.01 to 24 U/ml of incubation medium, gave very little
stimulation (less than 20% of that seen with 10% fetal
calf serum) of the incorporation of nH-thymidine into
DNA.
Figure 2 shows the effect of increasing concentrations
of PS material on the stimulation of the incorporation of
•1H-thymidine into the DNA of SMC. A dose-related
increase in stimulation was observed, particularly over the
range of 0.5-5 fig of PS protein/ml. At a level of 5 /xg/ml,
the stimulation of human SMC by PS was 67% as great as
that seen with 10% whole human serum (which gave
maximal stimulation).
405
Similar data on the mitogenic effects of PS on 3T3
mouse cells are also shown in Figure 2. The proliferation
of 3T3 cells was effectively stimulated by human PS, with
a dose-response pattern similar to that observed with the
human SMC. Both 3T3 cells and SMC showed the sharpest increase in stimulation between the concentration
range of from 1 to 5 /xg of PS protein/ml of medium, when
tested over a concentration (dose) range of from 0 to 50
/xg of PS protein/ml.
Since both SMC and 3T3 cells responded similarly to
the growth-promoting factor(s) released from human
platelets, further studies on the characteristics of the
release of the growth factor from platelets were carried
out using 3T3 cells for the assay of the growth factor
activity of various samples. This provided a more rapid
and simpler standard assay than that which could be
carried out with SMC.
Time Course of the Release of Platelet-Derived Growth
Factor (PDGF) from Human Platelets
Four pilot experiments were carried out to explore the
rate of release of PDGF in comparison with that of
serotonin. These experiments used platelets washed either
by serial centrifugation or by gel filtration on Sepharose
2B and aggregated with different doses of purified human
thrombin. In addition, an experiment was conducted to
measure the rate of release of PDGF and platelet-bound
l4
C-serotonin from platelet-rich plasma on addition of 4
/xg of collagen/ml. In each of these initial studies, the
release of PDGF and of serotonin followed a generally
similar pattern. Thrombin-induced release of PDGF and
.-—i
lOOr
PLATELET SUPERNATANT, jjq/ml
FIGURE 2 The effect of increasing concentrations of platelet
supernatant protein (indicated in ng/ml on the abscissa) on the
stimulation of the incorporation of 3H-thymidtne into DNA of
human arterial SMC and of3T3 cells. The figure shows the results
from two separate experiments with SMC (with each data point
determined in triplicate in each experiment) and five separate
experiments with 3T3 cells (with each data point determined in
duplicate in each experiment). Different batches of PS material
were involved in each experiment. The stimulation obtained with
10% whole human serum (for SMC) or with 10% fetal calf serum
(for 3T3 cells) was determined and designated as maximal stimulation. The results for each concentration of PS are expressed as
the percent of that (maximal) stimulation observed. Mean ± SEM
values arc shown for each data point.
CIRCULATION RESEARCH
406
serotonin (from washed platelets) was rapid, mainly occurring during the first 30 seconds. Collagen-induced
release (platelet-rich plasma) of both components was
slower and occurred mainly between 1 and 3 minutes.
More extensive experiments were then carried out to
compare the time course of the release of PDGF with that
of a number of other platelet-released products, namely
ADP, ATP, PF4, /3TG, and /3-N. Three experiments were
conducted. The results of one experiment are shown in
Figure 3, with similar results being obtained in the other
two experiments. PDGF, PF4, /3TG, serotonin, and ADP
all showed a rapid pattern of release during thrombininduced aggregation. ATP (data not shown) showed the
same pattern of release as did ADP and serotonin, while
/3-N (data not shown) showed a much slower rate of
release, as previously described by Holmsen et al. M M
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Dose-Response Studies of Thrombin-Induced Release of
Platelet-Derived Growth Factor
Detailed studies were carried out on the effect of
thrombin concentration on the release of PDGF from
human platelets in comparison with a number of other
released platelet components. We performed two detailed
dose-response experiments using GFP and concentrations
of thrombin ranging from 0.01 to 0.5 U/ml. We also
determined the total amount of each component present
in the platelets (using platelets that were frozen and
thawed four times) and the maximal amount releasable by
thrombin (5 U/ml for 5 minutes). The amount of each
component released at each concentration of thrombin
studied was then expressed in two ways, both as the
percent of the total amount released by repeated freezethawing and as the percent of the total amount released
after 5 minutes by thrombin, 5 U/ml. The results of the
two detailed experiments were similar, with the major
findings (see below) being confirmed by the results of four
additional, more limited experiments.
Figure 4 shows the results of one of the detailed
experiments, with the data for each component expressed
—1
50
45
60
90
II
T
120
II
|
ISO
T I M E , SECONDS
FIGURE 3 Time course of the release of platelet-derived growth
factor (PDGF), platelet factor 4 (PF4),
p-thromboglobulin
(f3TG), serotonin (5HT), and ADP from gel-filtered platelets after
addition of 0.05 U thrombin/ml of GFP. PDGF activity was
assayed as the stimulation of the incorporation of 'H-thymidine
into the DNA of 3T3 fibroblasts. For each released component,
the data are expressed as the percent of the amount of that
component released by a sample of GFP aggregated for 10 minutes
with the same concentration (0.05 U/ml) of thrombin.
VOL. 42, No. 3, MARCH
1978
IOOT
< n
60-
°z <t
2 V
40-
015 02 025 03
THROMBIN, UNITS/ml
FIGURE 4 The thrombin concentration dependence (dose-response) of the release of platelet-derived growth factor (PDGF),
platelet factor 4 (PF4), p-thromboglobulin (flTG), serotonin
(5HT), ADP, ATP, and p-N-acetylglycosaminidase (fi-N) from
gel-filtered platelets. Thrombin-induced release was determined
after 3 minutes of aggregation with different concentrations of
thrombin, as shown. The data are expressed as the percent of the
total amount of each component released from a 4-times freezethawed sample of GFP. For each of the components studied, a
sample of GFP aggregated for 5 minutes with thrombin, 5 U/ml,
released the following amounts (expressed as the percent of the
total amount released from the freeze-thawed sample). PDGF,
87%; PF4, 90%; /3TG, 106%; 5HT, 85%; ADP, 59%; ATP,
53%; and /3-/V, 66%. The absolute amounts of the various
components released from the repeatedly freeze-thawed GFP were:
0TG, 136 pmol/ml; PF4, 165 pmol/ml; /3-iV, 4 x I0'3 U/ml;
ADP, 9.01 nmol/ml; and ATP, 7.89 nmol/ml.
as the percent of the total amount released from repeatedly freeze-thawed platelets. The thrombin concentrationdependence of the release of PDGF, PF4, and /3TG was
found to be different from that of the release of the other
components measured. Thus, the major release of PDGF,
PF4, and /3TG occurred between 0.015 and 0.025 U of
thrombin/ml of GFP, with 70-80% of each of these three
components being released by a concentration of 0.025
U/ml. In contrast, the major portions of the known dense
granule components (serotonin, ADP, and ATP) were
released at thrombin concentrations between 0.02 and
0.05 U/ml, with only 15-20% of these components being
released at 0.025 U/ml. The dose-response curve of acid
hydrolase (/3-N) release was different from that of any of
the other components, and demonstrated both a higher
thrombin concentration necessary to achieve equivalent
release and less complete (30%) release with the highest
concentration (0.1 U/ml) shown. These latter findings are
similar to those described previously.30- M
The difference between the thrombin concentrationdependence of the release of PDGF, PF4, and /3TG and
that of the other released components was also evident
when the data were expressed as the percent of the
maximal amount of each component releasable by thrombin. Table 1 presents the percent of maximal release
obtained with 0.025 U of thrombin/ml in each of the two
detailed experiments, with the data expressed in each of
the two ways described. It is evident that, when either
method was used to express the data, the three compo-
RELEASE OF GROWTH FACTOR FROM HUMAN PLATELETS/W/»e et al.
TABLE
407
1 Percent of the Maximal Release Obtained with 0.025 U of Thrombin/ml ofGFP
Experiment 1
Experiment 2
Released
component
FT* - 100%
thrombint ~ 100%
FT* = 100%
thrombint = 100%
PDGF
PF4
/3TG
5HT
ADP
ATP
/8-N
81
72
76
15
20
18
8
71
78
71
18
36
34
14
81
76
82
19
28
16
20
77
89
62
21
34
36
40
5U
5U
* Data are calculated as the percent of that released by a control sample frozen and thawed 4 times.
t Data are calculated as the percent of that released by a control sample aggregated with 5 U of
thrombin/ml ofGFP.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
nents —PDGF, PF4, and /3TG —were nearly maximally
released after 3 minutes by 0.025 U of thrombin/ml,
whereas the dense granule components (serotonin, ADP,
and ATP) were released to a much smaller extent under
the same conditions.
Discussion
The experiments reported here were designed to examine some of the effects of the growth-promoting activity of
substances released from platelets on human arterial
smooth muscle cells in culture and to describe the characteristics of the release of the growth factor from human
platelets in relationship to other released platelet substances. This work represents the first phase of a project
that aims to explore the characteristics of human PDGF
and its effects on human arterial SMC.
In the present experiments, most of the growth-promoting activity of human serum was found in the lipoproteindeficient (density > 1.21) fraction of serum (see Fig. 1).
Although the addition of lipoproteins (particularly VLDL
and LDL) to the lipoprotein-deficient fraction did appear
to provide some enhancement of cell growth, lipoproteins
alone did not promote growth. These results confirm the
observations of Ross and Glomset with monkey SMC 4
and extend their findings to human SMC.
A number of investigators have studied the effects of
lipoproteins and of hyperlipidemic serum on the proliferation of arterial SMC in various tissue culture
systems.12'20' 34' M Fischer-Dzoga et al.35 used primary
explants of monkey and rabbit aortic media grown in 10%
serum plus 5% of either homologous hyperlipidemic serum or isolated lipoproteins. In this system, LDL from
hyperlipidemic serum and lymph stimulated SMC proliferation, whereas HDL or LDL from normal serum had no
effect. Brown et al.2" studied swine aortic SMC grown in
culture with 1.5% swine serum supplemented with various
lipoproteins. Swine serum VLDL, LDL, and HDL all
increased SMC growth rate, and there was no difference
between normo- and hyperlipidemic lipoproteins with
respect to cell growth rate. Factors appearing during the
platelet release reaction were found to contribute to the
SMC growth response. Ross and co-workers34 also observed stimulation of SMC growth by lipoproteins, particularly LDL, under appropriate conditions, in addition to
the stimulation provided by platelet factors. Taken to-
gether, these and other reports suggest that lipoproteins
may play an important role in stimulating arterial SMC
proliferation, although differences in the systems employed and the results found make it difficult to be more
certain about this role at the present time.
Various other substances in serum may also be involved
in stimulating the proliferation of cultured SMC. Insulin
was observed recently to stimulate the proliferation of
monkey arterial SMC.38 In addition, it has been reported
that diabetic rabbit serum, but not normal rabbit serum,
contains a factor other than insulin or lipoproteins that
stimulates growth of rabbit arterial SMC.37
The contribution of platelet-released substances to the
growth-promoting activity of whole serum can be estimated from the dose-response curves of growth stimulation observed on adding increasing amounts of serum to
cells (curves determined during the course of this work
but not reported here in detail), compared to those
observed for PS (Fig. 2). From these curves, assuming
that the platelet count in the blood from which the serum
was derived was 300,000/mm3, and having found that 10'°
platelets yield approximately 1 mg of PS protein, we
estimate that approximately two-thirds of the growthpromoting activity of normal serum can be attributed to
a factor(s) released from platelets.
Material released from human platelets is highly potent
in stimulating human SMC to proliferate, as shown by the
experiments summarized in Figure 2. A dose-response
pattern similar to that observed with the human SMC was
also found for mouse 3T3 cells. Only limited information
is available, however, about the range of types of cells
that may respond with growth stimulation to plateletreleased material. Growth stimulation by platelet factors
has been observed with monkey SMC,4-S monkey and
mouse (3T3) fibroblasts,5-6 and with human glial
cells.7'"•15 Studies with other cell types would be of
considerable interest.
A possible similarity between the growth-promoting
activity of platelets and of fibroblast growth factor (isolated from bovine pituitary or brain) has been suggested
by Gospodarowicz et al.3" Several other growth factors
from serum have been under study by many investigators,
including somatomedins, nonsuppressible insulin-like activity, serum factors SI and S2, and multiplication stimulating activity (MSA) (as discussed broadly in Ref. 39).
408
CIRCULATION RESEARCH
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
The possible relationships between these and other growth
factors and the growth factor derived from platelets warrant further investigation. Rutherford and Ross5 refer to
the growth-promoting activity of platelets simply as the
platelet factor(s), PF, whereas Busch et al.15 refer to it as
platelet MSA. In view of the evidence (Refs. 8, 9; the
present work; and L. Witte, unpublished observations)
suggesting that there is a definite substance released from
platelets that is responsible for most of the growth-promoting activity seen with cultured cells, we propose the
term "platelet-derived growth factor (PDGF)" for this
material.
Other investigators have speculated about the possible
normal and abnormal physiological roles played by platelet-derived growth factor. It has been suggested that
PDGF may play a role in maintaining normal vascular
integrity and in vascular and extravascular repair processes.4' 5' "•15 Pathologically, PDGF may play a critical
role in the pathogenesis of atherosclerosis, by promoting
the intimal smooth muscle cell proliferation that appears
to be a key element in the development of the early
atherosclerotic lesion ,"•10 as discussed above.
Human blood platelets are known to contain two morphologically distinct types of granules, the dense bodies
and the a-granules. The dense bodies contain serotonin,
ADP, ATP, and calcium.40 Earlier studies identified the
a-granules as the storage site for lysosomal enzymes (acid
hydrolases and cathepsins),41'" but more recent evidence
suggests that the platelet lysosomal enzymes are stored in
vesicular structures distinct from a-granules.43 During the
platelet release reaction, granule-bound substances are
extruded from the cell, while the contents of membranes,
cytoplasm, and mitochondria are retained.40 Different
mechanisms may be involved in the release of substances
from different types of granules. For example, the dense
granule-bound substances can be released rapidly by weak
release inducers, such as ADP and epinephrine, whereas
acid hydrolases are released more slowly and only by
"strong" release inducers, such as collagen and
thrombin.33
There is no evidence available from cell fractionation
studies about the localization of PDGF or /3TG within the
platelet. There is, however, indirect evidence from studies
of the release of these materials that they are contained in
granules different from those containing acid hydrolases.
Thus, Busch et al.15 found that the growth-promoting
activity for glial cells in culture was released by ADP and
epinephrine, and Kaplan et al.31 observed that /3TG was
released by ADP; acid hydrolases, in contrast, are not
released by ADP or epinephrine.33'40 Most of the studies
of the subcellular localization of PF4 have been conducted
with assays based on its heparin-neutralizing activity.
Initially, heparin-neutralizing activity was thought to be
localized in the dense bodies, since it could be released by
ADP and epinephrine,4446 and its release was inhibited
by aspirin.46-47 However, studies of patients with storage
pool deficiency have shown normal amounts of heparinneutralizing activity in the platelets of most of these
patients,4*' 49 and cell fractionation studies have localized
heparin-neutralizing activity to the a-granule fractions and
not the dense bodies.50-51 With specific assays for PF4,
VOL. 42, No. 3, MARCH
1978
there is indirect evidence that PF4 is in different granules
from acid hydrolases, since PF4 is released by ADP.3' Cell
fractionation studies have not been performed, using
specific assays for PF4.
The results presented here demonstrate that PDGF is a
product of the platelet release reaction and provide detailed information about the characteristics of PDGF
release as induced by varying concentrations of thrombin.
It was found that the dose-response pattern of release of
PDGF was similar to that of both PF4 and /3TG and
different from that of ATP, ADP, and serotonin, or of the
acid hydrolase /3-N. Thus, in 3 minutes, the major release
of PDGF, PF4, and £TG occurred at thrombin concentrations of 0.015-0.025 U/ml, whereas the dense body
substances were mainly released during the same interval
at thrombin concentrations of 0.02-0.05 U/ml. These
findings suggest that PDGF, PF4, and /3TG are localized
in granules other than the dense bodies or the acid
hydrolase-containing granules, possibly in a-granules.
Recent studies in patients with storage pool disease
support this latter possibility. The platelets of patients
with this disorder have diminished amounts of ATP,
ADP, serotonin, and calcium and show decreased numbers of dense bodies.29-30 In the majority of patients, the
number of a-granules and the content of PF4, )3TG, and
PDGF has been found to be normal. One unique patient,
however, showed a decreased content of all three substances and a strikingly diminished number of agranules.52 Four platelet lysosomal enzymes were previously shown to be normal.30 Thus, present evidence
suggests that ATP, ADP, serotonin, and calcium are
stored in dense granules, while PDGF, PF4, and /3TG are
stored in a-granules. The storage site of lysosomal enzymes is still uncertain. As reported here, the contents of
the a-granules (PDGF, PF4, /3TG) are particularly sensitive to release by low doses of thrombin. At present, it is
not known whether these various a-granule substances
are all present in a single granule or in a-granules of
varying composition. However, the parallel release of
PDGF, PF4, and /3TG in vitro suggests that these substances might be released in parallel in vivo, and, thus, an
assay of PF4 and /3TG in patient samples might reflect
release of PDGF as well. Further studies, including those
using subcellular fractionation techniques, in both normal
and abnormal platelets will be required to define in more
detail the properties of this population of granules.
Acknowledgments
We arc grateful to B. Adams, H. Brand. M. Drillings, and G Lesznik
for expert technical assistance.
References
1. Temin HA: Growth regulating substance* for animal cells in culture.
Wistar lnst Symp Monogr 7: 103-106, 1967
2. Holley RW, Kicman JA: "Contact inhibition" of cell division in 3T3
cells Proc Natl Acad So USA 60: 300-304, 1968
3. Balk SD: Calcium as a regulator of the proliferation of normal, but
not of transformed, chicken Fibroblasu in a plasma-containing medium. Proc Natl Acad Sci USA 68: 271-275. 1971
4. Ross RJ, Glomset J, Kanya B, Harker L: A platelet-dependent
serum factor that stimulates the proliferation of arterial smooth
muscle cells in vitro. Proc Natl Acad Sci USA 71: 1207-1210. 1974
5 Rutherford RB, Ross R: Platelet factors stimulate fibroblasts and
smooth muscle cells quiescent in plasma serum to proliferate. J Cell
Biol69: 196-203, 1976
RELEASE OF GROWTH FACTOR FROM HUMAN PLATELETS/VWwe el al.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
6. Kohler N, Lipton A: Platelets as a source of fibroblast growthpromoting activity. ExpCell Res 87: 297-301, 1974
7. Wcstermark B, Wastcson A- The response of cultured human normal
glial cells to growth factors. In Advances in Metabolic Disorders, vol
8, edited by K Hall, R Luft. New York, Academic Press, 1975, pp
85-100
8. Westermark B, Wasteson A: A platelet factor stimulating human
normal glial cells. ExpCell Res 98: 170-174, 1976
9. Ross R, Glomset JA: The pathogenesis of atherosclerosis. New Eng J
Med 295: 369-377, 420-425, 1976
10. Ross R, Harker L: Hyperlipidcmia and atherosclerosis. Chronic
hyperlipidemia initiates and maintains lesions by endothelial cell
desquamation and lipid accumulation. Science 193: 1094-1100. 1976
11 Geer JC, Haust MD: Smooth muscle cells in atherosclerosis. Monographs on Atherosclerosis, vol 2, Basel S. Karger, 1972
12. Ross, R, Glomset JA: Atherosclerosis and the arterial smooth muscle
cell. Proliferation of smooth muscle is a key event in the genesis of the
Iciions of atherosclerosis. Science 180: 1332-1339, 1973
13. Moore S, Friedman RF, Singal DP, Gauldie J, Blajchman M. Inhibition of injury-induced thromboathersclerotic lesions by antiplatelet
serum in rabbits. Thromb Diath Haemorrh 35: 70-81, 1976
14 Harker LA, Ross R, Slichter SJ, Scott CR: Homocystine-induced
arteriosclerosis. The role of endothelial cell injury and platelet response in its genesis. J din Invest 58: 731-741, 1976
15. Busch C, Wasteson A, Westermark B: Release of a cell growth
promoting factor from human platelets. Thromb Res 8: 493-500,
1976
16. Ross R: The smooth muscle cell. II Growth of smooth muscle in
culture and formation of elastic fibers. J Cell Biol 50: 172-186, 1971
17. Albers JJ, Bierman EL: The effect of hypoxia on uptake and degradation of low density lipoproteins by cultured human arterial smooth
muscle cells. Biochim Biophys Acta 424: 422-429, 1976
18. Coltoff-Schiller B, Goldfischer S, Adamany AM, Wolinsky H: Endocytosis by vascular smooth muscle cells in vivo and in vitro. Am J
Pathol 83: 45-60, 1976
19 Shannon JE, Macy JL: Freezing, storage, and recovery of cell stocks
In Tiuuc Culture: Methods and Applications, edited by PF Kruse Jr,
MK Patterson Jr. New York, Academic Press, 1973, pp 712-718
20. Brown BG, Mahlcy R, Assmann G: Swine aortic smooth muscle cells
in tissue culture. Some effects of purified swine lipoproteins on cell
growth and morphology. Circ Res 39: 415-424, 1976
21. Havel RJ, Eder HA, Bragdon JH: The distribution and chemical
composition of ultracentnfugally separated lipoproteins in human
serum. J Clin Invest 34: 1345-1353, 1955
22. KSser-Glanzmann R, Jakabovd M, LQscher EF: Isolation and some
properties of the heparin-neutralizing factor (PF4) released from
human blood platelets. Experientia 28: 1221-1223, 1972
23. Holmsen H, Day HJ, Setkowsky CA: Behaviour of adenine nucleotides during the platelet release reaction induced by adenosine diphosphate and adrenaline. Biochem J 129: 67-82, 1972
24 Lages B, Scrutton MC, Holmsen H. Studies on gel-filtered human
platelets. Isolation and characterization in a medium containing no
added Ca1+, Mg>\ or K+. J Lab Clin Med 85: 811-825, 1975
25. Harris M: Electronic enumeration and sizing of cells. B Tissue culture
cells. In Tissue Culture: Methods and Applications, edited by PF
Kruse Jr, MK Patterson Jr. New York, Academic Press, 1973, pp
400-405
26. Gospodarowicz D: Purification of a fibroblast growth factor from
bovine pituitary. J Biol Chem 250: 2515-2520, 1975
27. Rubin H: Inhibition of DNA synthesis in animal cells by ethylene
diamine tetraacetate, and its reversal by zinc. Proc Natl Acad Sci USA
69:712-716,1972
28 Holmsen H, Storm E, Day HJ: Determination of ATP and ADP in
blood platelets. A modification of the firefly luciferase assay for
plasma. Anal Biochem 46: 489-501, 1972
29. Weiss HJ, Tschopp TB, Rogers J, Brand H: Studies of platelet 5hydroxytryptamine (serotonin) in storage pool disease and albinism J
Clin Invest 54: 421-432, 1974
30. Holmsen H, Setkowsky CA, Lages B, Day HJ, Weiss HJ, Scrutton
MC: Content and thrombin-induccd release of acid hydrolases in gel-
31.
32.
33.
34
35.
36.
37.
38.
39.
40.
41.
42
43.
44.
45.
46.
47
48.
49.
50.
51.
52.
409
filtered platelets from patients with storage pool disease. Blood 46:
131-142, 1975
Kaplan KL, Nossel HL, Drillings M, Lesznik G: Radioimmunoassay
of platelet factor 4 and /3-thromboglobulin Development and application in in vitro studies of platelet release in relation to fibrinopeptide A generation. Br J Haematol (in press)
Chen LB, Buchanan JM: Mitogenk activity of blood components. I.
Thrombin and prothrombin. Proc Natl Acad Sci USA 72: 131-135,
1975
Holmsen H, Day HJ: The selectivity of the thrombin-induced platelet
release reaction. Subcellular localization of released and retained
constituents. J Lab Clin Med 75: 840-855, 1970
Ross R. Glomset J: Studies of pnmate artenal smooth muscle cells in
relation to atherosclerosis. In Arterial Mesenchyme and Arteriosclerosis, edited by WD Wagner, TB Clarkson. New York. Plenum Press,
1974, pp 265-279
Fischer-Dzoga K. Fraser R, Wissler RW: Stimulation of proliferation
in stationary primary cultures of monkey and rabbit aortic smooth
muscle cells. I. Effects of lipoprotein fractions of hyperlipemic serum
and lymph. Exp Mol Pathol 24: 346-359, 1976
Stout RW, Bierman EL, Ross R: Effect of insulin on the proliferation
of cultured pnmate arterial smooth muscle cells. Circ Res 36: 319327,1975
Ledet T, Fischer-Dzoga K, Wissler RW: Growth of rabbit aortic
smooth-muscle cells cultured in media containing diabetic and hyperlipemic serum. Diabetes 25: 207-215, 1976
Gospodarowicz D, Greene G, Moran J: Fibroblast growth factor can
substitute for platelet factor to sustain the growth of Balb/3T3 cells in
the presence of plasma. Biochem Biophys ResCommun 65: 779-787,
1975
Somatomedins and Some Other Growth Factors Advances in Metabolic Disorders, vol 8, edited by R Luft, K Hall. New York, Academic
Press, 1975
Weiss HJ: Platelet physiology and abnormalities of platelet function.
N Eng J Med 293: 531-541, 1975
Marcus AJ, Zucker-Franklin D, Safier LB, Ullman HL: Studies on
human platelet granules and membranes. J Clin Invest 45: 14-28,
1966
Day HJ, Holmsen H, Hovig T: Subcellular particles of human
platelets. A biochemical and electronmicroscopic study with particular
reference to the influence of fractionation techniques. Scand J Haematol Suppl 7: 1-35, 1969
Bcntfeld ME, Bainton DF: Cytochemical localization of lysosomal
enzymes in rat megakaryocytes and platelets J Clin Invest 56: 16351649, 1975
Niewiarowski S, Thomas DP: Platelet factor 4 and adenosine diphosphate release during human platelet aggregation. Nature 222: 12691270,1969
Youssef A, Barkhan P: Release of platelet factor 4 by adenosine
diphosphate and other platelet-aggregating agents. Br Med J 1: 746747, 1968
Harada K, Zucker MB: Simultaneous development of platelet factor
4 activity and release of MC-serotonin Thromb Diath Haemorrh 25:
41-46,1971
Youssef AH, Barkhan P: Inhibition by aspirin of release of antiheparin activity from human platelets. Br Med J 3: 394-396. 1969
Weiss HJ, Rogers J: Platelet factor 4 in platelet disorders-Storage
location and the requirement of endogenous ADP for its release. Proc
Soc Exp Biol Med 142: 30-35, 1973
Walsh PN, Gagnatelli G: Platelet antiheparin activity Storage site
and release mechanism. Blood 44: 157-168, 1974
Broekman MJ, Handin RI, Cohen P: Distnbution of fibrinogen, and
platelet factors 4 and XIII in subcellular fractions of human platelets.
Br J Haematol 31: 51-55, 1975
De Prada M, Jakfibova M, LQscher EF, Pletscher A, Richards JG:
Subcellular localization of the heparin-neutralizing factor in blood
platelets. J Physiol 257: 495-502, 1976
Weiss HJ, Lages BA, Witte LD, Kaplan KL, Goodman DS, Notsel
HL, Baumgartner HR: Storage pool disease. Evidence for clinical and
biochemical heterogeneity (abstr). Thromb Haemostasis 38: 3. 1977
Studies of the release from human platelets of the growth factor for cultured human arterial
smooth muscle cells.
L D Witte, K L Kaplan, H L Nossel, B A Lages, H J Weiss and D S Goodman
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Circ Res. 1978;42:402-409
doi: 10.1161/01.RES.42.3.402
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1978 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/42/3/402.citation
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/