The influence of oxidatively modified low density lipoproteins on expression of platelet-derived growth factor by human monocyte-derived macrophages.

Abstract

Platelet-derived growth factor (PDGF) is secreted by several cells that participate in the process of atherogenesis, including arterial wall monocyte-derived macrophages. Macrophages in human and non-human primate lesions have recently been demonstrated to contain PDGF-B chain protein in situ. In developing lesions of atherosclerosis, macrophages take up and metabolize modified lipoproteins, leading to lipid accumulation and foam cell formation. Oxidatively modified low density lipoproteins (LDL) have been implicated in atherogenesis and have been demonstrated in atherosclerotic lesions. The effects of the uptake of various forms of modified LDL on PDGF gene expression, synthesis, and secretion in adherent cultures of human blood monocyte-derived macrophages were examined. LDL oxidized in a cell-free system in the presence of air and copper inhibited the constitutive expression of PDGF-B mRNA and secretion of PDGF in a dose-dependent fashion. Oxidatively modified LDL also attenuated lipopolysaccharide-induced PDGF-B mRNA expression. These changes were unrelated to the mechanism of lipid uptake and the degree of lipid loading and were detectable within 2 h of exposure to oxidized LDL. The degree of inhibition of both basal and lipopolysaccharide-induced PDGF-B-chain expression increased with the extent of LDL oxidation. Monocyte-derived macrophages exposed to acetylated LDL or LDL aggregates accumulated more cholesterol than cells treated with oxidized LDL, but PDGF expression was not consistently altered. Thus, uptake of a product or products of LDL oxidation modulates the expression and secretion of one of the principal macrophage-derived growth factors, PDGF. This modulation may influence chemotaxis and mitogenesis of smooth muscle cells locally in the artery wall during atherogenesis.

Full text

Page 1

‘rHE JOURNAL
I(:’ 1991 by The American Society for Biochemistry and Molecular Biology, Inc.
OF BIOLOGICAL CHEMISTRY
Vol. 266, No. 21, Issue of July 25, pp. 13901-13907, 1991
Printed in U.S.A.
The Influence of Oxidatively Modified Low Density Lipoproteins on
Expression of Platelet-derived Growth Factor by Human Monocyte-
derived Macrophages*
(Received for publication, August 31, 1990)
L. Trevor MaldenSg, Alan Chaitll, Elaine W. RainesS, and Russell RossSII
From the Departments of $Pathology and IMedicine, University of Washington, Seattle, Washington 98195
Platelet-derived growth factor (PDGF) is secreted
by several cells that participate in the process
erogenesis, including arterial wall monocyte-derived
macrophages. Macrophages in human and non-human
primate lesions have recently been demonstrated to
contain PDGF-B chain protein in situ. In developing
lesions of atherosclerosis, macrophages take up and
metabolize modified lipoproteins, leading to lipid ac-
cumulation and foam cell formation. Oxidatively mod-
ified low density lipoproteins (LDL) have been impli-
cated in atherogenesis and have been demonstrated in
atherosclerotic lesions. The effects of the uptake of
various forms of modified LDL on PDGF gene expres-
sion, synthesis, and secretion in adherent cultures of
human blood monocyte-derived macrophages were ex-
amined. LDL oxidized in a cell-free system in the pres-
ence of air and copper inhibited the constitutive
expression of PDGF-B mRNA and secretion of PDGF
in a dose-dependent fashion. Oxidatively modified
LDL also attenuated
lipopolysaccharide-induced
PDGF-B mRNA expression. These changes were un-
related to the mechanism of lipid uptake and the degree
of lipid loading and were detectable within
exposure to oxidized LDL. The degree of inhibition of
both basal and lipopolysaccharide-induced PDGF-B-
chain expression increased with
oxidation. Monocyte-derived macrophages exposed to
acetylated LDL or LDL aggregates accumulated more
cholesterol than cells treated with oxidized LDL, but
PDGF expression was not consistently altered. Thus,
uptake of a product or products of LDL oxidation mod-
ulates the expression and secretion of one of the prin-
cipal macrophage-derived growth factors,
modulation may influence chemotaxis and mitogenesis
of smooth muscle cells locally in the artery wall during
atherogenesis.
of ath-
2 h of
the extent of LDL
PDGF. This
Atherosclerosis is the principal cause
States, Europe, and Japan, and hypercholesterolemia is one
of the major and most extensively studied risk factors for this
of death in the United
* This work was supported in part by National Institutes of Health
Grants HL-18645 (to R. R., E. W. R., and A. C.), HL-03174 (to R. R.
and E. W. R.), and HL-30086 and AM-02456 (to A. C.). The costs of
publication of this article were defrayed in part by the payment of
page charges. This article must therefore he hereby marked “aduer-
tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate
this fact.
1 Recipient of travel assistance from the National Heart Founda-
tion of Australia.
11 To whom correspondence should he sent: Dept. of Pathology,
SM-30, University of Washington, Seattle, WA 98195. Tel.: 206-685-
7441.
disease (National Heart, Lung, and
One of the earliest detectable cellular changes in hypercholes-
terolemic animal models of atherosclerosis is the adherence
of monocytes to an intact endothelium
migration from the artery lumen into the intima (Gerrity,
1981a, 1981b; Faggiotto et al., 1984; Faggiotto and Ross, 1984;
Masuda and Ross, 1990a, 1990b). These monocytes become
intimal macrophages; many of them accumulate cholesteryl
esters and, together with
T lymphocytes (Jonasson et al.,
1986; Munro et al., 1987; Emeson and Robertson, 1988; Ma-
suda and Ross, 1990a; 1990b), form the fatty streak, the first
and ubiquitous lesion of atherosclerosis.
During atherosclerotic lesion formation and progression
macrophages have the potential to produce growth-regulatory
molecules, including PDGF.’ Thus, macrophage infiltration
into the artery wall may play a critical role in the regulation
of smooth muscle cell migration and proliferation (Ross,
necessary for the development of advanced occlusive lesions.
Macrophage-derived growth-regulatory molecules also may be
involved in vessel repair (Ross et al., 1986), may regulate
lipoprotein metabolism (Chait et al., 1980; Chait and Maz-
zone, 1982; Oppenheimer et al., 1987), and may influence the
immunoregulatory response (Dinarello and Mier,
Raines et al., 1990; Roberts and Sporn, 1990). Macrophages
have been shown previously to express PDGF mRNA and to
secrete PDGF i n vitro after activation (Shimokado
1985; Martinet et al., 1986; Mornex et al., 1986). Several recent
in viuo studies have demonstrated that PDGF mRNA expres-
sion (Barrett and Benditt, 1987; Wilcox et al., 1988; Ross et
aL., 1990) is increased in human and non-human primate
lesions of atherosclerosis. The level of PDGF-B mRNA in
carotid artery lesions was shown to be &fold higher than the
constitutive level of expression detected in normal arteries
(Barrett and Benditt, 1987). Analysis of atherosclerotic
plaques in human (Barrett and Benditt,
human primates (Ross
et al., 1990) revealed a
between the levels of the mRNAs for PDGF-B and the mac-
rophage-specific marker, c-fms, suggesting
might be produced by macrophages.
A recent immunocytochemical study of human and non-
human primate lesions of atherosclerosis, using a monoclonal
antibody to a peptide near the C terminus of the PDGF-B
chain, demonstrated the presence of this protein in approxi-
mately 20% of the macrophages within the lesions. Protein
containing PDGF-B chain was observed in a perinuclear
Blood Institute, 1985).
followed by their
1986)
1987;
et al.,
1988) and in non-
correlation
that PDGF-B
The abbreviations used are: PDGF, platelet-derived
LDL, low density lipoprotein(s); IFN, interferon; BHT, butylated
hydroxytoluene; DTPA, diethylenetriaminepentaacetic acid TBARS,
thiobarbituric acid-reactive substance(s); SDS, sodium dodecyl sul-
fate; MOPS, 4-morpholinepropanesulfonic acid LPS, lipopolysaccha-
ride.
growth factor;
13901

Page 2

13902
PDGF Expression in Human Macrophages Treated with Oxidized LDL
location in non-foam
laden macrophage-derived foam cells had low or undetectable
staining for PDGF at all stages of lesions development (Ross
et al., 1990).
Since modifications of low density lipoproteins (LDL), es-
pecially oxidative modification, have been implicated in the
pathogenesis of atherosclerosis (Carew et al., 1987; Kita et al.,
1987; Steinberg et al., 1989), we have examined the effect of
the uptake of various modified forms of LDL on the expres-
sion of PDGF in adherent cultures
derived macrophages. The results are consistent with the
immunocytochemical studies of human atherosclerosis using
the antibody to PDGF-B in that the
and protein produced by cultured monocyte-derived macro-
phages decreases after exposure to and uptake of oxidatively
modified LDL.
cell macrophages whereas the lipid-
of human monocyte-
level of PDGF-B mRNA
EXPERIMENTAL PROCEDURES
Macrophage Isolation and Culture
Mixed mononuclear cells
volunteers were separated by the density gradient centrifugation
method of Boyum (Boyum, 1968). Twenty ml of blood (anticoagulated
with 10 units/ml sodium heparin) was layered over 15 ml of Ficoll-
Paque and was centrifuged at 500 X g for 30 min at 23 "C. The mixed
mononuclear cell band was removed by aspiration, and the cells were
washed twice with endotoxin-free RPMI 1640 culture medium (Whit-
taker, Baltimore, MD). Monocyte and lymphocyte counts were ob-
tained using a Coulter counter with Channelyzer (Coulter Electronic
Instruments, Hialeah, FL). The cells were resuspended and plated at
3 X 10' monocytes/6O-mm dish (Primaria, Falcon, Lincoln Park, NJ)
in 3.0 ml of culture medium. Monocytes were allowed to adhere for 2
h, and then the adherent cell population was washed three times to
remove loosely adherent cells.
The adherent cell population was cultured in the presence of 20%
autologous serum, with media changes on days 3 and 7. Macrophages
were primed with recombinant interferon-? (IFN-7) (100 units/ml,
lot K9079A) (Genentech, South San Francisco)
continued presence of serum unless specified otherwise. Lipid loading
was performed on day 9 in RPMI 1640 culture medium (containing
IFN-7) in the absence of serum.
from fasting blood samples of healthy
on day 7 in the
Lipoprotein Preparation
Low density lipoproteins (d = 1.019-1.063 g/ml) were prepared by
density gradient ultracentrifugation from plasma
a pool of five or six normolipidemic
overnight for 12-14 h as described previously (Heinecke et al., 1986).
After extensive dialysis against 150 mM NaC1, 1 mM EDTA, the
protein content of LDL was determined by the method of Lowry et
al. (1951). This LDL represents the native LDL
LDL was acetylated by the method of Basu et al. (1976); phospholi-
pase C-modified LDL was prepared as described previously, and the
alteration was assessed by turbidity (Suits et al., 1989). Oxidatively
modified LDL was prepared in a cell-free system (Steinbrecher et al.,
1984; Jialal and Chait, 1989; Boyd et al., 1989). Briefly, to prepare
oxidized LDL, native LDL was diluted to 1 mg of protein/ml with
phosphate-buffered saline and was dialyzed against phosphate-buff-
ered saline in the absence of EDTA at 4 "C for 14 h. The LDL was
sterilized by filtration (0.22-pm Millipore filter), diluted to a concen-
tration of 300 pg of protein/ml with phosphate-buffered saline, and
incubated at 37 "C in Corning tissue culture dishes in the presence
5 @M Cu. S04.5H20 for 24 h unless specified otherwise.
the completion of oxidation the free radical scavenger
hydroxytoluene (BHT) (50
p~ final concentration) was added to
inhibit further oxidation (Morel et al., 1983; Heinecke et al., 1986).
In some experiments the chelating agent diethylenetriaminepentaa-
cetic acid (DTPA) (25 p~ final concentration) also was added. The
oxidatively modified LDL was stored under nitrogen in the dark at
4 "C and used within 24 h. Control LDL was prepared in a manner
similar to oxidized LDL except that the 37 "C incubation step was
omitted. Control LDL contained copper, BHT, and DTPA in amounts
equivalent to those present in the oxidized LDL.
The extent of oxidative modification of LDL was evaluated in all
preparations by measurement of thiobarbituric acid-reactive sub-
(4 mM EDTA) from
volunteers who had fasted
used in these studies.
of
At
butylated
stances (TBARS), which gives a measure of lipid peroxides, by the
method of Buege and Aust (1978). TBARS values are expressed as
the malondialdehyde equivalents (nmol of malondialdehyde/mg of
LDL protein) derived from a standard curve obtained using freshly
prepared 1,1,3,3-tetramethoxyropane. Changes in the electropho-
retic mobility of LDL after oxidative modification were evaluated
using agarose gel electrophoresis (Henriksen et al., 1983).
The levels of endotoxin were measured with the limulus amebocyte
lysate assay (Whittaker). The final concentration of endotoxin in the
medium containing lipid usually was less than 0.1 ng/ml. However,
medium containing phospholipase C-modified LDL occasionally con-
tained higher levels.
Assays for Cytotoxicity
Lactate Dehydrogenase-The measurement of lactate dehydrogen-
ase in the conditioned medium reflects cell damage and is used as an
index of cytotoxicity. Lactate dehydrogenase levels in the 24-h con-
ditioned medium were measured routinely using a spectrophotometric
assay (Macomb Biotechnology, Inc., Romeo, MI).
Protein Synthesis-A second approach to evaluation of the toxic
effects of the modified lipoproteins upon the macrophages was to
measure changes in the rate of incorporation of [3H]leucine into
protein. After a 24-h exposure to control medium or modified lipo-
proteins the macrophages were washed with leucine-free RPMI and
then incubated with leucine-free RPMI containing ['Hlleucine (10
FCi/ml) (Du Pont-New England Nuclear)
subsequently washed with phosphate-buffered saline prior to precip-
itating the cell-associated proteins with 5% trichloroacetic acid. The
precipitated proteins were solubilized in 0.1 M sodium hydroxide for
the determination of protein content (Lowry et al., 1951) and meas-
urement of the incorporation of ['Hlleucine into protein.
for 2 h. The cells were
Preparation of Total Cellular RNA and Northern Blot Analysis
Total cellular RNAs
guanidinium isothiocyanate, homogenized (Bethesda Research Lab-
oratories), and purified using guanidinium-acid-phenol (Chomczynski
and Sacchi, 1987). Ten-pg aliquots
density) were heated to 60 "C for 5 min in running buffer (20 mM
MOPS, pH 7.0; 5 mM sodium acetate; 1 mM EDTA) containing 50%
formamide and 6% formaldehyde. Samples were electrophoresed at
85 volts for 3 h in 1 or 1.2% agarose gels containing 6% formaldehyde
and 1 X running buffer. Agarose gels were stained with ethidium
bromide (to verify equivalent RNA load in all lanes) prior to transfer
of RNA to nylon membranes (Nytran; Schleicher & Schuell) with 20
X standard saline citrate (1 X SSC = 150 mM NaC1, 15 mM sodium
citrate).
The human PDGF-B mRNA was identified using a 2,200-base pair
cRNA probe synthesized from the pGS-2
1 (Barrett and Benditt, 1988). A 1,330-base pair PstI fragment from
the 6-actin cDNA was purified from the bovine p-actin clone (Degen
et al., 1983). Glyceraldehyde-3-phosphate dehydrogenase mRNA was
detected using a 1,300-base pair fragment isolated from the chicken
clone pGAD-28 (Dugaiczyk et al., 1983).
cDNA Probe Hybridization-Isolated
dom prime labeled (Feinberg and Vogelstein, 1983)
deoxycytidine 5'-triphosphate (3,000 Ci/mmol; Du Pont-New Eng-
land Nuclear). Hybridizations were performed at 42 "C in a solution
containing 50% formamide (v/v), 0.25 M sodium phosphate (pH 7.21,
0.25 M NaC1, 1 mM EDTA, 7% sodium dodecyl sulfate (SDS), 10%
(w/v) polyethylene glycol (Mr 6,000), and 100 pg/ml salmon sperm
were isolated by lysis of cells in situ in
of RNA (as determined by optical
plasmid, derived from pSB-
cDNA fragments were ran-
with [cu-"P]
DNA. The filters were washed subsequently in 0.2 X SSC, 0.1% SDS
at 60 "C for 1 hand exposed to XAR-2 film (Kodak) with intensifying
screens (Cronex; Du Pont-New England Nuclear)
Composites shown are from a single film and were exposed for the
same time periods. Filters were stripped by washing at 100 "C in 0.01
x SSC, 0.1% SDS.
Single-stranded RNA Probe Hybridization-The
was linearized with HindIII, and single-stranded RNA probes were
synthesized using a commercially available
Madison, WI) in the presence of SP6 polymerase, RNasin, and [n-
32P]uridine 5"triphosphate (800 Ci/mmol; Du Pont-New England
Nuclear) (Melton et al., 1984). cRNA hybridizations were performed
at 60 "C in the hybridization buffer as above containing Escherichia
coli tRNA. Filters were washed at 60 "C in 0.1 X SSC, 0.1% SDS.
The filters hybridized with the pGS-2 cRNA were exposed to XAR-2
film to obtain preliminary exposures; then the filters were cut hori-
for 1-10 days.
plasmid pGS-2
kit (Promega Biotec,

Page 3

PDGF Expression in Human Macrophages Treated with Oxidized LDL
13903
zontally below the location of the 3.8-kilobase specific PDGF-B
signal, and the upper segment was subjected to RNase A treatment
(1 pg/ml in 2 X SSC at room temperature for 10 min) followed by a
second wash at 60 "C in wash buffer as above. The lower segment of
the filter could then be rehybridized with other specific probes.
Radioreceptor Assay for PDGF
Medium was removed from dishes and centrifuged at 800 rpm for
5 min to remove contaminating cells. The medium was concentrated
10-fold using Centriprep-10 concentrators (Amicon, W. R. Grace and
Co., Danvers, MA). The level of PDGF protein in the conditioned
medium was determined by PDGF radioreceptor assay using
labeled PDGF-AB and human dermal fibroblasts (Bowen-Pope and
Ross, 1985).
'"I-
RESULTS
PDGF Expression in Human Monocyte-derived Macrophages
To assess the basal level of expression of PDGF, human
monocyte-derived macrophages, maintained previously as ad-
herent cultures for 9 days in the presence of 20% autologous
serum, were evaluated by Northern blot analysis and dem-
onstrated expression of the 3.8-kilobase PDGF-B mRNA
transcript (Fig. 1). The level of the PDGF-B mRNA was
increased in macrophages treated with IFN-7 for 48 h (days
7-9) (Fig. 2). The day 9 macrophages produced a low or
undetectable level of PDGF-A chain mRNA (data not shown).
PDGF protein was also detected by PDGF radioreceptor
PDGFB *
Time(hours)
24
6
2 24
6
2
24
6
2 24
6
2
0
No
Add
Acetyl
LDL LDL
PLC-
LDL
Ox-
FIG. 1. PDGF-B chain mRNA expression in IFN-y-primed
human blood monocyte-derived macrophages incubated with
modified LDL. Human monocyte-derived macrophages were cul-
tured for 9 days and then exposed to control medium (No Add) or
modified lipoproteins in serum-free
phages were primed with IFN--y (100 units/ml) on day 7, and IFN--y
treatment was continued during lipid loading. RNA was obtained
from the cells 2, 6, and 24 h after lipid loading and was analyzed by
Northern blot using the PDGF-B cRNA
included acetylated LDL (Acetyl LDL; 100 pg of protein/ml), oxidized
LDL (Ox-LDL; 50 pg of protein/ml) prepared in a cell-free system,
and phospholipase C-modified LDL (PLC-LDL; 100 pg of protein/
ml) as described under "Experimental Procedures."
medium as indicated. Macro-
probe. Modified lipoproteins
assay in the 24-h conditioned media collected from day 9
macrophages. Conditioned medium from 60-mm dishes (orig-
inally plated at loG monocytes/ml medium) contained 1.029
* 0.7 ng of PDGF/ml (n = 10; range 0.35-2.75 ng of PDGF/
ml) .
The Influence of Different Forms of Modified LDL on PDGF
Expression
To evaluate the effect of lipid accumulation on the produc-
tion of PDGF by macrophages, lipid uptake was induced by
three different mechanisms. Day 9 adherent macrophage cul-
tures were exposed to oxidatively modified LDL (50 pg of
protein/ml), acetylated LDL (50-100 pg of
phospholipase C-modified LDL (50-100 pg of protein/ml) in
serum-free medium. In contrast to the stable level of expres-
sion of the PDGF-B mRNA in control macrophages (incu-
bated for identical time periods in serum-free medium devoid
of lipoprotein), incubation of macrophages with oxidatively
modified LDL usually resulted in a decrease in the level of
expression of PDGF-B mRNA. This decrease was evident as
early as 2 h after exposure of the macrophages to oxidatively
modified LDL (Fig. 1). Phospholipase C-modified LDL also
reduced the level of expression of the PDGF-B mRNA in
macrophages in some experiments whereas there usually was
either no significant change or a small increase in the PDGF-
B transcript level after exposure of the macrophages to acet-
ylated LDL (see Figs. 1 and 6). In 3 of 20 experiments there
was no detectable change in the level of the PDGF-B mRNA
expressed by the macrophages within the first 24 h of exposure
to oxidatively modified LDL (data not shown). The three
preparations of oxidized LDL did not differ in their extent of
modification as determined by TBARS.
Consistent with the observed decrease in the level of expres-
sion of PDGF-B mRNA, PDGF activity in the 24-h condi-
tioned medium was reduced by 90% after exposure of the
macrophages to oxidatively modified LDL. The other modi-
fied lipoproteins did not alter the
conditioned medium significantly as compared with levels
detected in control macrophage cultures (Fig. 3).
To assess whether the observed changes in the steady-state
levels of macrophage PDGF-B mRNA reflected a general
protein/ml), or
levels of PDGF in the
Time(hours)
IFN y
BHT/DTPA
g
ox-
&
FIG. 2. The influence of antioxidants and IFN-y on the
suppression of PDGF-B mRNA induced by oxidized LDL.
Northern blot analysis of RNA from human monocyte-derived mac-
rophages after exposure to oxidized LDL for 24 h with or without
IFN--y priming (48-h pretreatment). Macrophages were exposed to
control medium (No Add) or oxidized LDL ( Ox-LDL) with or without
BHT/DTPA for 24 h in serum-free medium. Hybridizations were
performed with the PDGF-B and @-actin probes.
Add
LDL LDL LDL LDL
No
Nallvc
Acclyi
PLC
Ox
FIG. 3. The influence of different modified forms of LDL on
cholesteryl ester mass, PDGF release, and [3H]leucine incor-
poration into protein. Cellular cholesteryl ester mass (@ pg/mg
cell protein) was measured in duplicate in day 10 macrophages after
exposure to control media (No Add) or 50 pg of protein/ml native
LDL, acetylated LDL (Acetyl LDL), phospholipase C-modified LDL
(PLC-LDL), and oxidized LDL (Ox-LDL) for 24 h. The conditioned
media also were collected at the end of the 24-h exposure, and PDGF
(0) was measured by radioreceptor assay. The values represent trip-
licate determinations using two different volumes of concentrated
conditioned medium in the assay. ['HILeucine incorporation into
trichloroacetic acid-precipitable,
measured in triplicate over a 2-h period
exposure of the macrophages to the lipoproteins.
cell-associated protein (B) was
after the 24-h period of

Page 4

13904
PDGF Expression in Human Macrophages Treated with Oxidized LDL
inhibition of mRNA the level of expression of two other
mRNA species, glyceraldehyde-3-phosphate dehydrogenase
and @-actin, were examined. The level of the @-actin mRNA
decreased in macrophages 24 h after the addition of oxida-
tively modified LDL (Fig. 2) and was independent of the
changes in PDGF-B mRNA (data not shown). The expression
of glyceraldehyde-3-phosphate dehydrogenase mRNA did not
decrease in the macrophages after exposure to oxidatively
modified LDL even when @-actin and PDGF-B mRNA levels
decreased (data not shown; see also Figs. 6 and 7). However,
oxidatively modified LDL increased the expression of glycer-
aldehyde-3-phosphate dehydrogenase mRNA in some exper-
iments when the cells were subsequently stimulated with LPS
(Fig. 6 and data not shown). Because of the variable effects
on the mRNA levels of genes often used to normalize for
differences in mRNA load, the optimal density of RNA sam-
ples was used to quantitate the RNA. Only equivalently loaded
lanes, as determined by ethidium bromide staining, were
evaluated. Consequently, small variations in the apparent
mRNA levels may not be significant.
Factors Influencing the Expression of PDGF-B mRNA
during Exposure to Oxidatively Modified LDL
IFN--, and Antioxidants-In an attempt to define factors
that may influence the reduction of the level of PDGF-B
mRNA in macrophages during exposure to oxidatively modi-
fied LDL, the roles of IFN--, priming, the free radical scav-
enger BHT, and the chelating agent DTPA were evaluated.
In these experiments, LDL was oxidized for 24 h and imme-
diately added to the cultures in either the presence or the
absence of BHT and/or DTPA. Parallel cultures were exam-
ined with or without prior exposure of the macrophages to
IFN--, (100 units/ml for 48 h). The basal level of PDGF-B
mRNA was higher in the IFN--,-primed macrophages (Fig.
2). However, with IFN--, priming there also was a significant
reduction in the level of expression of PDGF-B mRNA in
macrophages after exposure to oxidatively modified LDL (Fig.
2). Thus with IFN--, priming it did not appear that the
antioxidant BHT or the chelating agent DTPA could account
for the observed reduction in the levels of PDGF-B mRNA
induced by oxidized LDL in the macrophages.
Extent Of LDL Oxidation-During the process of copper-
induced oxidation of LDL performed at 37 "C in a cell-free
system, the TBARS routinely increased within the first 2 h,
plateaued at a maximum level after 3-4 h, and then decreased
by approximately 30% over the next 20 h (Fig. 4, A and B).
However, the electrophoretic mobility increased throughout
the period of oxidative modification (Fig. 4A).
To determine the influence of the degree of LDL oxidation
on the suppression of PDGF-B mRNA, macrophages were
exposed to LDL preparations that were oxidized to different
extents. The degree of LDL oxidation was controlled by
changing the time of incubation of LDL with copper. PDGF-
B mRNA expression remained stable when macrophages were
exposed to LDL that had been oxidized for periods up to 2 h
(Fig. 5). These preparations were mildly oxidized (Fig. 4A)
and may be similar to the previously described minimally
modified LDL based on TBARS levels and electrophoretic
mobility (Berliner et al., 1990). However, LDL that had been
oxidized for 2.25 h or longer decreased the level of PDGF-B
mRNA expression in macrophages (Figs. 1 and 5). Further
increases in the extent of LDL oxidation decreased the level
of expression of PDGF-B mRNA level in macrophages. These
findings suggest that the suppression of PDGF-B expression
is related to the extent of oxidation of LDL.
Concentration of Oxidized LDL-The effect of different
A
Origin c
TBARS
(nmol MDA equivalents
' per mg LDL protein)
Origin c
TBARS
(nmol MDA equivalents
per mg LDL protein)
FIG. 4. The electrophoretic mobilities of the lipoproteins
from different preparations of modified LDL were compared
using agarose gel electrophoresis. A, the change in electropho-
retic mobility after the oxidative modification of LDL for varying
periods of time in a cell-free system containing copper was measured
and compared with TBARS measurements.
and electrophoretic patterns are shown
LDL), control LDL, oxidized LDL (Ox-LDL; 4 h and 24 h), phospho-
lipase C-modified LDL (PLC-LDL), and native LDL.
R, representative TBARS
for the acetylated LDL
(Acetyl
P D G F B , O O W O W " -
I - I -
g ?
a -
.-
c u
-
$ 2
m
FIG. 5. Influence of the time and extent of LDL oxidation
on PDGF-B mRNA expression. LDL was oxidized for 1.25, 1.75,
2, 2.25, 2.5, and 4 h and then added to macrophages for 24 h. The
levels of PDGF-B chain expression were analyzed by Northern blot
and TBARS measured on the different LDL preparations. Evaluation
of ethidium bromide-stained gels demonstrated equivalent amounts
of total RNA in all lanes (data not shown).
-
4 Durationof LDL
2y42y2
2
Oxidation (hours)
TEARS
43
30
30
23
1y4
1y4
7
0
(nmol MDA equivalents/
mg LDL Protein)
0
35
cu
z
m o
concentrations of oxidized LDL also was analyzed. The extent
of inhibition of PDGF-B mRNA and protein production by
macrophages increased with higher concentrations of oxidized
LDL. The concentration of oxidized LDL required to induce
the suppression of PDGF-B mRNA and PDGF protein varied
among experiments; however, concentrations of oxidized LDL
greater than 25 pg of protein/ml were routinely associated
with suppression of PDGF-B mRNA levels (data not shown).
Degree of Lipid Loading-Although extensive oxidation of
LDL was associated with the observed suppression of PDGF-
B mRNA in macrophages there was no evidence for oxidation
of acetylated LDL or phospholipase C-modified LDL (as
measured by TBARS or changes in electrophoretic mobility)
(Fig. 4B). Nevertheless, phospholipase C-modified LDL also
suppressed PDGF-B mRNA expression in macrophages in
several experiments. Furthermore, there was no apparent
relationship between the levels of unesterified cholesterol

Page 5

PDGF Expression in Human Macrophages Treated with Oxidized LDL
13905
(data not shown) and cholesteryl ester accumulation in the
macrophages after exposure to the different forms of modified
LDL and to the differing effects of these LDLs on PDGF-B
mRNA levels in macrophages (Fig. 3).
Cytotoxicity-Oxidatively modified LDL produced varying
degrees of cytotoxicity in macrophage cultures as assessed by
lactate dehydrogenase release, inhibition of incorporation of
["Hlleucine into protein, and decreased protein recovery from
culture plates. In some experiments no evidence of cytotox-
icity was detected. However, in most experiments after ex-
posure of macrophages to oxidatively modified LDL for 24 h,
a reduction in the incorporation of ['HH]leucine into protein
was observed to varying extents with the oxidatively modified
LDL (Fig. 3).
Decreased ['Hlleucine incorporation into protein usually
correlated with an elevation of lactate dehydrogenase in the
conditioned medium after exposure of macrophages to oxi-
datively modified LDL (data not shown). Furthermore, in-
creases in the concentration of oxidatively modified LDL
produced progressive increases in the levels of lactate dehy-
drogenase detected in the conditioned medium corresponding
to the increased cytotoxicity of higher concentrations of oxi-
datively modified LDL (data not shown).
The extent of the cytotoxicity induced by oxidatively mod-
ified LDL varied independently of the observed changes in
PDGF-B chain mRNA. Cytotoxicity sometimes was detecta-
ble in the absence of an observable change in PDGF-B mRNA
(data not shown). Conversely, in some experiments, PDGF-
B mRNA and protein levels were markedly suppressed by
oxidized LDL, despite minimal cytotoxicity (Fig. 3). Cytotox-
icity was not observed in macrophages exposed to acetylated
LDL (Fig. 3) and was minimal in the case of macrophages
treated with phospholipase C-modified LDL (data not
The level of expression of glyceraldehyde-3-phosphate dehy-
drogenase mRNA in macrophages remained stable even in
the presence of apparent cytotoxicity (Fig. 6).
shown).
Rp - ; . -
PDGFB m
GAPDH +'
L P S I - - - - - - I + + + + + + )
FIG. 6. Modulation of LPS induction of PDGF-B mRNA in
macrophages by oxidized LDL. Day 9 macrophages were exposed
to control media (No Add), acetylated LDL (Acetyl LDL), native
LDL, control LDL (native LDL containing copper and BHT in
amounts equivalent to that present in oxidized LDL), and oxidized
LDL (Ox-LDL; 4 h and 24 h) at concentration of 50 pg of protein/ml
for 24 h, after which RNA was isolated. In a replicate set of cultures,
media were replaced by fresh media containing 1 pg/ml LPS, and at
the end of 2 h RNA was isolated and the media collected for deter-
mination of lactate dehydrogenase levels. ["HJLeucine incorporation
into trichloroacetic acid-precipitable, cell-associated protein was
measured in replicate cultures over a 2-h period after 24-h exposure
to the lipoproteins. Macrophage cultures exposed to oxidized LDL
demonstrated evidence of cytotoxicity, as assessed by increased lac-
tate dehydrogenase levels that were 8.25- and 9-fold above no-add
levels for 4 and 24 h oxidized LDL, respectively; decreased ["HI
leucine incorporation into protein at 55 and 46% of no-add values for
4 h and 24 h oxidized LDL, respectively.
Modulation of the Response of Macrophages to LPS by
Modified LDL
In the absence of added lipoproteins LPS led to a transient
induction of PDGF-B mRNA which was maximal 2 h after
exposure of the cells to LPS (data not
phages were exposed to oxidized LDL (50 pg of protein/ml)
for 24 h prior to a 2-h incubation with LPS (1 pglml), the
level of PDGF-B mRNA induced by LPS was reduced mark-
edly as compared with matched control macrophages stimu-
lated with LPS (Fig. 6). No inhibition of induction of PDGF-
B mRNA by LPS was observed when macrophages were
preincubated with native LDL or native LDL containing
copper and BHT in amounts equivalent to that present in
oxidized LDL (control LDL). Similarly, pretreatment of the
macrophages with acetylated LDL (50 pg of protein/ml) (Fig.
6) or phospholipase C-modified LDL (50 pg of protein/ml)
(data not shown) did not influence the subsequent induction
of PDGF-B mRNA in response to LPS. Modulation of LPS
induction of PDGF-B mRNA was observed with concentra-
tions of oxidized LDL (24 h) of less than 2 pg of protein/ml
and in the absence of demonstrable cytotoxicity. In contrast,
LDL that was oxidized for 1 h had no effect on the induction
of PDGF-B mRNA in response to LPS (Fig. 7).
shown). When macro-
DISCUSSION
PDGF Expression in Macrophages and the Influence of
Modified LDL-The factors regulating
PDGF-B mRNA in macrophages have not been well defined.
However, activation of cultured monocytes with endotoxin,
phorbol ester, concanavalin A, or cycloheximide increases
PDGF-B mRNA expression and PDGF secretion (Shimokado
et al., 1985; Martinet et al., 1986; Sariban and Kufe, 1988).
The expression of PDGF-B mRNA and protein by cultured
macrophages and the induction of PDGF after activation with
LPS observed in our study is consistent with these previous
reports. The increase in PDGF-B mRNA after IFN-7 priming
of macrophages is consistent with gene induction associated
with macrophage activation.
Decreased PDGF-B mRNA expression was observed after
exposure of monocyte-derived macrophages to oxidatively
modified LDL and occasionally was observed with phospho-
the expression of
- - - " - ~ ~ "
". , .. "_
PDGFB *
GAPDH -
z O b z O o o o +
8
0 0 0
- A A J E
7JJ7J6
FIG. 7. Concentration-dependent effect of oxidized LDL on
decreased expression of the PDGF-B mRNA after LPS stim-
ulation. Macrophages were exposed to control medium, oxidized
LDL (Ox-LDL; 1 h, 50 pg of protein/ml) or oxidized LDL (24 h; 2,
10, and 50 pg of protein/ml) for 24 h, then the medium was replaced
with RPMI containing 1 pdml LPS for 2 h. RNA was isolated, and
Northern blots were analyzed using the PDGF-B cRNA probe.
GAPDH, glyceraldehyde-3-phosphate dehydrogenase.