Volume 270,
Number 11,
Issue of March 17, 1995 pp. 6389-6395
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
Inhibition of
Platelet-derived Growth Factor-BB-induced Fibroblast Proliferation by
Plasmin-activated 
-Macroglobulin Is Mediated via an
-Macroglobulin Receptor/Low Density Lipoprotein
Receptor-related Protein-dependent Mechanism (*)
(Received for publication, July 22, 1994; and in revised form, November 28,
1994)
James C.
Bonner (§),
,
Annette
Badgett
,
Maureane
Hoffman
(1),
Pamela
M.
Lindroos
From the Laboratory of Pulmonary Pathobiology, National Institute of
Environmental Health Sciences, Research Triangle Park, North Carolina
27709 and Laboratory Services, Durham Veterans Affairs
Medical Center, Department of Pathology, Duke University Medical
Center, Durham, North Carolina 27705
ABSTRACT
- INTRODUCTION
- MATERIALS AND METHODS
- RESULTS
- DISCUSSION
- FOOTNOTES
- REFERENCES
ABSTRACT 
-Macroglobulin (M) is a
potentially important regulator of platelet-derived growth factor-BB
(PDGF-BB)-stimulated cell growth due to our previous observation that
PDGF-BB binds to M noncovalently (Bonner, J. C.,
Goodell, A. L., Lasky, J. A., and Hoffman, M. R.(1992) J. Biol.
Chem. 267, 12837-12844). We examined the in vitro effect of native and plasmin-activated (receptor-recognized)
M on the PDGF-BB-induced proliferation of mouse Swiss
3T3 and rat lung fibroblasts. Nondenaturing polyacrylamide gel
electrophoresis showed that plasmin converted M to its
electrophoretically ``fast'' form at a 2:1 molar ratio and
that 
I-PDGF-BB bound both M and
M-plasmin. PDGF-BB-induced growth was not affected by
native M (0.3 µM) or plasmin (0.6
µM). The combination of plasmin and M
(2:1 molar ratio) inhibited PDGF-BB-induced cell proliferation
80-90%. Complexes of PDGF-BB
M purified by
gel filtration chromatography retained growth promoting activity, but
the PDGF-BBM-plasmin complex did not.
Preincubation of fibroblasts (37 °C for 24 h) with
M-plasmin did not change I-PDGF-BB
binding or affect gene expression of the 6.5-kilobase PDGF-
receptor or 5.2-kilobase PDGF-
receptor mRNA. However,
preincubation with M-plasmin (0-4 °C for 4
h) increased I-PDGF-BB binding 2-fold, and this increase
was blocked by a receptor-associated protein antagonist of the
M-receptor/low density lipoprotein receptor-related
protein. The receptor-associated protein antagonist blocked I-M-methylamine binding, inhibited
PDGF-BB-M-plasmin uptake from fibroblast-cultured
supernatants, and abolished the inhibitory effect of
M-plasmin on PDGF-stimulated growth. These data
suggest that inhibition of PDGF-stimulated proliferation by
M-plasmin is mediated in part by clearance of
PDGF-BB-M-plasmin through the lipoprotein
receptor-related protein.
INTRODUCTION
Platelet-derived growth factor (PDGF) (
)and
homologues of PDGF secreted by macrophages(1, 2) ,
endothelial cells(3) , smooth muscle cells(4) , and
fibroblasts (5) are mesenchymal cell mitogens and
chemoattractants that operate in the normal processes of development,
wound healing, and tissue remodeling(6) . The aberrant
expression of PDGF has been postulated as a common feature in the
progression of fibroproliferative diseases such as atherosclerosis (7) and pulmonary fibrosis(8, 9) . PDGF is a
disulfide-linked dimer of two polypeptide chains termed A or B that
give rise to functional PDGF-AA, PDGF-AB, or PDGF-BB
isoforms(10) . Two subtypes of PDGF cell-surface receptors,
termed PDGF-R and PDGF-R are present on cells of mesenchymal
origin(5, 11) , and two related but distinct cDNAs
encoding and PDGF receptors have been
cloned(12, 13) . The PDGF-A chain binds only the
PDGF-R subtype, whereas the PDGF-B chain binds both PDGF-R
and PDGF-R(11, 14) . Apparently, PDGF receptor
dimerization, mediated by ligand binding, is required for signal
transduction(15) . The precise role for each of the different
PDGF isoforms and receptor subtypes is not well understood, but it has
been suggested that the different subtypes of PDGF and receptors could
allow for fine tuning of cellular responses, due to observations that
different cell types can vary greatly in the ratio of PDGF isoforms
secreted and in the receptor subtype composition that the responding
target cell possesses(11) .
The biological activity of PDGF
is also regulated by -macroglobulin
(M)(4, 16, 17, 18, 19) .
The function of M as a proteinase inhibitor has been
well described and the mechanism whereby native or electrophoretically
``slow'' M covalently entraps serine,
aspartic, cysteine, and metalloproteinases has been extensively
studied(20, 21, 22, 23, 24) .
Proteinases ``activate'' M through cleavage
at a specific ``bait region'' which inititates a series of
conformational changes in the molecule that entraps the
proteinase(20) . The conformational change reveals latent
receptor recognition sites on the molecule and also makes the
M more compact; conferring ``fast'' mobility
when subjected to nondenaturing gel electrophoresis as compared to the
native or slow form of M. The irreversible triggering
of the proteinase trap is mimicked by primary amines(25) , and
fast M-proteinase or M-amine complex,
but not native M, bind high affinity receptors on
fibroblasts(26, 27) , hepatocytes(28) , and
macrophages(29, 30) . This receptor, termed the
M receptor/low density lipoprotein receptor-related
protein (LRP), binds -migrating very low density lipoproteins
activated with apolipoprotein E as well as fast
Ms(31, 32) . It is synthesized as a
600-kDa precursor protein which undergoes post-translational processing
into a 515-kDa ligand-binding subunit and an 85-kDa transmembrane
subunit(33) . A 39-kDa protein that has been copurified with
LRP, termed receptor-associated protein (RAP), can reversibly bind to
the 515-kDa subunit and inhibit binding and uptake of ligands which
interact with the LRP(34) .
PDGFM
complexes have been isolated from plasma (17) and from
macrophage supernatants(18) . PDGF binds both native and
proteinase- or amine-activated forms of
M(35) , and PDGF-stimulated fibroblast
proliferation (19) and chemotaxis (36) are inhibited by
native M at concentrations above 0.3 µM.
Below this concentration, the native form has no significant effect on
the biological properties of PDGF. M activated with
methylamine synergistically enhances the growth promoting activity of
PDGF purified from human platelets (19) . M
also binds and modulates the biological activities of several other
growth factors, including transforming growth factor-
(TGF-)(37) , tumor necrosis factor- (TNF-) (38) , basic fibroblast growth factor(39) ,
interleukin-1 (IL-1) (40) , interleukin-6
(IL-6)(41) , nerve growth factor (NGF)(42) , and
vascular endothelial growth factor(43) . The action of native
and activated forms of M in regulating this wide
spectrum of growth factors has been reviewed(44) . However, the
precise role of M in regulating growth factor activity
is poorly understood.
The potential modulatory activity of
proteinase-activated M on PDGF-stimulated mitogenesis
has not yet been investigated. Because methylamine-modified and
proteinase-modified fast Ms are both receptor
recognized by fibroblasts, we hypothesized that both would possess
similar biological activities with regard to modulation of
PDGF-stimulated cell growth. On the contrary, we report that
plasmin-activated fast M inhibited the proliferation
of Swiss mouse 3T3 fibroblasts and rat lung fibroblasts induced by
PDGF-BB, and this inhibitory effect was blocked by the RAP antagonist
of LRP. We hypothesize that, in extravascular tissues, native
(nonreceptor recognized) M serves as a latent
reservoir for PDGF-BB, whereas proteinase-activated (receptor
recognized) M serves to clear PDGF-BB through the LRP.
MATERIALS AND METHODS
Reagents and Cells
Bovine M
(Boehringer Mannheim) was dialyzed against 100 volumes of distilled
water to precipitate fast M as has been described
previously (35) . Slow M was converted to fast
M by incubation with 25 mM methylamine
(Tris-HCl, 50 mM, pH 8.0) overnight at 25 °C, or by
incubation with a 4:1 molar excess of plasmin overnight at 37 °C.
Excess methylamine was removed from Mmethylamine
complexes by dialysis against 100 volumes of 50 mM Tris-HCl,
pH 8.2, at 4 °C. Excess plasmin was removed from
Mplasmin complexes by Superose 6 FPLC gel
filtration chromatography (see below). M preparations
were tested for PDGF contamination as described
previously(19) . A 39-kDa glutathione S-transferase
receptor-associated protein (RAP) antagonist of the LRP was the kind
gift of Dr. Dudley Strickland (American Red Cross, Rockville,
MD)(33) . Swiss mouse 3T3 fibroblasts were purchased from
American Type Culture Collection (Rockville, MD). Early passage rat
lung fibroblasts were isolated and characterized as described
elsewhere(19) .
Gel Electrophoresis
Electrophoresis of the
PDGF-BB-M mixtures was performed by nondenaturing gel
electrophoresis, due to our previous observation that PDGF-BB is
dissociated from M under denaturing
conditions(35) . Ten µg of M was mixed
with I-PDGF-BB (2 ng) and incubated overnight at 37
°C in a final volume of 25 µl. Samples were mixed with 10
µl of 6 
nondenaturing sample buffer containing Tris-borate
EDTA (TBE) glycerol, 1% xylene cyanol, and 1% bromphenol blue and
electrophoresed on a native 6% gel (Novex, Encinitas, CA). Native gels
were fixed in 40% methanol, 10% acetic acid, stained with Coomassie
Blue, destained with 25% methanol, and dehydrated using a gel dryer.
Dried gels were exposed to autoradiographic film (Amersham Corp.) to
visualize I-PDGF-BB bound to M.
Superose 6 FPLC Gel Filtration
Chromatography
PDGF-BBM and
PDGF-BBM-plasmin complexes were routinely
prepared by incubating 100 ng of human recombinant I-PDGF-BB with 1 mg of M or
M-plasmin for 24 h at 37 °C. These mixtures were
isolated by loading onto a gel filtration, molecular weight exclusion
column (Superose 6 FPLC, Pharmacia LKB Biotechnol) equilibrated in
phosphate-buffered saline, pH 7.5, operating at a flow rate of 0.5
ml/min.
Northern Analysis for PDGF- and - Receptor
mRNA
Fibroblasts were grown to confluence in 150-cm flasks in 10% FBS-DMEM and then rendered quiescent for 24 h in
serum-free defined medium (SFDM: Ham's F-12 with HEPES,
CaCl, and 0.25% BSA supplemented with an
insulin-transferrin-selenium mixture purchased from Boehringer Mannheim
Biochemicals). Fibroblasts were then treated with M or
M-plasmin for 24 h at 37 °C. Collection of total
RNA and Northern analysis was performed as described
previously(59) . The human cDNA probes for the PDGF-R
subunit and PDGF-R subunit, kindly provided by Dr. Carl-Henrik
Heldin (Ludwig Institute for Cancer Research, Uppsala, Sweden), were
labeled with a random primed DNA labeling kit (Boehringer Mannheim).
I-PDGF-AA and -BB Receptor Assay
Rat
lung fibroblasts or Swiss 3T3 fibroblasts were rendered quiescent for
24 h with SFDM. Plates of cells were either exposed to
M, M-plasmin, or media alone for an
additional 24 h prior to assay. The plates were then cooled to
0-4 °C for 15-30 min, the media aspirated, and the
cells washed twice with ice-cold binding medium (Ham's F-12 with
HEPES, CaCl, and 0.25% BSA). I-PDGF-BB
(DuPont NEN) or I-PDGF-AA (Biomedical Technologies,
Stoughton, MA) was diluted to 100 ng/ml in binding buffer. Binding
assays were performed by adding increasing amounts (0.5-20 ng/ml)
of I-PDGF-BB or -AA to cells in duplicate wells in a
final volume of 0.2 ml/well. After incubating on an oscillating
platform at 0-4 °C for 3 h, the cells were washed three times
with ice-cold binding buffer, the cell-bound radioactivity solubilized
with 1% Triton-X containing 0.1% BSA and 0.1 N NaOH, and the
radioactivity measured on a 
-counter.
-Macroglobulin Receptor
Assay
Fibroblasts were plated as described above for the PDGF
binding assay. Once confluent, the cells were rendered quiescent for 24
h in SFDM. Cultures were then cooled on ice for 15-30 min before
aspirating the 10% FBS and washing three times with a
serum-free-binding medium. Fibroblasts were incubated in binding medium
(1 ml/well) at 4 °C for 3-4 h, then washed three more times
before adding human I-M-methylamine
(custom iodinated by DuPont NEN, 8 mCi/mg) at a final concentration of
0.1 nM in the absence or presence of unlabeled bovine
M, M-plasmin, or the RAP antagonist
of the M receptor. Following a 3-4-h incubation
at 4 °C, the medium was aspirated and the cells were washed three
times with cold binding medium. The cell-bound radioactivity was
extracted with solubilization buffer (1% Triton X-100, 0.1 N NaOH) and counted on a -counter.
IPDGF-BBM-Plasmin
Uptake
Native M or plasmin-activated
M (20 µg) was incubated for 1 h at 37 °C with IPDGF-BB (2 ng) in a final volume of 50 µl.
These complexes were then purified by Superose 6 FPLC as described
above and the column fractions eluting between 20-28 min were
pooled and diluted in SFDM to a final concentration of 1 nM (M) in 20 ml. The complexes were then added to
confluent lung fibroblasts (
10 10
/150-cm flask) that had been rendered quiescent in SFDM for 24 h. Some
cultures of fibroblasts were pretreated with 100 nM RAP for 1
h prior to adding IPDGF-BBM-plasmin complex to
block uptake. The medium was collected at 0, 4, and 12 h
post-treatment, centrifuged 1400 g to remove
nonadherent cells, and the supernatant filtered (0.2 µm acrodisc,
Gelman Sciences). The filtrate was concentrated to 1 ml using 30-kDa
cutoff membrane (Amicon, Beverly, MA) and dialyzed on the Amicon
apparatus against 100 ml of deionized water three times. After
concentrating again to 1 ml, the sample was lyophilized using a Speed
Vac (Savant, Inc.) and reconstituted in 20 µl of water and 6 µl
of Tris-borate EDTA native sample buffer. The samples were
electrophoresed on 5% native gels, dried, and autoradiographed as
described above under ``Electrophoresis.'' In another set of
experiments, I-PDGF in the absence or presence of
M slow and fast forms was bound at 0-4 °C
for 3 h to confluent cultures of fibroblasts in 150-cm dishes prior to washing the cultures and warming to 37 °C to
measure release of radioactivity into the medium. The supernatants were
then run on a 4-20% SDS-polyacrylamide gel (Novex) along with
standard I-PDGF-BB, dried, and autoradiographed to assess
the molecular weight of the released radioactivity. This experiment
allowed us to determine whether M fast forms
facilitated recycling of intact PDGF following uptake via the LRP.
Cell Proliferation Assay
Fibroblasts suspended in
10% FBS-DMEM were seeded 2 ml/well in 12-well tissue culture plates at
20,000/well and allowed to proliferate for 24 h in 5% CO humidified air at 37 °C. These subconfluent cultures were
washed two times with SFDM, and fresh SFDM alone or containing
increasing concentrations of human recombinant PDGF-BB (UpState
Biotechnology, Lake Placid, NY) was added back to the wells. Parallel
incubations were coincubated with M, plasmin, or a
mixture of M and plasmin. In another experiment,
fractions from the Superose 6 FPLC column containing purified
PDGF-BBM or
PDGF-BBM-plasmin complex were added to the
cultures. After 3 days of incubation, the growth medium was aspirated
and the cells harvested by trypsinization and enumerated using a
Coulter Counter (Coulter Electronics, Hialeah, FL).
RESULTS
Fibroblast Proliferation Induced by PDGF-BB Is Inhibited by
A Mixture of M and Plasmin or by Purified
M-Plasmin Complex
We first sought to determine
the molar ratio of plasmin:M that would convert all of
the slow form M to electrophoretically fast,
receptor-recognized M. Increasing concentration of
plasmin progressively converted M slow form to its
electrophoretically fast conformation with complete conversion
occurring at a 2:1 molar ratio of plasmin to M (Fig. 1A). I-PDGF-BB bound to both native
slow M and fast plasmin-activated M (Fig. 1B). The PDGF-BB-stimulated proliferation of rat
lung fibroblasts and Swiss 3T3 fibroblasts was inhibited in the
presence of a mixture of 0.3 µM M and 0.6
µM plasmin (Fig. 2). M or plasmin
alone at these same concentrations did not affect PDGF-BB-induced
growth. Further experiments were performed on rat lung fibroblasts with
PDGF-BBM and
PDGF-BBM-plasmin complexes purified by Superose
6 FPLC chromatography. M and
M-plasmin bound approximately equal amounts of I-PDGF-BB (Fig. 3). However, no mitogenic activity
was observed at the molecular weight of the
PDGFM-plasmin complex, while the
PDGF-BBM complex stimulated the proliferation of
rat lung fibroblasts by as much as 40% above the number of cells
obtained from SFDM alone (Fig. 3).
Figure 1:
[I]PDGF-BB
binds to native and plasmin-activated M. Panel
A, Native M (10 µg) was treated with plasmin
for 24 h at 37 °C prior to loading onto a nondenaturing 6%
Tris-borate-EDTA gel as described under ``Materials and
Methods.'' Increasing plasmin concentrations resulted in a
progressive increase in receptor-recognized, electrophoretically fast (F) M with complete conversion occurring at a
2:1 molar ratio of plasmin to M. Panel B,
autoradiography of TBE gel demonstrating that M and
M-plasmin bound nearly equivalent amounts of I-PDGF-BB. Native and plasmin-activated
M (10 µg) were incubated with I-PDGF-BB (2 ng) at 37 °C for 24 h and
electrophoresis was performed as described
above.
Figure 2:
PDGF-BB-induced fibroblast proliferation
is inhibited by a combination of M and plasmin. The
proliferation of rat lung fibroblasts (panel A) and Swiss 3T3
fibroblasts (panel B) was measured as described under
``Materials and Methods.'' Subconfluent fibroblast cultures
were treated with increasing concentration of PDGF-BB in the absence or
presence of M (0.3 µM), plasmin (0.6
µM), or a combination of M and plasmin
(1:2 molar ratio) in serum-free defined medium. After 3 days in culture
the cells were removed from the plates by trypsin treatment and
enumerated with an electronic particle counter. Native
M (closed circles) or plasmin (open
triangles) had no significant affect on PDGF-stimulated cell
growth as compared to PDGF-BB alone (open circles, dashed
line), while the combination of M and plasmin (closed triangles) inhibited PDGF-BB-stimulated proliferation.
Results are the mean of six separate experiments (S.E. < 5% of the
mean). Each assay was performed in triplicate
wells.
Figure 3:
The
PDGF-BBM complex retains mitogenic activity,
while the PDGF-BBM-plasmin complex does not. One
mg of M or M-plasmin was incubated
with 200 ng of nonradioactive PDGF-BB for 24 h at 37 °C prior to
loading onto a Superose 6 FPLC column and elution in PBS, pH 7.4.
Column fractions were diluted 1:1 with serum-free defined medium and
tested for mitogenic activity on lung fibroblasts in a 3-day
proliferation assay as described under ``Materials and
Methods.'' The equivalent binding of I-PDGF-BB (2 ng) to either M or
M-plasmin (100 µg) was determined on Superose 6
FPLC in an identical manner as described above, and the column
fractions were -counted. A, PDGF-BB bound
M-plasmin (open triangles, dashed
line) but the PDGF-BB/M-plasmin did not elicit
mitogenesis of rat lung fibroblasts (solid line, closed
circles). A peak of mitogenic activity was apparent at the
approximate molecular mass of uncomplexed PDGF-BB. B, the
PDGF-BBM complex stimulated fibroblast
proliferation by as much as 40% above control cells maintained in SFDM
alone. Arrows indicate the elution of molecular weight mass
markers: 1, blue dextran (V), 2,
thyroglobulin (669 kDa, also marks the elution of M as
determined by ELISA), 3, apoferritin (440 kDa), 4,
-amylase (200 kDa), 5, BSA (66 kDa), 6, carbonic
anhydrase (29 kDa), and 7, cytochrome (12.4 kDa). The data are
typical of four separate experiments.
Preincubation with M and
M-Plasmin at 37 °C Does Not Alter PDGF Receptor
Gene Expression or Binding of I-PDGF-BB to Cell-surface
PDGF Receptors
We first postulated that
M-plasmin, which is receptor recognized (see Fig. 4), inhibited PDGF-BB-stimulated proliferation through
down-regulation of either the PDGF-R or the PDGF-R on
fibroblasts. Preincubation with M-plasmin for 24 h at
37 °C did not affect expression of the 6.5-kb PDGF-R gene on
Swiss 3T3 cells, nor did it significantly change the gene expression of
the 5.2-kb PDGF-R mRNA on rat lung fibroblasts (data not shown).
Furthermore, preincubation of fibroblasts with M or
M-plasmin for 24 h at 37 °C did not alter I-PDGF-BB or I-PDGF-AA binding on lung
fibroblasts or Swiss 3T3 fibroblasts. However, preincubation of rat
lung fibroblasts with methylamine-modified M increased I-PDGF-AA-binding sites 3-fold, indicating that this
fast form of M up-regulated the PDGF- receptor
subtype (Table 1).
Figure 4:
M-plasmin and a 39-kDa
RAP antagonist compete for the specific binding of I-M-methylamine to fibroblasts.
Confluent quiescent cultures of rat lung fibroblasts (panel A)
or Swiss 3T3 fibroblasts (panel B) were chilled to 0-4
°C for 30 min and assayed for specific binding of I-M-methylamine as described under
``Materials and Methods.'' Increasing concentration
of M-plasmin (open circles) and the 39-kDa
RAP antagonist (closed triangles) were added immediately prior
to the addition of 0.1 nMI-M-methylamine. A 200-fold molar
excess of native M did not inhibit the specific
binding of I-M-methylamine (closed
squares). The RAP antagonist inhibited specific binding (IC
= 0.1-0.2 nM) with about a 5-fold lesser
potency than that of M-plasmin (IC = 0.5-1.0 nM). The data are representative
of four experiments each performed in
triplicate.
Preincubation with M-Plasmin, But Not
M, at 0-4 °C Increases Cell-surface Binding
of I-PDGF-BB That Is Blocked by the RAP Antagonist of
LRP
As an alternate hypothesis, we postulated that
M-plasmin inhibited PDGF-BB-stimulated growth via the
LRP on fibroblasts. We first demonstrated that both fibroblast types
possessed LRP and that RAP antagonized M-plasmin
binding (Fig. 4). The specific binding of I-M-methylamine to rat lung fibroblasts
and Swiss 3T3 cells was inhibited in a concentration-dependent manner
by M-plasmin (IC between 0.2-1
nM). Similarly, the 39-kDa RAP antagonist inhibited I-M-methylamine binding (IC between 0.5-1 nM) for both types of fibroblasts.
Maximal inhibition of binding by the antagonist was reached at
approximately 10 nM RAP, a 100-fold molar excess over the
concentration of I-M-methylamine used in
the assay (Fig. 4). The RAP antagonist did not inhibit the
specific binding of I-PDGF-BB to rat lung fibroblasts (Fig. 5). We then preincubated fibroblasts with
M-plasmin at 0-4 °C for 4 h, which prevents
internalization of LRP-bound complex(34) , prior to performing
the PDGF receptor assay in order to determine if this
receptor-recognized form of M would increase the
binding of I-PDGF-BB at the cell surface.
M-plasmin, but not native M,
increased specific I-PDGF-BB binding 2-fold (Fig. 6). Importantly, this increased binding was observed only
at concentrations of I-PDGF-BB above 0.3 nM (>10 ng/ml), i.e. at concentrations of radioligand
above saturation of PDGF cell-surface receptors. These data indicated
that I-PDGF-BB bound to LRP-bound
M-plasmin, but only when PDGF cell-surface receptors
(predominantly PDGF-R) were saturated.
Figure 5:
The 39- kDa M RAP
antagonist does not compete for I-PDGF-BB binding to
fibroblasts. Increasing concentrations of PDGF-BB or the RAP
M-receptor antagonist were added with 1 ng/ml I-PDGF-BB to confluent rat lung fibroblasts rendered
quiescent in SFDM for 24 h. Receptor binding was assayed as described
under ``Materials and Methods.'' These data show
that RAP is a specific antagonist for the LRP and does not directly
interfere with PDGF-BB binding to its cell-surface receptor. The data
are representative of four experiments each performed in
triplicate.
Figure 6:
Preincubation of fibroblasts with
M-plasmin at 0-4 °C increases the specific
binding of I-PDGF-BB. Confluent, quiescent rat lung
fibroblasts were incubated with M or
M-plasmin, with or without the RAP antagonist, for 4 h
on ice prior to washing the cells three times with ice-cold binding
buffer and performing the radioligand binding assay for I-PDGF-BB (20 ng/ml) as described under ``Materials
and Methods.'' M-plasmin, but not native
M, caused a 2-fold increase in the specific binding of I-PDGF-BB that was significant (
p < 0.01 paired Student's t test), but only at
radioligand concentrations >10 ng/ml. The increase in I-PDGF-BB binding was blocked by co-incubation with a
100-fold excess of the RAP antagonist. These data indicate that I-PDGF-BB binds to surface-bound
M-plasmin, but only when the cell-surface PDGF
receptors are saturated. Data are the mean ± S.E. of four
separate experiments each performed in
triplicate.
Uptake of the IPDGF-BBM-Plasmin Complex
and Inhibition of PDGF-BB-stimulated Growth by
M-Plasmin Is Blocked by the RAP Antagonist of the
LRP
Because LRP-bound M-plasmin increased the
specific binding of I-PDGF-BB to fibroblasts, we proposed
that PDGF-BBM-plasmin complex could be taken up
through the LRP and that the mitoinhibitory effect of
M-plasmin on PDGF-BB-stimulated fibroblast growth
could be blocked by the RAP antagonist. We were unable to show uptake
of the IPDGF-BBM-plasmin
complex in the 3-day cell proliferation assay due to the low number of
cells in the assay (20,000/well at time 0) and the relatively low
specific activity of I-PDGF-BB bound to
M. While uptake of I-M-MA could have been assessed in the
cell proliferation assay, we sought to copurify I-PDGF-BB
bound to M by autoradiography of a nondenaturing gel
and visualize M by protein staining the same gel.
Therefore, the IPDGF-BBM or
the IPDGF-BBM-plasmin
complex was diluted to 1 nM in 20 ml of SFDM, and these
complexes were added to confluent lung fibroblasts (10
10/150-cm flask) in the absence or presence of
100 nM RAP. The IPDGF-BBM-plasmin complex
was taken up by rat lung fibroblasts in culture after 12 h, and this
uptake was blocked by RAP (Fig. 7). Furthermore, in the 3-day
cell proliferation assay, the RAP antagonist blocked
M-plasmin-induced inhibition of PDGF-BB-stimulated rat
lung fibroblast proliferation, while RAP alone did not affect
PDGF-BB-induced proliferation (Fig. 8). In another experiment,
we investigated the fate of I-PDGF-BB bound to
M-plasmin or M-methylamine following
uptake, since we previously reported that
M-methylamine synergistically enhanced PDGF-stimulated
growth of fibroblasts and this contrasted with our present observation
of M-plasmin inhibition of PDGF-stimulated
growth(19) . I-PDGF-BB was incubated with rat
lung fibroblasts at 0-4 °C for 3 h in the absence or presence
of 0.3 µM M-plasmin or
M-methylamine, then the cells were rinsed with binding
buffer and allowed to warm to 37 °C to allow uptake of growth
factor. At various time points (0.5, 1, 2, and 3 h) the supernatants
were removed and run on 4-20% SDS-polyacrylamide gel
electrophoresis to determine the amount of radioactivity released by
the cells and to determine if either fast form of M
was mediating recycling of intact 30-kDa PDGF-BB. Radioactivity
possessing a molecular mass below that of PDGF-BB (30 kDa) was released
in a time-dependent manner from fibroblasts that had internalized
surface-bound I-PDGF-BB (30 min, 2,333 ± 168 cpm;
1 h, 6,641 ± 170 cpm; 2 h, 20,916 ± 1,293 cpm; and 3 h,
22, 725 ± 781 cpm). The presence of M-plasmin
or M-methylamine did not significantly alter the
amount of radioactivity released into the medium nor was any intact I-PDGF-BB released (i.e. all released
radioactivity possessed a molecular mass below 30 kDa). These data
indicated that neither M-plasmin nor
M-methylamine mediate recycling of biologically active
PDGF.
Figure 7:
The uptake of IPDGF-BBM-plasmin complex by
fibroblasts in culture is inhibited by RAP. Confluent cultures of rat
lung fibroblasts grown in 10% FBS-DMEM in 150-cm flasks
were rendered quiescent with SFDM for 4-6 h at 37 °C and then
treated with BSA-free SFDM alone or supplemented with 100 nM RAP for an additional 1 h prior to adding 1 nMIPDGF-BBM or IPDGF-BBM-plasmin complex.
These complexes were prepared by incubating 20 µg of
M or M-plasmin with 0.3 ng of I-PDGF-BB in a final volume of 100 µl for 4 h at 37
°C prior to diluting in 10 ml BSA-free SFDM and adding to 10 ml of
pre-existing medium (with or without RAP) on the cultures. The final
concentration of M was 1 nM. At various time
points the media was removed, concentrated with a 100-kDa cutoff
filter, lyophilized, reconstituted to 50 µl, and electrophoresed on
a nondenaturing gel as described under ``Materials and
Methods.'' Following electrophoresis, the gels were dried
and stained with Coomassie Blue to visualize M slow (S) and fast (F) forms (panel A). The same
gel was exposed to autoradiographic film to visualize I-PDGF-BB bound to S and F forms of M (panel B). The IPDGF-BBM-plasmin complex,
but not the native complex, was taken up by the fibroblasts, and this
uptake was blocked the RAP antagonist.
Figure 8:
The RAP antagonist blocks inhibition of
PDGF-BB-induced growth by M-plasmin. Subconfluent
cultures of rat lung fibroblasts in 10% FBS-DMEM were washed three
times with SFDM and treated with PDGF-BB (10 ng/ml) in the absence or
presence of M (0.3 µM), plasmin (0.6
µM), or a combination of both (2:1 molar ratio of plasmin
to M) as described in Fig. 2. A 100-fold molar
excess of RAP antagonist significantly blocked the growth inhibitory
effect of the M-plasmin mixture on PDGF-BB-stimulated
growth (**p < 0.01 paired Student's t test).
Data are expressed as the percent increase in cell number above that
obtained in SFDM without PDGF-BB and are the mean ± S.E. from
six separate experiments each performed in triplicate. All treatments
shown contain 10 ng/ml PDGF-BB.
DISCUSSION
M has been described as a potential
modulator of several growth factors and cytokines. Its precise function
in regulating growth factor activity appears to vary greatly depending
on the specific growth factor, the conformation of M
to which the growth factor is bound, and the biological function
studied. In general, M could serve several regulatory
roles, including 1), inactivation and clearance of growth
factors, 2) potentiation of growth factor activity,
3) protection of growth factors against proteolytic
inactivation, and 4) as a latent extracellular reservoir of
growth factors. In this study we report that M
activated by plasmin is a negative growth regulator of PDGF-stimulated
fibroblast proliferation in vitro, and this appears to be
related to the clearance of PDGF-BBM-plasmin
complex through an LRP-dependent mechanism. In contrast, the
PDGF-BBM complex, which does not bind LRP or
PDGF receptors(18, 19) , retained growth-promoting
activity. This is likely explained by our previous observation that
PDGF is bound to M in a reversible, noncovalent
manner, and PDGF-BB dissociates from M with a
half-time of about 2 h(35) . PDGF-BB bound to
M does not bind PDGF-R or PDGF-R, apparently
because the receptor-recognized sites on the PDGF molecule are masked
by the M molecule (18) . Thus, PDGF-BB was
probably released from the PDGF-BBM complex
during the course of the 3-day growth assay to bind cell-surface PDGF
receptors (Fig. 3). The relative low affinity binding of PDGF-BB
to M in the micromolar range (47) as compared
to PDGF-BB binding either the PDGF-R or PDGF-R in the
nanomolar range (Fig. 5) supports our hypothesis that native
M serves as a reservoir of latent PDGF-BB that can be
released to bind high affinity PDGF cell-surface receptors on
mesenchymal cells.
Our data suggest that the
Mplasmin complex does not suppress PDGF action
by down-regulation of the PDGF receptor, since preincubation of
fibroblasts with M-plasmin did not alter PDGF-R
or PDGF-R gene expression. Furthermore, I-PDGF-BB
binding to fibroblasts was not affected by M-plasmin
preincubation for 24 h at 37 °C. However, preincubation of
fibroblasts with M-plasmin at 0-4 °C for 4
h, which prevents internalization of
M-plasminLRP complex, increased the specific
binding of I-PDGF-BB 2-fold. This increase in specific
binding was observed only at concentrations of I-PDGF-BB
above receptor saturation (0.3 nMI-PDGF-BB) (Fig. 6). Recently, Crookston and co-workers (47) reported a K
of 0.3 µM for PDGF-BB binding to M , indicating that
PDGF-BB binds to M with an affinity 1000-fold less
than PDGF-BB binding to its own receptor. In general, this is in
agreement with our observation that I-PDGF-BB will bind
to LRP-bound M-plasmin only when cell-surface PDGF
receptors are occupied (i.e. at concentrations of I-PDGF-BB above 10 ng/ml). We observed that the increase
in I-PDGF-BB binding following M-plasmin
preincubation on ice was blocked by the RAP antagonist of the LRP (Fig. 6). This proved that PDGF-BB bound to LRP-bound
M-plasmin at the cell surface, since excess
concentrations of RAP did not inhibit the binding of I-PDGF-BB binding to its own receptor on fibroblasts (Fig. 5). The RAP antagonist inhibited I-M-methylamine binding to fibroblasts (Fig. 4), inhibited uptake of IPDGF-BBM complex by
fibroblasts (Fig. 7), and abolished the inhibitory effect of
M-plasmin on PDGF-induced proliferation (Fig. 8). Thus, we concluded that the inhibitory effect of
M-plasmin on PDGF-BB-stimulated fibroblast growth is
mediated at least in part by uptake and degradation of
PDGF-BBM-plasmin complex through the LRP.
Our
findings are novel in that we have demonstrated that fibroblasts can
mediate growth factor clearance via their cell-surface LRP, and this
results in a mitoinhibitory effect. The concentration of
M used in this study (300 nM) is much lower
than that found in plasma (2-4 µM) (23, 24) and could be similar to concentrations found
in extravascular spaces. Activated macrophages produce
M(18) , and we have observed that human lung
fibroblasts in culture spontaneously secrete 0.5 ng
M/million cells. ()Thus, in vivo fibroblasts could have nanomolar levels of M in
their immediate environment. Several other investigators have reported
that activated Ms mediate the plasma clearance and
inactivation of growth factors and cytokines, including
PDGF-BB(48) , TGF-1(48, 49) , and
TNF-(38) . These studies are particularly relevant to
circumstances of vascular injury, where platelets degranulate and
release growth factors such as PDGF and TGF-. Crookston and
co-workers showed that, while PDGF-BB and TGF-1 bound both native
and methylamine-activated M, only the IPDGF-BBM-methylamine and ITGF-1M-methylamine
complexes were cleared from the circulation(48) . These
investigators suggested that native, non-receptor-recognized
M is the major carrier of TGF-1 and PDGF-BB in
the blood, whereas activated Ms mediate the rapid
clearance of these growth factors through the liver. This is a
reasonable assessment, since it has been established that TGF-1
and PDGF-BB bind to slow and fast Ms and because
M-proteinase complexes are cleared through the liver
with a half-time of minutes(28) . Our findings herein suggest
that M and M-proteinase complexes
could be critical mediators of growth factor clearance in extravascular
compartments, particularly during inflammation and tissue repair where
activated macrophages have accumulated and release numerous growth
factors (e.g. PDGF (1) , TGF-(50) ,
IL-1(51) ), M(18) ,and a wide
spectrum of proteolytic enzymes(52) . We reported earlier that
the majority of macrophage-derived PDGF in culture supernatants is
complexed to macrophage-derived homologues of plasma M
and that the macrophage-derived PDGFM complex
does not bind to PDGF receptors, nor is it recognized by antibodies
raised against PDGF(18) . However, these complexes retain
PDGF-like activity for fibroblasts. Like fibroblasts, macrophages also
possess LRP (29) and potentially could reclaim secreted
PDGFM complex in the presence of proteinases
through an LRP autocrine loop. This hypothetical situation would be
further complicated by inflammatory mediators (e.g. colony
stimulating factor, interferon-, and lipopolysaccharide) which
have been shown to either up-regulate or down-regulate the macrophage
LRP(53, 54) .
We previously reported that
M-methylamine enhanced the growth promoting activity
of PDGF for fibroblasts(19) . A similar effect of
